linux kernel tree branch, qualcomm-2.6.32.9, created, c78e46e6ee009c7d
g...@chromium.org
- Log -----------------------------------------------------------------
commit c78e46e6ee009c7d7c156dec021b77f2340467dc
Merge: 21cf46d ac3165f
Author: Willie Ruan <wr...@quicinc.com>
Date: Wed Jun 9 10:25:41 2010 -0700
Merge "msm: acpuclock-8x50a: Update 'sync with AXI' ACPU rate" into msm-2.6.32.9a
commit 21cf46d77e6ae0bb3a7a505b5ec5e54e31248596
Merge: 6eb4a9d 3b9bed8
Author: Saravana Kannan <ska...@quicinc.com>
Date: Tue Jun 8 20:32:19 2010 -0700
Merge "Merge remote branch 'cros/master'" into msm-2.6.32.9a
commit 6eb4a9d80e67be43be147dc3e6c27190216b47b1
Merge: a116b8c 41dc3d3
Author: Willie Ruan <wr...@quicinc.com>
Date: Tue Jun 8 17:59:07 2010 -0700
Merge "msm_fb: Improve HDMI insert/remove on ST1.5" into msm-2.6.32.9a
commit ac3165f6f9f703356133d68ee6af3613146df5d9
Author: Matt Wagantall <ma...@codeaurora.org>
Date: Tue Jun 8 17:33:45 2010 -0700
msm: acpuclock-8x50a: Update 'sync with AXI' ACPU rate
Update the cpu frequency table so that loops_per_jiffy and delays
are safe when booting with the ACPU clock sourced from AXI.
Signed-off-by: Matt Wagantall <ma...@codeaurora.org>
commit a116b8c93247fda177871a0391f0ae0511dcb447
Merge: 2a5b460 868a0da
Author: Willie Ruan <wr...@quicinc.com>
Date: Tue Jun 8 16:40:46 2010 -0700
Merge "msm: Add platform data for resetting HDMI on ST1.5" into msm-2.6.32.9a
commit 2a5b46038b25a9962e96b17ab628d4a7e7b413f5
Merge: 39281c0 46dc50d
Author: Willie Ruan <wr...@quicinc.com>
Date: Tue Jun 8 16:40:34 2010 -0700
Merge "msm: audio: Remove obsolete audio drivers." into msm-2.6.32.9a
commit 46dc50d9518eec13f5092867cf131dacbfd7223d
Author: Ashraya Arya <ashr...@codeaurora.org>
Date: Tue Jun 1 18:06:25 2010 -0700
msm: audio: Remove obsolete audio drivers.
Obsolete core audio drivers for the old firmware and corresponding
ALSA code removed.
Signed-off-by: Ashraya Arya <ashr...@codeaurora.org>
commit 39281c0b5d772e925aea40fbce9fd116833e1708
Merge: 52f55e3 da3f09e
Author: Willie Ruan <wr...@quicinc.com>
Date: Tue Jun 8 12:04:45 2010 -0700
Merge "msm: DRM: Add a reference count to the GPU mappings" into msm-2.6.32.9a
commit 52f55e35bf0f792fb5be56b0a478ae8d24392934
Merge: 2367d6e b382094
Author: Willie Ruan <wr...@quicinc.com>
Date: Tue Jun 8 12:04:21 2010 -0700
Merge "EHCI: msm: Avoid write to a read-only CONFIGFLAG register" into msm-2.6.32.9a
commit 2367d6e48431a2274919e0640678b788f4771c57
Merge: b892f76 fc0d901
Author: Pavan Kumar Kondeti <pkon...@qualcomm.com>
Date: Tue Jun 8 00:43:47 2010 -0700
Merge "tty: Allow RI, DSR and CTS modem status bits in tiocmset" into msm-2.6.32.9a
commit da3f09e05247446b90b3a7658ad20b24baf0b648
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Mon Jun 7 15:35:58 2010 -0600
msm: DRM: Add a reference count to the GPU mappings
Use a reference count to avoid unmapping a buffer object from
the GPU MMU until all users release it.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit b3820942a7af5bf4f0209ce6856051d36c186400
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Mon Jun 7 16:31:18 2010 +0530
EHCI: msm: Avoid write to a read-only CONFIGFLAG register
Companion controllers are not found in MSM hardware. All ports
are routed to EHCI controller by default.
Signed-off-by: Pavankumar Kondeti <pkon...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 3b9bed8f1dd4169b0a0e3492c99200a003be7eff
Merge: b892f76 c050c1a
Author: Matt Wagantall <ma...@codeaurora.org>
Date: Fri Jun 4 17:21:49 2010 -0700
Merge remote branch 'cros/master'
All 'cros/master' changes to MSM drivers are dropped as part of
this merge to avoid conflicts with the more actively maintained
'msm-2.6.32.9a versions of the drivers.
The splitconfigs have been manually merged and then regenerated.
Change-Id: I229f419c0d92aa4bc78426ab00eacb3f60e063e3
Signed-off-by: Matt Wagantall <ma...@codeaurora.org>
commit b892f76afff382a92c643ac42dbcfbfd28e8a4d5
Merge: ecc19f8 0fb007d
Author: Murali Palnati <paln...@qualcomm.com>
Date: Thu Jun 3 23:14:33 2010 -0700
Merge "msm: acpuclock-8x50a: Increase EBI1/AXI clock rates for better performance" into msm-2.6.32.9a
commit ecc19f866e1b889af8cbf521f56faf864b2e317f
Merge: 7695048 f19c9e5
Author: Murali Palnati <paln...@qualcomm.com>
Date: Thu Jun 3 23:14:28 2010 -0700
Merge "defconfig: Enable USB Audio Support" into msm-2.6.32.9a
commit 0fb007dc8393e4b36bfb12e9383e154dce7f644e
Author: Matt Wagantall <ma...@codeaurora.org>
Date: Wed Jun 2 17:28:43 2010 -0700
msm: acpuclock-8x50a: Increase EBI1/AXI clock rates for better performance
Change-Id: Ifdb3b58f509e04f70d3ccb6cb9a67bb5d2f10d3f
Signed-off-by: Matt Wagantall <ma...@codeaurora.org>
commit 41dc3d3d995275ccf83e6c5c027c06f7f3d1f8cd
Author: Philip Elcan <pel...@codeaurora.org>
Date: Thu Jun 3 11:34:19 2010 -0400
msm_fb: Improve HDMI insert/remove on ST1.5
The following changes improve reliability when inserting and removing
the HDMI cable on ST1.5:
Switch to using the Hotplug interrupt rather than the Rx Sense. This
eliminates kernel errors when reading the EDID from the monitor too
early.
Reset the HDMI chip in some lockup conditions observed during
stress testing.
Disable the HDMI video output when the cable is removed. This prevents
both the external display coming if the cable is incompletely inserted.
Signed-off-by: Philip Elcan <pel...@codeaurora.org>
commit 868a0daa9bb99fbb1fa037d61b02972bd4729128
Author: Philip Elcan <pel...@codeaurora.org>
Date: Thu Jun 3 11:33:29 2010 -0400
msm: Add platform data for resetting HDMI on ST1.5
Add platform data function to allow the HDMI driver to reset the chip on
ST1.5.
Signed-off-by: Philip Elcan <pel...@codeaurora.org>
commit 7695048ad2a15d7efff6be83a075ccbcbf8dcf72
Merge: 7dac253 9fe2d84
Author: Murali Palnati <paln...@qualcomm.com>
Date: Thu Jun 3 08:10:04 2010 -0700
Merge "usb: msm_gadget: Fix buffer corruption at ep0 status completion" into msm-2.6.32.9a
commit fc0d9012fdfa86b055225ae37a33bdb8e37c9d13
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu Jun 3 15:13:34 2010 +0530
tty: Allow RI, DSR and CTS modem status bits in tiocmset
Allow modem status bits like RI, DSR and CTS in tiocmset. TTY drivers,
for example USB function serial driver, which talks to DCE on behalf
DTE and running in DCE are interested in these bits.
Signed-off-by: Pavankumar Kondeti <pkon...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 7dac2538c414a01fa1ed161ccdbb02df9bcafdf6
Author: Ashish Bijlani <abij...@codeaurora.org>
Date: Thu May 20 16:30:27 2010 -0400
msm_kgsl: Fix GEM KMEM write-combine and non-cached mappings
This change fixes the KMEM WC mappings by flushing out existing
WB mappings. KMEM_NOCACHE defaults to NC mappings.
Change-Id: I0ee7ac92074d7fc9d263a7ed1673612d0be88d7a
Signed-off-by: Ashish Bijlani <abij...@codeaurora.org>
commit 687e5b02e855b62e7b408500e547ae4ac85a31a3
Author: Peter Lohmann <ploh...@codeaurora.org>
Date: Thu May 27 16:52:02 2010 -0700
msm: audio: Increase the size of device name buffer.
To accommodate multi-path devices increase the size of the device
routing name buffer.
Change-Id: Ib5fd178ae505928f7a5908407a62a342b59cece2
Signed-off-by: Peter Lohmann <ploh...@codeaurora.org>
commit 9fe2d8455ed94a4a470667f51c8748dc2bdd400d
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Wed Jun 2 16:17:10 2010 +0530
usb: msm_gadget: Fix buffer corruption at ep0 status completion
ep0 status complete handler is re-using the usb-request buffer from
function drivers for zero length status phase there by corrupting it.
Change the ep0 complete handler to use setup request buffer instead of
function driver's request buffer to fix the corruption.
Signed-off-by: Anji Jonnala <an...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit f19c9e50051a11dd501cb7b2c611fd9d3a94f624
Author: Mayank Rana <mr...@qualcomm.com>
Date: Wed Jun 2 14:44:29 2010 +0530
defconfig: Enable USB Audio Support
Signed-off-by: Mayank Rana <mr...@qualcomm.com>
commit a5db373688c46df1d6bbf42adda6277aa06cff3d
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Tue Jun 1 09:44:44 2010 -0600
msm_kgsl: Move to drivers/gpu/msm
Move the KGSL driver to drivers/gpu/msm to consolidate the
location with the Android driver and other vendors.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit fc32df8abdc67b5d067132a03a197fa24985d709
Author: Matt Wagantall <ma...@codeaurora.org>
Date: Tue May 25 12:30:31 2010 -0700
Generate qsd8650a-st1_5 splitconfig and refresh the others
Use ./chromeos/scripts/kernelconfig to generate a
qsd8650a-st1_5 splitconfig from qsd8650a-st1_5_defconfig, and to
refresh the other splitconfigs.
Signed-off-by: Matt Wagantall <ma...@codeaurora.org>
commit cbd3e7978d9483b4c79660ff8f99c36b88f0657c
Author: Matt Wagantall <ma...@codeaurora.org>
Date: Tue May 25 12:23:55 2010 -0700
defconfig: Sync-up qsd8x50a-st1_5 defconfigs with 2.6.32.9 kernel
Signed-off-by: Matt Wagantall <ma...@codeaurora.org>
commit c738ad42bd480bafbdd7c1ea172f989bef631a5d
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Tue May 25 08:47:56 2010 -0600
msm: Set the max AXI bus speed for 8650A
8650A has a maximum AXI bus speed of 192Mhz rather than
128Mhz on 8x50. Make sure that the GPU attempts to set
the proper speed for the platform.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit 8d0ae4ac0009d6891d52eebd217c5f61068dd57e
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Mon May 17 11:54:36 2010 -0700
Catch-up the DRM driver to the upstream use of 'unlocked_ioctl'
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit e8ce903d41b903d1c804e452ff0b5d1d75a23cd4
Author: Matt Wagantall <ma...@codeaurora.org>
Date: Tue May 11 17:29:14 2010 -0700
drm: Prevent link error when CONFIG_DRM_KMS_HELPER is undefined
Remove references to drm_async_list when its definition in
drm_crtc_helper.c is not present.
Signed-off-by: Matt Wagantall <ma...@codeaurora.org>
commit 8b17e70e177272d2485d40f9ac194ae3a467722e
Author: Stephen Boyd <sb...@codeaurora.org>
Date: Fri May 7 14:39:14 2010 -0700
msm: clock: Invert debugfs directory layout
There are currently 2 separate directories for clock debugging in
debugfs: clk_enable and clk_rate. Each of these directories
contains a list of clocks. This is rather annoying when you are
focusing on one clock and want to enable/disable it and then
check its rate. You either have to cd to the other directory or
cat ../clk_rate/<clk>.
Invert the layout so that there is one clock directory containing
a directory for each clock. Inside each respective clock
directory place an enable and rate file relating to the
clk_enable and clk_disable directories that exist today.
Change-Id: I4a457adfb088039d2961c9716102c575e031b049
Signed-off-by: Stephen Boyd <sb...@codeaurora.org>
commit 14d2970b9677ec5b6faa08874d79504e656dea54
Author: Jin Hong <ji...@codeaurora.org>
Date: Thu May 20 23:45:07 2010 -0700
msm: fix support for older version of TIMEREMOTE_PROG_VER
To support older remote rtc version, it needs to be explicitly
checked.
Change-Id: Ied28054c7ba4dfa9a122d5b770333ef966331e03
Signed-off-by: Jin Hong <ji...@codeaurora.org>
commit 4acdd5c5918336c7eb6ce12c5f4e4a6b333e85f4
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Mon Apr 19 15:25:33 2010 +0530
mmc: msm_sdcc: Do not check for DATA_BLK_END bit to exit the SDCC ISR.
In the SDCC ISR, when checking for the bits set in MCI_STATUS register
look for only those bits that correspond to interrupts enabled through
MCI_INT_MASK0 register. Interrupts on DATA_BLK_END are not enabled and
this bit should not prevent the control to exit from the SDCC ISR. If
this is not ensured, the control gets stuck in the ISR for longer time
till the DATA_END event occurs, until when, the DATA_BLK_END bit keeps
getting set for every data block that is transmitted.
Change-Id: Ic02e3b23af951cb8e5ace0231b26e2692f3d3274
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit 8dc6be077bdf07be6acec5058e6c1ddd71b767cf
Author: Murali Palnati <paln...@qualcomm.com>
Date: Tue Feb 23 15:19:06 2010 +0530
mmc: msm_sdcc: Correct the Module description for the driver.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 274bee1e666b261238ea62766046c5ff355deabe
Author: Murali Palnati <paln...@qualcomm.com>
Date: Tue Feb 23 17:03:43 2010 +0530
mmc: msm_sdcc: Change all printk calls to pr_ macros.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit e4d0acfd3dac86be49074340aaf341e7873c5cae
Author: Murali Palnati <paln...@qualcomm.com>
Date: Fri May 21 23:00:20 2010 +0530
mmc: msm_sdcc: Add support to enable SDIO to wakeup the MSM.
SDIO devices can wake up the MSM on reception of incoming data
via D1 line. This patch adds support in the msm_sdcc driver to
do this.
Change-Id: If56aa4d48795795bdc5966d5719f109aba8753f0
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 4bd5b6987ebfd77f80e2202ce548fa6878bfc59e
Author: Murali Palnati <paln...@qualcomm.com>
Date: Fri May 21 23:01:36 2010 +0530
mmc: msm_sdcc: Support for enabling SDIO to wakeup the MSM.
SDIO devices can wake up the MSM, on reception of incoming data
via D1 line. This patch adds the relevant field in the platform
data for this.
Change-Id: I5ae47a1dd6023b63e2f2e0f5430c1f62c69e96ef
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 738d35c100019de7f3b1185486c79beef97a2313
Author: Murali Palnati <paln...@qualcomm.com>
Date: Fri May 21 18:03:37 2010 +0530
mtd: msm_nand: Use the platform data pointer only if available.
Platform data pointer is used by the driver probe function to check for the
presence of the interleave mode. This pointer can be used only if it is not
NULL. Add a sanity check to check for platform data to be present.
Change-Id: If242ed4a3f8005f5e52bc945386aa01e18936a54
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit f225dac747488bd5f13d328f17b01f3c9991e5d6
Author: Murali Palnati <paln...@qualcomm.com>
Date: Wed May 12 15:02:19 2010 +0530
mtd: msm_nand: Cleanup the prints in the probe function.
There are some prints in the probe function which are really not necessary.
Also some of the prints needed some cleanup like eliminating spaces at the
end of the line and using __func__ instead of hard coded strings.
Change-Id: Id42feeb9c21854adc1781bc297c236d643016726
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit 20334c10cf32c28fb10e577ff304781da592b79e
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Mon May 17 14:47:34 2010 +0530
mtd: msm_nand: Correct the DMA buffer memory allocation in nand driver.
Add checks in the function nand_get_dma_buffer_request to ensure
that it does not succeed when requested size is greater than the
available memory.
Change-Id: I7536e7b4d4ac70dff0127e3b504bbebf2d4fc5f5
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit a436574db5a7a883ae73d959d0407ef6231542b9
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Tue May 18 16:46:43 2010 +0530
mtd: msm_nand: Enable interleave mode based on modem build ID.
Add build id detection to enable the interleave mode feature in the driver.
In the interleave mode image ( boot, userdata etc ) codewords are written
to the NAND devices in a interleave fashion. Hence modem and apps should
operate in the same mode to read/write the images properly.
Change-Id: I84107d665d72df320ff48441af117083c1c0034f
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit 9efe07dc1cd1a3326d8fd0ae7062d2e6fc359f11
Author: Neil Leeder <nle...@codeaurora.org>
Date: Mon May 17 18:05:28 2010 -0400
keyboard: update qcikbd for ST1.5 to use keypad keys
The directional keys no longer work in recent drops of the chrome
window manager. Replace them with the keypad keys which do work.
Change-Id: I8d1cc90f4aaf660fd9dca4d6afcff6a2a9c564f4
Signed-off-by: Neil Leeder <nle...@codeaurora.org>
commit 04e4b1744c31fe7d4ef7c3606dc7c6b4b96de482
Author: Jin Hong <ji...@codeaurora.org>
Date: Tue May 18 16:18:24 2010 -0700
msm: support multiple TIMEREMOTE_PROG_VER
The TIMEREMOTE_PROG_VER mismatch with modem causing rtc device not
created properly. Added logic to support new version and backward
compatible with old version.
Change-Id: Iace9143c0827506d2828afd540db215c38dc34ae
Signed-off-by: Jin Hong <ji...@codeaurora.org>
commit 322f5c23110ce5b19ab089b2d08e33c9fa672267
Author: Ashish Bijlani <abij...@codeaurora.org>
Date: Tue May 18 12:20:25 2010 -0400
msm: mdp: fixes race condition while serving new blit requests
This change fixes the race condition while creating new display jobs.
Change-Id: Iab4d199a0c13bce19105155bfcb42105550e40e2
Signed-off-by: Ashish Bijlani <abij...@codeaurora.org>
commit ae8e402fe569e7800bde25eb23a907aa7696ee04
Author: Ashish Bijlani <abij...@codeaurora.org>
Date: Tue May 18 11:27:30 2010 -0400
msm_fb: set LCDC underflow color to red on ST1.5
Changes the default displayed color in case of LCDC underflow to red
Change-Id: I44d241541872479c61fd5f62a21e05d083dc35b7
Signed-off-by: Ashish Bijlani <abij...@codeaurora.org>
commit 082284b3301c609976d046fd3bfe29007872c371
Author: Mohan Pallaka <mpal...@qualcomm.com>
Date: Tue May 4 19:58:54 2010 +0530
msm: handset: Add support for 3-switch handset
Add support for 3-switch handsets by receiving new keycodes
from modem through rpc. The corresponding Linux keycodes are
configured as VOLUP and VOLDOWN, and can be changed depending
on requirement.
Change-Id: Id99edc842b0195b2a331a153289e02e1b9b63d20
Signed-off-by: Mohan Pallaka <mpal...@qualcomm.com>
commit ef39a9cb64c2af7518b9fd0b1373b24e2b5049ab
Author: Mohan Pallaka <mpal...@qualcomm.com>
Date: Tue May 4 19:12:38 2010 +0530
msm: handset: Add support for HS 3.0
Add support for HS3.0 while maintaining the backward
compatibility.
Change-Id: If52ef79face053f161fcc404c63c283dd6551fc4
Signed-off-by: Mohan Pallaka <mpal...@qualcomm.com>
commit 56ec0a5e89771315c1729ede7f0c2f86ddbf4a53
Author: Peter Lohmann <ploh...@codeaurora.org>
Date: Mon May 17 16:18:15 2010 -0700
msm: audio: Implement HDMI audio support.
Signed-off-by: Peter Lohmann <ploh...@codeaurora.org>
commit 66553ff3d36e095074b94a22eb1b97dfd5fd57f4
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Mon May 17 14:27:48 2010 +0530
mtd: msm_nand: Convert all NAND_ defines to MSM_NAND_
Needed to prevent collisions with defines in <linux/mtd/nand.h>
Change-Id: I2c82b6517947b65a5750752fe173c43d888a42e9
Signed-off-by: Dima Zavin <di...@android.com>
[mnal...@qualcomm.com: Modified OneNAND defines too]
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit f01f6cd08f8d454175f1aef6ac2b87b923f01602
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Mon Feb 8 19:18:26 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Allow SCSI devices to suspend/resume asynchronously
PM: Allow SCSI devices to suspend/resume asynchronously
Set power.async_suspend for all SCSI devices, targets and hosts, so
that they can be suspended and resumed in parallel with the main
suspend/resume thread and possibly with other devices they don't
depend on in a known way (i.e. devices which are not their parents or
children).
The power.async_suspend flag is also set for devices that don't have
suspend or resume callbacks, because otherwise they would make the
main suspend/resume thread wait for their "asynchronous" children
(during suspend) or parents (during resume), effectively negating the
possible gains from executing these devices' suspend and resume
callbacks asynchronously.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 4cb077d93a57fb8bb64c277ec4ce637823cd76f4)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit e800ef043d163959b2531ed52f4cfe7643ae6eeb
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Mon Feb 8 19:18:16 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Allow USB devices to suspend/resume asynchronously
PM: Allow USB devices to suspend/resume asynchronously
Set power.async_suspend for USB devices, endpoints and interfaces,
allowing them to be suspended and resumed asynchronously during
system sleep transitions.
The power.async_suspend flag is also set for devices that don't have
suspend or resume callbacks, because otherwise they would make the
main suspend/resume thread wait for their "asynchronous" children
(during suspend) or parents (during resume), effectively negating the
possible gains from executing these devices' suspend and resume
callbacks asynchronously.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 927bc9165d3cd01054b472f9b25d91a8707209c8)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 8fd43326671e885d618e8ce61c1e77db567ad0ca
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Mon Feb 8 19:16:33 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Allow PCI devices to suspend/resume asynchronously
PM: Allow PCI devices to suspend/resume asynchronously
Set power.async_suspend for all PCI devices and PCIe port services,
so that they can be suspended and resumed in parallel with other
devices they don't depend on in a known way (i.e. devices which are
not their parents or children).
This only affects the "regular" suspend and resume stages, which
means in particular that the restoration of the PCI devices' standard
configuration registers during resume will still be carried out
synchronously (at the "early" resume stage).
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit a1e4d72cd3024999bfb6703092ea271438805c89)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 23e3df4cdff64ac32a17e3c0ff602bde998f64f4
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Sat Jan 23 22:25:31 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Start asynchronous resume threads upfront
PM: Start asynchronous resume threads upfront
It has been shown by testing that total device resume time can be
reduced significantly (by as much as 50% or more) if the async
threads executing some devices' resume routines are all started
before the main resume thread starts to handle the "synchronous"
devices.
This is a consequence of the fact that the slowest devices tend to be
located at the end of dpm_list, so their resume routines are started
very late. Consequently, they have to wait for all the preceding
"synchronous" devices before their resume routines can be started
by the main resume thread, even if they are "asynchronous". By
starting their async threads upfront we effectively move those
devices towards the beginning of dpm_list, without breaking their
ordering with respect to their parents and children. As a result,
their resume routines are started much earlier and we are able to
save much more device resume time this way.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 97df8c12995c5bac73e3bfeea4c5be155c1f4401)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 45d4b90c6b32ff2dcfc70813b243ee20d76d70f3
Author: Sameer Nanda <sna...@chromium.org>
Date: Wed May 19 13:33:42 2010 -0700
UPSTREAM: torvalds_linux-2.6.git: PM: Add facility for advanced testing of async suspend/resume
PM: Add facility for advanced testing of async suspend/resume
Add configuration switch CONFIG_PM_ADVANCED_DEBUG for compiling in
extra PM debugging/testing code allowing one to access some
PM-related attributes of devices from the user space via sysfs.
If CONFIG_PM_ADVANCED_DEBUG is set, add sysfs attribute power/async
for every device allowing the user space to access the device's
power.async_suspend flag and modify it, if desired.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 5a2eb8585f3b38e01e30aacaa8b985a1520a993d)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit be9e5f313703ea3020dc72d01b51fbac438ce976
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Sat Jan 23 22:25:15 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Add a switch for disabling/enabling asynchronous suspend/resume
PM: Add a switch for disabling/enabling asynchronous suspend/resume
Add sysfs attribute /sys/power/pm_async allowing the user space to
disable/enable asynchronous suspend/resume of devices.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 0e06b4a891c6a108412fe24b4500f499da2cf8a1)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 9b87324806dd606cd4d0056d05aa39143ab027b0
Author: Sameer Nanda <sna...@chromium.org>
Date: Wed May 19 10:57:56 2010 -0700
UPSTREAM: torvalds_linux-2.6.git: PM: Asynchronous suspend and resume of devices
PM: Asynchronous suspend and resume of devices
Theoretically, the total time of system sleep transitions (suspend
to RAM, hibernation) can be reduced by running suspend and resume
callbacks of device drivers in parallel with each other. However,
there are dependencies between devices such that we're not allowed
to suspend the parent of a device before suspending the device
itself. Analogously, we're not allowed to resume a device before
resuming its parent.
The most straightforward way to take these dependencies into accout
is to start the async threads used for suspending and resuming
devices at the core level, so that async_schedule() is called for
each suspend and resume callback supposed to be executed
asynchronously.
For this purpose, introduce a new device flag, power.async_suspend,
used to mark the devices whose suspend and resume callbacks are to be
executed asynchronously (ie. in parallel with the main suspend/resume
thread and possibly in parallel with each other) and helper function
device_enable_async_suspend() allowing one to set power.async_suspend
for given device (power.async_suspend is unset by default for all
devices). For each device with the power.async_suspend flag set the
PM core will use async_schedule() to execute its suspend and resume
callbacks.
The async threads started for different devices as a result of
calling async_schedule() are synchronized with each other and with
the main suspend/resume thread with the help of completions, in the
following way:
(1) There is a completion, power.completion, for each device object.
(2) Each device's completion is reset before calling async_schedule()
for the device or, in the case of devices with the
power.async_suspend flags unset, before executing the device's
suspend and resume callbacks.
(3) During suspend, right before running the bus type, device type
and device class suspend callbacks for the device, the PM core
waits for the completions of all the device's children to be
completed.
(4) During resume, right before running the bus type, device type and
device class resume callbacks for the device, the PM core waits
for the completion of the device's parent to be completed.
(5) The PM core completes power.completion for each device right
after the bus type, device type and device class suspend (or
resume) callbacks executed for the device have returned.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 5af84b82701a96be4b033aaa51d86c72e2ded061)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 5b9c6d3b9bfaae7c0674500a3e657638661f4964
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Sat Jan 23 22:03:29 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Add parent information to timing messages
PM: Add parent information to timing messages
Add parent information to the messages printed by the suspend/resume
core when initcall_debug is set.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 8cc6b39ff36b4bbce2d7471da088df122b0e9033)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 555571dad6ed1dc8a684527bde503a7bf03353f0
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Wed Jan 27 23:47:38 2010 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Allow device drivers to use dpm_wait()
PM: Allow device drivers to use dpm_wait()
There are some dependencies between devices (in particular, between
EHCI USB controllers and their OHCI/UHCI siblings) which are not
reflected by the structure of the device tree. With synchronous
suspend and resume these dependencies are taken into accout
automatically, because the devices in question are always registered
in the right order, but to meet these constraints with asynchronous
suspend and resume the drivers of these devices will need to use
dpm_wait() in their suspend/resume routines, so introduce a helper
function allowing them to do that.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit f8824cee405c62ba465b85365201166d9cf86a14)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 83176c4ba5514357b75bc231ba10f1e47fd2821a
Author: Rafael J. Wysocki <r...@sisk.pl>
Date: Fri Dec 18 01:57:31 2009 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Make the initcall_debug style timing for suspend/resume complete
PM: Make the initcall_debug style timing for suspend/resume complete
Commit f2511774863487e61b56a97da07ebf8dd61d7836
(PM: Add initcall_debug style timing for suspend/resume) introduced
basic timing instrumentation, needed for a scritps/bootgraph.pl
equivalent or humans, but it missed the fact that bus types and
device classes which haven't been switched to using struct dev_pm_ops
objects yet need special handling. As a result, the suspend/resume
timing information is only available for devices whose bus types or
device classes use struct dev_pm_ops objects, so the majority of
devices is not covered.
Fix this by adding basic suspend/resume timing instrumentation for
devices whose bus types and device classes still don't use struct
dev_pm_ops objects for power management. To reduce code duplication
move the timing code to helper functions.
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit 875ab0b74e85d6801a49392447d26e0b28688d86)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit d3411cc6f92f28155cc3c51e873f4bf8481383be
Author: Arjan van de Ven <ar...@infradead.org>
Date: Sun Dec 13 20:29:01 2009 +0100
UPSTREAM: torvalds_linux-2.6.git: PM: Add initcall_debug style timing for suspend/resume
PM: Add initcall_debug style timing for suspend/resume
In order to diagnose overall suspend/resume times, we need
basic instrumentation to break down the total time into per
device timing, similar to initcall_debug.
This patch adds the basic timing instrumentation, needed
for a scritps/bootgraph.pl equivalent or humans.
The bootgraph.pl program is still a work in progress, but
is far enough along to know that this patch is sufficient.
Signed-off-by: Arjan van de Ven <ar...@linux.intel.com>
Signed-off-by: Rafael J. Wysocki <r...@sisk.pl>
(cherry picked from commit f2511774863487e61b56a97da07ebf8dd61d7836)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 04df1b9f1b578639e605fce80b5f63e4401cea2a
Author: Kenji Kaneshige <kaneshi...@jp.fujitsu.com>
Date: Wed Nov 25 21:01:28 2009 +0900
UPSTREAM: torvalds_linux-2.6.git: PCI: portdrv: cleanup pcie_device registration
PCI: portdrv: cleanup pcie_device registration
In the current port bus driver implementation, pcie_device allocation,
initialization and registration are done in separated functions. Doing
those in one function make the code simple and easier to read.
Signed-off-by: Kenji Kaneshige <kaneshi...@jp.fujitsu.com>
Signed-off-by: Jesse Barnes <jba...@virtuousgeek.org>
(cherry picked from commit 52a0f24beabe9e89223e367c65a0156dff17265c)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit d1dfb95936676cda4f7980684aa6d9a893a831e5
Author: Sameer Nanda <sna...@chromium.org>
Date: Wed May 19 10:16:46 2010 -0700
UPSTREAM: torvalds_linux-2.6.git: PCI: portdrv: remove redundant pcie_port_device_probe
PCI: portdrv: remove redundant pcie_port_device_probe
We don't need pcie_port_device_probe() because we can get pci
device/port type using pci_is_pcie() and 'pcie_type' fields in struct
pci_dev. Remove pcie_port_device_probe().
Signed-off-by: Kenji Kaneshige <kaneshi...@jp.fujitsu.com>
Signed-off-by: Jesse Barnes <jba...@virtuousgeek.org>
(cherry picked from commit 898294c97500b1cdff6edce52fd34e024eb070ec)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit c481bd4345ac59803371f902b47081df8d864ff0
Author: Sameer Nanda <sna...@chromium.org>
Date: Wed May 19 10:14:33 2010 -0700
UPSTREAM: torvalds_linux-2.6.git: PCI: introduce pci_is_pcie()
PCI: introduce pci_is_pcie()
Introduce pci_is_pcie() which returns true if the specified PCI device
is PCI Express capable, false otherwise.
The purpose of pci_is_pcie() is removing 'is_pcie' flag in the struct
pci_dev, which is not needed because we can check it using 'pcie_cap'
field. To remove 'is_pcie', we need to update user of 'is_pcie' to use
pci_is_pcie() instead first.
Signed-off-by: Kenji Kaneshige <kaneshi...@jp.fujitsu.com>
Signed-off-by: Jesse Barnes <jba...@virtuousgeek.org>
(cherry picked from commit 7eb776c42e75d17bd8107a1359068d8c742639d1)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit 2fbc306c8d7897ea48c4e8e688e53b46e7d2181b
Author: Kenji Kaneshige <kaneshi...@jp.fujitsu.com>
Date: Thu Nov 5 12:05:11 2009 +0900
UPSTREAM: torvalds_linux-2.6.git: PCI: cache PCIe capability offset
PCI: cache PCIe capability offset
There are a lot of codes that searches PCI express capability offset
in the PCI configuration space using pci_find_capability(). Caching it
in the struct pci_dev will reduce unncecessary search. This patch adds
an additional 'pcie_cap' fields into struct pci_dev, which is
initialized at pci device scan time (in set_pcie_port_type()).
Signed-off-by: Kenji Kaneshige <kaneshi...@jp.fujitsu.com>
Signed-off-by: Jesse Barnes <jba...@virtuousgeek.org>
(cherry picked from commit 0efea0006335a2425b1a12a2ad35efad626fe353)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
(cherry picked from commit 72b784826eb3a076e511bc749982548a7f41e7da)
Signed-off-by: Sameer Nanda <sna...@chromium.org>
commit c762162cd9b771dd4a64f243740085c0562e17ee
Author: Olof Johansson <ol...@chromium.org>
Date: Wed May 19 14:50:01 2010 -0700
CHROMIUM: Enable CONFIG_FUNCTION_TRACER
Since it can be used to do nifty tracing of startup times
According to the Kconfig help text, it should have no measurable performance impact as long as it's not runtime enabled (boot time default is disabled).
Signed-off-by: Olof Johansson <ol...@chromium.org>
Review URL: http://codereview.chromium.org/2086007
commit 730411d6bb46aa8a443b4d8535178b10c9693442
Author: Girts Folkmanis <gi...@chromium.org>
Date: Wed May 19 13:14:22 2010 -0700
Add initrd support to all kernel configs.
Initrd image will be used by factory install shim.
Review URL: http://codereview.chromium.org/2092010
commit cd95cb64897cbd6ee74bb282fae2ded6562d1a43
Author: Jeongbae Seo <jeongb...@samsung.com>
Date: Tue May 18 15:50:10 2010 -0700
ARM: VOGUEV210: Add S5PV210 Watchdog Support
This patch adds Watchdog Support for VOGUEV210
Signed-off-by: Jeongbae Seo <jeongb...@samsung.com>
Signed-off-by: Daein Moon <moon...@samsung.com>
Acked-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/1848001
commit 0a4df21edd098bfb38e699f07fbcea578a297a22
Author: Jeongbae Seo <jeongb...@samsung.com>
Date: Tue May 18 15:47:43 2010 -0700
ARM: S5PV210: Add S5PV210 Watchdog Support
This patch adds S5PV210 Watchdog Support
Signed-off-by: Jeongbae Seo <jeongb...@samsung.com>
Signed-off-by: Daein Moon <moon...@samsung.com>
Acked-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/1848001
commit 627781c58d83bbee4caaaffa503256c6f8584592
Author: Geunjae Lee <gj2...@samsung.com>
Date: Wed May 12 16:06:45 2010 -0700
ARM: S5P: Add TV out device reference for Samsung S5P SoCs
This patch adds TV out device reference for Samsung S5P SoCs.
Signed-off-by: Geunjae Lee <gj2...@samsung.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/2044008
commit 73ff4753ddb06339242a39da0346b3b41d72995f
Author: Geunjae Lee <gj2...@samsung.com>
Date: Wed May 12 16:14:54 2010 -0700
ARM: S5PV210: Update clock register file
This patch updates regs-clock.h for Samsung S5PV210 TV out driver.
Signed-off-by: Geunjae Lee <gj2...@samsung.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/1996015
commit 20b95ca960e706f8c2039b0ecf75634b581acc9a
Author: Geunjae Lee <gj2...@samsung.com>
Date: Wed May 12 16:27:33 2010 -0700
ARM: S5PV210: Add TV out driver register definition files
This patch adds TV out driver register definition files for Samsung S5PV210.
Signed-off-by: Geunjae Lee <gj2...@samsung.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/2060003
commit a8fc6c0426e8b256432a8c9300813c3ff820ccc8
Author: Geunjae Lee <gj2...@samsung.com>
Date: Wed May 12 16:31:45 2010 -0700
ARM: S5P: Add TV out device definition for Samsung S5P SoCs.
This patch adds TV out device definition for Samsung S5P SoCs.
Signed-off-by: Geunjae Lee <gj2...@samsung.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/2010012
commit bfb2303a10f295f94833dadbddae260c125b867d
Author: Geunjae Lee <gj2...@samsung.com>
Date: Wed May 12 16:34:06 2010 -0700
ARM: S5PV210: Modify vogue machine file for TV out support
This patch modifies vogue machine file for TV out support.
Signed-off-by: Geunjae Lee <gj2...@samsung.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/2003015
commit c8ea79f03cabfab521a6ced2dffc55c4b3143504
Author: Mandeep Singh Baines <m...@chromium.org>
Date: Tue May 18 16:11:28 2010 -0700
CHROMIUM: config: run oldconfig
Cleanup due to recently added Kconfig options. No impact to build.
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/2075013
commit c3d9f7e4c8dabc663bbb23519481ea4d83644b12
Author: Bryan Freed <bfr...@google.com>
Date: Tue May 18 15:21:22 2010 -0700
Version 1.02 of the Realtek USB card reader driver fixes the lsusb hang.
The important part appears to be the mutex lock in rts51x_polling_func().
Review URL: http://codereview.chromium.org/2064009
commit b984af982a5744e56dc8272b937d1bb7cc6419cf
Author: Philip Elcan <pel...@codeaurora.org>
Date: Wed May 12 15:11:12 2010 -0400
msm_fb: Check external display capabilities for ST1.5
Read the EDID from an attached external HDMI display and verify it
supports the expected resolution.
Change-Id: I6eedc4bd71a3f393243f7cc6d6051b3090d9c118
Signed-off-by: Philip Elcan <pel...@codeaurora.org>
commit 5dfa101e420f1e5404ac93c5edf9f1f89aad436b
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Fri May 14 17:11:01 2010 +0530
EHCI: msm: shutting down VBUS during platform suspend
As part of platform suspend shutting down the VBUS and resuming back the
VBUS as part of platform resume.
Change-Id: Ieed07d66019ffee30d8a719212d077b8c49397cb
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit ec662b0a673fc2202f2a026c2ef4d979cb605ac3
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Thu May 13 15:56:35 2010 -0600
msm_fb: Fix incorrect pitch alighment math
The math to align the pitch to 32 pixels was incorrect. We were
instead aligning to 32 bytes which resulted in the wrong value
when the bit depth was 32bpp.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit a4949108452ea6047b66e13ec66d05c4c8ccbfe2
Author: Murali Palnati <paln...@qualcomm.com>
Date: Thu Jul 23 11:26:11 2009 +0530
mtd: Collect bad block count for ecc stats lazily.
Normally bad block counts for ECC stats are collected during boot time.
This can be done lazily when the ECCGETSTATS ioctl is invoked on the
partition. This can significantly decrease boot time, depending on the
size of the partition. Also rescanning on every ioctl invocation helps
in having the latest bad block count rather than depending on the count
that is collected during boot.
Change-Id: Ib6d445d5f74d832d09af0db34f1d8025e6ec2486
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit f2a31f20ad75c662a760a741d557490b82fec377
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu May 13 12:59:41 2010 +0530
usb: msm_gadget: Register with switch to send online/offline uevents
User space applications depends on uevents for knowing the state of
usb. Hence send uevents based on the connection of usb through switch
class. Add usb state sysfs entry support as well.
Send offline event when reset interrupt occurs in configured
state. Otherwise user space applications may still perform I/O
which eventually leads to high CPU usage.
This patch also fixes the bug where an incorrect event is sent
when no configuration is selected.
Change-Id: Ie421e099ea1e7611a00c8874a1885515df9e03ba
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 33c61d09ce3243ad19b151c5ca651602727fa8f4
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu May 13 14:47:10 2010 +0530
usb: gadget: Add rmnet function driver support
Rmnet function is an alternative to standard CDC-ECM and windows
proprietary RNDIS. This driver acts like a bridge between Host
and modem found in Qualcomm chipsets. Two SMD channels are used to
communicate with modem for control and data transfers respectively.
Provide two config options to specify control and data SMD channels.
Change-Id: I4feb2272865b189fdd3254de82b7b0af8db03d7e
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit b8678d3aac549d2806e30ed937b531dbaf9f7b96
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu May 13 14:42:49 2010 +0530
usb: gadget: Add serial driver support
Add serial driver support for chrome and add modem support in
serial driver.
Change-Id: Ifc158494df2a25221e604d11eb0aa5b273ee9a85
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 214763460ae52ce75597db5c82c895f8205a24ba
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Wed May 12 14:57:17 2010 -0600
msm_fb: Set the default color depth from the kernel config
Make the default color depth for the framebuffer configurable via
the kernel config. Default is RGB565, but ARGB8888 and RGBA8888
are available.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit 931e3e83db7926397196dd9f0390eba271db7dfc
Author: Murali Palnati <paln...@qualcomm.com>
Date: Wed May 12 18:05:19 2010 +0530
mmc: msm_sdcc: Remove CONFIG_MMC_MSM7X00A which is redundant.
Support for MSM SDCC controller can be selected by the option
CONFIG_MMC_MSM. CONFIG_MMC_MSM7X00A is the older option which
is redundant now and should be removed.
Change-Id: I999eb32134863a4e28ff72cd907d012fe831eb86
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 5f568b71dac538e2e680512d16438deb229b92aa
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Mon May 10 17:47:47 2010 -0600
msm_fb: Use varinfo information to choose the proper color config
Use the position of the transparent portion of the color information
in varinfo to determine if we should use RGBA8888 or ARGB8888 when
32bpp is selected.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit 3e9f65069eccf0a76b2558066bb09ec19390a694
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Tue May 11 11:44:52 2010 +0530
uvcvideo: Increase the streaming control time out
For the driver to configure the device correctly, the streaming
control timeout should be increased.
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit f1daeb7e039e62df548adc26cd8c9863f7b1fdf5
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Mon May 10 17:47:40 2010 -0600
msm_fb: Add support for 32bpp framebuffer color depth
Add support for RGBA_8888 mode at init time and to allow the
user to specify 32bpp color depth when changing the mode.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit c40a37f50c532580fedb95cf7b0b446dca107232
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Mon May 10 16:59:56 2010 -0600
msm: Bump up the default size of the ST15 framebuffer
Bump up the size of the ST15 framebuffer to accomodate
1376 x 1536 @ 32bpp.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit 22f126f481ed276994ea3e385b5b1880a2f80b9f
Author: Jordan Crouse <jcr...@codeaurora.org>
Date: Tue May 11 17:40:05 2010 -0600
msm_fb: Don't default to 32bpp for MDP 3.1
Go back to 16bpp as the default bit depth for MDP 3.1 until we have
more stable 32bpp support.
Signed-off-by: Jordan Crouse <jcr...@codeaurora.org>
commit 4dcd668ff21a165b7988b473cab22d2afe04c725
Author: Neeraj Kumar <c_ne...@quicinc.com>
Date: Mon May 10 16:59:54 2010 -0600
msm_fb: Enabled 32 bpp framebuffer for MDP 3.1
Modified DMA routined to support RGBA 8888 format
for 32 bit frame buffer.
Change-Id: I755da9c8be0a23ef175c840064f9981923df1316
Signed-off-by: Neeraj Kumar <c_ne...@quicinc.com>
commit e009c668f96be3484629094ac00180f8790b8681
Author: Neil Leeder <nle...@quicinc.com>
Date: Mon May 10 12:55:37 2010 -0400
msm_fb: disable HDMI interrupts while panel is off for ST1.5
The SiI9022 chip can generate spurious interrupts during
panel off processing. Make sure we don't attempt to process
these while the panel is off.
Change-Id: I497c24d0980517832fc5d30b51ed9870dd88e533
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit 99ce293acdc211022484e386644e705dafa0d51c
Author: Neil Leeder <nle...@quicinc.com>
Date: Mon May 10 12:54:51 2010 -0400
msm_fb: add support for HDMI cable detect on ST1.5
When an HDMI cable is inserted or removed work function updates
screen resolution for either HDMI or on-board panel. This is
communicated to userspace via a sysfs entry. Userspace daemon will
issue ioctl which results in panel off/on processing and
resolution change.
Change-Id: I8a17993ab4c3fff60fe1ce7457ae2831a659262a
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit 5a5b1c2f42ad7d8b0c64e82fb630f6ff937884b8
Author: Neil Leeder <nle...@quicinc.com>
Date: Mon May 10 12:54:27 2010 -0400
msm_fb: initialize HDMI chip during panel on processing on ST1.5
During power saving HDMI chip is turned off, so it needs to
be re-initialized during panel on processing. Add platform data
that adds a callback to re-do the i2c init sequence.
Change-Id: Ia9d9bad700f1ddcad733ef2cd2b2e21ed6873be5
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit 731e94b44438268741e77db8db3fc91ee9d9acb2
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Mon May 10 13:27:34 2010 +0530
usb: msm_otg: Fix race conditions while switching to peripheral mode
During switching the mode from host to peripheral, stopping the host
and going into lpm happens as part of work queue. usb remains in LPM
even after starting the peripheral, thereby causing the peripheral
mode failure.
