The driver(s) below do not use LINUX_VERSION_CODE nor KERNEL_VERSION. drivers/video/arkfb.c drivers/video/s3fb.c drivers/video/vt8623fb.c This patch removes the said #include <version.h>. Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com> diff --git a/drivers/video/arkfb.c b/drivers/video/arkfb.c index 4bd569e..314d186 100644 --- a/drivers/video/arkfb.c +++ b/drivers/video/arkfb.c @@ -11,7 +11,6 @@ * Code is based on s3fb */ -#include <linux/version.h> #include ...

The drivers below do not use LINUX_VERSION_CODE nor KERNEL_VERSION. drivers/misc/eeprom_93cx6.c This patch removes the said #include <version.h>. Signed-off-by: Huang Weiyi <weiyi.huang@gmail.com> diff --git a/drivers/misc/eeprom_93cx6.c b/drivers/misc/eeprom_93cx6.c index ea55654..15b1780 100644 --- a/drivers/misc/eeprom_93cx6.c +++ b/drivers/misc/eeprom_93cx6.c @@ -26,7 +26,6 @@ #include <linux/kernel.h> #include <linux/module.h> -#include <linux/version.h> #include ...

To the extent that it would result in this failing gracefully before some other part of the kernel becomes hideously upset, sure. I have a hard time caring, but I can easily fix it if other people do :) -- Matthew Garrett | mjg59@srcf.ucam.org --

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11346 Subject : kernel BUG at arch/x86/mm/pat.c:233! Submitter : Jean Delvare <khali@linux-fr.org> Date : 2008-08-15 02:10 (2 days old) Handled-By : Andi Kleen ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11263 Subject : Re: 2.6.27-rc2: uvcvideo WARNING after suspend to ram Submitter : Alan Jenkins <alan-jenkins@tuffmail.co.uk> Date : 2008-08-07 04:02 (10 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11279 Subject : 2.6.27-rc0 Power Bugs with HP/Compaq Laptops Submitter : Matt Parnell <mparnell@gmail.com> Date : 2008-08-07 14:57 (10 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11344 Subject : lockdep link failed Submitter : Ming Lei <tom.leiming@gmail.com> Date : 2008-08-14 9:58 (3 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11264 Subject : Invalid op opcode in kernel/workqueue Submitter : Jean-Luc Coulon <jean.luc.coulon@gmail.com> Date : 2008-08-07 04:18 (10 days old) --

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11355 Subject : Regression in 2.6.27-rc2 when cross-building the kernel Submitter : Larry Finger <Larry.Finger@lwfinger.net> Date : 2008-08-16 2:38 (1 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11210 Subject : libata badness Submitter : Kumar Gala <galak@kernel.crashing.org> Date : 2008-07-31 18:53 (17 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11282 Subject : Please fix x86 defconfig regression Submitter : Andi Kleen <andi@firstfloor.org> Date : 2008-08-07 20:46 (10 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11323 Subject : /proc/diskstats does not contain all disk devices Submitter : Andy Ryan <genanr@emsphone.com> Date : 2008-08-13 12:12 (4 days old) Handled-By : Greg Kroah-Hartman ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11219 Subject : KVM modules break emergency reboot Submitter : Zdenek Kabelac <zdenek.kabelac@gmail.com> Date : 2008-08-01 20:25 (16 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11215 Subject : INFO: possible recursive locking detected ps2_command Submitter : Zdenek Kabelac <zdenek.kabelac@gmail.com> Date : 2008-07-31 9:41 (17 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11209 Subject : 2.6.27-rc1 process time accounting Submitter : Lukas Hejtmanek <xhejtman@ics.muni.cz> Date : 2008-07-31 10:43 (17 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11205 Subject : x86: 2.6.27-rc1 does not build with gcc-3.2.3 any more Submitter : Mikael Pettersson <mikpe@it.uu.se> Date : 2008-07-30 11:02 (18 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11189 Subject : sky2 WOL broken Submitter : Rafael J. Wysocki <rjw@sisk.pl> Date : 2008-07-20 0:20:10 (28 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11341 Subject : 2.6.27-rc1 - ext4 e2fsck false prompting for fixing i_size of Inode Submitter : Kamalesh Babulal <kamalesh@linux.vnet.ibm.com> Date : 2008-08-13 6:56 (4 days ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11337 Subject : Warning in during hotplug on 2.6.27-rc2-git5 Submitter : Mark Langsdorf <mark.langsdorf@amd.com> Date : 2008-08-12 21:56 (5 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11340 Subject : LTP overnight run resulted in unusable box Submitter : Alexey Dobriyan <adobriyan@gmail.com> Date : 2008-08-13 9:24 (4 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11356 Subject : Linux 2.6.27-rc3 - build failure: undefined reference to `.lockdep_count_forward_deps' Submitter : Frans Pop <elendil@planet.nl> Date : 2008-08-16 19:11 (1 days ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11276 Subject : build error: CONFIG_OPTIMIZE_INLINING=y causes gcc 4.2 to do stupid things Submitter : Randy Dunlap <randy.dunlap@oracle.com> Date : 2008-08-06 17:18 (11 days ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11308 Subject : tbench regression on each kernel release from 2.6.22 -> 2.6.28 Submitter : Christoph Lameter <cl@linux-foundation.org> Date : 2008-08-11 18:36 (6 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11220 Subject : Heavy suspend and io problems in 2.6.27-rc1-00156-g94ad374 Submitter : Nico Schottelius <nico@schottelius.org> Date : 2008-07-31 21:05 (17 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11230 Subject : Kconfig no longer outputs a .config with freshly updated defconfigs Submitter : Josh Boyer <jwboyer@linux.vnet.ibm.com> Date : 2008-08-02 16:03 (15 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11316 Subject : severe performance regression for iptables nat routing Submitter : Alex Williamson <alex.williamson@hp.com> Date : 2008-08-12 22:04 (5 days old) Handled-By : Herbert Xu ...

