Arrr! Linux v2.6.14-rc2
Linus Torvalds
Ahoy landlubbers!
Here be t' Linux-2.6.14-rc2 release.
Not a whole lot o' excitement, ye scurvy dogs, but it has t' ALSA, LSM,
audit and watchdog merges that be missed from -rc1, and a merge series
with Andrew. But on t' whole pretty reasonable - you can see t' details in
the shortlog (appended).
Arrr!
Linus
---
Abhay Salunke:
dell_rbu: enhancements and fixes
Adam Kropelin:
ibmphp: Use dword accessors for PCI_ROM_ADDRESS
pciehp: Use dword accessors for PCI_ROM_ADDRESS
shpchp: Use dword accessors for PCI_ROM_ADDRESS
qla2xxx: Use dword accessors for PCI_ROM_ADDRESS
Adrian Bunk:
SECURITY must depend on SYSFS
Al Viro:
epca iomem annotations + several missing readw()
Alan Cox:
ide: clean up the garbage in eighty_ninty_three
Alexander Nyberg:
Fix fs/exec.c:788 (de_thread()) BUG_ON
Alexey Dobriyan:
[8021Q]: Add endian annotations.
Alok Kataria:
Fix slab BUG_ON() triggered by change in array cache size
Amy Griffis:
AUDIT: Prevent duplicate syscall rules
Andi Kleen:
Fix MPOL_F_VERIFY
Make BUILD_BUG_ON fail at compile time.
x86_64: Export end_pfn
Andrew Morton:
[WATCHDOG] driver-for-ibm-automatic-server-restart-watchdog-fix
i2c-keywest warning fix
set_current_state() commentary
schedule_timeout_[un]interruptible() speedup
s2io warning fixes
x86_64: e820.c needs module.h
x86_64: desc.h-needs smp.h
seclvl-use-securityfs tidy
dell_rbu tidy
joystick-vs-x.org fix
Andrey Panin:
[WATCHDOG] driver-for-ibm-automatic-server-restart-watchdog.patch
Anton Altaparmakov:
NTFS: Fix various bugs in the runlist merging code. (Based on libntfs
NTFS: Fix handling of compressed directories that I broke in earlier changeset.
NTFS: Fix ntfs_{read,write}page() to cope with concurrent truncates better.
Anton Blanchard:
hvc_console: start kernel thread before registering tty
ppc64: build fix
Antonino A. Daplas:
fbdev Kconfig fix
nv_i2c oops fix
savagefb: Fix load failure of the Twister chipset
vgacon: Fix sanity checking in vgacon_resize
vc: Use correct size on buffer copy in vc_resize
Arnaldo Carvalho de Melo:
[DCCP]: Check if already in the CLOSING state in dccp_rcv_closereq
[DCCP]: Handle SYNC packets in dccp_rcv_state_process
[DCCP]: More precisely set reset_code when sending RESET packets
[DCCP]: Introduce DCCP_SOCKOPT_SERVICE
[DCCP]: Add MAINTAINERS and CREDITS entries
[DCCP]: Move the ack vector code to net/dccp/ackvec.[ch]
[CCID3]: Introduce include/linux/tfrc.h
[DCCP]: Don't use necessarily the same CCID for tx and rx
[DCCP]: Introduce CCID getsockopt for the CCIDs
Arnaud Patard:
sata_sis: Fix typo in sata port2 initialisation
Arnaud Patard (Rtp):
[WATCHDOG] s3c2410_wdt.c-state_warning.patch
Badari Pulavarty:
AUDIT: Fix definition of audit_log_start() if audit not enabled
Bart De Schuymer:
[BRIDGE-NF]: Fix iptables redirect on bridge interface
Cal Peake:
Even more fallout from ATI Xpress timer workaround
Chris Mason:
reiserfs: use mark_inode_dirty instead of reiserfs_update_sd
Chuck Ebbert:
i386: Ignore masked FPU exceptions
Clemens Ladisch:
[ALSA] ad1889: add AD1889 driver
[ALSA] ad1889: add AD1889 driver docs
[ALSA] hdsp: always initialize card name
[ALSA] usb-audio: add SNDRV_PCM_INFO_BATCH flag
[ALSA] sparse address space annotations
[ALSA] opti93x: optimize a register access
[ALSA] rtctimer: add option to make RTC timer the default sequencer timer
[ALSA] set owner field in struct pci_driver
Con Kolivas:
vm: kswapd cleanup: use pgdat
Dave Hansen:
fix mm/Kconfig spelling
David Hardeman:
[WATCHDOG] i6300esb.patch
[WATCHDOG] i6300.h-removal-patch
David L Stevens:
[IPV6]: Fix per-socket multicast filtering in sk_reuse case
David S. Miller:
[NETLINK]: Reserve a slot for NETLINK_GENERIC.
[COMPAT]: Fixup compat_do_execve()
[LIB]: Consolidate _atomic_dec_and_lock()
[NET]: Do not leak MSG_CMSG_COMPAT into userspace.
[TG3]: Add AMD K8 to list of write-reorder chipsets.
David Woodhouse:
AUDIT: Allow filtering of user messages
AUDIT: Drop user-generated messages immediately while auditing disabled.
AUDIT: Really exempt auditd from having its actions audited.
AUDIT: Report lookup flags with path/inode records.
AUDIT: Spawn kernel thread to list filter rules.
AUDIT: Optimise the audit-disabled case for discarding user messages
Add audit subsystem to MAINTAINERS, for my sins.
AUDIT: Wait for backlog to clear when generating messages.
AUDIT: Remove stray declaration of tsk from audit_receive_msg().
AUDIT: No really, we don't want to audit auditd.
AUDIT: Return correct result from audit_filter_rules()
AUDIT: Clean up user message filtering
AUDIT: Use KERN_NOTICE for printk of audit records
AUDIT: Fix definition of audit_log() if audit not enabled
AUDIT: Stop waiting for backlog after audit_panic() happens
AUDIT: Really don't audit auditd.
Fix positioning of audit in MAINTAINERS.
AUDIT: Exempt the whole auditd thread-group from auditing
AUDIT: Avoid scheduling in idle thread
AUDIT: Fix compile error in audit_filter_syscall
AUDIT: Fix livelock in audit_serial().