When switching to peripheral mode via sysfs avoid putting usb to LPM.
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 38005262e4e8954876953b99cfa72c94b7b4d80b
Author: Murali Nalajala <mnal...@quicinc.com>
Date: Tue Apr 27 15:29:25 2010 +0530
msm: msm_nand: Correct the ONFI device parameter id match for MT29F4G08ABC
When ONFI device parameters are queried, the device model name is
received without any NULL character at the end of the string.
A normal string compare would fail on this. So replace that with
an strncmp.
Change-Id: Ib67109986d1759afd5222b7e041b254c328815c7
Signed-off-by: Murali Nalajala <mnal...@quicinc.com>
commit b1e96dd92b1a7e5b594b78851c79d73771eb0e7a
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue May 4 17:22:06 2010 -0400
msm: add reset pulse to HDMI for ST1.5
SiI9022 datasheet requires a 50us pulse on the reset line. Previously
the reset line was being released with no pulse.
Change-Id: I029f259e09bc36939ca697595695cc2fbe12549e
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit b8dd0abcad249d8aa76510de898b83fe30bcd1f0
Author: Philip Elcan <pel...@quicinc.com>
Date: Wed May 5 11:10:04 2010 -0400
msm: Move TPM to QUP bus for ST1.5
Newer ST1.5 RevB boards now support the TPM on the QUP I2C bus as
designed.
Change-Id: Ib79b43352b5a4479308eae196861bb49fe45d593
Signed-off-by: Philip Elcan <pel...@quicinc.com>
commit c4211f0a8798aa0a5f2a499f0495c0b84ca2f456
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Thu May 6 17:37:26 2010 -0600
msm_kgsl: Remove kernel.h from linux/msm_kgsl.h
kernel.h isn't needed by msm_kgsl.h, and removing it prevents
any problems with kernel specific headers used in userspace.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 0b38b68a242466a85b5195ca62baad940b88c1d1
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Thu May 6 17:37:25 2010 -0600
msm_kgsl: Fix idle check for GPU
GPU is really idle when rbbm_status register value is 0x110
Change-Id: I37fcb5da80b50d66c989a32e231a3d87d57a08ad
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit e3c56c6cec72b4bacb15c52f31829ce001da5dc8
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu May 6 18:20:18 2010 +0530
usb: ehci: EHSET Test-Fixture device driver for host compliance
An EHSET test fixture is used to initiate test modes on an embedded
host controller while performing the usb host compliance testing.
During enumeration by the USB host this test fixture presents a VID/PID
pair and this PID value corresponds to the requested testmode.
Add documentation for EHSET host compliance testing as well.
CRs-Fixed: 174266
Signed-off-by: Manu Gautam <mga...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 37557de6e2a414a9245d2b7f78667a3517f1e745
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Mon May 3 11:54:55 2010 -0600
qup_i2c: Enhance error reporting for NACKs.
Many I2C slave devices (e.g. camera sensors) are probed even if not
connected. In that case, reporting NACK as "error" may mislead
log-readers. So it is reported as device not connected with slave
address for better understanding of what I2C slave is not connected.
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 16d92c07c8be3298ae5fd7ec49936d6123ced7b8
Author: Bryan Huntsman <bry...@quicinc.com>
Date: Tue May 4 15:35:30 2010 -0700
msm: pm2: fix arm_pm_restart assignment warning
kernel/arch/arm/mach-msm/pm2.c: In function 'msm_pm_init':
kernel/arch/arm/mach-msm/pm2.c:1610: warning: assignment from incompatible pointer type
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
commit 8e93e4fe71eaf2fd4d1a59006cf62ca99e4cfa57
Author: Bryan Huntsman <bry...@quicinc.com>
Date: Tue May 4 16:32:11 2010 -0700
defconfig: remove qsd8650-st1 defconfigs
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
commit 2cf69324a5ef8db6fe4da4ac62cd8aa12949d275
Author: Tianyi Gou <tg...@quicinc.com>
Date: Tue May 4 14:43:43 2010 -0700
msm: add irq value for battery driver for ST1.5
Signed-off-by: Tianyi Gou <tg...@quicinc.com>
commit 461788d03d4c1e8ab54fc58f351ec23ed0f753fe
Author: Manu Gautam <mga...@qualcomm.com>
Date: Wed Feb 24 18:42:17 2010 +0530
usb: ehci: Add support for SINGLE_STEP_SET_FEATURE test of EHSET
Embedded High-speed Host Electrical Test or EHSET defines following
tests for a USB Host port:
1: TEST_SE0
2: TEST_J
3: TEST_K
4: TEST_PACKET
5: HS_HOST_PORT_SUSPEND_RESUME
6: SINGLE_STEP_GET_DEV_DESC
7: SINGLE_STEP_SET_FEATURE
Tests 1-4 needs support from EHCI i.e setting test control bits in
PORTSC register.
Tests 5&6 dont require any additional support from EHCI stack.
Where as 7th test requires support in EHCI and is being added in this
patch.
For SINGLE_STEP_SET_FEATURE test, first the SETUP request of GetDesc
is sent which is followed by the IN stage after a delay of 15 secs.
SINGLE_STEP_SET_FEATURE test can be initiated like any other 1-4 Test
by issuing a SetFeature request to the root hub with Test-Selector
value = 0x06.
Change-Id: I6ad2bd5fc14c31c80f0291a82fef40c2c5a9045a
Signed-off-by: Manu Gautam <mga...@qualcomm.com>
commit 426408072b27584c072f84b5f44a293dd1251234
Author: Susan Brownhill <sbro...@quicinc.com>
Date: Wed Apr 14 16:37:12 2010 -0700
msm_q6venc: pass encoder buffer source to dsp
Introduce a field to record buffer source info, and use it
to pass output buffer info to dsp.
Signed-off-by: Susan Brownhill <sbro...@quicinc.com>
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit ab553b378370088dce68f4a0483a7938f5a2af7a
Author: Rajesh Sastrula <vra...@quicinc.com>
Date: Tue Apr 27 12:18:00 2010 -0700
msm: qdsp6: Allow only one instance of video decoding
Do not allow more than one instance since it needs more pmem
and multiple instances is not tested well.
Signed-off-by: Rajesh Sastrula <vra...@quicinc.com>
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit 14161a0dca1f4d31b11ba7352508d346c777f526
Author: Derek Kulinski <dkul...@quicinc.com>
Date: Fri Apr 23 15:05:38 2010 -0700
msm: audio: configuration option to remove voice driver
ChromeOS doesn't need voice driver for its operation.
This should speed up the boot process by ~180ms
Change-Id: I6f370ff545f35e61b5b09baa8035ad3f1cd61581
Signed-off-by: Derek Kulinski <dkul...@quicinc.com>
commit c5a1eab97dec86362c3ed49289ca9f8ec288450c
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue Apr 20 13:10:37 2010 -0400
msm: add cable detection for ST1.5 HDMI display
Register a callback with the display driver which will select the
correct display to enable when an HDMI cable plug/unplug
event is detected by the display driver on ST1.5
Change-Id: Iaad972781e1f9555a73ef8056a8517d43f63f3ec
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit 2a007a7d2ca1d2c33426ab2247535082d7161bba
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Apr 28 08:59:11 2010 -0600
msm_kgsl: Fix a typo in an #ifdef to enable SMI memory regions
Change CONFIG_PMEM_SMI_REGION to the correct CONFIG_KERNEL_PMEM_SMI_REGION
which will let the DRM module create blocks of memory in SMI.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 34483a29fedd7f08d0ada0c806028adfa5140364
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Apr 28 08:56:21 2010 -0600
msm_kgsl: Reset the hardware status when G12 is closed
Make sure to reset the hardware status in case G12 was closed while
it was idle. This prevents the device from being reopened in a
bad state.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 0768bacfe077204d0e83241b6e43124167854c84
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Apr 28 08:51:48 2010 -0600
msm_kgsl: Remove the G12 bottom half handler
The interrupt handler for the Z180 is very lightweight, and
we don't need the additional overhead of the bottom half
handler.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 624e63afca0a3c3e828d9165abfb3208802c9b96
Author: Lucille Sylvester <lsyl...@quicinc.com>
Date: Wed Apr 28 08:49:03 2010 -0600
msm: kgsl: Corrected open/close times of the 2D GPU
With idle detection in place we can now turn on/off the core at more
appropriate times.
Change-Id: If4bbb7bc27557125be0b5a86c4f26cafe7c0d06e
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
commit 6e937cb0a097c01cf96e268b228aecbe1b1de567
Author: Jon Ashburn <jash...@quicinc.com>
Date: Wed Apr 28 08:49:02 2010 -0600
msm_kgsl: Fix for proper clearing of context mask on destroys
Context ID mask was being cleared incorrectly by improper bit shift.
Change-Id: I64901741a53cee31f034787fc28c8000899ca669
Signed-off-by: Jon Ashburn <jash...@quicinc.com>
commit fbe314bff8de84edff1b5fd6c4bb0bd273d9c863
Author: Lucille Sylvester <lsyl...@quicinc.com>
Date: Wed Apr 28 08:49:01 2010 -0600
msm: kgsl: 2D GPU core idle detection
Set timers after receiving 2D interrupts to check for equal timestamps.
If the core has no more work in its queue, timestamps will be equal and
we can turn off 2D clocks and associated resources. Check that clocks
are on before allowing 2D commands to use the GPU.
Change-Id: Ie13a338cdd7cc652d0495498e6a3a6ed43c3e8bb
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
commit 4d9d474b8987a5988bf19de751582f16a7351a27
Author: Jon Ashburn <jash...@quicinc.com>
Date: Wed Apr 28 08:48:59 2010 -0600
msm_kgsl: Add MMU support for Z180 core
Add MMU support for Z180 core. This includes some restructuring
of existing default pagetable code for 3D core to handle shared pagetable.
CRs-fixed: 231150
Change-Id: Ic76f493ba1cea82520ff1c6524e425ea9670755a
Signed-off-by: Jon Ashburn <jash...@quicinc.com>
commit b3e2a66a7605c7f74f85ea79824206468a209778
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Wed Apr 28 08:48:56 2010 -0600
msm_kgsl: Dump register state on timeout for waittimestamp.
Also, if the user asks for no timeout, force it to be 10 seconds so we
don't end up blocked forever if GPU hangs.
Change-Id: Id3eea15d3fa53b0736d68f90bb2d4ea644fdbc58
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 75a4938bc0535535997eb3bb6e5c67e73e281c07
Author: Lucille Sylvester <lsyl...@quicinc.com>
Date: Tue Apr 27 17:29:53 2010 -0600
msm: kgsl: Vote for minimum AXI when idle
On targets without a 2D core we need to vote for default AXI speed
when the 3D core goes idle.
Change-Id: I417bdb05851179670acd1d1f841f432b9777b08b
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
commit 54df117c42b9cc4e8649192c776a2245f1c1d19d
Author: Jin Hong <ji...@quicinc.com>
Date: Thu Apr 29 13:28:45 2010 -0700
defconfig: sync up with kernel 2.6.31.12
Signed-off-by: Jin Hong <ji...@quicinc.com>
commit ff3ce95bb99d73a5308016984d81bb5a5662e5e9
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Tue Apr 27 17:29:07 2010 -0600
msm_kgsl: Reuse vmalloced memory buffers
Memory buffers that are vmalloced and can be freed are put
into a free list and reused later if an allocation of same size
is required.
Change-Id: Ie65a87051fe231873485580bc30261c6ddd8b5e6
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 35ec90005a530a210e7eb3501d3ac1b5979fd57d
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Tue Apr 27 11:04:38 2010 -0600
msm_kgsl: Optimize TLB flushing
Only flush TLB when it is absolutely nessasary as opposed to
flushing it on every map and unmap
Change-Id: I086a025b80490ebdd37bfbb5e98bba3bcb533d89
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 49b225018bccd15b3d1e0342ab73fb5d3a0866d3
Author: Jon Ashburn <jash...@quicinc.com>
Date: Tue Apr 27 11:02:39 2010 -0600
msm_kgsl: Only enable Z180 device when used
Separate G12 and Yamato Open/Release functions to support different
open/close times. Add checks for G12 HW usage and only open the
device and turn on clocks and IRQs at that time.
CRs-fixed: 228181
Change-Id: I40c52318c0cf5f1a4f87e53efc0fc2b41494e0da
Signed-off-by: Jon Ashburn <jash...@quicinc.com>
commit 17e9bf5a4009a438f0d7463c3b2f73808db804c7
Author: Jon Ashburn <jash...@quicinc.com>
Date: Mon Apr 26 10:51:49 2010 -0600
msm_kgsl: Add support for Open VG and Z180
Change-Id: I98b2e6e4d6714118e79fbc7fb8abc57f3fca44a2
Signed-off-by: Jon Ashburn <jash...@quicinc.com>
commit e9a4909b08990e8d0524f5529f0652de5fb99ed0
Author: Stephen Caudle <sca...@quicinc.com>
Date: Thu Apr 22 16:14:11 2010 -0400
input: qci_kbd: Enable keyboard auto repeat
Change-Id: I545d133acdcd136e46c3130aaa15e32f5a6772ea
Signed-off-by: Stephen Caudle <sca...@quicinc.com>
commit 7fe35d20c66f5914a7db2caae0aadc02d61f7264
Author: Naveen Ramaraj <nram...@quicinc.com>
Date: Thu Dec 3 15:08:04 2009 -0800
pmem: Fix Bitmap allocator to use correct meta data sizes
The PMEM bitmap allocator makes use of bitm_alloc to keep track of memory
allocation requests. When the pmem device memory usage increases, the
number of bitm_allocs should be increased accordingly. The reallocation
should therefore be in multiples of bitm_alloc sizes. Add a sanity check
to make sure the number of bitm_allocs is within the maximum and also,
intialize the quanta to zero.
Change-Id: I06d6e1c3557b41143af92cc0c96ba6bcef4c8cb6
Signed-off-by: Naveen Ramaraj <nram...@quicinc.com>
commit 2be7a4d6a3f8b0f09646182018271b6b3e0b3915
Author: Murali Palnati <paln...@qualcomm.com>
Date: Wed Apr 21 22:45:05 2010 -0700
msm: msm_nand: Use uncached, unbuffered memory for ONFI buffers.
Buffers needed for reading ONFi identifier and the parameter page need
to be present in the coherent memory so that cache handling over head
can be avoided. Moreover cacheable memory is not needed for these ONFi
buffers as they are used only once.
Change-Id: I37b3037d728d296e996d5b2c75a7482e326534f4
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 4c49a4b02e07aec2b70e67008d532be92b66b5ba
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Apr 21 12:54:16 2010 -0600
msm_kgsl: Don't attempt to write to the GPU MMU if it isn't started.
When the KGSL device closes it takes down enough of the MMU structure
to cause a segmentation fault when the GEM objects unmap themselves.
Since the MMU is gone, there isn't any value to be had in trying to
flush anything.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 7e929b4b617004c5956b22be523313553afe1b9b
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Tue Apr 13 17:01:17 2010 -0600
msm: 8x50a: Match remote lock name with modem processor.
For 8x50A, modem side uses DAL remote lock.
Change-Id: I71fbe65ab022e7a819e097a8b679b0cad11739e7
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit f4a68e6faa3fd4c5ebf7f62c9773b7c8a58523ae
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Apr 23 16:41:05 2010 -0600
msm_kgsl: Add config option for using only mapped MMU addresses
Added a config option that controls if the GPU can use
physical addresses or not when MMU is turned on
Change-Id: Iae1ab0bd67b5feda5cfd6a1f2e82bc1f67f596c4
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 8077d17ea3ae77eea840d05e5d2c9bcc4f409c69
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Fri Apr 23 16:41:05 2010 -0600
msm_kgsl: Set up the GPU to fault when non-MMU address is used
Change-Id: Iae7814b3c80cf832380e4eb2755dc5498490ddce
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit e8c835f44410cc010aa49068d3c6067ee7e96c23
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Fri Apr 23 16:41:05 2010 -0600
msm_kgsl: Add writecombine flag to map vmalloced memory
Adding writecombine flag to map vmalloc memory into user space
improves performance
Change-Id: I298602a6004e24efe7e367185e993bdab3ad5251
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 84a4827382419709e48b9b349966eee217810f1d
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Apr 23 16:41:05 2010 -0600
msm_kgsl: Add new support ioctls to kgsl_drm.h
Add the new ioctls to support KGSL GEM operations to kgsl_drm.h.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 29f246874b9a7ceb8fcfc6f959ea561b2e3fd247
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Apr 23 16:41:05 2010 -0600
drm: Add drm_vm_close_locked()
Add drm_vm_close_lock() to allow GEM to free VM
entries while already holding the device drm mutex.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e2dea9a2a6c1a115d90bb20975be5e7cc2ae9910
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Apr 23 16:41:04 2010 -0600
DRM: Use platform appropriate indexes for the GEM fake mmap offset
GEM uses "fake" mmap offsets to map structures. The offset member
of mmap() is an unsigned long, so we need to make sure the offsets
are properly formatted depending on the architectural length of an
unsigned long.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 439f0cd1fca3bea7c3b3b38235040dcfc99bf2cc
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Apr 23 10:11:41 2010 -0600
drm: Add support for platform devices to register as DRM devices
Allow platform devices without PCI resources to be DRM devices.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 69c2d0ead35cb01475c66b295537e433643310b8
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Tue Apr 20 22:43:42 2010 +0530
mmc: core: Add auto suspend functionality
Provide an interface for host drivers to register a callback function that
the core will call if the MMC bus is idle for a certain amount of time.
Host drivers can enter into low power mode upon receiving this callback.
The MMC Core takes care of resuming the host upon getting new requests from
block layer.
A sysfs entry is added for configuring the idle timeout value. If it is
set to a -ve value, core resumes the host and never suspends it again.
A module parameter "idle_timeout" is also added to set the idle timeout
through the kernel command-line. The default value is set to 20 sec.
Signed-off-by: Sahitya Tummala <stum...@qualcomm.com>
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit f87e84a309d795d5e3cbb2fc5c5e44cb02273536
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Tue Apr 20 18:47:55 2010 +0530
mmc: core: Accept EXT_CSD revision upto 1.5
For cards that comply to MMC specification 4.4, the ext_csd_rev upto 1.5 is
acceptable.
Signed-off-by: Sahitya Tummala <stum...@qualcomm.com>
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit 285dbe26fb94b7b8167dbb1fdb61e7024cbcd4a6
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Tue Apr 20 13:51:10 2010 +0530
mmc: Add concept of an 'embedded' SDIO device.
This is required to support chips which use SDIO for signaling/
communication but do not implement the various card enumeration registers
as required for full SD / SDIO cards.
mmc: sdio: Fix bug where we're freeing the CIS tables we never allocated when using EMBEDDED_SDIO
mmc: Add max_blksize to embedded SDIO data
Signed-off-by: San Mehat <s...@google.com>
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit 079530ce75d0230f66378fc5a47c694b4ca2db14
Author: David Brown <dav...@quicinc.com>
Date: Fri Apr 16 11:46:26 2010 -0700
Fixes to enable ehci-msm on .32 kernel.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 07e573f86724d8ec7e41c849d6286eaad237c525
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Thu Apr 15 17:40:28 2010 -0600
msm_kgsl: Change memory allocation size for MMU pagetables
When global pagetable is used with MMU then maximum memory
that can now be allocated is 128MB. With per process
pagetables each process can allocate 32MB of system memory.
Change-Id: Iee686b1310dee40026f3aabaddd51c41023bd096
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 1b74f739e3d62718792278600419fa0bdf21ea83
Author: Dima Zavin <di...@android.com>
Date: Thu Apr 15 17:40:17 2010 -0600
[ARM] msm: kgsl: Fix a typo when checking pmem region size
Change-Id: Icab8179e15569c77762e24764380f8f97d0d43e2
Signed-off-by: Dima Zavin <di...@android.com>
[sa...@quicinc.com: Changed the offset where patch is applied]
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 1652fce64e31a0082f0e750b03d8436b06a3967a
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Wed Apr 14 14:16:13 2010 -0600
msm_kgsl: Add idle packet before flushing TLB via command stream
Adding an idle packet before flushing the GPU TLB via command
stream may prevent a hang.
Change-Id: I04cfd542c1d89963e295b0f15e82b29b459832eb
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit cac12730a563a57be9b1badf83c880116fb2dce0
Author: David Brown <dav...@quicinc.com>
Date: Fri Apr 16 11:39:15 2010 -0700
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
commit 09ca1d925607b8ad7df4df9fe14a25257f5c3ffe
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Thu Mar 11 17:04:06 2010 -0700
i2c_msm: Power management for I2C drivers.
Enabling/disabling clocks can be a remote call and doing that per
transaction may incur penalty for every transaction.
To avoid that, the driver uses activity-timer which disables clocks
upon expiration. Clocks are enabled during first transaction after
timer has expired. Timer is reset after every transaction.
Change-Id: If2f5f39f00ef6ab60ff647fa0755035b43c9c68c
Signed-off-by: Sagar Dharia<sdh...@quicinc.com>
commit 1046d9469c54edca64b36a0f18765563de5bdcb5
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Thu Mar 25 16:47:38 2010 -0600
msm_i2c: Clear interrupt generated by transaction on other processor
I2C hardware broadcasts interrupt to both processors when the HW is
shared. If the I2C HW is shared, the interrupt is acknowledged
(and thus cleared) before beginning the transaction on Linux side to
avoid spurious interrupt handling.
Change-Id: I6148329d2b3ed6ace3350a53b84814e69d3ee938
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 5b943c4e47bece572b8866a9383fb248854a45c9
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Wed Apr 14 14:16:14 2010 -0600
msm_kgsl: Restore the shader when CTXT_FLAGS_SHADER_RESTORE flag is set
Change-Id: I0b05099c73dea130109310e368e9bf4b48a6e3eb
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 13c52f7699945bcfe1cb4f584829b4338bd7b3a1
Author: Dima Zavin <di...@android.com>
Date: Wed Apr 14 14:16:14 2010 -0600
[ARM] msm: kgsl: allow the framebuffer to be mapped into the mmu
The from_pmem ioctl can now accept an fd for the framebuffer, and
create an mmu mapping so that the GPU can render straight into
the fb.
Change-Id: I8845eb9ee3ce793349e71785311655c8b096e6eb
Signed-off-by: Dima Zavin <di...@android.com>
[sa...@quicinc.com: Patch applied to a different driver version]
Signed-off-by: Shubhraprakash Das <sa...@android.com>
commit 07e24bfd3a24b8ff0fa5788ff94d17972b3b34f1
Author: Jin Hong <ji...@quicinc.com>
Date: Wed Apr 14 11:41:20 2010 -0700
defconfig: enable INPUT_MOUSEDEV for ST1.5
With CONFIG_INPUT_MOUSEDEV trackpad is working without any change to /dev
Change-Id: I98c9c3369189c77fd9574f1c30f763bddcb5d640
Signed-off-by: Jin Hong <ji...@quicinc.com>
commit f9787ada9a0c7bd1e3c404179bd50f44f0e2188f
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Tue Apr 13 13:33:24 2010 -0600
msm_kgsl: Adding feature per fd pagetables for MMU
This change adds a feature to use per file descriptor
pagetables with the GPU's MMU unit.
Change-Id: I6598dfb31086a8aec3e8c9496e33ead288642bab
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 618d84b07daa6c98143492d8b28cb1c81d518c8f
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Wed Apr 14 17:57:46 2010 +0530
defconfig: Enable usb selective suspend
Change-Id: Icb99840639a235d13c6ede49406a22e712d378c0
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 8a17adcd881dc8720ce80cd34e8cb651a812d2b9
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue Apr 13 14:49:04 2010 -0400
Revert "msm_fb: initialize HDMI chip during panel on processing on ST1.5"
This reverts commit 6fbdf927c9cd6fe4e0530b8881200d1b94107ce5.
The commit being reverted relied on the HDMI chip working. This is not
the case on ST1.5 rev A boards. On these boards neither the HDMI nor the
on-board display will work.
Change-Id: I3bfaadd1db0901d8643032a146163bd32d56e60c
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit 37d005ab786a07b10235f125ee22c1033aca858a
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Apr 8 17:49:59 2010 -0400
msm: for ST1.5, add settle time after HDMI power off
During mode switch, the video driver calls the panel off/on
functions in rapid succession, which results in power to the
HDMI chip being removed then reapplied. It needs a brief delay
to stabilize between these events.
Change-Id: Iad94c928e6d64a9e82f95c048f4519f30f412a0a
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit ccd7ef4623f6095da7bb45b15276e10e3c927254
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Fri Apr 2 10:41:37 2010 -0600
msm: 8x50a: Register HDMI with I2C QUP bus.
HDMI is connected on I2C QUP bus and not Legacy I2C bus.
Change-Id: I4ff13fd5b3502deadd3097fc9f051e4500fcb162
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 94b872a01d9abb9801d0adb15923b7742d109221
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Thu Apr 8 15:13:52 2010 -0600
qup_i2c: Incorporate better RESET and interrupt clearing logic
Doing a SW_RESET takes 1 QUP HW logical cycle so we poll to ensure
that RESET has happened before writing other registers.
Level triggered interrupts are cleared by putting mini-core in RESET
Change-Id: I301725fd33016e1e497147d870b8adfd80f6388c
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 4c717bd578bc2178565bb64f5135e8b3359619dd
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Wed Mar 31 15:21:02 2010 +0530
msm: qsd8x50: Add Support for MMC Hardware based detection for ST1_5.
Configure the MSM GPIO 100 for SDCC Detection on qsd8x50 for slot1 and
request for a MSM GPIO IRQ to handle the interrupt.
Change-Id: I5df66530a5bdb35efc26a5dfe57803bef73bd8c2
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit 612eeb34e833931333e7b2af9cc904242e2b34e7
Author: Neil Leeder <nle...@quicinc.com>
Date: Mon Apr 5 12:49:02 2010 -0400
msm: blank ST1.5 screen correctly
Blank the screen correctly - to black - instead of leaving it white.
Change-Id: I3cbef3308dda1f456bfbcf0d682dfcf2512e73a6
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit dd95043e91c824da389b0b279447636724859435
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Mar 24 16:37:53 2010 -0600
msm: 8x50a: Enable QUP I2C on 8x50a board
Change-Id: Iafcf9db96fec08124324f7961488d2aff45bbb17
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 89d3be7108581b3e3621ffc899334cfc373157f9
Author: Krishna Konda <kko...@quicinc.com>
Date: Sun Apr 11 23:04:45 2010 -0700
defconfig: ST1.5 perf defconfig
The perf defconfig had a few debug options turned ON which
were turned OFF in this patch along with enabling oprofile
config options.
Change-Id: I84b276a67c24e45da16b853f799079948b669d40
Signed-off-by: Krishna Konda <kko...@quicinc.com>
commit 8e2e964fe5c477c38face5c86e07b94e81f3dabe
Author: Jason Wessel <jason....@windriver.com>
Date: Mon Oct 13 08:33:42 2008 -0500
NET_POLL: Pass skb via NET_POLL rx routine
This patch is a change to the NET POLL api so that the MAC address or
IP frame information can be parsed by the driver that wishes to use
it. Speficially for kgdboe, it is used to automatically collect the
MAC address of the GDB host.
Origin -
http://git.kernel.org/?p=linux/kernel/git/jwessel/linux-2.6-kgdb.git;a=commit;h=8e0075215391f5ef3203b3333d9e018001a8021b
Signed-off-by: Jason Wessel <jason....@windriver.com>
Acked-by: Matt Mackall <m...@selenic.com>
Signed-off-by: Shubha Nagaraja Rao <shu...@quicinc.com>
commit e7cb582eb19e645d82df045f4a307a959481226a
Author: Jason Wessel <jason....@windriver.com>
Date: Mon Oct 13 08:33:42 2008 -0500
kgdb: Add the ability to schedule a breakpoint via a tasklet
Some kgdb I/O modules require the ability to create a breakpoint tasklet.
Origin -
http://git.kernel.org/?p=linux/kernel/git/jwessel/linux-2.6-kgdb.git;a=commit;h=aaee3b9c384febeccf4dc66dd6c1ca04101876ad
Signed-off-by: Jason Wessel <jason....@windriver.com>
Signed-off-by: Shubha Nagaraja Rao <shu...@quicinc.com>
commit 33078ccc5e57f09901bbe0643e6c2d521ef790a5
Author: Jason Wessel <jason....@windriver.com>
Date: Mon Oct 13 08:33:42 2008 -0500
kgdb: Implement kgdb over ethernet with NET_POLL
At one point this was very similar to the version Matt Mackall wrote
to allow for the use of KGDB over ethernet. Since then it has been
reworked to fit into the unified KGDB model.
This kgdb I/O driver can work as a kernel built-in or a kernel
module.
Origin -
http://git.kernel.org/?p=linux/kernel/git/jwessel/linux-2.6-kgdb.git;a=commit;h=8e42fbe863d4800492be01be28a2c3f94df81aeb
Signed-off-by: Tom Rini <tr...@kernel.crashing.org>
Signed-off-by: Jason Wessel <jason....@windriver.com>
Acked-by: Matt Mackall <m...@selenic.com>
[shu...@quicinc.com: Resolved cherry-pick conflicts]
Signed-off-by: Shubha Nagaraja Rao <shu...@quicinc.com>
commit 2a16a4daec5e2cc087acb23b2e83e2d0dff00d09
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Tue Apr 6 10:28:02 2010 -0600
msm_kgsl: Enable 3D clocks before power for all Adreno platforms
7X30 and 8650A require that 3D clocks be turned on before the
power rails. 8X50 and 7X27 work fine with the same sequence, so
do away with the #ifdefs.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit e165285e73ef02c942ca251fea973b8521100e1b
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Apr 8 17:50:56 2010 -0400
msm_fb: initialize HDMI chip during panel on processing on ST1.5
During power saving HDMI chip is turned off, so it needs to
be re-initialized during panel on processing. Add platform data
that adds a callback to re-do the i2c init sequence.
Change-Id: I78a35b03e50651c2dc37b3b7ad95b829929d3228
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit b3b5590d4ec16a6dad0e32487e486a09d90707b9
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Apr 8 15:33:53 2010 -0400
msm_fb: add sii9022 HDMI support to st1.5 driver
On ST1.5 support for HDMI and the on-board display is
provided through the same LCDC bus. Chrome only uses one device
and will drive both panels through it. This provides support
for both panels in one driver.
These changes are largely taken from hdmi_sii9022.c.
Change-Id: I4560df47edb3a45d42e8add01eae96c74be18b7c
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit f43a95fba593699aec5e0010ca846913e3072b51
Author: Stephen Boyd <sb...@quicinc.com>
Date: Wed Apr 7 17:11:19 2010 -0600
msm: internal_power_rail: Add GRP_2D, ROTATOR, and VPE
Add some more rail ids so we can use them in the future.
Change-Id: Ice4b641e948407a2e022ef3b31ec7b76b634f88e
Signed-off-by: Stephen Boyd <sb...@quicinc.com>
commit 70fbad6037eb6ce27ad9560d3015cea13082bae5
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Tue Apr 6 10:28:15 2010 -0600
msm: kgsl: Add barriers and flush the tlb when changing MMU mappings
Memory barriers are required whenever we map/unmap entries from
pagetables because the pagetables are allocated from normal
noncached memory pool.
The tlb of the GPU needs to be flushed when entries from its
pagetable are unmapped as we do not want old/invalid entries to remain
in the tlb.
Change-Id: I2bfb9bc68215c5bf1395a2a4c45a31fe279e0d10
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
commit 4e6fc35530504fadd33631df0b3e165fcbfc3172
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Fri Apr 2 10:42:05 2010 -0600
qup_i2c: Enable QUP I2C mini core driver for QSD8x50A
Change-Id: I376b787cfcd449236b33ffec5c4346bf86387881
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 3254b5a47eacf65ee9423159908c55d9ad6670bc
Author: David Brown <dav...@quicinc.com>
Date: Wed Apr 14 13:48:49 2010 -0700
Update ST1.5 defconfig to disable features not yet working.
Disabling: kgsl, msm_serial_hs, and ehci.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7afec0748b54a75f5b81629ac36aae6f39186c38
Author: David Brown <dav...@quicinc.com>
Date: Wed Apr 14 13:31:29 2010 -0700
Cleanup merge errors.
These files have merge errors.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9b3fdd1e5f93df350241afa1a78331fc8a8d6779
Author: David Brown <dav...@quicinc.com>
Date: Wed Apr 14 13:30:32 2010 -0700
msm: timer: Fix timer calls for .32 kernel.
The kernel cyc2ns api has changed. Instead of passing in a
clocksource directly to it, it takes two arguments of two of the
fields.
Signed-off-by: David Brown <dav...@quicinc.com>
commit a42aa6fc48b054d583b041c4a6126ba406d2f32c
Author: David Brown <dav...@quicinc.com>
Date: Wed Apr 14 13:27:16 2010 -0700
msm: Remove references to early suspend.
Remove unneeded references to early/late suspend, which works
differently in the .32 kernel series.
Signed-off-by: David Brown <dav...@quicinc.com>
commit a26cdc33f36bdab851e9a1795bd142c5b42c5f6d
Author: Hanyu Cui <han...@quicinc.com>
Date: Mon Apr 5 12:52:04 2010 -0400
msm_fb: Correct image rectangle calculation for dma cache operations
The original calculation is incorrect and could result in cache
operations performed on incorrect memory regions.
Signed-off-by: Praveen Arkeri <par...@quicinc.com>
Signed-off-by: Hanyu Cui <han...@quicinc.com>
commit 3f10bb7ac6f0eab23fd2151d749c255a356ab17e
Author: Rajesh Sastrula <vra...@quicinc.com>
Date: Fri Feb 5 18:17:16 2010 -0800
msm: qdsp6: Move to the old Vdec Interface Version 2.0
- Q6 supports the backward compatible interface,
hence moving to the old scorpion/qdsp6 interface (2.0).
- Deprecating the intermediate driver versions.
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit e38303cfb6d673c8e25544157a0c2bfd5595fa89
Author: Susan Brownhill <sbro...@quicinc.com>
Date: Thu Feb 11 10:17:28 2010 -0800
msm_q6venc: Update to the latest driver.
- Add encoder interface version check
- clean up driver
- add stop in release function
- fput to fix pmem leak issue
- fix to handle user application crash
- acquire wakelock during video encoding
- add an ioctl to get driver version
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit f9ba9b557df50290647094da3f56d996f8547d68
Author: Dinesh Garg <c_d...@quicinc.com>
Date: Thu Mar 4 14:39:23 2010 -0800
msm: 8x50a: Replace gpu0 region with smipool and remove unused gpu1 region.
Replace gpu0 pmem region with smipool region.
3D Graphics has transitioned to using system memory instead of pmem,
hence remove the gpu1 pmem region that is not required.
Change-Id: Ief3c5730211f0255e9f18af6bb17f43ea7ac38a0
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit a0e7930abcbf9824e2fe3b2b06d54a308b298c13
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Apr 5 11:25:06 2010 -0700
msm: acpuclock-8x50a: Set safe ACPU voltages for all frequencies.
Until further characterization data is available, specify ACPU
voltages that are sufficiently high for stability, yet safe for
the transistor technology.
Also, ensure the initial ACPU voltage matches the table at boot.
Change-Id: I219e3d11bf9e36cc1c652bf4bc5b84fb17cf08eb
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit 11333fe3c1d063a4c009ed6a1c1fcbb86fe9b23d
Author: Jin Hong <ji...@quicinc.com>
Date: Wed Mar 31 17:12:15 2010 -0700
defconfig: enable CONFIG_EFI_PARTITION for ST 1.x boards
enable EFI_PARTITION to support rootfs partitioned using EFI GPT
Change-Id: Ibdeb7d4e6bedeca16cf24063358f9188e4dbc407
Signed-off-by: Jin Hong <ji...@quicinc.com>
commit e35863141f66e728d29be15d0fea36d0aa04a590
Author: Jin Hong <ji...@quicinc.com>
Date: Tue Mar 30 18:07:38 2010 -0700
defconfig: build Quanta keyboard by default for ST 1.x boards
Change-Id: If8cef1e4adc6aca5ae6e4c2dd9796ed59e2f42f4
Signed-off-by: Jin Hong <ji...@quicinc.com>
commit da35a3ad08cf21635c15c1d80a6e0539b76a6dcc
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Wed Mar 31 11:53:48 2010 +0530
msm: Correct dma memory allocate and free size in probe and remove methods
Correct mismatch in allocated and freed dma coherent memory size in probe
and remove methods of driver.
Change-Id: Ibe9fc55f3f0f5ddb3576b9ae70ddd801dbd31647
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit b45d087b8c20bb854f6da7b4f2bc9a558fbceaa8
Author: Mandeep Singh Baines <m...@chromium.org>
Date: Mon Mar 29 15:12:14 2010 -0700
MSM: add keyboard driver in st1.5
BUG=NONE
TEST=Built and ran on st1.5
Signed-off-by: Horace Fu <hora...@quantatw.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/723001
commit 86becb6b78a28624caacad7076f551bdb6a6fd34
Author: Trevor Bourget <tbou...@quicinc.com>
Date: Sat Mar 27 18:11:36 2010 -0700
defconfig: use Quanta trackpad by default for st1q variants
Change-Id: I8df57817f8257031813d948c3a8a8ca900121fd0
Signed-off-by: Trevor Bourget <tbou...@quicinc.com>
commit f4fb8a8d1a99fa21872690a42e27f3d1a9347ae9
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Thu Mar 25 18:24:53 2010 -0600
qup_i2c: Support reads, writes greater than FIFO-size number of bytes
If read-transfer size is greater than in input-fifo, block mode is
used to read the data. For writes greater than output FIFO size, we
ensure that the output FIFO isn't overrun by using output block mode.
Change-Id: Ia4b99b2798d645e2d10163f43448737d509da5d9
CRs-fixed: 229442
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit d3eccfd17c2c7f6f2e75c40f2854233cbc1305d8
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Feb 24 15:46:58 2010 -0700
qup_i2c: Make driver compatible with new QUP I2C core.
8650a and upcoming chipsets have new QUP I2C core which has 1
interrupt instead of 3. The interrupt is level triggered.
We rely on boards file and interrupt resource to determine how many
interrupts we have. We rely on QUP registers rather than interrupt
number in ISR to figure out interrupt reasons.
The driver changes are backwards compatible and work on chipsets with
older QUP core. (e.g. 7x30)
Change-Id: I9316b544139433cf6f6dc8a2fb1991552baf9f9e
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 861aa81119d5180af1a66932dd9018dc0d33665e
Author: Mandeep Singh Baines <m...@chromium.org>
Date: Wed Mar 24 21:04:32 2010 -0700
MSM: add trackpad driver in st1.5
BUG=NONE
TEST=Built and ran on st1.5
Signed-off-by: Horace Fu <hora...@quantatw.com>
Signed-off-by: Mandeep Singh Baines <m...@chromium.org>
Review URL: http://codereview.chromium.org/724001
commit 4cb3a6654362a55f6afc810728291690273d1487
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Tue Mar 23 17:22:09 2010 -0700
8x50: audio: Enable speaker and mic on ST1.5.
Default to a valid device path so that audio can be rendered.
Do not reject the modem based on ACDB or ADIE version since
the final modem ADIE and ACDB version is not determined. For
now it must also be backward compatible.
Change-Id: I0603c025d0b36d5fce9fe6b5b2d9f91b6c8ecf8d
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 5cee5238aed97471b8342e2859ff579b77d21b23
Author: David Brown <dav...@quicinc.com>
Date: Thu Mar 25 12:25:14 2010 -0700
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
commit c408c5c75caf239d0aaeb105d9683627b8e1085c
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Tue Feb 9 15:34:01 2010 -0700
qup_i2c: Use platform specific PCLK, I2C clock name.
QUP hardware block on some boards need PCLK as well. QUP driver will
use PCLK if it's specified by the platform data structure in the
boards file. I2C clock name is added to platform data structure since
targets with multiple I2C controllers may require controller specific
clock names.
Change-Id: I8a9fb37b684197da71dd8777178c3b555ec67ef7
CRs-fixed: 224242
Signed-off-by: Kenneth Heitke <khe...@quicinc.com>
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit c7ca0804db5de2346cc2d53ab53d0cbe70f44ecd
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Tue Feb 9 15:34:01 2010 -0700
qup_i2c: Use platform specific PCLK, I2C clock name.
QUP hardware block on some boards need PCLK as well. QUP driver will
use PCLK if it's specified by the platform data structure in the
boards file. I2C clock name is added to platform data structure since
targets with multiple I2C controllers may require controller specific
clock names.
Change-Id: I8a9fb37b684197da71dd8777178c3b555ec67ef7
CRs-fixed: 224242
Signed-off-by: Kenneth Heitke <khe...@quicinc.com>
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 84219be3377fea525b7411b8efe27188a610d51f
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Feb 10 14:46:59 2010 -0700
qup_i2c: Use DEBUG correctly as a flag and not as a value.
The DEBUG flag was being used inconsistently as value and flag,
which was causing printing of messages when not desired. This commit
takes care of that inconsistency.
Correct usage of DEBUG as flag also cleans up the code and eliminates
computation that was being done when debugging was disabled.
Change-Id: I7542b39d07fc562f375d1f977ded938bfe6fbfd4
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 5b48da7bde164d26c55ce0b290e08b99c2aac41f
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Tue Mar 23 17:18:53 2010 -0700
alsa: soc: 8k: Initialize the audio device upon ALSA open.
Change-Id: I00ce3b9da100ac464063e97a575e1e5d92875484
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 9ce25a74fc8df1c5917da93b68d76cdd209b66f5
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue Mar 23 16:36:12 2010 -0400
msm: add GPIO IRQ number for ST1, ST1.5 k/b, touchpad
Change-Id: I87c6da3e0a24837bd99c5c22df00e92113bd0d6a
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit f079ddfe61f41833938b9420e5818cd102dd5172
Author: Anshul Gupta <ans...@qualcomm.com>
Date: Tue Mar 23 16:04:56 2010 +0530
Kbuild: export msm headers to userspace
This provides the needed msm headers by userspace multi-media
Signed-off-by: Anshul Gupta <ans...@qualcomm.com>
commit 60b3621512eca84d2faf4d05e74b4c3e66451721
Author: Trevor Bourget <tbou...@quicinc.com>
Date: Mon Mar 22 15:54:23 2010 -0700
defconfig: enable SECURE_FILE_CAPABILITIES for st1 and st15
enable SECURITY_FILE_CAPABILITIES for use by chromium-os minijail process
Change-Id: I7f107261ae3411b881dbed59a8d600b417c8cc9b
Signed-off-by: Trevor Bourget <tbou...@quicinc.com>
commit 9d173edd95c6cc29fafa114c1858d253af39942c
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Tue Feb 2 12:20:15 2010 -0700
qup_i2c: Move QUP register initialization to first transfer
QUP registers are also clocked by QUP client clock. The register
reads may fail if the client clock is not on during QUP probe.
Moving the register reads to first transfer ensures that the client
has called QUP transfer after it has enabled QUP client clock and
the register reads won't fail.
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit ce0eb9427af9080de30c3143f1881cbeba64dcd0
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Fri Dec 4 11:24:48 2009 -0700
qup_i2c: Disable QUP interrupts when QUP is not in use.
QUP (and its interrupts) is only used when a client initiates an I2C
transaction. We disable QUP interrupts when QUP is not in use
so that idle power collapse can be done.
Change-Id: Ieffe892ad3fd1c55efc655913e94037a13023211
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 03e3560ef4eb1be5bb3e9cdf54884100acfdd894
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Sep 30 17:27:06 2009 -0600
qup_i2c: Make sure that READ start and READ size tags are in same FIFO
There is a known hardware restriction with I2C mini core, where for a
READ command, start and size tags have to be specified in same FIFO.
Failing to do so may cause QUP to generate extra read cycle.
To avoid this, 1 NOP is introduced at the end of WRITE preceding READ,
if those WRITE command(s) leave only 1 space in the FIFO.
Change-Id: I5a715f34b35edef97ec2867b3bfcbc14674dea5d
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit baa36d0dd9b135bc3197a0e3d7acba4b67671a04
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Mon Mar 22 18:42:19 2010 +0530
usb: gadget: diag: Fix to send correct pid and serial number
pid and serial number are taken from Device descriptor, which is
not intialized before diag bind gets called, but properly updated
in composite driver structure.
Take pid and serial number from composite driver structure rather than
Device descriptor.
Change-Id: I736e3947d549ceca15f02eb389509e2e3cbb7d67
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
Signed-off-by: Vamsi Krishna <vskr...@qualcomm.com>
commit b8f1e617ff8d567a4225d46dcf21ed7d4cfc3ddb
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Thu Mar 11 10:38:05 2010 -0800
8x50: audio: Enable ST1.5 speaker phone.
Change-Id: I0ee7b9ab19d4102b311abd0718668a86401e8958
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 0c0a695c655f09bcfbde73b57ae254682fc09ff6
Author: Shalabh Jain <shal...@quicinc.com>
Date: Thu Mar 18 19:03:02 2010 -0700
diag: Make DIAG char driver static
DIAG driver needs to be static to be loaded automatically. The config
file made it dynamic and hence QPST/QXDM were not working.
Change-Id: If0be223ca419a3c5ccea89ebb92fc554a1417237
Signed-off-by: Shalabh Jain <shal...@quicinc.com>
commit 647b0202ee9e6665d2c31579ee6d5540f290d387
Author: Philip Elcan <pel...@quicinc.com>
Date: Tue Feb 23 09:11:57 2010 -0500
tpm: Add support for ST ST19NP18 I2C TPM device
Add support for the STMicroelectronics I2C Trusted Platform
Module (TPM).
