Linux 2.6.13-rc2

Linus Torvalds

unread,

Jul 6, 2005, 2:05:12 AM7/6/05

to Linux Kernel Mailing List

Ok,
-rc3 is pretty small, with the bulk of the diff being some defconfig
updates, and cleanup of xtensa (notably removal of another copy of zlib).

But there are ia64/arm/ppc64 updates and the TSO update from Davem is
probably worth pointing out to people. And various smaller things which
are more easily just seen from the shortlog.

Among the one-liners of note is the silly block level spinlock bugfix that
obviously hit -rc1 and made itself felt on SMP and preempt under moderate
IO loads.

Linus

----
Adrian Bunk:
drivers/ide/Makefile: kill dead CONFIG_BLK_DEV_IDE_TCQ entry
MMC: wbsd cleanups

Alexey Dobriyan:
[NET]: Remove __ARGS from include/net/slhc_vj.h

Andrei Konovalov:
ppc32: add Freescale MPC885ADS board support

Andrew Morton:
fatfs sectioning fix
reiserfs: handle_attrs() fix

Andy Whitcroft:
gregkh-pci-pci-assign-unassigned-resources fix

Arnd Bergmann:
ppc64: simplify nvram partition scanning code

Ben Dooks:
ARM: 2785/1: S3C24XX - serial calls request_irq() with IRQs disabled
ARM: 2783/1: Remove omnimeter_defconfig as there is no kernel support
ARM: 2765/1: S3C24XX - small cleanups in arch/arm/mach-s3c2410
ARM: 2764/1: S3C24XX - Common PM functions for Simtec boards

Bjorn Helgaas:
[IA64] Recognize HP sx2000 chipset

Bruce Losure:
[IA64-SGI] Altix patch to tiocx, add subsys_initcall

Catalin Marinas:
ARM: 2784/1: Fix the block cache flush operation range
ARM: 2780/1: AFS partition length calculation fix
ARM: 2779/1: Fix the V bit setting for the ARM1020x CPUs
ARM: 2778/1: Add -mno-thumb-interwork to CFLAGS_ABI
ARM: 2777/1: Fix broken comment arch/arm/mm/proc-arm1020.S
ARM: 2769/1: cpu_init() stack setup fix

Chris Zankel:
xtensa: Fix asm macro
xtensa: Removed local copy of zlib and fixed O= support
xtensa: Added mm/Kconfig to get a flat memory layout
xtensa: cleanups for errno and ipc.

Christoph Hellwig:
[SHAPER]: Switch to spinlocks.
[SPARC]: bpp: remove sleep_on usage
udf_find_entry() cleanup

Colin Ngam:
[IA64-SGI] Fix TIO IOSPACE MMR Addres

Cornelia Huck:
driver core: add bus_find_device & driver_find_device functions

David Chau:
[NET]: improve readability of dev_set_promiscuity() in net/core/dev.c

David Mosberger-Tang:
[IA64] Replace stale KDB-code with useful MAGIC_SYSRQ code in simserial.c
[IA64] Speed up lfetch.fault [NULL]
[IA64] Fix convert_to_non_syscall() so gdb inferior calls work again
[IA64] Merge audit fix for fsyscalls with syscall-optimizations
[IA64] need r29=psr *after* rsm psr.i
[IA64] use srlz.d instead of srlz.i in ia64_leave_kernel()
[IA64] Annotate fsys_bubble_down() with McKinley dispatch info.
[IA64] Reschedule fsys_bubble_down().
[IA64] Annotate __kernel_syscall_via_epc() with McKinley dispatch info.
[IA64] Reschedule __kernel_syscall_via_epc().
[IA64] Reschedule break_fault() for better performance.
[IA64] In ia64_leave_syscall(), fix comments and whitespace only.
[IA64] Schedule ia64_leave_syscall() to read ar.bsp earlier
[IA64] In syscall-entry, use st8 instead of stf8 to clear pt_regs.r8
[IA64] On return from syscall, hint b7 with __kernel_syscall_via_epc().
[IA64] Schedule fp-clearing insns at least 6 cycles after reading ar.bsp.
[IA64] Use dynamic prediction for RSE-clearing branches.
[IA64] __ia64_syscall() is no longer used anywhere in the kernel. Remove it.

David S. Miller:
[SPARC64]: Fix UltraSPARC-III fallout from membar changes.
[TCP]: Never TSO defer under periods of congestion.
[TCP]: Move to new TSO segmenting scheme.
[TCP]: Break out send buffer expansion test.
[TCP]: Do not call tcp_tso_acked() if no work to do.
[TCP]: Kill bogus comment above tcp_tso_acked().
[TCP]: Fix send-side cpu utiliziation regression.
[TCP]: Eliminate redundant computations in tcp_write_xmit().
[TCP]: Break out tcp_snd_test() into it's constituent parts.
[TCP]: Fix __tcp_push_pending_frames() 'nonagle' handling.
[TCP]: Fix redundant calculations of tcp_current_mss()
[TCP]: tcp_write_xmit() tabbing cleanup
[TCP]: Kill extra cwnd validate in __tcp_push_pending_frames().
[TCP]: Add missing skb_header_release() call to tcp_fragment().
[TCP]: Move __tcp_data_snd_check into tcp_output.c
[TCP]: Move send test logic out of net/tcp.h
[TCP]: Fix quick-ack decrementing with TSO.
[TCP]: Simplify SKB data portion allocation with NETIF_F_SG.
[TG3]: Update driver version and reldate.
[SKGE]: Fix build on big-endian
[SPARC64]: Fix IRQ retry interval timer value on sparc64 PCI controllers.
[SPARC64]: Small Schizo PCI controller programming tweaks.
[SPARC64]: Do proper DMA IRQ syncing on Tomatillo
[SPARC64]: Add support for IRQ pre-handlers.