The problem is not evident in 2.6.27-rc3 Grant. --

Hi Rafael, Andi's patch still needs to be pushed to Linus. -- Jean Delvare --

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11333 Subject : Rewrite SSB DMA API breaks compilation on ARM Submitter : Martin Michlmayr <tbm@cyrius.com> Date : 2008-08-10 12:16 (7 days old) References : ...

This message has been generated automatically as a part of a report of recent regressions. The following bug entry is on the current list of known regressions from 2.6.26. Please verify if it still should be listed and let me know (either way). Bug-Entry : http://bugzilla.kernel.org/show_bug.cgi?id=11278 Subject : 2.6.27-rc2: Very odd top: '5124095h kthreadd' display Submitter : Grant Coady <grant_lkml@dodo.com.au> Date : 2008-08-07 7:03 (10 days old) References : ...

On Sat, 16 Aug 2008 21:30:27 +0100 how does this interact with the backlight control done via ACPI ? -- If you want to reach me at my work email, use arjan@linux.intel.com For development, discussion and tips for power savings, visit http://www.lesswatts.org --

This driver adds backlight and rfkill support for Dell laptops. It uses the DCDBAS driver to trigger the system management calls required for this, and parses the DMI tables itself in order to find the appropriate tokens. In future it should be possible to add LED control, but I don't have any appropriate machines right now for testing. -- Matthew Garrett | mjg59@srcf.ucam.org --

Dell's ACPI implementation appears to be implemented on top of this at a firmware level, so both are consistent. However, in terms of cleanliness, we should wait for Thomas's patches to be merged so this driver can disable the backlight support if there's an ACPI backlight present. The reason I implemented it is that there's no ACPI backlight support on some currently shipping Dells, never mind the older hardware. -- Matthew Garrett | mjg59@srcf.ucam.org --

It's well may be that I'm just missing something obvious. It seems to me that find_first_bit/find_next_bit functions return their offsets "base 1" - first set bit is "1" and last is "bitmap size". This means that if only the last bit in the bitmap is set, the returned value will be indistinguishable from no bits set situation. Moreover, bit manipulation functions appear to use "base 0" bit addresses, adding to the inconvenience. Is this a desired behavior? And, if yes, how is one supposed to ...

Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com> --- arch/x86/kernel/apic_64.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/arch/x86/kernel/apic_64.c b/arch/x86/kernel/apic_64.c index 18c0b8c..5e07381 100644 --- a/arch/x86/kernel/apic_64.c +++ b/arch/x86/kernel/apic_64.c @@ -174,7 +174,7 @@ u64 xapic_icr_read(void) icr2 = apic_read(APIC_ICR2); icr1 = apic_read(APIC_ICR); - return (icr1 | ((u64)icr2 << 32)); + return icr1 | ((u64)icr2 << 32); } ...

That's a quantitative rather than a qualitative limitation though. I can see no reason why limiting the number of CPUs would preclude the use of the x2APIC mode altogether -- the extra CPUs will simply be ignored and never waken up. And with smaller systems and a 32-bit kernel one may still want to get the small performance gain from using the MSRs rather than going through the TLB for example. Similarly there is some RAM size limitation with the 32-bit configuration too, but that does ...

Please review - any comments are welcome! For now it's like code bloating - but it's just preliminary series to make apic_*.c code more or less similar. And it's still a bit far from being ready to be merged down. - Cyrill - --

[Maciej W. Rozycki - Sat, Aug 16, 2008 at 08:52:07PM +0100] | On Sat, 16 Aug 2008, Cyrill Gorcunov wrote: | | > +#ifdef CONFIG_X86_64 | > + if (x2apic) | > + enable_x2apic(); | > + else | > +#endif | | Hmm, x86-64 hardware can run a 32-bit kernel, so it might be a good idea | to take the opportunity of the merge and extend x2APIC support to the | 32-bit configuration too. It should be mostly a mechanical change. Just | a suggestion though -- feel free to ignore if you'd rather not dive ...

Hmm, x86-64 hardware can run a 32-bit kernel, so it might be a good idea to take the opportunity of the merge and extend x2APIC support to the 32-bit configuration too. It should be mostly a mechanical change. Just a suggestion though -- feel free to ignore if you'd rather not dive into it. ;) Maciej --

Rearrange functions and comments to find differences easier. Also use apic_printk in setup_boot_APIC_clock for 64bit mode. Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com> --- arch/x86/kernel/apic_32.c | 69 +++++++++++++++++++++++++------------------- arch/x86/kernel/apic_64.c | 48 ++++++++++++++++++++++++------ 2 files changed, 77 insertions(+), 40 deletions(-) diff --git a/arch/x86/kernel/apic_32.c b/arch/x86/kernel/apic_32.c index 6cb8aaa..9e8702e 100644 --- ...

On Sat, 16 Aug 2008 09:34:26 -0700 monitor/mwait is rather really expensive.. are we really sure we want to use this? (from an Intel cpu perspective the answer is very likely no; but I don't know what AMD does here) -- If you want to reach me at my work email, use arjan@linux.intel.com For development, discussion and tips for power savings, visit http://www.lesswatts.org --

The intended use is when you're going to be waiting for a while (on the order of microseconds or more). In the Xen case, I use this to block the vcpu if we pass a few iterations without the condition being true. While the mwait patch doesn't do this at present, it could. J --

Add an x86-specific implementation. This patch doesn't do much other than wire it up to paravirt_ops, and lay the groundwork for the upcoming monitor/mwait implementation. Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Rusty Russell <rusty@rustcorp.com.au> --- arch/x86/Kconfig | 3 + arch/x86/kernel/Makefile ...

There are various places in the kernel which wish to wait for a condition to come true while in a non-blocking context. Existing examples of this are stop_machine() and smp_call_function_mask(). (No doubt there are other instances of this pattern in the tree.) Thus far, the only way to achieve this is by spinning with a cpu_relax() loop. This is fine if the condition becomes true very quickly, but it is not ideal: - There's little opportunity to put the CPUs into a low-power state. ...