AUDIT: Reduce contention in audit_serial()
AUDIT: Fix task refcount leak in audit_filter_syscall()
AUDIT: Speed up audit_filter_syscall() for the non-auditable case.
Fix missing audit_syscall_exit() on ppc64 sigsuspend exit path
[AUDIT] Allow filtering on system call success _or_ failure
Fix build failure on ppc64 without CONFIG_AUDIT
Davide Libenzi:
[ALSA] hda-codec - Bring the Vaio's RA826G HDA (82801) to life ...
epoll: fix delayed initialization bug
Deepak Saxena:
[ARM] 2917/1: Make IXP4xx mach_desc's depend on config options
Denis Lukianov:
[MCAST]: Fix MCAST_EXCLUDE line dupes
Dipankar Sarma:
Fix the fdtable freeing in the case of vmalloced fdset/arrays
files: fix preemption issues
Domen Puncer:
Remove arch/arm26/boot/compressed/hw-bse.c
Dominik Brodowski:
pcmcia: warn on IOCTL usage
Florin Malita:
[BOND]: Fix bond_init() error path handling.
Frank Pavlic:
s390: ctc driver fixes
s390: TSO related fixes in qeth driver
s390: qeth driver fixes
George G. Davis:
[ARM] 2896/1: Add sys_ipc_wrapper to pass 'fifth' argument on stack
Greg KH:
add securityfs for all LSMs to use
H. Peter Anvin:
RAID6 Altivec fix
Harald Welte:
[NETFILTER]: Fix oops in conntrack event cache
[NETFILTER] Fix Kconfig dependencies for nfnetlink/ctnetlink
[NETFILTER] move nfnetlink options to right location in kconfig menu
[NETFILTER]: Solve Kconfig dependency problem
[NETFILTER]: Add new PPTP conntrack and NAT helper
[NETFILTER]: Export ip_nat_port_{nfattr_to_range,range_to_nfattr}
Heiko Carstens:
s390: bl_dev array size
s390: show_cpuinfo fix
Herbert Xu:
[TCP]: Compute in_sacked properly when we split up a TSO frame.
[TCP]: Handle SACK'd packets properly in tcp_fragment().
Hidetoshi Seto:
[IA64] mca_drv cleanup
Hugh Dickins:
error path in setup_arg_pages() misses vm_unacct_memory()
Ian E. Morgan:
[WATCHDOG] New SBC8360 watchdog driver (revised)
Ingo Molnar:
Fix spinlock owner debugging
Ivan Kokshaysky:
yenta oops fix
Alpha: ISA IRQs fixup for dp264
James Chapman:
[WATCHDOG] mv64x60_wdt.patch
James Courtier-Dutton:
[ALSA] snd-ca0106: Tidy up volume controls
[ALSA] snd-ca0106, snd-emu10k1: Add symlink in the sys tree.
[ALSA] snd-emu10k1: ALSA bug#1297: Fix a error recognising the SB Live Platinum.
Jan Beulich:
free initrd mem adjustment
minor fbcon_scroll adjustment
fbcon: constify font data
matroxfb adjustments
x86_64: NMI watchdog frequency calculation adjustments
Jaroslav Kysela:
[ALSA] version 1.0.10rc1
Jay Vosburgh:
bonding: plug reference count leak
Jean Delvare:
i2c: kill an unused i2c_adapter struct member
Jeff Dike:
uml: _switch_to code consolidation
uml: breakpoint an arbitrary thread
uml: Remove an unused file
uml: Remove a useless include
uml: Remove some build warnings
uml: preserve errno in error paths
uml: move libc code out of mem_user.c and tempfile.c
uml: merge mem_user.c and mem.c
uml: return a real error code
uml: remove include of asm/elf.h
uml: UML/i386 cmpxchg fix
Jeff Garzik:
[libata] fix PIO completion race
Jens Axboe:
fix pf request handling
Jens Osterkamp:
net: fix spider_net media detection
Jimi Xenidis:
ppc64: Fix recent regression
Jiri Slaby:
[WATCHDOG] removes pci_find_device from i6300esb.c
[WATCHDOG] i6300esb.c-2-bugs-little-cleanup.patch
[ALSA] pci_find_device remove
drivers/base/*: use kzalloc instead of kmalloc+memset
John W. Linville:
e1000: correct rx_dropped counting
e100: correct rx_dropped and add rx_missed_errors
ixgb: correct rx_dropped counting
pci: only call pci_restore_bars at boot
Jose Miguel Goncalves:
[WATCHDOG] w83977f-watchdog-driver.patch
Julian Anastasov:
[IPVS]: Really invalidate persistent templates
[IPVS]: ip_vs_ftp breaks connections using persistence
Karsten Keil:
i4l: Sedlbauer speed star II V 3.1 exist with various subversions
Add PCI IDs for Sitecom DC-105
cleanup whitespace in pci_ids.h
Fix ST 5481 USB driver
Keith Owens:
Correct xircom_cb use of CONFIG_NET_POLL_CONTROLLER
[IA64] Add Documentation/ia64/mca.txt
Komuro:
pcmcia: add another orinoco_cs id
KOVACS Krisztian:
[NETFILTER] CLUSTERIP: introduce reference counting for entries
[NETFILTER] CLUSTERIP: use a bitmap to store node responsibility data
Krzysztof Halasa:
[WAN] hdlc_cisco: Fix regression introduced by skb->tail changes.