Denis Vlasenko:
ide: fix line break in ide messages

Dominik Brodowski:
pcmcia: update Documentation
pcmcia: fix modalias attribute in sysfs

Eric Dumazet:
[IPV4]: Bug fix in rt_check_expire()
[IPV4]: Use the fancy alloc_large_system_hash() function for route hash table
[NET]: Hashed spinlocks in net/ipv4/route.c

Eric Paris:
selinux_sb_copy_data() should not require a whole page

Eugene Surovegin:
ppc32: explicitly disable 440GP IRQ compatibility mode in 440GX setup

Greg KH:
Merge rsync://rsync.kernel.org/.../torvalds/linux-2.6

Greg Kroah-Hartman:
PCI: clean up dynamic pci id logic
PCI: Fix up PCI routing in parent bridge
driver core: change bus_rescan_devices to return void
driver core: Add the ability to bind drivers to devices from userspace
driver core: Add the ability to unbind drivers to devices from userspace

Hannes Reinecke:
PCI: Remove newline from pci MODALIAS variable

Heiko Carstens:
s390: fix finish_arch_switch

Herbert Xu:
[IPV6]: Makes IPv6 rcv registration happen last during initialisation.
[IPV4]: Fix crash in ip_rcv while booting related to netconsole
ide: hotplug mark __devinit via82cxxx.c
ide: hotplug mark __devinit triflex.c
ide: hotplug mark __devinit slc90e66.c
ide: hotplug mark __devinit sl82c105.c
ide: hotplug mark __devinit sc1200.c
ide: hotplug mark __devinit opti621.c
ide: hotplug mark __devinit ns87415.c
ide: hotplug mark __devinit it8172.c
ide: hotplug mark __devinit cy82c693.c
ide: hotplug mark __devinit cs5530.c
ide: hotplug mark __devinit amd74xx.c
ide: hotplug mark __devinit alim15x3.c

Hugh Dickins:
Fix get_request nastiness

Ingo Molnar:
x86: i8253/i8259A lock cleanup

Ivan Kokshaysky:
PCI: pci_assign_unassigned_resources() on x86
PCI: handle subtractive decode pci-pci bridge better
alpha smp fix (part #2)
alpha smp fix

Jack Steiner:
[IA64-SGI] - new macros for SGI SN simulator

Jay Lan:
Improper initrd failure message at boot time

Jean Delvare:
PCI: Add PCI quirk for SMBus on the Asus P4B-LX

Jeff Mahoney:
reiserfs: enable attrs by default if saf
reiserfs: Check if attrs are enabled for attr ioctls

Jesper Juhl:
xtensa: use valid_signal()

john stultz:
ppc32: stop misusing NTP's time_offset value

John W. Linville:
pci: cleanup argument comments for pci_{save,restore}_state

Kumar Gala:
ppc32: Fix pointer check for MPC8540 ADS device

Kylene Jo Hall:
tpm: fix bug introduced by the /proc/misc

Linus Torvalds:
Linux v2.6.13-rc3
Merge master.kernel.org:/.../davem/sparc-2.6
Merge master.kernel.org:/.../davem/net-2.6
Merge rsync://rsync.kernel.org/.../davem/sparc-2.6
Merge master.kernel.org:/.../gregkh/pci-2.6
Merge master.kernel.org:/home/rmk/linux-2.6-arm
Merge master.kernel.org:/home/rmk/linux-2.6-serial
Merge master.kernel.org:/home/rmk/linux-2.6-mmc
Merge master.kernel.org:/home/rmk/linux-2.6-arm
If ACPI doesn't find an irq listed, don't accept 0 as a valid PCI irq.
Merge master.kernel.org:/home/rmk/linux-2.6-arm
Merge master.kernel.org:/home/rmk/linux-2.6-serial
Merge master.kernel.org:/.../gregkh/driver-2.6
Merge rsync://rsync.kernel.org/.../paulus/ppc64-2.6
Merge rsync://rsync.kernel.org/.../aegl/linux-2.6
Merge master.kernel.org:/home/rmk/linux-2.6-serial
Merge master.kernel.org:/home/rmk/linux-2.6-arm
Remove duplicate mention of "edd" in Documentation/kernel-parameters.txt

long:
PCI: acpi tg3 ethernet not coming back properly after S3 suspendon DellM70

Mark Maule:
[IA64-SGI] altix: enable vgacon support
[IA64-SGI] pcdp: add PCDP pci interface support

Michael Chan:
[TG3]: support for ethtool -C

Michael Ellerman:
ppc64: Replace custom locking code with a spinlock
ppc64: Formatting cleanups in arch/ppc64/kernel/ItLpQueue.c
ppc64: Cleanup whitespace in arch/ppc64/kernel/ItLpQueue.c
ppc64: Cleanup proc printing of event types
ppc64: Simplify counting of lpevents, remove lpevent_count from paca
ppc64: Don't count number of events processed for caller
ppc64: Rename ItLpQueue_* functions to hvlpevent_queue_*
ppc64: Rename xItLpQueue to hvlpevent_queue
ppc64: Move definition of xItLpQueue
ppc64: Make two ItLpQueue related functions static
ppc64: Move xItLpQueue proc code into ItLpQueue.c
ppc64: Move initialisation of xItLpQueue into ItLpQueue.c
ppc64: Don't pass the pointers to xItLpQueue around
ppc64: Reorganise the paca initialisation macros
ppc64: Move set_spread_lpevents() into ItLpQueue.c
ppc64: Spread lpevents by default on iSeries
ppc64: Remove lpqueue pointer from the paca on iSeries