Linus, Please pull the latest core-fixes-for-linus git tree from: git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git core-fixes-for-linus Thanks, Ingo ------------------> Andrew Morton (1): lockdep: spin_lock_nest_lock(), checkpatch fixes Arjan van de Ven (1): lockdep: use WARN() in kernel/lockdep.c Ingo Molnar (1): lockdep: build fix Stephen Hemminger (1): lockdep: fix build if CONFIG_PROVE_LOCKING not defined ...

Regression since commit 73cf60232ef16e1f8a64defa97214a1722db1e6c, "ieee1394: use class iteration api": The two loops for (1.) driver updates and (2.) driver probes were replaced by a single loop with bogus needs_probe checks. Hence updates and probes were now intermixed, and especially sbp2 updates (reconnects) held up longer than necessary. While we fix it, change the needs_probe flag to bool type for clarity. Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de> --- I will make ...

sbp2 was too quick to run away screaming "Failed to reconnect to sbp2 device!" when additional nodes on the same bus came online. Signed-off-by: Stefan Richter <stefanr@s5r6.in-berlin.de> --- drivers/ieee1394/sbp2.c | 25 ++++++++++++++++++------- 1 file changed, 18 insertions(+), 7 deletions(-) Index: linux/drivers/ieee1394/sbp2.c =================================================================== --- linux.orig/drivers/ieee1394/sbp2.c +++ linux/drivers/ieee1394/sbp2.c @@ -733,15 ...

Yes, that was indeed the intention. I didn't know until quite recently about the Kbuild file. And I was just about to make a patch adding these headers when your email arrived :-) So here is my SoB for the ivtv headers: Signed-off-by: Hans Verkuil <hverkuil@xs4all.nl> Thanks! --

After disabling kernel support for "Group CPU scheduler" and applying 'echo -1 > /proc/sys/kernel/sched_rt_runtime_us' the behaviour is as expected. chrt -f 99 ./a.out average:13 usec min. jitter:0 usec max. jitter:29 usec chrt -o 0 ./a.out average:153 usec min. jitter:0 usec max. jitter:37035 usec So the problem is located first in the new sched_rt_runtime_us default value and second in the "Group CPU scheduler". A last question: I though that the kernel will never break ...

I haved tried your suggestion on my 2.6.26 pentium notebook. Nothing is changing. After applying 'echo -1> /proc/sys/kernel/sched_rt_runtime_us' the SCHED_FIFO jitter are still higher than with SCHED_OTHER. Here are the results of my notebook time chrt -f 99 /tmp/a.out time chrt -o 0 /tmp/a.out average: 212 average: 13 min. jitter: 0 usec min. jitter: 0 usec max. jitter: 50013 usec max. jitter: 33 usec The kernel was startet with ...

Agreed. I think Alexey's patch is broken. The thing is, BUILD_BUG_ON() is a different thing. It says "this is a build error", while BUG_ON() says "this is an error if we reach it". Very different. The fact that you broke BUG_ON(1) should have made you think. Sometimes the "1" isn't necessarily a constant one. It might be if (something_that_can_never_happen_in_some_configuration) { ... BUG_ON(CONFIG_XYZZY); ... } where the BUG_ON(1) is absolutely *not* the same thing ...

BUILD_BUG_ON should have never existed -- BUG_ON could upgrade itself to compile-breaking version if compiler has enough information and this is what patch does. The only downside is that one can't write BUG_ON(1) anymore. Note: this compile-breaks jbd, jbd2, tipc, what this code is supposed to do? journal = handle->h_transaction->t_journal; if (!journal_set_features(journal, 0, 0, JFS_FEATURE_INCOMPAT_REVOKE)) { J_ASSERT (!"Cannot set revoke feature!"); ^^^^ return ...

merge io_apic_xx.c Thanks Yinghai Lu --

use MACRO for 32 bit too Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com> --- arch/x86/kernel/io_apic.c | 80 ++++++++++++---------------------------------- 1 file changed, 21 insertions(+), 59 deletions(-) Index: linux-2.6/arch/x86/kernel/io_apic.c =================================================================== --- linux-2.6.orig/arch/x86/kernel/io_apic.c +++ linux-2.6/arch/x86/kernel/io_apic.c @@ -608,18 +608,7 @@ static void __init replace_pin_at_irq(un add_pin_to_irq(irq, ...

Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com> --- drivers/pci/intr_remapping.c | 2 +- include/asm-x86/hw_irq.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) Index: linux-2.6/include/asm-x86/hw_irq.h =================================================================== --- linux-2.6.orig/include/asm-x86/hw_irq.h +++ linux-2.6/include/asm-x86/hw_irq.h @@ -115,7 +115,7 @@ extern asmlinkage void smp_invalidate_in #endif #ifdef CONFIG_X86_32 -extern void (*const ...

Hi, I would also prefer to maintain symmetry here. Your argument is moot, why diverge a small part of one API just because it is not used much? Everyone using the spin_lock functions learns the weird interface pretty fast. If you are in a rare situation where you have to use the trylock versions, you would really expect them to be used equivalently. I agree that this argument of Ingo's is not a very good one... ;) Hannes --

Sparc _used_ to save/restore the whole processor flags word with the irq flags. That includes, iirc, things like the register window crap, so if you did a save/restore flags in a function, it would get all that wrong and things would blow up. However, I don't think sparc has actually done that for a _loong_ time However, I refuse to see this crap. Like it or not, the 'flags' argument has always been pass-by-reference, and not a pointer. It does that because it used to make a huge ...

1) de-macro, remove ({ usages as side-effect, 2) change calling convention to not accept "flags" by value -- trylock functions can modify them, so by-value is misleading, and number of users is relatively low. 3) de-macro spin_trylock_irq() for a change. Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> --- drivers/infiniband/ulp/ipoib/ipoib_main.c | 2 - drivers/net/atl1e/atl1e_main.c | 2 - drivers/net/atlx/atl1.c | 2 - ...

Only slightly, reader is hinted that flags can be changed, otherwise We can't really, and I don't propose that: ~8700 usages of spin_lock_irqsave, ~1300 usages of local_irq_save. However for code which has small number of users, why not? The prehistory of this patch is that I'm deeply in spinlock and irqflags.h headers for clean irq_flags_t conversion and overall implession is that they're horrible. Just the joke with local_irq_enable() defined via raw_local_irq_enable() and several ...