Kumar Gala:
ppc32: remove use of asm/segment.h
Lennert Buytenhek:
[ARM] 2911/1: ixp2000_reg_{read,write} accessors
[ARM] 2909/1: remove IXP2000_PROD_ID
[ARM] 2904/1: update ixp2000 defconfigs to 2.6.13
[ARM] 2905/1: enable the ixp2000 i2c bus driver in ixp2000 defconfigs
Linus Torvalds:
Fix up more strange byte writes to the PCI_ROM_ADDRESS config word
Fix yenta error message when unable to find a bus assignment
Partially revert "Fix time going twice as fast problem on ATI Xpress chipsets"
x86-64/smp: fix random SIGSEGV issues
Make fsnotify possibly work better for the inode removal case
Linux v2.6.14-rc2
Marcel Holtmann:
[Bluetooth] Add support for extended inquiry responses
[Bluetooth] Prevent RFCOMM connections through the RAW socket
[Bluetooth] Add ignore parameters to the HCI USB driver
Marcelo Tosatti:
relayfs documentation typo
Mark J Cox:
raw_sendmsg DoS on 2.6
Martin Habets:
[ALSA] Several fixes for the Sun DBRI driver
Martin Schwidefsky:
s390: default configuration
s390: crypto driver patch take 2
matthieu castet:
airo : fix channel number in scan
Mauro Carvalho Chehab:
v4l: experimental Sliced VBI API support
v4l: fixup on cx88_dvb for Dvico HDTV5 Gold
Michael Chan:
[TG3]: Fix 4GB boundary tx handling
Michael Kerrisk:
PR_GET_DUMPABLE returns incorrect info
Michal Piotrowski:
dontdiff: add asm_offsets
Mike Miller:
cciss: new controller pci/subsystem ids
cciss: busy_initializing flag
cciss: new disk register/deregister routines
cciss: direct lookup for command completions
cciss: bug fix in cciss_remove_one
cciss: fix for DMA brokeness
cciss: One Button Disaster Recovery support
cciss: SCSI tape info for /proc
Mitsuru KANDA:
[IPV6]: Check connect(2) status for IPv6 UDP socket (Re: xfrm_lookup)
Naveen Gupta:
[WATCHDOG] i6300esb.c-WDT_ENABLE-bug
[WATCHDOG] i6300esb-set_correct_reload_register_bit
[WATCHDOG] i6300esb.c-pci_dev_put+nowayout-patch
Neil Brown:
nfsd4: printk reduction
nfsd4: move replay_owner
nfsd4: fix open seqid incrementing in lock
nfsd4: fix setclientid unlock of unlocked state lock
Code cleanups in calbacks in svcsock
Nicolas Pitre:
[ARM] 2910/1: missing Lubbock audio device declaration
OGAWA Hirofumi:
FAT: miss-sync issues on sync mount (miss-sync on write)
Pantelis Antoniou:
ppc32 8xx: flush_tlb_range() declaration uses wrong pointer type
Patrick McHardy:
[NETFILTER]: Use correct type for "ports" module parameter
[NETFILTER]: Simplify netbios helper
[NETFILTER]: Fix rcu race in ipt_REDIRECT
[NETFILTER]: Fix DHCP + MASQUERADE problem
[NETFILTER]: Rename misnamed function
Paul Mackerras:
ppc64: Make eeh_init function again
Pavel Machek:
Tell people not to use pm_register()
Pekka J Enberg:
CodingStyle: memory allocation
Peter Chubb:
[IA64] Remove warnings for gcc 4.0 IA64 compilation.
Peter Hagervall:
[TG3]: Sparse fixes for tg3
Peter Oberparleiter:
s390: kernel stack corruption
Peter Osterlund:
Remove unnecessary check_region references in comments
pktcdvd: fix bogus BUG_ON
pktcdvd: documentation update
pktcdvd: more accurate I/O accounting
pktcdvd: use kcalloc and kzalloc
pktcdvd: BUG_ON cleanups
Peter Staubach:
open returns ENFILE but creates file anyway
Randy Dunlap:
use add_taint() for setting tainted bit flags
Doc: update oops-tracing.txt (Tainted flags)
feature removal of io_remap_page_range()
Richard Purdie:
SharpSL: Abstract c7x0 specifics from Corgi SSP
SharpSL: Add cxx00 support to the Corgi LCD driver
SharpSL: Abstract c7x0 specifics from Corgi Touchscreen driver
SharpSL: Abstract model specifics from Corgi Backlight driver
SharpSL: Add new ARM PXA machines Spitz and Borzoi with partial Akita Support
SharpSL: Add an input keyboard driver for Zaurus cxx00 series
[ARM] 2915/1: SA1100 Collie: Correct scoop device calls
[ARM] 2912/1: PXA Corgi: Cleanup some unneeded code
[ARM] 2913/1: PXA Poodle: Cleanup some unneeded code
[ARM] 2914/1: PXA Poodle: Add MMC and UDC support
SharpSL: Add missing hunk from backlight update
MTD: Update SharpSL partition definitions
Fix up some pm_message_t types
Robert Love:
hdaps driver update
Robert Olsson:
[IPV4]: fib_trie tnode stats refinements
[IPV4]: fib_trie RCU refinements
Roland Dreier:
add PCI IDs so RME32 and RME96 drivers build
Russell King:
[ARM SMP] Add timer/watchdog defines for MPCore
[I2C] Add a functionality method, and remove algorithm ids
[ARM SMP] Add timer/watchdog defines for MPCore
[ARM] Remove PFN_TO_NID for !DISCONTIGMEM
[ARM] Tighten pfn_valid() test.