Nathan Lynch:
ppc64: don't create spurious symlinks under node0 sysdev

Nicolas Pitre:
ARM: 2723/2: remove __udivdi3 and __umoddi3 from the kernel

Nishanth Aravamudan:
ARM: replace schedule_timeout() with msleep()

Olaf Hering:
ppc32: use correct register names in arch/ppc/kernel/relocate_kernel.S
remove duplicate printf in arch/ppc64/boot/main.c
remove printk usage in arch/ppc64/boot/prom.c
remove unused arch/ppc64/boot/mknote.c
remove unused arch/ppc64/boot/piggyback.c

Patrick McHardy:
[IPV4]: Handle large allocations in fib_trie
[NET]: net/core/filter.c: make len cover the entire packet
[NET]: Consolidate common code in net/core/filter.c
[NET]: Remove redundant code in net/core/filter.c
[NET]: Fix signedness issues in net/core/filter.c

Patrick Mochel:
Driver core: Use klist_del() instead of klist_remove().

Pekka Enberg:
freevxfs: minor cleanups
freevxfs: remove 2.4 compatability
freevxfs: fix buffer_head leak

Pekka J Enberg:
fat: fix slab cache leak

Peter Chubb:
[IA64] Fix another IA64 preemption problem

Pierre Ossman:
MMC: wbsd delayed insertion

Prakash Punnoor:
Don't fill up log with atxp1 vcore messages change message

Prarit Bhargava:
[IA64] sparse cleanup of TIOCA files
[IA64] sparse cleanup of shub_mmr.h

rajes...@intel.com:
PCI: Increase the number of PCI bus resources

Raphael Assenat:
[SPARC64/COMPAT]: Add some compat ioctl for ppdev

Richard Purdie:
ARM: 2768/1: PXA: Add a required header file for LL_DEBUG

Rob Punkunus:
amd74xx: support MCP55 device IDs

Robert Olsson:
[IPV4]: Add LC-Trie implementation notes
[IPV4]: More broken memory allocation fixes for fib_trie

Russell King:
ARM: Fix new-ABI layout of struct stat64
ARM: Fix non-standard PXA io_pg_offst initialisers
ARM: Change 'param_offset' to 'boot_params'
Serial: Fix console port spinlock initialisation
ARM: Remove machine description macros
MMC: Fix divdi3 reference in mmci.c
ARM: Make the magic values in head.S more obvious
Serial: Fix small CONFIG_SERIAL_8250_NR_UARTS
ARM: Acornfb: Don't claim IRQ fbcon for cursor
ARM: Don't try to send a signal to pid0
ARM: Don't force SIGFPE
ARM: Fix VFP to use do_div()
Serial: Split 8250 port table (part 2)
Serial: Disable OX950 transmitter for flow control
Serial: Check status of CTS when using flow control
ARM: Remove nmi_tick from integrator platforms
ARM: Convert ARM timer implementations to use readl/writel
Serial: Adjust serial locking
Merge with ../linux-2.6-smp
ARM SMP: __xchg support
ARM SMP: Add IPI support code for SMP TLB flushing
ARM SMP: Use local_flush_tlb* where we really want to be local
ARM SMP: TLB implementations only affect local CPU

Rusty Lynch:
kprobes: fix namespace problem and sparc64 build

Thomas Graf:
[PKT_SCHED]: Blackhole queueing discipline
[DECNET]: Fix memset overflow on 64bit archs while dumping decnet routing rules
[PKT_SCHED]: Report rate estimator configuration errors during qdisc allocation
[PKT_SCHED]: Cleanup qdisc creation and alignment macros
[PKT_SCHED]: Move sch_generic.c prototypes to correct header file
[NET]: Reduce size of sk_buff by 4 bytes
[NET]: Remove unused security member in sk_buff

Todd Poynor:
ARM: 2782/1: PXA27x MDREFR K0DB4 define
ARM: 2781/2: PXA27x Standby mode take 2

Tony Lindgren:
ARM: 2771/1: Dynamic Tick support for OMAP, take 4

Tony Luck:
[IA64] Update zx1_defconfig
[IA64] Update tiger_defconfig
Auto merge with /home/aegl/GIT/linus
Auto merge with /home/aegl/GIT/ia64-test
Auto merge with /home/aegl/GIT/linus
Merge with temp tree to get David's gdb inferior calls patch

Jan Dittmer

unread,

Jul 6, 2005, 4:50:05 AM7/6/05

to Linus Torvalds, Linux Kernel Mailing List

Linus Torvalds wrote:
> Ok,
> -rc3 is pretty small, with the bulk of the diff being some defconfig

..
> Linus Torvalds:
> Linux v2.6.13-rc3

Confused?!

--
Jan

Matt Mackall

unread,

Jul 6, 2005, 5:44:08 AM7/6/05

to Linus Torvalds, Linux Kernel Mailing List

On Tue, Jul 05, 2005 at 09:32:34PM -0700, Linus Torvalds wrote:
>
> Ok,
> -rc3 is pretty small, with the bulk of the diff being some defconfig
> updates, and cleanup of xtensa (notably removal of another copy of zlib).

Hmm.

-rc2:
in title, in tags, in makefile, in patch file name
-rc3:
in git commit message and in body of email

-rc3 appears to be outvoted.

--
Mathematics is the supreme nostalgia of our time.

Florian Weimer

unread,

Jul 6, 2005, 7:39:12 AM7/6/05

to Linus Torvalds, Linux Kernel Mailing List

* Linus Torvalds:

> Ok,
> -rc3 is pretty small,

Is it -rc2 or -rc3? (Makefile and log message don't agree, either.)