I don't think we should have anything in the inode. We don't want to What controls somewhat writing such a tag on media remotely ? Locally you can do this (although you are way too specialized in design - an LSM hook for controlling tag setting or a general tag reservation sysfs interface User space problem. Set flags 'dirty', then set bit 'scanning' clear 'dirty' then clear 'scanning' when finished. If the dirty flag got set while you were scanning it will still be set now you've ...

Not on this occasion - fixed patch at the bottom. I suspect you of having a wicked sense of humour. (Once upon a time some types got separated out from mm.h to mm_types.h: probably mm.h's VM_flags should have travelled with vm_area_struct, even though they're not types. And on another occasion, a forward [PATCH] mm: VM_flags comment fixes Try to comment away a little of the confusion between mm's vm_area_struct vm_flags and vmalloc's vm_struct flags: based on an idea by Ulrich ...

Commit 163f6876f5c3ff8215e900b93779e960a56b3694 missed one, resulting in the following compile error: AS arch/powerpc/kernel/misc_32.o arch/powerpc/kernel/misc_32.S: Assembler messages: arch/powerpc/kernel/misc_32.S:902: Error: unsupported relocation against KEXEC_CONTROL_CODE_SIZE make[2]: *** [arch/powerpc/kernel/misc_32.o] Error 1 make[1]: *** [arch/powerpc/kernel] Error 2 make: *** [vmlinux] Error 2 I grepped arch/ and found no further instances. Signed-off-by: Paul Collins ...

Have you looked at my sleping rwlock trial thing? It's very different from a spinning one, but I think the fast path should be identical, and that's the one I tried to make fairly optimal. See http://git.kernel.org/?p=linux/kernel/git/torvalds/rwlock.git;a=summary for a git tree. The sleeping version has two extra words for the sleep events, but those would be irrelevant for the spinning version. The fastpath is movl $4,%eax lock ; xaddl %eax,(%rdi) testl $3,%eax jne ...

Right. Which is exactly why I'd suggest using the extra space for saying "this CPU is busy-looping waiting for a write lock", and then the read-lock contention case can say - ok, there's a pending write lock holder on _my_ CPU, so I need to just succeed right away despite the fact that there is contention. In other words, there are a few cases: - you actually *got* the write lock interrupts obviously have to be disabled here, because any reader can simply not get the ...

This is correct; this is in fact true for all instructions except "mov". Whether it's sign- or zero-extending is sometimes subtle, but not in these cases. Do you happen to know if this is a manifest bug in the current kernel (i.e. if there is anywhere we're using more than ±2 GB as a constant to these functions?) Either way, I'll queue this up to tip:x86/urgent if Ingo hasn't already since this is a pure bug fix. -hpa --

From: Grant Coady <gcoady.lk@gmail.com> ATA Kconfig cleanup 3/3: A few text changes and break a couple long lines. Signed-off-by: Grant Coady <gcoady.lk@gmail.com> --- Kconfig | 15 +++++++++------ 1 files changed, 9 insertions(+), 6 deletions(-) --- linux-2.6.27-rc3b2/drivers/ata/Kconfig 2008-08-16 14:20:08.177922891 +1000 +++ linux-2.6.27-rc3b3/drivers/ata/Kconfig 2008-08-16 14:50:16.454342447 +1000 @@ -3,7 +3,7 @@ # menuconfig ATA - tristate "Serial ATA (prod) and ...

Grant Coady writes: > > From: Grant Coady <gcoady.lk@gmail.com> > > ATA Kconfig cleanup 1/3: Hide ATA_SFF, add PATA and SATA driver menu > enable items, and add dependency on the enabled PATA/SATA lines to > all the ATA drivers except the PATA_PLATFORM items. ... > @@ -147,7 +166,7 @@ > > config SATA_PROMISE > tristate "Promise SATA TX2/TX4 support" > - depends on PCI > + depends on ATA_SFF_SATA && PCI > help > This option enables support for Promise Serial ...

From: Grant Coady <gcoady.lk@gmail.com> ATA Kconfig cleanup 1/3: Hide ATA_SFF, add PATA and SATA driver menu enable items, and add dependency on the enabled PATA/SATA lines to all the ATA drivers except the PATA_PLATFORM items. Signed-off-by: Grant Coady <gcoady.lk@gmail.com> --- Kconfig | 169 +++++++++++++++++++++++++++++++++++----------------------------- 1 files changed, 94 insertions(+), 75 deletions(-) --- linux-2.6.27-rc3/drivers/ata/Kconfig 2008-07-14 07:51:29.000000000 ...

08-15-2008 Changelog: Patch 1/1 remove module base conditionals. remove any reference to any file owned or maintained by Ingo Molnar based on his statement he has no interest in being involved in decisions involving debuggers. Fixed UP problems with generic genapic settings. I will be maintaining this debugger and posting releases for EACH AND EVERY Linux kernel version posted to the public, including EACH AND EVERY release candidate. Any of these releases at any time may be ...

add debugger KConfig materials. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/lib/Kconfig.debug 2008-08-15 21:26:27.000000000 -0600 +++ b/lib/Kconfig.debug 2008-08-15 15:41:57.000000000 -0600 @@ -134,6 +134,39 @@ Say Y here if you are developing drivers or trying to debug and identify kernel problems. +config MDB + tristate "Merkey's Kernel Debugger" + depends on DEBUG_KERNEL + select KALLSYMS + select KALLSYMS_ALL + default m + help + ...

