With the recent progression of your arch tree, and the wakelock/suspend-blocker stuff, what's the need for these files to still be in the staging directory at all? Can't it be deleted and it go through your tree? thanks, greg k-h --

Hi Jörn, After merging the logfs tree, today's linux-next build (x86_64 allmodconfig) produced this warning: fs/logfs/file.c: In function 'logfs_fsync': fs/logfs/file.c:225: warning: unused variable 'super' Introduced by commit c0c79c31c9d5fcc19812c6c35f842baf50ee788a ("logfs: fix sync"). -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

No major change since V6, just looks nicer. Changelog since V6 o Make the nested anon_vma lock taking prettier Changelog since V5 o Have rmap_walk take anon_vma locks in order starting from the "root" o Ensure that mm_take_all_locks locks VMAs in the same order Changelog since V4 o Switch back anon_vma locking to put bulk of locking in rmap_walk o Fix anon_vma lock ordering in exec vs migration race Changelog since V3 o Rediff against the latest upstream tree o Improve ...

vma_adjust() is updating anon VMA information without locks being taken. In contrast, file-backed mappings use the i_mmap_lock and this lack of locking can result in races with users of rmap_walk such as page migration. vma_address() can return -EFAULT for an address that will soon be valid. For migration, this potentially leaves a dangling migration PTE behind which can later cause a BUG_ON to trigger when the page is faulted in. With the recent anon_vma changes, there can be more than one ...

Ingo, Please pull the latest tip/tracing/core tree, which can be found at: git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-2.6-trace.git tip/tracing/core Li Zefan (1): tracing: Make the documentation clear on trace_event boot option Steven Rostedt (1): tracing: Fix tracepoint.h DECLARE_TRACE() to allow more than one header Thiago Farina (1): tracing: Fix "integer as NULL pointer" warning. ---- Documentation/trace/events.txt | 3 +- ...

From: Thiago Farina <tfransosi@gmail.com> kernel/trace/trace_output.c:256:24: warning: Using plain integer as NULL pointer Signed-off-by: Thiago Farina <tfransosi@gmail.com> LKML-Reference: <1264349038-1766-3-git-send-email-tfransosi@gmail.com> Signed-off-by: Steven Rostedt <rostedt@goodmis.org> --- kernel/trace/trace_output.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index 8e46b33..2404c12 ...

The Microblaze implementations of get_user() and (MMU) put_user() evaluate the address argument more than once. This causes unexpected side-effects for invocations that include increment operators, i.e. get_user(foo, bar++). This patch also removes the distinction between MMU and noMMU put_user(). Without the patch: $ echo 1234567890 > /proc/sys/kernel/core_pattern $ cat /proc/sys/kernel/core_pattern 12345 Signed-off-by: Steven J. Magnani <steve@digidescorp.com> --- diff -uprN ...

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 This addresses all known CONFIG_PROVE_RCU warnings (modulo 3 fixes pending in maintainer trees). Thanks, Ingo ------------------> David Howells (2): KEYS: Fix an RCU warning KEYS: Fix an RCU warning in the reading of user keys Li Zefan (4): cgroup: Fix an RCU warning in cgroup_path() cgroup: Fix ...

Pulled, thanks Paul! Ingo --

Yes, it is described implicitly in the EHCI spec, section 4.10.3. Transaction errors cause the status field to be updated to reflect the type of error, but the transaction continues to be retried until the Active bit is set to 0. When the error counter reaches 0, the Halt bit is set and the Active bit is cleared. Alan Stern --

But how do you know that processes won't get unfrozen until all the pending IRQs have been handled? Imagine something like this: CPU 0 CPU 1 ----- ----- Wake up non-boot CPUs Resume devices Invoke the IRQ handler [ CPU 0 should wait here for the handler to finish, but it doesn't ] Defrost threads Handler running... Workqueue routine runs Start another suspend Handler enables a suspend blocker, but it's too late Alan Stern --

I have many different good quality lists from various sources. The prices are relatively low as well. Drop me a line here: Adan.Barlow@realresults.co.cc I'll get you all the details and samples. By emailing disappear@realresults.co.cc you will have your email taken off --

There were some reports[1] of isp1760 USB driver malfunctioning with high speed devices, noticed on Blackfin and PowerPC targets. These reports indicated that the original Philips 'pehcd'[2] driver worked fine. We've noticed the same issue with an ARM RealView platform. This happens under load (with only some mass storage devices, not all, just as in another report[3]): error bit is set in DW3 error bit is set in DW3 error bit is set in DW3 usb 1-1.2: device descriptor read/64, ...