Signed-off-by: Philip Elcan <pel...@quicinc.com>
commit 2a0269c863481fbe62ffdb8228b2015e84c4a491
Author: Philip Elcan <pel...@quicinc.com>
Date: Tue Feb 23 09:36:06 2010 -0500
msm: Add ST1-specific changes for ST TPM device
Create platform data definition and board-file definitions specific
for support of the tpm_st_i2c device on ST1.
Signed-off-by: Philip Elcan <pel...@quicinc.com>
commit c5ba313375b8dd4ede419213638d266a5321a1d4
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Thu Mar 18 18:11:00 2010 -0600
headers: Add msm_kgsl.h to the list of headers to be installed
Add msm_kgsl.h to the list to be copied with the other headers
for use by userspace applications.
Change-Id: Ica055b20d64c66827491bfdbda9a1cf6d708685e
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 943ed24e8618f32e8f801f63dc16cc511b25bbf6
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Wed Mar 17 20:01:23 2010 -0700
defconfig: Enable ALSA SOC for ST1.5.
Change-Id: I5b85ea5f1102264352cfff96168392a95eada4a9
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 8228be8a3d966258286967dd834aef4ec96c51b6
Author: David Brown <dav...@quicinc.com>
Date: Thu Mar 18 08:46:48 2010 -0700
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1d4eff6acfde482763d43d5f2e2c4ad7d83b7653
Author: Ashwin Chaugule <ash...@quicinc.com>
Date: Thu Feb 25 16:37:47 2010 -0500
PMIC: Add driver for TPS65023
This driver scales the Scorpion voltage according to the acputable
frequencies.
Change-Id: Id8b2e8cbd231246143045c7650721f3ef0b63501
Signed-off-by: Ashwin Chaugule <ash...@quicinc.com>
commit f607dcff6f4f5334e1eebbeca0feea554e9ef12e
Author: Ashish Bijlani <abij...@quicinc.com>
Date: Tue Mar 16 19:37:32 2010 -0400
msm_kgsl: Add support for caching GEM objects
This patch provides necessary support to cache GEM objects.
GEM objects allocated both from PMEM EBI and KMEM areas are
by default WC
Change-Id: Idade39f8877f6c592fb34b896225a4b30ee5ff42
Signed-off-by: Ashish Bijlani <abij...@quicinc.com>
commit c4d1e958ab817f0b932c0fa1b68e3431ebcb8056
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Jan 27 13:50:15 2010 -0800
msm: acpuclock-8x50: Use PLL0 instead of AXI under all scenarios.
Some boards may experience stability issues when jumping from an ACPU
frequency of 128MHz to 998MHz. Hence, use PLL0 (246MHz) instead of
AXI (128MHz) while reconfiguring PLL3, or entering power collapse or
SWFI.
Change-Id: I9cc0e567249486cd19d7d346cac2b0b4f0bbef22
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit d9b0be87d7123148a09853a0f28c67703f9c55f6
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Jan 13 12:57:42 2010 -0800
msm: acpuclock-8x50: Select ACPU VDD based on max freq EFUSE data.
The ACPU voltages are updated to improve stability. Since the optimum
voltages differs with the maximum ACPU frequency, seperate ACPU
frequency tables are introduced for 768 MHz and 998 MHz SoCs. The
correct table is selected from EFUSE data. If no EFUSE data is
available, the 998 MHz table is selected.
Change-Id: Id6f8ea0d513042c339653a946b99cd19e8ee93bc
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: Bryan Huntsman <bry...@codeaurora.org>
commit 8a8a0487d603fb5f36c0781c001a9cd3371e5d50
Author: Matthew Wilson <mtwi...@quicinc.com>
Date: Wed Mar 10 10:24:39 2010 -0600
serial: export interfaces for sleep
Bluetooth sleep driver needs this interface
to control UART flow.
Change-Id: I9f51a34759e28b36609afa778ef9ef6c4636f949
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit d787202431376c4877bf1198ab8986cfadc5e465
Author: Matthew Wilson <mtwi...@quicinc.com>
Date: Wed Mar 10 10:16:24 2010 -0600
Bluetooth: export interfaces for sleep
Bluetooth sleep driver needs these interfaces
to determine idle state.
Change-Id: Ib8980ff9e91e26159a2e5c21c33f70707196a1d4
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit 248a53b4b5bb65c28d6e1f5fc5448f54ea363d64
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu Mar 11 12:33:30 2010 +0530
defconfig: Enable USB DCD driver
Change-Id: If3ae7255ba7c6b1831f1b624c61d2410bc59c494
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit fa8224b7a927a4408ceece57a29dea9531413121
Author: Vamsi Krishna <vskr...@qualcomm.com>
Date: Wed Mar 3 19:40:05 2010 +0530
usb: msm_gadget: Interface support with otg
Minimal changes to support interface support to existing otg
framework
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 12ab2a0f7eaef9a8a04ec32b4c6d9b312935e74d
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Mar 11 18:36:42 2010 -0800
msm: clock: Add GRP_PCLK to 8x50A.
Change-Id: I9ee26407e8fecc6ebfd63b961421d2e7c68c2816
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
commit 932f0ee6267f6736657ecc50ac081f2be921e93a
Author: Hanyu Cui <han...@quicinc.com>
Date: Thu Mar 11 17:22:43 2010 -0500
msm_fb: Fix error handling code for asynchronous hardware blit
Resources are released and states are properly restored when errors are
detected and the blit needs to exit early. It fixes these bugs below:
Bug 1: does not call up() when msmfb_async_blit() returns error, which
results in fb lock-up for future fb accesses. Fixed by removing the
unnecessary early return.
Bug 2: curr_djob is not set to zero when returning early due to error,
which causes a false BUG_ON(). Neither are resources released properly,
which could result in memory leak and other potential problems. Fixed by
rewriting the error handling code.
Signed-off-by: Hanyu Cui <han...@quicinc.com>
commit 16890e83b40f48fa43eccd131bb2f43589be2fe2
Author: Mike Lockwood <lock...@android.com>
Date: Wed Feb 17 16:45:04 2010 -0500
msm: usb: gadget: Fix reference after free bug in handle_endpoint()
Change-Id: I5ca4121f43d002719c16fa8c0bcbdd67695bb189
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 0215cdeec2830f030d1af6bde7dc6073f9fdb7e9
Author: Mike Lockwood <lock...@android.com>
Date: Mon Feb 8 05:35:19 2010 -0500
msm: usb: gadget: Set max packet size to 64 for non-bulk endpoints.
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 3306a5a6e7974330602216e33b773666b9140de9
Author: Vincent Kao <vince...@htc.com>
Date: Thu Feb 4 20:08:30 2010 -0800
msm: usb: reset the PHY before suspend
Signed-off-by: Iliyan Malchev <mal...@google.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 2d0b332b14fc70751d6584df4e8c036fba153af9
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu Mar 11 11:10:01 2010 +0530
msm: usb: gadget: misc
[ARM] msm: usb: gadget: Reset the controller before and after resetting
the PHY
Change-Id: If0a415b4727608a78d7c667465cfa4aa0bf424e8
Signed-off-by: Mike Lockwood <lock...@android.com>
[ARM] msm: usb: gadget: use dev_set_name()
Replacing strcpy() to device->bus_id which no longer exists in 2.6.32
Change-Id: I85d14a5569d8a70107e87bc275625cff11b836f6
Signed-off-by: Mike Lockwood <lock...@android.com>
[ARM] msm: usb: gadget: Integrate recent changes from msm_hsusb function
driver
Change-Id: I58d7cd8ea6622aef2e7d4fd0d2e47c08592aca49
Signed-off-by: Mike Lockwood <lock...@android.com>
[ARM] msm: usb: gadget: Add support for interrupt and isochronous
endpoints
Change-Id: Ica1c8e51883dff8d55156d79f831053807a2034e
Signed-off-by: Mike Lockwood <lock...@android.com>
[ARM] msm: usb: gadget: Fix receiving ep0 out requests with length > 0
Change-Id: I4a476fcf326bbf0ea4bbafa8cbb703af8674925c
Signed-off-by: Mike Lockwood <lock...@android.com>
[ARM] msm: usb: gadget: Another fix for ep0 requests with length > 0
Signed-off-by: Mike Lockwood <lock...@android.com>
[cvel...@qualcomm.com: Resolved merge conflicts]
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit e4f8cef8e087bb26f8b4db2658813f78f205f19c
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Sep 23 13:27:27 2009 -0600
qup_i2c: Initial implementation of I2C mini-core driver for QUP
QUP (Qualcomm Universal Peripheral engine) hardware provides FIFO
based data path to mini cores like I2C. i2c-qup driver communicates
with the QUP and its FIFOs.
Advantages of this driver include 1-interrupt per FIFO/BLOCK number
of byte unlike 1-interrupt-per-byte of i2c-msm.
Since the QUP hardware communicates with I2C mini core, and this
driver communicates with QUP hardware, driver communicating with
I2C hardware directly (like i2c-msm) can't be used to communicate
with QUP. QUP has FIFO mode and BLOCK mode. FIFO mode can be used
if the data transfer size is less than FIFO size. BLOCK mode should
be used for transfers greater than FIFO size.
FIFO mode is supported in this initial implementation.
Change-Id: I13360d8c3ebb2bc60f59705fb43d712094990def
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
commit 57cd5714640683431dcdc88924e5c7eeb857ccfd
Author: Kenneth Waters <kwa...@chromium.org>
Date: Wed Mar 10 11:07:09 2010 -0800
msm_kgsl: Implement the old version of the KGSL_SHAREDMEM_FROM_PMEM ioctl.
BUG=NONE
TEST=Built and run on st1q
Review URL: http://codereview.chromium.org/668184
commit bc2851bf289da1947e5b7ed2e14529bfeece5be6
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Wed Mar 10 14:00:25 2010 +0530
msm: ONFI probe fix for MT29F4G08ABC device.
Nand device on ST 1.x hardware is 8 bit ONFi device
but it is advertising itself as a 16 bit device via
ONFi parameter page. This device is identified as the
one from the initial ONFi devices supplied by the vendor
and is known to be deviating from the ONFi Specification.
However since these devices already got into ST1.x boards
we need to have this hack to enable our software to work
with such devices.
Change-Id: Ie12955e218ac2e0e3ea617567edadeebacbf8446
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit 69af82c8d078baf7571ce3c2459e558f22898bc4
Author: Rajesh Sastrula <vra...@quicinc.com>
Date: Fri Feb 19 11:09:02 2010 -0800
msm: qdsp6: Fix pmem leak when application crash
When application crashes vdec_release is not calling put_pmem_file on
the pmem file descriptors that are used before with get_pmem_file call
that results in pmem leak.
Signed-off-by: Rajesh Sastrula <vra...@quicinc.com>
commit e1f2512172573cb79e107f808a0c2de3fd2fb380
Author: stevel <ste...@sklembedded.com>
Date: Tue Jun 16 15:27:45 2009 -0400
msm: usb: gadget: Configure endpoints properly in msm72k_enable
Fixes a problem with full-speed USB support.
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 95ad01760fc6451a2a97d4cdf9c7ed39f503b833
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Wed Mar 3 19:33:15 2010 +0530
msm: usb: gadget: Move udc header file to platform folder
As the same header file will be used by OTG, Host and gadget device
Move header file to a shared location(platform-folder)
Patch provided by Qualcomm
Signed-off-by: Mike Lockwood <lock...@android.com>
[cvel...@qualcomm.com: Removed file renaming change from original patch]
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit f3ebd8927be49215143c1b865da1f15a88cc0fac
Author: Mike Lockwood <lock...@android.com>
Date: Sat Mar 14 15:04:11 2009 +0530
msm: usb: gadget: Disable pullup when vbus is off
Patch provided by Qualcomm
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 68079c2dafbf7544504a6abc7124a01aef531aae
Author: Mike Lockwood <lock...@android.com>
Date: Sat Mar 14 15:04:12 2009 +0530
msm: usb: gadget: Call _pullup function to enable pullup after reset
Pullup needs to enabled only when vbus is valid and gadget driver
is registered. Reset function was enabling pullup directly w/o any
checks. Call enable pullup function which performs the required
checks before enabling the pullup.
Patch provided by Qualcomm
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit f4ff56193f662fe32b6edadeeb23c531a635bebd
Author: Mike Lockwood <lock...@android.com>
Date: Sat Mar 14 15:04:10 2009 +0530
msm: usb: gadget: Add check to enable pullup if gadget and VBUS are valid
Pullup on D+ should be enabled when both vbus and gadget are valid.
Patch provided by Qualcomm
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 3d62e4fe248d633e4188550f770bdf26bea4b932
Author: Mike Lockwood <lock...@android.com>
Date: Sat Mar 14 15:04:09 2009 +0530
msm: usb: gadget: Save and restore otgsc register at h/w reset
As otgsc register is used by transceiver to read the ID pin status
and for interrupts like ID pin change and vbus(B session valid) changes.
The state of this register is cleared once the link is reset and has to
be restored to the previous state the transceiver is aware of. Hence
save the state before the reset and restore the same after reset.
Patch provided by Qualcomm
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit bc765185996d535146bcdbf5a3b362415760dd98
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Tue Feb 10 21:17:07 2009 +0530
msm: usb: gadget: Add support to initiate Remote Wakeup through Sysfs
Add support to initiate Remote Wakeup through sysfs interface. Writing data
to this file results in Remote Wakeup
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit fc6354be90cf5ea6afad21fcd34168435d79e6bc
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:41 2009 +0530
msm: usb: gadget: support remote wakeup GET/SET/CLEAR features
Support GET/SET/CLEAR features on remote wakeup so that host can
read, enable or disable remote wakeup support
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 3f9aaa0c92587e75ac4bfd1ad9c99d7caa282a15
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:40 2009 +0530
msm: usb: gadget: usb electrical test modes support
Add support for TEST_SE0_NAK, TEST_J, TEST_K, TEST_PKT electrical
test modes. To facilitate compliance testing, high-speed capable
usb devices must support above mentioned test modes.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit f9393243704fb1519e61d3f4cbaea37c97a5facc
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:39 2009 +0530
msm: usb: gadget: support usb protocol GET STATUS cmd
USBCV-CH9 requires support for GET STATUS command which can be used
to get status of device, endpoint or interface. It is used by USBCV
to verify if previous SET command was successful or not.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit f34befad0df93f1dccb0f6651bd0a28e38265f00
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:38 2009 +0530
msm: usb: gadget: support usb protocol SET FEATURE cmd
USBCV-CH9 requires to support SET FEATURE command to be supported.
SET FEATURE command can be used to set the HALT on an endpoint and
to enable/disbale remote wakeup. Currently we are supporting only
SET FETURE on endpoint only.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 12bb86b0938b81d30b12d47f9f491004c310034b
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:37 2009 +0530
msm: usb: gadget: add support for interrupt endpoint type
Add support for interrupt endpoint. CDC/ACM requires
interrupt endpoint support to work as modem.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit c32e8dd745877757ce8d35537cad6a3e99ef384e
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:36 2009 +0530
msm: usb: gadget: clear the stall bit on CLEAR FEATURE cmd
CLEAR FEATURE command on an endpoint was enabling endpoint instead
of clearing the stall bit.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 655af524e89f56be0fef64704e7f2c46e17563f0
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:35 2009 +0530
msm: usb: gadget: STALL support for endpoint
If device does not support any control request or non-control
endpoint requests, device can stall that particular endpoint.
Add support to stall endpoint.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 4aab336d4eb2480e53000198ffabaaea2552b3d0
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Fri Feb 6 19:28:34 2009 +0530
msm: usb: gadget: Send ACK handshake to CLEAR_FEATURE req for EP0
When device STALL's control endpoint, host sends CLEAR_FEATURE request
to clear STALL on endpoint-0. For Endpoint-0, hardware clears STALL
immediately after receiving SETUP token on EP0, hence software does
not need to explicitly clear the STALL for EP0 in hardware, but,
software must need to configure Hardware to send ACK handshake to
CLEAR_FEATURE request."
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 1a3b0ef3971f5ed3f0941a990b0662ba15d7c641
Author: Krishna, Vamsi <vskr...@qualcomm.com>
Date: Wed Mar 3 19:29:38 2010 +0530
msm: usb: gadget: support for ep0 out requests
Support for transfer on ep0-out endpoint.CDC/ACM requries
this feature to process set line coding commands.
Acked-by: Allam, Suresh Reddy <sal...@qualcomm.com>
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 2fcfcb812e1898473044f24e55e326737d52a1c6
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Mon Mar 8 11:32:36 2010 +0530
msm: msm72k_udc: gadget driver for msm7k
Change-Id: Id99283f74d0f28b3d2af698e998b2fe7a1c7d44c
Signed-off-by: Mike Lockwood <lock...@android.com>
[cvel...@qualcomm.com: Resolved merge and compilation issues]
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 7503c33e688cdb68f26c7d0920170d9c94dccbcf
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Mon Mar 8 13:29:38 2010 -0700
msm_kgsl: Use dmac_flush_range to flush cached GEM memory
Use dmac_flush_range to flush cached memory allocated by GEM rather
than flush_dache_page.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 6ea2c55dc7566b0b191989f307cb20613d98f94a
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Mar 5 19:43:05 2010 -0800
defconfig: Enable PPP support for ST1 and ST1.5.
Change-Id: I4cf75b484cfffb93ab4989bd08549ccbb6b11973
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit 58758d9caefefcad9a1f862e4f96a6f3f2c526b2
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Fri Mar 5 18:19:33 2010 -0800
alsa: soc: 8k: Fix build and config options for audiov2 interface.
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 7c2c4c1ba94adf7a39fecd626fe8a93c683edc53
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Mon Mar 8 11:29:44 2010 +0530
msm: msm72k_udc: gadget driver for msm7k
Change-Id: I15f5254b14dbb789920303603537fc32d81aa737
Signed-off-by: Mike Lockwood <lock...@android.com>
[cvel...@qualcomm.com: patch is split]
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 3c219d3c40f7ffe69d7d3ea2c10cad98ca37b5e3
Author: Michael K. Edwards <medw...@quicinc.com>
Date: Sun Mar 7 02:12:05 2010 -0800
vfp: change vfp context save logging to pr_debug to suppress excess spew
Change-Id: I8afb9bc251aa1d7e999e9ba929c3024a521b3c67
Signed-off-by: Michael K. Edwards <medw...@quicinc.com>
commit 71b02600e2e9eec4acdd2037dc05c5e5c4c9a085
Author: David Brown <dav...@quicinc.com>
Date: Fri Mar 5 17:17:31 2010 -0800
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4102098f465844f45e9f5257068c63c50d15e632
Author: Hanyu Cui <han...@quicinc.com>
Date: Fri Mar 5 19:19:53 2010 -0500
defconfig: enable asynchronous hardware bliting
Signed-off-by: Hanyu Cui <han...@quicinc.com>
commit 39e0415c353c5b13e48c94bf275ac8e355e0cb6a
Author: Anuradha Chandramouli <chan...@qualcomm.com>
Date: Fri Mar 5 13:35:52 2010 -0800
defconfig: Enable Libra SDIO Interface Module for ST1.5
CONFIG_LIBRA_SDIOIF=y
Acked-by: Anuradha Chandramouli <chan...@qualcomm.com>
Signed-off-by: Yunsen Wang <yun...@quicinc.com>
commit 686293249723c86a9545a669c2c1268402c94fe0
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Wed Mar 3 19:25:07 2010 -0800
msm: audio: 8x50: allow either legacy or new audio driver
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 563243796c2b1fdf8904fe2d8216a639bf60cfc0
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Thu Mar 4 23:19:16 2010 -0800
8x50: audio: Fix compile errors in audiov2 driver.
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit aed66db314d4e99ff81142c13e0308d318dba575
Author: Hanyu Cui <han...@quicinc.com>
Date: Fri Feb 26 15:15:04 2010 -0500
msm: mdp: Add support for asynchronous hardware bliting
EXA layer makes MSMFB_BLIT ioctl call into the MDP driver to perform
operations like copy bliting and compositing on a batch of ROIs. However,
the current MDP driver code for MSMFB_BLIT ioctl serves bliting requests in
a loop, synchronizing with the MDP PPP engine by blocking the user
application for every request. This results in huge overall ioctl call
latency and CPU underutilization. Profiling the driver code for video shows
the total bliting time to be around 21ms per frame. As a result video is
jerky and sluggish. We also see approx 30% of video frames getting dropped
despite only 30% CPU usage.
This patch adds a new ioctl for bliting, MSMFB_ASYNC_BLIT. Compared to
MSMFB_BLIT, its call latency is reduced by having enqueuing support for
blit requests and making the ioctl asynchronous. Another new ioctl,
MSMFB_BLIT_FLUSH, has also been added to let EXA layer explicitly
synchronize with the MDP PPP engine when needed. With these changes we
saw video becoming smooth (with no frame drops). Profiling shows approx
58% FPS improvement with compositing. Auto scrolling test time reduces
by about 7% under firefox.
Signed-off-by: Ashish Bijlani <abij...@quicinc.com>
Signed-off-by: Hanyu Cui <han...@quicinc.com>
Signed-off-by: Greg Reid <gr...@quicinc.com>
commit 4ea6fd482136d368f6a3fe763c7dc88a5467f537
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Wed Mar 3 15:53:28 2010 -0800
alsa: soc: 8k: Add support for qsd audio v1 and v2 devices.
The ALSA driver will use the native QSD driver to control the
SOC device. This driver will be one of two designs, a legacy
driver or a newer modular driver.
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit e29a8f72f8a2aa52c687879321ef89a67b052c1a
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Thu Mar 4 21:01:09 2010 -0800
qdsp6: add audiov2 pcm open routines
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
commit 3a159ed9f944dd610a1d7201cd3ee3a35b5bbf31
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu Mar 4 13:49:57 2010 +0530
defconfig: Enable usb host CDC ACM and ECM support
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 63aff97696b16370777c6ab19adcb14d0809d1e9
Author: Yunsen Wang <yun...@quicinc.com>
Date: Tue Mar 2 11:27:52 2010 -0800
wireless: libra: SDIO interface module
Libra SDIO Interface driver module. Adds interface functions
to interact with the SD/MMC bus driver.
Acked-by: Anuradha Chandramouli <chan...@qualcomm.com>
Signed-off-by: Yunsen Wang <yun...@quicinc.com>
commit 05be67c49da7a6a177db73257a0ce9046d017868
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Tue Mar 2 18:20:21 2010 -0800
qdsp6: add audiov2 driver
This driver will support:
* PCM playback and capture
* device switching
* volume control, mute/un-mute
* voice call
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 7722d284c460574053ddac577fd98a3eb132f87a
Author: Anirudh Ghayal <agh...@qualcomm.com>
Date: Wed Mar 3 15:38:40 2010 +0530
defconfig: Enable gpio_keys driver for Lid Sensor on ST1.5
CONFIG_KEYBOARD_GPIO=y
Signed-off-by: Anirudh Ghayal <agh...@qualcomm.com>
commit ed89369e292f5d3fc76a1681c614a09f056252aa
Author: Hanyu Cui <han...@quicinc.com>
Date: Thu Feb 25 16:32:22 2010 -0500
vfp: Handle vfp context preserve/restore during signal
Add vfp preserve and restore code to signal handler. Expose vfp
context to signal handlers. Use modified magic number to avoid
conflict with mainline userspace vfp API.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: Hanyu Cui <han...@quicinc.com>
commit 74e6902a7582021880802155e3f364b6228e5321
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Tue Mar 2 09:39:53 2010 -0700
msm_kgsl: Fix logic in a macro in the DRM driver
The TYPE_IS_PMEM() macro had broken logic which causes very old
version of the X driver to not work.
Change-Id: I6f8447bf5bad4b739cf47e0ba1f8b43fdd7c4e4b
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 4c9b365129d11014cf836598e64203198be4cb91
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Mon Mar 1 18:03:26 2010 -0800
8x50a: audio: Add audio GPIO for ST1.5.
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit 75b6be93a380c6cdfc75cbecc043e4d31dfb5e42
Author: David Brown <dav...@quicinc.com>
Date: Tue Mar 2 12:26:02 2010 -0800
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0a7d8aaf5257a1d40ab0f8b8b30b60a42a8a0377
Author: Ashwin Chaugule <ash...@quicinc.com>
Date: Mon Mar 1 11:38:11 2010 -0500
defconfig: Perf defconfigs for ST1 and ST1.5
- Removed all high overhead kernel debugging options.
- Removed all driver debug.
- Removed all drivers/options which we don't support/use.
Signed-off-by: Ashwin Chaugule <ash...@quicinc.com>
commit 1ac1feeba22fe26208b7fb5153ca2fc64a5acd31
Author: Michael K. Edwards <medw...@quicinc.com>
Date: Mon Mar 1 11:56:47 2010 +0530
msm: sdcc: ports 1 and 2: Increase clock drive strength to 16MA.
On ST 1.5 hardware, due to some differences in the hardware, 8MA drive
strength seems insufficient which is resulting in CRC errors. Increase
the drive strength to 16MA for CLK lines to get rid of these errors.
Signed-off-by: Michael K. Edwards <medw...@quicinc.com>
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 504820348fd09d7f369ed759c3abc5ad5713aa4d
Author: Michael K. Edwards <medw...@quicinc.com>
Date: Mon Mar 1 11:57:53 2010 +0530
msm: sdcc: ports 1 and 2: Reduce the clock frequency to 40.96MHz
On ST 1.5 hardware, due to some differences in the hardware, CRC errors are
observed with the clock running at 50 MHz. With clock running around 40 MHz
these errors disappeared.
Signed-off-by: Michael K. Edwards <medw...@quicinc.com>
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit f092ee67acdbf0ce3c478df02a879945cc502e3c
Author: Murali Palnati <paln...@qualcomm.com>
Date: Tue Mar 2 11:58:39 2010 +0530
[ARM] msm: Defer partition address and size computation.
Since both 2K and 4K page devices are to be supported, the partition
information that is obtained from SMEM or ATAGs which contains the
starting block number and the number of blocks, is insufficient to
determine the partition start and end addresses. This computation is
now deferred till the Nand device type is identified during the Nand
probe phase.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 69a7e26a54e0e1635944d48a0d166b30decc4171
Author: Anirudh Ghayal <agh...@qualcomm.com>
Date: Tue Mar 2 10:25:00 2010 +0530
msm: Add support for LID open/close detection on ST1.5
PT3661 Hall Sensor is used for sensing the open/close position
of the Lid. The msm GPIO used for detection is configured by the
gpio-keys driver. Based on the change in the status of the GPIO,
the gpio-keys driver reports the event (as SW_LID or !SW_LID)
to the input subsystem.
Signed-off-by: Anirudh Ghayal <agh...@qualcomm.com>
commit 7c991cf84dd452acbf647a3480b6eacb05a0bf4a
Author: Annamaria Pyreddy <c_ap...@quicinc.com>
Date: Wed Feb 17 19:26:57 2010 -0800
msm_q6video: remove the video encoder unused header file
Change-Id: Ibd22016e437509c7bb8fe8d07e9890b6d5256994
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit 4394a74fa0bd147b32ecfbcfef0603b42693ceeb
Author: Annamaria Pyreddy <c_ap...@quicinc.com>
Date: Wed Feb 17 16:04:52 2010 -0800
msm_q6venc: data structure changes for new encoder interface 2.0
New interface changes and move two qdsp6 buffers from qdsp6 to scorpion
to reduce qdsp6 image size,add state checks for dalsys calls.
This is not backward compatible.
Change-Id: I8d7d34d505004e900b62898c76399b35a9a4a43c
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit 5e6c950f7e65abeac9c1c6a0f9ec55ef2f0cdbee
Author: Annamaria Pyreddy <c_ap...@quicinc.com>
Date: Wed Feb 17 15:45:42 2010 -0800
msm: qdsp6: Implement new Scorpion/QDSP6 interface
- Scorpion/QDSP6 interface is changed for video decoder from version
2 to version 3.
- Add VDEC_IOCTL_INITIALIZE_V3, VDEC_IOCTL_SETBUFFERS_V3,
VDEC_IOCTL_FREEBUFFERS_V3, VDEC_IOCTL_GETDECATTRIBUTES_V3 ioctls
to retain the backward compatibility with version 2.
- Changes to pass the sequence header from pmem in vdec_initialize
instead of pointer to argument in rpc call.
- Add new ioctl VDEC_IOCTL_GETINTERNALBUFREQ to get the QDSP6
internal buffer requirements, previously we used to get these as
part of vdec_initialize call. This is to give flexibility when
number of internal buffers at QDSP6 side changes.
- Add new ioctls VDEC_IOCTL_GETPROPERTY, VDEC_IOCTL_SETPROPERTY to
get and set the different properties of various decoders.These
are not implemented but added for future use.
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit 70fe799d1dde7d509253e090733a0d0ddb830c19
Author: Annamaria Pyreddy <c_ap...@quicinc.com>
Date: Mon Mar 1 12:35:08 2010 -0800
Enable the config option for qdsp6 compilation.
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit aeeee28bbe4bf9914f4f799013b8cb4aaa8b7551
Author: Jin Hong <ji...@quicinc.com>
Date: Fri Feb 26 13:21:00 2010 -0800
defconfig: turn off PMEM_GPU0 for ST1 and ST1.5
Signed-off-by: Jin Hong <ji...@quicinc.com>
commit d77f3f0d496f570068739a238be1c0a38fe27dff
Author: Yunsen Wang <yun...@quicinc.com>
Date: Fri Feb 26 16:59:51 2010 -0800
defconfig: Enable Libra WLAN for st1 and st1.5
CONFIG_LIBRA_SDIOIF=y
CONFIG_CFG80211=y
CONFIG_WIRELESS_EXT=y
CONFIG_MMC_MSM_PROG_DONE_SCAN=y
PROG_DONE_SCAN is for SDIO after cmd53 - dummy cmd52 sequence.
Signed-off-by: Yunsen Wang <yun...@quicinc.com>
commit ad06f8ccbe4feed8ef3df02c0dd85881e601e2e5
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Fri Feb 26 13:56:30 2010 +0530
usb: gadget: use clk_reset API's for resetting hardware blocks
clk_reset API's are more reliable than direct Proc-comm calls
for resetting hardware blocks.
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit d71478a0a86422d14a15d88c02bead0da1a097e7
Author: Matthew Wilson <mtwi...@quicinc.com>
Date: Fri Feb 26 17:36:30 2010 -0600
defconfig: Enable HSUART module for ST-1 and ST-1.5
Change-Id: I68b058383873af77b515698e2c32d862d4fc15f1
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit 2cc95bb4cd2c1789c1412d6c74351c1c8405b860
Author: Matthew Wilson <mtwi...@quicinc.com>
Date: Fri Feb 26 18:10:43 2010 -0800
serial: Add MSM high speed UART driver
Change-Id: I77fae26eaa1da7d4f51fc39397af926d25d41b84
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit 7341741bb12583b4381baf2012fd60c798a447dd
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Wed Jan 6 23:07:17 2010 -0800
msm_rmnet: kill the tasklet when the network device is stopped
Calling tasklet_kill instead of tasklet_disable when the device is
stopped removes the tasklet from the scheduled list. tasklet_disable
is intended for temporarily deferring the execution of scheduled
tasklet and should not be used when rmnet device is stopped.
Acked-by: Jimi Shah <ji...@qualcomm.com>
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
commit d9691fb9821020b917a0fb0c2376e8ce0062f5aa
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Mon Dec 28 12:58:56 2009 -0800
msm_rmnet: set start of transmission time
The kernel TCP/IP stack calls timeout if trans_start + watchdog
timeout is beyond the current time. Set the trans_start just
before sending each packet to avoid timeout problems.
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
commit 4e3d3ef561a90697e1c2d30b903325ddd8b74d1b
Author: Ashwin Chaugule <ash...@quicinc.com>
Date: Fri Feb 26 11:23:19 2010 -0500
defconfig: Enable CPU frequency scaling for ST1
- Enable SVS
- CONFIG_QSD_PMIC_DEFAULT_DCDC1 = 1300
- Set max freq = 768Mhz
Change-Id: Ia8f99cddfe6057a9807cedfc4d25799a4c3f2bc9
Signed-off-by: Ashwin Chaugule <ash...@quicinc.com>
commit d130e5a2e304d75170683086ca236624cafe3d6c
Author: Jin Hong <ji...@quicinc.com>
Date: Thu Feb 25 17:01:07 2010 -0800
defconfig: enable DRM on ST1
Turn on DRM and KGSL_DRM for graphics.
Change-Id: Id1abf714200757184f12ced8cb3b3ffe1947b41c
Signed-off-by: Jin Hong <ji...@quicinc.com>
commit c7aeb9886c745a026ee0e17ccad7892ae414ba6a
Author: David Brown <dav...@quicinc.com>
Date: Fri Feb 26 12:17:03 2010 -0800
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9d057a5b2ec98a64eb387a5aee0d7b3c084a4980
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Feb 25 13:25:14 2010 -0800
diag: Various fixes for Qualcomm SoC diagnostics driver
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit e9675c6813f44d983c14aeb1cc17ccd19388d6a6
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Tue Feb 16 13:52:46 2010 -0700
msm_fb: Add support for changing the scan-out address
Allow appliations to change the buffer address that the
MDP displays from via FBIOPAN_DISPLAY.
Change-Id: I962e56939c6f676bfda4e21208852402a29c4a1a
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 16b70998926d05cb04598986b6114628068f0b80
Author: Suren Eda Naarayana Kulothungan <seda...@quicinc.com>
Date: Fri Feb 26 09:21:15 2010 -0500
msm: CP register access tool for Read/Write to CP registers
The tool can be used to read/write to CP registers by
passing the CP parameters through /sys interface.
Signed-off-by: Suren Eda Naarayana Kulothungan <seda...@quicinc.com>
commit 0f798e27956e0cce4c1f8b5008e75bd312a162ea
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Fri Feb 26 08:57:43 2010 +0530
defconfig: Enable diag on gadget
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 3aa0c2ed39ae3361fd354d890589b54db9a90448
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Tue Nov 24 10:22:09 2009 -0800
driver: interrupt based flow control in msm_rmnet driver
In the event when rmnet_xmit can not write to smd channel, save
the data to be sent and disable the up link flow. Use the generic smd
callback function to check if smd channel is available, and if so,
schedule a tasklet to transmit the saved data and enable the up
link flow. In the event when a rmnet device is stopped, disable
the corresponding tasklet and stop the data flow.
Acked-by: Jimi Shah <ji...@qualcomm.com>
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
commit cb1f099dd4d5b13af38d08bd6dfc439d3cf4493f
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Mon Oct 12 17:17:09 2009 -0700
driver: fix msm_rmnet driver problem with dropping packets
In the event with rmnet_xmit can not write to smd channel, it schedules
a timer based callback only once to attempt the write later. If the smd
channel is clogged when the callback is invoked the timer is rescheduled
once again. This single reschedule of the timer continues until the
write succeeds, essentially introducing uplink flow control in the driver.
CRs-Fixed: 180014
Acked-by: Jimi Shah <ji...@qualcomm.com>
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
commit edfab4c59e3a2de88933fef6b364c1df8f595d3c
Author: Iliyan Malchev <mal...@google.com>
Date: Thu Dec 10 15:06:39 2009 -0800
[ARM] qsd8k: initialize the spinlock in dal_client
Change-Id: Ia276c39f494a6321180fe66b44f2ba067cf3eba1
Signed-off-by: Iliyan Malchev <mal...@google.com>
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0cfaeae630617c88ab695705e6f08e420ef6e611
Author: Brian Swetland <swet...@google.com>
Date: Mon Nov 16 21:39:02 2009 -0800
qsd8k: audio: enhance stuck dal debugging
Dump the packet we were in the act of sending, if we timed out while
waiting for a reply. This covers the situation of failing on a
different channel than one being traced.
Change-Id: I33c00325446d6c0e3d95880ad492cbffe08e28ee
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d708c0ab911b8867b970f32b4024e29b58b70988
Author: Brian Swetland <swet...@google.com>
Date: Tue Nov 10 21:07:14 2009 -0800
qsd8k: dal: add tracing capabilities to DAL channels
If requested, log the last 32 calls, replies, and/or events.
Change-Id: Iee69b275430d86abd3818dae09611dbb8c3a68c3
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3f7d5ce968f0081990ebb773ee43855db7c97704
Author: Ashraya Arya <ashr...@quicinc.com>
Date: Thu Feb 25 17:52:04 2010 -0800
msm: audio: 8x50: New Voice call interface for Audio driver
New voice call interface to support CVD 1.0 associated with the new
q6 firmware
Change-Id: I1a32a8e7cbfea1c07db8a60c9b4e03cf7c8dea8e
Signed-off-by: Ashraya Arya <ashr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8690aa330463d6f3883dc55eff2b301c2b326be4
Author: Brian Swetland <swet...@google.com>
Date: Thu Feb 25 17:47:33 2010 -0800
qsd8k: audio: tracing, and stuck dsp panic code
If the dsp doesn't respond in 5s, it will probably never respond,
leaving the device in a bad state -- panic in that case.
Add some debug chatter on pcm in/out open and close as well.
Change-Id: Id9d0e0084a01089b58a9d98a33672a9c54bdc6b0
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7a0613762ccba9180c9666a169e5a04e5f0cee76
Author: Alex Wong <wa...@quicinc.com>
Date: Thu Feb 25 17:43:32 2010 -0800
msm: audio: 8x50: Add dal interface version check in audio driver
Add a new dal function to get the dal interface version. Update the
audio driver to check interface version for each dal driver,
if either one of the dal driver doesn't match the driver version.
The audio driver will fail at init time and allow kernel to boot up.
Change-Id: I11b8cf3a542338347f58c5ccacb744c5beed4d91
Signed-off-by: Alex Wong <wa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0f4e582d3c8eb079d61935ae1d71cccb43e8e104
Author: Rajesh Sastrula <vra...@quicinc.com>
Date: Tue Dec 15 14:55:52 2009 -0800
msm: qdsp6: Add convenience wrapper for rpc idx 14
New convenience wrapper dal_call_f14 is added since the new video
interface changes requires it.
Change-Id: If8ea78203688da9fe4025c972639e5cfe5076169
Signed-off-by: Rajesh Sastrula <vra...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 434bc40ea94096b8ecc95fb87a4afa38c41012dc
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Thu Feb 25 17:41:12 2010 -0800
[ARM] msm: qdsp6: Support Qualcomm's SMD Driver
The qdsp6 uses DAL which uses SMD. To support Qualcomm's SMD driver a
CPU destination parameter was added to dal_attach and the portname
used to open the DAL channel (SMD channel) was updated.
Change-Id: I750920afd4a46c73c50b678ea463994fe23b0f4c
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit af018e5ceb5b7e9654271dca1e771bc00a52b482
Author: Dima Zavin <di...@android.com>
Date: Thu Feb 25 17:39:12 2010 -0800
[ARM] msm: Add qdsp6 config option and makefile
Change-Id: I6120d03a96339139d4d88f32e165d26b75ac805e
Signed-off-by: Dima Zavin <di...@android.com>
[dav...@quicinc.com: only include changes in dal.c]
Signed-off-by: David Brown <dav...@quicinc.com>
commit 82e0f3a0226b76f46ea9f544235c80763cdc1710
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Wed Nov 12 17:11:18 2008 -0700
msm_rmnet: Fix SMD port names
After the SMD named port merge, all users of SMD pass a SMD name
to open the port. This name has to match the name used in
shared memory. This fix aligns these.
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
commit 8f895e8a381ce56592ffa78c5c1404fe9576cb0b
Author: Brian Swetland <swet...@google.com>
Date: Tue Feb 9 22:46:19 2010 -0800
[ARM] msm: driver for rmnet virtual ethernet interface
The MSM7X00A baseband makes up to 3 "virtual ethernet" channels available,
which allow ethernet packets to be exchanged with the cellular network, once
an appropriate data connection is established.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: San Mehat <s...@android.com>
[ARM] msm_rmnet: HACK: do not count ARP packets
The android network traffic watchdog is tricked into thinking that
data traffic is working at times when it isn't, due to ARP traffic
between the apps and modem processor. Don't count ARP packets in
link statistics to avoid this problem.
[ARM] msm: rmnet: Add stat tracking for number of radio wakeups occur.
There are two paramaters that appear for all rmnet devices.
rmnet0 for example:
/sys/devices/virtual/net/rmnet0/timeout (RW)
/sys/devices/virtual/net/rmnet0/wakeups (RO)
timeout is configured by userspace for the proper network timeout values
wakeups is the number of radio wakeups that occured.
By default timeout is zero which means the stats are disabled.
MSM_RMNET_DEBUG must be set.
Signed-off-by: Mike Chan <mi...@android.com>
[ARM] msm: rmnet: Track wakeups due to xmit/rcv instead of globally.
Signed-off-by: Mike Chan <mi...@android.com>
[ARM] msm: rmnet: Different stat timeouts when screen is on/off.
Timeout for modem powerdown can differ when the screen is on/off.
Allow timeout to change via early suspend/resume hooks.
Signed-off-by: Mike Chan <mi...@android.com>
[ARM] msm: rmnet: Lock a wakelock for half a second when receiving data.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
[ARM] msm: rmnet: Tracks total awake time when the rmnet is active.
Exports data in /sys/devices/virtual/net/rmnet0/awake_time_ms
in time expressed as ms awake.
Signed-off-by: Mike Chan <mi...@android.com>
[ARM] msm: rmnet: Fix compilation issue when MSM_RMNET_DEBUG is not set.
Signed-off-by: Dima Zavin <di...@android.com>
msm_rmnet: fix to build on 2.6.32
Change-Id: Ic6a4903dd12ea83723354d00f639ae2f9375167f
msm_rmnet: ensure packet writes are atomic
Use the smd_write_atomic() function to prevent concurrent
packet writes to the transport from stepping on each other.
Signed-off-by: Brian Swetland <swet...@google.com>
commit d9ca763e3092063a843ca3bbd680bb701ffe7ba0
Author: Brian Swetland <swet...@google.com>
Date: Mon Dec 7 15:28:08 2009 -0800
msm: smd: provide atomic channel writes
Some smd clients may write from multiple threads, in which case it's
not safe to call smd_write without holding a lock. smd_write_atomic()
provides the same functionality as smd_write() but obtains the smd
lock first.
Signed-off-by: Brian Swetland <swet...@google.com>
commit f1a912879889b804f0b484ebbb1ce94face6e71a
Author: Brian Swetland <swet...@google.com>
Date: Tue Jul 14 11:26:42 2009 -0700
[ARM] msm: qsd8k: Device Access Layer (DAL) RPC
Device Access Layer (DAL) RPC is yet another remote interface to software
on the baseband and dsp processors of the msm7k/qsd8k family. It uses SMD
as a transport.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
commit e847e6f783796de4027789592f3f6cdd65c7c4c8
Author: Premal Shah <ps...@quicinc.com>
Date: Wed Feb 24 15:18:13 2010 -0800
pmem: mark pmem_adsp region as cached
Flash10 player on ChromeOS does color conversion in the cpu and hence
for performance reasons we want to make the decoder buffers as cached.
Subsystems that use pmem_adsp region (vdec/camera) can still use
this area by using O_SYNC flag which will give the original
uncached behavior.
Change-Id: Ibfe098ff047d945e80a754b3fbf78d313fe81788
Signed-off-by: Premal Shah <ps...@quicinc.com>
commit 77de472135773ca2db5589fb7e505606c1a6d237
Author: Ashwin Chaugule <ash...@quicinc.com>
Date: Thu Feb 25 14:44:49 2010 -0500
defconfig: Removed unnecessary options from ST1.5 defconfig
Removed:
- Drivers for which h/w doesnt exist on the ST1.5
- Needless options for FB logo
Signed-off-by: Ashwin Chaugule <ash...@quicinc.com>
commit e8750550ea1b3ae881d1d48e7626a967d8646bf3
Author: Ashwin Chaugule <ash...@quicinc.com>
Date: Thu Feb 25 14:40:06 2010 -0500
defconfig: Removed unnecessary options from ST1 defconfig
Removed:
- Drivers for which h/w doesnt exist on the ST1
- Needless options for FB logo
Signed-off-by: Ashwin Chaugule <ash...@quicinc.com>
commit 7c2d1d32a4262cf391e488b392ebe86c9d70bd82
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Thu Feb 25 11:35:39 2010 +0530
USB: gadget: Add chrome composite gadget
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
commit 590ae6617ea46e45cae2b221fcd0effa7de3ddbb
Author: Wilson, Matt <mtwi...@quicinc.com>
Date: Fri Feb 19 16:57:21 2010 -0600
defconfig: enable Bluetooth power driver for ST-1.0 and ST-1.5
Configures Bluetooth power driver as built-in modules
for QSD8650 and QSD8650A on both ST1 and ST1-5
Change-Id: I4018eebb3c3d2d7c9c5a5f63c071bed675ee58c9
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit 4bb69d1e0049721fba102c9a703b28de95517550
Author: Wilson, Matt <mtwi...@quicinc.com>
Date: Tue Jan 12 14:06:58 2010 -0600
bluetooth: remove initial power off for Bluetooth
hci_core in Bluetooth requires rfkill non-blocked.
Remove initial blocked state from power switch driver to
allow hci_core to start.
Change-Id: I655a312bcc3e5157799b38927b77ef134072ee91
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit a39be01773ba31f1850f3706c5dc869e2683bcb4
Author: Wilson, Matt <mtwi...@quicinc.com>
Date: Fri Feb 19 16:37:55 2010 -0600
bluetooth: Add Bluetooth power driver
Bluetooth power control based on rfkill
Change-Id: I3e7de7804a5f1a2a049d6dfe301a5e8dc30599df
Signed-off-by: Philip Elcan <pel...@quicinc.com>
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit 03de72ecbdfb4ff050bd6c143aa8002c8ac66cbb
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Tue Feb 16 09:30:15 2010 -0700
msm_fb: Make the pitch for the framebuffer 32 pixel aligned
The adreno GPU uses 32 pixel aligned buffers, so make sure the
framebuffer alignment is also 32 pixel aligned to make it easier
to do flipping.