Got it. Added to the list. I will Instrument it. Jeff --

export clocksource watchdog to suppress bogus soft lockup reports while processors are held in the debugger. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/kernel/time/clocksource.c 2008-08-15 21:26:27.000000000 -0600 +++ b/kernel/time/clocksource.c 2008-08-15 15:41:57.000000000 -0600 @@ -242,6 +242,7 @@ { clocksource_resume_watchdog(); } +EXPORT_SYMBOL_GPL(clocksource_touch_watchdog); /** * clocksource_get_next - Returns the selected clocksource --

remove conditional exporting of console redirection handle, and just export it. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/drivers/char/vt.c 2008-08-15 21:26:13.000000000 -0600 +++ b/drivers/char/vt.c 2008-08-15 15:41:57.000000000 -0600 @@ -187,6 +187,7 @@ int last_console; int want_console = -1; int kmsg_redirect; +EXPORT_SYMBOL_GPL(kmsg_redirect); /* * For each existing display, we have a pointer to console currently visible --

export the kallsyms lookup functions to the module based debugger. add debugger specific lookup to search for matching strstr text fragments for entered search terms. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/kernel/kallsyms.c 2008-08-15 21:26:27.000000000 -0600 +++ b/kernel/kallsyms.c 2008-08-15 15:41:57.000000000 -0600 @@ -160,6 +160,7 @@ } return module_kallsyms_lookup_name(name); } +EXPORT_SYMBOL_GPL(kallsyms_lookup_name); static unsigned ...

disable the console warning checks if the debugger is active because the debugger is dependent in the console output functions in order to provide console display. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/include/linux/console.h 2008-08-15 21:26:27.000000000 -0600 +++ b/include/linux/console.h 2008-08-15 15:41:57.000000000 -0600 @@ -141,11 +141,16 @@ void vcs_remove_sysfs(struct tty_struct *tty); /* Some debug stub to catch some of the obvious ...

The obvious alternative would be to just set oops_in_progress while the debugger runs. -Andi --

updated and included operating system dependent defines Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-os.h 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-os.h 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,106 @@ + +/*************************************************************************** +* +* Copyright (c) 2008 Jeff V. Merkey All Rights Reserved. +* 1058 East 50 South +* Lindon, Utah 84042 +* ...

added mathematical logic parser and conditional breakpoint support. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-logic.c 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-logic.c 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,2375 @@ + +/*************************************************************************** +* +* Copyright (c) 2008 Jeff V. Merkey All Rights Reserved. +* 1058 East 50 South +* Lindon, Utah 84042 +* ...

added linux specific support. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-os.c 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-os.c 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,1125 @@ + +/*************************************************************************** +* +* Copyright (c) 2008 Jeff V. Merkey All Rights Reserved. +* 1058 East 50 South +* Lindon, Utah 84042 +* jmerkey@wolfmountaingroup.com +* +* This program is ...

added ia32 disassembler and branch detection. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-ia32-support.c 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-ia32-support.c 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,1935 @@ + +/*************************************************************************** +* +* Copyright (c) 2008 Jeff V. Merkey All Rights Reserved. +* 1058 East 50 South +* Lindon, Utah 84042 +* ...

added includes and external defines for keyboard specific code Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-keyboard.h 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-keyboard.h 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,137 @@ + + +/*************************************************************************** +* +* Copyright (c) 2008 Jeff V. Merkey All Rights Reserved. +* 1058 East 50 South +* Lindon, Utah 84042 +* ...

Need to add TSS gates to all exceptions in ia32 since current GDT only uses trap gates. TSS gates allow the processor to switch to a known good stack rather than expect the double fault handler to do so. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-ia32.c 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-ia32.c 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,5651 ...

add architecture independent external defines. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/mdb/mdb-base.h 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/mdb/mdb-base.h 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,296 @@ + +/*************************************************************************** +* +* Copyright (c) 2008 Jeff V. Merkey All Rights Reserved. +* 1058 East 50 South +* Lindon, Utah 84042 +* ...

add a debug directory makefile for loadable kernel debugger modules. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/debug/Makefile 1969-12-31 17:00:00.000000000 -0700 +++ b/debug/Makefile 2008-08-15 15:41:57.000000000 -0600 @@ -0,0 +1,5 @@ +# +# Makefile for kernel debugger modules +# + +obj-$(CONFIG_MDB) += mdb/ --

remove module base conditionals. remove any reference to any file owned or maintained by Ingo Molnar based on his request. Signed-off-by: Jeffrey Vernon Merkey (jmerkey@wolfmountaingroup.com) --- a/arch/x86/kernel/reboot.c 2008-08-15 21:26:12.000000000 -0600 +++ b/arch/x86/kernel/reboot.c 2008-08-15 16:52:30.000000000 -0600 @@ -502,7 +502,10 @@ { machine_ops.restart(cmd); } - +EXPORT_SYMBOL_GPL(machine_restart); +#if defined(CONFIG_SMP) +EXPORT_SYMBOL(genapic); +#endif void ...

The justification is questionable. The fact that the irq number reported for msi-x vectors is only 16bits is a bug. Everywhere else in the kernel an irq number is stored in a 32bit field. We also have a few other bits of craziness in the msix interface. The entry field in msix_entry is unnecessary. In fact now that we have a linked list of irq entries we stop passing the structure entirely. Eric --

So that the driver can release the two request_irq() allocated handlers during tear-down (via qla24xx_disable_msix()->free_irq()). Beyond caching (vector/irq) what's returned during pci_enable_msix(), is there some other mechanism a driver can use to get the IRQ number? --

Yes, which is no bad thing for some platforms. There are some driver assumptions like that but those have also been stomped. Alan --

current code static unsigned int build_irq_for_pci_dev(struct pci_dev *dev) { unsigned int irq; irq = dev->bus->number; irq <<= 8; irq |= dev->devfn; irq <<= 12; return irq; } int arch_setup_msi_irq(struct pci_dev *dev, struct msi_desc *desc) { unsigned int irq; int ret; unsigned int irq_want; irq_want = build_irq_for_pci_dev(dev) + 0x100; irq = create_irq(irq_want); domain is not used ...