I'd prefer to take it in -tip, partly because I would like to rework a few things -- including preexisting braindamage like c->x86_hyper_vendor (since when was hypervisor vendor a per-cpu property!?) I'll try to get to it today, if not feel free to bug me. -hpa -- H. Peter Anvin, Intel Open Source Technology Center I work for Intel. I don't speak on their behalf. --

Probably just make it a global variable, or somesuch. -hpa --

This provides basic DIO support for reads only. It does not do any of the work to recover from mismatching checksums, that will come later. A few design changes have been made from Jim's code (sorry Jim!) 1) Use the generic direct-io code. Jim originally re-wrote all the generic DIO code in order to account for all of BTRFS's oddities, but thanks to that work it seems like the best bet is to just ignore compression and such and just opt to fallback on buffered IO. 2) Fallback on buffered ...

What about making sure that dio->submit_io is always set in These hunk seem like they're unrealted to the actual hook, I'd rather Please don't add another wrapper. At this point I'd suggest just using __blockdev_direct_IO for everything but the trivial blockdev_direct_IO and also kill blockdev_direct_IO_no_locking. --

This comes down mostly to efficiency. Although the suspend blocker patch does the actual suspending in a workqueue thread, AFAIK there's no reason it couldn't use a user thread instead. The important difference lies in what happens when a suspend fails because a driver is busy. Without suspend blockers, the kernel has to go through the whole procedure of freezing userspace and kernel threads and then suspending a bunch of drivers before hitting the one that's busy. Then all that work has ...

I've not received any patches relating to the above, but I seem to have received everything else, including stuff for PPC and x86. Maybe you've negated the list of people to send patches to? --

move_W_to_insn.patch From: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Move Macro W to asm/insn.h Macro W used to know if the instructions are valid for user-space/kernel space. This macro is used by kprobes and user_bkpt. (i.e user space breakpoint assistance layer.) So moving it to a common header file asm/insn.h. TODO: replace macro W with bits in inat table. (suggested by Masami Hiramatsu) Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com> --- ...

x86_support_for_user_bkpt.patch From: Srikar Dronamraju <srikar@linux.vnet.ibm.com> x86 support for user breakpoint Infrastructure Changelog: set USER_BKPT_FIX_SLEEPY if post_xol might sleep. This patch provides x86 specific userspace breakpoint assistance implementation details. It uses the "x86: instruction decoder API" patch to do validate and analyze the instructions. This analysis is used at the time of post-processing of breakpoint hit to do the necessary fix-ups. Almost all ...

uprobes_core.patch From: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Uprobes Implementation Changelog from v2: - Introduce TIF_UPROBED flag. - uprobes hooks now in fork/exec/exit paths instead of tracehooks. - uprobe_process is now part of the mm struct and is shared between processes that share the mm. - per thread information is now allocated on the fly. * Hence allocation and freeing of this information is lockless. - For now run the handler only ...

samples.patch From: Srikar Dronamraju <srikar@linux.vnet.ibm.com> Uprobes Samples This provides an example uprobes module in the samples directory. To run this module run (as root) insmod uprobe_example.ko vaddr=<vaddr> pid=<pid> Where <vaddr> is the address where we want to place the probe. <pid> is the pid of the process we are interested to probe. example: - # cd samples/uprobes [get the virtual address to place the probe.] # vaddr=0x$(objdump -T /bin/bash |awk ...

uprobes_trace_events.patch From: Srikar Dronamraju <srikar@linux.vnet.ibm.com> The following patch implements trace_event support for uprobes. In its current form it can be used to put probes at a specified text address in a process and dump the required registers when the code flow reaches the probed address. This is based on trace_events for kprobes to the extent that it may resemble that file on 2.6.34-rc3. The following example shows how to dump the instruction pointer and %ax ...