Change-Id: I04c3e115a0e3d58c53ce544a512f80a3fe9ffcb9
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 49ffbece108a33a8ee36b2652a40978fa268c9bd
Author: Wilson, Matt <mtwi...@quicinc.com>
Date: Mon Feb 22 14:51:17 2010 -0600
[ARM] msm: bluetooth: power switch hardware support
Add power switch support on ST-1.0 and ST-1.5 boards.
Change-Id: I3c1c22da
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
commit b8c93de905f85860cbb58ee2ae6be20775401a5f
Author: Nick Pelly <npe...@google.com>
Date: Fri Mar 20 13:53:00 2009 -0700
msm: msm_serial_hs: Disable datamover when clocking off uart.
Signed-off-by: Nick Pelly <npe...@google.com>
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
commit 57a101241a31c7d5d7f3b69086216de79be6a9f0
Author: Nick Pelly <npe...@google.com>
Date: Wed Dec 17 10:43:57 2008 -0800
msm: msm_serial_hs: Hold wakelock while rx dma in progress.
Use RXLEV with watermark level of 0 to acquire wakelock.
Signed-off-by: Nick Pelly <npe...@google.com>
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
commit 02e8c4ca229c82d214d194d4f6ee1cf35e978ba3
Author: Nick Pelly <npe...@google.com>
Date: Thu Dec 11 15:22:57 2008 -0800
msm: msm_serial_hs: Add optional low power wakeup.
Low power wakeup is configured using msm_serial_hs_platform_data. This allows
us to clk_enable() the uart on another interrupt (for example a gpio rx line).
It also allows us to inject an char into the rx tty layer when this occurs.
This is useful for uart attached devices that have low power modes independant
of the linux CPU, for example bluetooth chip.
Currently the interface to control low power mode is through
msm_hs_request_clock_on()
msm_hs_request_clock_off()
Ideally we want to use the tty/serial layer but we do not have a good solution
to this yet.
A few other misc changes in here
- Use _locked convention for functions that hold spinlock
- request_irq() and free_irq() in startup/shutdown not probe/release
- Use BIT(x) helper in the hw_reg header. Add missing bitmasks.
Signed-off-by: Nick Pelly <npe...@google.com>
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
commit d16879f966847a759f9303f5e3e6e918674ec6ae
Author: Nick Pelly <npe...@google.com>
Date: Tue Dec 9 09:38:21 2008 -0800
msm: msm_serial_hs: Wakeup on uart1dm irq
Other minor fixes: disable_irq in the isr is redundant because of spinlock,
removed. Removed a stray debug comment.
Signed-off-by: Nick Pelly <npe...@google.com>
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
commit 081faa33ec6e352e6b2e81cbc954f8545683578d
Author: Nick Pelly <npe...@google.com>
Date: Mon Dec 1 22:41:17 2008 -0800
msm_serial_hs: Redo hs uart driver for MSM7k
- Do not use workqueues
- Use dma_sync instead of dma_map in rx and tx callbacks
- Fix bug where last RX is repeated into tty on shutdown
- Fix membase / mapbase mix-up
- Fix shutdown and remove paths. Tested as module
- UART2 now works
- Use container_of and correct types instead of casts
- Use msm_hs_ prefix for local functions and structs
- Remove the non-static pm interface
- Acquire uart spinlock in irq handler, dma callback and set_termios
- Use waitqueue instead of completion
- Use get_resource helpers
- Remove orphanned Kconfig options
- Use new clockname/device node key for clocks
- Plenty more misc cleanup
Signed-off-by: Nick Pelly <npe...@google.com>
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
commit c00adb5704378644b5bb81c2bb7d655067a57894
Author: Mike Lockwood <lock...@android.com>
Date: Fri Nov 14 12:19:45 2008 -0500
msm_serial_hs: HSUART driver from Qualcomm
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: Sankalp Bose <sank...@qualcomm.com>
commit 9a5fdc69987274dcbe3d84ecaee4791ddce828ee
Author: Sheetal Sahasrabudhe <shee...@quicinc.com>
Date: Wed Feb 24 08:43:40 2010 -0500
KSAPI: Fix the global thread and AXI counters
This fixes an issue with the global thread and AXI counters where
the counters are not updated correctly under some conditions.
Change-Id: I2f14c8fd749787ec44e313c1579bacac3ae92338
Signed-off-by: Sheetal Sahasrabudhe <shee...@quicinc.com>
commit 8bd1b4cba2d618e3e6d910e113d038d6f2450823
Author: Stephen Boyd <sb...@quicinc.com>
Date: Fri Feb 12 09:29:11 2010 -0800
msm: clock: Add clocks for 8x50a
On 8x50a the GSBI clocks are used by the QUP I2C block. To keep
the clock names consistent with other targets, we simply call
these qup clocks.
While we're here add the GRP2D and GRP2D_P clocks.
Change-Id: Ie73ecc2a2f33ff44eced658736d2370519f629f0
Signed-off-by: Stephen Boyd <sb...@quicinc.com>
commit 82a6fc5f0903d51e741d0169e55166a6f5a6bd95
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Wed Feb 24 12:05:01 2010 +0530
mtd: Enable MTD tests by default.
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit d4a11fdf25652b6798ad2f50810d00fbeec21ec2
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Nov 3 10:35:41 2009 -0800
msm: acpuclock: Improve debug messages.
Use cpufreq's debug message framework to print debug messages. Also
make consistent use of pr_err(), pr_info() and pr_warning() macros.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit 03b9f96f250328449e5279604f9789db497f3f81
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Feb 4 15:18:41 2010 -0800
msm: Move Power Collapse and SWFI ACPU freqs into acpuclock drivers.
Power Collapse and SWFI ACPU frequencies shouldn't be modified
separately from the acpuclock drivers. Hence, move the constants
from the board files into the driver files.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit 3d52d080ce1e85d0d0dfc56c4e5eb17d14ddd5f2
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Nov 5 13:48:30 2009 -0800
msm: acpuclock: Fix ACPU frequency units.
Remove repeated conversions between Hz and KHz. The power_collapse_khz
and wait_for_irq_khz variables are also corrected to contain values in
KHz instead of Hz.
acpuclk_set_rate() now takes a rate argument in KHz. Likewise,
acpuclk_get_rate(), acpuclk_power_collapse(), and acpuclk_wait_for_irq()
return rates in KHz.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit 8a37863dee396d460cc1a519730859fc346e1af8
Author: David Brown <dav...@quicinc.com>
Date: Tue Feb 23 10:38:05 2010 -0800
defconfig: automatic update
Change-Id: I08eb1ffdbadccfbb4a6a2ffc1c9b6dc9a13fcd0c
Signed-off-by: David Brown <dav...@quicinc.com>
commit fbfe9475430ad4f1db20d8bafb991f8d9af185c8
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Dec 9 16:30:09 2009 -0800
msm: smem_log: Check for remote spinlock initialization failures.
smem_log_init() should return an error code if remote spinlock
initialization fails.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit ab12e38d0b59778a438853c1176ebd3881d4dc5a
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Dec 8 10:26:59 2009 -0800
msm: remote-spinlock: Return error if DAL remote lock init fails.
The DAL remote lock initialization function should print an error
message and return -EINVAL if the remote lock is not found in SMEM.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
commit 23f23f4de153cea2507b2e87bb6da52f24954f62
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Tue Feb 23 14:19:52 2010 +0530
sdio: Claim and release host in the sdio_reset_comm API
Change-Id: I1f127a2e483464ea939ea159e6388574e609ac1a
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit 6f41c7be8828e7ec740b080fb8994753b1507a57
Author: Murali Nalajala <mnal...@qualcomm.com>
Date: Tue Feb 23 19:44:20 2010 +0530
msm-nand: Add support for msm nand driver.
Add support for msm-nand driver.
Signed-off-by: Murali Nalajala <mnal...@qualcomm.com>
commit e4ea61de85558b716aff4d4fa7e795746e4e06b2
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Thu Feb 18 11:35:51 2010 +0530
sdio: Activate high speed clock after SDIO reset.
sdio_reset_comm API resets the SDIO card. After the SDIO reset,
enable high speed mode and set the clock to the high speed.
Fixes CRC errors caused by some WiFi cards after waking up
from a low power mode.
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
commit 881e6b0d1a70e22a347ab974fbbeba15b3eb23bf
Author: Philip Elcan <pel...@quicinc.com>
Date: Fri Feb 19 18:15:11 2010 +0530
mmc: msm_sdcc: Add support in msmsdcc driver for dummy 52 reads.
Certain SDIO devices require the host to send a dummy cmd52 read
for each data transaction. This patch adds support in the msmsdcc
driver to enable this.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 607779580fb063943837439a80581fbee41721b6
Author: Philip Elcan <pel...@quicinc.com>
Date: Fri Feb 19 18:13:31 2010 +0530
mmc: Add support in board file to enable dummy 52 reads.
Certain SDIO devices require the host to send a dummy cmd52 read
for each data transaction. This patch adds support in the board
files to enable this.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 9fb39a2c2084fa1a4e78f422f21ca5c3713f0728
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Tue Feb 16 20:54:18 2010 +0530
msmsdcc: Fix to prevent a compilation issue when DEBUG_FS is disabled.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit fde0261e545244f7f6e3d246ecfd7f0ab549b519
Author: Sahitya Tummala <stum...@qualcomm.com>
Date: Fri Jan 22 12:43:56 2010 +0530
mmc: msm_sdcc: reset SDCC in case of data transfer errors
1. If it is a DMA operation, reset it after the callback from data mover
is received for flush/error condition.
2. If it is a non DMA operation, reset it in SDCC interrupt handler for
the error case.
CRs-fixed: 222926
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
commit 722fa6846f836443286a67974e2e4a4768263b21
Author: Naresh Babu Saladi <c_ns...@quicinc.com>
Date: Wed May 20 15:04:17 2009 -0700
[ARM] oprofile: Add scorpion support to Oprofile kernel driver
Qualcomm Scorpion based SOCs contain additional hardware
performance counters that are not present on other devices.
Add support for these additional counters to the Oprofile monitor.
Change-Id: I3de2db8fc794a7ea867fd2373d878b4240f057f4
Signed-off-by: Sheetal Sahasrabudhe <shee...@quicinc.com>
commit 1c4c9ef994828c286589ff93cbb9ae385b2b383a
Author: Annamaria Pyreddy <c_ap...@quicinc.com>
Date: Wed Feb 17 15:29:18 2010 -0800
msm_q6video: remove the video decoder and encoder drivers
New video driver variant is merged at arch/arm/mach-msm/qdsp6
folder hence removing the existing one to resolve conflicts.
Signed-off-by: Annamaria Pyreddy <c_ap...@quicinc.com>
commit 019d3eaac50b811e36c8bd50d3e73bad7162b6d1
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Feb 5 14:50:21 2010 -0700
msm_kgsl: Add support for double buffered GEM objects
Add support for creating multiple buffers within a GEM object to
facilitate double buffering and flipping for 3D applications.
Change-Id: I909d82531e1e2dc5da3ec79aa0f1adc8a9a504f7
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 95d06bea66f5c363621821bb0dcac87ad17692a9
Author: David Brown <dav...@quicinc.com>
Date: Sun Feb 14 07:30:31 2010 -0800
defconfig: automatic update
Change-Id: I128ed5debb30e004578f16a21f94a3baca619e61
Signed-off-by: David Brown <dav...@quicinc.com>
commit e0e0b2f8e77b0d8ec20b99e3aba506606497be25
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Feb 3 16:38:26 2010 -0500
msm_fb: add SII9022 HDMI driver support
Use 24-bit (RGB888) support for ST1.5. Also rearrange
makefile order to ensure HDMI panel appears as secondary
panel if selected along with another primary display.
Change-Id: Ie0c87f195c4d2434e8f9ef131861e11f0935c2cc
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit f799482879a80bf2868be06afa9870bc016fa07e
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Feb 3 16:55:16 2010 -0500
msm: Add HDMI support for ST1.5
Change-Id: Ia00bd7c341665fedba596cdf0177ab09ae834e65
Signed-off-by: Neil Leeder <nle...@quicinc.com>
commit cede91b9a42f0b77d8626f78a36416d83bc2e1ed
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Feb 5 15:00:20 2010 -0700
pmem: Add range checking to pmem_kapi_free_index_bitmap
pmem_kapi_free_index_bitmap was returning a bit index for all memory
addreses passed to it, even those it didn't own. Add range checking
and return -1 if it doesn't belong.
Change-Id: I57d30d95e7950ed2b337fac40382e8fc9d9b409d
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
commit 98d5856c41be284c8794ac4a645361e7b3b7ba94
Author: Saravana Kannan <ska...@quicinc.com>
Date: Fri Feb 5 13:15:06 2010 -0800
msm: acpuclock-8x50a: Bump up max CPU freq to 1.19 GHz.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
commit b59a778a76b0dc86047028def15d1c9ef398458a
Author: Naveen Ramaraj <nram...@quicinc.com>
Date: Fri Feb 5 14:54:02 2010 -0700
pmem: Remove extra IO-remap for In-Kernel PMEM memory
In-Kernel PMEM allocation returns a physical address which would
be IO-remapped by callers if they need to obtain a kernel virtual
address. Remove the unnecessary IO-remap in the pmem driver during
initialization which diminishes the kernel virtual address range.
Also, explicitly log the memory type during initialization.
Change-Id: I326452f5025feeb110b89a555c0aa6bc452b44f7
Signed-off-by: Naveen Ramaraj <nram...@quicinc.com>
commit 9f14512ff4ea7e5ca1beef431cf13f8832f3dfc1
Author: Bryan Freed <bfr...@google.com>
Date: Tue Feb 16 18:19:35 2010 -0800
Add the msm_mdp.h file to the list of headers to be installed.
This header is needed by the msm xf86 driver.
Review URL: http://codereview.chromium.org/608010
This is a cherry-pick from 2.6.31.5 to 2.6.31.12.
Review URL: http://codereview.chromium.org/600163
commit 36b8233b05c1ce725a28d0be6447f96c8e327445
Author: Bryan Freed <bfr...@google.com>
Date: Tue Feb 16 16:43:49 2010 -0800
Update the ARM config file to support USB.
This gets USB keyboard and mouse working in the 2.6.31.12 kernel.
Review URL: http://codereview.chromium.org/601084
commit 8352379d8deafa07725d2af0cfd847a4892a9f3f
Author: Bryan Freed <bfr...@google.com>
Date: Tue Feb 9 16:41:40 2010 -0800
Update an ARM config file for the kernel ebuild.
Review URL: http://codereview.chromium.org/594021
commit d6607c26d94c90ef8b690a60e032d2384e90528a
Author: Jeffrey Ohlstein <johl...@quicinc.com>
Date: Wed Oct 7 10:08:53 2009 -0700
msm: fix warnings introduced in gcc 4.4.0
Change-Id: I977873d6d84417c47e10404bea10def7182eb703
Signed-off-by: Jeffrey Ohlstein <johl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eff5c572d04ffca676896e13c9ffd6f62fe19f63
Author: David Brown <dav...@quicinc.com>
Date: Mon Feb 1 13:59:33 2010 -0800
defconfig: Enable kgsl for st1.5
Signed-off-by: David Brown <dav...@quicinc.com>
commit ca9ffac2e0a0643f4a3ed966c879f387e3dd0b50
Author: David Brown <dav...@quicinc.com>
Date: Fri Jan 29 11:04:22 2010 -0800
defconfig: autoupdate
Signed-off-by: David Brown <dav...@quicinc.com>
commit 96cd3411c619365b6576869be76947413a1f0979
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Fri Jan 29 15:18:32 2010 -0800
msm_kgsl: fix module init call for builtin
Has to be called as a late_initcall.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit fc74dc975fbbc03b5643b9c049ef6cff6c000475
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Fri Jan 29 14:31:21 2010 -0800
msm: drm: couple stray bit removed to compile
A few bits that got in which prevent compiling of DRM.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 25d131073779c3279e8e9bebe00627232171ddd5
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Jan 27 12:57:16 2010 -0700
msm_kgsl: Allow greater control over GEM memory types
Turn the GEM memory type #defines into a mask to give userspace
more control over the caching scheme that should be used.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit be12e6094d4cd3c6d84cdf968b0bfdcfe71decf5
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Tue Jun 2 10:19:28 2009 -0600
msm_kgsl: Clean up memory when timestamps are freed
Call yamato_runpending when a timestamp is freed to ensure that
the memory actually gets cleaned.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5d91dbe31d62f0891903d08b8c6e63dc84e2a49b
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Nov 20 11:37:40 2009 -0700
msm: Add in-kernel SMI PMEM region
Create a SMI region accessible by the in-kernel PMEM API.
The region is optional and requires disabling the PMEM GPU0
device.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1e1cf5d27f2b10c2417adf8b49abd097dda7034d
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Fri Nov 20 11:37:38 2009 -0700
msm: Make the GPU0 PMEM region optional
Allow the user to disable the GPU0 PMEM region at config time.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bc11f24dabd411eca7e6ab92c1bc31a2a2b9e2c8
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Fri Jan 29 12:34:44 2010 -0800
msm: defconfig: qsd8650 st1 turn on graphics support
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 9dac10eac9612f60e4fd6747cf06dc069dcdede9
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 28 10:03:49 2010 -0800
kgsl: trivial compile fixes
Just some fixes to get everything compiling.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 4e2f0799d78dedbcd2254a56083e34e1511221d0
Author: Susan Brownhill <sbro...@quicinc.com>
Date: Thu Jun 4 12:47:09 2009 -0700
msm_q6venc: enable cache for 8k video encoder
Add rlc_buf_map to record virtual addresses to be used
in cache invalidation function. Add rlc_buf_ptr/rlc_buf_len
to record current rlc buffer. The other(non-current) rlc
buffer gets invalidated before each encode command is sent
to q6.
Signed-off-by: Susan Brownhill <sbro...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5e626b182affd34b49249f2dff7d5e37be97cc97
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Tue Jan 19 11:04:25 2010 -0700
msm_kgsl: Map all memory allocations through the MMU
Map all memory allocations through the MMU. The
only allocations that are not mapped are the allocations for MMU
page table and a dummyspace allocation whose address is written
to the REG_MH_MMU_TRAN_ERROR register.
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 830de8798458a3ad7369c16d03822f784d609ecf
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Jan 13 09:48:18 2010 -0700
msm_kgsl: Only allow MDP blitting from from PMEM
Only PMEM regions can be used with the MDP, so don't allow
KMEM regions to be returned from kgsl_gem_obj_addr().
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 625345ee5363b95b22f2fddd2f8316846e9219b8
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Thu Nov 5 17:05:14 2009 -0800
kgsl: add barriers for pmem noncached changed to Normal
Changing pmem noncached mapping from Strongly Ordered to Normal
Noncached memory means it can have out of order accesses and can
also have speculative fetches. Ensure correctness using barriers
Change-Id: Idc7b6e4ad3efeb4470103e60b1d169a6e6246677
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit fce87174a7bee93bbe2f83475a47d86fc40410a0
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Tue Dec 22 13:41:12 2009 -0700
msm_kgsl: Disable shadow writes
Shadow write feature has multicontext bugs and hence we need to disable
this feature.
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4830b039000aa96cf3fd84222f998a005fd522bb
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Fri Dec 11 14:26:30 2009 -0700
msm_kgsl: Do not wait for the REG_CP_RB_WPTR to be updated
kgsl_yamato_regread can cause rbbm read errors and it is not nessasary
to wait for REG_CP_RB_WPTR to be updated here anyway.
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 950dce513e10b105ff910c63782bfb2b10ac64db
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Fri Dec 11 14:29:39 2009 -0700
msm: kgsl: Call kgsl_yamato_runpending before allocating memory.
Calling kgsl_yamato_runpending before we allocate memory will cause
the driver to free all pending buffers which will improve the chance
of memory allocation on low memory conditions.
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4010e4c406d25cdb46db5b98a4d3d24262bb7817
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Dec 16 14:05:12 2009 -0700
msm_kgsl: Flush cached buffers in GEM
Before executing a command buffer in the GPU, flush all cached buffers
mapped in the GPU by DRM/GEM.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f3246e4bda130f97c782c55b124663b4d7ad797c
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Dec 16 14:05:10 2009 -0700
msm_kgsl: Add GPU bind and unbind ioctls
Add ioctls to bind and unbind buffers in GPU memory.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0dafc51962faf57ea273cd58c31d9461a8be536f
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Dec 16 14:05:09 2009 -0700
msm_kgsl: Enable GEM to allocate memory from different sources
Enable GEM to allocate its memory from a number of different sources
including the in-kernel PMEM regions or from general purpose kernel memory.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ac8194e15799c1a1f59b4e1caa958b5e88c7b412
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Dec 16 14:05:07 2009 -0700
msm_kgsl: Add generic pagetable allocation functions
Add generic functions for creating and ref-counting pagetables so that
DRM/GEM can access the pagetable structures.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b1000f493c9e1f6a091620e5fecb9fa853fa6473
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Dec 16 14:05:06 2009 -0700
msm_kgsl: Move kgsl_sharedmem_init to module load
kgsl_sharedmem_init/kgsl_sharedmem_close only need to be called once,
so move them to the the module load and close functions so they can be
available through the life of the module.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 330a88e9cefed309885d044c509465f6beaa7113
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Dec 16 14:04:47 2009 -0700
drm: Add drm_vm_close_locked()
Add drm_vm_close_lock() to allow GEM to free VM
entries while already holding the device drm mutex.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1882ba91d0ed4c34ca5231e6284b5d618b09e78f
Author: Kiran Kandi <kka...@quicinc.com>
Date: Fri Dec 11 22:54:55 2009 -0700
msm_kgsl: Add outer cache flushing if it exists.
Flush outer cache if it exists and rename the kgsl flush
routines to more appropriate names
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 85cefb2244bfdb4cc70e1e2dda201dd0e115e803
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Fri Dec 11 23:01:16 2009 -0700
msm_kgsl: Fixes multicontext screen corruption issue
This change fixes an issue in multicontext environment
where some part of texture of one apllication gets displayed on
the screen surface of another application
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 76cb539e77f761fb2d7c046d1c6a3a0537f47af1
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Fri Dec 11 13:44:23 2009 -0700
msm_kgsl: Set correct values for PM4_DRAW_INDX_BIN and PM4_DRAW_INDX_2_BIN
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7f2b7eae7616c1cb563c16f1e779dbf2c17f7d90
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Mon Nov 23 14:16:13 2009 -0700
msm_kgsl: Add graphics MMU support
Add MMU support to kgsl driver to use vmalloc and a kernel
configuration to turn MMU on/off.
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9806376ee5663736d1cbffd33899708a0f868aec
Author: Mahesh Vojjala <c_mv...@quicinc.com>
Date: Tue Dec 8 19:44:05 2009 -0800
kgsl: Change sequence of clock calls to fix boot up issue on some 7x30s.
MSM 7x30 has a HW bug and needs the clocks turned on/off before the
power rails.
Signed-off-by: Mahesh Vojjala <c_mv...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 50a57bc49f8cf96febd9fd91a81057eb1e1bd2c8
Author: Shubh Das <sa...@quicinc.com>
Date: Mon Nov 30 13:39:36 2009 -0700
msm_kgsl: Properly restore the bin base offset on context switches
The correct bin base offset is set for the context whenever it
changes after the binning pass so that it can be restored on
context switch.
Signed-off-by: Simon Wilson <swi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b4b971bd79bb25f5507980ad45caa2dee539e877
Author: Lucille Sylvester <lsyl...@quicinc.com>
Date: Mon Dec 7 11:41:00 2009 -0700
msm: kgsl: Disable IRQ when GPU is idle
Disabling our interrupt check when the GPU is known to be idle allows
system wide idle-power-collapse to take place.
Change-Id: I5f94c8565270193ac49090618d14d3b198ad7605
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7545533f12e74a62b837e70955cb3561147fb7dd
Author: Sylvester, Lucille <lsyl...@quicinc.com>
Date: Wed Nov 25 14:36:17 2009 -0700
msm: kgsl: Increase idle timeout counter
Idle timeout errors are occassionally observed upon initialization.
Increasing the timer allows initialization to proceed and doesn't impact
standard usage.
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a2433d3d32afa2379f476ecd9f4929c6075ed00a
Author: Sylvester, Lucille <lsyl...@quicinc.com>
Date: Wed Nov 25 14:54:21 2009 -0700
msm: kgsl: Remove AXI vote when the GPU is idle
Include the AXI vote with other clock changes regulated by
suspend/resume and GPU idle.
CRs-fixed: 218129
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 26fbf3807db6c1590f1ada328c4dd155c1eb1950
Author: Shubh Das <sa...@quicinc.com>
Date: Tue Nov 24 12:32:15 2009 -0700
msm_kgsl: save gmem based on correct use of flags
Corrects the way we save gmem. Renderboy informs kgsl via a flag
when the gmem should be saved by kgsl on a context switch and when
it does not need to be saved.
Signed-off-by: Simon Wilson <swi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8f62211e7eed61886baf1e16e65504997555fd48
Author: Naomi Luis <nl...@quicinc.com>
Date: Mon Nov 9 14:59:46 2009 -0800
msm_kgsl: Add multi-context and cmdstream handling fixes
Add multi-context and cmdstream handling fixes, and register write
changes
Signed-off-by: Naomi Luis <nl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7b9398950c440ae7288ff0c709b28318c2867c95
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Mon Oct 5 15:41:46 2009 -0600
msm_kgsl: Put back the file descriptor in kgsl_gem_obj_addr
kgsl_gem_obj_addr was not putting back the file structure it got
with fget(), so the DRM device was never releasing.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 17f30777f9d2bedc74ff8480c2af0e32ae96a710
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Thu Sep 24 14:20:11 2009 -0400
msm_fb: char: enable MDP blit from GEM
From: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b6b406bea0aa0da716fbc86a3d9001b2a49844a3
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Sep 16 14:50:57 2009 -0700
drm: msm_kgsl: memory allocation changes
drm_calloc, drm_alloc, drm_free have been removed in
2.6.31, replace them with kzalloc, kfree
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 876cf3664eb0be472c1ad56baa6fe201efe57c28
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Sep 16 13:03:19 2009 -0700
msm_kgsl: Add GEM support to the DRM device
Add GEM support to the msm_kgsl DRM device.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f83800dc27cf5e7a327dae83c917911b7c4cde51
Author: Jeff Boody <jbo...@quicinc.com>
Date: Wed Oct 7 21:43:16 2009 -0600
msm_kgsl: Fix a double free and prevent errors when freeing null gpuaddr
CRs-Fixed: 181367
Signed-off-by: Jeff Boody <jbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e62abb880752772d552dae555d2c3661df114096
Author: Sylvester, Lucille <lsyl...@quicinc.com>
Date: Fri Sep 18 17:19:09 2009 -0600
msm: kgsl: Power savings through GPU clock control
- Turn off all GPU clocks during suspend/resume
- When the GPU goes idle turn the clocks off until new
commands are issued. This change allows TCXO shutdown to happen
when apps are closed but the processes remain alive.
CRs-fixed: 189566
Signed-off-by: Lucille Sylvester <lsyl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 190e48a76b73f130a89b9fc3c21e07d3fbcb40fa
Author: Shubhraprakash Das <sa...@quicinc.com>
Date: Thu Sep 24 11:40:29 2009 -0600
msm_kgsl: Multi context bug fix
Before this fix when a process would close its handle to kgsl driver,
it used to destroy all contexts including ones not owned by it. This
fix ensures that a process only destroys contexts it created/owned.
Signed-off-by: Shubhraprakash Das <sa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 39abeddc84e589951ee0c90e29135ff26fab7e98
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Thu Sep 3 09:39:52 2009 -0600
msm_kgsl: Add GEM support to the DRM device
Add GEM support to the msm_kgsl DRM device.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 894364350dcc381b6535b663b68a86f85fd88bb2
Author: c_od...@quicinc.com <c_od...@quicinc.com>
Date: Fri Aug 28 15:22:38 2009 -0700
msm_kgsl: Bypass SQ RTR for host accesses.
Host access are blocked based on status of SQ block
this may cause 3ms delay on the bus to access a register.
This delay can lock out GSM and cause fatal error in MDSP.
So, bypass SQ RTR for host register access.
Signed-off-by: Omprakash Dhyade <c_od...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1d5f5f072434485647d9a55defcb27adb584a48e
Author: c_od...@quicinc.com <c_od...@quicinc.com>
Date: Wed Aug 26 21:47:20 2009 -0700
msm_kgsl: Vote for maximum AXI bus speed
AXI bus doesn't always run at the highest possible frequency.
Hence vote for the highest AXI frequency when the graphics core
is in use and and remove the vote when it's not in use.
Also making sure the AXI vote is dropped during suspend state
to avoid unnecessarily running the AXI at max frequency.
Signed-off-by: Omprakash Dhyade <c_od...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit da65ac72a90638b025306dcc0853562fade72d15
Author: Simon Wilson <swi...@quicinc.com>
Date: Mon Jun 29 12:45:00 2009 -0600
msm_kgsl: report rev 2.1.1 instead of 2.1 due to hardware bug
Hardware rev 2.1.1 is incorrectly reported as 2.1 when reading
the chip ID register. This causes triangle strips to be made
larger by the usermode driver and thus reduces triangle rate
in applications such as Powerlift. The chip ID is changed back
to 2.1.1 to avoid this.
CRs-fixed: 187187
Signed-off-by: Simon Wilson <swi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 60d2984f7ee9f37a8a64defa160f08cdd40a42c6
Author: Omprakash Dhyade <c_od...@quicinc.com>
Date: Tue Jun 23 13:32:21 2009 -0700
msm_kgsl: add grp_pclk handling
msm7x27 has an additional "grp_pclk". Add appropriate functionality to
manage this clock.
Signed-off-by: Omprakash Dhyade <c_od...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 764e9ca21e8afa69c2d9dc148ad0167e3567fb4d
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Tue Jun 2 10:26:08 2009 -0600
msm_kgsl: Add DRM support
Add support for msm_kgsl to register as a DRM device.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1e8601ee3e5df4fc73ef90ba722fbc6931b89851
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Wed Jan 27 10:05:27 2010 -0800
drivers: char: msm video drivers
this add the kgsl driver, and the v6 encoder/decoder
parts from the Initial Contribution.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit c23544863669f580e3667506eba1a3ff83d61761
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Wed Sep 2 11:07:35 2009 -0700
mm: dma_cache operations for non linearly mapped buffers
dma_cache_*_ops() perform cache maintenance operations on linearly and
non-linearly mapped buffers.
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4098dc3067142c859c7fa892a3952ae3f39948b7
Author: David Brown <dav...@quicinc.com>
Date: Fri Jan 29 13:36:40 2010 -0800
diag: Qualcomm SOC diagnostics driver.
Character driver to make diagnostic messages from modem processor
available within Linux.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1c8d5f7d71edcdbb3a0d5b50db58ed12ab72b430
Author: David Brown <dav...@quicinc.com>
Date: Fri Jan 29 13:30:20 2010 -0800
defconfig: Enable EHCI_ROOT_HUB_TT
Signed-off-by: David Brown <dav...@quicinc.com>
commit a392039cd93dc4daff243323c82596d4cabcd050
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Tue Jan 26 01:17:21 2010 +0530
usb: msm_otg: sysfs support to switch between host & peripheral
Adding support for switching between host & peripheral through sysfs.
To switch to host mode, after plugging in the A cable, type in below
echo host > /sys/devices/platform/msm_otg/mode
To switch to device mode, after plugging in the B cable, type in below
echo peripheral > /sys/devices/platform/msm_otg/mode
To check, Which mode user is currently operating in, type in below
cat /sys/devices/platform/msm_otg/mode
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 90cf773b0cacfc2dd41eed0eec72895f1f499fe5
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Jan 28 15:49:19 2010 -0500
net: add support to supress random MAC address generation in smsc911x
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 12337b1700529d933a79d8ea346ac9fdf8b8ac3a
Author: Brent DeGraaf <bdeg...@quicinc.com>
Date: Tue Jan 19 16:47:20 2010 -0500
mmc: Handle requests properly when block queue is plugged
If the block queue becomes plugged (usually under heavy loading), an old
request pointer will be used, possibly leading to corrupted data on the
card. This change properly initializes the request pointer at every
iteration so that a plugged block queue will be properly handled.
Signed-off-by: Brent DeGraaf <bdeg...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b8cd29418fccdfb280dd4a219d490a40ee0bf964
Author: Sahitya Tummala <stum...@qualcomm.com>
Date: Fri Jan 15 11:28:55 2010 +0530
mmc: msm_sdcc: reset the controller in case of data CRC error
Perform a SW reset of the controller if a data CRC error is detected
during the data transfer. This ensures that the data path state machine
is clean for the next data trasfer.
Signed-off-by: Sahitya Tummala <stum...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2b5caf55fe4f1c624a3c253938725ebbc4027b7f
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue Jan 26 17:29:06 2010 -0500
msm: put ST1.5 framebuffer back in SMI
No longer needed in DDR now that SMI is running at proper speed
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4402b7676ec4a0f8164fb677ff92127d8313d3d4
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 28 14:20:09 2010 -0800
defconfig: Enable USB OTG
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2ebab3957a15176ed0418a17f85443b14936bb7d
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 28 14:17:12 2010 -0800
USB: MSM OTG support.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 35b903dd7880106a24c1c2c690e2340158ec5a87
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 28 13:33:48 2010 -0800
msm: Update hsusb RPC
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 847f3410fbb07b982cf5913a9dcd04577da7b54e
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 28 10:07:06 2010 -0800
defconfig: Enable Qualcomm EHCI device on st1.5
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e96ae18ca217d0d9210c73251718bad56975cf7c
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 28 10:03:02 2010 -0800
USB: EHCI: Qualcomm MSM USB EHCI driver.
Add support for the EHCI device in Qualcomm MSM SOCs.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 167274a72dd329901ccf1fcb0b7216e59d4e8441
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 28 10:14:57 2010 -0800
msm: usb: Add rpc calls for USB control.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 070f235015990358e468ff01e558469620b39a87
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Wed Jan 27 09:45:19 2010 -0800
msm: qsd8x50 defconfig: enable audio
Enable qdsp, and alsa support.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit d3bc6fcd39a37559fe44078971db5034ee2160fe
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Tue Jan 26 16:44:15 2010 -0800
sound: soc: msm: a few updates for 2.6.31
codec got pushed into the card structure.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 1ad7e93931f04062e652150719f199ad9d56c1c1
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Thu Dec 24 16:36:45 2009 +0530
alsa: soc: 8x50: new framework with dual interrupt sources
Added timer interrupt source to handle non periodic interrupts from DSP
for data done event. The driver now depends on two interrupt sources to
sync the hw_ptr.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 208229ed19f146f1c4877d67d2f146f4dadcb1b4
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Wed Dec 23 08:57:56 2009 +0530
alsa: mixer: snd device dynamic frequency change support
The Tx devices can support dynamic frequency change.
Added support to poll for accepted frequency by voice and audio sessions.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 78948381c4e01f4a10903a14649b784730c33714
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon Dec 14 17:11:18 2009 +0530
alsa: soc: msm7x30: ASoc driver for pcm playback/capture
The driver plays pcm audio through HPCM interface and records through
AUDREC interface. Supports copy operation.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7be5c8f774b907dcf2745afd01850c77483e087c
Author: Patrick Lai <pl...@quicinc.com>
Date: Wed Dec 2 19:16:07 2009 -0800
audio: 7x30: Increase sound device list size
There will be more sound devices coming as list of 7x30 audio
features grows. Currently, the list can only accomodate up to 16
sound devices which would not be enough in near future. Update
device list size macro to 64
Signed-off-by: Patrick Lai <pl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2e8395c0a70db37b9437420e22697c669701c419
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Thu Dec 3 13:17:25 2009 +0530
audio: 7x30: Add listener code to mixer interface
All decoder/encoder and voice drivers registers to this listener and
would be informed of any event changes like routing or device
enable/disable events.
The drivers are then required to use this to take appropriate action
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a4efb03968c955c53ac407f31b1c9b64b59a8de9
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Sat Dec 5 17:45:36 2009 +0530
alsa: soc: msm7x30: Fix for blocked voice call thread
The fix wakes up the kernel voice thread after device is enabled.
The voice thread waits on wait queue which should wake up when the
device is enabled else it waits for 30 sec to timeout.This results in
delay in voice call even when the device is already enabled.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ed5b0e244f4ada4fe11d97a44734b1406405d9bf
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Fri Nov 20 17:59:30 2009 +0530
alsa: soc: Add support for multisession playbacks.
Add subdevice support to support multisession playback.
It now supports 4 playback streams and one record stream.
CRs-Fixed: 217559,217847,217863
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 79eb880030625e3835290c88219049cf56ed1bcc
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Fri Nov 20 17:58:54 2009 +0530
alsa: soc: 8k: Fix to avoid I/O error due to early stream start
CAD does not support copying buffer before the stream is started.
This results in race condition as ALSA receives interrupt before trigger
is called. ALSA expects to do stream start after copying atleast one
buffer.
CRs-Fixed: 214911
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7ee849a66feb7f3b5512d8c9eb10035d11d77708
Author: Sriranjan Srikantam <c_ss...@qualcomm.com>
Date: Fri Nov 20 23:36:37 2009 +0530
msm: audio: 7x30: Update the audpp commands with correct parameters
Update the pause, flush and post processing commands with correct
parameters according to the new TIS specification. Update the pcm and
alsa mixer driver set volume commands according to new interface.
Signed-off-by: Sriranjan Srikantam <c_ss...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3a982f6b908ae1eb8e0d5dbe9fb0bef63dafb597
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Sat Nov 21 08:43:21 2009 +0530
alsa: 7x30: Fix the extra new line added to device name.
The current mixer code wrongly adds newline to all audio device names.
This is unwanted and removed.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dbdf12af47017548a2a812ee408671815af81c68
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Wed Nov 18 19:54:34 2009 +0530
alsa: soc: msm: routing amixer controls for audio playback/voice
The driver is designed to do routing configuration for audio
playback and voice path. This also is responsible for enabling/disabling
audio endpoints.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1bf33bef478ff70b6d792718f49d6231956ccd34
Author: Stephen Boyd <sb...@quicinc.com>
Date: Thu Jul 23 15:48:00 2009 -0700
pmem: fix dependencies when disabling pmem
MSM sound, graphics, and camera drivers all depend on PMEM being
enabled. If pmem is disabled, these drivers should not be
included.
Also fix some warnings:
drivers/video/msm/mdp_ppp.c:1141: warning: unused variable 'vstart'
drivers/video/msm/mdp_ppp.c:501: warning: 'get_len' defined but not used
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d5485f6095bcc32c474b652e42bd64f29e6f365f
Author: Ben Romberger <brom...@quicinc.com>
Date: Wed Sep 16 10:12:36 2009 -0700
msm: audio: 8k: Q6 API change
Q6 open structure, command structures & format blocks changed.
Format block is no longer stored in shared memory.
Signed-off-by: Ben Romberger <brom...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1e9af67f46ad35fd24de65d22adb50f2934afcbc
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Thu Oct 1 03:16:17 2009 +0530
alsa: soc: Fix to include correct header file.
Some part of alsa code is shared between different architectures.
The code is not appropriately including the correct header files
based on the architecture.The fix puts a check to architecture and
includes correct headers.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 201799eac6f3fe7e3bc019adc864b53b4e0c2b08
Author: Stephen Boyd <sb...@quicinc.com>
Date: Fri Jul 17 11:45:22 2009 -0700
sound: msm-soc: add missing newlines to some printk()'s
output from dmesg:
<3>[ 1.599748] msm_soc:ALSA MSM Mixer Setting<3>mmc1: Command
CRC error
Also modernize a bit with pr_err().
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 75ac216a399355bb9f4db695cdcec9373b48ba7c
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Wed Aug 26 17:56:52 2009 +0530
alsa: soc: 7k: Fix to ensure graceful exit if not yet enabled.
If the application calls close immediately after calling open,
the driver needs to ignore wait for EOS. This might happen if
the pcm sample header has some issue and needs to be prematurely
closed.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e2667ee54a15d2abe67f33561d0f77887d2729df
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon Aug 24 12:33:32 2009 +0530
alsa: soc: 8k: Fix to ensure graceful exit if not yet enabled.
If the application calls close immediately after calling open, the driver
needs to ignore wait for EOS. This might happen if the pcm sample header
has some issue and needs to be prematurely closed.
CRs-Fixed: 193613
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b15160b6d00d375a768a5ba07dcee34790360f5b
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Thu Aug 20 15:42:59 2009 +0530
alsa: soc: 8k: Fix underrun errors in alsa concurrent playback record.
The first underrun occurs due to the volume update forcefully done
for stream even if there is no change from the default. Removed this
forceful update of volume.
Then onwards the underrun is triggered because in prepare routine
pcm_irq_pos is not reset in the second instance of the function call.
This results in a continuous mismatch in the hw/sw pointers and they
result in repeated underrun errors. Resetng the pcm_irq_pos solves the
issue.
CRs-Fixed: 192951
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a15355bed397f880f0538f81b3899f3dc9454788
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Wed Aug 19 10:35:01 2009 +0530
alsa: soc: 8k: Remove unwanted check of device list support.
Alsa driver is currently checking a list of device and maintains the
device list.This is not required as cad layer now caches the device list.
CRs-Fixed: 193176
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 14860a0fdf90ac8df3ddab6e9581ebc7c96b49cd
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon Aug 10 12:09:51 2009 +0530
alsa: soc: 8k: Initialize default values to volume and device settings
Defaults are set currently to 0 which results in the alsactl to fail as
the device number 0 is not supported by cad.
CRs-Fixed: 192012
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3c582b78f88373438ed05276c78d534c78d9208b
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon Jul 27 15:38:12 2009 +0530
alsa: soc: 7k: Add EOS support in playback
The current implementation tries to close the session without confirming
the actual playback of complete data. Adding support to ensure
complete playback completion before termination.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 98a1aa5da0a799e94bbe4b2c4150ca56d67daf75
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon Jul 27 14:45:14 2009 +0530
alsa: soc: 8k: Add EOS support in playback
The current implementation tries to close the session without confirming
the actual playback of complete data. Adding support to ensure
complete playback completion before termination.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e77287aada9c034d337037b64539a6ec5f247555
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Fri Jul 24 11:55:38 2009 +0530
alsa: soc: 8k: Remove stream volume update for TX path.
The current q6 does not yet support the tx stream volume update. This
may result in crash scenario if the user tries to update the stream
volume.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4e2efdd2324577739969f7d02aa236cf67b7a19c
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Thu Jul 16 18:18:39 2009 +0530
alsa: soc: 8k: Add device volume control support for RX/TX path.
Current implementation has a volume control to RX path only.
Split this to support volume control for both RX/TX path.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 62973646cc6aacd10eead8c878d1262ed2fd97c5
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Thu Jul 2 11:59:48 2009 +0530
alsa: soc: 7k: Fix to max dB value permissible for volume control.
The command that we send to dsp for the DB value is crossing the
16 bit (unsigned short) boundary for the values 17 and 18.
The max dB permissible value is hence 16.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a91ea421c34b922521923064c95ccac7abe76c84
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Tue Jun 30 20:39:26 2009 +0530
alsa: soc: 7k: Fix to initial bit stream configuration passed to DSP.
Problem Description:
Current code sends first two buffer size as configuration rather than
the valid sample count as expected by DSP.This results in corrupted
data being played in case the sample count is less than the buffer size.
Also the audio playback needs to be started after first data copy is
done.To do this we need to send the second buffer sample count as zero.
But the applictaion tries to write to second buffer thus breaking the
sync with DSP.
Solution
The first buffer is kept with zero sample count instead of the second
and sending actual bitstream starting from second buffer thus writing
zero sample size to first.This helps in maintaining the sync with DSP
and also to enable dsp with only one buffer copy.Sending actual sample
count solves the glitch due to playback of corrupted data.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6e1acb0b4e29e81b3ec7487e7e46c70ce06d4695
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Wed Jun 24 18:15:34 2009 +0530
alsa: soc: 8k: Ensure that buffer size is a multiple of period size
In case alsa application does not give both buffer-size or buffer-time
we need to make sure from the driver to always used buffer-size as
multiple of period-size.For this the driver adds a rule to enforce this.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e8adfa08129bd83ce7919ee6c7cab76e54faab22
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Wed Jun 24 18:09:25 2009 +0530
alsa: soc: 7k: Ensure that buffer size is a multiple of period size
In case alsa application does not give both buffer-size or buffer-time
we need to make sure from the driver to always used buffer-size as
multiple of period-size.For this the driver adds a rule to enforce this.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 431d2c67acddd4b74192a42323657943b6b5a2c9
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Fri Jun 19 12:01:20 2009 +0530
alsa: soc: 8k: Modify MUTE control to support playback/Capture path
The cad code has recently added support to provide mute control for
playback/Capture path seperately. Support for amixer to handle mute for
playback/Capture.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c3e50f38df465c089d3b70c230f08821e6df9b69
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Fri Jun 19 11:48:02 2009 +0530
alsa: soc: 8k: Add volume caching support.
cad volume interface now supports caching of last set volume for
currently running stream. Update alsa accordingly.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 110deb7974f2beb4aee37165d08fe34d1f718bfe
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Tue Jun 16 13:58:37 2009 +0530
alsa: soc: 8k: Fix to alsa playback and capture concurrency
The runtime is maintained specific to playback and capture stream and
hence should be initialised based on the stream direction.The current
implementation overwrites the runtime by the last open session without
the check for playback/capture direction.
