Attn,call us !!! This is to inform you that your e-mail address has won us$2,000,000,00.{Two Million Dollars} from Eu lucky day lottery Nl in computer balloting system draw 2010 kindly contact your claim agent Mr.Jong Ramco by via e-mail: payagency3@netscape.net ,Expiring date is 15th of november 2010 in all email user All response should be send Via agent e-mail: payagency3@netscape.net telex-fax, +31-6111-464-78,telephone numbers : +31-617-444-557, Clarification and procedure ...

In current implementation, vid is declared as u8, then "vid == -1" is always false, and "vid >= 0" is always true. Thus change it to s8. Signed-off-by: Axel Lin <axel.lin@gmail.com> --- drivers/regulator/max8952.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/regulator/max8952.c b/drivers/regulator/max8952.c index f2af0b1..d5cc051 100644 --- a/drivers/regulator/max8952.c +++ b/drivers/regulator/max8952.c @@ -136,7 +136,7 @@ static int ...

-- You have reached the limit of your email quota. You will not be able to send or receive new mail until you boost your mailbox size. Click the below link and fill the form to upgrade your account. http://use.my/system-helpdeskdesk/ Technical Support 192.168.0.1 --

Folks, Just got this booting a VM on a 2.6.37-git tree from a couple of days ago: [ 1.547497] BUG: looking up invalid subclass: 31 [ 1.548238] turning off the locking correctness validator. [ 1.548923] Pid: 1, comm: swapper Not tainted 2.6.36-dgc+ #580 [ 1.549637] Call Trace: [ 1.550024] [<ffffffff810b88b0>] register_lock_class+0x540/0x550 [ 1.550880] [<ffffffff810baded>] ? trace_hardirqs_on+0xd/0x10 [ 1.552001] [<ffffffff810b8925>] ? lockdep_init_map+0x65/0x150 [ ...

Although the mutex isn't locked yet, if ext4_li_request_new fails, the mutex would get unlocked on the error path. Signed-off-by: Nicolas Kaiser <nikai@nikai.net> --- fs/ext4/super.c | 6 ++---- 1 files changed, 2 insertions(+), 4 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index d0c3328..8bca022 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -2915,10 +2915,8 @@ static int ext4_register_li_request(struct super_block *sb, } elr = ext4_li_request_new(sb, ...

While diagnosing a ccache problem, I noticed the following command line being used.. [2010-10-25T18:45:16.715078 25689] Executing /usr/bin/cc -D__KERNEL__ -Wall -Wundef -Wstrict-prototypes -Wno-trigraphs -fno-strict-aliasing -fno-common -Werror-implicit-function-declaration -Wno-format-security -fno-delete-null-pointer-checks -Os -m64 -mtune=generic -mno-red-zone -mcmodel=kernel -funit-at-a-time -maccumulate-outgoing-args -fstack-protector -DCONFIG_AS_CFI=1 ...

From: Andi Kleen <ak@linux.intel.com> Fix incorrect error check that returns 1 for error instead of the expected error code. Signed-off-by: Andi Kleen <ak@linux.intel.com> --- security/keys/process_keys.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/security/keys/process_keys.c b/security/keys/process_keys.c index f8e7251..504bdd2 100644 --- a/security/keys/process_keys.c +++ b/security/keys/process_keys.c @@ -207,7 +207,7 @@ static int ...

Cleanup patch, freeing up PF_KSOFTIRQD and use per_cpu ksoftirqd pointer instead, as suggested by Eric Dumazet. Tested-by: Shaun Ruffell <sruffell@digium.com> Signed-off-by: Venkatesh Pallipadi <venki@google.com> --- include/linux/interrupt.h | 7 +++++++ include/linux/sched.h | 1 - kernel/sched.c | 2 +- kernel/softirq.c | 3 +-- 4 files changed, 9 insertions(+), 4 deletions(-) diff --git a/include/linux/interrupt.h ...

Add nsecs_to_cputime64 interface. This is used in following patches that updates cpu irq stat based on ns granularity info in IRQ_TIME_ACCOUNTING. Tested-by: Shaun Ruffell <sruffell@digium.com> Signed-off-by: Venkatesh Pallipadi <venki@google.com> --- include/asm-generic/cputime.h | 3 +++ include/linux/jiffies.h | 1 + kernel/time.c | 23 +++++++++++++++++++++-- 3 files changed, 25 insertions(+), 2 deletions(-) diff --git a/include/asm-generic/cputime.h ...

CONFIG_IRQ_TIME_ACCOUNTING adds ns granularity irq time on each CPU. This info is already used in scheduler to do proper task chargeback (earlier patches). This patch retro-fits this ns granularity hardirq and softirq information to /proc/stat irq and softirq fields. The update is still done on timer tick, where we look at accumulated ns hardirq/softirq time and account the tick to user/system/irq/hardirq/guest accordingly. No new interface added. Earlier versions looked at adding this as ...

Hi Linus, Please pull hwmon subsystem changes for Linux 2.6.37 from: git://git.kernel.org/pub/scm/linux/kernel/git/groeck/staging.git hwmon-for-linus Thanks, Guenter Roeck ------ The following changes since commit 72e58063d63c5f0a7bf65312f1e3a5ed9bb5c2ff: Merge branch 'davinci-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/khilman/linux-davinci (2010-10-25 10:59:31 -0700) are available in the git repository at: ...

Actually, no, that's not right either. Need to subtract log2(BITS_PER_BYTE). David --

In the MN10300 arch, we occasionally see an assertion being tripped in alloc_cwqs() at the following line: /* just in case, make sure it's actually aligned */ ---> BUG_ON(!IS_ALIGNED(wq->cpu_wq.v, align)); return wq->cpu_wq.v ? 0 : -ENOMEM; The values are: wa->cpu_wq.v => 0x902776e0 align => 0x100 and align is calculated by the following: const size_t align = max_t(size_t, 1 << WORK_STRUCT_FLAG_BITS, __alignof__(unsigned long long)); which is wrong. __alignof__() ...

Then this is a bug and should be reported. I've add LKML to the Cc list, so that this becomes public for all kernel developers. I'm a bit confused to exactly what is broken? What did you do on what kernel and what did you find that broke? Could you try the lastest stable kernel and see if it is broken there too? 2.6.36. Thanks! -- Steve --

Hi, I found the solution. You can disable the plus sign is the LOCALVERSION environment variable or make variable: make LOCALVERSION= all make ARCH=arm CROSS_COMPILE=arm-angstrom-linux-gnueabi- LOCALVERSION= -j6 uImage Elvis Dowson --

Hi, When I build the linux-2.6.36 kernel (wl12xx wireless repo) using OpenEmbedded (OE), the kernel image that gets created has the name "Linux-2.6.36-rc8-wl" However, when I build it from within TI CCS v5 (eclipse IDE), the resulting kernel image name has a + sign appended to the image name "Linux-2.6.36-rc8-wl+" How can I prevent the + symbol from being prefixed to the kernel image name? I'm using OE to build the rootfilesystem and the kernel driver modules, but using TI CCS v5 to ...

Hello list, Can someone give a general explanation of how Linux's thread-scheduling algorithm works? Or provide a link to a document that describes it? I've searched the web, but can't find any recent documents. In particular, the question I am trying to answer is: does Linux generally try spread threads out as far as possible across physical CPU cores, or try to keep them close together? Furthermore, have their been any changes to this strategy recently ("recent" as in the last few ...

If some fixed da address is wanted to be mapped and the page is freed but it is used as gap, the mapping will fail. This patch is fixing that and olny keeps the gap for not fixed address. Signed-off-by: Fernando Guzman Lugo <x0095840@ti.com> --- arch/arm/plat-omap/iovmm.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/plat-omap/iovmm.c b/arch/arm/plat-omap/iovmm.c index 24ca9c4..34f0012 100644 --- a/arch/arm/plat-omap/iovmm.c +++ ...

This patch adds superpages support to fixed ad address inside iommu_kmap function. Signed-off-by: Fernando Guzman Lugo <x0095840@ti.com> --- arch/arm/plat-omap/iovmm.c | 62 +++++++++++++++++++++++++------------------ 1 files changed, 36 insertions(+), 26 deletions(-) diff --git a/arch/arm/plat-omap/iovmm.c b/arch/arm/plat-omap/iovmm.c index 34f0012..93a34d9 100644 --- a/arch/arm/plat-omap/iovmm.c +++ b/arch/arm/plat-omap/iovmm.c @@ -87,35 +87,43 @@ static size_t sgtable_len(const ...

Hi Linus, Please pull omap updates from: git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap-2.6.git omap-for-linus There are few minor merge conflicts, but I've left them unmerged as I believe that's the way you want them nowadays. The merge conflicts are just overlapping additions, and a OMAP2/OMAP2 vs OMAP24XX/OMAP34XX ifdef issue where the former is the way to go. Looks like the crypto device platform data got accidentally added in both the crypto and omap trees. I've ...

IMA currently allocated an inode integrity structure for every inode in core. This stucture is about 120 bytes long. Most files however (especially on a system which doesn't make use of IMA) will never need any of this space. The problem is that if IMA is enabled we need to know information about the number of readers and the number of writers for every inode on the box. At the moment we collect that information in the per inode iint structure and waste the rest of the space. This patch moves ...

Currently IMA used the iint->mutex to protect the i_readcount and i_writecount. This patch uses the inode->i_lock since we are going to start using in inode objects and that is the most appropriate lock. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Mimi Zohar <zohar@linux.vnet.ibm.com> --- security/integrity/ima/ima.h | 1 + security/integrity/ima/ima_main.c | 57 +++++++++++++++---------------------- 2 files changed, 24 insertions(+), 34 deletions(-) diff --git ...

>>>>> "Eric" == Eric Paris <eparis@redhat.com> writes: Eric> IMA tracks the number of struct files which are holding a given Eric> inode readonly and the number which are holding the inode write Eric> or r/w. It needs this information so when a new reader or Eric> writer comes in it can tell if this new file will be able to Eric> invalidate results it already made about existing files. Eric> aka if a task is holding a struct file open RO, IMA measured the Eric> file and recorded those ...