Rafael J. Wysocki

unread,

Jul 6, 2005, 8:43:49 AM7/6/05

to Linus Torvalds, Linux Kernel Mailing List

On Wednesday, 6 of July 2005 06:32, Linus Torvalds wrote:
>
> Ok,
> -rc3 is pretty small, with the bulk of the diff being some defconfig
> updates, and cleanup of xtensa (notably removal of another copy of zlib).
>
> But there are ia64/arm/ppc64 updates and the TSO update from Davem is
> probably worth pointing out to people. And various smaller things which
> are more easily just seen from the shortlog.
>
> Among the one-liners of note is the silly block level spinlock bugfix that
> obviously hit -rc1 and made itself felt on SMP and preempt under moderate
> IO loads.

On my box (Asus L5D, Athlon 64 + nForce3, 64-bit kernel) there is a problem
with PCMCIA:

albercik:~ # cardmgr
cardmgr[4702]: no sockets found!

but all of the relevant modules are loaded. The output of dmesg is attached.

It works fine with -rc1.

Greets,
Rafael

--
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
-- Lewis Carroll "Alice's Adventures in Wonderland"

dmesg.out

Eyal Lebedinsky

unread,

Jul 6, 2005, 9:05:25 AM7/6/05

to Linus Torvalds, Linux Kernel Mailing List

CC [M] sound/pci/bt87x.o
sound/pci/bt87x.c: In function `snd_bt87x_detect_card':
sound/pci/bt87x.c:807: error: `driver' undeclared (first use in this function)
sound/pci/bt87x.c:807: error: (Each undeclared identifier is reported only once
sound/pci/bt87x.c:807: error: for each function it appears in.)
sound/pci/bt87x.c: At top level:
sound/pci/bt87x.c:910: error: `driver' used prior to declaration

--
Eyal Lebedinsky (ey...@eyal.emu.id.au) <http://samba.org/eyal/>

Tomasz Torcz

unread,

Jul 6, 2005, 9:30:26 AM7/6/05

to Linus Torvalds

CC [M] sound/pci/bt87x.o
sound/pci/bt87x.c: In function `snd_bt87x_detect_card':
sound/pci/bt87x.c:807: error: `driver' undeclared (first use in this function)
sound/pci/bt87x.c:807: error: (Each undeclared identifier is reported only once
sound/pci/bt87x.c:807: error: for each function it appears in.)

make[2]: *** [sound/pci/bt87x.o] Error 1
make[1]: *** [sound/pci] Error 2
make: *** [sound] Error 2

--
Tomasz Torcz Only gods can safely risk perfection,
zdzichu@irc.-nie.spam-.pl it's a dangerous thing for a man. -- Alia

Paolo Ciarrocchi

unread,

Jul 6, 2005, 12:40:10 PM7/6/05

to Florian Weimer, Linus Torvalds, Linux Kernel Mailing List

2005/7/6, Florian Weimer <f...@deneb.enyo.de>:

> * Linus Torvalds:
>
> > Ok,
> > -rc3 is pretty small,
>
> Is it -rc2 or -rc3? (Makefile and log message don't agree, either.)

-rc2

--
http://paoloc.blogspot.com

Linus Torvalds

unread,

Jul 6, 2005, 4:06:42 PM7/6/05

to Dominik Brodowski, ak...@osdl.org, Rafael J. Wysocki, Greg KH, Linux Kernel Mailing List

On Wed, 6 Jul 2005, Greg KH wrote:
>
> On Wed, Jul 06, 2005 at 11:28:49AM +0200, Rafael J. Wysocki wrote:
> > PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.0
> > PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.1
> > PCI: Failed to allocate I/O resource #7:1000@e000 for 0000:02:01.1
> > PCI: Failed to allocate I/O resource #8:1000@e000 for 0000:02:01.1
>
> Do you also get the above errors on booting with -rc1?

I'd assume so - there are basically no resource changes in -rc2, but -rc1
has a lot of PCMCIA updates. Dominik - the full dmesg is in the original
report by Rafael on linux-kernel, mind taking a look?

However, the above also seems to have tried to allocate a 64-bit resource,
and I wonder if that's right. Rafael, what does lspci say? Is that 2:1.1
device actually 64-bit capable? Sounds unlikely (they are pretty rare),
and hat would be a PCI layer bug if not.

> > PCI: Bus 3, cardbus bridge: 0000:02:01.0
> > IO window: 0000b000-0000bfff
> > IO window: 0000c000-0000cfff
> > PREFETCH window: 30000000-31ffffff
> > PCI: Bus 7, cardbus bridge: 0000:02:01.1
> > PREFETCH window: 32000000-33ffffff
> > PCI: Bridge: 0000:00:0a.0
> > IO window: b000-dfff
> > MEM window: f8a00000-feafffff
> > PREFETCH window: 30000000-34ffffff
> > PCI: Bridge: 0000:00:0b.0
> > IO window: disabled.
> > MEM window: f6900000-f89fffff
> > PREFETCH window: c6800000-e67fffff
> > PCI: Setting latency timer of device 0000:00:0a.0 to 64
> > PCI: Device 0000:02:01.0 not available because of resource collisions
> > PCI: Device 0000:02:01.1 not available because of resource collisions
>
> And these? As this is your carbus bridge, I'm guessing that is why -rc2
> is failing for you.

Indeed, if 2:1.1 is your cardbus bridge, I can pretty much guarantee that
it's not using 64-bit windows. So Greg, I think you might be at least
partially involved..

Linus

Rafael J. Wysocki

unread,

Jul 6, 2005, 4:50:14 PM7/6/05

to Greg KH, Linus Torvalds, Linux Kernel Mailing List

On Wednesday, 6 of July 2005 18:47, Greg KH wrote:
> On Wed, Jul 06, 2005 at 11:28:49AM +0200, Rafael J. Wysocki wrote:
> > PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.0
> > PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.1
> > PCI: Failed to allocate I/O resource #7:1000@e000 for 0000:02:01.1
> > PCI: Failed to allocate I/O resource #8:1000@e000 for 0000:02:01.1
>
> Do you also get the above errors on booting with -rc1?