Could you also rebase your patch on the latest tip tree? :) Hmm, I'm not sure why we have to specify both of PID and exec name. Is that possible to give a symbol as below? (omit exec name) Hmm, the event name should be simpler, as like as zfree_3329. or, we might better make a new event group for each process, as like as Looks great! :D Some comments on the patch. Hmm, if user gave --list with -p, what happened? Again, it must be checked that the combination of -p option ...

From: Myron Stowe <myron.stowe@hp.com> Update core IPMI driver printk()'s with dev_printk(), and its constructs, to provide additional device topology information. An example of the additional device topology for a PNP device - ipmi_si 00:02: probing via ACPI ipmi_si 00:02: [io 0x0ca2-0x0ca3] regsize 1 spacing 1 irq 0 ipmi_si 00:02: Found new BMC (man_id: 0x00000b, prod_id: 0x0000, ... ipmi_si 00:02: IPMI kcs interface initialized and for a PCI device - ipmi_si 0000:01:04.6: ...

wbr, jirka Signed-off-by: Jiri Olsa <jolsa@redhat.com> --- drivers/video/atafb.c | 2 -- drivers/video/matrox/matroxfb_DAC1064.c | 5 +---- 2 files changed, 1 insertions(+), 6 deletions(-) diff --git a/drivers/video/atafb.c b/drivers/video/atafb.c index f3aada2..5b2b5ef 100644 --- a/drivers/video/atafb.c +++ b/drivers/video/atafb.c @@ -1718,11 +1718,9 @@ static int falcon_setcolreg(unsigned int regno, unsigned int red, (((red & 0xe000) >> 13) | ((red & ...

wbr, jirka Signed-off-by: Jiri Olsa <jolsa@redhat.com> --- arch/s390/include/asm/processor.h | 5 ----- arch/s390/kernel/vdso.c | 4 ---- 2 files changed, 0 insertions(+), 9 deletions(-) diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h index 73e2598..ced1714 100644 --- a/arch/s390/include/asm/processor.h +++ b/arch/s390/include/asm/processor.h @@ -52,8 +52,6 @@ extern int get_cpu_capability(unsigned int *); #endif /* __s390x__ */ ...

wbr, jirka Signed-off-by: Jiri Olsa <jolsa@redhat.com> --- drivers/ata/ahci.c | 2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 8b502f4..5326af2 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -3179,7 +3179,6 @@ static bool ahci_broken_online(struct pci_dev *pdev) #ifdef CONFIG_ATA_ACPI static void ahci_gtf_filter_workaround(struct ata_host *host) { -#ifdef CONFIG_ATA_ACPI static const struct ...

Ah, it's finally come through. This is wrong. There is a preceding #ifdef CONFIG_ARCH_OMAP2PLUS, but there's also a preceding #endif terminating that conditional section. --

wbr, jirka Signed-off-by: Jiri Olsa <jolsa@redhat.com> --- fs/ext4/super.c | 2 -- include/linux/socket.h | 4 ---- kernel/sched.c | 6 ------ kernel/srcu.c | 2 -- security/capability.c | 2 -- sound/pci/au88x0/au88x0_core.c | 4 ---- 6 files changed, 0 insertions(+), 20 deletions(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index e14d22c..afba6f0 100644 --- a/fs/ext4/super.c +++ ...

wbr, jirka Signed-off-by: Jiri Olsa <jolsa@redhat.com> --- arch/powerpc/include/asm/vdso_datapage.h | 2 -- arch/powerpc/kernel/traps.c | 2 -- 2 files changed, 0 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/include/asm/vdso_datapage.h b/arch/powerpc/include/asm/vdso_datapage.h index 13c2c28..7c9cfe9 100644 --- a/arch/powerpc/include/asm/vdso_datapage.h +++ b/arch/powerpc/include/asm/vdso_datapage.h @@ -114,9 +114,7 @@ struct vdso_data { #endif /* ...