CRs-Fixed: 185996
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ac66eb10ee529361511c9ebb575098ee47764529
Author: Michael Bohan <mbo...@quicinc.com>
Date: Thu May 28 10:10:06 2009 -0700
[ARM]: msm: Apply new MSM naming scheme
The current naming scheme consisting of ARCH_MSM and ARCH_QSD no
longer properly describes our targets. We have new targets that
are functionally hybrids between 7k and 8k, and as such we need a
finer level of granularity to express such nuances.
The new scheme consists of four different levels. At the top is
ARCH_MSM, which now encapsulates all other levels.
The next three levels are user-defined. The first is the
"CPU Subsystem" (ARCH_MSM_*), which specifies which peripheral block
(clock control, timers, interrupt controller, etc.) one requires.
There may only be one SS configured at a time. It is selected
implicitly when selecting the SoC.
The SoC represents a CPU core. The SoC may have subtle differences
between other SoCs that use the same SS, and can be expressed
with preprocessor conditions. There may only be one SoC
configured at a time. Despite the rather specific SoC names (eg.
ARCH_MSM7X01A), it is possible for a SoC to describe multiple
targets (eg. MSM7201A, MSM7X25).
The last level is the "Machine" specification. They depend on the
SoC that has been chosen. Any number of machines can be selected
at once, assuming they correspond to the same SoC.
example:
ARCH_MSM -> ARCH_MSM_SCORPION -> ARCH_MSM7X30 -> MACH_MSM7X30_SURF
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ec24465e8a8b93f2da81b7b370acebf3884d24f6
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon Jun 8 10:31:17 2009 +0530
alsa: soc: 8k: Fix to stream volume control
The stream volume control ignored volume gain level zero.
Comparison to wrong variable resulted in the bug.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 61ec1aa984d5be983663880246f0d625ce9c072e
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Tue May 26 19:02:27 2009 +0530
alsa: soc: 7k: Fix to connect snd endpoint once during init
Current non optimal implementation connects, calls and
closes the RPC connection everytime snd_msm_device_put is called.
Now the driver connect once during init then use the same endpoint
for all the following operations and close the connection in exit.
msm_rpc_call_reply is replaced with msm_rpc_call as no response is
expected. Correction made to endianness of parameters to this function.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7f83f82c114137551d037c073fd96a81790736d5
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon May 18 17:59:58 2009 +0530
alsa: soc: Support continuous range of sampling rates in qsd8k
Support continuous range of sampling rates for both playback and record
in qsd8k alsa.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b5a3dfe11a0f552195fcd1788b24f623ac1e0c56
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon May 18 16:28:27 2009 +0530
alsa: soc: Port alsa 8k driver to kernel-2.6.29
Port qsd8k alsa driver to 2.6.29.
CRs-Fixed: 182615
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cb7767208acdd0d09c9c1ca0193aa7cef3b75e34
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Mon May 18 11:00:42 2009 +0530
alsa: soc: Port alsa driver to kernel-2.6.29
Port msm7k alsa driver to 2.6.29.
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8ba6dcefd3cab9b46f86e5a9205448f653357cb7
Author: Asish Bhattacharya <asi...@qualcomm.com>
Date: Tue May 12 12:45:35 2009 +0530
alsa: soc: Remove pause before closing cad session.
Stream PAUSE before closing cad session is redundant. Hence removing cad
ioctl call of STREAM PAUSE before closing cad session.
CRs-Fixed: 181219
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 66ddc3e3513cd69e39b0e11c658331138696e673
Author: Peter Lohmann <ploh...@quicinc.com>
Date: Tue May 12 11:15:01 2009 -0700
msm8k_audio: Use a mute_info struct to pass path and mute info.
Signed-off-by: Peter Lohmann <ploh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d796122bde83cce81620bc079b07bd43e7aa6bf4
Author: Bhattacharya, Asish <asi...@qualcomm.com>
Date: Wed May 6 09:18:16 2009 +0530
alsa: soc: Fix the uninitialized usage of mutex
Signed-off-by: Asish Bhattacharya <asi...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ad91dc6236632b0d20f85bf5fbf555f9efb54258
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Tue Jan 26 11:06:05 2010 -0800
sound: soc: add msm alsa support
Parts of alsa support in the Initial Contribution.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 731ebf6ae8418af9204ce9916d9896dfd46d8fe1
Author: Varun Colbert <c_vc...@quicinc.com>
Date: Tue Jan 19 15:45:11 2010 -0800
qdsp6: Replace existing 3.1 audio drivers with 3.2 CAD drivers
Signed-off-by: Varun Colbert <c_vc...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2aad690a7fbe66ca60474ac49cedb0f839e22b08
Author: Varun Colbert <c_vc...@quicinc.com>
Date: Tue Jan 19 15:29:27 2010 -0800
qdsp6: Replace 720P 2.0 video encoder with 1.0 driver
Signed-off-by: Varun Colbert <c_vc...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8fb42566c36485793049d3c961d3f0505e8c9201
Author: Varun Colbert <c_vc...@quicinc.com>
Date: Tue Jan 19 15:24:44 2010 -0800
qdsp6: Replace existing 3.0 VDEC drivers with 2.0 VDEC drivers
Signed-off-by: Varun Colbert <c_vc...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 69125c3986a7ee34c6baf9519ea4ef235fc1acdc
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Tue Jan 26 10:55:43 2010 -0800
msm: audio: enable audio in the makefile
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 8ffe1b42054d9ef96818a57c7fa24b2b0bf799e7
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Tue Jan 26 10:56:12 2010 -0800
add msm_audio_aac.h
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 51dec5ff44dd37e7da03d5160fe21c7c7993793b
Author: David Brown <dav...@quicinc.com>
Date: Wed Jan 27 16:35:28 2010 -0800
defconfig: autoupdate
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 84381e6123fc1461d4409ad37d0bd88779459d02
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Dec 11 17:32:22 2009 -0800
msm: Set default QSD8x50A PHYS_OFFSET Kconfig.
On 8x50A, the kernel is located at a physical offset of 0x00000000.
Change-Id: I71182feff96a66892a69b1d3ba94b287665c3fc7
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f09347ef1e6692ca599155f923f783be1dd218c1
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Nov 25 11:00:31 2009 -0800
msm: socinfo: Add SoC IDs and cpu_is_qsd8x50a() function for QSD8x50A.
SoC IDs are added for QSD8250A (ID 64) and QSD8650A (ID 65).
Change-Id: If61e8f214672f0228739eb5ecd35e6cf4be69607
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 355dd780e0eb257b856c4740419a2afbb9339762
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Dec 17 14:02:11 2009 -0800
msm: Makefile.boot: Add support for QSD8x50A.
Change-Id: I8194c686efd9ebeb6a8ad2eaad360cfa694b1189
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b20853251c6853de54b40a7a3b37b4eaeec11e3f
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Dec 16 19:12:12 2009 -0800
msm: Makefile.boot: Select linker parameters based ARCH_* Kconfigs.
Linker parameters are now determined by the ARCH_* Kconfigs, rather
than by their CPU type and whether or not their memory is stacked.
This makes the file more readable while also adding the flexibility
to handle arbitrary memory maps for specific architectures.
Change-Id: Id8f4654e2bb8d7bcca1d97361c85ff0faf03671f
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8b5d402c37ad079c983f827f354d15d26a3ca562
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Nov 25 12:18:45 2009 -0800
msm: Select kernel PHYS_OFFSET based on ARCH_* Kconfigs.
Using Kconfigs to select the kernel's physical base address is
cleaner than using ifdef's in mach/memory.h.
Change-Id: I7856861786e10996d5e19472a4f066a23fd0c538
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 35245f466569ad6799a953715c3aa66940e52d0c
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 26 14:39:41 2010 -0800
defconfig: Enable switch class for ST1.0.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8e497510dba12940f13d88e42faaac9e52bcf04b
Author: Anirudh Ghayal <agh...@qualcomm.com>
Date: Thu Nov 26 18:37:44 2009 +0530
msm: handset: Add support to report headset status to modem
Add rpc call to indicate the status of headset (connect/disconnect) to
the modem processor. This is required when the headset detection logic
is present on the apps processor.
Signed-off-by: Anirudh Ghayal <agh...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit daf926c2c7e44aa6d8465ff572f13889665f03f0
Author: Dennis Taul <dt...@quicinc.com>
Date: Thu Oct 1 14:17:24 2009 -0400
switch: port switch class and GPIO drivers from 2.6.29.
Signed-off-by: Dennis Taul <dt...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit af62b40fa35c676433b5c97987b6c8f3456a1c2a
Author: Trilok Soni <ts...@qualcomm.com>
Date: Wed Jun 10 11:25:28 2009 +0530
[ARM] msm: handset: Merge msm_handset driver into rpc_server_handset
msm_handset driver introduces un-necessary reverse
dependancy due to msm_get_handset_input_dev() API and acts
only as dummy driver to which rpc_server_handset was pushing
the events. Due to this reverse dependancy msm_handset driver is
not able to work as loadable module.
Signed-off-by: Trilok Soni <ts...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 44458c5d3566b779a6c70c6020302f9da660ca7e
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 26 14:27:40 2010 -0800
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e37a8a0baf839972a74ce057f821afb67538351c
Author: Ashwin Chaugule <ash...@quicinc.com>
Date: Mon Jan 18 17:42:20 2010 -0500
defconfig: Disable AVS and update default ACPU voltage to 1300mV for ST1
AVS is disabled for stability reasons since it incorrectly steps down
the voltage at 1GHz leading to dcache parity errors. Additionally, the
ACPU voltage is updated to match what's currently used by acpuclock-8x50
driver.
Change-Id: I9d58bb30425e883b228eac9470736fe453a0fdb8
Signed-off-by: Ashwin Chaugule <ash...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f0d991766e0fb85e7d1bd497222f01248d7303cd
Author: Saravana Kannan <ska...@quicinc.com>
Date: Sat Jan 23 17:43:21 2010 -0800
defconfig: Choose the correct LCDC panel for ST 1.5.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1078a52aa5ac1b74368b2fc5852fef6a5dffe29d
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 26 09:42:38 2010 -0800
defconfig: automatic update
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d97038b5206a06201667f7c7087f1f410e013734
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Jan 22 15:29:59 2010 -0800
defconfig: Enable CPUFreq for ST 1.5.
The 'Performance' governor is enabled by default.
Change-Id: I09e455b4642d475a4b1ef0801809e9205874fa9f
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4312e8bb4181df4c31601866ae1ae362c47d5004
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Jan 18 13:01:04 2010 -0800
msm: acpuclock-8x50a: Add QSD8x50A ACPU frequency scaling driver.
QSD8x50A requires different procedures than QSD8x50 for SCPLL
initialization and SCPLL frequency switching. Hence, a new driver
based on acpuclock-8x50 is introduced.
Change-Id: I83efdcfdb9b596c9be3f4ec96431f7e60dd942c4
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit efbc0a5c8b6fbf1fe4dfc446781ccfc471726081
Author: Chiranjeevi Velempati <cvel...@qualcomm.com>
Date: Fri Jan 22 12:07:12 2010 -0800
msm: configure and set the gpios as per the platform
ST1.5 and ST1 uses different gpio lines for configuring the
analog switch to select the host mode. Hence configuring & setting
the gpio based on the platform.
Change-Id: I28d01c1d4ebb3e5e0beb18f34658a51306f97724
Signed-off-by: Chiranjeevi Velempati <cvel...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 153777142ddd8a7b55095b50aa5ad572ee2f324c
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Jan 21 10:54:50 2010 -0800
msm: board-qsd8x50: Register ST 1.5 I2C devices.
Change-Id: I906dbdad1a8431092345fc40885ba48c9c4f3756
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 776ada57a54b381ae14273f2f973e4b911c12bad
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Jan 14 20:09:39 2010 -0800
msm: board-qsd8x50: Add Ethernet support for ST 1.5.
The physical base address and the IRQ GPIO for the SMSC LAN9221
controller are different on the ST 1.5 and ST 1.0. Change the
board init code to handle this appropriately.
Additionally, ST 1.5 hardware supports use of the controller in
32-bit mode. This is enabled for improved performance.
Change-Id: I1d41f321d8ccbe446e3dbcd84ce694a53971d389
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b3aacda85e04228179f2841b1684fe8732becd18
Author: Sheetal Sahasrabudhe <shee...@quicinc.com>
Date: Fri Jan 22 12:05:18 2010 -0500
KSAPI: Performance monitoring tool for Snapdragon linux
KSAPI records performance statistics for Snapdragon linux platform.
It uses the /proc FS as a means to exchange configuration data and counter
statistics. It can monitor the counter statistics for Scorpion processor
supported hardware performance counters on a per thread basis or AXI
counters on an overall system basis.
Signed-off-by: Sheetal Sahasrabudhe <shee...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c3e49a1525a98023fbf59d571ce19dfc4c43f47e
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Jan 22 00:38:28 2010 -0800
defconfig: Add initial ST 1.5 defconfig.
Change-Id: I59ed2fef07b683ff3af3cfb43c418528081d7369
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 28c5cf7b269e492b864d91fffabe3b99e767d490
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Jan 18 13:24:35 2010 -0800
msm: acpuclock-8x50: Fix ACPU frequency detection when running off AXI.
acpuclk_init() is updated to select the correct frequency table row
when booting with AXI as the ACPU clock source.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e30ef3be41e858176a39c783ef4ca2ee0a56e8fd
Author: Murali Palnati <paln...@qualcomm.com>
Date: Mon Jan 11 09:52:07 2010 +0530
mmc: core: Add support for MMC cards with CSD structure revision 3.
CSD structure version 3 and above are defined by MMC specification
as reserved, to be specific, CSD version is encoded in the EXT_CSD
register. Some manfacturers like SanDisk use this version for some
of their latest MMC 4.3 plus, iNand cards. This patch adds support
for these cards.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b4169fe898d13fff96625c5f24a9b1928b8e402d
Author: Brent DeGraaf <bdeg...@quicinc.com>
Date: Mon Dec 21 15:45:36 2009 -0500
mmc: msm_sdcc: Eliminate forced reset of SD controller port
Forced resets of the SD controller's command path state machine can
cause cards to abort their internal flash maintenance and terminate a
write to flash after the card has indicated the command is complete but
before to the data has actually been fully written. This change
eliminates one such occurance.
Signed-off-by: Brent DeGraaf <bdeg...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 34fa18caf2d923d685f2e7b6b278665ba2c24fbd
Author: Brent DeGraaf <bdeg...@quicinc.com>
Date: Mon Dec 21 15:12:25 2009 -0500
mmc: msm_sdcc: Fix sdcc resume sequence
A missing set of brackets creates an inadvertent logic change to the
msm_sdcc resume sequence if CONFIG_MMC_MSM7X00A_RESUME_IN_WQ is not
configured. The redundancy in the resulting code is likely to be
optimized out by the compiler, but the source code is misleading due to
indentation. This change inserts the missing brackets.
Signed-off-by: Brent DeGraaf <bdeg...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3ee99dca29fbba499ada292eae3dbe0bb933232c
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Jan 18 22:22:44 2010 -0800
msm: Remove unused file 'keypad-surf-ffa.h'.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dd84dc474a496048a0f96db87c7f1e7003e65f3a
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jan 20 17:04:07 2010 -0500
msm: add platform support for ST1.5 display
This includes a temporary move of the framebuffer from SMI
to DDR for ST1.5 only, until SMI timing issues are resolved.
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 755e6cc6775bc0282217a45f1004bfb565f1e5aa
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jan 20 15:34:16 2010 -0500
msm_fb: Add display driver for ST1.5
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5b93e303e9dfdf8de8ba0927b14298e8e882b455
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Jan 15 16:58:58 2010 -0800
msm: Fix Kconfig to prevent building MACH_QSD8X50_ST1 for QSD8x50A.
Change-Id: I334335ceafd2fe9de7767173bba62b97d2ec546f
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e48cc534f046de20545911cdaace8322bdbd986b
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Jan 13 13:04:35 2010 -0700
qsd8x50: Allow the size of the in-kernel PMEM region to be configured
Add a config option to allow the in-kernel EBI1 PMEM region to be adjusted
at build time.
Change-Id: I55b4941636d0dadf14f577b9146c98f56f0beb91
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6addc2121e7d9be640ce43af92eb1676f1ebcfcc
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed Jan 13 23:31:59 2010 -0800
msm: board: Fix SDCC VREG assignments for ST 1.5.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3003ad0df42ca1e7a94f600315c8fe41a36ee536
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Jan 14 14:42:47 2010 -0800
msm: Add board support for ST 1.5.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d0ba95b5d0d272f5feeec7f5af2efcd0638914b2
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Jan 13 09:48:35 2010 -0700
msm_fb: Fix two MDP flags with the same value
Two defines for the MDP flag parameter had the same value, resulting in
confusion.
Change-Id: I0ed322f26d88a379fa8e0352744b5bfdacdf9482
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 99611709f34282b9a2db54487267247f0b17d991
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Jan 5 11:57:43 2010 -0800
msm: Enable MMU corner case fixes only for 8x50 and not 8x50a.
The fixes put in to handle MMU corner case issues on 8x50 are
not necessary for 8x50a as they have been fixed in the new h/w.
Change-Id: Ia7ead9e1d604d2c7bebed844bb7d121114793655
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3dabbe2658b3cac07bac255268428ede2db45e48
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Dec 16 16:56:46 2009 -0800
msm: Introduce QSD8x50A Kconfig options.
MACH_QSD8X50A_SURF and MACH_QSD8X50A_FFA are introduced. These
options are configured to depend on MSM_SOC_REV_A, which is added by
this patch to distinguish between 8x50 and 8x50A SoCs. Similarly, the
MACH_QSD8X50_* options are modified to depend on MSM_SOC_REV_NONE.
Change-Id: I4599d69703e0c532e00219d88c4e029237c3dea6
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d651b7e38cb3495d317d69dbef3ae30eca55563d
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Dec 11 17:11:13 2009 -0800
msm: Kconfig: Add SoC revision config parameter.
Some SoCs are completely or mostly backwards compatible with others,
but add new features. It is sometimes necessary to distinguish
between these SoCs at compile time.
Introducing new ARCH_MSM/QSD_* Kconfigs for all similar SoCs would
result in excessively long #ifdef expressions. Instead, MSM_SOC_REV_*
Kconfigs are added to distinguish between them.
Change-Id: Id0ca01bb83f71e9d707fce630732fd7f8363bf38
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5f53295498a5831ae475c39562a98d636c2673fe
Author: Murali Palnati <paln...@qualcomm.com>
Date: Mon Jan 11 16:26:51 2010 +0530
msm: board-qsd8x50: Modify the check for Write Protect switch support.
Support for Write Protection switch detection using FPGA registers is
applicable only for SURFs. For FFA and other hardware this support is
not applicable. This patch fixes the condition check and enables this
support only for SURFs and prevents this support from unintentionally
getting enabled for other platforms.
Signed-off-by: Murali Palnati <paln...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 63102e11bbbbbd5b324840f74104d74f17c3838a
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Fri Jan 8 10:58:22 2010 -0800
msm_fb: Add configuration of background transparent color key
Configure background transparent color key registers if
background color transparency is enabled.
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2aa824f1824f301e2a2ee9c3b65f9dc0bf9f3f64
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Mon Dec 28 16:11:54 2009 -0800
msm_fb: align height of 64X32 video tile format to 32 instead of 64
CRs-fixed: 222165
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f6d28806406477eedcd818d9e63bf4b0f6c97155
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Jan 4 10:59:04 2010 -0800
msm: board-qsd8x50: Support shared memory (SMI) for QSD8x50A.
The SMI RAM on the QSD8x50 and QSD8x50A are located at different
physical addresses. Select the appropriate one based on the
MSM_SOC_REV_A Kconfig.
Change-Id: I72cdc86494bc5ec198f69a2c5a4533b275f4d336
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a28c540e614347acaccc19b4d2f6850b4c955d0f
Author: Naveen Ramaraj <nram...@quicinc.com>
Date: Thu Jan 7 11:06:05 2010 -0800
pmem: Fix Bitmap allocator to allocate the final bit for the device
Fix an off by one error that prevents the last free bit from being
allocated by removing the condition that the end bit cannot align
with the total_bits(n). When the entire device size is allocated in
one shot, this ensures that bits 0 to n-1 are set. The code already
prevents any overflow errors by enforcing free_quanta > 0.
Signed-off-by: Naveen Ramaraj <nram...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ca22c9feefb8835fc08d67d914033a0ecdb64271
Author: Bryan Huntsman <bry...@quicinc.com>
Date: Wed Jan 6 19:22:59 2010 -0800
PM: NPA should be off by default
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eb1cdc0ef89b0b9259592178838180eb6b57c9a8
Author: Praveen Chidambaram <pchi...@quicinc.com>
Date: Mon Sep 21 09:41:06 2009 -0600
add Node Power Architecture (NPA) support
NPA is a software framework for effectively managing resources and
requests for resource states. NPA enables better power management
while meeting client QoS requirements. NPA provides mechanisms to
handle concurrent state requests effectively and notifications upon
state change.
Signed-off-by: Praveen Chidambaram <pchi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ba77ec3646efa9a6ae9369a1223efe190d21e086
Author: Naveen Ramaraj <nram...@quicinc.com>
Date: Tue Jan 5 17:36:01 2010 -0800
pmem: Add In-Kernel API for handling cache operations
Add a new API to pmem that will allow any kernel components that
use PMEM memory to perform cache operations. Some of the drivers
may need to perform cache operations from kernel space to ensure
correctness. This new API will perform such required operations
by abstracting away the underlying device memory type. The new
API's signature will be similar to the existing pmem_ioctl exposed
to user space. The ioctl has been refactored to call the new API.
Signed-off-by: Naveen Ramaraj <nram...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 775413ce7d65c84949893c2bc3a811cbbc0344fa
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Jan 5 16:53:14 2010 -0800
msm: Change type of msm_shared_ram_phys to 'unsigned int'.
The __phys_to_pfn(msm_shared_ram_phys) macro is called, which performs
a right-shift of its argument. If the argument is of a signed type,
then its MSB is used for sign extension, corrupting the address if
msm_shared_ram_phys's MSB is set to '1'.
Hence, change the type to 'unsigned int' to support addresses which
have their MSB set.
Change-Id: I8282671ae8353f87275313424e347e79f347b208
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cebf7f34fffa539f872d9760e7f9ba6b922aaf7c
Author: Neeraj Kumar <c_ne...@quicinc.com>
Date: Mon Jan 4 12:01:20 2010 -0800
msm_fb: Added data packet size option in mddi_host_register_write
Provided option for flexible data packet size of
4, 8,12,16 and 24 bytes for a register write
command from MDDI Host. This changes the data packet
size but does not add support for multi-register write
CRs-fixed: 212143
Signed-off-by: Neeraj Kumar <c_ne...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 772767986af09b876a8c68630aa8a2091cf86459
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Wed Dec 30 09:18:49 2009 -0800
msm_fb: add MSMFB_OVERLAY_GET ioctl
To allow user space application to retrive overlay pipe
configuration set by previous MSMFB_OVERLAY_SET request.
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 48b1b966dc8d64d47c5a1d3517d5d966e180156b
Author: David Brown <dav...@quicinc.com>
Date: Mon Jan 18 16:25:50 2010 -0800
defconfig: Initial ST1 defconfig.
Provide initial support for the following devices on ST1.
- Enable MSM RTC driver.
- Enable Pmem.
- Enable Framebuffer.
- Disable Comet target.
- Enable SD/MMC.
- VIC.
- remote spinlock.
- i2c.
- ST1 ethernet.
- Panel selection for ST1.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 09cab39d8c224a1b5f46513d3c025786db8e67fd
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Wed Oct 14 17:20:31 2009 -0700
dma: barrier api for coherent memory before and after dma
Coherent memory in ARMv6+ could be StronglyOrdered or Normal. On ARMv7
StronglyOrdered guarantees program order execution only within 1KB and
Normal memory could have speculative fetches on them. Hence we need
barrier operations before and after dma for coherent memory.
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Change-Id: I33a5f37af7114a7bf13d6b6706c4eca1340b5e41
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3b52e7fac0be59caa2db63df56fed92d2b13efa3
Author: David Brown <dav...@quicinc.com>
Date: Fri Jan 22 10:52:46 2010 -0800
Update to latest MSM framebuffer driver.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7712cc42274817c471fd51ed06e103c59e120cd2
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Tue Nov 10 14:39:48 2009 -0800
mddi: Add barriers for coherent memory accesses
Memory returned by dma_alloc_coherent could be StronglyOrdered or Normal.
Access to StronglyOrdered memory does not guarantee program order beyond
1KB and access to Normal memory may have speculative fetches in addition to
reordering. Introduce the dma_coherent_pre/post_ops at appropriate places.
Change-Id: I1e4fd372e18611134fa7d4acfdadba5a21561bba
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2334a9bbb5f72db43fd34a0f7dd3d6af57958ed0
Author: Ai Li <a...@quicinc.com>
Date: Wed Nov 25 12:36:25 2009 -0700
msm: timer: halt when Modem fails to respond to time sync
Instead of failing silently and causing timer instability, halt the
Application Processor when it times out waiting for Modem's time sync.
A config option controls the timeout behaviour. The choices
MSM_PM_TIMEOUT_HALT and MSM_PM_TIMEOUT_RESET_MODEM are useful during
development and debugging. MSM_PM_TIMEOUT_RESET_CHIP can be used on
production builds. This patch also increases the timeout value to
avoid false detection.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b76ab056e83ebef1065c238515d227744a79176a
Author: David Brown <dav...@quicinc.com>
Date: Thu Nov 5 13:57:01 2009 -0800
msm: st1: Add experimental flag.
Certain revs of the ST1 hardware are not wired correctly. The
CONFIG_ST1_EXPERIMENTAL flag allows these devices to be used.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d402730744c833ba484fce43258a2f3c6469280b
Author: Shubha Nagaraja Rao <shu...@quicinc.com>
Date: Wed Nov 18 09:02:52 2009 -0500
smsc911x: Add Ethernet support for ST1
Add resources for smsc911x ethernet driver in board file. Also
add temporary software workaround for wrong hardware wiring of
address bus bits on ST1 boards in the driver file.
Signed-off-by: Shubha Nagaraja Rao <shu...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 17196511587b81a6a8db26b25dc0047d665d517c
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Nov 25 13:01:34 2009 -0800
msm: sirc: Add support for 8x50A interrupts.
The INT_SC_TEMP_SENSOR is also renamed to INT_TEMP_SENSOR to be more
general. The location of the temperature sensor associated with this
interrupt differs between 8x50 and 8x50A. Presently, it is unused by
software.
Change-Id: I23e4dc01a4287e0d578e3caae7cab4da1f90a692
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1bcc27df383b70ce0f3c4a40c70d003dfa32b5b4
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Fri Nov 20 16:46:03 2009 -0700
msm_i2c: I2C controller should be in receive-data state during reads.
Between writing address and polling for 'receive-data' state, I2C
controller seems to transition from 'address' state to 'tx-data'
state and then to 'receive-data' state. So polling during reads
for just 'data' state may be incorrectly over when the controller
is in 'tx-data' state and not really 'receive-data' state.
This commit fixes that problem by making sure that I2C controller
is in 'receive data' state and not just 'data state'.
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1a05b920429ca2c0057320cce38bf92bb117a37b
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Dec 2 18:09:17 2009 -0700
msm_i2c: Workaround for hardware setup time issue.
Due to a hardware timing issue, data line setup time may be reduced to
less than recommended 250 ns. This happens when next byte is written
in a particular window of clock line being low and master not
stretching the clock line. We introduce delay of just over 1/2 clock
cycle to ensure master stretches the clock line thereby avoiding
setup time violation. The delay is introduced only if I2C clock FSM
is low. Delay is not needed if I2C clock FSM is high or FORCED_LOW.
CRs-fixed: 215856
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b66315ab96dfc4ac14842f1e7e12129186ceb0a5
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Tue Nov 10 14:33:45 2009 -0700
msm_i2c: Use less restrictive inter-processor remote mutex locks
Remote spinlock is more restrictive since it prevents preemption.
If other processor doesn't release the remote lock quickly,
spin-locks become in-efficient since CPU can't run other ready tasks.
Remote spinlocks in i2c-msm are replaced by less restrictive remote
mutex-locks.
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5f8cf04698c192fb68b3cfdc6b3a4f1cf3d73b9d
Author: Vladimir Kondratiev <vkon...@qualcomm.com>
Date: Tue Nov 17 09:11:11 2009 +0200
msm: io: Extend HWIO area to cover EBI2 CS[67]
On MSM arch, devices mapped to the HWIO area should be set as
MT_DEVICE_NONSHARED. CS6 and CS7 areas on EBI2, mapped to
physical address range [0xd0000000 .. 0xefffffff], also fall into
this category.
There are devices which need to use EBI2 CS6 and CS7 memory
mappings. When these areas are mapped without
MT_DEVICE_NONSHARED, system hangs on memory access.
Extend platform override to fix MT_DEVICE -> MT_DEVICE_NONSHARED
for areas above.
Change-Id: I5de71019d3d342494bb6b5551e1e90a4960685b5
Signed-off-by: Vladimir Kondratiev <vkon...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2fb4ab0ccb431287fa70b98a1cb9f5267eef86d8
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 21 14:10:55 2010 -0800
Rest of the IRQ code refactoring.
Generalize the irq handling in preparation for future MSM
devices.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e0eff14e0f99fda8dc2b34f0e1a2906cb41f2c6c
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Wed Jan 20 16:31:29 2010 -0800
msm: Compilation fixes
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 00ba9668357ae628396df0540745cd5a53d4be66
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Dec 16 10:44:55 2009 -0800
msm: remote-spinlock: Add 'MSM' prefix to REMOTE_SPINLOCK_* Kconfigs.
The remote spinlock Kconfigs are specific to mach-msm. Their names
should reflect this.
Change-Id: I894bffa90c1e489f80c1caec8673d8c4f1fd23aa
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6d5f3cdd1088e6be94c730db1ea64db4b4ed4be2
Author: Steve Muckle <smu...@quicinc.com>
Date: Wed Dec 9 14:03:13 2009 -0800
msm: irq: rename existing irq support to irq-vic
The existing MSM irq support is specific to a VIC implementation.
Renaming this makes room for irq support based on other interrupt
controllers.
Change-Id: I2ba20ab146aa3c449ff21d163e2ec67515460719
Signed-off-by: Steve Muckle <smu...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d206265c145efb2d6916396d538de2b02c81a2a3
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Nov 19 20:36:16 2009 -0800
remote_spinlock: Add remote mutex support.
Remote mutex is similar to remote spinlock but allows interrupts and
other threads to preempt the current thread.
The remote mutex would be preferred over remote spinlocks in cases
where the remote processor holds the lock for long durations or when
the local processor needs to hold the lock for long durations.
Change-Id: If913cc2009758d42d4264761b647e21d0935fc48
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5679786615f4ad0e1c954bc10df89709799b4f78
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Tue Nov 10 13:23:50 2009 -0700
remote_spinlock: Add trylock support for remote spin locks.
Change-Id: I6e7da977063617c815649781911fb3a036e2945e
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a752916f1bc866fe3801cd2517be4ba6633cda8b
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Nov 10 15:42:12 2009 -0800
msm: remote-spinlock: Fix Dekker's algorithm.
A set self_lock flag indicates a request for the lock. Also make
sure the interpretation of the next_yield field is consistent with
other users of the locks.
Change-Id: I80d27d069e18c81e123854955eeee94e078a724d
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c5ee5854c4156d2e21a2188ebe964add466a0707
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Oct 27 20:14:12 2009 -0700
msm: remote-spinlock: Use Dekker's Algorithm for 7x30 SoCs.
The 7x30 SoC does not support shared memory (SMEM) locking using
LDREX/STREX instructions, and an atomic SWP instruction is also
unavailable. A software mutual exclusion protocol is used instead.
Change-Id: I17efbbeb78c5977138a44f0483a54bb6b093d2e6
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a6f7953f58c9d43d8063b7976d3d3518f278d01c
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Oct 15 09:37:20 2009 -0700
msm: remote-spinlock: Add support for DAL remote spinlocks.
DAL spinlocks operate similarly to the existing remote-spinlocks,
but are located in the DAL region of SMEM. The locks are found in
DAL chunk headers, which are identified by string names.
The remote-spinlock initialization API is modified to use strings
to specify which lock to use. Strings of the following forms are
valid:
"D:.{1,12}" for DAL spinlocks, where what follows the
colon is a DAL chunk name.
"S:[0-9]" for SMEM spinlocks, where what follows the
colon is the SMEM spinlock number.
Change-Id: Iaa90af0bfee33877e64ca19dac88f4038e1971a8
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dffccd4d3ee2e64229aab93ac067698deec02b54
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Aug 19 17:51:42 2009 -0600
msm_i2c: Add I2C Auxiliary interface.
MSM I2C hardware supports primary and auxiliary interface lines.
This commit adds support for auxiliary interface lines using
a different I2C adapter so that devices connected to the
auxiliary lines can be communicated with. Since both interfaces
share a common HW controller, only one adapter can be actively
used at any time.
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8c916d5768f0f12180286be4c1fd7670eb2b78e9
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Aug 12 14:33:52 2009 -0600
msm_i2c: Recovery function to recover from unexpected HW errors
Unexpected spikes from some slave devices (eg Touchscreen) cause
I2C HW controller to go in bus_busy state.
The function recovers I2C HW controller from that state.
CR fixed: 192442
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 09a437a74b006cdc3ae63d4c1c9ca82f5b45a222
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Jul 22 19:19:07 2009 -0600
msm_i2c: Add multi-processor concurrency support for I2C driver.
On some boards (e.g. 8x50), I2C bus is accessed by Modem and Apps
processor. The multi-processor lock prevents contention.
CR Fixed: 188034
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 67497174d90f43f40ef05e7ef9d972bedda3cff6
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Wed Jun 10 08:41:47 2009 -0600
i2c_msm: Derive PRI and AUX interface numbers from adapter ID
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 05aea73bf78ef30f42382bd7b692ca65888ebe37
Author: Sagar Dharia <sdh...@quicinc.com>
Date: Thu Jun 4 23:07:56 2009 -0600
[ARM] msm_i2c: Fix Last Byte issues while reading.
Slow slave devices reading 1-byte now go through interrupt.
Per MSM I2C HW requirements, LAST_BYTE is written before reading
2nd-to-last byte in multi-byte reads.
Signed-off-by: Sagar Dharia <sdh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ef1120f662c4c9d6803c32bcc47f87b03a699ab3
Author: Kenneth Heitke <khe...@quicinc.com>
Date: Wed May 27 20:24:25 2009 -0600
[ARM] msm: i2c: Configure I2C GPIOs from Linux side
I2C GPIOs are configured to use HW controller during probe.
Functionality is added so that GPIOs can be switched between
HW Controller functionality and SW-IO controller functionality.
CRs-Fixed: 176489
Signed-off-by: Kenneth Heitke <khe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 31112bf54fde1645f8ab0e7c9bd74ebbdc7cebd2
Author: Kenneth Heitke <khe...@quicinc.com>
Date: Mon May 18 17:47:09 2009 -0600
[ARM] msm: i2c: Don't send STOP between write and read operations
Implements the repeated START condition as specified by the I2C bus
specification. This is required for some I2C devices that are expecting
the repeated START condition rather than the STOP/START condition.
CRs-Fixed: 176188
Signed-off-by: Kenneth Heitke <khe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7d20538c26d0f545329df240ce3ce3e621ba6fc7
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Tue Jan 19 13:06:46 2010 -0800
i2c-msm changes from the initial contribution.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9e3404b793f2e65815341d65f83623b83a31bce1
Author: Arve Hjønnevåg <ar...@android.com>
Date: Tue Mar 24 14:45:41 2009 -0700
[ARM] msm: i2c: Send another byte if we are still the bus master in msm_i2c_recover_bus_busy.
If a transfer timed out before we told the controller to stop writing the
controller may be waiting for more data to write. Write 0xff as the last
byte, after switching the lines to gpio mode, to trigger a failure.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c54dba2e6647660c54b2e60f7bf6135c417d60b5
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Mar 20 17:06:01 2009 -0700
[ARM] msm: i2c: Only send STOP for last for the last write message in the group.
This allows drivers to perform a write followed by a read with a repeated
start. A repeated start after a read is still not possible.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 28fb39bf20899a0b0dfaf0451391b9c112fe6c26
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Mar 20 16:57:38 2009 -0700
[ARM] msm: i2c: Always read at least two bytes
If we tell the controller to stop at the same time as it finished reading
the first byte, it may ignore the stop request, loose bus access or fail
to start the next read or write. We now always read an extra byte if a
single byte read was requested.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit aebdde283fd9903df5a314ee91d67dce2c6f7802
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Mar 20 16:32:40 2009 -0700
[ARM] msm: i2c: Improve error recovery
- Reduce bus busy timeout.
- Call recovery function if the bus is busy at the start of a transaction.
- Wait for bus busy to clear after a transaction.
- Move read stop on timeout to recover function.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 81680c992af6d97b7158d586efcd0ad53d8f742a
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Mar 20 16:12:50 2009 -0700
[ARM] msm: i2c: Prevent infinite reads.
The contoller sometimes ignores our stop request. Send a new stop request
before flushing every other byte.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 09d57995c0dadf6f726dad8c9ed51f02d5f4ea04
Author: Arve Hjønnevåg <ar...@android.com>
Date: Thu Mar 12 22:10:29 2009 -0700
[ARM] msm: i2c: Perform entire i2c transfer at interrupt context.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a03a256d0715aa020437ce49f4ad40e911e04a58
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Tue Jan 19 13:01:54 2010 -0800
msm: i2c: add msm_set_i2c_mux function
This adds a function, msm_set_i2c_mux, to change i2c mux settings since we do
not have mux api.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit c03ab6c67dc0d59e7d4ecfa8f1891354825aafa6
Author: Arve Hjønnevåg <ar...@android.com>
Date: Thu Mar 12 18:43:33 2009 -0700
[ARM] msm: i2c: Try to recover if the i2c bus is busy after a failed transfer.
If the bus is busy, switch the i2c pins to gpio mode and drive the bus
manually. This will clear the busy condition with some clients.
This adds a function, msm_set_i2c_mux, to change i2c mux settings since we do
not have mux api.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2750832e2f3dc9abd21d89bacfc5b471eecdd505
Author: Arve Hjønnevåg <ar...@android.com>
Date: Thu Mar 12 17:27:08 2009 -0700
[ARM] msm: i2c: Try to read the requested number of bytes.
If the requested read length is two or greater, this should always work.
It the read length is one, an extra byte may be read if the interrupt
handler is not called in time. The previous version always read an extra
byte.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d64aaa9a7bb2674d5447bd29b80b77131245433f
Author: Thomas Tsai <thoma...@htc.com>
Date: Tue Feb 17 14:56:34 2009 -0800
[ARM] msm: I2C: Send stop command to finish i2c transfer on timeout
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7665cb9a1e5e868b281f4048261d994978dc43df
Author: Arve Hjønnevåg <ar...@android.com>
Date: Thu Oct 30 15:52:52 2008 -0700
[ARM] msm: driver for MSM7X00A on-chip I2C controller.
San: Re-write driver to write/read to/from buffer in IRQ context
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c31c1f1af116bd1071f5411d600964000e111cae
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:34:32 2010 -0800
msm: qsd8x50 platform data updates.
Bring in numerous changes to the qsd8x50 platform data. In order
to allow the drivers themselves to be pulled in cleanly, bring
the board and device files to the latest versions.
Basicaly we're breaking the cyclic dependencies of commits by
bringing in header files first (commits before this), then the
board files themselves, and this will be followed by the drivers.
This boardfile describes devices that aren't yet committed, but
will follow in subsequent patches.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d3a2b4361bc4f1701a2d4b5f912f33d03f3ec86c
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Fri Nov 20 12:12:54 2009 -0700
msm: Add SMEM_SEFS_INFO smem item
Add SMEM_SEFS_INFO smem item required for
sharing RAMFS information.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b2d233dd6f92ed34fe28fcf446e1aaa68ae391b8
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Oct 15 13:57:39 2009 -0600
msm: Refactor SMD driver. debug interface, separate channel list.
- Move debug interfaces to smd_debug.c file
- Keep separate list for modem and q6 smd channels
- Renamed some variables
- Simplify interface to access smsm items
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 61c82e915ddf133ba78f7b765001708d64d7b0ca
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:23:53 2010 -0800
msm: Reset callback for I2C keyboard.
Add a reset callback to the I2C keyboard, used on some targets to
externally reset the keyboard through a GPIO or other means.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 31b233f744c2be16429796b0415b5ab951f8f9f4
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:20:05 2010 -0800
Add internal mach-types definitions for 8650A targets.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 96fcd6757755348d3b587d088db24c990a34a600
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM boardfile support.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2cc2147dbb09f589c19616fc1ab6f3d5639ab97e
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM kgsl driver.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c16383e0d13fccaf65ce5052c4a766bbb0e173fd
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM battery driver.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit fce93b7a633570bfd12f5589f12e8f103124b593
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM TSIF driver.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 85f3384e65b24163f0392bde15ed148736c885fc
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM s1r72v05 driver.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 49d853403e4845337aa1846decbb16f50077b6cf
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM SPI drivers.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0301dea1264ccfcf26fc565d06925da78be059bb
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM camera drivers.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 97deeb3defcfbe46abec658faeb76c141c77ad11
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM high-speed uarts.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ac01cc6106f062b442e18178091beaa68540fc98
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:03:39 2010 -0800
Device data for MSM hsusb devices.
Needed to break cyclic dependencies with the platform board
files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b8990ac8f8bf862298a2d891d944a725ddcc0f70
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 09:01:56 2010 -0800
BMA150 platform data.
Need platform data for BMA150 driver to boostrap dependencies
with platform files.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e925c3e55aa6854f23d148d6839b767958430355
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 08:58:52 2010 -0800
msm: mfd/tps65023 header.
Needed early to break cyclic dependencies between drivers and
platform data.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2d48a7a3269dd21b8e1527007e907f6a8924fefe
Author: David Brown <dav...@quicinc.com>
Date: Tue Jan 19 08:52:12 2010 -0800
usb: Add platform data for mass storage function.
Based on work by Android/Google. The platform data is needed so
that the platform data can compile, and is coming before the
driver itself to break the dependency chain.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 878084b60f390f0ba4c6c99e7ef248aa47200880
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Jan 14 14:41:58 2010 -0800
arm: Add machine id for ST 1.5.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 394ac03b38915f78fdadd487242a11fe7ca8ac34
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Sat Dec 5 11:35:26 2009 +0530
msm: sdcc: Add support for write protect switch detection.
SD cards have a write protect switch, depending on the position
of which, the host software can put the card in read-only mode,
thereby preventing any write accesses to the card.
CRs-fixed: 211278
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 37e086f6b17dcb43c5caea1ae277a447fe6882e1
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Sep 26 22:10:56 2008 -0700
PM: Add early suspend api.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7de1f300ef2d4fe4df239807730462643c28a2f1
Author: David Brown <dav...@quicinc.com>
Date: Mon Aug 24 15:30:38 2009 -0700
msm_sdcc: Initial port of Qualcomm MSM MMC driver.
Original 7K driver by San Mehat <s...@google.com>, modified by
Qualcomm Innovation Center to support QSD860 MMC/SD/SDIO
controller.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit fd847a47f785c9ac6f115671c1580024495c91a5
Author: Dmitry Shmidt <dimi...@google.com>
Date: Tue Nov 11 11:22:38 2008 -0800
trout: Add functions for WiFi
Signed-off-by: David Brown <dav...@quicinc.com>
commit cfd293c013e8f4434b39e3135f74e508ecdd9b7b
Author: Yunsen Wang <yun...@quicinc.com>
Date: Tue Sep 22 21:57:13 2009 -0700
mmc: core: Decrease minimum length requirements for SDIO CIS_FUNCE
Decreasing min length to 34, to temporarily fix Libra wlan SDIO chipset
detection, will be removed once we move to new Libra wlan chipsets.
Signed-off-by: Yunsen Wang <yun...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0bb46996548db8b2d3c1a0b066225773e3684e60
Author: Dmitry Shmidt <dimi...@google.com>
Date: Fri Mar 6 09:34:40 2009 -0800
sdio: fix sdio_reset_comm() error path
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 54ab1275d090d51c527389527f9bfb98296ceab2
Author: San Mehat <s...@android.com>
Date: Thu May 15 09:15:37 2008 -0700
mmc: Add new API call 'sdio_reset_comm' for resetting communication with an SDIO device
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d7f4bbe2a1b4c970a129850378f847024b81e95f
Author: Dima Zavin <di...@android.com>
Date: Wed Mar 25 21:29:45 2009 -0700
mmc: sd: Remove debugging printk
Signed-off-by: Dima Zavin <di...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 50483bf86c70cd56ece8a4b55efc91f332a23519
Author: San Mehat <s...@android.com>
Date: Mon Dec 1 08:52:34 2008 -0800
mmc: sd: When resuming, try a little harder to init the card
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a70a669a0592810dac2c366be5540ffbbd5ef400
Author: San Mehat <s...@android.com>
Date: Thu Dec 4 11:18:00 2008 -0800
mmc: sd: Add retries in re-detection
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0cf12e3292cdcba15cfb1a31d8e2af731df61136
Author: San Mehat <s...@google.com>
Date: Tue Nov 11 09:35:36 2008 -0800
mmc: sd: Add new CONFIG_MMC_PARANOID_SD_INIT for enabling retries during SD detection
Signed-off-by: San Mehat <s...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 86bf17e74054a613a3dad46ef51dd509145cd07b
Author: Aparna Mallavarapu <apa...@qualcomm.com>
Date: Mon Nov 23 22:53:07 2009 +0530
mmc: core: Add 1ms delay after sending CMD6 to MMC cards.