No, I don't.

> > PCI: Bus 3, cardbus bridge: 0000:02:01.0
> > IO window: 0000b000-0000bfff
> > IO window: 0000c000-0000cfff
> > PREFETCH window: 30000000-31ffffff
> > PCI: Bus 7, cardbus bridge: 0000:02:01.1
> > PREFETCH window: 32000000-33ffffff
> > PCI: Bridge: 0000:00:0a.0
> > IO window: b000-dfff
> > MEM window: f8a00000-feafffff
> > PREFETCH window: 30000000-34ffffff
> > PCI: Bridge: 0000:00:0b.0
> > IO window: disabled.
> > MEM window: f6900000-f89fffff
> > PREFETCH window: c6800000-e67fffff
> > PCI: Setting latency timer of device 0000:00:0a.0 to 64
> > PCI: Device 0000:02:01.0 not available because of resource collisions
> > PCI: Device 0000:02:01.1 not available because of resource collisions
>
> And these?

No ...

> As this is your carbus bridge, I'm guessing that is why -rc2
> is failing for you.

I guess so.

The dmesg output from -rc1 is attached for reference.

dmesg-rc1.out

Rafael J. Wysocki

unread,

Jul 6, 2005, 4:50:14 PM7/6/05

to Linus Torvalds, Dominik Brodowski, ak...@osdl.org, Greg KH, Linux Kernel Mailing List

On Wednesday, 6 of July 2005 19:16, Linus Torvalds wrote:
>
> On Wed, 6 Jul 2005, Greg KH wrote:
> >
> > On Wed, Jul 06, 2005 at 11:28:49AM +0200, Rafael J. Wysocki wrote:
> > > PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.0
> > > PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.1
> > > PCI: Failed to allocate I/O resource #7:1000@e000 for 0000:02:01.1
> > > PCI: Failed to allocate I/O resource #8:1000@e000 for 0000:02:01.1
> >
> > Do you also get the above errors on booting with -rc1?
>
> I'd assume so - there are basically no resource changes in -rc2, but -rc1
> has a lot of PCMCIA updates. Dominik - the full dmesg is in the original
> report by Rafael on linux-kernel, mind taking a look?
>
> However, the above also seems to have tried to allocate a 64-bit resource,
> and I wonder if that's right. Rafael, what does lspci say? Is that 2:1.1
> device actually 64-bit capable? Sounds unlikely (they are pretty rare),
> and hat would be a PCI layer bug if not.

'lspci -v' says:

0000:02:01.0 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
Subsystem: ASUSTeK Computer Inc.: Unknown device 1854
Flags: bus master, medium devsel, latency 64
Memory at fd200000 (32-bit, non-prefetchable) [size=4K]
Bus: primary=02, secondary=03, subordinate=06, sec-latency=176
Memory window 0: 30000000-31fff000 (prefetchable)
Memory window 1: fc600000-fd1ff000
I/O window 0: 0000b000-0000bfff
I/O window 1: 0000c000-0000cfff
16-bit legacy interface ports at 0001

0000:02:01.1 CardBus bridge: Ricoh Co Ltd RL5c476 II (rev ab)
Subsystem: ASUSTeK Computer Inc.: Unknown device 1854
Flags: bus master, medium devsel, latency 64
Memory at fa200000 (32-bit, non-prefetchable) [size=4K]
Bus: primary=02, secondary=07, subordinate=0a, sec-latency=176
Memory window 0: 32000000-33fff000 (prefetchable)
Memory window 1: f9600000-fa1ff000
I/O window 0: 0000b000-0000b7ff
I/O window 1: 0000b800-0000bfff
16-bit legacy interface ports at 0001

Greets,
Rafael

--
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
-- Lewis Carroll "Alice's Adventures in Wonderland"

Ivan Kokshaysky

unread,

Jul 6, 2005, 5:50:52 PM7/6/05

to Rafael J. Wysocki, Linus Torvalds, Linux Kernel Mailing List

On Wed, Jul 06, 2005 at 11:28:49AM +0200, Rafael J. Wysocki wrote:

> albercik:~ # cardmgr
> cardmgr[4702]: no sockets found!

..

> PCI: Device 0000:02:01.0 not available because of resource collisions
> PCI: Device 0000:02:01.1 not available because of resource collisions

Thanks for the report.
Does the appended one-liner fix that?

Ivan.

--- 2.6.13-rc2/drivers/pci/setup-bus.c Thu Jul 7 01:30:58 2005
+++ linux/drivers/pci/setup-bus.c Thu Jul 7 01:32:43 2005
@@ -74,6 +74,7 @@ pbus_assign_resources_sorted(struct pci_
idx = res - &list->dev->resource[0];
if (pci_assign_resource(list->dev, idx)) {
res->start = 0;
+ res->end = 0;
res->flags = 0;
}
tmp = list;

Rafael J. Wysocki

unread,

Jul 6, 2005, 6:33:26 PM7/6/05

to Ivan Kokshaysky, Linus Torvalds, Linux Kernel Mailing List

On Wednesday, 6 of July 2005 23:43, Ivan Kokshaysky wrote:
> On Wed, Jul 06, 2005 at 11:28:49AM +0200, Rafael J. Wysocki wrote:
> > albercik:~ # cardmgr
> > cardmgr[4702]: no sockets found!
> ..
> > PCI: Device 0000:02:01.0 not available because of resource collisions
> > PCI: Device 0000:02:01.1 not available because of resource collisions
>
> Thanks for the report.
> Does the appended one-liner fix that?