wbr, jirka Signed-off-by: Jiri Olsa <jolsa@redhat.com> --- arch/h8300/include/asm/tlb.h | 13 ------------- 1 files changed, 0 insertions(+), 13 deletions(-) diff --git a/arch/h8300/include/asm/tlb.h b/arch/h8300/include/asm/tlb.h index 3dea80a..6339a02 100644 --- a/arch/h8300/include/asm/tlb.h +++ b/arch/h8300/include/asm/tlb.h @@ -7,17 +7,4 @@ #define tlb_flush(tlb) do { } while(0) -/* - include/asm-h8300/tlb.h -*/ - -#ifndef __H8300_TLB_H__ -#define ...

I'm pretty sure I fixed those so there should be a patch on its way to or in mainline somewhere. johannes --

applied, along with achiang's lenovo list. looking forward to deleting this list in 2.6.35... thanks, Len Brown, Intel Open Source Technology Center --

From: Paul E. McKenney <paulmck@linux.vnet.ibm.com> The memory barriers must be in the SMP case, not in the !SMP case. Also add a barrier after the atomic_inc() in order to ensure that other CPUs see post-synchronize_sched_expedited() actions as following the expedited grace period. Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Signed-off-by: Tejun Heo <tj@kernel.org> --- kernel/sched.c | 21 +++++++++++---------- 1 files changed, 11 insertions(+), 10 deletions(-) diff ...

Currently migration_thread is serving three purposes - migration pusher, context to execute active_load_balance() and forced context switcher for expedited RCU synchronize_sched. All three roles are hardcoded into migration_thread() and determining which job is scheduled is slightly messy. This patch kills migration_thread and replaces all three uses with cpu_stop. The three different roles of migration_thread() are splitted into three separate cpu_stop callbacks - migration_cpu_stop(), ...

Implement a simplistic per-cpu maximum priority cpu monopolization mechanism. A non-sleeping callback can be scheduled to run on one or multiple cpus with maximum priority monopolozing those cpus. This is primarily to replace and unify RT workqueue usage in stop_machine and scheduler migration_thread which currently is serving multiple purposes. Four functions are provided - stop_one_cpu(), stop_one_cpu_nowait(), stop_cpus() and try_stop_cpus(). This is to allow clean sharing of resources ...

Hi When I was examining a reported performance degradation, I found out that the reason was buggy firmware in Maxtor Atlas 15K2 SCSI disks. If the requests are larger than 256kB, read rate drops excessively, even to one half of the normal read rate. Performance on these disks improves if the request size is limited with echo 256 >/sys/block/sdc/queue/max_sectors_kb I remember seeing similar problem (increased transfer size => decreased performance) with IDE and SATA Maxtor disks some ...

On Thu, 6 May 2010 12:15:58 -0400 (EDT) ATA MAXTOR STM316021 3.AA 64 kB: 78.9 MB/s 128 kB: 79.1 MB/s 192 kB: 79.1 MB/s 256 kB: 78.6 MB/s 320 kB: 79.1 MB/s 384 kB: 78.9 MB/s 448 kB: 79.0 MB/s 512 kB: 79.1 MB/s 576 kB: 79.0 MB/s 640 kB: 79.1 MB/s 704 kB: 79.0 MB/s 768 kB: 79.1 MB/s 832 kB: 79.0 MB/s 896 kB: 79.0 MB/s 960 kB: 79.1 MB/s 1024 kB: 78.8 MB/s -- Paolo Ornati Linux 2.6.34-rc6-00063-gbe1066b on x86_64 --

Can this be duplicate of https://bugs.freedesktop.org/show_bug.cgi?id=27822 ? -- Rafał --

Yes this is a duplicate of : https://bugs.freedesktop.org/show_bug.cgi?id=27822 The patch attached in the bug should fix your issue. Sorry for the trouble. Cheers, Jerome --

>From 0692c53dfb745df62fd805fe089786951a8d96ec Mon Sep 17 00:00:00 2001 From: Stefan Achatz <erazor_de@users.sourceforge.net> Date: Thu, 15 Apr 2010 17:24:45 +0200 Subject: [PATCH] HID: add chardev to propagate special events of Roccat hardware to userland This mail was first sent on april 15th but was not noticed. This Patch adds a chardev that is used to report special events like request for on-screen-display of actual mouse-setting informations, request for execution of macro sequences not ...