Some eMMC cards require a delay of around 1ms after sending
the CMD6 command, before accepting further commands.
Signed-off-by: Aparna Mallavarapu <apa...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 87bc48519495eb970cf8f5bbbba1b7beedfa4f1a
Author: San Mehat <s...@android.com>
Date: Mon Mar 23 12:20:37 2009 -0700
mmc: core: Hold a wake lock accross delayed work + mmc rescan
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7323faf0946d51b7403f9564c845b931d5577d02
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Sep 26 22:10:56 2008 -0700
PM: Add wake lock api.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f33e490ef31a79bb6eb033df7965a98054bc24ba
Author: San Mehat <s...@android.com>
Date: Wed Dec 3 10:22:59 2008 -0800
mmc: mmcblk: Add new feature 'CONFIG_MMC_BLOCK_PARANOID_RESUME'
With this feature enabled, mmcblk will check the card-status before issuing
a transaction *only* after being resumed. This protectes us from issuing
transactions before the sdcard is ready (which can occur if the host driver
deferrs mmc_resume_host() to reduce resume latency)
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 806da8003c59a4af75e44bdb5246e02edc63123b
Author: Sahitya Tummala <stum...@qualcomm.com>
Date: Tue Oct 6 12:11:33 2009 +0530
mmc: block: extend support for more than 8 partitions
For cards that can be used as bootable devices, there is a possibility
for the number of partitions to exceed 8. So increasing MMC_SHIFT to 4
so that up to 16 partitions can be supported.
Signed-off-by: Sahitya Tummala <stum...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cf2e21d4324ddd8e38d8c7041f2856eeba0d24e7
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 14 14:08:27 2010 -0800
msm: dma: Export flush/stop commands.
These commands are needed by the high-speed msm UART driver.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 37d4e20856078178b641327810aabd7a73cd894c
Author: Brent DeGraaf <bdeg...@quicinc.com>
Date: Tue Nov 10 11:19:34 2009 -0500
msm: Extend datamover callback infrastructure
The datamover supports channels which can be shared amongst devices.
As a result, the actual data transfer may occur some time after the
request is queued up. Some devices such as mmc host controllers
will timeout if a command is issued too far in advance of the actual
transfer, so if dma to other devices on the same channel is already
in progress or queued up, the added delay can cause pending transfers
to fail before they start. This change extends the api to allow a
user callback to be invoked just before the actual transfer takes
place, thus allowing actions directly associated with the dma
transfer, such as device commands, to be invoked with precise timing.
Without this mechanism, there is no way for a driver to realize
this timing. Also adds a user pointer to the command structure for use
by the caller to reference information that may be needed by the
callback routine for proper identification and processing associated
with that specific request. This change is necessary to fix problems
associated with excessive command timeouts and race conditions in the
mmc driver.
Signed-off-by: Brent DeGraaf <bdeg...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1915118275d9a655afeff61656b4dc34fa620703
Author: Shubha Nagaraja Rao <shu...@quicinc.com>
Date: Wed Nov 18 09:02:52 2009 -0500
smsc911x: Add Ethernet support for ST1
Add resources for smsc911x ethernet driver in board file. Also
add temporary software workaround for wrong hardware wiring of
address bus bits on ST1 boards in the driver file.
Signed-off-by: Shubha Nagaraja Rao <shu...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit fa911ca860453457ea23700512d776dbb92e9d23
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Mon Jan 18 11:52:36 2010 -0800
msm: msm_fb: add msm framebuffer support
This add basic support for framebuffer devices, and specific support
for qsd8x50.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 67b50b1a7e5236e86785ec480fef6f8d86f37fd9
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Mon Jan 18 12:08:54 2010 -0800
msm: Allow the size of the in-kernel PMEM region to be configured.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8daba594ea4fb3470b5302ffab0c359d14dccb60
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Fri Jan 15 10:33:58 2010 -0800
msm: qsd8x50: check pmem size.
Make sure the pmem region is not zero before doing the allocation.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit b5edfce50384867ab93918b5784ff64942a762ad
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Fri Jan 15 10:06:27 2010 -0800
msm: qsd8x50: update pmem region to bitmap
Use the new bitmap allocator for qsd8x50.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 2cc43a35129eda45a76e66e666a85c8d07d7f7b6
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue Dec 15 15:38:24 2009 -0500
msm_fb: correct external WXGA name in Kconfig
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c8f5e7f58c48121749451ae580a0732f00c3b50e
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Dec 10 11:34:52 2009 -0500
msm_fb: correct porch values for ST1 WXGA display
Correct porch values are needed for external display.
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1e5eada158820c0113071e5b6305e25ad2d2b622
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Nov 5 13:02:09 2009 -0500
msm_fb: Add wxga drive for ST1 panel
Change-Id: I168ba988404aed774e82daec99d44dae8b568523
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ba638559e4fd1741211499ee2eaabd2acb1657ff
Author: Neil Leeder <nle...@quicinc.com>
Date: Tue Sep 15 17:41:18 2009 -0700
msm: separate grapefruit function by machine type
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 686554b01a2ee18768fad46eeaa969c56d9b4cbd
Author: Neil Leeder <nle...@quicinc.com>
Date: Fri Sep 11 14:53:02 2009 -0700
msm_fb: add grapefruit framebuffer support
Adds new video device to support grapefruit LCDC display
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d69b1de8bbdb30e9e9eade3c167974a913bd4031
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Mon Jan 18 13:19:09 2010 -0800
msm_fb: use driver specific Kconfig
Remove the drivers/video/Kconfig entries, and convert to using
drivers/video/msm/Kconfig .
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit b89fd93eb0e5e6c0905359782f5210a97c28b791
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Thu Oct 1 15:27:43 2009 -0700
msm_fb: add MDP4 RGB pipe as overlay base layer
Add utilization of MDP4 RGB pipe as overlay base layer in addition to
existing DMA_P DMA engine to fetch pixel images from Frame Buffer and
deliver images to display panel. This is the foundation toward to
completion of MDP4 overlay implementation.
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 67f2df1729fc9c108dc6a4ebf58fac02cec7e217
Author: David Neiss <dne...@quicinc.com>
Date: Thu Oct 15 00:39:30 2009 -0700
msm_fb: Added frame_count to debugfs dir
Added a frame_count file which contains a running count of number of
frames updated. Useful for computing FPS.
Signed-off-by: David Neiss <dne...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit db243785b26f40713211b325206629cd619415df
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Oct 8 15:29:15 2009 -0700
msm_fb: enable mdp hw vsync
Uses mdp hw vsync to synchronize to mddi lcd panel.
mdp hw has internal vsync counter to be used for
synchronizing dma update. For lcdc, it uses
different hw timer for pixel synchronization.
Change-Id: I1c446445810d6d4b430128e0e8ca1d80059d08c4
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit aff2c783c539da39f4e1e1fbc50fe3e2acb6d552
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Oct 8 16:02:20 2009 -0700
msm_fb: mddi toshiba init routine bug fix
Fixes mddi init routine with correct timing.
Change-Id: I35081f596762d2b636d64258d26826af1cbddf1a
CRs-fixed: 211786
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 353601f23091e2300597ae3210e01592679cd877
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Thu Oct 8 08:04:47 2009 -0700
msm_fb: increase brightness of Toshiba wvga display panel
increase VCS voltage to increase brightness of Toshiba wvga
display panel (LTM030DD40)
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0fc79133bcb56bf2da99ee8864e7622a161d5141
Author: Stephen Boyd <sb...@quicinc.com>
Date: Thu Jul 23 15:48:00 2009 -0700
msm_fb: fix dependencies when disabling pmem
MSM sound, graphics, and camera drivers all depend on PMEM being
enabled. If pmem is disabled, these drivers should not be
included.
Also fix some warnings:
drivers/video/msm/mdp_ppp.c:1141: warning: unused variable 'vstart'
drivers/video/msm/mdp_ppp.c:501: warning: 'get_len' defined but not used
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f7ccb5626edc78fe1959ecb7a9db10d8a97afaa3
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Mon Oct 5 15:41:19 2009 -0700
msm_fb: fix display panel driver auto detection error
Have lcdc_toshiba_wvga_pt display panel driver check its name tag
against board files dispaly panel's database. Do not launch itself
if name tag checking failed to prevent two lcdc display panle drivers
active at same time.
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5b429fe2b3cd3795494fbb040674d34cb82215f3
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Tue Sep 29 17:15:51 2009 -0700
msm_fb: Reduce AXI bus frequency to 58 Mhz from 64 Mhz.
Reduce AXI bus frequency further to 58 Mhz for lcdc panels to save
~20mA power during idle screen mode/limited sleep mode.
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4fbfd5505236e45df5ec75bcfb8371bf7e22948a
Author: Michael Bohan <mbo...@quicinc.com>
Date: Tue Sep 29 10:52:56 2009 -0700
msm_fb: Add support for pclks
7x30 has additional clocks (pclks) that must be
managed along with the traditional core clocks.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 33dba84d10d9552735843e8f02b84c1fce15476c
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Wed Sep 23 09:38:58 2009 -0700
msm_fb: add display panel auto detection
add display panel auto detection between MDDI and LCDC
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2aa6bf6fdfe42fc6af93fdd459afd65c0397ae76
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Wed Sep 16 08:56:17 2009 -0700
msm_fb: add mdp4 lcdc feature
add Toshiba wega display panel (LTM030DD40) driver to support
mdp4 lcdc feature
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f9ee2224a718be3d4905eee05ded80934abebe29
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Thu Sep 17 14:03:57 2009 -0700
msm_fb: Reduce AXI bus frequency to 62 Mhz from 128 Mhz
Reduced AXI bus frequency to 62 Mhx to save power during idle
screen mode/limited sleep mode.
CRs-fixed: 210721
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4210b88f65cbacffddc014c876c544dcc27648a6
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Sep 2 14:04:09 2009 -0700
msm_fb: disable mddi host timer under suspend
mddi host timer kicks off mddi transaction every
5 seconds which needs to be stopped when mddi host
is disabled.
CRs-fixed: 194174
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f39105a362013f41e98900a2ecf496bc18fea2be
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Sep 2 13:27:26 2009 -0700
msm_fb: turn off mddi power when there is no mddi panel
Adds mddi power platform function under mddi_disable which
gets called after mddi panel detection routine finds no
mddi client panel attached to the device at runtime.
Change-Id: I2c5b4c0758b4b1c82e9cd2f91974fe50d1769097
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1ce543ca4519a2bcf5deb6665a455f67fb594229
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Thu Sep 3 15:45:32 2009 -0700
msm_fb: Support for dynamically changing mdp ccs matrix.
Added two new ioctls's to set and get mdp color conversion matrix
from user space at runtime.
CRs-fixed: 190625
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6c6ca4655e80a4f1207a4f4abdc86fb3763d03c5
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Sep 2 11:36:52 2009 -0400
msm_fb: add S-video option to tv output
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 22d9a8abac9bf1131f7aebc78a4108ec2288da03
Author: Stephen Boyd <sb...@quicinc.com>
Date: Thu Jul 16 13:37:06 2009 -0700
msm_fb: cleanup Makefile to use CONFIGs directly
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 80cc9299c5c8a7617bf560ba6975b424fba75fb1
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Aug 27 10:36:47 2009 -0700
msm_fb: add MDP_BLUR support
Adds ability to apply blur effect via scale tables.
CRs-fixed: 190785
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 13992978823d78ab314bd39ef4452f34716c8899
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Aug 27 17:34:55 2009 -0700
msm_fb: add sharpening support for non-scaled image
Forces image to go through scaling block as mdp hw
only adds sharpening through scale filter.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 45d36ed9f45c694225ccb42db80ac2f097072fb0
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Aug 27 14:34:03 2009 -0700
msm_fb: fix lcdc ystride value
Needs to use X virtual resolution for ystride instead of
regular resolution.
CRs-fixed: 193932
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8245025430a4c57e367c5541a4347b9fd7cd75d7
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Aug 20 14:17:23 2009 -0700
msm_fb: move mddi power save under on/off routine
Moves mddi power save platform routine from suspend/resume
to mddi on/off function in order to save power even when
a user manually turns off the display panel, not necessarily
from suspend/resume function.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8cf8ac6c8330d639168a2b63abe28af24bc79e88
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Aug 20 13:43:35 2009 -0700
msm_fb: 8650 FFA suspend/resume display bug fix
Invalid mddi client register (PWM0OFF) access has caused
mddi client device to fail to operate after suspend/resume.
It's not necessary to access client registers when pmic
platform function is available. Therefore, it returns
right away.
CRs-fixed: 186622
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b0643475d2968fc343a1e11d8702acdb0905b369
Author: Michael Bohan <mbo...@quicinc.com>
Date: Tue Aug 18 19:28:56 2009 -0700
msm_fb: mdp4: Add support for 128 MHz AXI bus
Switching the AXI bus from 64 Mhz to 128 MHz causes the MDP to
hang during large data transfers on the msm7x30 SURF.
Fix:
-Reduce the DMA_P burst size from 16 to 8.
-Reduce the number of pending commands for the AXI port to 3.
The data shows that either of these changes alone
allows the display to work correctly, but for the time being we will set
both to be conservative.
According to the hardware folks, if the MDP clock is less than
the AXI clock, then we must use a max of three pending requests
-- otherwise we should be able to use 8. We should make a change
in the future to support this with proper testing.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 982a372f3bb0b812987f8262be9e8f97185b1e07
Author: Michael Bohan <mbo...@quicinc.com>
Date: Wed Aug 12 22:55:49 2009 -0700
msm_fb: Add Toshiba WEGA initialization for msm7x30
Proper bringup sequence is necessary for this chip. This
includes supplying the correct voltage and then reseting.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0b269baec43a61d6860b1feeb5e60ceb5242c230
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Fri Aug 14 14:48:50 2009 -0700
msm_fb: turn ON FFA Display after Bootup
changed dynamic range of backlight to turn it ON
after initial Bootup sequence.
CRs-fixed: 192264
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4ed2cb626cf57a71d1393b05e18846363c673053
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Aug 12 18:07:25 2009 -0700
msm_fb: disable mdp irq during idle
Under lcdc mode, mdp constantly fetches framebuffer memory
and sends out to the panel which never disable its clock
until suspend happens.
As we had irq logic attached to clock logic, this prevented us
from going into idle power collapse.
This patch is to separate mdp irq control from mdp clock control
for idle power collapse even under lcdc mode.
CRs-fixed: 188662
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 26a4e595a94267719d5e05dcbbb984f973e6aca2
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Thu Aug 13 11:49:49 2009 -0700
msm_fb: set AXI bus frequency to default after lcdc/tvenc turns off.
set AXI bus frequency back to default value after lcdc/tvenc
turns off completely, to avoid any underrun issues with PCLK.
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 441a32beb05e426d8294a9b8548605063dcdb224
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Aug 13 10:58:33 2009 -0700
msm_fb: remove the notch filter in tv encoder
The notch filter is misplaced in the tv encoder hw
that causes a ghosting artifact on tv.
And sw workaround is to disable the notch filter.
CRs-fixed: 183380
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6e6f3dc384a5e3a6be3f38af69517c525890a8d5
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Aug 12 18:06:42 2009 -0700
msm_fb: tvout frame done bug fix
Need to clear up TV frame start interrupt bit before
enabling it as it already might have been set from
the previous update.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8de814e714749346ea41e0d42cf9505cc2656553
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Tue Aug 11 10:48:03 2009 -0700
msm_fb: add MDP4 features into existing MDP driver
Implement MDP4-FrameBuffer-Push mode by utilizing DMA_P dma engine of MDP4
and MDDI interface to Display
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b30aeb7d0f4da33736b3eeed7cd9240909993c96
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Wed Aug 12 11:25:45 2009 -0700
msm_fb: Add AXI bus QoS API calls in pull drivers.
Add AXI bus QoS API calls to request a minimum AXI bus
frequency,which is always above the PCLK.
CRs Fixed: 186223
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ce05b9e0a4ee12c87953a59eb80cf2199fe15216
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Aug 12 11:01:46 2009 -0700
msm_fb: add timeout during mddi wait
Add timeout for mddi register read/write completion.
CRs-fixed: 183794
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e8f3eb6fbc9347a1c9537bd6095586642a72c905
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Tue Aug 11 08:44:18 2009 -0700
msm_fb: clean up code which does not comply with linux coding standard
Clean up existing video driver code which does not comply with
linux coding standard.
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ce0d55d288647ec0d0b7f20b30b9558f4bd5151b
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Thu Jul 23 18:18:51 2009 -0700
msm_fb: add backlight on/off support for msm7x27 lcdc
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 379d849f16b914bfbce4d3da4d3262d6795a48e1
Author: Bryan Huntsman <bry...@quicinc.com>
Date: Mon Jul 27 18:30:44 2009 -0700
msm_fb: fix BSD license statement in headers
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 944bd5a6014bbe6f58ee654d5d85d76a92a65639
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Jul 22 17:23:30 2009 -0700
[ARM] msm_fb: fix h2v2 image output
Add horizontal slice of 2 (h2) for correct h2v2 output image.
CRs-fixed: 188904
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 311de74bf08cf86a35c5a04619198ebe7ef847d4
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Jul 22 10:53:32 2009 -0700
[ARM] msm_fb: remove 24bpp check for lcdc
Support 16bpp and 18bpp lcdc panel.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 52d9cb296ce0c26c85b25b734eb2acff547fe5d4
Author: Joon Rhee <jr...@quicinc.com>
Date: Tue Jul 14 15:08:25 2009 -0700
[ARM] msm_fb: lcdc frame done bug fix
Need to clear up LCDC frame start interrupt bit before
enabling it as it already might have been set from
the previous update.
Also reset wait flag after updating hw register for lcdc
frame address to avoid ISR for sending an complete
event too early.
CRs-fixed: 189181
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bed98c004c733a34af77625a3321bda547a6e80c
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 15:05:01 2010 -0800
msm: add 7x27 lcdc_power_save callback
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 6588c9b8f80a84738ee7d13c9969dff41edf7a15
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Wed Jul 15 14:28:57 2009 -0700
msm_fb: added LCD_3V regulator on/off support in lcdc.
PMIC LCDC_3V regulator is now switched on/off with
lcdc on/off functions to reduce power leakage.
CRs-fixed: 188202,186467
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 11a2c851491577406b1bf07a5b8a8b18c2b1ca1b
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Mon Jul 13 15:41:46 2009 -0700
msm_fb: better fix for suspend resume issue
improve state transition conditions in lcdc panel
on/off function
CRs Fixed: 186467,182625
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5e9d8e01fe93d4bdf3c8c70c471176d278728a68
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Jul 9 13:27:54 2009 -0700
[ARM] msm_fb: csc pre limit vector fix
Fix csc pre limit vector set to 2.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b9712272782f08f4a14b96333754c7db5837761f
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Thu Jul 2 13:31:19 2009 -0700
video: Fix failure to fill in file pointer upon bad pmem return.
If the call to get_pmem_file in get_img function fails for any
reason, as it may well, the continuation of the function
to get the image file information was not filling in the file
pointer. The resultant null file pointer was then being passed
to flush_pmem_file, another PMEM function.
Due to testing of new features in PMEM, kernel panics
were being caused by a temporary failure of PMEM to protect itself
adequately against other kernel drivers invoking its functions such
as flush_pmem_file that may have passed in a null file pointer as one
of its arguments.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 07cab3a429bb438913bf9c25f109a8e80fa27be6
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 14:50:18 2010 -0800
msm: msm_fb: update 7x25 to use mddi prescan
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 797fbb3cc02d5786526544acea706a9c19c875c4
Author: Joon Rhee <jr...@quicinc.com>
Date: Fri Jun 19 14:24:36 2009 -0700
[ARM] msm_fb: enable mddi client detection on lcdc platform
For non-mddi platform, we need to scan for mddi client and see
if we can disable mddi host for power saving.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3559f597290db71a2fed636b3fd296ab11b5bb70
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Jul 1 14:29:11 2009 -0700
[ARM] msm_fb: move LUT block after histogram
For ABL calculation, LUT must be enabled after histogram
as it will accumulate the post-LUT result on top of
histogram.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a3e9a0e9f01f81049513bdfb414b16637efa1b2c
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jul 1 11:55:32 2009 -0400
[ARM] msm fb: rename external SVGA mode to WXGA
External LCDC mode is 1280*720, which is WXGA.
Update all defconfigs which reference this option.
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 91ee984a248bc2e0ab2faf71184216e213254f4a
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jul 1 13:35:07 2009 -0400
[ARM] msm fb: only call client_id function when it's defined
mddi_get_client_id is only defined when MDDI autodetect is
configured, so only call it when correct config option is set
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bf9443afb1151b502cd31582a0483ffabfcad1f8
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu Jun 18 15:10:08 2009 -0700
[ARM] msm_fb: change mdp idle wait time to 1 sec
This is to allow idle power collapse to happen
but not changing mdp power state too frequently.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5bd4c2c5a9312fe481b892fbe3afa7cb21f207cc
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Jun 10 17:51:57 2009 -0600
msm_fb: Try to detect a hung MDP
Before starting a new blit operation, ensure that the previous
operation has completed. If the operation timeouts, then print
a log message to the user.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 74c0073e3fbf7bc4cf65a07e55ad2e91b45908a3
Author: Abhishek Kharbanda <c_ak...@quicinc.com>
Date: Wed Jun 10 09:47:41 2009 -0700
msm_fb: fix for suspend resume issue
Remove few control registers in power off mode
that affect resume.
CRs Fixed: 182625
Signed-off-by: Abhishek Kharbanda <c_ak...@quicinc.com>
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7c66388579303036379a965b556c03819ff21a0d
Author: Joon Rhee <jr...@quicinc.com>
Date: Wed Jun 10 14:53:59 2009 -0700
[ARM] msm_fb: disable mddi from panel detection.
Disable mddi host when there is no mddi panel detected.
This is to allow tcxo shutdown for lcdc/rgb panel device
by turning off mddi hw.
CRs-fixed: 184798
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b479799607680eac97116862c0c62346d6bf65b6
Author: Jordan Crouse <jcr...@quicinc.com>
Date: Wed Jun 10 17:46:48 2009 -0600
msm_fb: Replace interruptible waits with killable waits
wait_for_completion_killable only exists on a critical signal -
this was the original intented behavior for the code. Using
wait_for_completion_interruptible will exit on any signal which is
undesirable in some situations.
Signed-off-by: Jordan Crouse <jcr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 77f17720fa11f59e0a9a0a068c5e97b3fedc1836
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu May 28 15:46:28 2009 -0700
[ARM] msm_fb: add histogram support
Add support for user app to retrieve histogram data
for each R/G/B component.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a2442dc5b031b7263a4bc63741047f87529da96b
Author: Joon Rhee <jr...@quicinc.com>
Date: Thu May 28 14:56:07 2009 -0700
[ARM] msm_fb: add gamma curve support
Add support for user app to set gamma (or LUT=lookup table)
for each R/G/B color component.
Signed-off-by: Joon Rhee <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 26228a40fe984dac075d41a82fb3e094b834882b
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 14:21:34 2010 -0800
Initial Contribution changes for msm_mdp.h
Signed-off-by: David Brown <dav...@quicinc.com>
commit cbefc2fc0f5f8e35b851080d89aa37766e1d02de
Author: Neil Leeder <nle...@quicinc.com>
Date: Fri May 29 18:02:12 2009 -0400
[ARM] msm_fb: add support for Comet2 display
Adds support for new display options available on Comet2
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4627efe49ac94c75ff1948d9c3e554fcac280838
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 14:13:57 2010 -0800
Add video Kconfig changes
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 90d13e6e9a54a4b4628506b33273a1f654b54edc
Author: Rhee, C. Joon <jr...@quicinc.com>
Date: Fri May 22 14:43:53 2009 -0700
[ARM] msm_fb: increase toshiba backlight level
Increase toshiba lcd backlight max level to support
adaptive backlight feature which requires at least
32 different settings.
Signed-off-by: Rhee, C. Joon <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9f7a43c71ea1c2fca61dafe66d65b5dd297cad88
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 13:22:42 2010 -0800
msm: msm_fb: update platform data backlight callback arguments.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 3869d138bf7ab2df74d6d8af4719ada4611988ec
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 13:19:41 2010 -0800
msm: msm_fb: add platform data into board.h
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 8f0048dcea8167c5d0b4d589ee0e88fcd989882b
Author: Rhee, C. Joon <jr...@quicinc.com>
Date: Fri May 22 14:22:48 2009 -0700
[ARM] msm_fb: change backlight to use max and min value
Change backlight platform function to use max and min value
to adjust the backlight level.
Signed-off-by: Rhee, C. Joon <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 38df09cab31130492b0bfde331a00297cec164f7
Author: Abhishek <c_ak...@quicinc.com>
Date: Fri May 8 15:04:10 2009 -0700
msm_fb : bug fix for MDDI/LCDC auto detection feature.
CRs-Fixed: 182179
Signed-off-by: Abhishek <c_ak...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6f405c34280ae29be23cd3a4b1f2cf789f1ec0f8
Author: Rhee, C. Joon <jr...@quicinc.com>
Date: Wed May 6 17:38:32 2009 -0700
[ARM] msm_fb: fix mddi resume lock up
Bring up mddi pad control after enabling mddi clock
CRs-fixed: 178051
Signed-off-by: Rhee, C. Joon <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a537ab2280a74ebd9d3fd8e15c86901954371ccd
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 12:57:10 2010 -0800
Initial Contribution for msm framebuffer.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7f1775916e4e883768b13c7a4a3d535c60c72ca2
Author: Dima Zavin <di...@android.com>
Date: Mon Feb 23 13:57:18 2009 -0800
msmfb: specify partial updates support as a panel capability.
Not all panels will be able to support partial updates.
Signed-off-by: Dima Zavin <di...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit fc7f800c2211df44e61028e0e0332a43f0aab576
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 12:36:38 2010 -0800
msm: msm_fb: add capabilities field
This add a capabilities field to the panel data structure.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit a5ca165b57175bbb98d042cbc3d82cdc203b8f50
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 12:34:14 2010 -0800
msm: msm_fb: update mach-msm msm_fb.h
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 3e9ea904fa4d10adcc538f8e03fa61d82452f51b
Author: Rebecca Schultz Zavin <reb...@android.com>
Date: Tue Feb 17 16:14:18 2009 -0800
[ARM] msm: mdp: Missing image type when checking if src has alpha
The workaround for a hardware bug in tile fetch is requried only if the
src image has an alpha channel. This code failed to check for one
image type when checking for alpha.
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dfe16c0fff291b03e6127e2a703f152dc2ee9e52
Author: Rebecca Schultz Zavin <reb...@android.com>
Date: Thu Feb 5 17:29:45 2009 -0800
[ARM] msm: mdp: Refer to image memory by file struct instead of fd
When down refing the file data used to blit, use the file struct instead
of the fd in case the file descriptor was deleted during the blit
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 18e5f230bec8540e128a3febf590f57dbfe5a300
Author: Rebecca Schultz Zavin <reb...@android.com>
Date: Wed Jan 28 18:16:58 2009 -0800
[ARM] msm: fb: Fixing bug in fb driver when invalid update is encountered.
When an invalid zero size update is sent to the fb driver the driver correctly
just marks that frame as done, but doesn't wake any waiters that were waiting
for that frame to complete.
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e51b7ae22b453396b194c1f75a184d09b3709dea
Author: Rebecca Schultz Zavin <reb...@android.com>
Date: Tue Dec 2 13:19:24 2008 -0800
[ARM] msm_fb: Fix bugs in dummy client driver, clean args for client functions.
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 13f0b90aa070ba928fefc3476869cf0fec92bcd6
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Nov 24 21:20:44 2008 -0800
[ARM] msm_fb: Don't crash if platform data is missing.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3ea10f4147dc034a8bcf802cd8b1557d0e77213f
Author: Rebecca Schultz <rsch...@google.com>
Date: Wed Nov 19 16:24:43 2008 -0800
[ARM] msm: Adding display drivers.
Adds mddi, mdp and msm_fb drivers.
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6b69d441be6cf0327d9a408b3e589e1aa29b4284
Author: David Brown <dav...@quicinc.com>
Date: Thu Jan 14 10:51:37 2010 -0800
pmem: Add Kconfig description.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a73594c6e1d7a9093cea11b39689bf98d4261706
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Jan 14 10:11:06 2010 -0800
pmem: add 8k board support
This adds some basic pmem regions in the 8k board file.
Signed-off-by: Daniel Walker <c_dw...@quicinc.com>
commit 293ca1b1ebd18cf48096dd8030c8c3a5163e6c5e
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Sun Sep 13 23:27:46 2009 -0700
pmem: Use efficient bit manipulation for bitmap allocator
Discard bit-by-bit bitmap manipulators in favor of custom built
functions. These new functions use parallel bit masking as much as
possible. This is feasible due to the contiguous nature of the
allocations.
Empirical testing shows dramatic improvement of execution speed with
this change; almost double, in fact. The test consists of two
different executables differing only in whether it uses the old or
this current algorithm for bit manipulations. Each executable runs
100,000 sets of simulated allocations. The runs showed a time
decrease from 5.596s using the old algorithm to 3.051s using this
new algorithm.
Also change bitmap allocator out-of-memory message to be more
understandable.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4e48759753b7e62511c8f56dbf54ef0be2b274d7
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Thu Sep 3 10:08:37 2009 -0700
pmem: minor code fixes
Further localize variables in a couple of places in the code.
Make sure that the return value of ioremap is fed back to original
caller.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b079a8a2fc95f9479aaf28b2e81f2b42b616606e
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Thu Sep 3 10:07:38 2009 -0700
pmem: fix output formatting and code cosmetics
Add function name to all output where possible.
Minor fixes to indentation and comments.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a2b8dc0874ff759aac330d4b402740b4db081c2e
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Mon Aug 31 11:44:37 2009 -0700
pmem: fail device initialization if passed in a zero size
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6980bb0a85754d07d68dcdd0f8ba66529dafd10e
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Wed Aug 26 14:20:19 2009 -0700
pmem: fix buddy free memory corruption bug
Enforce explicit upper limit for buddy index during memory free
operations. This index is generated during memory free operations of
the buddy bestfit allocator. The index points to each free buddy that
should be coalesced into a greater free memory area. The upper limit
now enforced is the number of entries in the current PMEM region.
Without this fix, pmem_buddy_bestfit_free would continue to corrupt
bytes in ever higher addresses, by power of 2, of memory until
stopped by the high bit being set in the currently examined byte.
Unless the high bit was set immediately at the first invalid byte,
this could have caused memory corruption past the current pmem region
metadata in kernel memory space.
One manifestation of this problem was a corner case where
non-existent buddies were incorrectly coalesced.
In a 12MB region, the buddy free for 8MB tried to coalesce the freed
buddy metadata with a non-existent 8MB buddy which actually belonged
to the following region. This caused the order number in the metadata
belonging to the following region to be incorrectly incremented past
the region boundary. Soon after, there was a subsequent invalid
allocation in that following region past the region boundary.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 997e6f890ea8cae395e2df7eafe562f4b378dbd3
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Thu Aug 20 11:37:17 2009 -0700
pmem: add sysfs component for status of regions
NOTE: Macro usage in the code causes checkpatch to fail in two
places. David Brown has seen this and agrees that it is a false
positive and that this will have to be manually verified and
integrated.
Add read-only elements under /sys/kernel for seeing status of PMEM
regions.
Merge debugfs functionality into the sysfs area, removing debugfs
completely.
The top directory is /sys/kernel/pmem_regions with each named PMEM
region having its own directory underneath.
In each PMEM region directory, there are files that can be read that
contain current values of various parameters and statuses for the
region desired. The file names and types change according to which
allocator the region was defined at in the board file.
Currently, all files are read-only. There is no capability provided
as of yet to enable modification of any PMEM parameters via sysfs.
Some sysfs files are common across all regions:
base -> shows the base address of the region
size -> shows the total size of the region in bytes
allocator_type -> prints a text string defining the allocator type
mapped_regions -> displays a table of all regions mapped according
to process id holding the mapping. This functionality had
been exposed via debugfs.
If the allocator type of the region is ALL_OR_NOTHING, a single
sysfs file appears under the directory of that region:
allocated -> shows a 1 if this region is allocated, 0 if not
If the allocator type of the region is either BUDDY_BESTFIT or
BITMAP, two additional sysfs files appear under the directory
of that region:
total_entries -> shows the total number of quanta present in the
region, allocated or not
quantum_size -> shows the size in bytes of the smallest allocatable
unit
If the allocator type of the region is BUDDY_BESTFIT, a single
additional sysfs file appears under the directory of that region:
buddy_bitmap_dump -> shows a list of all indices available to the
system and whether they are allocated or not
If the allocator type of the region is BITMAP, two additional
sysfs files appear under the directory of that region:
free_quanta -> shows the number of quanta available to be allocated
bits_allocated -> displays a table listing the current allocations
per bit with size, index, and number of quanta allocated
So, an example sysfs hierarchy directory tree could look like this:
/sys/kernel/pmem_regions --> top level directory
| + pmem_region1 --> e.g., an ALL_OR_NOTHING allocator region
| | + base
| | + size
| | + allocator_type --> prints the string 'All or Nothing'
| | + mapped_regions
| | - allocated
| + pmem_region2 --> e.g., a BUDDY_BESTFIT allocator region
| | + base
| | + size
| | + allocator_type --> prints the string 'Buddy Bestfit'
| | + mapped_regions
| | + total_entries
| | + quantum_size
| | - buddy_bitmap_dump
| + pmem_region3 --> e.g., a BITMAP allocator region
| | + base
| | + size
| | + allocator_type --> prints the string 'Bitmap'
| | + mapped_regions
| | + total_entries
| | + quantum_size
| | + free_quanta
----- bits_allocated
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit adced56b2623d364c4b1bfb0a7026348bc841b49
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Thu Aug 20 11:40:44 2009 -0700
pmem: verify alignment of incoming values
Check with all allocators that incoming size and start address
parameters from the board initialization routines are aligned
correctly and fail driver initialization if not.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4a47db8ab49f42df3670272853fd0d41b737934e
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Fri Aug 14 17:52:52 2009 -0700
pmem: fix spinlock recursion messages
Fix worrisome (but ultimately benign, at least in this case) output
from kernel related to spinlock debugging as a result of attempting
to acquire more than one spinlock of the same type as a result of
PMEM connect action.
Two different file pointers have different internal data
pointers and different spinlock instances. But, the spinlocks are
of the same class, hashed based on their initial name which is the
same for all internal file data spinlocks. Thus, grabbing more than
one internal data spinlock causes kernel debugging code to activate
due to restrictions in the kernel deadlock detection debug code on
acquisition of more than one spinlock of the same type or class.
Fix here is to save needed data from the private data of the source
file into a local variable under spinlock, then relinquish that
spinlock immediately before then acquiring the spinlock of the other
file pointer to store that saved data in the private data of the
other file pointer.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e3a477ecec50290c1aa629eb4dca2c75c576272e
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Wed Aug 12 10:36:24 2009 -0700
pmem: add more error checking in connect function.
Make sure that caller error cases like passing in file descriptor
that has the same file pointer as the passed in file pointer and
other issues (such as a null file pointer) are caught and rejected.
This fixes a recursive spinlock kernel oops that happened because
connect was invoked with the file descriptor having the same file
pointer as passed-in.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6093f729706a4e024172a43867aef1618e621cdb
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Fri Jul 24 13:27:13 2009 -0700
pmem: Make number of allocated regions dynamic
Make the number of allocated regions dynamic by growing the array
keeping track of allocations by a power of 2 every time needed,
failing only when system memory is exhausted.
This change fixes the problem of hardcoded value of 64 allocations
being too few for many use cases; any more allocations than 64
without a free first was causing an artificial out-of-memory
condition.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4cd40ae333f1680f5ebfe66ddfbad4baaf28ee24
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Wed Jun 17 15:54:48 2009 -0700
pmem: add api for PMEM alloc and free from kernel space
int32_t pmem_kalloc(const size_t size, const uint32_t flags);
Attempts to allocate contiguous physical memory with the
specified alignment from the requested memory type.
This returns a physical address aligned to either 4k page boundary
or 1M boundary (if specified in the flags argument) or a negative
error number (i.e. TRUE when IS_ERR() is invoked on the return
value) if the allocation failed.
Size of the returned memory will be rounded up to the next full 4K
page.
Possible errors are -EINVAL (bad flag or size) and -ENOMEM (out of
memory).
Parameters:
size - is specified in bytes.
flags - are an or'ed bitmap of MEMTYPE and ALIGNMENT values which
define the type of memory requested and the alignment of the
allocated memory:
Valid values for MEMTYPE are:
PMEM_MEMTYPE_EBI1 (currently the only allowed type for in-kernel
allocations)
Valid values for ALIGNMENT are:
PMEM_ALIGNMENT_4K (the default) and PMEM_ALIGNMENT_1M
----------------------------
int32_t pmem_kfree(const int32_t physaddr);
This frees memory previously allocated by pmem_kalloc.
The entire size allocated starting at this address will be freed.
This returns 0 if the free succeeded or a negative kernel error
number if it failed.
Possible error values may be -EINVAL if a bad physaddr was provided.
"physaddr" must be a valid value that has been returned from
pmem_kalloc. This will be checked.
If, for some reason, a passed-in value manages to pass all the
internal checks but is an invalid value, and is submitted to the
freeing functions, undefined behavior may ensue, possibly including
corruption of memory arenas.
Example setup of the board files have been provided as part of this
submission.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 63aede7834f50b290962e501c091f8f52bf2affc
Author: Stephen Biggs <sbi...@quicinc.com>
Date: Tue Jul 7 13:31:59 2009 -0700
pmem: Add bitmap quantasized allocation algorithm.
This new algorithm is necessary due to limitations in the current two
algorithms (all-or-nothing, or buddy/bestfit "power of 2") which can
lead to waste and/or fragmentation of memory space on odd size
allocations.
This new algorithm should be used if odd allocation sizes are needed
that require more efficient used of memory and finer control of
granularity. As used from user space, upon successful allocation,
this algorithm will give contiguous physical memory with size
rounded up to the nearest 4K page.
The format for the board file PMEM specifications has changed. The
"no_allocator" field of unsigned type in "struct
android_pmem_platform_data" has been replaced with field
"allocator_type" of type "enum pmem_allocator_type" which is newly
defined in include/linux/android_pmem.h. The allocator types are
now PMEM_ALLOCATORTYPE_ALLORNOTHING,
PMEM_ALLOCATORTYPE_BUDDYBESTFIT, and PMEM_ALLOCATORTYPE_BITMAP.
Thus:
static struct android_pmem_platform_data foo {
/* .no_allocator = 0, becomes: */
.allocator_type = PMEM_ALLOCATORTYPE_BUDDYBESTFIT,
};
and...
static struct android_pmem_platform_data bar {
/* .no_allocator = 1, becomes: */
.allocator_type = PMEM_ALLOCATORTYPE_ALLORNOTHING,
};
Note that all board file allocator types that were
".no_allocator = 0" that now have the default of
PMEM_ALLOCATOR_BUDDYBESTFIT.
Signed-off-by: Stephen Biggs <sbi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6e45cd9418b914944d5dad43e5948ff2b5bbc180
Author: Larry Bassel <lba...@quicinc.com>
Date: Thu Jun 11 17:13:34 2009 -0700
pmem: add ioctls to clean and/or invalidate caches
Three new ioctls have been added:
PMEM_CLEAN_INV_CACHES which will clean and invalidate
all caches for all or part of a PMEM region,
PMEM_CLEAN_CACHES which will clean
all caches for all or part of a PMEM region,
and PMEM_INV_CACHES which will invalidate
all caches for all or part of a PMEM region,
This also adds the low level support for these
ioctls.
Signed-off-by: Larry Bassel <lba...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 22d0d3e9bdc109a1f2b5d009dd9f5e4882bebcf4
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Wed Jan 13 15:56:49 2010 -0800
pmem: various updates.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d4f639f39debfb5ac4c9008f281d0ea059e75d91
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Fri May 8 09:30:34 2009 -0700
[ARM] msm: pmem: PMEM_DEBUG defined according to defconfig
PMEM_DEBUG is defined with xxxx_defconfig and undefined with
xxxx-perf_defconfig
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 816631250c0f7460cc6e1786228026c10b7a8e9d
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Wed Jan 13 15:47:51 2010 -0800
pmem changes from Initial Contribution.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2cc1b4f06330a9be7ba7181a5912ab0bd31521b0
Author: Rebecca Schultz <rsch...@google.com>
Date: Thu Jul 24 11:22:53 2008 -0700
pmem: Add pmem driver
Signed-off-by: Rebecca Schultz <rsch...@google.com>
pmem: Use the thread group leader insted of the current thread.
Instead of keeping track of the current thread, use the thread group leader
Signed-off-by: Rebecca Schultz <rsch...@google.com>
pmem: Add some apis to reference and flush pmem files by file struct
The api to refer to pmem files by fd should be depricated, it can
cause problems if a processes fd table changes while the kernel is processing
data in a pmem file. This change adds the safer api.
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
pmem: Remove unused depricated fd api to pmem.
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
pmem: Remove error message when calling get_pmem_addr
This call is used from the mdp driver to determine if the memory
is in pmem or in the fb. We will encounter this case during normal operation
so this error message should be removed.
Signed-off-by: Rebecca Schultz Zavin <reb...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e7a48576d26a2d89c322e7008c922f7cc7dc0fd7
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Mon Nov 9 19:41:37 2009 -0800
msm: change dma_coherent_pre/post_ops() to use COHERENT_IS_NORMAL
Clean the #if nesting by using the COHERENT_IS_NORMAL flag. Introduce a
compiler barrier() in the pre case when COHERENT_IS_NORMAL is 0 and arch is
not coherent. Note that for Xscale we will have to force dmb() as it uses
kmalloc for coherent memory.
Change-Id: I1753fc62f5dfa3333c65269ab1815cd29e5698f7
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5a8656119546493eac4a0699afec1a429ca584a5
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Wed Nov 4 11:40:22 2009 -0800
mm: Make coherent memory as normal noncached on all MSM's
Since Normal memory mapping are less restrictive than Strongly Ordered,
they are faster and should be used as coherent memory. Change coherent
to Normal for all MSM's. Change the dma_coherent_pre_ops and
dma_coherent_post_ops to do a dmb() on all MSMs.
Change-Id: I9e7d96e58ce9af2c42c2d4b880071adbf1413f5c
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c79f3a6aee487fda2d3ba4df3d6343797ad5a37e
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Tue Nov 10 11:31:27 2009 -0800
system: update pgprot_device to use pgprot_modify
pgprot_device is used for ioremapping peripheral addresses.
Use pgprot_modify macro and also mark the memory as not executable.
Change-Id: Ib81de0875ca297775bdeb33c321b8baac2201507
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b500e196b52d83af96bc73c3739c30853d7e97af
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Wed Dec 16 13:05:34 2009 -0800
mm: create pgprot macros to support various cacheable memory types
The framebuffer driver needs to map memory that has either
writethrough cache, writeback cache or writeback write
allocate cache attribute. Add macros to support such
mapping types.
Change-Id: I89b76779b6d2352e1d5a1d8b9555c092fc9a7904
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 433a5150d7599383a758b3c22e7f614ecbde3c37
Author: Russell King - ARM Linux <li...@arm.linux.org.uk>
Date: Wed Oct 14 12:28:38 2009 -0700
ARM: dma-mapping: switch ARMv7 DMA mappings to retain 'memory' attribute
On ARMv7, it is invalid to map the same physical address multiple times
with different memory types. Since system RAM is already mapped as
'memory', subsequent remapping of it must retain this attribute.
However, DMA memory maps it as "strongly ordered". Fix this by introducing
'pgprot_dmacoherent()' which provides the necessary page table bits for
DMA mappings.
Signed-off-by: Russell King <rmk+kernel at arm.linux.org.uk>
Change-Id: Ic0f02f8e85f02abf6ad1542bb23fbefd13122c00
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e03852c2e9e6ad2d0124a268591457615c30fa8e
Author: Larry Bassel <lba...@quicinc.com>
Date: Wed Dec 2 16:40:31 2009 -0800
Include tlbflush.h in pgtable.h
The page migration code calls update_mmu_cache(), which on ARM is
defined in arch/arm/include/asm/tlbflush.h.
Change-Id: Ib5ee7520f82fb1c6a3e342db6a2971053bfcd3dc
Signed-off-by: Larry Bassel <lba...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d16ae02af681904d11230dbd0e658cbeb3726750
Author: San Mehat <s...@google.com>
Date: Mon Mar 31 20:38:24 2008 -0700
msm: RTC driver for Qualcomm MSM chipsets.
The MSM devices contain an RTC that is accessed via an RPC call
to the modem processor.
Original-patch-by: San Mehat <s...@android.com>
Cc: Murali Palnati <paln...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d24d76b4e22de254d8284285b176817a53aaaee4
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri Jul 25 20:58:15 2008 -0700
rtc: Try to prevent RTC errors from accumulating.
When we resume we only know how many whole seconds has elapsed.
These errors would accumulate in delta. We now only set the delta
if it would change by more than two seconds. If we drift back by
by more than a second add one in resume.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 369101f97b154dbd3429d5ed723403e0e527c70e
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Mon Dec 21 15:17:04 2009 -0800
Remove more driver related stuff.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7e5596e33797739ec64250285ac9777c3a708a04
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Mon Dec 21 11:27:15 2009 -0800
Add missing bits for compilation.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 786a34f8967793e48963dc7d1b9375d8b3ddac72
Author: Dave Estes <ces...@quicinc.com>
Date: Thu Jun 25 18:03:02 2009 -0400
[ARM] vfp: Add additional vfp interfaces
Refactor common code to vfp_flush_context() and vfp_reinit(). Allow
use by other client beside suspend/resume. Currently intended for
idle power collapse.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cb45e9ffea20d93e6c10c5a23ccca40b7d612a8a
Author: Praveen Chidambaram <pchi...@quicinc.com>
Date: Thu Jun 18 16:15:42 2009 -0600
pm_qos: Add QoS param, minimum system bus frequency
In some systems, the system bus speed can be varied, usually
based on the current CPU frequency. However, various device
drivers and/or applications may need a faster system bus for I/O
even though the CPU itself may be idle.