[ARM] Fix warning in asm/futex.h
[ARM] Fix warning in arch/arm/kernel/semaphore.c
[ARM] Fix warning in arch/arm/mach-sa1100/generic.c
Serge Hallyn:
seclvl: use securityfs (fix)
se...@us.ibm.com:
seclvl securityfs
Srivatsa Vaddagiri:
CPU hotplug breaks wake_up_new_task
Stephen Hemminger:
sk98lin: remove PCI id info for cards for conflicting devices
skge: gmac register access errors in dual port
8139cp: allocate statistics space only when needed
Takashi Iwai:
[ALSA] Fix EAPD for MSI S270
[ALSA] atiixp - Fix PM resume
[ALSA] intel8x0 - Add quirk for IBM NetVisa A30p
[ALSA] Fix ALC658D support
[ALSA] Add snd_card_set_dev()
[ALSA] hda-intel - Check validity of DMA position
[ALSA] Update/fix ALSA document
[ALSA] Introduce snd_card_set_generic_dev()
[ALSA] Add snd_card_set_generic_dev() call
[ALSA] Add snd_card_set_generic_dev() call to ISA drivers
[ALSA] hda-codec - Assign audio PCMS first
[ALSA] Fix DocBook warnings
[ALSA] Remove superfluous PCI ID definitions
[ALSA] hda-intel - Fix modem PCM creation
[ALSA] powermac - Add AUTO_DRC config
[ALSA] pcm-oss - Add bugg-yptr option
[ALSA] intel8x0 - Add buggy_semaphore option
[ALSA] hda-codec - Added ASUS A6
[ALSA] Replace with kzalloc() - core stuff
[ALSA] Replace with kzalloc() - seq stuff
[ALSA] Replace with kzalloc() - isa stuff
[ALSA] Replace with kzalloc() - pci stuff
[ALSA] Replace with kzalloc() - others
[ALSA] Another fix for DocBook
[ALSA] Add missing sound PCI IDs to pci_ids.h
Thomas Maguin:
scsi_ioctl: Add WRITE_LONG_2 as write safe command
Timothy Thelin:
ide: fix null request pointer for taskfile ioctl
Tobias Klauser:
arch/i386: Replace custom macro with isdigit()
drivers/video: Replace custom macro with isdigit()
Tom Rini:
ppc32: discard *.exit.text and *.exit.data sections
Tony Luck:
[IA64] fix circular dependency on generation of asm-offsets.h
Uwe Koziolek:
sata_sis: uninitialized variable
Victor Fusco:
[AUDIT] Fix sparse warning about gfp_mask type
Vincent Pelletier:
[ALSA] Correct detection of iBook G4 1420Mhz soundcard
Vincent Sanders:
[ARM] 2907/1: GCC 4 serial driver compile fixes
Vitaly Bordug:
ppc32: Add ppc_sys descriptions for PowerQUICC I devices
Vivek Goyal:
More documentation, minor cleanup in kdump.txt
Volker Sameske:
s390: diag 0x308 reipl
Wim Van Sebroeck:
[WATCHDOG] Kconfig+Makefile-clean2
[WATCHDOG] driver-for-ibm-automatic-server-restart-watchdog-fix2.patch
[WATCHDOG] sbc8360+w83977f_wdt-consolidate_CONFIG_WATCHDOG_NOWAYOUT_handling
[WATCHDOG] pcwd_pci-include+WDIOC_SETOPTIONS-patch
Yasuyuki Kozakai:
[NETFILTER] ip6tables: remove duplicate code
Zach Brown:
Add smp_mb__after_clear_bit() to unlock_kiocb()
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Nick Piggin
>Ahoy landlubbers!
>
>Here be t' Linux-2.6.14-rc2 release.
>
>Not a whole lot o' excitement, ye scurvy dogs, but it has t' ALSA, LSM,
>audit and watchdog merges that be missed from -rc1, and a merge series
>with Andrew. But on t' whole pretty reasonable - you can see t' details in
>the shortlog (appended).
>
>Arrr!
>
>
I hope you've christened her with an appropriately seafaring
name? If not, then I vote for white whale.
Hmm, too late: I see Ben has already named 2.6.14 :P
Send instant messages to your online friends http://au.messenger.yahoo.com
Gene Heskett
>Ahoy landlubbers!
>
>Here be t' Linux-2.6.14-rc2 release.
>
>Not a whole lot o' excitement, ye scurvy dogs, but it has t' ALSA, LSM,
>audit and watchdog merges that be missed from -rc1, and a merge series
>with Andrew. But on t' whole pretty reasonable - you can see t' details
> in the shortlog (appended).
>
>Arrr!
:-)
You've been watching entirely too much tv Linus. That commercial
is one of the better examples of the "vast wasteland" that is todays
tv.
Also, where can this wondrous new patch be found as its not made its
way to kernel.org as of 00:45 EDT?
[...]
--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.35% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
Patrick McFarland
> You've been watching entirely too much tv Linus. That commercial
> is one of the better examples of the "vast wasteland" that is todays
> tv.
D'arr, the 19th of Septembarrr tis International Talk Like A Pirate Day.
( http://talklikeapirate.com/ ).
--
Patrick the Bloody, Captain of the Red Sea Deamon, Yarr!
"Computer games don't affect kids; I mean if Pac-Man affected us as kids, we'd
all be running around in darkened rooms, munching magic pills and listening to
repetitive electronic music." -- Kristian Wilson, Nintendo, Inc, 1989
Gene Heskett
>On Tuesday 20 September 2005 12:50 am, Gene Heskett wrote:
>> You've been watching entirely too much tv Linus. That commercial
>> is one of the better examples of the "vast wasteland" that is todays
>> tv.
>
>D'arr, the 19th of Septembarrr tis International Talk Like A Pirate Day.
>( http://talklikeapirate.com/ ).
Yeah :-), but where is this new patch-2.6.14-rc2.gz to be
found. Its still not made it to kernel.org as of 10:05 EDT.
Russell King
> On Tuesday 20 September 2005 09:31, Patrick McFarland wrote:
> >On Tuesday 20 September 2005 12:50 am, Gene Heskett wrote:
> >> You've been watching entirely too much tv Linus. That commercial
> >> is one of the better examples of the "vast wasteland" that is todays
> >> tv.
> >
> >D'arr, the 19th of Septembarrr tis International Talk Like A Pirate Day.
> >( http://talklikeapirate.com/ ).
>
> Yeah :-), but where is this new patch-2.6.14-rc2.gz to be
> found. Its still not made it to kernel.org as of 10:05 EDT.
The mutinous bots have deserted master.kernel.org and aren't doing
what they're supposed to be doing.
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of: 2.6 Serial core
Gene Heskett
>On Tue, Sep 20, 2005 at 10:05:49AM -0400, Gene Heskett wrote:
>> On Tuesday 20 September 2005 09:31, Patrick McFarland wrote:
>> >On Tuesday 20 September 2005 12:50 am, Gene Heskett wrote:
>> >> You've been watching entirely too much tv Linus. That commercial
>> >> is one of the better examples of the "vast wasteland" that is
>> >> todays tv.
>> >
>> >D'arr, the 19th of Septembarrr tis International Talk Like A Pirate
>> > Day. ( http://talklikeapirate.com/ ).
>>
>> Yeah :-), but where is this new patch-2.6.14-rc2.gz to be
>> found. Its still not made it to kernel.org as of 10:05 EDT.