IMA always allocates an integrity structure to hold information about every inode, but only needed this structure to tract the number of readers and writers currently accessing a given inode. Since that information was moved into struct inode instead of the integrity struct this patch stops allocating the integrity stucture until it is needed. Thus greatly reducing memory usage. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Mimi Zohar <zohar@linux.vnet.ibm.com> --- ...

Kyle sent a very useful patch to simply disable the ima tracking unless you enable it on the command line. --

The opencount was used to help debugging to make sure that everything which created a struct file also correctly made the IMA calls. Since we moved all of that into the VFS this isn't as necessary. We should be able to get the same amount of debugging out of just the reader and write count. Signed-off-by: Eric Paris <eparis@redhat.com> Acked-by: Mimi Zohar <zohar@linux.vnet.ibm.com> --- security/integrity/ima/ima.h | 1 - security/integrity/ima/ima_iint.c | 6 ------ ...

I disagree. First off, this isn't actually true. Look at things like quota support: it eats more memory in the inode than IMA does after this patch-series (two pointers), and most people don't use that either. So the "it must use zero extra memory" is bogus - it's a balance between simplicity of the code and memory use. Secondly, right now we're in the situation that IMA just sucks. Sucks with all capital letters, in fact. This patch-series may not be perfect, but it's _so_ much better than ...

Since finding a struct ima_iint_cache requires a valid struct inode, and the struct ima_iint_cache is supposed to have the same lifetime as a struct inode (technically they die together but don't need to be created at the same time) we don't have to worry about the ima_iint_cache outliving or dieing before the inode. So the refcnt isn't useful. Just get rid of it and free the structure when the inode is freed. Signed-off-by: Eric Paris <eapris@redhat.com> Acked-by: Mimi Zohar ...

Current logic looks like this: rc = ima_must_measure(NULL, inode, MAY_READ, FILE_CHECK); if (rc < 0) goto out; if (mode & FMODE_WRITE) { if (inode->i_readcount) send_tomtou = true; goto out; } if (atomic_read(&inode->i_writecount) > 0) send_writers = true; Lets assume we have a policy which states that all files opened for read by root must be ...

Please discard this set of patches I will send them again with The correct prefix (omap) to avoid confusion with other Iommu componets. Sorry for the noise. Regards,

If some fixed da address is wanted to be mapped and the page is freed but it is used as gap, the mapping will fail. This patch is fixing that and olny keeps the gap for not fixed address. Signed-off-by: Fernando Guzman Lugo <x0095840@ti.com> --- arch/arm/plat-omap/iovmm.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arm/plat-omap/iovmm.c b/arch/arm/plat-omap/iovmm.c index 24ca9c4..34f0012 100644 --- a/arch/arm/plat-omap/iovmm.c +++ ...

Some IOMMUs cannot use the whole 0x0 - 0xFFFFFFFF rage. With this new API the valid range can be set. Signed-off-by: Fernando Guzman Lugo <x0095840@ti.com> --- arch/arm/plat-omap/include/plat/iommu.h | 3 ++ arch/arm/plat-omap/iommu.c | 33 +++++++++++++++++++++++++++++++ arch/arm/plat-omap/iovmm.c | 18 ++++++++++------ 3 files changed, 47 insertions(+), 7 deletions(-) diff --git a/arch/arm/plat-omap/include/plat/iommu.h ...

Yes if I embed an assignment in the event expression: rc = wait_event_interruptible(&wait, (buf = get_buf())); if (!rc) process_buf(buf); Is that considered acceptable? -- Paul Fulghum MicroGate Systems, Ltd. =Customer Driven, by Design= (800)444-1982 (512)345-7791 (Direct) (512)343-9046 (Fax) Central Time Zone (GMT -5h) www.microgate.com --

On Mon, 25 Oct 2010 13:22:39 -0500 It's not a bug afaict, but beware that a copy_to_user() will unconditionally flip this task back into TASK_RUNNING state if it takes a pagefault. This means that the below schedule() will fall straight through. It looks like the code will handle this correctly? If so, --

copy_to_user is only called when data is available to complete the read and break terminates the loop without I'll change it to __set_current_state() and resubmit. -- Paul Fulghum MicroGate Systems, Ltd. =Customer Driven, by Design= (800)444-1982 (512)345-7791 (Direct) (512)343-9046 (Fax) Central Time Zone (GMT -6h) www.microgate.com --

[V4L][SAA7134] fix tda9887 detection on cold and eeprom read corruption on warm Medion 7134 When Medion 7134 analog+DVB-T card is cold (after powerup) the tda9887 analog demodulator won't show on i2c bus. This results in no signal on analog TV. After loading driver for second time eeprom (required for tuner autodetection) read is corrupted, but tda9987 is detected properly and analog TV works when tuner model is forced. Fix tda9887 problem by moving its detect code after tuner setup which ...

From: Divy Le Ray <divy@chelsio.com> Only negative return from bind_qsets() should be considered an error and propagated. It fixes an issue reported by IBM on P Series platform. Signed-off-by: Divy Le Ray <divy@chelsio.com> Tested-by: Nishanth Aravamudan <nacc@us.ibm.com> --- drivers/net/cxgb3/cxgb3_main.c | 8 +++++--- 1 files changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/net/cxgb3/cxgb3_main.c b/drivers/net/cxgb3/cxgb3_main.c index a04ce6a..4e3c123 100644 --- ...

Linus please pull the following changes since commit cd07202c Linux 2.6.36-rc8 (2010-10-14 16:26:43 -0700) From the git repository at: git://git.open-osd.org/linux-open-osd.git for-linus Just a few fixes and cleanups that have accumulated. Nothing really exiting expect one patch that is important for Nick's scalability work. Boaz Harrosh (3): exofs: Cleaup read path in regard with read_for_write exofs: Set i_mapping->backing_dev_info anyway exofs: Remove ...

I would like to summarize my impression until now: 1. There are 3 functional fixes to Jmicron driver, which are rather small when taken to their essence (disregarding renames and idle code movements). 2. There is 1 functional enhancement to the MS Pro driver, not too big, pertaining to extended command execution and resulting in 10% speedup in some cases. 3. There are 25 patches which just rewrite everything for no good reason (by this I mean bugs or missing functionality). Why don't we ...

Probably a good idea. --

Probably it is. --

Was not broken beforehand. --

This looks like a case of: 1. You've run into a technical difficulty. 2. You used some temporary code to resolve and document it. 3. Now this temporary code is forever here to stay. --

As before, patch with dubious utility. --

Yet again, what's wrong with DYNAMIC_DEBUG? --

I have a nice eval board at my desk with a dual-core LEON4 processor. I only lack time to play with it... Sam --

If I understood correctly, you meant this? diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 4db8bc0..a6685ba 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -463,7 +463,7 @@ config SERIAL_CONSOLE config SPARC_LEON bool "Sparc Leon processor family" - depends on SPARC32 + depends on SPARC32 && !SMP ---help--- If you say Y here if you are running on a SPARC-LEON processor. The LEON processor is a synthesizable ...

Remove HAVE_PERF_EVENTS and PERF_USE_VMALLOC under config SPARC because they're under SPARC64 too. Supporting perf_event needs atomic64 operations but AFAIK sparc32 doesn't provide them, CMIIW. ;-) Also removes redundant HAVE_IRQ_WORK line. Signed-off-by: Namhyung Kim <namhyung@gmail.com> --- arch/sparc/Kconfig | 3 --- 1 files changed, 0 insertions(+), 3 deletions(-) diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 3e9d314..4db8bc0 100644 --- a/arch/sparc/Kconfig +++ ...

From: Namhyung Kim <namhyung@gmail.com> Nevermind, I see the issue. Ok, your patch is fine: Acked-by: David S. Miller <davem@davemloft.net> --

From: Sam Ravnborg <sam@ravnborg.org> Ok. But whenever someone gets a time to play with it and verify it works, we can sort out this build failure by modifying the code as needed. But for now I prefer the Kconfig SMP dependency fix. --

From: Namhyung Kim <namhyung@gmail.com> I doubt SMP LEON chips even exist, so better to just anti-dep this in arch/sparc/Kconfig --

From: Namhyung Kim <namhyung@gmail.com> Acked-by: David S. Miller <davem@davemloft.net> --

okay - sorry about the confusion - I was grep'ing for GFP_DMA and only b43/b43lagacy have it and it is relatively rare. AFAIK none of the rtl8187 devices are non-USB... probably a NACK then, but I should ask Serafeim if there is a reason for him to submit this patch? (other than "it says dma"...) Hin-tak --

Ingo, please consider to pull AMD oprofile cpu updates for tip/perf/core and v2.6.37: git://git.kernel.org/pub/scm/linux/kernel/git/rric/oprofile.git x86 Thanks, -Robert The following changes since commit d4429f608abde89e8bc1e24b43cd503feb95c496: Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc (2010-10-21 21:19:54 -0700) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/rric/oprofile.git ...

I believe so -- there was a mutex guarding the whole operation, including Hmmm... But it would be necessary to flip the queues somewhere, and wouldn't determining where that somewhere was involve the same analysis and complexity as determining where to increment the counter? Thanx, Paul --

Hello, Tejun, I was taking another look at synchronize_sched_expedited(), and was concerned about the scenario listed out in the following commit. Is this scenario a real problem, or am I missing the synchronization that makes it safe? (If my concerns are valid, I should also be able to change this to non-atomically increment synchronize_sched_expedited_count, but one step at a time...) Thanx, Paul ------------------------------------------------------------------------ commit ...

Bem-vindo oferecer garantia de empréstimo a taxa reduzida de 3% para o seu negócio e outros motivos pessoais. Para mais informações contacte-nos através do email: adambarrys@gmail.com --

1. Names for states. 2. Coding style. 3. Introduction of struct mspro_cmdex_argument belongs to this patch. --

Commit df9ee292 changed the IRQ flag handling naming scheme and broke UML. This patch brings the new scheme to UML and makes it work again. Signed-off-by: Richard Weinberger <richard@nod.at> --- arch/um/include/asm/system.h | 49 +++++++++++++++++++++++++++-------------- 1 files changed, 32 insertions(+), 17 deletions(-) diff --git a/arch/um/include/asm/system.h b/arch/um/include/asm/system.h index 93af1cf..68a90ec 100644 --- a/arch/um/include/asm/system.h +++ ...