Signed-off-by: Praveen Chidambaram <pchi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 361a8842267da7d8bf9119c25b6768a11c7e37d9
Author: Jeffrey Ohlstein <johl...@quicinc.com>
Date: Thu Oct 1 15:30:40 2009 -0700
sched: Extend completion api to allow io_wait time tracking
Adds a function wait_for_completion_io which behaves like
wait_for_completion, except it calls io_schedule instead of schedule. This
indicates that the process waiting on the completion is waiting on an io
event, and keeps statistics accordingly.
Signed-off-by: Jeffrey Ohlstein <johl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 115107532f033650c48c6ed114e31f592c017468
Author: Jeff Ohlstein <johl...@quicinc.com>
Date: Wed Nov 18 14:22:08 2009 -0800
msm: gpio: Fix misnamed ifdefs
CONFIG_MACH_SMD appears several time in our gpio driver. This symbol
is incorrect, and the right feature name is CONFIG_MSM_SMD.
Signed-off-by: Jeff Ohlstein <johl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 09088bb61bb81aca07c7789dcffa0456be318fac
Author: David Brown <dav...@quicinc.com>
Date: Thu Nov 5 13:56:10 2009 -0800
msm: Add QSD8x50 ST1 machine type.
Change-Id: Iaa42379f24716be07f27b797605d2309683b1c4d
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit aef6ce18090e297591f2848367e26cd15ae339aa
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Sep 16 16:41:47 2009 -0700
msm: add config option for grapefruit platform
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2a2cd1987c9e4e64d2273f1b860b4fdb6efafca1
Author: David Brown <dav...@quicinc.com>
Date: Tue Sep 15 15:02:36 2009 -0700
msm: add grapefruit qsd8x50 support.
Add machine-type support for the qsd8x50 "grapefruit" internal
development platform.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2d9897ac69ef2e9d61a5cc32301c2015ab4a4210
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Mon Aug 31 15:39:27 2009 -0700
mach_msm: dma: Use an ISR safe disable IRQ call
Kernel 2.6.31 brought out a latent bug in the dma ISR,
disable_irq was being used instead of disable_irq_nosync.
This caused the kernel to lock up at boot.
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2bbf03b3f830c43f0dd0c8f33730c8cfd98221c6
Author: David Brown <dav...@quicinc.com>
Date: Mon Aug 24 15:54:16 2009 -0700
mach_msm: Conditionalize SMD support in GPIO code.
Remove dependency on SMD from GPIO code. Modem CPU will not be
notified of GPIO states for proper sleep without this feature.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ec96df3bc7595eedcc316e2e885f9b73d6f1dcbe
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Oct 29 16:06:37 2009 -0700
msm: clock: Add support for AXI_ROTATOR and HDMI clocks.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 927f0aad58171204e6610fd69279793dd6826764
Author: Willie Ruan <wr...@quicinc.com>
Date: Sun Oct 18 16:12:42 2009 -0700
msm: add pmic8058-mpp.c to support PM8058 MPP in gpiolib framework
PM8058 has MPPs which can be used by Linux drivers. The goal is to use
the kernel's gpiolib but right now we have to use Google's gpiolib
because there can't be two gpiolib implementations. Porting to kernel
gpiolib will occur after Google's generic_gpio is ported.
Only digital input function of a MPP is supported. Drivers should have
configured PM8058 MPPs using mpp_config_digital_in before calling
gpiolib APIs.
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 37ac2a6c639747a4de251fd5ab230a4e97f68485
Author: Willie Ruan <wr...@quicinc.com>
Date: Sun Oct 18 15:03:04 2009 -0700
msm: add pmic8058-gpio.c to support PM8058 GPIO in gpiolib framework
PM8058 has GPIOs which can be used by Linux drivers. The goal is to use
the kernel's gpiolib but right now we have to use Google's gpiolib
because there can't be two gpiolib implementations. Porting to kernel
gpiolib will occur after Google's generic_gpio is ported.
Drivers should have configured PM8058 GPIOs using pm8058_gpio_config
with full parameters before calling gpiolib APIs.
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 79e741d613902e3bef428e30d0f960c7f624eba2
Author: Ai Li <a...@quicinc.com>
Date: Thu Oct 15 19:12:33 2009 -0600
msm: timer: use tiered method to sync DGT after power collapse
The time-sync protocol is not intended to support DGT. When sclk wraps,
it can cause DGT to jump. In the tiered method, GPT syncs to sclk via
the time-sync protocol, and DGT syncs to GPT. Because both DGT and GPT
are local, syncing DGT to GPT is fast and can occur more often to detect
GPT wrapping around.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a4e694f762d4de30cc3b23cc2025aa3bab3e13d3
Author: Karthik Parsha <kpa...@quicinc.com>
Date: Wed Oct 14 17:36:27 2009 -0700
msm: acpuclock-8x50: Remove support for frequencies based on PLL1.
PLL1 might not always be on. Also, it doesn't offer any frequencies that
are useful when compared to the rest of the available frequencies. Hence
removing frequencies based on PLL1.
Change-Id: I823906bdbd137b883cfd116f425735eaae555a09
Signed-off-by: Karthik Parsha <kpa...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a3695edfd8e6add78e1fa5fd2643147024f6557f
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Oct 7 12:37:28 2009 -0700
msm: clock-7x30: Add support for local PLL enable/disable.
The 7x30 SoC includes h/w that allows each processor to vote on
enabling/disabling a PLL. Use the voting h/w to enable/disable PLLs
instead of making proc-comm calls to the modem processor.
Also, create an API that can be called from the acpuclock code.
Change-Id: I825d11787c3310adcd20aa4db65a077c906fcc7d
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f04b049bf5ca9b511bf9acfdbc4c150aa677cb16
Author: Chandan Uddaraju <chan...@quicinc.com>
Date: Thu Oct 22 11:44:15 2009 -0700
msm: socinfo: add SoC ID of 61 of ESM7x27 to 7x27 list.
ESM7x27 is a EDGE target of 7x27 family. It is using a new SoC id.
Signed-off-by: Chandan Uddaraju <chan...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0d335a14076692f78dd99c543e99225c2d3f06b0
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Oct 7 15:35:23 2009 -0600
msm: Add module to test proc comm system
Add moulde to test the proc comm system.
Module sends a predefined proc comm command to modem
and expect a known return.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4746b95bfa93a7753ed628a71421636b1db3de32
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Oct 16 17:08:27 2009 -0700
msm: clock-7x30: Add image rotator clock support.
Change-Id: I4e07e0da136889f2741a2e74463c04e146fcad75
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e05dad56f41970543537687af9d68131bf6ae591
Author: Willie Ruan <wr...@quicinc.com>
Date: Sat Oct 17 23:15:14 2009 -0700
msm: gpio: move two msm_gpio structs from gpio.c to gpio_chip.h
There is a requirement to add PM8058 GPIOs and MPP(Multi Purpose Pin)s
on 7x30 SoC to kernel gpio framework. To allow another file to really
use register_gpio_chip() these structs must be moved to gpio_chip.h.
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bc443da3951df0b6846f940d0accf0a679a5dca8
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Oct 12 13:24:13 2009 -0700
msm: clock: Implement clock reset API.
Some drivers require the ability to assert and deassert reset signals
to hardware blocks associated with clocks. An API is added to do this
using proc_comm.
Change-Id: Ib29fa5b5fae6e4892ada2015b831583d5b8d87f4
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9311a895e06892b9267d1eea3ab35a8981e28818
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Oct 7 12:17:37 2009 -0700
msm: acpuclock: Move CPU-specific defines to correct driver files
CPU-specific defines are moved from the common acpuclock.h header
file into the CPU-specific drivers. Doing so keeps acpuclock.h
generic, allowing it to be used for all acpuclock drivers.
Change-Id: I4cf9633271a077d31dc793e96f0b721bab286e04
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6db3ab93ffc4c1f01349b86dd73072e7a9da5b3a
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Oct 20 17:49:10 2009 -0700
msm: internal_power_rail: Add MFC power rail control support
Change-Id: I040d655baf470a0cb276dab1b5f06872614a1c3b
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 176cb0d9717abc895a4c059ec5fe7154e131fec8
Author: Michael Bohan <mbo...@quicinc.com>
Date: Tue Oct 13 15:10:25 2009 -0700
msm: Add MSM7x30 FFA and MSM7x30 FLUID board types
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 112aa2462b408119a31273e40ca0b1a5d84d693b
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Oct 7 12:46:11 2009 -0700
msm: iomap: Add 7x30 ACC register region
The ACC register region contains Scorpion CPU clock control/distribution
registers. It is required for Scorpion clock source selection.
Change-Id: I1bcb412043f7cea403387fe249cf9be3fa5c7459
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b61d055762cad29a25b1133dcca8a6757c28f71e
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Oct 16 14:43:43 2009 -0700
msm: clock: Put clock lists in alphabetical order
Also, rename the 7x30 'cam_mclk_clk' clock to 'cam_m_clk', to
comply with the naming convention.
Change-Id: If703b93cbcff26e4baf48970b1d5fa2b25b0b0f5
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6ceb695ccd4915b61960009ece663872e70ed850
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Oct 16 14:11:16 2009 -0700
msm: clock: Fix proc_comm 'OFF' flags for several 7x30 clocks
Add 'OFF' flags to several clocks. Missing 'OFF' flags for clocks
that are not enabled at boot can cause crashes when attemping to read
clock enable/disabled status (from debugfs, for example) if the clock
is not enabled first.
Unnecessary 'OFF' flags are also removed from several clocks that are
known to be on at boot.
Change-Id: I7c952414d4cac68d51d0c9f6827928f8fc20c3ad
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 50c7b8fe9e402bbc897b86a2ccc3512c75f48f00
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Oct 16 10:22:56 2009 -0700
msm: acpuclock-8x50: Change "a11" naming convention to "acpu"
8K application processors are Scorpion-based, not ARM11-based.
Replace the ARM11 naming convention with something more generic.
Change-Id: Ie77b9cdc354f988ffe4446fda8c430b8f5f61142
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3914f317e4cd1783c6092dc49372fc10908a707d
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Oct 16 10:04:08 2009 -0700
msm: acpuclock-8x50: Clean-up PLL0 fix-up
The PLL0 fix-up should not assume there is only one PLL0 row in the
frequency table. The fix-up now checks the row's frequency before
changing it.
Change-Id: I1a459cfb916cf8bf8a84dc7aac9ba7745dff6978
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5112c9b31c16763c23af3ae87458e52819d0d976
Author: Willie Ruan <wr...@quicinc.com>
Date: Mon Oct 12 22:40:44 2009 -0700
msm: mpp: add MPP as digital input configuration API
New hardware design, like 7x30 FLUID, needs to use MPP as digital
input on application processor. FLUID uses one MPP for flip detection.
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 16eb3c8baccb84d4f74fdee81fe9f6ae7ed29520
Author: Willie Ruan <wr...@quicinc.com>
Date: Mon Oct 12 22:37:25 2009 -0700
msm: proc_comm: fix typo of POCM for PCOM
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2243d21be0cc4ab883ffe67ff8cbd64ca7a653aa
Author: Michael Bohan <mbo...@quicinc.com>
Date: Thu Oct 8 15:34:19 2009 -0700
msm: socinfo: Add support for hw_smem v3
Version 3 includes support to show which 'hardware platform'
we're runnning on. Current possibilities are SURF, FFA, and
Fluid.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eb74e8c457d03dbbefa188175c4ff2b58908878a
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Oct 7 15:33:22 2009 -0600
msm: Add OEM proc comm commands enumeration
OEM proc comm commands start from base 0x10000000
Thie update helps OEMs to add their own commands easily.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 56bbe1ef18c1264a6e0b45f236bbf0e1ec160297
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Oct 7 14:48:12 2009 -0600
msm: Handle SMD packets with 0 length
Users might call smd_write with lenght 0 in which case
SMD should not write anything (including packet header).
Also, if SMD receives a 0 length packet from other processor,
it should skip it and move on to next packet.
CRs-fixed: 212417
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c19c8c9ed992f1bd4d34b84c291513be47446481
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Oct 5 17:22:26 2009 -0700
msm: timer: Remove MSM_DGT_SHIFT for Scorpion processors
All 32 bits of Scorpion debug timer "COUNT_VAL" registers are useful,
unlike ARM11 debug timers, for which the least-significant bits are
unreliable and marked as reserved.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 92c1d267ec94395af9063242b28e6367dd17e34a
Author: Matt Wagantall <ma...@quicinc.com>
Date: Wed Sep 16 10:03:56 2009 -0700
msm: timer: Fix 7x30 debug timer register offsets and clock source
The debug timer's register addresses are corrected. To avoid future
confusion, the MSM_GPT_* iomap macros are renamed to MSM_TMR_*. This
clarifies that they represent the address of the first timer register,
which is not necessarily the first GPT register on all targets.
Additionally, the debug timer frequency is changed to 24.576 MHz, since
the 7x30 uses LPXO as its debug timer clock source.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6bb5106dea99585792e0b4c07fb199c792f3e18f
Author: Jeffrey Ohlstein <johl...@quicinc.com>
Date: Thu Oct 1 16:56:30 2009 -0700
msm: dma: Track waiting for requests as io_wait time
Using wait_for_completion_io instead of wait_for_completion improves the
kernel's statistics about time spent waiting for io.
Signed-off-by: Jeffrey Ohlstein <johl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2146057e640bf30ef8199227b7633eb5c16c34db
Author: Ruigrok, Richard <rrui...@quicinc.com>
Date: Fri Oct 2 14:26:02 2009 -0600
msm: smd_rpcrouter: fix bug in pacmark messages sequence.
Bug in router causes packet corruption with multiple packet
message is pre-empted with a 2nd write of single packet message.
The first message (A) being a multi-packet message is started.
which contains packet A1 and A2.
A second message (B) is sent during the write sequence of message (A)
between A1 and A2 packets. In the buggy version, the packet (B1) would
use the same MID as (A1) and would cause message (A) to be
corrupted at the far end when re-assembled. Also more, the
last packet of (A2) would have a different mid, causing the far-end
to never receive the completed message. The failures caused by sending
an ill-formatted packet at the far-end are undeterministic. On the local
end, The user-space thread will be locked as no reply will be received.
CRs-fixed: 211851
Signed-off-by: Richard Ruigrok <rrui...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 41a98f23c79415a0d06c0808405f299a734350a3
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Sep 25 11:43:10 2009 -0700
msm: clock: Add camera clocks to 7x30 clock list
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 25758c79091ae185e2d56806014db943cd63338b
Author: Ai Li <a...@quicinc.com>
Date: Mon Aug 31 12:33:55 2009 -0600
msm: timer: set SMSM bits for time sync during initialization
Setting the bits during initialization also removes the need to set
them again during each power collapse.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 52d47a4b78798ebe130e2833c569c0de920f6160
Author: Ai Li <a...@quicinc.com>
Date: Mon Aug 31 12:11:59 2009 -0600
msm: clear SMSM interrupt masks during initialization
By the time power collapse or time sync occurs, kernel has already
disabled interrupts. Both power collapse and time sync protocols
poll for the SMSM bit changes. Clear the SMSM masks so that Modem
don't bother sending the unnecessary interrupts.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 237adaf4113ee5f5aec12a57dcbb7a733c832664
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Sep 15 13:39:25 2009 -0700
msm: Add API for clock power rail control using proc_comm
Previously, an API did not exist to control power rails independent of
clocks. The new API provides this functionality using two exported
interfaces:
int internal_pwr_rail_ctl(unsigned rail_id, bool enable);
int internal_pwr_rail_mode(unsigned rail_id, enum rail_ctl_mode mode);
where mode is either PWR_RAIL_CONTROL_AUTO or PWR_RAIL_CONTROL_MANUAL.
If the control mode is never set, the rail will be controlled automatically
(ie. turned on and off with the clock).
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f565b5ac2771fb6b03c3dedb57089b082adeddc1
Author: Ramesh Garimella <c_rg...@qualcomm.com>
Date: Thu Sep 24 18:32:33 2009 +0530
msm: dma_test: allocate minor number dynamically
CRs-Fixed: 210475
Signed-off-by: Ramesh Garimella <c_rg...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5390fdb13559e2c3357d3bf3246aa98aa53e81d3
Author: Lena Salman <esa...@qualcomm.com>
Date: Tue Sep 29 10:40:00 2009 +0200
msm: dma: Add crypto engine DM channels.
SPI core can use crypto channels as predetermined by the SPI mux value.
Since SPI core doesn't own its own channels, the mux will determine which
channels it will take - crypto, uart1 or uart2.
Signed-off-by: Lena Salman <esa...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 51d109722b53510b25fbd7b9739dae2d7a07d5d8
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Sep 11 17:36:41 2009 -0700
msm: remote-spinlock: Remove WFE instruction.
Use of the WFE instruction avoids excessive polling of memory by putting
the application processor in a wait state. The application processor,
however, expects the remote processor to execute an SEV instruction to
cause the application processor to come out of the wait state.
The remote processors do not support the SEV instruction (older ARM
architectures, etc.) and so using the WFE instruction can lead to the
application processor waiting forever.
Hence, the WFE instruction is removed in favor of continuously polling
the remote spinlock's location in memory.
CRs-fixed: 211016, 211034
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dfc2899805027819bbe21ff50d21c85c2a1c8417
Author: Dave Estes <ces...@quicinc.com>
Date: Mon Aug 24 14:12:42 2009 -0400
msm: qsd8x50: Add adaptive voltage scaling (AVS)
The purpose of AVS is to minimize the power used by the CPU by
determining and setting the optimal voltage. The optimal
voltage is actively determined for each frequency and
temperature. AVS actively adjusts the CPU voltage as the
CPU frequency and temperature changes.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 99138aaf8e35b21e0481195a919befcc3933928e
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Sep 24 13:27:09 2009 -0700
msm: acpuclock-8x50: Fix PLL0 speed for CDMA-only targets
PLL0 runs at 235MHz instead of 245MHz for CDMA-only targets.
The corresponding ACPU frequency table entry is corrected during
initialization if the slower rate is detected.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4750802c7162f3b161ef034570b713c3acaccdc5
Author: Pavankumar Kondeti <pkon...@qualcomm.com>
Date: Tue Sep 22 13:39:48 2009 +0530
msm: Fix UART base address for MSM7x30
CRs-fixed: 210856
Signed-off-by: Pavankumar Kondeti <pkon...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 86d648adfa15b636f52b05c86892f0c34619662f
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Sep 22 11:23:13 2009 -0700
msm: clock: Add rotator clocks to 7x30 clock list
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 222a8ec2daabc0142442a109ae93615f5546e95f
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Sep 8 11:26:56 2009 -0700
msm: acpuclock: Grab lock when changing ACPU freq.
It is possible for multiple threads to concurrently execute
acpuclk_set_rate(). To prevent race conditions, code using data from
the drv_state struct must be locked.
The mutex is acquired prior to accessing drv_state data, but only when
acpuclk_set_rate() is called for a reason other than Power Collapse or
SWFI. In these situations a mutex cannot be acquired because interrupts
are disabled, and is not needed because the CPU is known to be idle.
CRs-fixed: 194883
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c3bf5354c91b968c58f8a24562e48f256d80a0ce
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Aug 21 12:14:41 2009 -0700
msm: acpuclock: Rewrite frequency stepping
For certain configurations of SoC and modem firmware, given a starting
frequency, the list of optimal stepping frequencies to use will vary
depending on the target frequency. This is now handled by having, for
each frequency in the table, one 'up' and one 'down' entry for each PLL.
The up/down indices are also replaced with pointers to the corresponding
table entries. These pointers are calculated during boot since manual
maintenance of indices is error prone.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c0ff0fa7cebcc37272fc59b0d5a1c6966730c113
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Sun Aug 23 15:22:52 2009 -0600
msm: Add more support in RPC server infrastucture
Add machanism to send payload along with reply message.
Add support for servers to do callback requests.
1)The below interface is to be used to initiate a reply message
void *msm_rpc_server_start_accepted_reply(struct msm_rpc_server *server,
uint32_t xid,
uint32_t accept_status);
where,
'server' is pointer to server data structure,
'xid' the xid; has to be same as in request message,
'accept_status' is acceptance status.
it returns pointer to buffer where payload can be attached.
2)The below interface is to be used to send a reply message
int msm_rpc_server_send_accepted_reply(struct msm_rpc_server *server,
uint32_t size);
where,
'server' is pointer to server data structure,
'size' is size of the payload attached.
3)The below interface to be used to do a callback request
int msm_rpc_server_cb_req(struct msm_rpc_server *server,
struct msm_rpc_client_info *clnt_info,
uint32_t cb_proc,
int (*arg_func)(struct msm_rpc_server *server,
void *buf, void *data),
void *arg_data,
int (*ret_func)(struct msm_rpc_server *server,
void *buf, void *data),
void *ret_data, long timeout);
where,
'server' is pointer to server data structure,
'clnt_info' is pointer to client information data structure.
callback will be sent to this client.
'cb_proc' is the callback procedure number.
'arg_func' is argument function pointer. 'buf' is where arguments needs
to be filled. 'data' is arg_data
'ret_func' is return function pointer. 'buf' is where returned data should
be read from. 'data' is ret_data
'arg_data' is passed as an input parameter to argument function.
'ret_data' is passed as an input parameter to return function.
'timeout' is timeout for reply wait in jiffies.
If -ve timeout is specified a default timeout of 10s is used.
it returns 0 on success, otherwise an error code is returned.
4)The below interface to be used to get requesting client's information.
This information is needed while doing callback requests.
void msm_rpc_server_get_requesting_client(
struct msm_rpc_client_info *clnt_info)
where,
'clnt_info' is pointer to client information data structure.
user of this call has to handle allocating and freeing memory
for this data structure.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e41efa802bb4e7804ebb3a03a9943845f504f283
Author: Dave Estes <ces...@quicinc.com>
Date: Fri Aug 21 14:41:48 2009 -0400
mm: Add SW emulation for ARM domain manager feature
Do not set domain manager bits in cp15 dacr. Emulate using SW. Add
kernel hooks to handle domain changes, permission faults, and context
switches.
This feature is required by ARCH_QSD8x50 to fix a problem with page
crossing memory accesses. Set ARCH_QSD8X50 to select emulator.
CRs-fixed: 193589
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2156f14b81ab28c52dd050ee275a779e528f3bbe
Author: Dave Estes <ces...@quicinc.com>
Date: Tue Aug 11 17:58:49 2009 -0400
mm: qsd8x50: Fix incorrect permission faults
Handle incorrectly reported permission faults for qsd8650. On
permission faults, retry MVA to PA conversion. If retry detects
translation fault. Report as translation fault.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bf9348875ab8032d56c722bafcc6a128b01b1c3e
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Sep 1 12:18:31 2009 -0700
msm: acpuclock: Restore VDD when returning from power collapse
Restore the apps processor's VDD level when coming out of power collapse,
since the modem may have changed it after the apps processor went down.
CRs-fixed: 193493
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 382e82734b64988e5446ca1854da878114584259
Author: Anoop Krishnan Jayasankaran <c_aj...@quicinc.com>
Date: Mon Sep 14 11:57:55 2009 -0700
msm: acpuclock: Reduce AXI freq to 61.44 MHz at ACPU freq of 245.76 MHz
Reduce AXI freq to 61.44 MHz when ACPU runs at 245.76 MHz.
This provides significant power savings during audio playback.
Tests indicate that this change doesn't impact performance since
the actions of other drivers end up increasing the AXI frequency
when the display is on.
Signed-off-by: Anoop Krishnan Jayasankaran <c_aj...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dc63913933d0471f5ce17784b6c3c7e55271099a
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Sep 11 01:17:21 2009 -0700
msm: clock: Fix AXI rate voting use of magic value
PM QOS passes only s32 integer rates through to the clock driver. The
MSM_AXI_MAX_FREQ magic value, which is used to vote for the maximum
supported AXI bus rate, is changed to respect this restriction.
The behavior and use of the magic value is also changed so that it
represents the maximium AXI rate in the correct units, regardless of
whether it's passed to ebi1_clk_set_min_rate() directly, or used
with PM QOS.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d5ffc6e2e65138f86b249eb05beeb05947c9fe32
Author: Stephen Boyd <sb...@quicinc.com>
Date: Wed Jul 22 10:23:18 2009 -0700
msm: smd: split into separately configurable pieces
Some users don't want the whole SMD package, instead they want
just the bare essentials. Add kconfig options to compile out
pieces of the SMD driver. The new options cover: MSM_SMD_TTY,
MSM_SMD_QMI, and MSM_SMD_CTL.
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f4a2c29b3455a719633658f0c3993d56d0451074
Author: Stephen Boyd <sb...@quicinc.com>
Date: Tue Jul 21 15:31:21 2009 -0700
msm: Add kconfig option to disable NMEA driver
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2bb8825dfb6b215a035718a5788e56c8fd59b8bf
Author: Stephen Boyd <sb...@quicinc.com>
Date: Thu Jul 16 14:39:42 2009 -0700
msm: allow users to disable smem logger
add/remove: 0/38 grow/shrink: 0/8 up/down: 0/-17104 (-17104)
This also slightly reduces kernel init time by removing
smem_log_init().
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1c1927694af67bf216724a62044306c4ae08f057
Author: Lai, Patrick <pl...@quicinc.com>
Date: Tue Sep 8 17:39:00 2009 -0700
msm: clock: Add MI2S codec clocks in 7x30 clock list
Signed-off-by: Patrick Lai <pl...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 02308e7002832b7aca17e50880ac5b5175fa3e15
Author: Ruigrok, Richard <rrui...@quicinc.com>
Date: Tue Sep 8 17:08:07 2009 -0600
msm: proc_comm add commands for clkctl
PCOM_CLKCTL_RPC_RAIL_ENABLE
PCOM_CLKCTL_RPC_RAIL_DISABLE
PCOM_CLKCTL_RPC_RAIL_CONTROL
PCOM_CLKCTL_RPC_MIN_MSMC1
Signed-off-by: Richard Ruigrok <rrui...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b06e35aa8b5cfa40435e997803f0a5ff56d87bb8
Author: Ruigrok, Richard <rrui...@quicinc.com>
Date: Thu Sep 3 17:16:21 2009 -0600
msm: rpc: Add support for PING control message.
Modem processor may send a PING control message
if the application processor does not respond for more
than 20ms to a flow control request.
Support for this feature only requires that we receive the message.
Signed-off-by: Richard Ruigrok <rrui...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f0067c7195a74370338ad26a261c76aee06b0db7
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Sep 3 15:12:33 2009 -0600
msm: Check for correct RPC version and program in callback request
Check for correct RPC version and program in callback request.
This check is done if callback is processed in a separate thread.
Add the check when callback is processed in client's read thread.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 81c9992a78debdc4b5f0aebed28208f8fc23efd4
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Sep 3 14:49:14 2009 -0600
msm: Change some SMD info messages to dbg messages to reduce logging
Change some SMD info messages which are of low importance
to dbg messages. By default only info messages are turned on.
Hence will reduced smd message logging.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4c6657e08020159792261553ff008e05174f60e1
Author: Willie Ruan <wr...@quicinc.com>
Date: Thu Sep 3 11:44:29 2009 -0700
msm: vreg: fix a typo for lvsw1
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 14300d50726449d9b0361ed1a131ec4507fe5b47
Author: Matt Wagantall <ma...@quicinc.com>
Date: Tue Sep 1 17:30:30 2009 -0700
msm: clock: Add SoC/board independent APIs to set/get max AXI frequency.
Some drivers need to set the AXI frequency to the maximum frequency
supported by the board. Add a MSM_AXI_MAX_FREQ_KHZ magic value that
allows them to achieve that in a SoC/board independent manner.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 233ec50c90f5dd3dabdf3b259633ae393023ea66
Author: Matt Wagantall <ma...@quicinc.com>
Date: Fri Aug 28 16:00:41 2009 -0700
msm: clock: Correct MI2S clock divider for 7x30
The value written to the MI2S clock divider register is corrected so
that it accomplishes the divide-by-4 specified in the frequency plan.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c6fb069d7260590aebe6dd97f41dc8b01c3c762d
Author: Matt Wagantall <ma...@quicinc.com>
Date: Thu Aug 13 11:36:51 2009 -0700
msm: remote-spinlock: Use SWP-based locking with ARM11 apps cores
Use a remote spinlock mechanism based on the SWP instruction instead of
one based on the local spinlock's LDREX/STREX implementation. This is
required for MSM7XXX SoC's equipped with an ARM11 application processor.
The memory controllers in these SoC's do not support tagging shared
memory addresses for exclusive access and therefore LDREX and STREX cannot
be used for locking between the modem and apps processors.
For SoC's supporting exclusive tagging of shared memory, as well as for
spinlocks local to the application processor, the LDREX/STREX
implementation is preferable and therefore still used.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 64c647ece79a89320511bc947551a65c219eb984
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Tue Sep 1 10:55:31 2009 -0600
msm: Fix bug in ping modem client in callback reply
Ping modem client while replying to register callback
requests, is wrongly setting accept_status to
RPC_ACCEPTSTAT_SYSTEM_ERR instead of RPC_ACCEPTSTAT_SUCCESS
when user callback function returns success.
Fix it to correctly set accpet status to RPC_ACCEPTSTAT_SUCCESS.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7aab75e2683acc6ef4915ec94a971d6cfd5c5770
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Sat Aug 15 22:27:19 2009 -0700
msm: dma: disable dma clk when idle
Add a timer to defer dma clk disable until idle to negate
the immediate disable performance penalty. In addition add
a suspend_late call to cancel the timer when suspending
and execute dma clk disable if all channels are idle.
CRs-fixed: 190215
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ea59bf2f22f3ce92eb42840f36893ae1b225aca2
Author: Stephen Boyd <sb...@quicinc.com>
Date: Fri Jul 24 15:29:04 2009 -0700
msm: io: refactor msm_map_*_io() and compile out unused code
add/remove: 1/6 grow/shrink: 0/1 up/down: 108/-928 (-820)
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4403d8cf1747608e367c336e1006aa0dda117372
Author: Stephen Boyd <sb...@quicinc.com>
Date: Tue Jul 14 13:31:21 2009 -0700
smem_log: add missing newline in printk, use __func__
output from dmesg
<3>[ 0.361046] smem_log_init: no power log or log_idx
allocated, smem_log disabled<6>smd_alloc_channel() 'DATA5_CNTL'
cid=38
While we're here, use __func__ instead of hand writing the
function name and modernize a bit with pr_err().
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 686f1455036edcbbc6eb5b473d0da35f9eb1c5bb
Author: Stephen Boyd <sb...@quicinc.com>
Date: Wed Jul 15 14:25:38 2009 -0700
rpcrouter: cleanup printk messages
Output before change:
rpcrouter:Server create rejected, version = 0program (3000fe00)
Output after change:
rpcrouter: Server create rejected, version = 0, program = 3000fe00
Also modernize a bit with pr_err().
Acked-by: Stephen Boyd <sb...@quicinc.com>
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7c52b25356f48a83e198236ff8e2887d71c98d1e
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Aug 27 16:24:57 2009 -0600
msm: provide interface to set/clear smsm_intr_mask
APPS processor can specifify which smsm interrupts
it wants to receive by setting/clearing smsm_intr_mask
table entry for a particular smsm_entry and APPS host.
Provide interface from smd driver to get and set/clear
these fields.
Provide following interfaces.
int smsm_change_intr_mask(uint32_t smsm_entry,
uint32_t clear_mask, uint32_t set_mask);
where 'set_mask' and 'clear_mask' are set and clear masks and
the status whether set/clear is successful.
int smsm_get_intr_mask(uint32_t smsm_entry, uint32_t *intr_mask);
where required 'smsm_entry' value for APPS host is returned in
'*intr_mask' and status of read is returned.
Also correct initializing the smsm_intr_mask table.
Initialize SMSM_APPS host field for all smsm entries
instead of initializing only apps owned entries.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d282353f865c0bfe0b68cfffe1f73ac01cae42c4
Author: Ai Li <a...@quicinc.com>
Date: Mon Aug 31 10:12:38 2009 -0600
msm: pm: update shared memory bits
Update the shared memory bits to be in sync with newer modem
values.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f61e70dce046200a79eaf58ce66481241917099f
Author: Ai Li <a...@quicinc.com>
Date: Mon Aug 31 10:09:23 2009 -0600
msm: pm: fix residency math to match expiration time unit
Timer expiration is in units of nanoseconds. residency is in units
of microseconds. Fix the math so the units match.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c23755abe6c19a376e9701ed6b643f67c1e27639
Author: Ai Li <a...@quicinc.com>
Date: Mon Aug 31 10:03:27 2009 -0600
msm: pm: correct CLK_SLEEP_EN bit value for 7x27
If bit 4 of CLK_SLEEP_EN is set before entering SWFI, the GP_Timer
occasionally misses expiration and not fires the timer interrupt.
Change code to make sure bit 4 is not set before SWFI.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 22aef584b4c2b8629cacc6f93a989038055cfce0
Author: Zeng, Helen <xiao...@quicinc.com>
Date: Thu Aug 27 13:09:57 2009 -0700
Use msm_rpc_connect_compatible API in pmic function
Replace msm_rpc_connect with msm_rpc_connect_compatible
in pmic_rpc_req_reply function.
Change-Id: I396953d289ba50d5e2eb46c91a7bc7fa0a5e837e
Signed-off-by: Zeng, Helen <xiao...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dbc3978d8e28939c5d96a5017624ee7d79af4fea
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Aug 27 13:59:01 2009 -0600
msm: set low_latency flag for DATA1 smd port tty interface
smd7 which is tty interface for smd DATA1 port can be used
in place of smd0 interface. Increase the throughput effeciency
of smd7 interface by setting low_latency flag similar to
smd0 interface.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d45f0b411b31fb11934af45efa1123da6215913f
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Aug 27 13:50:23 2009 -0600
msm: allow smd_ctl_read to attempt reading more than maximum packet size
remove checking 'count' value in smd_ctl_read.
user trying to read more than maximum packet size is valid.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b9e9c84e6b2241861a2f3cff0573d28e4d486674
Author: Matt Wagantall <ma...@quicinc.com>
Date: Mon Aug 24 16:50:24 2009 -0700
msm: clock-7x30: Handle AXI_LI_ADSP_A clock dependencies.
The AXI_LI_ADSP_A clock feeds the AXI arbiter that arbitrates between
the USB and MDH h/w blocks. When any one of those h/w block are active,
the arbiter clock needs to be turned on to allow the use of the AXI bus.
Also, enable the UMDX_P_CLK clock during initialization.
Signed-off-by: Matt Wagantall <ma...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 61fbde4ad28e04b2aa9a8f7e9dc96301d9916a40
Author: Saravana Kannan <ska...@quicinc.com>
Date: Tue Aug 25 16:18:22 2009 -0700
msm: acpuclock-8x50: Avoid configuring PLL divider mux when not using it.
Configuring the PLL divider mux unnecessarily could result in running
off of wrong PLLs for a very short duration. Doing so will cause a
crash if that PLL is not on.
One example case is when switching from running off of modem PLL to
run off of the AXI clock source.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d5246958fee5d191583afb5e2ea226a4131c525b
Author: Saravana Kannan <ska...@quicinc.com>
Date: Fri Aug 21 19:08:58 2009 -0700
msm: clock: Updated supported clock list for 7x30.
Add MFC, LPA, SPI and GRP 2D clocks.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b6b94f02d10550391e9764dd15340cfac5e0988a
Author: Vladimir Kondratiev <vkon...@qualcomm.com>
Date: Thu Aug 20 15:15:32 2009 +0300
msm: smc91x: off-by-one in memory resource
In memory resource, .end should be last used byte,
i.e. .start + size - 1
Signed-off-by: Vladimir Kondratiev <vkon...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a55a45b80085a6e3ed75f3ed2a3f84dcd44395d8
Author: Michael Bohan <mbo...@quicinc.com>
Date: Mon Aug 17 18:55:51 2009 -0700
msm: socinfo: Add support for msm7x30
Without this change, flash booted msm7x30 SURFs will dereference
bad memory.
ID 59 is for MSM7630.
ID 60 is for MSM7230.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5448931ec894531f57f13d6985ec3abfd0449b27
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed Jul 15 18:05:31 2009 -0700
msm: clock: Implement direct clock control for 7x30.
The other existing targets use proc_comm calls to the modem processor
to control all the clocks. In 7x30 based SoCs, the application
processor has read/write permissions to most clock registers. This
allows direct control of most clocks in 7x30 based SoCs.
Since the shadow registers aren't accessible at this time (requires
modem change), the init code has been commented out to avoid trampling
over registers that are used by the modem.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7c199d19907b51cdf9293e82fc59a1780047e03e
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Fri Jul 31 16:40:35 2009 -0700
msm: acpuclock-8x50: Couple CPU freq and AXI freq.
The memory throughput is directly proportional to the AXI frequency. The
AXI freq is coupled with the CPU freq to prevent the memory from being
the bottleneck when the CPU is running at one of its higher frequencies.
This will cause an increase in power consumption when the CPU is running
at higher frequencies, but will give a better performance/power ratio.
Signed-off-by: Abhijeet Dharmapurikar <adha...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bd0bc7669823487e4c82a27281bb9da793874417
Author: Ruigrok, Richard <rrui...@quicinc.com>
Date: Mon Aug 17 15:20:45 2009 -0600
msm: add proc_comm commands for clkctl_rpc_rail on/off
New proc_comm commands:
PCOM_CLKCTL_RPC_RAIL_ON
PCOM_CLKCTL_RPC_RAIL_OFF
Signed-off-by: Richard Ruigrok <rrui...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6890b91021d2e858fd5b40f6c8ec75021bc4ea78
Author: Kenneth Heitke <khe...@quicinc.com>
Date: Mon Aug 17 12:26:58 2009 -0600
msm: spi: Add SPI clock support for 7x30
Signed-off-by: Kenneth Heitke <khe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 918a5b01652bc80b7bd0479a09c8bb5456df9196
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed Aug 12 23:02:38 2009 -0700
msm: clock: Update supported clocks list for 7x30.
Add USB and display related clocks.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d4e0d88dc7593cb1e6fea7d8eba2259bd60db24d
Author: Michael Bohan <mbo...@quicinc.com>
Date: Fri Aug 7 14:01:56 2009 -0700
msm: gpio: Add support for msm7x30
Note that we are using GPIO shadow region 2 per the modem
configuration.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7594da0939495acd80118162a904f5381b88d739
Author: David Brown <dav...@quicinc.com>
Date: Tue Aug 11 09:36:00 2009 -0700
msm: dma: Use proper EXPORT_SYMBOL.
Use EXPORT_SYMBOL instead of EXPORT_SYMBOL_GPL to match the other
instances in the driver.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b086d300ce2272d61489dc26d1a3ea8d3b30505f
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed Aug 12 22:57:34 2009 -0700
msm: clock: Add new proc-comm clock ids for 7x30.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 87bf33ba7db0260492413eb9edcf4388321e20ee
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Aug 12 18:28:49 2009 -0600
msm: Add new smem items
Add following new smem items.
SMEM_SMD_BRIDGE_ALLOC_TABLE ;to be used in SMD bridge layer for SPS
SMEM_SMDLITE_TABLE ;to be used for SMDLITE implementation
SMEM_SD_IMG_UPGRADE_STATUS ;not used in linux, keeping API in sync
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d9b634f12119922a7cac85e2986e4fbd61bb663e
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Aug 12 18:23:20 2009 -0600
msm: Fix bugs in oem_rapi driver
Size of each data item transmitted over RPC should be in multiple
of 4 bytes. Strings passed are not padded to make their size a
multple of 4. Pad the strings with 0s to make its size a multiple of 4.
As per the marshalling protocol, the size of variable sized array should
only be sent if it is a valid pointer. Correct code to send output_size
only if output_valid is true.
Buffer pointer is not incremented after reading out_len.
Correct buffer increment after reading out_len parameter.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4c87ec5aaedfbfb4b8306021afd8289d8b97729f
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Aug 12 17:39:39 2009 -0600
msm: Add smd tty device interface for SMD port 'DATA21'
Use smd tty driver for SMD port 'DATA21'.
Also remove smd port enumeration, as ports are dynamically allocated.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a5e24b7033ff818df2b4e765b99bb9bc6419efac
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed Aug 12 12:25:53 2009 -0700
msm: clock: Add support for clk_round_rate().
Clock control over proc-comm doesn't support this API, but other
implementations of clock control could support this API.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 61bf8a6bd5823fb347337ebf049f30d51bf1d287
Author: Ai Li <a...@quicinc.com>
Date: Fri Aug 7 15:58:05 2009 -0600
msm: pm: Add config option to choose power management timeout action
The action indicates what will happen when the Application Processor
times out waiting for Modem's handshake. The choices MSM_PM_TIMEOUT_HALT
and MSM_PM_TIMEOUT_RESET_MODEM are useful during development and
debugging. MSM_PM_TIMEOUT_RESET_CHIP can be used on production builds.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 020dc853404353892db1dfff39530371d41e0a20
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Mon Aug 3 16:04:48 2009 -0600
msm: Make logging SMD debug messages a config option
Add config item MSM_SMD_DEBUG, enable by default.
Also allow runtime switch on/off of debug/info messages.
Module parameter msm_smd_debug_mask has following bit definition
bit 0 -> enable SMD DEBUG messages when set; default 0;
bit 1 -> enable SMSM DEBUG messages when set; default 0;
bit 2 -> enable SMD INFO messages when set; default 1;
bit 3 -> enable SMSM INFO messages when set; default 1;
CRs-Fixed: 188680
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ba1e9b22bfb974e6bd5db39834cf47cbb037c41a
Author: Neil Leeder <nle...@quicinc.com>
Date: Thu Aug 6 14:28:37 2009 -0400
msm: correct gpio register address for 8K
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cba283c5329c7daf47aed96bed33c5ee70e3de89
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed Aug 5 16:26:28 2009 -0600
msm: Add support for NULL RPC callback function
Callback Id 0xffffffff (-1) should be used when client specifies a
NULL callback function. This is required for other processor
not to send callback requests.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6fdff7df3cffc8a6d266351be86ee54c7168ac66
Author: Ai Li <a...@quicinc.com>
Date: Wed Jul 29 15:02:34 2009 -0600
msm: pm: reset chip when Modem fails to respond to handshake (n-way)
When Modem fails to respond to the power collapse handshake, Modem and
Apps are out of sync and Modem is in a bad state. There is no way of
recovery other than resetting the chip. This patch also increases the
timeout value dramatically to avoid false detection.
This patch applies to the n-way protocol implementation.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8f5206eeb239bb49887de09ad06e8ce2b3178e4f
Author: Anoop Krishnan Jayasankaran <c_aj...@quicinc.com>
Date: Wed Aug 5 10:50:51 2009 -0700
msm: irq: Fix msm_irq_enter_sleep2 to return correct status.
Fixed msm_irq_enter_sleep2 to return error code when there
are pending interrupts.
CRs-Fixed: 191845
Signed-off-by: Anoop Krishnan Jayasankaran <c_aj...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a45681fb021861899c4c01f6041087625dc3d538
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Tue Aug 4 15:42:33 2009 -0600
msm: Include required header files in oem_rapi_client.h
Include mising header files in oem_rapi_client.h
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d5d2ceb736909db6ae1bc3983017e9eabd41e6e6
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jun 3 10:30:30 2009 -0400
msm: move DRAM physical address for 8K QSD processors
Several address constants are changed to match the new
physical DRAM region which is now based at 0x20000000
CRs-fixed: 189579
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 142146e3ff133aa7dc43d8ccf841cd9ec2e8698a
Author: Willie Ruan <wr...@quicinc.com>
Date: Fri Jul 31 16:07:54 2009 -0700
msm: pmic: add HSED APIs
Add HSED control APIs for 7x30 target using PMIC8058. Add the use of
new RPC version to keep backward compatible.
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 551ef2d8dec13b5f6ee7c3a3ba6e296a9bdf86ba
Author: Iliyan Malchev <mal...@google.com>
Date: Wed Apr 1 17:29:08 2009 -0700
[ARM] msm: AMSS >= 6350 swaps VFE_CLK and VFE_MDC_CLK
Signed-off-by: Iliyan Malchev <mal...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7fed371fd675ac7aaccfbdcce0079d976872149c
Author: Steve Muckle <smu...@quicinc.com>
Date: Tue Jul 28 18:29:05 2009 -0700
add arch_has_speculative_dfetch() to address dma caching issue
The Scorpion CPU speculatively reads data into the cache. This
may occur while a region of memory is being written via DMA, so
that region must be invalidated when it is brought under CPU
control after the DMA transaction finishes, assuming the DMA
was either bidirectional or from the device.
Currently both a clean and invalidate are being done for
DMA_BIDIRECTIONAL in dma_unmap_single. Only an invalidate should be
required here. There are drivers that currently rely on the clean
however so this will be removed when those drivers are updated.