Yes, it does, but I'm still getting these in dmesg:

PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.0
PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.1
PCI: Failed to allocate I/O resource #7:1000@e000 for 0000:02:01.1
PCI: Failed to allocate I/O resource #8:1000@e000 for 0000:02:01.1

Greets,
Rafael

--
- Would you tell me, please, which way I ought to go from here?
- That depends a good deal on where you want to get to.
-- Lewis Carroll "Alice's Adventures in Wonderland"

Ivan Kokshaysky

unread,

Jul 6, 2005, 7:09:25 PM7/6/05

to Rafael J. Wysocki, Linus Torvalds, Linux Kernel Mailing List

On Thu, Jul 07, 2005 at 12:15:19AM +0200, Rafael J. Wysocki wrote:
> > Does the appended one-liner fix that?
>
> Yes, it does, but I'm still getting these in dmesg:
>
> PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.0
> PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.1
> PCI: Failed to allocate I/O resource #7:1000@e000 for 0000:02:01.1
> PCI: Failed to allocate I/O resource #8:1000@e000 for 0000:02:01.1

It's OK. Eventually we move such stuff into "debug" level.

Thanks,

Ivan.

Greg KH

unread,

Jul 6, 2005, 10:12:31 PM7/6/05

to Rafael J. Wysocki, Linus Torvalds, Linux Kernel Mailing List

On Wed, Jul 06, 2005 at 11:28:49AM +0200, Rafael J. Wysocki wrote:
> PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.0
> PCI: Failed to allocate mem resource #10:2000000@100000000 for 0000:02:01.1
> PCI: Failed to allocate I/O resource #7:1000@e000 for 0000:02:01.1
> PCI: Failed to allocate I/O resource #8:1000@e000 for 0000:02:01.1

Do you also get the above errors on booting with -rc1?

> PCI: Bus 3, cardbus bridge: 0000:02:01.0

> IO window: 0000b000-0000bfff
> IO window: 0000c000-0000cfff
> PREFETCH window: 30000000-31ffffff
> PCI: Bus 7, cardbus bridge: 0000:02:01.1
> PREFETCH window: 32000000-33ffffff
> PCI: Bridge: 0000:00:0a.0
> IO window: b000-dfff
> MEM window: f8a00000-feafffff
> PREFETCH window: 30000000-34ffffff
> PCI: Bridge: 0000:00:0b.0
> IO window: disabled.
> MEM window: f6900000-f89fffff
> PREFETCH window: c6800000-e67fffff
> PCI: Setting latency timer of device 0000:00:0a.0 to 64

> PCI: Device 0000:02:01.0 not available because of resource collisions
> PCI: Device 0000:02:01.1 not available because of resource collisions

And these? As this is your carbus bridge, I'm guessing that is why -rc2
is failing for you.

thanks,

greg k-h

Greg KH

unread,

Jul 6, 2005, 10:26:14 PM7/6/05

to Linus Torvalds, Eyal Lebedinsky, Matthias Andree, Linux Kernel Mailing List

On Wed, Jul 06, 2005 at 09:22:05AM -0700, Linus Torvalds wrote:
>
>
> On Wed, 6 Jul 2005, Greg KH wrote:
> >

> > --- gregkh-2.6.orig/sound/pci/bt87x.c 2005-07-06 08:48:29.000000000 -0700
> > +++ gregkh-2.6/sound/pci/bt87x.c 2005-07-06 08:48:54.000000000 -0700
> > @@ -798,6 +798,8 @@
> > {0x270f, 0xfc00}, /* Chaintech Digitop DST-1000 DVB-S */
> > };
> >
> > +static struct pci_driver driver;
> > +
>
> Hmm.. Shouldn't you at a _minimum_ initialize the name and owner fields?

It's done so a few functions down. This is just a "forward" reference
to the real thing there.

Linus Torvalds

unread,

Jul 6, 2005, 10:28:47 PM7/6/05

to Greg KH, Eyal Lebedinsky, Matthias Andree, Linux Kernel Mailing List

On Wed, 6 Jul 2005, Greg KH wrote:
>
> --- gregkh-2.6.orig/sound/pci/bt87x.c 2005-07-06 08:48:29.000000000 -0700
> +++ gregkh-2.6/sound/pci/bt87x.c 2005-07-06 08:48:54.000000000 -0700
> @@ -798,6 +798,8 @@
> {0x270f, 0xfc00}, /* Chaintech Digitop DST-1000 DVB-S */
> };
>
> +static struct pci_driver driver;
> +

Hmm.. Shouldn't you at a _minimum_ initialize the name and owner fields?

Linus

Linus Torvalds

unread,

Jul 6, 2005, 11:01:59 PM7/6/05

to Jan Dittmer, Linux Kernel Mailing List

On Wed, 6 Jul 2005, Jan Dittmer wrote:
>
> Linus Torvalds wrote:
> > Ok,
> > -rc3 is pretty small, with the bulk of the diff being some defconfig

> ...

> > Linus Torvalds:
> > Linux v2.6.13-rc3
>
> Confused?!

Constantly.

Let's hope that commit naming bug was the worst part of the release..

Linus

Jan Dittmer

unread,

Jul 6, 2005, 11:13:34 PM7/6/05

to Linus Torvalds, Linux Kernel Mailing List, Greg KH

Linus Torvalds wrote:
>
> On Wed, 6 Jul 2005, Jan Dittmer wrote:
>
>>Linus Torvalds wrote:
>>
>>>Ok,
>>> -rc3 is pretty small, with the bulk of the diff being some defconfig
>>
>>...
>>
>>>Linus Torvalds:
>>> Linux v2.6.13-rc3
>>
>>Confused?!
>
>
> Constantly.
>
> Let's hope that commit naming bug was the worst part of the release..