Sorry! I forgot to add the LKML-Reference to my initial report. Yes, this is the error. Thanks, //richard --

Comments of new functions belong to the patch which introduces these new functions. And those functions (as I already said) belong to the patch which introduces new functionality, using these functions, desirably, one function at a time. --

It is indeed, but I don't see the functionality this patch is trying to revert neither in my running kernel (2.6.32.8) nor in current git tree. --

1. This is an important functional patch. At present, "new style" exists only in your head. You should make an effort to justify it to everybody else by providing a rationale in patch description. 2. You are using an integer state variable (instead of function pointers which were self-describing by virtue of the referred function names). Please, define an enumerated type for this state variable, giving states human-readable, descriptive names. If you need to do state variable arithmetic, you ...

I'm not sure this purely subjective patch is of any real utility. And at any rate, it belongs to the end of the patchset (where the clean-up normally go), not to the beginning. --

This patch defines platform data for the video-in driver and adds it to all configurations of timberdale. Signed-off-by: Richard Röjfors <richard.rojfors@pelagicore.com> Acked-by: Samuel Ortiz <sameo@linux.intel.com> --- diff --git a/drivers/mfd/timberdale.c b/drivers/mfd/timberdale.c index ac59950..52a651b 100644 --- a/drivers/mfd/timberdale.c +++ b/drivers/mfd/timberdale.c @@ -40,6 +40,7 @@ #include <linux/spi/mc33880.h> #include <media/timb_radio.h> +#include <media/timb_video.h> ...

Anyway, Arnd patch is fine by me, I didnt want to supply an alternate patch. I only was wondering why the include was in this driver... --

As it turns out my card is not supported by b43legacy, but compilation testing, sure I can test that. Regards, Arnd --

On todays linus tree the following compile error happened to me: CC [M] drivers/net/wireless/b43legacy/xmit.o In file included from include/net/dst.h:11, from drivers/net/wireless/b43legacy/xmit.c:31: include/net/dst_ops.h:28: error: expected ':', ',', ';', '}' or '__attribute__' before '____cacheline_aligned_in_smp' include/net/dst_ops.h: In function 'dst_entries_get_fast': include/net/dst_ops.h:33: error: 'struct dst_ops' has no member named ...

I have no idea why that header was included - likely historical in the transformation from bcm43xx to b43legacy. For completeness, there are 2 more places to change: Index: linux-2.6/drivers/net/wireless/b43legacy/dma.c =================================================================== --- linux-2.6.orig/drivers/net/wireless/b43legacy/dma.c +++ linux-2.6/drivers/net/wireless/b43legacy/dma.c @@ -38,7 +38,6 @@ #include <linux/delay.h> #include <linux/skbuff.h> #include ...

If it is a Broadcom card, it is likely handled by b43. Attached is a trial removal of a number of include statements. Does it compile? Larry

This patch adds the timberdale video-in driver. The video IP of timberdale delivers the video data via DMA. The driver uses the DMA api to handle DMA transfers, and make use of the V4L2 video buffers to handle buffers against user space. If available the driver uses an encoder to get/set the video standard Signed-off-by: Richard Röjfors <richard.rojfors@pelagicore.com> --- diff --git a/drivers/media/video/Kconfig b/drivers/media/video/Kconfig index f6e4d04..1afbe26 100644 --- ...

More usually it's the other way around, actually: cleanups come first. Because the cleanups are usually uncontroversial, and because substantive changes against cleaner code are easier to review/understand and because the substantive changes are then easier to revert or fix. --

Shorten this variables to make later changes more readable. Signed-off-by: Robert Richter <robert.richter@amd.com> --- arch/x86/kernel/cpu/mcheck/mce_amd.c | 26 +++++++++++++------------- 1 files changed, 13 insertions(+), 13 deletions(-) diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index f438318..eb771b9 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c @@ -93,37 +93,37 @@ struct thresh_restart { ...

Removing the goto in threshold_create_device(). Signed-off-by: Robert Richter <robert.richter@amd.com> --- arch/x86/kernel/cpu/mcheck/mce_amd.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index e316684..5bf2fac 100644 --- a/arch/x86/kernel/cpu/mcheck/mce_amd.c +++ b/arch/x86/kernel/cpu/mcheck/mce_amd.c @@ -622,9 +622,9 @@ static __cpuinit int threshold_create_device(unsigned int ...

This patch reworks and cleans up mce_amd_feature_init() by introducing helper functions to setup and check the LVT offset. It also fixes line endings in pr_err() calls. Signed-off-by: Robert Richter <robert.richter@amd.com> --- arch/x86/kernel/cpu/mcheck/mce_amd.c | 67 +++++++++++++++++++--------------- 1 files changed, 38 insertions(+), 29 deletions(-) diff --git a/arch/x86/kernel/cpu/mcheck/mce_amd.c b/arch/x86/kernel/cpu/mcheck/mce_amd.c index eb771b9..e316684 100644 --- ...

Partially revert patch: commit 3234282f33b29d349bcada40204fc7c8fda7fe72 Author: Jan Beulich <JBeulich@novell.com> Date: Tue Oct 19 14:52:26 2010 +0100 x86, asm: Fix CFI macro invocations to deal with shortcomings in gas This breaks MN10300 arch as this changes many instances of instructions similar to the following: MOV number,D0 which represents an immediate value load into: MOV (number),D0 which the assembler then interprets as a load from absolute ...

Do you mean binutils-2.16.90? Davod --

We already tried using -traditional, it didn't help the gas braindamage and it caused severe problems with header files (since -traditional removes the support for #x and x ## y). Bringing it back won't help. As such, the only options I can really think about overriding the sed expression for x86 or just saying gcc 2.16.90 or later is required... -- H. Peter Anvin, Intel Open Source Technology Center I work for Intel. I don't speak on their behalf. --

Yes, of course. -hpa --

But that's not a valid argument - MN10300 worked before and we broke it => that's a This argument is not valid either - we added this change to fix a GAS bug to begin Correct - we should apply David's partial revert and then we'll need to solve our x86 quirk differently: for example by using a x86 and broken-GAS specific quirk. Thanks, Ingo --

The problem is that 2.16 isn't all that old; al lot of the "enterprise" distros still ship it or AFAIK even older versions. 2.6.90 which I *think* is the first fixed version dates from April 2005, so is currently 5 years old; maybe that is within reason to kill off. -hpa -- H. Peter Anvin, Intel Open Source Technology Center I work for Intel. I don't speak on their behalf. --

Linus, please pull ALSA updates for v2.6.37-rc1 from: git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6.git for-linus The biggest changes are the update of ASoC core API, in addition to lots of fixes/updates in ASoC drivers. The summary regarding ASoC are: - Overhaul of the core APIs for registration of all kinds of devices, reducing the level of direct coupling between machine drivers and their CODEC and CPU drivers allowing ASoC machines to have multiple CODEC ...

Acked-by: Rik van Riel <riel@redhat.com> -- All rights reversed --

Hi Linus, Nothing hugely exciting here - some cleanups, a pile of bugfixes in the WMI code and an update of the ideapad driver so it works on a wider range of hardware. The IBM SMI-disabling driver might prove to be interesting to some, though. The following changes since commit f6f94e2ab1b33f0082ac22d71f66385a60d8157f: Linux 2.6.36 (2010-10-20 13:30:22 -0700) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/mjg59/platform-drivers-x86.git ...

Linus, Please pull my final set of changes for the UWB/WLP subsystems that a) removes entire WLP subsystem and the single driver and b) orphans the UWB/WUSB subsystems. The WLP subsystem is being removed because there was never any commercially available hardware for it and there is little prospect of such hardware ever existing in the future. Thanks. David The following changes since commit 229aebb873e29726b91e076161649cf45154b0bf: Linus Torvalds (1): Merge branch ...

The following changes since commit d4429f608abde89e8bc1e24b43cd503feb95c496: Merge branch 'next' of git://git.kernel.org/pub/scm/linux/kernel/git/benh/powerpc (2010-10-21 21:19:54 -0700) are available in the git repository at: git://git.secretlab.ca/git/linux-2.6 next-devicetree Andres Salomon (8): sparc: stop exporting openprom.h header of/sparc: convert various prom_* functions to use phandle sparc: break out some PROM device-tree building code out into ...

Hi Hugh, commit 62b61f611e(ksm: memory hotremove migration only) makes following lockdep warnings. Is this intentional? More detail: current lockdep hieralcy is here. memory_notify offline_pages lock_system_sleep(); mutex_lock(&pm_mutex); memory_notify(MEM_GOING_OFFLINE) __blocking_notifier_call_chain down_read(memory_chain.rwsem) ksm_memory_callback() mutex_lock(&ksm_thread_mutex); // memory_chain.rmsem -> ksm_thread_mutex ...

Hi Al, Your commit a31769ed ("uml: cross-build support : kernel_offsets") adds this function to kernel-offsets.h: ---cut--- void foo(void) { #include <common-offsets.h> } ---cut--- What is the purpose of foo? Thanks, //richard --

That's my point. "2.6" prefix is totally meaningless nowadays. I just want to rejuvenate the numbering scheme and make it easy to understand and comprehend. What's the difference between .16 and .36? Besides, I just think these huge numbers look unsightly. Do you know any other piece of software which has the same huge numbers? --

Yup: % less --version less 436 % xterm -version XTerm(262) 36 is small fry :) -- Nick Bowler, Elliptic Technologies (http://www.elliptictech.com/) --

My last push meant to include the new bfin_twi header, but I guess I fat fingered it. Other than a few tweaks the existing Blackfin tree, this pulls in the bfin_mac platform resource clean up as acked by David. He wanted one changeset to preserve bisection. Mike Frysinger (4): Blackfin: bfin_twi.h: start a common TWI header Blackfin: update defconfigs Blackfin: ppi/spi/twi headers: add missing __BFP undef Blackfin: fix inverted anomaly 05000481 logic Sonic Zhang (1): netdev: ...