>
>The mutinous bots have deserted master.kernel.org and aren't doing
>what they're supposed to be doing.
Humm, what are they holding out for, more ram or more cpu?:-)
FWIW, http://master.kernel.org doesn't show it either just now.
--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.35% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
-
Russell King
> On Tue, September 20, 2005 10:25 am, Gene Heskett said:
>
> > Humm, what are they holding out for, more ram or more cpu?:-)
> >
> > FWIW, http://master.kernel.org doesn't show it either just now.
>
>
> While kernel.org snapshots will no doubt be working again shortly, you
> might want to consider using git. It reduces the amount you have to
> download for each release a lot.
This doesn't help when the bots get stuck - neither the git repository
nor the ftp space get updated when this happens. I believe both use
the same lock and are probably the same script.
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of: 2.6 Serial core
Sean
> Humm, what are they holding out for, more ram or more cpu?:-)
>
> FWIW, http://master.kernel.org doesn't show it either just now.
Gene,
While kernel.org snapshots will no doubt be working again shortly, you
might want to consider using git. It reduces the amount you have to
download for each release a lot.
It's really easy to grab a copy of git and use it to grab the kernel:
mkdir kernel
cd kernel
wget http://kernel.org/pub/software/scm/git/git-core-0.99.7.tar.bz2
tar -xvjf git-core-0.99.7.tar.bz2
cd git-core-0.99.7
make install
cd ..
git clone \
rsync://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git \
linux
cd linux
git checkout
The above is given as an attachment as well because of annoying word wrap
issues with the long url's. Anyway, after that you can stay current with
the latest Linus release with a simple "git pull".
Cheers,
Sean
Alexander Nyberg
> On Tue, September 20, 2005 10:25 am, Gene Heskett said:
>
> > Humm, what are they holding out for, more ram or more cpu?:-)
> >
> > FWIW, http://master.kernel.org doesn't show it either just now.
>
> Gene,
>
> While kernel.org snapshots will no doubt be working again shortly, you
> might want to consider using git. It reduces the amount you have to
> download for each release a lot.
>
> It's really easy to grab a copy of git and use it to grab the kernel:
>
> mkdir kernel
> cd kernel
> wget http://kernel.org/pub/software/scm/git/git-core-0.99.7.tar.bz2
> tar -xvjf git-core-0.99.7.tar.bz2
> cd git-core-0.99.7
> make install
> cd ..
>
> git clone \
> rsync://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git \
> linux
>
> cd linux
> git checkout
>
ketchup <version>
Sean
> ketchup <version>
"git pull" is actually simpler in that you don't need to specify a
version. And it will keep you current with HEAD even between official
releases.
Sean
Jan Dittmer
> On Tue, September 20, 2005 11:32 am, Alexander Nyberg said:
>
>
>>ketchup <version>
>
>
> "git pull" is actually simpler in that you don't need to specify a
> version. And it will keep you current with HEAD even between official
> releases.
$ ketchup 2.6-git
and you've the plus of very well defined checkpoints.
Jan
Sean
> Sean wrote:
>> On Tue, September 20, 2005 11:32 am, Alexander Nyberg said:
>>
>>>ketchup <version>
>>
>> "git pull" is actually simpler in that you don't need to specify a
>> version. And it will keep you current with HEAD even between official
>> releases.
>
> $ ketchup 2.6-git
>
> and you've the plus of very well defined checkpoints.
Huh? Have you ever used git? Not only do you get very well defined
checkpoints you can instantiate a tree down to any specific commit. And
you get the plus of a complete detailed changelog etc.. "git log".
Really, ketchup doesn't come close to git.
Sean
Jan Dittmer
> On Tue, September 20, 2005 12:18 pm, Jan Dittmer said:
>
>>Sean wrote:
>>
>>>On Tue, September 20, 2005 11:32 am, Alexander Nyberg said:
>>>
>>>
>>>>ketchup <version>
>>>
>>>"git pull" is actually simpler in that you don't need to specify a
>>>version. And it will keep you current with HEAD even between official
>>>releases.
>>
>>$ ketchup 2.6-git
>>
>>and you've the plus of very well defined checkpoints.
>
>
> Huh? Have you ever used git? Not only do you get very well defined
> checkpoints you can instantiate a tree down to any specific commit. And
> you get the plus of a complete detailed changelog etc.. "git log".
> Really, ketchup doesn't come close to git.
I know, but for multiple people testing daily releases it's much easier to
say -git1 worked -git2 didn't. Sure, for searching the patch `git bisect`
is priceless but for regular testing the -gitx thing comes very handy.
Otherwise you can get a arbitrary intermediate state of linus tree if
you're pulling at the wrong moment. It's actually also faster I suppose
to get one patch than running `git pull` - at least with a cold cache
(it used to be in the 0.1 days of git).
Just my .02,
Jan
Sean
> I know, but for multiple people testing daily releases it's much easier to
> say -git1 worked -git2 didn't. Sure, for searching the patch `git bisect`
> is priceless but for regular testing the -gitx thing comes very handy.
> Otherwise you can get a arbitrary intermediate state of linus tree if
> you're pulling at the wrong moment. It's actually also faster I suppose
> to get one patch than running `git pull` - at least with a cold cache
> (it used to be in the 0.1 days of git).
> Just my .02,
That's a good point. Guess it would be useful if the HEAD commit was
documented along with each -gitX release.
Sean
Jan Dittmer
> On Tue, September 20, 2005 12:44 pm, Jan Dittmer said:
>
>
>>I know, but for multiple people testing daily releases it's much easier to
>>say -git1 worked -git2 didn't. Sure, for searching the patch `git bisect`
>>is priceless but for regular testing the -gitx thing comes very handy.
>>Otherwise you can get a arbitrary intermediate state of linus tree if
>>you're pulling at the wrong moment. It's actually also faster I suppose
>>to get one patch than running `git pull` - at least with a cold cache
>>(it used to be in the 0.1 days of git).
>>Just my .02,
>
>
> That's a good point. Guess it would be useful if the HEAD commit was
> documented along with each -gitX release.