Nah, compared to git7 you (or greg?) managed to break alpha, sparc and x86_64.

Jan

Comparing 2.6.13-rc1-git7 to 2.6.13-rc2 (defconfig)
===================================================

- alpha: broke
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: request for member `node' in something not a structure or union
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:157: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:159: error: dereferencing pointer to incomplete type
make[3]: *** [drivers/pci/pci-driver.o] Error 1
make[2]: *** [drivers/pci] Error 2
make[1]: *** [drivers] Error 2
make: *** [_all] Error 2

Details: http://l4x.org/k/?d=5184

- sparc: broke
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: request for member `node' in something not a structure or union
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:157: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:159: error: dereferencing pointer to incomplete type
make[3]: *** [drivers/pci/pci-driver.o] Error 1
make[2]: *** [drivers/pci] Error 2
make[1]: *** [drivers] Error 2
make: *** [_all] Error 2

Details: http://l4x.org/k/?d=5204

- x86_64: broke
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: error: request for member `node' in something not a structure or union
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:157: error: dereferencing pointer to incomplete type
/usr/src/ctest/rc/kernel/drivers/pci/pci-driver.c:159: error: dereferencing pointer to incomplete type
make[3]: *** [drivers/pci/pci-driver.o] Error 1
make[2]: *** [drivers/pci] Error 2
make[1]: *** [drivers] Error 2
make: *** [_all] Error 2

Details: http://l4x.org/k/?d=5208

- arm26: still broken
Details: http://l4x.org/k/?d=5186

- cris: still broken
Details: http://l4x.org/k/?d=5187

- frv: still broken
Details: http://l4x.org/k/?d=5188

- m68k: still broken
Details: http://l4x.org/k/?d=5193

- m68knommu: still broken
Details: http://l4x.org/k/?d=5195

- parisc: still broken
Details: http://l4x.org/k/?d=5197

- s390: still broken
Details: http://l4x.org/k/?d=5201

- sh: still broken
Details: http://l4x.org/k/?d=5202

- sh64: still broken
Details: http://l4x.org/k/?d=5203

- sparc64: still broken
Details: http://l4x.org/k/?d=5205

- v850: still broken
Details: http://l4x.org/k/?d=5207

- xtensa: still broken
Details: http://l4x.org/k/?d=5209

Summary: 9 ok, 14 failed

Steven Cole

unread,

Jul 6, 2005, 11:15:58 PM7/6/05

to Linus Torvalds, Linux Kernel Mailing List, Greg KH

Linus Torvalds wrote:
> Ok,
> -rc3 is pretty small, with the bulk of the diff being some defconfig
> updates, and cleanup of xtensa (notably removal of another copy of zlib).
>

> Greg Kroah-Hartman:

> PCI: clean up dynamic pci id logic
> PCI: Fix up PCI routing in parent bridge

Without CONFIG_HOTPLUG and CONFIG_HOTPLUG_PCI, I got this:

CC drivers/pci/pci-driver.o
drivers/pci/pci-driver.c: In function `pci_match_device':

drivers/pci/pci-driver.c:156: error: dereferencing pointer to incomplete type

drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `type name'

drivers/pci/pci-driver.c:156: error: request for member `node' in something not a structure or union

drivers/pci/pci-driver.c:156: warning: type defaults to `int' in declaration of `__mptr'
drivers/pci/pci-driver.c:156: warning: initialization from incompatible pointer type
[snipped similar errors/warnings]
make[2]: *** [drivers/pci/pci-driver.o] Error 1
make[1]: *** [drivers/pci] Error 2
make: *** [drivers] Error 2

Setting CONFIG_HOTPLUG=y and CONFIG_HOTPLUG_PCI=y allowed 2.6.13-rc2 to build/boot/run.

Steven

--
This e-mail contains no programmatic content requiring independent ADC review.

Greg KH

unread,

Jul 6, 2005, 11:18:22 PM7/6/05

to Linus Torvalds, Eyal Lebedinsky, Matthias Andree, Linux Kernel Mailing List

On Wed, Jul 06, 2005 at 08:42:16AM -0700, Linus Torvalds wrote:

>
>
> On Wed, 6 Jul 2005, Eyal Lebedinsky wrote:
> >
> > CC [M] sound/pci/bt87x.o
> > sound/pci/bt87x.c: In function `snd_bt87x_detect_card':
> > sound/pci/bt87x.c:807: error: `driver' undeclared (first use in this function)
> > sound/pci/bt87x.c:807: error: (Each undeclared identifier is reported only once
> > sound/pci/bt87x.c:807: error: for each function it appears in.)

> > sound/pci/bt87x.c: At top level:
> > sound/pci/bt87x.c:910: error: `driver' used prior to declaration
>

> This seems to be a thinko by Greg. That line got changed from
>
> supported = pci_match_device(snd_bt87x_ids, pci);
>
> to
>
> supported = pci_match_device(driver, pci);
>
> but as far as I can tell it _should_ be
>
> supported = pci_match_id(snd_bt87x_ids, pci);

No, I wanted it to be "driver", but forgot to build the code, sorry.
Try the following patch instead:

thanks,

greg k-h

-------------------

Fixes bt87x.c build problem.