From: Andrea Arcangeli <aarcange@redhat.com> Page migration requires rmap to be able to find all migration ptes created by migration. If the second rmap_walk clearing migration PTEs misses an entry, it is left dangling causing a BUG_ON to trigger during fault. For example; [ 511.201534] kernel BUG at include/linux/swapops.h:105! [ 511.201534] invalid opcode: 0000 [#1] PREEMPT SMP [ 511.201534] last sysfs file: /sys/block/sde/size [ 511.201534] CPU 0 [ 511.201534] Modules linked in: ...

One conceptual problem here. By taking the oldest anon_vma, instead of the anon_vma of the page, we may end up searching way too many processes. Eg. if the page is the page of a child process in a forking server workload, the above code will end up searching the parent and all of the siblings - even for a private page, in the child process's private anon_vma. For an Apache or Oracle system with 1000 clients (and child processes), that could be quite a drag - searching 1000 times as many ...

[ Removed '-' lines to show the actual end result ] That loop is f*cked up. In the "goto next_vma" case, it will then test the 'ret' from the _previous_ iteration after having unlocked the anon_vma. Which may not even exist, if this is the first one. Yes, yes, 'ret' is initialized to SWAP_AGAIN, so it will work, but it's still screwed up. It's just _waiting_ for bugs to be introduced. Just make the "goto out" case unlock thngs properly. Have a real exclusive error return case ...

CPU notifier register functions also exist if CONFIG_SMP is disabled. This change is part of hotplug code rework and also necessary for later patches. Cc: Andi Kleen <andi@firstfloor.org> Signed-off-by: Robert Richter <robert.richter@amd.com> --- arch/x86/oprofile/nmi_int.c | 7 +------ 1 files changed, 1 insertions(+), 6 deletions(-) diff --git a/arch/x86/oprofile/nmi_int.c b/arch/x86/oprofile/nmi_int.c index 2458204..c5df8ee 100644 --- a/arch/x86/oprofile/nmi_int.c +++ ...

This fixes a NULL pointer dereference that is triggered when taking a cpu offline after oprofile was initialized, e.g.: $ opcontrol --init $ opcontrol --start-daemon $ opcontrol --shutdown $ opcontrol --deinit $ echo 0 > /sys/devices/system/cpu/cpu1/online See the crash dump below. Though the counter has been disabled the cpu notifier is still active and trying to use already freed counter data. This fix is for linux-stable. To proper fix this, the hotplug code must be rewritten. ...

This patch reworks oprofile cpu hotplug code as follows: Introduce ctr_running variable to check, if counters are running or not. The state must be known for taking a cpu on or offline and when switching counters during counter multiplexing. Protect on_each_cpu() sections with get_online_cpus()/put_online_cpu() functions. This is necessary if notifiers or states are modified. Within these sections the cpu mask may not change. Switch only between counters in nmi_cpu_switch(), if counters ...

Current IBS code is not hotplug capable. An offline cpu might not be initialized or deinitialized properly. This patch fixes this by removing on_each_cpu() functions. The IBS init/deinit code is executed in the per-cpu functions model->setup_ctrs() and model->cpu_down() which are also called by hotplug notifiers. model->cpu_down() replaces model->exit() that became obsolete. Cc: Andi Kleen <andi@firstfloor.org> Signed-off-by: Robert Richter <robert.richter@amd.com> --- ...