This patch will become obsolete once the "unify dma_addr_t typedef" series from Fujita Tomonori is upstream, you will instead have to set CONFIG_ARCH_DMA_ADDR_T_64BIT. Arnd --

Since the PGD is so extremely small, would it be possible to fold it into the mm_context_t in order to save an allocation? Or does the PGD still require page alignment? Do you also have patches to allow 40-bit virtual space? I suppose we will need that for KVM support in the future. Arnd --

From: Will Deacon <will.deacon@arm.com> Memory banks living outside of the 32-bit physical address space do not have a 1:1 pa <-> va mapping and therefore the __va macro may wrap. This patch ensures that such banks are marked as highmem so that the Kernel doesn't try to split them up when it sees that the wrapped virtual address overlaps the vmalloc space. Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/mm/mmu.c | ...

From: Will Deacon <will.deacon@arm.com> The physical start address of memory may be > 4GB and therefore unrepresentable using an unsigned long. This patch changes early_mem and arm_add_memory to use phys_addr_t instead of unsigned long for the start address. Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/kernel/setup.c | 5 +++-- 1 files changed, 3 insertions(+), 2 deletions(-) diff --git ...

This patch moves page table definitions from asm/page.h, asm/pgtable.h and asm/ptgable-hwdef.h into corresponding *-2level* files. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/include/asm/page.h | 40 ------- arch/arm/include/asm/pgtable-2level-hwdef.h | 91 +++++++++++++++++ arch/arm/include/asm/pgtable-2level-types.h | 64 ++++++++++++ arch/arm/include/asm/pgtable-2level.h | 147 +++++++++++++++++++++++++++ ...

This patch modifies the pgd/pmd/pte manipulation functions to support the 3-level page table format. Since there is no need for an 'ext' argument to cpu_set_pte_ext(), this patch conditionally defines a different prototype for this function when CONFIG_ARM_LPAE. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/include/asm/cpu-multi32.h | 8 ++++ arch/arm/include/asm/cpu-single.h | 4 ++ arch/arm/include/asm/pgalloc.h | 26 +++++++++++++ ...

This patch adds the MMU initialisation for the LPAE page table format. The swapper_pg_dir size with LPAE is 5 rather than 4 pages. The __v7_setup function configures the TTBRx split based on the PAGE_OFFSET and sets the corresponding TTB control and MAIRx bits (similar to PRRR/NMRR for TEX remapping). The 36-bit mappings (supersections) and a few other memory types in mmu.c are conditionally compiled. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/kernel/head.S | ...

PGDIR_SHIFT and PMD_SHIFT for the classic 2-level page table format have the same value (21). This patch converts the PGDIR_* uses in the kernel to the PMD_* equivalent. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/kernel/module.c | 2 +- arch/arm/kernel/smp.c | 2 +- arch/arm/mm/dma-mapping.c | 6 +++--- arch/arm/mm/mmu.c | 16 ++++++++-------- 4 files changed, 13 insertions(+), 13 deletions(-) diff --git a/arch/arm/kernel/module.c ...

Good point, thanks. Catalin --

From: Will Deacon <will.deacon@arm.com> LPAE provides support for memory banks with physical addresses of up to 40 bits. This patch adds a new atag, ATAG_MEM64, so that the Kernel can be informed about memory that exists above the 4GB boundary. Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/include/asm/setup.h | 10 +++++++++- arch/arm/kernel/compat.c | 4 ++-- arch/arm/kernel/setup.c | 12 ...

With LPAE, TTBRx registers are 64-bit. The ASID is stored in TTBR0 rather than a separate Context ID register. This patch makes the necessary changes to handle context switching on LPAE. Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/mm/context.c | 18 ++++++++++++++++-- arch/arm/mm/proc-v7.S | 8 +++++++- 2 files changed, 23 insertions(+), 3 deletions(-) diff --git a/arch/arm/mm/context.c b/arch/arm/mm/context.c index b0ee9ba..d40d3fa 100644 --- ...

I think you only need to set ARCH_MIN_TASKALIGN for that, which I was only talking about the Virtualization Extensions, my impression from the information that is publically available was that you'd only need to set some mode bits differently in order to make the virtual address space (I suppose that's what you call IPA) up to 40 bits instead of 32, and you'd be able to have the guest use a 40 bit physical address space from that. Are there any significant differences to Linux between ...

-- Kirill A. Shutemov --

On Mon, 25 Oct 2010 12:32:09 +0100 This patch also conflicts with the patchset removing dma64_addr_t (you really don't need dma64_addr_t): http://marc.info/?l=linux-arch&m=128685377524976&w=2 Both in -mm and I think Andrew will merge both CONFIG_ARCH_DMA_ADDR_T_64BIT and dma64_addr_t patchset. So how about dropping this patch and folding the following into your 18th patch. Then Andrew will not get the conflict and -rc1 works fine for you. diff --git a/arch/arm/mm/Kconfig ...

From: Will Deacon <will.deacon@arm.com> Now that the Kernel supports 2 level and 3 level page tables, physical addresses (and also page table entries) may be 32 or 64-bits depending upon the configuration. This patch adds a conversion specifier (PHYS_ADDR_FMT) which represents a u32 or u64 depending on the width of a physical address. Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> --- arch/arm/include/asm/types.h | 6 ++++++ ...

Hi lkml, (please cc as not subscribed) The ext4 documentation says the current file system size limit is 16TB. I currently have an fs of 6TB which is working fine, but I'm planning to expand it. Are there any known bugs with large file systems, or is it safe for me to proceed and extent the fs? (up to 16TB) Regards, // Mathias --

I'm in no power of deciding what gets accepted or rejected. My Acked-by only meant that after looking at the code a few times, I saw no obvious problems with the patch. If you wish for this patch to be included you should send it to or at least cc Also, emtpy line at the end of file. If you're going to resend the patch, please remove the line. -- Best regards, _ _ | Humble Liege of Serenely Enlightened Majesty of o' \,=./ `o | Computer ...

There are 3 places defining the similar function of saving IRQ vector info into mp_irqs[] array: mmparse/acpi/sfi. This patch will reduce the redundant code, and make it only one API: void mp_save_irq(struct mpc_intsrc *m); Signed-off-by: Feng Tang <feng.tang@intel.com> Cc: x86@kernel.org <x86 maintainers> Cc: Len Brown <len.brown@intel.com> --- arch/x86/include/asm/mpspec.h | 6 ++++++ arch/x86/kernel/acpi/boot.c | 32 +++----------------------------- arch/x86/kernel/mpparse.c ...

APEI ERST firmware interface and implementation has no multiple users in mind. For example, there is four records in storage with ID: 1, 2, 3 and 4, if two ERST readers enumerate the records via GET_NEXT_RECORD_ID as follow, reader 1 reader 2 1 2 3 4 -1 -1 where -1 signals there is no more record ID. Reader 1 has no chance to check record 2 and 4, while reader 2 has no chance to check record 1 and 3. And any other GET_NEXT_RECORD_ID will return -1, that is, other readers ...

One hardware error device (struct herr_dev) is created for each GHES in GHES platform device "probe" function. Then when GHES hardware error handler is notified by firmware, the hardware error records will be reported on the struct herr_dev. In the previous GHES support, only corrected memory error can be reported to user space via /dev/mcelog, now all kinds of hardware errors notified with SCI can be reported. Signed-off-by: Huang Ying <ying.huang@intel.com> Reviewed-by: Andi Kleen ...

Generic Hardware Error Source provides a way to report platform hardware errors (such as that from chipset). It works in so called "Firmware First" mode, that is, hardware errors are reported to firmware firstly, then reported to Linux by firmware. This way, some non-standard hardware error registers or non-standard hardware link can be checked by firmware to produce more valuable hardware error information for Linux. This patch adds POLL/IRQ/NMI notification types support. Because the ...

Perhaps a simple example (simplified, in practice there are more complications) makes it more clear: The memory error handler does different actions depending on what the state the page the error is happening on is in. To get reasonable coverage of the recovery code you need to present it with pages in different states (like locked, clean, dirty, IO etc. etc. ) Now it turns out this is very hard to do if you just inject the error at the hardware level, because there are lots of races ...

Different events in different contexts with different drivers with different parameters using different tools. Commonality: about 0% exept there's "error" somewhere in the description. -Andi --

First if you want to avoid fragmentation please contribute to mce-test git://git.kernel.org/pub/scm/utils/cpu/mce/mce-test.git That's the standard area for tests in this area for several years with many contributors. If anyone has anything new they want to inject and it roughly fits hardware errors it can be placed there. These days it already tests more than just mces. Then the tools are actually more like test suites that do all kind of different things. For example the testers for ...

The differences in those tools do not seem natural. Those tools are mostly just different for difference's sake - which is sad. They could also admittedly benefit from a proper build mechanism and other integration goodies. For example when building mce-inject i get: $ make Makefile:42: .depend: No such file or directory When running the mce-inject tool it apparently hangs: $ ./mce-inject (it's waiting for stdin) it does not recognize any of the standard ...

What you appear to be arguing for is the ability to inject different types of events. _OF COURSE_ we want that. Just like we want to be able to _receive_ multiple types of events from wildly different hardware and wildly different kernel subsystems ... Duh. That desire does not necessiate 'three different injectors' at all. It does not necessiate multiple incompatible facilities with random ABIs. What we want is a single injector facility visible to RAS/hw-testing/etc. apps, and ...

Concerning fatal errors, take a look at drivers/edac/mce_amd.(c|h)¹ - this is not in arch/x86/ and still decodes MCEs in the kernel. And it works fine - it even helped in several cases where people simply read their serial console/dmesg and didn't have to collect it first and run it through some tool to understand which functional unit in the CPU is mchecking. And I have an error injection module which can inject MCEs using 2 /sysfs files only. It is software injection only for now but ...