It is in /pub/linux/kernel/v2.6/snapshots/patch-2.6.x-gity.id
Jan
Alexey Dobriyan
> On Tue, September 20, 2005 12:44 pm, Jan Dittmer said:
> > say -git1 worked -git2 didn't. Sure, for searching the patch `git bisect`
> > is priceless but for regular testing the -gitx thing comes very handy.
> > Otherwise you can get a arbitrary intermediate state of linus tree if
> > you're pulling at the wrong moment. It's actually also faster I suppose
> > to get one patch than running `git pull` - at least with a cold cache
> > (it used to be in the 0.1 days of git).
> > Just my .02,
>
> documented along with each -gitX release.
Look for 41-byte *.id files in snapshots dir.
Linus Torvalds
On Tue, 20 Sep 2005, Sean wrote:
>
> That's a good point. Guess it would be useful if the HEAD commit was
> documented along with each -gitX release.
It is. Just get the "id" file that is associated with a snapshot, and it
gives the git commit ID for that state.
So for example, the 2.6.14-rc1-git3 snapshot is associated with the ID
file patch-2.6.14-rc1-git3.id, which contains
v2.6/snapshots(0)$ cat patch-2.6.14-rc1-git3.id
065d9cac98a5406ecd5a1368f8fd38f55739dee9
so once you know that something broke between rc1-git3 and rc1-git4, you
can now do
git bisect start
git bisect good 065d9cac98a5406ecd5a1368f8fd38f55739dee9
git bisect bad bc5e8fdfc622b03acf5ac974a1b8b26da6511c99
and off you go..
Linus
Sean
> On Tue, 20 Sep 2005, Sean wrote:
>>
>> That's a good point. Guess it would be useful if the HEAD commit was
>> documented along with each -gitX release.
>
> It is. Just get the "id" file that is associated with a snapshot, and it
> gives the git commit ID for that state.
>
Great! Would there be anything standing in the way of having -gitX tags
automatically added to the git repository too?
Sean
Horst von Brand
> On Tue, September 20, 2005 10:25 am, Gene Heskett said:
>
> > Humm, what are they holding out for, more ram or more cpu?:-)
> >
> > FWIW, http://master.kernel.org doesn't show it either just now.
> While kernel.org snapshots will no doubt be working again shortly, you
> might want to consider using git. It reduces the amount you have to
> download for each release a lot.
Only that it doesn't work either today. Kernel stays at 2.6.14-rc1 as of
yesterday (latest were a few NTFS patches), everything up to date.
BTW, the cogito repository is hosed, cg-update can't get needed object
69ba00668be16e44cae699098694286f703ec61d. Fetching the contents by rsync
gives the same mess.
--
Dr. Horst H. von Brand User #22616 counter.li.org
Departamento de Informatica Fono: +56 32 654431
Universidad Tecnica Federico Santa Maria +56 32 654239
Casilla 110-V, Valparaiso, Chile Fax: +56 32 797513
Sean
> Only that it doesn't work either today. Kernel stays at 2.6.14-rc1 as of
> yesterday (latest were a few NTFS patches), everything up to date.
Yeah, Russell pointed the same thing out a bit earlier. There are 13
commits MIA.
> BTW, the cogito repository is hosed, cg-update can't get needed object
> 69ba00668be16e44cae699098694286f703ec61d. Fetching the contents by rsync
> gives the same mess.
For simply tracking the kernel there isn't much reason to use cogito.
Using native git means fewer problems right now since both cogito and git
are developing quickly with inevitable version skew etc..
Sean
Russell King
> On Tue, September 20, 2005 1:59 pm, Horst von Brand said:
>
> > yesterday (latest were a few NTFS patches), everything up to date.
>
> commits MIA.
>
> > 69ba00668be16e44cae699098694286f703ec61d. Fetching the contents by rsync
> > gives the same mess.
>
> Using native git means fewer problems right now since both cogito and git
> are developing quickly with inevitable version skew etc..
Also note that the public git repository is probably suffering from
the same problem as the lack of -rc2. I wonder whether anyone's
reported it to the ftp admins yet (if it hasn't been fixed)?
--
Russell King
Linux kernel 2.6 ARM Linux - http://www.arm.linux.org.uk/
maintainer of: 2.6 Serial core
Gene Heskett
With my luck, then git will be modified the next day in a
non-backwards compatible way. Murphy's Law you know.
But I'll give it a shot anyway. Any idea how much disk it might need?
Gene Heskett
>On Tue, September 20, 2005 10:25 am, Gene Heskett said:
>> Humm, what are they holding out for, more ram or more cpu?:-)
>>
>> FWIW, http://master.kernel.org doesn't show it either just now.
>
>Gene,
>
>While kernel.org snapshots will no doubt be working again shortly, you
>might want to consider using git. It reduces the amount you have to
>download for each release a lot.
>
>It's really easy to grab a copy of git and use it to grab the kernel:
>
>mkdir kernel
>cd kernel
check
>wget http://kernel.org/pub/software/scm/git/git-core-0.99.7.tar.bz2
check
>tar -xvjf git-core-0.99.7.tar.bz2
check
>cd git-core-0.99.7
check
>make install
check
Well, things marched right along till I got to the line for git clone
etc.
[root@coyote kernel]# git clone
rsync://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
linux
bash: git: command not found
Humm, back up & find it put git stuff in /root/bin ?????
[root@coyote kernel]# /root/bin/git clone \
rsync://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
linux
/root/bin/git-clone: line 2: git-init-db: command not found
* git clone [-l [-s]] [-q] [-u <upload-pack>] <repo> <dir>
So why did it put it in /root/bin? It's there, but not in my $PATH.
And theres no "make uninstall" to clean up the install so I might be
able to cleanly fix the miss-fire... Murphy rides again..
[...]
Matthew Dharm
Matt
--
Matthew Dharm Home: mdhar...@one-eyed-alien.net
Maintainer, Linux USB Mass Storage Driver
I'm just trying to think of a way to say "up yours" without getting fired.
-- Stef
User Friendly, 10/8/1998
Gene Heskett
>~/bin is in many people's paths in many distributions.