Here's a completely new issue. When using opportunistic suspends on an SMP system, it could happen that the system gets a wakeup event and this routine starts running again before the event's IRQ handler has finished (or has enabled a suspend blocker). The system would re-suspend too soon. Is there anything in the PM core that waits until all outstanding IRQs have been handled before unfreezing processes? If there is, I can't find it. Should such a thing be added? Alan Stern --

Reviewed-by: Rik van Riel <riel@redhat.com> --

Add the Controller reset functions to the Driver. This is first part of the Online Controller Reset. The driver supports XScle and Gen2 OCR. The reset functions for those two chips added to the driver. Signed-off-by Bo Yang<bo.yang@lsi.com> --- megaraid_sas.c | 235 +++++++++++++++++++++++++++++++++++++++++++++++++++------ megaraid_sas.h | 40 ++++++++- 2 files changed, 248 insertions(+), 27 deletions(-) diff -rupN old/drivers/scsi/megaraid/megaraid_sas.c ...

This is third part of the Online Controller Reset. In ISR routine, if driver receive the FW state change interrupt And the online controller reset support enabled in FW, driver will read the controller register. Driver will issue Controller reset if the FW register state failure. Also driver will back up the pending cmds for the re-fire after the reset finished. Also Driver added the CTIO support to this patch. Signed-off-by Bo Yang<bo.yang@lsi.com> --- megaraid_sas.c | 412 ...

Update the version and documentation. Signed-off-by Bo Yang<bo.yang@lsi.com> --- Documentation/scsi/ChangeLog.megaraid_sas | 42 ++++++++++++++++++++++++++++++ drivers/scsi/megaraid/megaraid_sas.h | 6 ++-- 2 files changed, 45 insertions(+), 3 deletions(-) diff -rupN old/Documentation/scsi/ChangeLog.megaraid_sas new/Documentation/scsi/ChangeLog.megaraid_sas --- old/Documentation/scsi/ChangeLog.megaraid_sas 2010-05-05 02:08:10.000000000 -0400 +++ ...

Driver add the input parameters support for max_sectors for our gen2 chip. Customer can Set the max_sectors support to 1MB for gen2 chip during the driver load. Signed-off-by Bo Yang<bo.yang@lsi.com> --- megaraid_sas.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ megaraid_sas.h | 1 2 files changed, 75 insertions(+) diff -rupN old/drivers/scsi/megaraid/megaraid_sas.c new/drivers/scsi/megaraid/megaraid_sas.c --- old/drivers/scsi/megaraid/megaraid_sas.c 2010-05-03 ...

With this patch, _none_ of the wakeup bits are enabled. That should work, right? But it leads to another question: Later on, when the controller is put into D3hot, the new code will try to turn on wakeup bits for ports that have already been suspended. It should be safe because ehci_set_wakeup_flags() will enable PORT_WKDISC_E | PORT_WKOC_E for connected ports, and PORT_WKCONN_E | PORT_WKOC_E for unconnected ports. Still, since this happens _after_ the ports are suspended and the ...

The Linux Power-Management Summit is scheduled for August 9, at the start of the LinuxCon meeting in Boston. This would make an excellent topic for discussion. Alan Stern --

Sure, but it will make scheduling much more expensive. The current scheme will only ever check the first N events because it stops at the first that fails, and since you can max fix N events on the PMU its constant time. To fix this issue you'd have to basically always iterate all events and only stop once the PMU is fully booked, which reduces to an O(n) worst case algorithm. But yeah, I did think of making the thing an RB-tree and basically schedule on service received, that should fix ...

Not sure if this would be any better. --

I'm not sure why you're arguing, you asked why it did as it did, I gave an answer ;-) I agree its not optimal, but fixing it isn't trivial, I would very much like to avoid a full O(n) loop over all events, esp since creating them is a non-privilidged operation. So what we can look at is trying to do better, and making it a service based scheduler instead of a strict RR should at least get a more equal distribution. Another thing we can do is quit at the second or third fail. --

Schedule those events that got scheduled least, if because of constraints we didn't fully utilize the PMU it is very likely that strict RR (like we do now) will not end up giving equal service to each counter/group. Therefore, if you sort them in a tree, based on the amount of time they got on the PMU, and always schedule the leftmost, you do get fairness. Sure you can go add weights to them, but that's not the immediate goal. --