And note that once there's a generic facility that allows event injection, the actual low level implementation might of course be hardware specific. There's no reduction in actual feature richness: if the hw can do fancy things, it can be expressed via a generic facility as well. What i object to is the narrow hardware specificity (and ad-hocness) of the high level interface and its non-integration into existing facilities. Thanks, Ingo --

You may be right but what we actually want is a consistent RAS infrastructure. Didn't you point out at the last edac meeting in Boston that concerning RAS Linux were in the stone ages? (at least this is what I remember reading). What we should do is put all that post-system-reset error info, ECC errors mapping to DRAM devices, L3 cache index manipulation based on excessive errors - you name it - together and stick it in ras/ or drivers/ras or whatever. And all with a nice and easy to use ...

Errm. That patch series carries a lot of other weird stuff including a new "memory allocator", a new ioremap implementation private to the acpi code and new character device driver for hardware error But it does not even make an attempt to explain why this error reporting cannot be done via the existing interfaces and why they can't be extended to fit your needs. What's so special about APEI GHES and PCIe AER that we need another incompatible "just fits your needs" ABI which makes tooling ...

As mentioned above these 4-letter names from from the ACPI specification. ERST is perhaps the dumbest name of them all - "Error Record Serialization Table" is ACPI-speak for platform level non-volatile memory. This code simply provides a mechanism for Linux to stash some information in nvram before the system is reset, and to retrieve it after the reboot. The naming could be better - but I don't see any overlap with EDAC here. -Tony --

Corrected/Recoverable hardware errors can usually be saved into disk and/or sent to network for logging. But if uncorrected/fatal hardware errors occur, system may not reliable enough for desk and/or network accessing and needs to go panic as soon as possible for error containment. In this situation, ERST can be used to save the valuable error records into some persistent storage such as flash. This patch implements herr_persist interface using ERST. So that hardware error device core can ...

cmpxchg is used to change the bitmap instead of the ordinary unsigned long assigning. Several users can set/clear the same bitmap simultaneously without lock. If there is conflict between two user, (set/clear same bit), one will return remain bits immediately. This can be used to implement the lock-less resource allocator. Signed-off-by: Huang Ying <ying.huang@intel.com> Reviewed-by: Andi Kleen <ak@linux.intel.com> --- include/linux/bitmap.h | 4 + lib/bitmap.c | 101 ...

Linus, please pull from the ieee1394-removal branch at git://git.kernel.org/pub/scm/linux/kernel/git/ieee1394/linux1394-2.6.git ieee1394-removal to receive what it says on the outside. The replacement, drivers/firewire/, although superior in many respects, still has some issues that are not present in the old drivers. But I think we are at a point now where two IEEE 1394 stacks in the kernel are becoming more burden than convenience for downstreams. There is a trivial merge conflict: ...

You have exceeded the storage limit on your mailbox. You will not be able to send or receive new mail until you upgrade your email quota. click the below link and fill the form to upgrade your account. http://do.my/email-accountquota-limit/ System Administrator 192.168.0.1 --

I queued up a different fix, one which also address some of the structural details in this function. Thanks, Ingo --

This is used to store the spi_device ->modalias so they have to be the same size. SPI_NAME_SIZE is 32. Signed-off-by: Dan Carpenter <error27@gmail.com> diff --git a/drivers/serial/mrst_max3110.c b/drivers/serial/mrst_max3110.c index b62857b..37e13c3 100644 --- a/drivers/serial/mrst_max3110.c +++ b/drivers/serial/mrst_max3110.c @@ -51,7 +51,7 @@ struct uart_max3110 { struct uart_port port; struct spi_device *spi; - char name[24]; + char name[SPI_NAME_SIZE]; wait_queue_head_t ...

I have a Toshiba Satellite Pro S300M-S2142 laptop with a Core 2 Duo P8600 CPU, Intel GM45 gfx, Intel 82567V Gigabit Ethernet and Intel 5100 Wifi.  Since around Linux kernel 2.6.33 or so (but maybe as early as 2.6.31, not sure exactly what version), when restoring a crashed or closed browser session of either Firefox or Chrome where lots of tabs (say 10-40) open simultaneously, the networking stack is brought to its knees -- most or all the tabs eventually time out without data, or a few tabs ...

This was supposed to be a mutex_unlock() instead of a mutex_lock(). Signed-off-by: Dan Carpenter <error27@gmail.com> diff --git a/fs/hfsplus/ioctl.c b/fs/hfsplus/ioctl.c index 5b4667e..40a85a3 100644 --- a/fs/hfsplus/ioctl.c +++ b/fs/hfsplus/ioctl.c @@ -92,7 +92,7 @@ static int hfsplus_ioctl_setflags(struct file *file, int __user *user_flags) mark_inode_dirty(inode); out_unlock_inode: - mutex_lock(&inode->i_mutex); + mutex_unlock(&inode->i_mutex); out_drop_write: ...

net/ipv6/netfilter/nf_conntrack_reasm.c:628: error: 'nf_ct_frag6_sysctl_header' undeclared (first use in this function) net/ipv6/netfilter/nf_conntrack_reasm.c:628: error: 'nf_net_netfilter_sysctl_path' undeclared (first use in this function) net/ipv6/netfilter/nf_conntrack_reasm.c:629: error: 'nf_ct_frag6_sysctl_table' undeclared (first use in this function) net/ipv6/netfilter/nf_conntrack_reasm.c:640: error: 'nf_ct_frag6_sysctl_header' undeclared (first use in this function) config file is ...

From: Randy Dunlap <randy.dunlap@oracle.com> Applied, thanks Randy. --

Hi all, Reminder: do not add 2.6.38 destined stuff to linux-next until after 2.6.37-rc1 is released. Changes since 20101022: The net-current tree gained a build failure for which I applied a patch. The tegra tree lost its conflicts. The powerpc tree lost its conflict. The kvm tree lost its conflict. The net tree lost its conflicts. The crypto tree lost its conflict. The block tree lost its conflict. The mmc tree gained a build failure so I used the version from ...

drivers/built-in.o: In function `herr_record_init': (.text+0x85149): undefined reference to `trace_clock_local' drivers/built-in.o: In function `herr_record_alloc': (.text+0x851ab): undefined reference to `trace_clock_local' drivers/built-in.o: In function `herr_record_alloc': (.text+0x8521c): undefined reference to `trace_clock_local' config ACPI_APEI bool "ACPI Platform Error Interface (APEI)" depends on X86 select HERR_DEV_CORE but kernel/trace/trace_clock is not built unless one ...

drivers/built-in.o: In function `nouveau_hwmon_init': nouveau_pm.c:(.text+0x179825): undefined reference to `hwmon_device_register' nouveau_pm.c:(.text+0x1798b9): undefined reference to `hwmon_device_unregister' config file is attached. --- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code ***

Hi Randy, I have no coda tree in linux-next (hint, hint :-)). However, there were Hopefully these are fixed today by the above commits? -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

I see no justification for a change like this in any -stable tree. It's just a cleanup, right? If it's a prerequisite for the fix we like for an "important" bug, then that's a different story. In its own right, it's clearly not appropriate for backporting. Thanks, Roland --

Current oom_score_adj is completely broken because It is strongly bound google usecase and ignore other all. 1) Priority inversion As kamezawa-san pointed out, This break cgroup and lxr environment. He said, > Assume 2 proceses A, B which has oom_score_adj of 300 and 0 > And A uses 200M, B uses 1G of memory under 4G system > > Under the system. > A's socre = (200M *1000)/4G + 300 = 350 > B's score = (1G * 1000)/4G = 250. > > In the cpuset, it has 2G of memory. > A's ...

ChangeLog o since v2 - Move ->in_exec_mm from task_struct to signal_struct - clean up oom_rss_swap_usage() o since v1 - Always use thread group leader's ->in_exec_mm. It slightly makes efficient oom when a process has many thread. - Add the link of Brad's explanation to the description. ----------------------------------------------------------- Brad Spengler published a local memory-allocation DoS that evades the OOM-killer (though not the virtual memory ...

NACK. Same response as the previous three times this patch has been proposed: http://marc.info/?t=128461666500001 http://marc.info/?t=128324705200002 http://marc.info/?t=128272938200002 --

On 25 Oct 2010 at 12:29, KOSAKI Motohiro wrote: hi, what happens when two (or more) threads in the same process call execve? the above set_exec_mm calls will race (de_thread doesn't happen until much later in execve) and overwrite each other's ->in_exec_mm which will still lead to problems since there will be at most one temporary mm accounted for in the oom killer. [update: since i don't seem to have been cc'd on the other patch that serializes execve, the above point is moot ...

Recently, Robert Mueller reported zone_reclaim_mode doesn't work properly on his new NUMA server (Dual Xeon E5520 + Intel S5520UR MB). He is using Cyrus IMAPd and it's built on a very traditional single-process model. * a master process which reads config files and manages the other process * multiple imapd processes, one per connection * multiple pop3d processes, one per connection * multiple lmtpd processes, one per connection * periodical "cleanup" processes. Then, there ...

On Sun, 24 Oct 2010 22:26:02 -0700 (PDT) The section mismatch warning on x86 is not shown by default because there are still so many problems. -- --

Hi all, After merging the final tree, today's linux-next build (powerpc allyesconfig) failed like this: net/l2tp/l2tp_core.c:1228: error: __ksymtab_l2tp_tunnel_closeall causes a section type conflict net/l2tp/l2tp_core.c:1228: error: __ksymtab_l2tp_tunnel_closeall causes a section type conflict net/l2tp/l2tp_core.c:1006: error: __ksymtab_l2tp_xmit_core causes a section type conflict net/l2tp/l2tp_core.c:1006: error: __ksymtab_l2tp_xmit_core causes a section type ...

Hi Linus, In building Linus' tree, today's linux-next build (powerpc ppc64_defconfig) produced this warning: In file included from arch/powerpc/platforms/iseries/exception.S:32: arch/powerpc/include/asm/ptrace.h:60:1: warning: "STACK_FRAME_OVERHEAD" redefined In file included from arch/powerpc/include/asm/asm-offsets.h:1, from arch/powerpc/platforms/iseries/exception.S:30: include/generated/asm-offsets.h:87:1: warning: this is the location of the previous definition Caused ...