And that seems to have fixed it, its now pulling in a bunch of what
looks like uuenocded stuffs.Thanks Matt
>> >t \ linux
Gene Heskett
Ok, once all thats done, then anytime I want a snapshot, go there and
do a git checkout, then snapshot copy that dir to
/usr/src/linux-$VERSION, copy the old .config in to this copy, add my
build scripts and go? Anyway, I've just done the above, so we'll
test it out right now.
Sean
> It is. Just get the "id" file that is associated with a snapshot, and
> it gives the git commit ID for that state.
>
> So for example, the 2.6.14-rc1-git3 snapshot is associated with the ID
> file patch-2.6.14-rc1-git3.id, which contains
>
> v2.6/snapshots(0)$ cat patch-2.6.14-rc1-git3.id
> 065d9cac98a5406ecd5a1368f8fd38f55739dee9
>
> so once you know that something broke between rc1-git3 and rc1-git4,
> you can now do
>
> git bisect start
> git bisect good 065d9cac98a5406ecd5a1368f8fd38f55739dee9
> git bisect bad bc5e8fdfc622b03acf5ac974a1b8b26da6511c99
>
> and off you go..
The attached patch grabs all the .id files from the snapshot directory on
kernel.org and converts them into tags in a local git repository. So
after running "gtags", your example becomes:
git bisect start
git bisect good v2.6.14-rc1-git3
git bisect bad v2.6.14-rc1-git4
Sean
#----[gtags]-----
#!/bin/sh
cd .git/refs/tags/ || exit
lftp http://www.kernel.org <<\EOF
cd /pub/linux/kernel/v2.6/snapshots/
mget patch*.id
cd /pub/scm/linux/kernel/git/torvalds/linux-2.6.git/refs/tags/
mget *
EOF
rename patch- v patch-*.id
rename .id "" v*.id
Martin J. Bligh
>
> Here be t' Linux-2.6.14-rc2 release.
>
> Not a whole lot o' excitement, ye scurvy dogs, but it has t' ALSA, LSM,
> audit and watchdog merges that be missed from -rc1, and a merge series
> with Andrew. But on t' whole pretty reasonable - you can see t' details in
> the shortlog (appended).
>
> Arrr!
SCSI is broken in several cases by a lack of the patch below, which means
some of the regular test boxes can't - James, any chance of getting that
one upstream? (it's cut and pasted, but then you wouldn't want to apply
it anyway without correct "flow" ;-)).
Thanks,
M.
On Wed, 2005-09-14 at 18:06 +1000, Anton Blanchard wrote:
> And in particular it looks like the scsi_unprep_request in
> scsi_queue_insert is causing it. The following patch fixes the boot
> problems on the vscsi machine:
OK, my fault. Your fix is almost correct .. I was going to do this
eventually, honest, because there's no need to unprep and reprep a
command that comes in through scsi_queue_insert().
However, I decided to leave it in to exercise the scsi_unprep_request()
path just to make sure it was working. What's happening, I think, is
that we also use this path for retries. Since we kill and reget the
command each time, the retries decrement is never seen, so we're
retrying forever.
This should be the correct reversal.
James
diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -140,14 +140,12 @@ static void scsi_unprep_request(struct r
* commands.
* Notes: This could be called either from an interrupt context or a
* normal process context.
- * Notes: Upon return, cmd is a stale pointer.
*/
int scsi_queue_insert(struct scsi_cmnd *cmd, int reason)
{
struct Scsi_Host *host = cmd->device->host;
struct scsi_device *device = cmd->device;
struct request_queue *q = device->request_queue;
- struct request *req = cmd->request;
unsigned long flags;
SCSI_LOG_MLQUEUE(1,
@@ -188,9 +186,8 @@ int scsi_queue_insert(struct scsi_cmnd *
* function. The SCSI request function detects the blocked condition
* and plugs the queue appropriately.
*/
- scsi_unprep_request(req);
spin_lock_irqsave(q->queue_lock, flags);
- blk_requeue_request(q, req);
+ blk_requeue_request(q, cmd->request);
spin_unlock_irqrestore(q->queue_lock, flags);
scsi_run_queue(q);
James Bottomley
> SCSI is broken in several cases by a lack of the patch below, which means
> some of the regular test boxes can't - James, any chance of getting that
> one upstream? (it's cut and pasted, but then you wouldn't want to apply
> it anyway without correct "flow" ;-)).
It already is ... unfortunately just after 2.6.14-rc2, but if you use
the latest git head you should get it.
James
Nish Aravamudan
> > Ahoy landlubbers!
> >
> > Here be t' Linux-2.6.14-rc2 release.
> >
> > Not a whole lot o' excitement, ye scurvy dogs, but it has t' ALSA, LSM,
> > audit and watchdog merges that be missed from -rc1, and a merge series
> > with Andrew. But on t' whole pretty reasonable - you can see t' details in
> > the shortlog (appended).
> >
> > Arrr!
>
> SCSI is broken in several cases by a lack of the patch below, which means
> some of the regular test boxes can't - James, any chance of getting that
> one upstream? (it's cut and pasted, but then you wouldn't want to apply
> it anyway without correct "flow" ;-)).
FYI, test.kernel.org was able to verify this because 2.6.14-rc2 is
available via ftp, even though the website and git (maybe I'm not
using it correctly) don't seem to be updated yet.
Thanks,
Nish
Martin J. Bligh
--On Tuesday, September 20, 2005 18:12:02 -0500 James Bottomley <James.B...@HansenPartnership.com> wrote:
> On Tue, 2005-09-20 at 16:03 -0700, Martin J. Bligh wrote:
>> SCSI is broken in several cases by a lack of the patch below, which means
>> some of the regular test boxes can't - James, any chance of getting that
>> one upstream? (it's cut and pasted, but then you wouldn't want to apply
>> it anyway without correct "flow" ;-)).
>
> It already is ... unfortunately just after 2.6.14-rc2, but if you use
> the latest git head you should get it.
Spiffy - it should catch that tonight then ...
Thanks,
M.
Gene Heskett
>Sean <sean...@sympatico.ca> wrote:
>> On Tue, September 20, 2005 10:25 am, Gene Heskett said:
>> > Humm, what are they holding out for, more ram or more cpu?:-)
>> >
>> > FWIW, http://master.kernel.org doesn't show it either just now.