ULE (Unidirectional Lightweight Encapsulation RFC 4326) decapsulation code has a bug that incorrectly treats ULE SNDU packed into the remaining 2 or 3 bytes of a MPEG2-TS frame as having invalid pointer field on the subsequent MPEG2-TS frame. This patch was generated and tested against v2.6.34-rc6. I suspect that this bug was introduced in kernel version 2.6.15, but had not verified it. Care has been taken not to introduce more bug by fixing this bug, but please scrutinize the code ...

i think it'd be a better idea to do something like: if (spi->bits_per_word != 16) { if (spi->bits_per_word) { dev_err(&spi->dev, "Invalid SPI settings; bits_per_word must be 16\n"); return -EINVAL; } spi->bits_per_word = 16; spi_setup(spi); } -mike --

With no word size given in the users platform data, a generic spi host controller driver will assume a default word size of eight bit. This causes transmission to be performed bytewise, which will fail on little endian machines for sure. Failure on big endian depends on usage of slave select to mark word boundaries. Anyway, ad7877 is specified to work with 16 bit per word, so unconditionally set the word size accordingly. Signed-off-by: Oskar Schirmer <os@emlix.com> Signed-off-by: Daniel ...

Greetings, Michael Analog Devices GmbH Wilhelm-Wagenfeld-Str. 6 80807 Muenchen Sitz der Gesellschaft Muenchen, Registergericht Muenchen HRB 4036 Geschaeftsfuehrer Thomas Wessel, William A. Martin, Margaret Seif

Use the resource_size function instead of manually calculating the resource size. This reduces the chance of introducing off-by-one errors. Signed-off-by: Tobias Klauser <tklauser@distanz.ch> --- drivers/dma/mpc512x_dma.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/dma/mpc512x_dma.c b/drivers/dma/mpc512x_dma.c index 84326c9..201e6e1 100644 --- a/drivers/dma/mpc512x_dma.c +++ b/drivers/dma/mpc512x_dma.c @@ -658,7 +658,7 @@ static int __devinit ...

This patch fixes a long line length in comedidev.h to make checkpatch.pl happy Signed-off-by: Mark Rankilor <reodge@gmail.com> --- drivers/staging/comedi/comedidev.h | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-) diff --git a/drivers/staging/comedi/comedidev.h b/drivers/staging/comedi/comedidev.h index c38ebb4..4eb2b77 100644 --- a/drivers/staging/comedi/comedidev.h +++ b/drivers/staging/comedi/comedidev.h @@ -374,7 +374,9 @@ enum subdevice_runflags { SRF_RUNNING = ...

This is a patch to include log levels and fix some over length lines in printk calls in drivers.c Signed-off-by: Mark Rankilor <reodge@gmail.com> --- drivers/staging/comedi/drivers.c | 16 ++++++++++------ 1 files changed, 10 insertions(+), 6 deletions(-) diff --git a/drivers/staging/comedi/drivers.c b/drivers/staging/comedi/drivers.c index 1f48b6d..ffb0ad4 100644 --- a/drivers/staging/comedi/drivers.c +++ b/drivers/staging/comedi/drivers.c @@ -95,7 +95,8 @@ static void ...

Rename 'root_count' to 'active_count' in kvm_mmu_page, since the unsync pages also will use it in later patch Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> --- arch/x86/include/asm/kvm_host.h | 7 ++++++- arch/x86/kvm/mmu.c | 14 +++++++------- arch/x86/kvm/mmutrace.h | 6 +++--- 3 files changed, 16 insertions(+), 11 deletions(-) diff --git a/arch/x86/include/asm/kvm_host.h b/arch/x86/include/asm/kvm_host.h index ed48904..86a8550 100644 --- ...

In current code, shadow page can become asynchronous only if one shadow page for a gfn, this rule is too strict, in fact, we can let all last mapping page(i.e, it's the pte page) become unsync, and sync them at invlpg or flush tlb time. This patch allow more page become asynchronous at gfn mapping time Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> --- arch/x86/kvm/mmu.c | 81 +++++++++++++++++++++++---------------------------- 1 files changed, 37 insertions(+), 44 ...