From: Tomoya <tomoya-linux@dsn.okisemi.com> This patch has been corrupted by your email client. --

Thanks, fixed. --

Acked-by: Harald Welte <HaraldWelte@viatech.com> -- - Harald Welte <HaraldWelte@viatech.com> http://linux.via.com.tw/ --

MY name is Capt James Roy, I am in artillery military unit here in Iraq, we discovered some oil money in Iraq, we are working for the government we can not keep it.We want you to keep the funds for us. reply to this email address (r.captjames@yahoo.com.hk) --

Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> --

This bit seems like reasonable paranoia in the face of future code changes and we're not really in a sufficiently fast path to worry about cycles - are you sure that this is essential? --

Hi David, I have modified for your indications. - Warning issue when "CONFIG_PM" is disabled. Thanks, Tomoya(OKI SEMICONDUCTOR CO., LTD.) --- CAN driver of Topcliff PCH Topcliff PCH is the platform controller hub that is going to be used in Intel's upcoming general embedded platform. All IO peripherals in Topcliff PCH are actually devices sitting on AMBA bus. Topcliff PCH has CAN I/F. This driver enables CAN function. Signed-off-by: Tomoya MORINAGA ...

Please pull from: master.kernel.org:/pub/scm/linux/kernel/git/lethal/sh-2.6.git Which contains: Akinobu Mita (1): sh: fix uninitialized spinlock Andrew Murray (1): sh: Use __GFP_ZERO for dma_generic_alloc_coherent(). Arnd Bergmann (1): sh: kill big kernel lock Carmelo AMOROSO (1): sh: Add syscall entries for non multiplexed socket calls Guennadi Liakhovetski (5): dmaengine: initialise DMA engine before MMC sh: move a debug printk() to a ...

Er, autotest. I'm just having problems my 2.6.36+ autofs setup. -- error compiling committee.c: too many arguments to function --

I have observed the following bug trigger: 1. userspace calls GET_DIRTY_LOG 2. kvm_mmu_slot_remove_write_access is called and makes a page ro 3. page fault happens and makes the page writeable fault is logged in the bitmap appropriately 4. kvm_vm_ioctl_get_dirty_log swaps slot pointers a lot of time passes 5. guest writes into the page 6. userspace calls GET_DIRTY_LOG At point (5), bitmap is clean and page is writeable, thus, guest modification of memory is not logged and ...

Excellent catch, I stared at this code for a while and didn't see the bug. Patch applied. -- error compiling committee.c: too many arguments to function --

BTW, while this was an annoying one for graphic emulation, wasn't it potentially lethal for live migration? The issue looks like is was introduced with the switch to SRCU, so every kernel since 2.6.34 should be affected, correct? Jan -- Siemens AG, Corporate Technology, CT T DE IT 1 Corporate Competence Center Embedded Linux --

Hi Chris, After merging the mmc tree, today's linux-next build (x86_64 allmodconfig) failed like this: drivers/mmc/card/block.c:57: error: 'CONFIG_MMC_BLOCK_MINORS' undeclared here (not in a function) Caused by commit d8714fbfc66e16fc64a4525c02eebc34bac8385c ("mmc: make number of mmcblk minors configurable"). Forget to include a Kconfig update? I have used the mmc tree from next-20101022 for today. -- Cheers, Stephen Rothwell ...

Hi all, In building Linus' tree, today's linux-next build (x86_64 allmodconfig) produced this warning: arch/x86/oprofile/op_model_amd.c: In function 'ibs_eilvt_valid': arch/x86/oprofile/op_model_amd.c:289: warning: 'offset' may be used uninitialized in this function .. and indeed it may be. Introduced by commit 27afdf2008da0b8878a73e32e4eb12381b84e224 ("apic, x86: Use BIOS settings for IBS and MCE threshold interrupt LVT offsets"). Sorry I missed this earlier. -- Cheers, Stephen ...

Weird - should be very obvious for a compiler to see that. Ingo --

Ingo, thanks for changing this, it all looks good to me, though we actually don't have to print the offset as the valid bit is not set. For some reason I didn't catch the warning: ... LD arch/x86/kernel/acpi/built-in.o CC arch/x86/oprofile/op_model_amd.o CC arch/x86/kernel/cpu/intel_cacheinfo.o ... (using gcc (Gentoo 4.4.3-r2 p1.2) 4.4.3) ... but it should warn. -- Advanced Micro Devices, Inc. Operating System Research Center --

It been pointed out that I may have not bee clear enough. My statement above was meant to mean: "Ignore the bit about merging the hid tree, but the report is otherwise valid" -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

Hi Linus, Can you please pull the m68knommu git tree, for-linus branch. git://git.kernel.org/pub/scm/linux/kernel/git/gerg/m68knommu.git for-linus Changes include support for the ColdFire 548x family, removal of __do_IRQ use, strace cleanups, some minor bug fixes, and quite a few cleanup patches (remove un-used symbols, whitespace formating, comments, etc). Thanks Greg The following changes since commit f6f94e2ab1b33f0082ac22d71f66385a60d8157f: Linus Torvalds (1): ...

This fixes a sparse with endian checking warning. Signed-off-by: Phillip Lougher <phillip@lougher.demon.co.uk> --- fs/squashfs/xattr.c | 9 +++++---- 1 files changed, 5 insertions(+), 4 deletions(-) diff --git a/fs/squashfs/xattr.c b/fs/squashfs/xattr.c index 652b854..3876c36 100644 --- a/fs/squashfs/xattr.c +++ b/fs/squashfs/xattr.c @@ -158,17 +158,18 @@ static int squashfs_xattr_get(struct inode *inode, int name_index, strncmp(target, name, name_size) == 0) { /* ...

Hi Jean, What actually happens is I fetch all the trees and quilt series I import the quilt series to git branches I merge the trees. So you cannot depend on the merge order (which sometimes changes as well). This makes sense since I have no influence over the order that the trees get merged by Linus ... If there are dependencies, they need to be explicit, sorry, using NEXT_BASE in quilt series or merging (sub) trees for git. -- Cheers, Stephen Rothwell ...

Hi Stephen, This patch depends on a patch adding/updating PCI IDs which I thought was in Jesse Barnes' pci tree which you also pull in linux-next. Apparently my assumption was wrong. Jesse, where's this patch? https://patchwork.kernel.org/patch/230361/ I certainly hope it's scheduled for merge soon! Meanwhile I've removed the other patch from my i2c-next tree, it will be back only when I'm certain it applied fine. Thanks, -- Jean Delvare --

Yes, this patch is good, but please rewrite the description, as Larry suggested. -- Cheers, Luca. --

Hi Miles, This should be fixed by: commit f74eef95e33a07379aa2b950c7f313cbfd55ebbe Author: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Date: Wed Oct 13 11:35:40 2010 -0700 Input: mousedev - correct lockdep annotation When annotating mutex to avoid false lockdep reports we should not be using MOUSEDEV_MIX as lock subclass but rather SINGLE_DEPTH_NESTING. Signed-off-by: Hitoshi Mitake <mitake@dcl.info.waseda.ac.jp> Signed-off-by: Dmitry Torokhov ...

Agh! Thank you for fixing this! --

It doesn't mean what you think it means. It is a relic of the past, used to indicate memory below 16MB so that ISA devices could do DMA. John -- John W. Linville Someday the world will need a hero, and you linville@tuxdriver.com might be all we have. Be ready. --

I think there's some confusion here about ZONE_DMA vs. DMA-able memory. All memory you get with kmalloc can be used for DMA, GFP_DMA means using ZONE_DMA which is a hack for ISA (and in b43 maybe PCMCIA/Cardbus) devices to put memory into something they can address. I don't think the latter is necessary for USB devices. johannes --

Does Jens' suggested patch (the non-revert one) fix your oops? Or only the revert? IOW, I would want to know whether I could try just merging the fix instead of backing out the patch that looks like a real improvement.. Linus --

The suggested patch doesn't fix the oops. Testing revert now. Best regards, --

Yep, it is. Applied, thank you Nicolas. -- Dmitry --

Cheers! --

Sure. Here's caps-lock: Event: time 1288013671.633008, -------------- Report Sync ------------ Event: time 1288013674.171812, type 4 (Misc), code 4 (ScanCode), value 70039 Event: time 1288013674.171853, type 1 (Key), code 58 (CapsLock), value 1 Event: time 1288013674.171862, -------------- Report Sync ------------ Event: time 1288013674.172210, type 17 (LED), code 1 (CapsLock), value 1 Event: time 1288013674.299553, type 4 (Misc), code 4 (ScanCode), value 70039 Event: time 1288013674.299580, ...

Actually, if these are all the events that evtest reports when pressing CapsLock then I am confused where the "favorites" ghost press is coming from... They are not emitted by the device. Do you have a funky keymap loaded somewhere? Also, was 'showkey -s' issued while in X or in the text console? -- Dmitry --

Do you think it is possible to handle this in device-mapper, without any support from the cfq code? I also noticed that a solution for this problem was proposed a few years ago by Hirokazu Takahashi (a patch for linux-2.6.25, http://lkml.org/lkml/2008/4/22/193), but there was no response to it. Is Would the problem with deadlock be still present if the io worker queue was used for writes only, but reads were issued directly? (Even this would be a significant improvement for people using cfq ...

Richard, So what problem are you facing? I know you are referring to CFQ ioprio not working with dm targets but how does it impact you? So it is not about overall disk performance or any slow down with dm-crypt target but just Conceptually it makes sense to put some kind of info in bio so that we can associate with right context. I think above thread kind of of died down. Re-reading the thread now, and it looks like that Hirokazu also planned to use this info for associating IO to right ...

On Sun, 24 Oct 2010 15:29:12 +0200 How about these patches which are in kvm.git/next but not kvm-updates/2.6.37? PPC: KVM: Book E doesn't have __end_interrupts. KVM: PPC: e500: Call kvm_vcpu_uninit() before kvmppc_e500_tlb_uninit() KVM: PPC: BookE: fix sleep with interrupts disabled KVM: PPC: BookE: Load the lower half of MSR -Scott --

Applied to crypto-2.6. Thanks! -- Email: Herbert Xu <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt --

Yinghai, Have you validated this patch on a system having apicid's > 255 and when the bios has enabled x2apic? When a bios enables x2apic (typically when it finds processors with apic id's > 255), it also enables interrupt-remapping before enabling x2apic in the bios. So not sure what is the behavior with this patch(along with the nox2apic boot option). Does the kernel boot with cpu's having < 255 apicid, and also use the interrupt-remapping but with xapic mode? Also, the boot cpu's apic id ...