CRs-fixed: 190032,186612
Signed-off-by: Steve Muckle <smu...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 333560e57d1321ad38bdc46f4014b48a6a8c316a
Author: Bryan Huntsman <bry...@quicinc.com>
Date: Mon Jul 27 18:30:44 2009 -0700
arm: msm: fix BSD license statement in headers
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 13eb7111ec208e7a984eca15e5d8eb768b4feb64
Author: Brent DeGraaf <bdeg...@quicinc.com>
Date: Wed Jul 29 20:12:04 2009 -0400
msm: rpc: fix warnings when disabling MSM_ONCRPCROUTER_DEBUG
Signed-off-by: Brent DeGraaf <bdeg...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9f6d6a47770fb45fcd81ec1b00bc5cbba2fcb7c7
Author: Willie Ruan <wr...@quicinc.com>
Date: Fri Jul 24 17:21:21 2009 -0700
[ARM] msm: add new vreg IDs for MSM7X30
Signed-off-by: Willie Ruan <wr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8678f823a9ca86986a47f655804e3f1bfb076bde
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Mon Jun 22 11:23:38 2009 -0600
msm: Add OEM RAPI RPC Client Driver.
Add oem rapi rpc client module. This module implements the
oem rapi client rpc apis and also provides open/close/read/write
test access to the userspace. This module can be used by OEMs to
implement their RPC client drivers in the kernel.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5660eefd7d9293fa505cff30cdbc617a5c3430eb
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Mon Jun 22 11:17:09 2009 -0600
msm: Cleanup Ping MDM client driver
Change name to ping_mdm_rpc_client.c
Cleanup Ping MDM Client driver to a well formatted structure.
Remove ioctl debug interface as read/write is enough to run tests.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a52b0a3bed56be71dab26dd57376f4804a4a7871
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Mon Jun 22 11:07:27 2009 -0600
msm: Add more support in RPC Client Framework
Add support to piggy back payload along with reply
by providing msm_rpc_start_accepted_reply and
msm_rpc_send_accepted_reply apis.
Provide machanism/interface in framework for
'callback Id Vs Callback function' table for a client.
Change msm_rpc_client_req to accept timeout value from user.
Update the ping/handset/fsusb drivers as per RPC Client Framework change.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 06c8dbe1af774614860b64df0d75218cddd83814
Author: Ruigrok, Richard <rrui...@quicinc.com>
Date: Mon Jul 27 14:19:16 2009 -0600
[ARM] msm: add proc_comm commands for clkctl_rpc_reset assert/deassert
New proc_comm commands:
PCOM_CLKCTL_RPC_RESET_ASSERT
PCOM_CLKCTL_RPC_RESET_DEASSERT
Signed-off-by: Richard Ruigrok <rrui...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 64ec389af205189bc56227b4d05ab5a9d802df61
Author: Ai Li <a...@quicinc.com>
Date: Wed Jul 22 18:36:38 2009 -0600
[ARM] msm: idle: add memory barriers, cache flush, TLB invalidation
The 1:1 MMU mapping code is missing cache flushes, TLB invalidations,
and branch predictor invalidations, which are required when changing
MMU mappings. The missing TLB invalidation, in particular, causes
the Android framework to crash randomly. The patch also adds data,
instruction memory barriers as needed, and moves SWFI to a separate
function so that the error return path of power collapse correctly
restores registers.
CRs-Fixed: 188703
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b8bc1af0c3804efd855cdb0b4ba4f1eba02cbadd
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Jul 16 12:48:11 2009 -0700
[ARM] msm: clock: Add new list of supported clocks for 7x30.
The list of clocks is just a preliminary list and will change soon.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d207afa89a2dc0a8d11d0aee57d801700ee0711d
Author: Saravana Kannan <ska...@quicinc.com>
Date: Tue Jul 14 19:57:50 2009 -0700
[ARM] msm: clock: Refactor code to allow SoC specific implementation.
Each clock now has a list of function pointers that are used to enact
the standard clock APIs on that clock. So, clocks that need SoC
specific code will have their "clock_ops" function pointers
initialized differently.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7ee9cfd77d7f2ad72d92870d1c74c7f733579114
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed Jul 22 12:57:51 2009 -0700
[ARM] msm: pm: Fix up existing workaround to enable code refactoring.
An already existing temporary workaround includes and uses restricted
APIs to go around the clock driver. That prevents refactoring of the
clock code. So, change the workaround to not use the #defines that
will be refactored.
The workaround is expected to be removed soon.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3eaa9c6f5999df653bf221d3c43ce5c2e3fa0702
Author: Dave Estes <ces...@quicinc.com>
Date: Wed Jul 22 12:26:30 2009 -0400
[ARM] msm: Update residencies and latencies for the sleep modes on 8x50.
Update residencies and latencies for the sleep modes based on tests
performed on 8k hardware. This improves power and performance.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5d71b3fa17e984b65d930b4f6d967a15bf700007
Author: Abhijeet Dharmapurikar <adha...@quicinc.com>
Date: Tue Jul 21 18:15:34 2009 -0700
[ARM] msm: moved cpufreq to late_initcall
The cpufreq driver calls pmic before it is initialized. This change moved
the cpufreq to late_initcall (level 7) so that it will call pmic after it
is initialised.
CRs-fixed: 184897
Signed-off-by: Bryan Huntsman <bry...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a455a682f73f3c57c231702f5d81e427c326e75f
Author: Ai Li <a...@quicinc.com>
Date: Thu Jul 16 18:28:58 2009 -0600
[ARM] pm: reset chip when Modem fails to respond to handshake
When Modem fails to respond to the power collapse handshake, Modem and
Apps are out of sync and Modem is in a bad state. There is no way of
recovery other than resetting the chip. This patch also increases the
timeout value dramatically to avoid false detection. The wait for Modem
setting its RUN bit while Apps exiting power collapse is removed. The
wait is not necessary and keeping it could lead to a Apps-Modem deadlock.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d12d74945a0b68daf47bac364509f25370ce05c0
Author: Vladimir Kondratiev <vkon...@qualcomm.com>
Date: Thu Jul 16 15:52:51 2009 +0300
[ARM] msm: GPIO: group operations
Provide facility to operate with groups of GPIO's.
All GPIO information written in single structure, containing all
bits as required for TLMM in GPIO_CFG() style.
GPIO pin, as required by other functions, extracted from
GPIO_CFG() style - this removes data duplication.
Group functions provide accurate error checking and
rollback in case of error
Signed-off-by: Vladimir Kondratiev <vkon...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6c7a0f4c11d7d8484255b56922bfa363cd4646c8
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jul 1 12:06:23 2009 -0400
[ARM] msm: allow external LCDC display for 8K surf
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eb7ff72791e7260626e384c73a6f01011fee70f1
Author: Dudani, Ajay <adu...@quicinc.com>
Date: Mon Jun 29 12:07:05 2009 -0700
[ARM] msm: move DRAM physical address for qsd8k
Update physical addresses used for kernel decompression such that it
does not get overwritten by other values. Since all ARCH_MSM_SCORPION
targets now use a base of 0x24000000, the CONFIG_QSD_BASE_24000000 can
be removed
Signed-off-by: Ajay Dudani <adu...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c685ac573022b4febd29520583d1315be1d7afe7
Author: Dave Estes <ces...@quicinc.com>
Date: Mon Jun 29 15:26:30 2009 -0400
[ARM] msm: Add jtag debug register save & restore
Add ARM specified JTAG debug register save & restore code.
This allows JTAG debug sessions to work through power collapse.
Breakpoints etc. are preserved. Adds new jtag-v7.S file.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8ae9df07be81349d376a9e4cc2bd940a07f1d752
Author: Wilson, Matt <mtwi...@quicinc.com>
Date: Mon Jun 29 14:13:04 2009 -0500
[ARM] msm: add vreg reference count
Support independent enable and disable by clients for common
vreg. First enable switches on and last disable switches off.
This change has no check for voltage level so clients
must agree on level for common vreg.
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 187d24885c402f649589bb6ad9376137c3476e4a
Author: Dave Estes <ces...@quicinc.com>
Date: Thu Jun 25 18:09:24 2009 -0400
[ARM] msm: idle: Use vfp subsystem to preserve vfp context
Use vfp subsystem interface to preserve vfp context through power
collapse. Remove old assembly vfp preserve code in msm idle-v7.S
implementation.
Signed-off-by: Dave Estes <ces...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d86b04b5bc7d538146230eb2b4f6b0af10a54d2a
Author: Praveen Chidambaram <pchi...@quicinc.com>
Date: Fri Jun 26 09:16:30 2009 -0600
[ARM] msm: clock: Use system bus QoS to vote for min AXI freq.
The application processor usually sets the EBI1 clock frequency to
match the processor frequency. In some cases, a lower CPU frequency
but a higher AXI frequency may be required.
The drivers and user space applications can vote for a min EBI1 clock
frequency using the system bus QoS parameter. The QoS parameter
is to be provided in KHz. Note, that this does not guarantee a bus
bandwith. The resultant clock frequency will be atleast greater
than or equal to the QoS value.
CRs-fixed: 185976
Signed-off-by: Praveen Chidambaram <pchi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b312cdc8703551bdb8448ac1ace841f0bcfd55a8
Author: Saravana Kannan <ska...@quicinc.com>
Date: Mon Jun 22 19:25:26 2009 -0700
[ARM] msm: clock: Add support for voting on EBI1 min rate.
There are two ways of voting on EBI1 min rate. One is the PM QoS API
(which might sleep) and the other is the ebi1_clk_set_min_rate() API
(which won't sleep). The 2nd API is needed to aggregate the acpuclock
vote and the final PM QoS vote since the acpuclock code can't afford
to sleep in the context of the suspend code (one of the callers).
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cb8e34f586aece14dbb16fdefb284753da1d2a88
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Thu Jun 25 14:52:18 2009 -0700
[ARM] msm: pm: pm2: Enable msm_pm_restart restart handling.
This function's reset hook or PCOM_RESET_CHIP modem command
was ifdef'd out for bringup. This patch removes the
ifdef, allowing this function to execute the reset hook or
modem commanded restart.
This function is executed when the user runs /sbin/shutdown
or kill 1 (kill init).
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit f399a0c58b8587c2fcaeaa429c4db8be0424e8c1
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Tue Jun 23 23:07:47 2009 -0600
[ARM] msm: acpuclock: Disable PLLs during 7x30 power collapse.
Unlike the older 7k series of SoCs, the application processor
in the 7x27 SoC directly controls the PLLs. So, when the
application processor goes into power collapse, it has to take
care of turning off the unused PLLs and can't depend on the
modem processor to do it.
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c2401b66747ac596528ea3fc6499f744a1fd88b2
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Thu Jun 25 10:43:19 2009 -0600
[ARM] msm: Remove loopback driver
Remove loopback driver as tty driver is being used
to drive loopback port.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0854128d3cc767e1e60fbc802f78ada1ac740761
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Mon Jun 22 13:36:19 2009 -0700
[ARM] msm: pmic: fix endless read operation during debugfs read
return 0 to indicate an end-of-read condition to debugfs during
read operation
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 43be2fa9a46853d59c78f1db61da9b56fc9819c4
Author: Michael Bohan <mbo...@quicinc.com>
Date: Mon Jun 22 12:15:22 2009 -0700
[ARM] msm: proc_comm: Add support for msm7x30
msm7x30 uses a bit offset into a single register to interrupt the
modem processor.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit dae3feaae536ab513b7e1040c5a88d18bba670c0
Author: Michael Bohan <mbo...@quicinc.com>
Date: Wed Jun 17 16:39:16 2009 -0700
[ARM] msm: smd: Add error checking to smem_find calls
It's better to get a warning message than a NULL pointer
dereference.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cfce782bdf5b876eb2233457edfe165d9da765cf
Author: Michael Bohan <mbo...@quicinc.com>
Date: Wed Jun 17 16:20:36 2009 -0700
[ARM] msm: smd: Add support for msm7x30
msm7x30 uses a bit offset into a single register to interrupt
the modem processor.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 13309de36e1c49052333e37800a17595ac99f4d7
Author: Zach Pfeffer <zpfe...@quicinc.com>
Date: Tue Jun 16 22:37:17 2009 -0600
[ARM] msm: smd_ctl2: Use per-inst buf and inc tx/rx buf size from 1k to 2k
The unused rx_buf associated with each smd_ctl2 instance is now used
instead of a stack allocated buffer in the read.
In addition to using a per-instance buffer each smd_ctl2 port's transmit
and a receive buffer's size has been increased. It is an error to read or
write more than this limit.
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bb6cdb2ad2c8b9a1b56b1e2da0bb44954e89f14f
Author: Praveen Chidambaram <pchi...@quicinc.com>
Date: Tue Jun 9 12:24:42 2009 -0600
[ARM] msm: svs: Make PMIC_DEFAULT_DCDC1 configurable
Signed-off-by: Praveen Chidambaram <pchi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 19cf972de576216754bfaabe06802fd389c0da29
Author: Michael Bohan <mbo...@quicinc.com>
Date: Tue Jun 16 16:35:57 2009 -0700
[ARM] msm: irq: Fix incorrect bounds checking
The current code is clobbering several hundred bytes of memory.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9c86225c9d7b6407337aa957d9dd9023baa1b94d
Author: Michael Bohan <mbo...@quicinc.com>
Date: Sun Jun 14 18:43:40 2009 -0700
[ARM] msm: Fix GPIO2 base address for ARCH_QSD8X50
The respective MSM_GPIO2_PHYS values for QSD8X50 and MSM7X30 are confused
with each other.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2039ac454489bbe01177c46cd563a3ea343f6ba4
Author: Michael Bohan <mbo...@quicinc.com>
Date: Thu May 28 14:33:11 2009 -0700
[ARM]: msm: Add base board support for msm7x30
The MSM7X30 SoC includes a Scorpions apps processor with 512 KB of L2
cache.
This patch includes support the MSM7X30 SURF board. FFA board
support is not included in this patch.
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c004bf2d3409a453aee8497cf7e0bcb829ec6cf5
Author: Michael Bohan <mbo...@quicinc.com>
Date: Thu May 28 10:10:06 2009 -0700
[ARM]: msm: Apply new MSM naming scheme
The current naming scheme consisting of ARCH_MSM and ARCH_QSD no
longer properly describes our targets. We have new targets that
are functionally hybrids between 7k and 8k, and as such we need a
finer level of granularity to express such nuances.
The new scheme consists of four different levels. At the top is
ARCH_MSM, which now encapsulates all other levels.
The next three levels are user-defined. The first is the
"CPU Subsystem" (ARCH_MSM_*), which specifies which peripheral block
(clock control, timers, interrupt controller, etc.) one requires.
There may only be one SS configured at a time. It is selected
implicitly when selecting the SoC.
The SoC represents a CPU core. The SoC may have subtle differences
between other SoCs that use the same SS, and can be expressed
with preprocessor conditions. There may only be one SoC
configured at a time. Despite the rather specific SoC names (eg.
ARCH_MSM7X01A), it is possible for a SoC to describe multiple
targets (eg. MSM7201A, MSM7X25).
The last level is the "Machine" specification. They depend on the
SoC that has been chosen. Any number of machines can be selected
at once, assuming they correspond to the same SoC.
example:
ARCH_MSM -> ARCH_MSM_SCORPION -> ARCH_MSM7X30 -> MACH_MSM7X30_SURF
Signed-off-by: Michael Bohan <mbo...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c2852a43c189918d85d6d59738b903abbcaf146e
Author: Neil Leeder <nle...@quicinc.com>
Date: Wed Jun 10 14:08:26 2009 -0400
[ARM] msm: move DRAM physical address for 8K QSD
Several address constants are changed to match the new 8K QSD
physical DRAM region which is now based at 0x24000000, if
new config option QSD_BASE_24000000 is set.
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2f73a344b455e4b35392a70c351b6cad3bba8410
Author: Saravana Kannan <ska...@quicinc.com>
Date: Mon Jun 8 20:02:49 2009 -0700
[ARM] msm: acpuclock-8x50: Fix efuse magic value bit length.
Turns out the magic value is only 4 bits long. So don't look at the
other 4 bits anymore.
CRs-fixed: 183437
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2872fff0640a1105e8e05216492913353b8506cb
Author: Saravana Kannan <ska...@quicinc.com>
Date: Fri Jun 5 13:52:52 2009 -0700
[ARM] msm: acpuclock-8x50: Fix SCPLL init and freq switching code.
Use HOP switch method instead of SHOT switch method when changing
SCPLL frequency (except for the very first time) to work around a
hardware bug that manifests when using SHOT switch at temperatures
around -20 C. Add a few delays between SCPLL state changes.
CRs-fixed: 180376
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4f4c5d5824c0bec8b2c94a73ac7a1ca7b9562f62
Author: Li, Ai <a...@quicinc.com>
Date: Thu Jun 4 21:52:30 2009 -0400
[ARM] msm: suspend: fix suspend restore code
Added code to preserve critical CP15 registers. Update MMU restore
function to properly follow ARMv7 TEX remapping procedures. Clean
up code.
Signed-off-by: Li, Ai <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5a595d18f7ddb9050d5a0708e4900ddf33137dae
Author: Li, Ai <a...@quicinc.com>
Date: Tue May 5 16:04:44 2009 -0600
[ARM] msm: pm: set CLK_SLEEP_EN register before SWFI
Setting the register before SWFI allows several clocks to
be turned off during SWFI. This could yield additional
power savings.
Signed-off-by: Li, Ai <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6f8ae4af277b1f599b4919e0a7bda17b314db235
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Fri Jun 5 09:58:29 2009 -0600
[ARM] msm: Save test results across reads and writes
This module contains many tests that are launched by writing the test
name into the device node. Test results are retrieved by reading the
device node.
The test return value was being allocated in the open and freed in the
close. This meant that the typical open, read, close and open, write,
close of echo and cat deallocated the return value. This fix statically
allocates the return value so that its value is saved across echo and cat.
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ec8a757e21bcb2ae9787bb6be85293237c1859fc
Author: Trilok Soni <ts...@qualcomm.com>
Date: Fri Jun 5 15:30:29 2009 +0530
[ARM] msm: handset: Add handset callback api support
Add support for receiving handset events from modem through
RPC callback api mechanism.
Enabled only for 7x27 SURF and FFAs.
Signed-off-by: Trilok Soni <ts...@qualcomm.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5657c79cc9f8890c0bec97d3fadf099d92781af9
Author: Neil Leeder <nle...@quicinc.com>
Date: Fri May 29 13:30:55 2009 -0400
[ARM] msm: keep vmalloc space location with 2G/2G memory split
The 2G/2G memory split option moves the kernel starting address
PAGE_OFFSET down 1GB from its default location. This keeps
vmalloc space at the same location, ending at 0xE0000000.
Signed-off-by: Neil Leeder <nle...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5b437e4a8431aecd84239bfc3a3b4bbb31af13e0
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Fri May 22 13:30:33 2009 -0600
[ARM] msm: Add write and read to execute tests and return results
Reading the device node returns the return code of the last test run.
If no tests have been run the node will return 0. The return code is
returned as an ASCII coded signed integer.
Tests and Test Modification Commands
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
null_test
reg_test
data_reg_test
data_cb_reg_test
use_multi_clients
use_default_client
Execute Test Example
~~~~~~~~~~~~~~~~~~~~
echo null_test > /dev/ping_test
Read Last Test Example
~~~~~~~~~~~~~~~~~~~~~~
cat /dev/ping_test
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit bfff4654aaf87d16dd4eef79497c72c350485f56
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu May 21 12:02:00 2009 -0700
[ARM] msm: clock: Add new clocks for 7x27 and 8x50.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c4463f5d639a8d18f9602e100b7160f70fd1ffdd
Author: Kiran Kandi <kka...@quicinc.com>
Date: Thu Apr 2 23:51:59 2009 -0700
msm: export msm_rpc_create_server
Signed-off-by: Kiran Kandi <kka...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 107a548e2c3f6f8b9e94d0c9d48477df289fc83e
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Wed May 20 15:11:40 2009 -0600
[ARM] msm: Add smd rpc ping module
Add smd rpc ping module. Currently, the ping module implements
the ping modem client rpc apis and also provides open/close/ioctl
test access to the userspace. This is useful as a test rpc client
in the kernel.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ccc61a88010ffef6b372739995b9ae25498b5b39
Author: Wilson, Matt <mtwi...@quicinc.com>
Date: Thu May 14 08:08:02 2009 -0700
[ARM] msm: bluetooth: power switch hardware support
Add power switch support on newer 8K FFA keypads with BT and WLAN.
No change is needed for any 7k FFA or SURF.
CRs-Fixed: 174505, 177193
Signed-off-by: Matthew Wilson <mtwi...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 0014108c4f73566be99bdc8b90cb460ff4b50561
Author: Ai Li <a...@quicinc.com>
Date: Mon May 4 16:51:50 2009 -0600
[ARM] msm: timer: reduce rounding error in time sync calculation
Because the frequency of the debug timer is not a multiple of
that of the slow clock, the order of the calculation leads to
a large rounding error. It, in turn, causes the system time
to jump ahead intermittenly. The order of calculation is
changed to minimizes the rounding error.
Signed-off-by: Ai Li <a...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1d661b63aaa8324d1e680ab031838173c7cc453e
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Fri May 15 15:12:52 2009 -0700
[ARM] msm: pmic: add new modem control APIs
Implement pmic modem control APIs to provide modem control
path to others
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 75f5eb0bae7413ce18f42d7bc05d66c093f42058
Author: Saravana Kannan <ska...@quicinc.com>
Date: Wed May 13 16:38:16 2009 -0700
[ARM] msm: acpuclock-8x50: Fix typo in efuse magic values.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cf59a702efb914b7d5cf0bc66a4fcc7cf3e68bc7
Author: Kuogee Hsieh <c_kh...@quicinc.com>
Date: Wed May 13 08:24:01 2009 -0700
[ARM] msm: pmic: prefix modem control APIs with pmic_
prefixed pmic_ to all modem control APIs to work with new modem control
APIs prvoided by pmic
Signed-off-by: Kuogee Hsieh <c_kh...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 13f6638073202013d6aa56384fcd40fc116da37c
Author: Vishwanathapura, Niranjana <nvis...@quicinc.com>
Date: Mon May 11 16:04:24 2009 -0600
[ARM] msm: Add rpc client operation framework
Add rpc client operation framework to smd_rpcrouter driver
to abstract kernel smd rpc clients opertions into a common
framework.
Signed-off-by: Niranjana Vishwanathapura <nvis...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a67e650f324a1cbdb9e403d5dff5fefbaf8c0509
Author: Pfeffer, Zach <zpfe...@quicinc.com>
Date: Tue May 12 12:20:01 2009 -0600
[ARM] msm: smem_log: Add power log debugfs node
Add a debugfs node that dumps the separate logging area used on
some modem builds to log power events. Output nothing if the power
event logging area is not present. Two nodes are present a raw node
and a symbol resolved node.
This change necessitated a minor refactoring to allow an arbitrary
number of logging areas to be added with ease.
Signed-off-by: Zach Pfeffer <zpfe...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 5ae29cab088e2032e7573722f9fc84c10224bbc7
Author: Saravana Kannan <ska...@quicinc.com>
Date: Fri May 8 20:18:56 2009 -0700
[ARM] msm: acpuclock: Rewrite runtime clock plan selection.
The run time fix up of a single ACPU clock plan table is difficult to
understand for many people and also couldn't handle certain clock
plans elegantly.
Make clock plan selection code more maintainable by using a separate,
mostly-static table for each supported clock plan. Minor fixups are
still used if the fix ups are simple and avoid adding several tables.
Signed-off-by: Saravana Kannan <ska...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3cee2f0b28165e72980788ea5c2fb604b83c6a76
Author: Rhee, C. Joon <jr...@quicinc.com>
Date: Wed May 6 17:36:54 2009 -0700
[ARM] msm: remove msme2 vreg control
msme2 needs to be controlled through pmic instead of vreg api.
pmic_lp_mode_control() api is needed for msme2 power line control.
Signed-off-by: Rhee, C. Joon <jr...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b4f056cb82ddb6b4bb1d37cf3a4285c0e3e85c08
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Dec 17 11:30:34 2009 -0800
Large generic arm update.
Signed-off-by: David Brown <dav...@quicinc.com>
commit b228e85062ae05ffa074ed1aaa80fb7f7d60dab8
Author: Daniel Walker <c_dw...@quicinc.com>
Date: Thu Dec 17 11:28:23 2009 -0800
Large update patch for arch/arm/mach-msm
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ee638ca9318c7f3c71e9f3ad13b0f1924ce9a226
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Mar 23 19:27:45 2009 -0700
[ARM] msm: smd_rpcrouter: Hold a wakelock while while reading a packet, and while packets are queued.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eed8697874a3c99a02e36e92f5f907715c13b3ed
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Mar 23 19:24:56 2009 -0700
[ARM] msm: smd_rpcrouter_servers: Hold a wakelock while calling rpc_call.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d6babc9b8620a7c2beb86f2ad42f1ca6835c385c
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Mar 23 19:17:14 2009 -0700
[ARM] msm: smd_tty: Lock a wakelock for half a second when receiving data.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 363cd47b99b59f12333661d822bf454f27b4ad18
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Mar 23 19:12:36 2009 -0700
[ARM] msm: smd_qmi: Add a wakelock with a timeout to prevent sleep after receiving a qmi notification.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit db7aa4ff30e41fffa145932acc4f08ef030807f1
Author: Mike Chan <mike...@android.com>
Date: Mon Jan 5 23:00:02 2009 -0800
[ARM] msm: acpuclock: Enable PLLs of stepping freqs when necessary.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 800e41f6bbe8cd14ac757bb316fca8d0ae71baa7
Author: Mike Chan <mike...@android.com>
Date: Tue Jan 6 17:09:30 2009 -0700
[ARM] msm: Move cpufreq code from acpuclk.c to cpufreq.c
The acpuclk.c has function calls used to notify the cpufreq framework
when the acpuclk frequency is changed. These notifications are cpufreq
specific and independent of acpu clk.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9cafa1183358e70476b62c2305902176e29a74b1
Author: Saravana Kannan <ska...@quicinc.com>
Date: Mon Dec 22 20:24:17 2008 -0800
[ARM] msm: acpuclock: Couple minimum AXI freq to the current CPU freq.
This patch comes from qualcomm with some minor changes by mi...@android.com.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b15e074f2bc1c40cfc972392f89c0719c8fcaba7
Author: Steve Muckle <smu...@quicinc.com>
Date: Fri Dec 5 17:03:15 2008 -0500
[ARM] msm: make clkctl speed struct internal
The clkctl_acpu_speed structure is not used outside the acpuclock
implementation.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 363b96306d64a0530effbf5aec9cd05a3e00bc76
Author: Mike Lockwood <lock...@android.com>
Date: Fri Feb 20 15:56:06 2009 -0500
[ARM] msm: htc: Clean up USB support in trout and sapphire board files
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4ac77d985be68f0164145aa1e6cf8e3a292ccdbd
Author: San Mehat <s...@android.com>
Date: Wed Feb 18 09:16:50 2009 -0800
[ARM] sapphire: Initial import of Sapphire board support
Signed-off-by: Thomas Tsai <thoma...@htc.com>
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 70e320974b0dd5bf574d7056f4775f16b8997184
Author: Thomas Tsai <thoma...@htc.com>
Date: Wed Feb 18 09:03:23 2009 -0800
[ARM] msm: iomap: Add entries for HSUSB_BASE, SSBI, and TSSC
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 790e50ae2473773791d4cf61014e4cbcd4eed1b0
Author: Thomas Tsai <thoma...@htc.com>
Date: Wed Feb 18 08:34:13 2009 -0800
[ARM] msm: FIQ: Add ability to unselect an FIQ IRQ
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 243c51875fb0381f06c1895d6643110fc30405b7
Author: Thomas Tsai <thoma...@htc.com>
Date: Tue Feb 17 15:16:30 2009 -0800
[ARM] msm: vibrator: Initial support for MSM PMIC vibrator
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c32ffe026cf39be253ca422293bab5550d94d3d0
Author: Dmitry Shmidt <dimi...@google.com>
Date: Tue Feb 17 14:35:27 2009 -0800
[ARM] msm: htc_pwrsink: Generify driver
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4bc8d778c0ae3707d180f287e556d6f303432d3b
Author: Saravana Kannan <ska...@quicinc.com>
Date: Fri Nov 14 19:33:36 2008 -0800
[ARM] msm: acpuclock: Remove VDD check during init.
The VDD register reflects the status of the PMIC only after the
application processor has written to it. This is not the case after a
reset or power collapse since it is the modem processor that sets up
the PMIC. Hence, the VDD register is no longer used to verify the
current VDD level during init (boot up).
Acked-by: Steve Muckle <smu...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 791b93f458cc8b788b8971025b5c09aa9568393d
Author: Iliyan Malchev <mal...@google.com>
Date: Sat Feb 14 13:33:24 2009 -0800
[ARM] msm: htc_acoustic: use smem and rpc to write acoustic table
Signed-off-by: Iliyan Malchev <mal...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 22ce50df3911e4825c192026a96281214a5ff57c
Author: Dima Zavin <di...@android.com>
Date: Fri Feb 6 16:43:24 2009 -0800
[ARM] msm: move clock definitions to devices.c
Also, add the begininngs for having a master list of clocks, and a
bitmask for which sub-arch's they are used by.
Signed-off-by: Dima Zavin <di...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1b4398eb7a95a2cced993b868540736b00b0e908
Author: San Mehat <s...@android.com>
Date: Tue Feb 17 14:03:22 2009 -0800
[ARM] msm: trout: Switch to new headset driver in trout
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4be3b3514e153dd7432fec7dd97c483ede0922aa
Author: Thomas Tsai <thoma...@htc.com>
Date: Tue Feb 17 14:00:09 2009 -0800
[ARM] msm: htc_headset: Import new headset driver from HTC
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit d8931f040eedc4bf7f541e34b58742253fe137b6
Author: Ben Cheng <bcc...@google.com>
Date: Fri Feb 6 16:10:34 2009 -0800
[ARM] msm: oprofile: Enable OProfile and preserve performance counters.
Signed-off-by: Ben Cheng <bcc...@gmail.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 00c4a5e2a32fbc43a8bc334d126a31e5f14b31b3
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Dec 10 16:01:22 2008 -0800
[ARM] msm: Don't call ktime_get from sched_clock
With some kernel debug options sched_clock gets called with xtime_lock
locked for write resulting in a deadlock.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1af51675c4e7a24cd410af67c6151fc46fcf85ca
Author: Arve Hjønnevåg <ar...@android.com>
Date: Thu Dec 11 19:34:10 2008 -0800
[ARM] msm: Prevent clocksource read from jumping back after power collapse.
If the clock events device is shut down or we are idle, don't read from
the hardware, but return a cached value. It would be better to trigger
this from the clocksource api, but there is no hook to enable or disable
a clocksource.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 50159d03ea28fe395738500a4beba5c635d2a94f
Author: Arve Hjønnevåg <ar...@android.com>
Date: Thu Dec 11 20:10:39 2008 -0800
[ARM] msm: timer: Don't call ktime_get from msm_timer_sync_smem_clock
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit deca3b3e46fe42beb494484faa114685aab91539
Author: Saravana Kannan <ska...@quicinc.com>
Date: Thu Jan 29 19:22:50 2009 -0800
[ARM] msm: acpuclock: Fix conditional compile symbol in acpuclock.h.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b6665a9715298cf5bcc3bf72414d244c3874e1f8
Author: Iliyan Malchev <mal...@google.com>
Date: Fri Dec 19 13:30:47 2008 -0800
[ARM] msm: halibut: add list of SND endpoints
Signed-off-by: Iliyan Malchev <mal...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eeea7c92dfba923a075062c68f2bbbb7fa2c303c
Author: Dmitry Shmidt <dimi...@google.com>
Date: Thu Dec 18 12:40:04 2008 -0800
[ARM] trout: Add abstract wifi control functions support
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8842151dfec514dfacc9a1b7a4f247ae7df6ffd7
Author: Mike Chan <mi...@android.com>
Date: Wed Dec 10 13:22:14 2008 -0800
[ARM]: msm: clock: Adding PLL2 clocks for 528 power collapse stepping.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 255e77c07ac2428af1a34e52c8ed0b6b150dfe32
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Dec 10 19:51:50 2008 -0800
[ARM] msm: gpio: Don't call generic_handle_irq with interrupts enabled
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit c29fa878987f4e26e8e37b6e93ef91ee6e2307d5
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Dec 10 19:48:05 2008 -0800
[ARM] msm: generic_gpio: Don't pass GFP_KERNEL to kmalloc when holding a spinlock
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a0af2459b171258ca519db1252f3e539c6a7af1c
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Dec 10 19:45:39 2008 -0800
[ARM] msm: trout-h2w: Don't call request_irq with interrupts disabled.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8ce020dac542241558dfda38857a63113afa0b1d
Author: Mike Chan <mi...@android.com>
Date: Fri Dec 5 15:56:40 2008 -0800
[ARM] msm: clock: Fix cpu speed ramping down from 528mhz.
Should not jump from 528mhz -> 245mhz this is greater than 256mhz spec.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit eed13125cd5e2b76b162c016e17c52aed4893573
Author: Mike Chan <mi...@android.com>
Date: Wed Dec 3 11:06:41 2008 -0800
[ARM] msm: clock: Support for freq table in speed selection.
Change/disable full range of frequencies from /sys in userspace.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit aabfb6415a855d5f8a638f21c57511de3666c78c
Author: Dmitry Shmidt <dimi...@google.com>
Date: Thu Dec 4 10:17:50 2008 -0800
[ARM] msm: Turn ON DMA support for SDC1
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cc5514b9f4ff793f51b3d879e2a8c0b036601f54
Author: Nick Pelly <npe...@google.com>
Date: Mon Dec 1 22:22:13 2008 -0800
[ARM] msm: Fix bugs preventing modular build
Signed-off-by: Nick Pelly <npe...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e72c97d8d05b9c89748aed1a9e45e5fc11faf17e
Author: Arve Hjønnevåg <ar...@android.com>
Date: Tue Dec 2 19:25:19 2008 -0800
[ARM] msm: lldebug: Fix waituart
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 8b78a82cf5375c2b6cda00d5907fda7e1a7ccb13
Author: Arve Hjønnevåg <ar...@android.com>
Date: Tue Nov 18 14:58:08 2008 -0800
[ARM] msm: Add MSM_DEBUG_UART option.
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit fd6cea85103123faead1f0864cfb5b71fca6ae01
Author: Dmitry Shmidt <dimi...@google.com>
Date: Mon Nov 17 20:08:23 2008 -0800
[ARM] msm: trout: Temporarily Disable DMA for SDC1
Signed-off-by: David Brown <dav...@quicinc.com>
commit ab85e5573ffa1eb47a2300d2141b1adc1c40ec1d
Author: Nick Pelly <npe...@google.com>
Date: Mon Nov 17 19:16:13 2008 -0800
[ARM] msm: trout: bluetooth: Use rfkill API to turn bluetooth chipset on/off.
Signed-off-by: David Brown <dav...@quicinc.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b4f2a059ea64266613014c48268df1ba51cb631a
Author: Rebecca Schultz <rsch...@google.com>
Date: Wed Oct 1 10:43:10 2008 -0700
[ARM] msm: trout: Adds trout Power Sink Driver
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 64410f388e41e880f10e280e11f938e994097cc6
Author: San Mehat <s...@google.com>
Date: Mon Nov 10 17:57:27 2008 -0800
[ARM] trout_mmc: Initial MMC support for Trout on 2.6.27
Signed-off-by: San Mehat <s...@google.com>
trout: mmc: Re-enable WIFI SDIO code
Signed-off-by: San Mehat <s...@google.com>
trout: mmc: Re-enable wifi debugfs code
Signed-off-by: San Mehat <s...@google.com>
trout: mmc: Add maxblksize specifier to embedded data for wifi
Signed-off-by: San Mehat <s...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4da1530e63c2e5446e235d0ac920fddb572f7818
Author: San Mehat <s...@google.com>
Date: Mon Nov 10 16:30:27 2008 -0800
[ARM] msm: common: Add SDC device runtime registration
Signed-off-by: San Mehat <s...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 133cac52de0ecc3fc02e553ed9cee51173b35e34
Author: Mike A. Chan <mi...@android.com>
Date: Fri Nov 7 12:39:29 2008 -0800
[ARM] msm: clock: Adding ondemand cpufreq scaling.
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 117b1744f47ebfbe1185aeb39f667787a42f74b9
Author: Rebecca Schultz <rsch...@google.com>
Date: Wed Oct 29 16:57:59 2008 -0700
[ARM] msm: trout: Add msm_fb support to trout architecture
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit cec49691a39dc0b140bccb9ffe4ec04b81056256
Author: Rebecca Schultz <rsch...@google.com>
Date: Mon Sep 29 18:24:57 2008 -0700
[ARM] trout: htc platform battery driver
Uses rpc to communicate with battery state/charging service on the
baseband processor.
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a9bdc7cd3da80e5c891145a2e0eb8e03944abda0
Author: Brian Swetland <swet...@google.com>
Date: Thu Mar 27 18:18:01 2008 -0700
[ARM] msm: add hook for vbus state notification
The msm_hsusb_set_vbus_state() function allows the board file,
battery/charger/power code, etc to inform the USB driver about
the presence of VBUS. The driver now puts the USB client
controller online when VBUS is present and suspends the
PHY when it goes away.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit ff3bddf747040b6fbeb164e812e684d112da28f1
Author: Nick Pelly <npe...@google.com>
Date: Mon Sep 29 18:19:37 2008 -0700
[ARM] msm: trout: htc2wire driver to support htc headsets
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 157b4e0b0c22700e41240e9664b755e5612dbf6b
Author: Rebecca Schultz <rsch...@google.com>
Date: Mon Sep 29 17:58:03 2008 -0700
[ARM] msm: trout: htc acoustic driver
Allows userspace to set acoustic settings via memory mapped region
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 300c0779e118767c88103df575973592c2055fcc
Author: Dmitry Shmidt <dimi...@google.com>
Date: Mon Sep 29 17:55:10 2008 -0700
[ARM] msm: trout: Code to extract wifi calibration data from ATAGs
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4571e0f34715a1d284c3e992478f033db7eee7a5
Author: Rebecca Schultz <rsch...@google.com>
Date: Mon Sep 29 17:51:38 2008 -0700
[ARM] msm: trout: obtain calibration data for the compass sensor
Extracts compass calibraton from the ATAGs set up by the bootloader
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 729d49aa928587d8824150f425792b175da0450e
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Sep 29 17:44:12 2008 -0700
[ARM] msm: trout: platform specific keypad driver
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 9bc05097b9ab05cb8f88ec46d4b69a0c00ac346f
Author: Arve Hjønnevåg <ar...@android.com>
Date: Mon Sep 29 11:16:12 2008 -0700
[ARM] msm: trout: CPLD GPIO driver
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 3a20066078b4b49cd9f5e61db7d9935a46e05b5a
Author: Rebecca Schultz <rsch...@google.com>
Date: Mon Sep 29 17:32:13 2008 -0700
[ARM] msm: trout: Add trout (aka Dream/G1) machine target and some defines
Signed-off-by: Rebecca Schultz <rsch...@google.com>
Signed-off-by: Mike Chan <mi...@android.com>
Signed-off-by: Mike Lockwood <lock...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a225ac6222b3d918c2a1743ee26a9afe581f0e21
Author: Arve Hjønnevåg <ar...@android.com>
Date: Sat Dec 1 19:01:30 2007 -0800
[ARM] msm: add keypad for halibut board
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 45f1e3687ed00089729dcf7155d72af2a3455da2
Author: Brian Swetland <swet...@google.com>
Date: Mon Apr 7 17:59:14 2008 -0700
[ARM] msm: halibut: support fiq mode kernel debugger if configured
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 10c70443a99087e60f8cf7ef997931eec2f5dec8
Author: Mike A. Chan <mike...@google.com>
Date: Wed Oct 29 13:18:57 2008 -0700
[ARM] msm: clock: Simple cpufreq scaling based on screen ON/OFF
Signed-off-by: Mike A. Chan <mike...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit a0d8b5dd2a3157df71666501d57a143da4b66a9b
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Oct 29 21:45:46 2008 -0700
[ARM] msm: halibut: Add memory configuration.
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2b37578e0c0dd47566e6ed9375c75e57fdb86636
Author: Brian Swetland <swet...@google.com>
Date: Mon Apr 7 17:50:22 2008 -0700
[ARM] msm: enable installation of an FIQ handler
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 09a7a9623e6277ff8345d609450024afc5522109
Author: Brian Swetland <swet...@google.com>
Date: Sat Dec 1 19:54:04 2007 -0800
[ARM] msm: pick up partition table from bootloader via private ATAG
It's the wrong way to do this (according to rmk), but we've got
bootloader code that's working this way, so here's support for it.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 2cbe86ec9a2fd093e972c1eb8ab300b0e9be4cdc
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Oct 22 20:46:57 2008 -0700
[ARM] msm: power management - add wakeup support and implement suspend to ram as wait for interrupt
Enter low power states from idle.
Add idle statistics.
If enabled /proc/msm_pm_stats reports time spent in
different idle states.
Spin waiting for interrupts for a while before entering a lower power state.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 1a0b6c83855f90b86686326d2dbe688f4af87abb
Author: Mike A. Chan <mike...@google.com>
Date: Wed Jul 23 18:55:01 2008 -0700
[ARM] msm: clock: Cpu clk stepping.
Jump in 256mhz increments, acpu_freq_tbl has up/down fields
to specify next target freq if greater than 256mhz. Less loop spinning.
Signed-off-by: Mike A. Chan <mike...@google.com>
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6bac17d513d38e1c4523b864dfa4ec98fdea793f
Author: Arve Hjønnevåg <ar...@android.com>
Date: Wed Oct 22 21:32:21 2008 -0700
[ARM] msm: irq: Add sleep support.
Add shadow registers for type and polarity and restore them in resume.
Add wakeup support to main interrupt controller.
Mask interrupt as soon as disable_irq is called. This is needed in some cases to allow power collapse from idle.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 6c0922ac252eef5ecd927c2ef75f2ac88e8857dc
Author: Arve Hjønnevåg <ar...@android.com>
Date: Sat Dec 1 18:05:26 2007 -0800
[ARM] msm: irq: fix lockdep issues
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 7b655b80fafcc9351a00782893910076d2e30c19
Author: Arve Hjønnevåg <ar...@android.com>
Date: Fri May 2 15:30:32 2008 -0700
[ARM] msm: Timer fixes.
Restart timer when entering idle if it appears to have missed.
If the timer should have fired more than a millisecond ago, we
assume that the interrupt was lost and reprogram the timer.
Sync the GP timer with the other core.
Get the time from the other core after low power idle.
Return time until next alarm when entering idle.
Add config option to select default timer.
Fix shift value so multiplier does not overflow.
Add workarounds for some hardware bugs.
Add sched_clock.
Increase min_delta_ns so that we don't get stuck if running at low speed.
Limit number of alarm-already-expired messages.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit b62e1a6a422e475df64baa9e3b1771ca0cd34178
Author: Iliyan Malchev <i...@android.com>
Date: Mon Dec 31 13:11:28 2007 -0800
[ARM] msm_rpc: support rpc servers
- time_remote_mtoa server receives notification of time bases
- dog_keepalive server to handle watchdog events
Signed-off-by: Iliyan Malchev <i...@android.com>
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 4921eb1024f42f9acfcc190f01080fc431efd1f8
Author: Brian Swetland <swet...@google.com>
Date: Sun May 25 22:05:07 2008 -0700
[ARM] msm: provide AMSS version selection mechanism
Some drivers have dependencies on AMSS versions that are not
easily resolved at runtime. Provide a mechanism for selecting
the AMSS version so they can be compiled appropriately.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 55483f7677179a38519fa3484105cbd319aaacb3
Author: San Mehat <s...@android.com>
Date: Sat Dec 1 19:14:34 2007 -0800
[ARM] msm: shared memory rpc router
The RPC router provides access to RPC services running on the baseband
core from kernel and userspace.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: San Mehat <s...@android.com>
Signed-off-by: Iliyan Malchev <i...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 37ec77f5782f35c016e45d844cc6d4917eff428f
Author: Brian Swetland <swet...@google.com>
Date: Sat Dec 1 19:11:31 2007 -0800
[ARM] msm: shared memory qmi interface
The QMI control stream is used to configure the virtual ethernet
network interface to the baseband processor / cellular network.
Publish qmi0, qmi1, and qmi2 (which control the
rmnet0, rmnet1, and rmnet2 interfaces).
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 12e5b0ec87a3178aaee1c1a17c51d74660e1dbf3
Author: Brian Swetland <swet...@google.com>
Date: Sat Dec 1 19:10:02 2007 -0800
[ARM] msm: shared memory tty interface
Provides a simple tty interface to the shared memory "virtual serial" streams.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit e9e316d7b62ee8e80a42c38126567c3406ff19b6
Author: Arve Hjønnevåg <ar...@android.com>
Date: Sat Dec 1 18:47:49 2007 -0800
[ARM] msm: gpio support
Provide an implementation of the generic gpio API for the MSM family.
An extended API provides single-step configuration and access to
edge detect status.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: Arve Hjønnevåg <ar...@android.com>
Signed-off-by: David Brown <dav...@quicinc.com>
commit 483ce1a025f3933376a2abcadb1af8b3b7e1877b
Author: Brian Swetland <swet...@google.com>
Date: Mon Sep 29 16:00:48 2008 -0700
[ARM] msm: shared memory interface for baseband processor ipc
This code provides the low level interface to the "shared memory
state machine" (smsm), and the virtual serial channels (smd), used
to communicate with the baseband processor. Higher level transports
(rpc, ethernet, AT command channel, etc) ride on top of this.
Signed-off-by: Brian Swetland <swet...@google.com>
Signed-off-by: David Brown <dav...@quicinc.com>
-----------------------------------------------------------------------
The branch, qualcomm-2.6.32.9 has been created
This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "linux kernel tree".
hooks/post-receive
--
linux kernel tree