On Sat, Aug 16, 2008 at 4:09 PM, James Bottomley msix_entry.vector == irq. and cpu have per_cpu(vector_irq), and use that cpu get irq number. when irq migratition, different cpu could use different vector for same irq. or say phys_flat mode, same vector in different cpus could server different irq. it seems x86 (64 bit) and ia64 already did that. YH --

The 28 bits aren't enough, are they: we need domain as well (and surely we can have more than 16 domains?) -hpa -- Sent from my mobile phone (pardon any lack of formatting) -----Original Message----- From: Yinghai Lu <yhlu.kernel@gmail.com> Sent: Friday, August 15, 2008 19:36 To: Jesse Barnes <jbarnes@virtuousgeek.org>; James Bottomley <James.Bottomley@hansenpartnership.com>; Ingo Molnar <mingo@elte.hu>; Thomas Gleixner <tglx@linutronix.de>; H. Peter Anvin <hpa@zytor.com>; Eric W. ...

Obviously I'm not communicating very well. Your apparent assumption is that irq number == vector. What I'm saying is that's not what we've done for individually vectored CPU interrupts in other architectures. In those we did (cpu no, irq) == vector. i.e. the affinity and the irq number identify the vector. For non-numa systems, this is effectively what you're interested in doing anyway. For numa systems, it just becomes a sparse matrix. James --

Well, given that 2.6.27 uses a compact irq space, probably not ... No, what I mean is that msis can trip directly to CPUs, so this is an affinity thing (that MSI is directly bound to that CPU now), so in the matrixed way we display this in show_interrupts() with the CPU along the top and the IRQ down the side, it doesn't make sense to me to encode IRQ affinity in the irq number again. So it makes more sense to assign the vectors based on both the irq number and the CPU affinity so that if ...

One of my i386 systems used just for testing can take up to 18 hours to build a new kernel with about 200 modules. When performing a bisection, this gets rather tedious. As a result, I moved the kernel sources to an NFS volume, do the majority of the work as a cross-build on my x86_64 computer, and only switch to the slow machine to install the kernel and modules. Somewhere between 2.6.27-rc1 and -rc2, the module installation broke because the build leaves scripts/basic/fixdep as a 64-bit ...

we are using 28bit pci (bus/dev/fn + 12 bits) as irq number, so the cache for irq number should be 32 bit too. Signed-off-by: Yinghai Lu <yhlu.kernel@gmail.com> Cc: Andrew Vasquez <andrew.vasquez@qlogic.com> --- drivers/scsi/qla2xxx/qla_def.h | 2 +- include/linux/pci.h | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) Index: linux-2.6/drivers/scsi/qla2xxx/qla_def.h =================================================================== --- ...

The new design of pre-recorded mcounts and updating the code outside of kstop_machine has changed the way the records themselves are protected. This patch uses the ftrace_lock to protect the records. Note, the lock still does not need to be taken within calls that are only called via kstop_machine, since the that code can not run while the spin lock is held. Also removed the hash_lock needed for the daemon when MCOUNT_RECORD is configured. Also did a slight cleanup of an unused ...

Due to the new mcount recording method, the assumptions about the records that hold the mcount locations have changed. Now we need to protect reading and writing of the records between module load and reading of the debugfs files. There is not much contention here, but they need to be protected none the less. I simply use the ftrace_lock to accomplish this. The ftrace_lock should never be held under interrupt context so disabling interrupts should not be necessary. I've run this under lockdep ...

From: David Brownell <dbrownell@users.sourceforge.net> Add a new mechanism to the rtc-cmos platform glue: an irq_handler_t can now "filter" the interrupts. Update rtc-cmos to use that mechanism where it's been set up, instead of "HPET emulation". http://bugzilla.kernel.org/show_bug.cgi?id=11153 shows the problem: that emulation makes update interrupts misbehave wildly, and thus makes NTP synch tools malfunction. Make ACPI glue always set this using ACPI "events". That seems to work ...

From: Gary Shi <garyu.shi@xxxxxxxxxx> When I read sysrq-t call trace, I find the collected call trace for some cases is not a still snapshot, but a moving one. After checking show_state src, I realize other cpus are not frozen when one cpu is doing show_state. This moving call traces make debugging much more difficult, or even impossible for some cases, since during sysrq-t, some tasks have been switched in/out, threads can jump from kernel space to user space(or vice versa), or just keep ...

From 306bb46f92d326c5afae3fa04860bdccb879b6b8 Mon Sep 17 00:00:00 2001 From: Ingo Molnar <mingo@elte.hu> Date: Sat, 16 Aug 2008 10:18:31 +0200 Subject: [PATCH] x86: io_apic.c, build fix fix: arch/x86/kernel/io_apic.c: In function 'MPBIOS_trigger': arch/x86/kernel/io_apic.c:1105: error: implicit declaration of function 'default_MCA_trigger' the MCA defines got lost in the unification. Signed-off-by: Ingo Molnar <mingo@elte.hu> --- arch/x86/kernel/io_apic.c | 6 ++++++ 1 files ...

ah. EISA got too copies /* ISA interrupts are always polarity zero edge triggered, * when listed as conforming in the MP table. */ #define default_ISA_trigger(idx) (0) #define default_ISA_polarity(idx) (0) /* EISA interrupts are always polarity zero and can be edge or level * trigger depending on the ELCR value. If an interrupt is listed as * EISA conforming in the MP table, that means its trigger type must * be read in from the ELCR */ #define ...

Here is a kernel log with full debugging output enabled for the iwl drivers: [ 8.881684] asus-laptop: M50Vm model detected [ 8.901015] acpi device:1e: registered as cooling_device2 [ 8.901307] input: Video Bus as /class/input/input7 [ 8.902500] Registered led device: asus::mail [ 8.902689] Registered led device: asus::touchpad [ 8.964753] ACPI: Video Device [VGA] (multi-head: yes rom: no post: no) [ 9.030505] iwlagn: Intel(R) Wireless WiFi Link AGN driver for Linux, ...

At Fri, 15 Aug 2008 15:02:21 -0400, Actually I hit the related Oops with the latest linux-next tree. Since this bug appears first on that tree and didn't occur on the previous tree, it was apparently a regression. So, I looked through changes over knfsd since the previous tree, found a suspicious one, and tried to fix. The linux-next tree is pretty helpful to catch a regression if one eats own dog foods everyday. BTW, I've been building linux-next SUSE kernel packages at every linux-next ...

have just tried a test-merge, and got a few crashes during -tip testing. One is a hard lockup during bootup, with this config: http://redhat.com/~mingo/misc/config-Sat_Aug_16_16_49_05_CEST_2008.bad no console output. I've pushed out this integrated tree into tip/x86/time.broken: note that i had to do a few manual merges due to the io_apic.c unification - maybe that interacted. there's a 32-bit failure as well: ...