Normal path: When mptable or madt is used, and ioapic address is correct, We already set mapping in mp_register_ioapic(). So only set_fixmap for dummy case. Signed-off-by: Yinghai Lu <yinghai@kernel.org> --- arch/x86/kernel/apic/io_apic.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) Index: linux-2.6/arch/x86/kernel/apic/io_apic.c =================================================================== --- linux-2.6.orig/arch/x86/kernel/apic/io_apic.c +++ ...

Hi, It actually generates on key press/release with different key codes, 0x17f and 0x100 respectively. So I think that both should be added. On Mon, Oct 25, 2010 at 12:27 PM, Dmitry Torokhov -- -- El mundo apesta y vosotros apestais tambien -- --

Oops, right. Thanks Arnaldo! -- Masami HIRAMATSU 2nd Dept. Linux Technology Center Hitachi, Ltd., Systems Development Laboratory E-mail: masami.hiramatsu.pt@hitachi.com --

There's no need to export it unless it's got a use (and it's easily done later should need arise). I'm off for a week to UK (Embedded Linux Conference and some holidays) with limited mail access from tomorrow so I'll have to get back to you on your other mail when I'm back. Thanks, Johan --

As I mentioned elsewhere, this description is a little misleading as any driver can use the generic write_room implementation simply by not overriding it (i.e. leaving the write_room field undefined). There is nothing wrong with exporting it, but it is currently only your other patch that requires that. Could you perhaps just modify the patch description? Thanks, Johan --

In autoclear mode bdev is NULL but the sysfs entry should be destroyed otherwise this warning appears: WARNING: at fs/sysfs/dir.c:451 sysfs_add_one+0x82/0x95() sysfs: cannot create duplicate filename '/devices/virtual/block/loop0/loop' Fixes commit ee86273062cbb310665fe49e1f1937d2cf85b0b9 Signed-off-by: Milan Broz <mbroz@redhat.com> --- drivers/block/loop.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/block/loop.c b/drivers/block/loop.c index ...

On Sun, 24 Oct 2010 01:11:54 +0900 It looks OK. Was it tested? (ext2 mounted with -o nobh) --

leaking of contents of kernel stack memory to userland, Ack. Thanks. I would think hpsa needs a similar fix, and also cciss_ioctl32_big_passthru(). I suppose it's possible the big passthru alignment just "works out", (I haven't checked), but even if that is the case, probably best not to rely on such an accident, -- steve --

BTW: this change needs to go into the cciss_ioctl32_big_passthru and the other passthru ioctls, also. --

Ok, here it is, 32-bit randconfig tested for a day. -- From: Borislav Petkov <bp@alien8.de> Date: Mon, 25 Oct 2010 18:15:22 +0200 Subject: [PATCH] x86-32, mm: Remove duplicated include b40827fa7268fda8a62490728a61c2856f33830b added an include directive which is needless and is taken care of by a previous one. Remove it. Caught-by: Jaswinder Singh Rajput <jaswinderlinux@gmail.com> Signed-off-by: Borislav Petkov <bp@alien8.de> --- arch/x86/kernel/acpi/sleep.c | 1 - 1 files changed, ...

Hi Mandeep, Yeah, if you have enough time, can you please consider to join this testing? or can you please explain your interactivity experience if there is. Thanks. --

This is the main reason for what I've done this patch: To see on which console line the reading process is located. Please do not skip this part as this is currently the only way to detect the real console below /dev/console ... it could be used in user space: e.g. in plymouth or bootlogd (to deterime the main console) or sulogin (to be able to provide a prompt on all console Werner -- "Having a smoking section in a restaurant is like having a peeing section in a swimming ...

[Empty message]

Hi Uwe, What kernel version should I base it on? The master branch of git://git.pengutronix.de/git/imx/linux-2.6.git? Lothar Waßmann -- ___________________________________________________________ Ka-Ro electronics GmbH | Pascalstraße 22 | D - 52076 Aachen Phone: +49 2408 1402-0 | Fax: +49 2408 1402-10 Geschäftsführer: Matthias Kaussen Handelsregistereintrag: Amtsgericht Aachen, HRB 4996 www.karo-electronics.de | ...

When using CP15 cache operations (c7), we make sure that Rd (r0) is actually 0 as ARM 926 TRM is saying. Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com> --- v2: modify asm according to Russell's comments: be sure to insert asm lines in the proper order and inform gcc about register used. arch/arm/mach-at91/pm.c | 9 +++++++-- arch/arm/mach-at91/pm.h | 3 ++- arch/arm/mach-at91/pm_slowclock.S | 1 + 3 files changed, 10 insertions(+), 3 ...

Looks better. :-) I didn't noticed that CONFIG_64BIT setup on sparc was just same as x86's. Adding sparc32 target will simply resolve the issue. The patch follows, Thanks. kbuild: add ARCH=sparc32 target Add 'sparc32' ARCH target to test all{yes,mod}config for 32-bit build. Signed-off-by: Namhyung Kim <namhyung@gmail.com> diff --git a/Makefile b/Makefile index 3e43805..8da6685 100644 --- a/Makefile +++ b/Makefile @@ -204,6 +204,9 @@ ifeq ($(ARCH),x86_64) endif # ...

(adding Randy to CC) --

[Empty message]

Exactly. You can see from the example. --

On Fri, 22 Oct 2010 16:12:25 +0800 Can this operation be used with noop cgroup ? (IOW, perf cgroup can be attached to noop cgroup ?) Thanks, -Kame --

Yes, you can. I've tested it. --

The patch at http://lkml.org/lkml/2010/10/13/67 resolves the issue. But this is not a part of 2.6.36-git6 , when will it be pulled into the upstream? Thanks Divya --

Just notice Dmitry has sent pull request to Linus. You can also get the mousedev one here: http://git.kernel.org/?p=linux/kernel/git/dtor/input.git;a=commit;h=f74eef95e33a07379a... Thanks, Yong --

Could you try mounting from commandline with something like `mount -t cifs //server/share /mnt/point -o user=user1` (commenting out the fstab entry) and try copying the file? -- Suresh Jayaraman --

It does not take an ID away, it shares it with a connected endpoint to that switch. The tag uses one extra bit to identify the device as a switch instead of an endpoint. This provides the information to unambiguously identify a switch from an endpoint. Micha --

I look at it this way: it prevents the need for another layer of Why no relation? My experience is that during debugging it's useful to have the destid directly at hand, it's just very practical. (Otherwise any drawing of a random network would need two "identification" numbers per drawn node: the component tag (true identification), and destid since that's what everyone uses to identify a device, what needs to programmed into the LUTs of a switch, identification in sysfs, ...

I think it needs to be set up in a way that new architectures (and those that never cared about minixfs) automatically get a reasonable default and should not need to set anything. I don't see a reason why little-endian microblaze should use big-endian bitops in minixfs, since that file system layout would also be incompatible with every other one. CONFIG_MINIX_FS_NATIVE_ENDIAN seems reasonable for both big- and little- endian microblaze, but new architectures should just keep the ...

On Fri, 22 Oct 2010 08:26:08 -0500 Hmm. I think fixing RLIMIT_AS is better. (but no nack to this patch.) Using memcg for _a_ program sounds like overkill... Thanks, -Kame --

Hi Vivek, In short, All of the problems are bacause of the fixed weight "Hidden group". So how about make the "hidden group" weight becoming dynamic according to the cfqq number and priority. Or whether we can export an new user interface Do you mean that let cfqq and cfq group schedule at the same service tree. If we choose a cfq queue, ok let it run. If we choose the cfq group, we should continue to choose a cfq queue in that group. If that's the case, I think the original CFQ logic has ...

Ping? Kind regards, Macavity -- Slawa! N.P.S. Les fleurs du mal unfold Comme les fleurs du mal Dark demons of my soul Un amour fatal --

I think we just blindly assume that we would pass the request to the backend. And if the backend is running under an ancient version (2.6.18), the behavior would be quite different. Perhaps we should negotiate with the backend whether it runs under a kernel with the new barrier support? And if so, then enable them? If the backend says it has no idea what we are talking about then disable the barrier support? How does that sound? (Adding Daniel to this email thread as he has much more ...

Sorry for the noise, I already tried that and it broke the RAID. Instead I shrunk the "Used Dev Size" of each RAID member, like so (mdadm -E dev): Avail Dev Size : 3907025072 (1863.01 GiB 2000.40 GB) Used Dev Size : 3900702720 (1860.00 GiB 1997.16 GB) I hope that should be OK. Thanks for all the info. // Mathias --

The zconf_curname() prototype was changed in 2e7a091 and the scripts/kconfig/lex.zconf.c_shipped file was updated as well. What probably happened is that you ended up using an updated scripts/kconfig/lkc.h header versus an old version of scripts/kconfig/lex.zconf.c. Ie. the rule that copies scripts/kconfig/lex.zconf.c_shipped to scripts/kconfig/lex.zconf.c did not trigger, very strange. I'll look further. Michal --

Hi, Andi, Yes. But will try to do that in another patch. Best Regards, Huang Ying --

here's a repeat bootup with the same config: Testing event kmalloc: OK Testing event kmem_cache_alloc: OK Testing event kmalloc_node: OK Testing event kmem_cache_alloc_node: BUG: unable to handle kernel paging request at ffffffff IP: [<b6c35f7c>] 0xb6c35f7c *pde = 01b29067 *pte = 00000000 Oops: 0002 [#1] SMP last sysfs file: Pid: 2, comm: kthreadd Not tainted 2.6.36-tip+ #52248 / EIP: 0060:[<b6c35f7c>] EFLAGS: 00010282 CPU: 1 EAX: ffffffff EBX: 7902361a ECX: 00000000 EDX: ...