>>
>> While kernel.org snapshots will no doubt be working again shortly, you
>> might want to consider using git. It reduces the amount you have to
>> download for each release a lot.
>
>Only that it doesn't work either today. Kernel stays at 2.6.14-rc1 as of
>yesterday (latest were a few NTFS patches), everything up to date.
Thats odd. I followed the directions in the message near the head of
this thread, and after I fixed my $PATH to include ~/bin:, it
just worked, and I made a copy of it to my /usr/src dir & built
it. I've been running 2.6.14-rc2 for about 00:02:17 now. And it seems
to be running nominally so far. Or at least thats the version in the
Makefile. :-)
>BTW, the cogito repository is hosed, cg-update can't get needed object
>69ba00668be16e44cae699098694286f703ec61d. Fetching the contents by rsync
>gives the same mess.
--
Cheers, Gene
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
99.35% setiathome rank, not too shabby for a WV hillbilly
Yahoo.com and AOL/TW attorneys please note, additions to the above
message by Gene Heskett are:
Copyright 2005 by Maurice Eugene Heskett, all rights reserved.
-
Gene Heskett
What is the lifetime of this script? The reason I am asking is that
somehow, the 'linux' directory got emptied when I used mc to make a
copy of it to /usr/src/linux-2.6.14-rc2, which then built just fine
and I'm running it.
But editing the script so that it only does the stuff from the rsync
line on, I have not been able to duplicate its flawless performance,
instead getting this near the end of the proceedure:
------
[...]
tags/v2.6.14-rc2
sent 563 bytes received 2778 bytes 954.57 bytes/sec
total size is 779 speedup is 0.23
rsync: link_stat
"/scm/linux/kernel/git/torvalds/linux-2.6.git/objects/info/alternates"
(in pub) failed: No such file or directory (2)
rsync error: some files could not be transferred (code 23) at
main.c(812)
------
I have now re-ran this 4 times, each time doing an rm -fR on the
'linux' directory, with the same results as above each time.
Is this normal, or am I missing something important (and probably
a 'duh' moment too) that it takes to do a re-run?
Bill Davidsen
> On Tue, Sep 20, 2005 at 10:05:49AM -0400, Gene Heskett wrote:
>
>>On Tuesday 20 September 2005 09:31, Patrick McFarland wrote:
>>
>>>On Tuesday 20 September 2005 12:50 am, Gene Heskett wrote:
>>>
>>>>You've been watching entirely too much tv Linus. That commercial
>>>>is one of the better examples of the "vast wasteland" that is todays
>>>>tv.
>>>
>>>D'arr, the 19th of Septembarrr tis International Talk Like A Pirate Day.
>>>( http://talklikeapirate.com/ ).
>>
>>Yeah :-), but where is this new patch-2.6.14-rc2.gz to be
>>found. Its still not made it to kernel.org as of 10:05 EDT.
>
>
> The mutinous bots have deserted master.kernel.org and aren't doing
> what they're supposed to be doing.
>
I guess having a software error in the kernel would constitute bug-ery
on the high C.
--
-bill davidsen (davi...@tmr.com)
"The secret to procrastination is to put things off until the
last possible moment - but no longer" -me
Bill Davidsen
> On Tue, September 20, 2005 10:25 am, Gene Heskett said:
>
>
>>Humm, what are they holding out for, more ram or more cpu?:-)
>>
>>FWIW, http://master.kernel.org doesn't show it either just now.
>
>
> Gene,
>
> While kernel.org snapshots will no doubt be working again shortly, you
> might want to consider using git. It reduces the amount you have to
> download for each release a lot.
>
> It's really easy to grab a copy of git and use it to grab the kernel:
>
> mkdir kernel
> cd kernel
> wget http://kernel.org/pub/software/scm/git/git-core-0.99.7.tar.bz2
> tar -xvjf git-core-0.99.7.tar.bz2
> cd git-core-0.99.7
> make install
> cd ..
>
> git clone \
> rsync://www.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git \
> linux
>
> cd linux
> git checkout
>
>
> The above is given as an attachment as well because of annoying word wrap
> issues with the long url's. Anyway, after that you can stay current with
> the latest Linus release with a simple "git pull".
But that pulls the current tree, doesn't it? Not the git release?
For purposes of bug reporting and fixing, I would think that having some
reproducible version is superior. If I say git3, anyone can get it, it
would appear that "git pull" isn't as deterministic.
Sean
> But that pulls the current tree, doesn't it? Not the git release?
It pulls the latest git release plus any extra commits that have been made
since, if any. Now, if you have a specific reason to test a -gitX
release you can simply lookup which commit corresponds to the -gitX
release and check it out of the git repository and build it. Actually, I
just posted a script that lets you check out a build by -gitX name instead
of having to manually lookup the corresponding commit id.
> For purposes of bug reporting and fixing, I would think that having some
> reproducible version is superior. If I say git3, anyone can get it, it
> would appear that "git pull" isn't as deterministic.
There are cases where using the -gitX release helps a bit. But there's
nothing magic about a -gitX release, it's just a collection of whatever
commits existed when the timed build procedure ran.
Using git you get the choice of using -gitX builds or tracking directly
from the most recent commit as the situation demands. Therefore if Linus
commits a fix beyond the latest -gitX release you happen to need, you're
good to go. If you want to track the latest -gitX release, you can do
that with git too.
Sean
Martin J. Bligh
--On Tuesday, September 20, 2005 16:13:46 -0700 "Martin J. Bligh" <mbl...@mbligh.org> wrote:
>
>
> --On Tuesday, September 20, 2005 18:12:02 -0500 James Bottomley <James.B...@HansenPartnership.com> wrote:
>
>> On Tue, 2005-09-20 at 16:03 -0700, Martin J. Bligh wrote:
>>> SCSI is broken in several cases by a lack of the patch below, which means
>>> some of the regular test boxes can't - James, any chance of getting that
>>> one upstream? (it's cut and pasted, but then you wouldn't want to apply
>>> it anyway without correct "flow" ;-)).
>>
>> It already is ... unfortunately just after 2.6.14-rc2, but if you use
>> the latest git head you should get it.
>
> Spiffy - it should catch that tonight then ...
Works. Thanks,