At Fri, 15 Aug 2008 21:49:54 +0200, Hm, OK. Originally I fixed it because our build system has a sanity check against bogus symlinks and refused to package the latest kernel. If you don't want to put it in, we can keep it locally, of course. But, a bogus symlink is definitely not pretty... thanks, Takashi --

On Sat, 16 Aug 2008 13:57:50 +1000 The problem TALPA is trying to solve is only part of the puzzle. Everyone recognizes that. It's a very relevant part of the puzzle (in corporate context at least), but it's very much so not a complete puzzle. Does that mean we shouldn't deal with this just because it's incomplete? Absolutely not! (nor should we do something that has no value.. but that's not the case; the model that Erik described is quite well defined as "do not give ''bad' content to ...

The threat module you are looking at does not cover all the real world usage even worse detection of unknown real world threats. Currently if we have a unknown infection on a windows partition that is been shared by linux the scanner on Linux cannot see that the windows permissions has been screwed with. OS with badly damaged permissions is a sign of 1 of three things. 100 percent incompetent admin, failing harddrive or system is breached. Now if system is breached on a partition it is ...

On Fri, 15 Aug 2008 21:09:42 PDT, Arjan van de Ven said: Many security experts believe that a false sense of security is worse than no security at all. In other words, unless the design team is *honest* with themselves about what the proposal does and doesn't cover, and has at least an *idea* of how the uncovered parts will function, you're not adding to the *real* security. The problem with saying stuff like "Oh, our threat model explicitly rules out anything done by root" is that all too ...

These file system caches are internal permissions caching points where the driver decides what you can and cannot see. Before conversion to normal inode structs. Others have own internal buffers for transfers. Yes everything is stored on the page cache but it does not have to be in any shape you would normally id as a file. I have a bad habit of putting buffers and caches in the same box. Thinking that some file system drivers are smart enough to use the same buffer if they get the same ...

Thanks to everyone that answered, probably the solution is an update to the driver, hardware support for this chip was added in 27-rc. My distro still uses 26.2. I'm going on holydays for a week now, will make more tests on return, and continue the thread. Thanks everyone. -- J.A. Magallon <jamagallon()ono!com> \ Software is like sex: \ It's better when it's free Mandriva Linux release 2009.0 (Cooker) for i586 Linux ...

Hi Takashi and thanks, This patch is first one you sent, except differing line numbers (rediffed against 2.6.27-rc3?). You could try attaching the patch to the mail too, to avoid the line wrapping, silent whitespace conversion and any gmail/mailer 'hide quoted text' mangling... Thanks, Daniel -- Daniel J Blueman --

Hi Takashi, The updated patch validates fully also. Checking everything with codecgraph and varying both capture inputs looks as expected, so thanks again! Tested-by: Daniel J Blueman <daniel.blueman@gmail.com> Daniel -- Daniel J Blueman --

Looks good to me. I'll test it on Monday. I don't know if I have a system with memory holes to test on, but I take it you do? I notice some of our 32-bit code is using lmb_enforce_memory_limit() to enforce an address limit, which is technically broken, but is probably Acked-by: Michael Ellerman <michael@ellerman.id.au> cheers --=20 Michael Ellerman OzLabs, IBM Australia Development Lab wwweb: http://michael.ellerman.id.au phone: +61 2 6212 1183 (tie line 70 21183) We do not ...

Simplest fix is to shuffle Makefile. But better is to create an acpi "module" so the namespacing just works, something like below. Overriding MODULE_PREFIX only works for builtin code anyway. (Which makes sense: moving a parameter from one module to another isn't a change we can cover up). (Sam: foo-objs-y would make this neater) diff -r 5f7194400572 drivers/acpi/Makefile --- a/drivers/acpi/Makefile Sat Aug 16 13:23:26 2008 +1000 +++ b/drivers/acpi/Makefile Sat Aug 16 13:44:17 2008 ...

Ah, missed that. Thanks Sam! I had another bug: if both options were off, compile would fail. Put everything in acpi.o to fix this. Rusty. diff -r 79f0d024aebc drivers/acpi/Makefile --- a/drivers/acpi/Makefile Fri Aug 15 11:23:34 2008 +1000 +++ b/drivers/acpi/Makefile Sat Aug 16 15:55:10 2008 +1000 @@ -21,10 +21,15 @@ obj-$(CONFIG_X86) += blacklist.o # # ACPI Core Subsystem (Interpreter) # -obj-y += osl.o utils.o reboot.o\ +obj-y += acpi.o \ dispatcher/ events/ ...

sysfs does this anyways, doesn't it. We would just need to teach it to not BUG() in this case, perhaps with a special entry point. Also a BUG() in general seems a little harsh for this, surely a WARN_ON should be enough. -Andi --

Yeah, they are stored in a global array __start___param, and the current code expects the prefixes to be sorted, so the sysfs directories can be They have been module options, not prefixed kernel parameters so far, and the prefix was just the module name. So it just strikes back, that acpi uses generic names for the modules, there would have been no problem if "power" would be called "acpi_power" and the options would just be "acpi.acpica_version" and "acpi_power.nocheck". But well, there ...