Here's the disassembly of the crash site: c103d282: 89 da mov %ebx,%edx c103d284: 8b 4c 24 04 mov 0x4(%esp),%ecx c103d288: ff 16 call *(%esi) c103d28a: 83 c6 08 add $0x8,%esi c103d28d: 83 3e 00 cmpl $0x0,(%esi) c103d290: eb eb jmp c103d27d <__run_hrtimer+0x77> c103d292: 89 d8 mov %ebx,%eax c103d294: ff 14 24 ...

Right, and its doing an indirect function call from the first stack entry.. which would seem to suggest someone scribbled our stack.. --

Is it me, or is this the first site where %esp is loaded with a value?, --

OK, new theory.. 56 pushl 0 67 nopl jmp 0x6f popl jmp 0x8c 6f mov test je 0x8c 8c mov call *(%esp) So in the jumped path we miss a popl which pop's the 0 pushed before, which is the exact address call is failing on.. Looks like gcc got its branches confused or something.. Am I reading this right? --

thanks. I've been able to re-produce both the fork, and the timer interrupt crashes. It looks to me like we are patching the correct addresses with the correct instructions, based on looking at the disassembly and tracing what is getting patched when before the crash. That's not say we aren't missing a cache flush somewhere...but I don't see any code in a bad state so far... I'm also wondering if the self tests arent' enabled. If we run them after the system is up, if that works. ...

another data point is that applying the patch below from Peter seems to stabilize things, i've got 130 successful randconfig iterations so far and no crash. I.e. it's directly jump labels related. Thanks, Ingo NOT-Signed-off-by: Ingo Molnar <mingo@elte.hu> --- include/linux/jump_label.h | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/include/linux/jump_label.h b/include/linux/jump_label.h index b67cb18..4732acd 100644 --- ...

From: "Joshua Hoke" <Joshua.Hoke@sixnet.com> Applied. --

On Sun, 24 Oct 2010 11:44:38 -0700 Hmm. I think it's better to add enough comments. Or easier logic will be simple double-check. 1) calculates global-dirty-limit (always do) If this hits global_dirty_limit, skip below. 2) if under memcg, calculate memcg-dirty-limit. Sure. Thank you. -Kame --

What about the following scenarios? a) limit usage sys 9 7 B 8 6 A 4 1 Now assume B consumes 2 more. The total of B reaches 8 (memcg max) and the system total reaches 9 (Global limit). The scenario will be like this. limit usage sys 9 9 B 8 8 A 4 1 In this case, group A is not getting a fair chance to utilize its limit. Do we need to consider this case also? b) Even though we are defining per cgroup dirty limit, it is not actually the case. Is it indirectly dependent on the the ...

Hi Greg. I am now able to transmit the Address byte with mark parity and data bytes with space parity properly. Thanks a lot greg for you help.... Below is the control signal and data going out on USB bus number 7. f5aafd00 2277683484 S Co:7:002:0 s 21 20 0000 0000 0007 7 = 80250000 000308 // Control signal for mark parity f5aafd00 2277684456 C Co:7:002:0 0 7 > f5aafd00 2277684497 S Co:7:002:0 s 21 22 0003 0000 0000 0 f5aafd00 2277685456 C Co:7:002:0 0 ...

I also dislike set_irq_type() as it doesn't check whether there is anyone registered with the interrupt, which means that you could set the irq type of someone else's irq. I wonder if we should pass a struct resource instead, in case there are multiple interrupt sources, as well as having it registered with the right resource systems. -- Ben Q: What's a light-year? A: One-third less calories than a regular year. --

Not really. If someone implements atomic64 on 32-bit they may do it by putting a spinlock in the atomic64_t. Or they might use hashed spinlocks, in which case that'll work. Probably hashed spinlocks, given the (realtively new) convention that the all-zeroes pattern is a legit way of initialising an atomic_t. --

I'd say that reporting -1 for unknown is ugly. You can have -1A current easily (charging at 1A), and I've seen machines reporting <0 current -- when charging. Logical and well-defined. -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html --

I think the number is not necessarily "never before used", but rather "not currently used". -hpa --

Signed-off-by: Ian Campbell <ian.campbell@citrix.com> --- drivers/xen/events.c | 102 ++++++++++++++++++++++++++++++++------------------ 1 files changed, 65 insertions(+), 37 deletions(-) diff --git a/drivers/xen/events.c b/drivers/xen/events.c index c8f3e43..33fae3d 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -129,46 +129,76 @@ static struct irq_chip xen_dynamic_chip; static struct irq_chip xen_percpu_chip; static struct irq_chip xen_pirq_chip; -/* Constructor ...

I suspect what we should really be doing is to segregate the different classes of event channel in IRQ space. I _think_ this new stuff is happy with a discontinuous (but presumably clustered) IRQ space, I should probably check. e.g. regular interdomain event channels, VIRQs and the like should probably request allocations from some range higher than nr_hw_irqs, thus avoiding conflicts with hardware PIRQ event channels which would ask for a 1-1 mapping with the GSI (i.e. same interrupt numbers ...

Removes nr_irq sized array allocation at start of day. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> --- drivers/xen/events.c | 50 +++++++++++++++++++++++++++++++++++++++++--------- 1 files changed, 41 insertions(+), 9 deletions(-) diff --git a/drivers/xen/events.c b/drivers/xen/events.c index 94055ea..9b58505 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -56,6 +56,8 @@ */ static DEFINE_SPINLOCK(irq_mapping_update_lock); +static ...

Encapsulate allocate and free in xen_irq_alloc and xen_irq_free. Signed-off-by: Ian Campbell <ian.campbell@citrix.com> --- drivers/xen/events.c | 68 ++++++++++++++++++++----------------------------- 1 files changed, 28 insertions(+), 40 deletions(-) diff --git a/drivers/xen/events.c b/drivers/xen/events.c index 97612f5..c8f3e43 100644 --- a/drivers/xen/events.c +++ b/drivers/xen/events.c @@ -394,41 +394,29 @@ static int find_unbound_pirq(void) return -1; } -static int ...

So I am curious what the /proc/interrupts looks?The issue (and the reason for this implementation above) was that under PV with PCI devices we would overlap PCI devices IRQs with Xen event channels. So we could have a USB device at IRQ 16 _and_ also a xen_spinlock4 handler. That would throw off the system since the xen_spinlock4 was an edge type handler while the USB device was an level (at least on my box). But with this shinny sparse_irq rework, maybe this is not an issue anymore? Can we ...

Ah, oh, and I added my review to yours :-(. Sorry, for the inconvenience it caused: https://lists.berlios.de/pipermail/socketcan-core/2010-October/004787.html Anyway, Ohtake should have received the review. The driver is *not* yet ready for inclusion. Therefore, please fix the remaining issues asap. Thanks, Wolfgang. --

We tripped over another issue in Fedora: https://bugzilla.redhat.com/show_bug.cgi?id=646027 I think I see the problem, and posted a patch for testing. When it's resolved, I'll post an updated series. Bjorn --

On Mon, 25 Oct 2010 15:08:48 -0600 Ok thanks, I'll drop it from -next then. -- Jesse Barnes, Intel Open Source Technology Center --

Not really, CPUs can stay idle for _very_ long times (!x86 cpus that don't have crappy timers like HPET which roll around every 2-4 seconds). But all CPUs staying idle for a long time is exactly the scenario you fix before using the decay_load_misses() stuff, except that is for the load-balancer per-cpu load numbers not the global cpu load avg. Won't a similar approach work here? --

[Empty message]

This hunk conflicts with the update Mike Miller sent commit 6362beea8914cbd4630ccde3617d944aeca2d48f Author: Mike Miller <mike.miller@hp.com> Date: Tue Oct 19 09:40:34 2010 +0200 cciss: fix PCI IDs for new Smart Array controllers And which is now mainline. James --

And a locking macro/structure is better in self documenting than a helper function which was proposed by Christoph? Thanks, tglx --

And add a nice bit of overhead to a hot path... Making it per-cpu would help, but then we'd end up with the same ifdef mess as mnt_writers and still a non-zero overhead. But what's the point anyway, the per-sb write count always equals SUM(mnt->mnt_writers). How about this variant? It addresses the vfsmount vs. sb read only status independence: Index: linux-2.6/fs/internal.h =================================================================== --- ...

On Mon, 25 Oct 2010 00:52:34 +0800 Hi, thanks for the patch. I'll make a few changes to what I finally apply - for example we don't really need mempools in r1buf_poll_alloc as that isn't on the writeout path - so I'll tidy that up first. Also I'll avoid making changes to fs/bio.c at first. It may still be a good idea to have a bio_clone_bioset, but that should be a separate patch - there are at least 3 places that would use it. Thanks - I'll try to get this into the current ...

Hello Frederic, Sorry for delay, the patch have been accepted by Jason. http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=c1bb9a9c19... Dongdong --

This is pure conjecture, I have not measured it, but I suspect in fact that we could just change the composite nops in nops.h to use a 3E prefix instead of a separate 90 nop. Some platforms will take a penalty on the prefix, but that would be balanced against handling two instructions. The P5 core and others of the same generation might suffer, as it might have been able to do U+V pipe pairing on two instructions which it wouldn't for prefixes. -hpa --

Cool, thanks for the info! Steven and Jason should probably update their respective infrastructure to use the 32-bit 5-byte nop you propose rather than the 5-byte jump. Mathieu -- Mathieu Desnoyers Operating System Efficiency R&D Consultant EfficiOS Inc. http://www.efficios.com --

-Os unfortunately drops a bunch of optimizations. With gcc 4.5.1 there is actually a way to guarantee to get rid of double jumps, which is that you tell gcc that it is branching to one of two targets: asm goto("1: .byte 0xe9 ; .long %l[t_no]-2f\n" "2:\n" /* patching infrastructure goes here */ : : "i" (bit) : : t_no, t_yes); __builtin_unreachable(); t_no: return false; t_yes: return ...

Looks good to me. Reviewed-by: Tejun Heo <tj@kernel.org> Thanks. -- tejun --