Signed-off-by: Greg Kroah-Hartman <gre...@suse.de>

---
sound/pci/bt87x.c | 2 ++
1 files changed, 2 insertions(+)

--- gregkh-2.6.orig/sound/pci/bt87x.c 2005-07-06 08:48:29.000000000 -0700
+++ gregkh-2.6/sound/pci/bt87x.c 2005-07-06 08:48:54.000000000 -0700
@@ -798,6 +798,8 @@
{0x270f, 0xfc00}, /* Chaintech Digitop DST-1000 DVB-S */
};

+static struct pci_driver driver;
+

/* return the rate of the card, or a negative value if it's blacklisted */
static int __devinit snd_bt87x_detect_card(struct pci_dev *pci)
{

Linus Torvalds

unread,

Jul 6, 2005, 11:27:13 PM7/6/05

to Eyal Lebedinsky, Matthias Andree, Greg KH, Linux Kernel Mailing List

On Wed, 6 Jul 2005, Eyal Lebedinsky wrote:
>

> CC [M] sound/pci/bt87x.o
> sound/pci/bt87x.c: In function `snd_bt87x_detect_card':
> sound/pci/bt87x.c:807: error: `driver' undeclared (first use in this function)
> sound/pci/bt87x.c:807: error: (Each undeclared identifier is reported only once
> sound/pci/bt87x.c:807: error: for each function it appears in.)
> sound/pci/bt87x.c: At top level:
> sound/pci/bt87x.c:910: error: `driver' used prior to declaration

This seems to be a thinko by Greg. That line got changed from

supported = pci_match_device(snd_bt87x_ids, pci);

to

supported = pci_match_device(driver, pci);

but as far as I can tell it _should_ be

supported = pci_match_id(snd_bt87x_ids, pci);

does that fix it for you?

Linus

Greg KH

unread,

Jul 6, 2005, 11:29:28 PM7/6/05

to Jan Dittmer, Linus Torvalds, Linux Kernel Mailing List

On Wed, Jul 06, 2005 at 05:53:05PM +0200, Jan Dittmer wrote:
> Linus Torvalds wrote:
> >
> > On Wed, 6 Jul 2005, Jan Dittmer wrote:
> >
> >>Linus Torvalds wrote:
> >>
> >>>Ok,
> >>> -rc3 is pretty small, with the bulk of the diff being some defconfig
> >>
> >>...
> >>
> >>>Linus Torvalds:
> >>> Linux v2.6.13-rc3
> >>
> >>Confused?!
> >
> >
> > Constantly.
> >
> > Let's hope that commit naming bug was the worst part of the release..
>
> Nah, compared to git7 you (or greg?) managed to break alpha, sparc and x86_64.

This was a CONFIG_HOTPLUG issue, I'll fix it in a bit...

thanks,

greg k-h

Alexandre Buisse

unread,

Jul 6, 2005, 11:59:31 PM7/6/05

to Tomasz Torcz, linux-...@vger.kernel.org

Tomasz Torcz wrote:

> CC [M] sound/pci/bt87x.o
>sound/pci/bt87x.c: In function `snd_bt87x_detect_card':
>sound/pci/bt87x.c:807: error: `driver' undeclared (first use in this function)
>sound/pci/bt87x.c:807: error: (Each undeclared identifier is reported only once
>sound/pci/bt87x.c:807: error: for each function it appears in.)
>make[2]: *** [sound/pci/bt87x.o] Error 1
>make[1]: *** [sound/pci] Error 2
>make: *** [sound] Error 2
>
>

Hi,

The attached patch should fix it.

Regards,
Alexandre

Signed-off-by: Alexandre Buisse <Alexandr...@ens-lyon.fr>

bt87x-driver-2.6.13-rc2.5.patch

Takashi Iwai

unread,

Jul 7, 2005, 5:49:15 AM7/7/05

to Greg KH, Linus Torvalds, Eyal Lebedinsky, Matthias Andree, Linux Kernel Mailing List

At Wed, 6 Jul 2005 08:51:03 -0700,

Greg KH wrote:
>
> On Wed, Jul 06, 2005 at 08:42:16AM -0700, Linus Torvalds wrote:
> >
> >
> > On Wed, 6 Jul 2005, Eyal Lebedinsky wrote:
> > >
> > > CC [M] sound/pci/bt87x.o
> > > sound/pci/bt87x.c: In function `snd_bt87x_detect_card':
> > > sound/pci/bt87x.c:807: error: `driver' undeclared (first use in this function)
> > > sound/pci/bt87x.c:807: error: (Each undeclared identifier is reported only once
> > > sound/pci/bt87x.c:807: error: for each function it appears in.)
> > > sound/pci/bt87x.c: At top level:
> > > sound/pci/bt87x.c:910: error: `driver' used prior to declaration
> >
> > This seems to be a thinko by Greg. That line got changed from
> >
> > supported = pci_match_device(snd_bt87x_ids, pci);
> >
> > to
> >
> > supported = pci_match_device(driver, pci);
> >
> > but as far as I can tell it _should_ be
> >
> > supported = pci_match_id(snd_bt87x_ids, pci);
>
> No, I wanted it to be "driver", but forgot to build the code, sorry.
> Try the following patch instead:

Well, I don't think pci_match_device() makes sense here.
snd_bt87x_detect_card() checks the sample rate of a known device
(listed in the static pci id table), and doesn't need to check the
dynamic devices.
Calling pci_match_id() would be more feasible.

Takashi

Greg KH

unread,

Jul 7, 2005, 3:44:44 PM7/7/05

to Takashi Iwai, Linus Torvalds, Eyal Lebedinsky, Matthias Andree, Linux Kernel Mailing List

Ok, that's fine with me if you want to do that instead.

thanks,

greg k-h

Takashi Iwai

unread,

Jul 7, 2005, 4:30:20 PM7/7/05

to Greg KH, Linus Torvalds, Eyal Lebedinsky, Matthias Andree, Linux Kernel Mailing List

At Thu, 7 Jul 2005 11:39:29 -0700,

Since Linus already applied the fix patches, I'll leave as it is ;)

pci_match_device() should work identically like pci_match_id() in the
above case since driver_data is 0.

thanks,

Takashi