Actually, I seem to remember one cause of this: the vdso code creates an unusual stackframe where the pointer to the previous frame is stored in an unusual place. I think the sysprof kernel code needs to take this into account, or alternatively send sufficient information to userspace that it can figure it out itself. Soren --

That is not ok, but it could happen if the process in question was not compiled with framepointers. Or it could be a bug in the kernel code. Soren --

Hm, sysfs hasn't changed any in 2.6.27-rcX that I know of. thanks, greg k-h --

My guess is that MCE does somthing that is not allowed by sysfs any more. Thanks, Rafael --

Sorry for the delayed response, after passing the slub_debug as the command line parameter while booting the kernel oops, the kernel faults with message Kernel command line: root=/dev/sdb1 selinux=0 elevator=cfq slub_debug IDENT=1218735938 . . <snip> . . Unable to handle kernel paging request for data at address 0x6b6b6b6b6b6b6be3 Faulting instruction address: 0xc0000000002afb98 Oops: Kernel access of bad area, sig: 11 [#1] SMP NR_CPUS=128 NUMA pSeries Modules linked in: NIP: ...

correct - assuming drivers/telephony/ is maintained. Maybe Andrew could mark this patch as v2.6.27-must-have - to make sure it hits upstream in this cycle? Ingo --

Actually, since it's unused at the moment, we can define it however we want. But note that this is an ABI; while the kernel-internal definitions are fluid, this semantic must stay the same (even if the actual values differ). So we should probably put an explicit mapping function there anyway. Thanks, Rusty. --

I tried just the writeback_index patch and got only 4 fragmented files on ext4 after a compilebench run. Then I tried again and got 1200. Seems there is something timing dependent in here ;) By default compilebench uses 256k buffers for writing (see compilebench -b) and btrfs_file_write will lock down up to 512 pages at a time during a single write. This means that for most small files, compilebench will send the whole file down in one write() and btrfs_file_write will lock down pages for ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: memcpy export needs to follow host declaration to the 2.6.25-stable tree. Its filename is uml-memcpy-export-needs-to-follow-host-declaration.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:15:05 2008 From: Jeff Dike <jdike@addtoit.com> Date: Tue, 5 Aug ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: stub needs to tolerate SIGWINCH to the 2.6.25-stable tree. Its filename is uml-stub-needs-to-tolerate-sigwinch.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:15:05 2008 From: Jeff Dike <jdike@addtoit.com> Date: Tue, 5 Aug 2008 16:14:09 -0400 Subject: ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: Fix boot crash to the 2.6.25-stable tree. Its filename is uml-fix-boot-crash.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:14:58 2008 From: Jeff Dike <jdike@addtoit.com> Date: Tue, 5 Aug 2008 16:14:10 -0400 Subject: uml: Fix boot crash To: ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: missing export of csum_partial() on uml/amd64 to the 2.6.25-stable tree. Its filename is uml-missing-export-of-csum_partial-on-uml-amd64.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:14:50 2008 From: Al Viro <viro@ftp.linux.org.uk> Date: Tue, 5 Aug ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: missed kmalloc() in pcap_user.c to the 2.6.25-stable tree. Its filename is uml-missed-kmalloc-in-pcap_user.c.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:14:38 2008 From: Al Viro <viro@ftp.linux.org.uk> Date: Tue, 5 Aug 2008 16:14:05 -0400 Subject: ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: deal with inaccessible address space start to the 2.6.25-stable tree. Its filename is uml-deal-with-inaccessible-address-space-start.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:14:42 2008 From: Tom Spink <tspink@gmail.com> Date: Tue, 5 Aug 2008 ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: deal with host time going backwards to the 2.6.25-stable tree. Its filename is uml-deal-with-host-time-going-backwards.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:14:23 2008 From: Jeff Dike <jdike@addtoit.com> Date: Tue, 5 Aug 2008 16:14:06 ...

This is a note to let you know that we have just queued up the patch titled Subject: uml: fix bad NTP interaction with clock to the 2.6.25-stable tree. Its filename is uml-fix-bad-ntp-interaction-with-clock.patch A git repo of this tree can be found at http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary From stable-bounces@linux.kernel.org Tue Aug 5 13:14:46 2008 From: Jeff Dike <jdike@addtoit.com> Date: Tue, 5 Aug 2008 16:14:03 ...

It would be better to do it unconditionally if possible - remove the Please avoid using the sys_ namespace. That is for system calls. How does this differ from sys_utime()? --

thank you Ingo. ciao, -- Paolo http://paolo.ciarrocchi.googlepages.com/ --

You should have started out by discussing your design - the document just rambles a bit about some implementation details - it doesn't talk about how it maps to the PI problem space. Anyway - from what I can make of the code, you managed to convert the pi graph walking code that used to be in rt_mutex_adjust_prio_chain() and was iterative, into a recursive function call. Not something you should do lightly.. --

Ok, so a specialized version of printk will do the work to calculate the hash. Only, what will we do if there ever is a conflict? The message tag has to be unique. The shorter the hash is, the more likely a collision gets. Don't know if 6 hash digits is enough to just ignore the Sounds doable to me. -- blue skies, Martin. "Reality continues to ruin my life." - Calvin. --

The kmsg id is just that, an index into the table of message for a That is what we are trying to achieve as well. -- blue skies, Martin. "Reality continues to ruin my life." - Calvin. --

The invention of dev_printk was a good thing. It adds structure to the messages if it is about a device. Before we had a simple printk that did or did not include information about the device driver and the device name. dev_printk makes sure that this information is always added. BUT, what dev_printk does not do is to make sure that the message can be uniquely identified so that the description of the message can be found in the message catalog. Our approach is to combine the driver name and ...

On Fri, 15 Aug 2008 20:00:48 +0200 Ah, sorry, @@ -262,7 +264,11 @@ static dma_addr_t dma_map_area(struct device *dev, dma_addr_t phys_mem, static dma_addr_t gart_map_simple(struct device *dev, phys_addr_t paddr, size_t size, int dir) { - dma_addr_t map = dma_map_area(dev, paddr, size, dir); + dma_addr_t map; + unsigned long align_mask; + + align_mask = (roundup_pow_of_two(size) >> PAGE_SHIFT) - 1; + map = dma_map_area(dev, paddr, size, dir, align_mask); This code doesn't work with ...

I finally got a chance to test 2.6.27-rc2 and confirm that my laptop correctly powers off with "shutdown method = platform now". Do you have any idea which patch might have fixed this? It would be nice to include it in Debian's 2.6.26 kernel. -- Martin Michlmayr http://www.cyrius.com/ --

Yep, this one has problems as I just noticed. Reworking... --