Split kvm_sync_page() into kvm_sync_page() and kvm_sync_page_transient() to clarify the code address Avi's suggestion kvm_sync_page_transient() function only update shadow page but not mark it sync and not write protect sp->gfn. it will be used by later patch Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> --- arch/x86/kvm/mmu.c | 28 ++++++++++++++++++++++++---- 1 files changed, 24 insertions(+), 4 deletions(-) diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c index ...

1.Name in full2.Address 3.Phone --

Currently all fpu state access is through tsk->thread.xstate. Since we wish to generalize fpu access to non-task contexts, wrap the state in a new 'struct fpu' and convert existing access to use an fpu API. Signal frame handlers are not converted to the API since they will remain task context only things. Signed-off-by: Avi Kivity <avi@redhat.com> --- arch/x86/include/asm/i387.h | 115 ++++++++++++++++++++++++++++---------- arch/x86/include/asm/processor.h | 6 ++- ...

This last sentence is of course incorrect. EP0 has been tested as well as the first sentence implies. Again forgot to edit the whole cover letter. Sorry about the confusion. -- Best regards, _ _ | Humble Liege of Serenely Enlightened Majesty of o' \,=./ `o | Computer Science, Michał "mina86" Nazarewicz (o o) +----[mina86*mina86.com]---[mina86*jabber.org]----ooO--(_)--Ooo-- --

This is a resend of the pull request for 2.6.34-rc6 with the addition of an lcdcfb patch. It hasn't been rebased, so if you've already pulled the previous changes then you should just get that patch. Please pull from: master.kernel.org:/pub/scm/linux/kernel/git/lethal/sh-2.6.git sh/for-2.6.34 Which contains: Guennadi Liakhovetski (3): SH: fix TS field shift calculation for DMA drivers SH: fix error paths in DMA driver sh: fix a number of Oopses and leaks in SH ...

In the early days we passed the mmap sem around. That became the "int fshared" with the fast gup improvements. Then we added "int clockrt" in places. This patch unifies these options as "flags" and cleans up various calls which had fshared in their signature but no longer used it. V6: Whitespace fixes V4: Fix an operator precedence bug Signed-off-by: Darren Hart <dvhltc@us.ibm.com> --- kernel/futex.c | 215 +++++++++++++++++++++++++++----------------------------- 1 files changed, 104 ...

Prepare for FUTEX_LOCK by refactoring futex_lock_pi_atomic() into lock_futex_atomic() and lock_pi_futex_atomic(). The name change is meant to reflect the naming convention in futex.c that futex_*() functions map directly to futex op codes and the others are internal helper functions. V6: Whitespace cleanup Signed-off-by: Darren Hart <dvhltc@us.ibm.com> =================================================================== --- kernel/futex.c | 79 ...

Hello, after some nfs server start/stop i hit kernel bug on 2.6.26. There is hope[1] that this bug is fixed[2], and fix is merged[3] into 2.6.29. But there is missing bugfix for older stable kernels. Reports for this bug can also be found on distro bug systems[4][5][6][7]. Sadly these bugs are still open. Please backport fix for stable kernels too. roma1390 1. http://linux-nfs.org/pipermail/nfsv4/2009-March/010018.html 2. ...

What does it mean? Should I take care about getting it upstream? --peter; --

BTW, the CHROMIUM in the subject is just used for internal bookeeping for patches not yet accepted upstream. --

Hello all, this patch is for s5pv210(or s5pc110) support and adds resizing and default window features and also fixes distortendness situation on pixel format more then 24bpp. patch contents are as following. 1. resizing feature. - when application tries to change framebuffer attributes(pixel format, resolution and so no), fb_set_var() is called by linux framebuffer(fbmem.c) at that time, s3c-fb sets attributes for setting porch and pixelclock to registers so setting would ...

CONFIG_ACPI depends on CONFIG_PCI, so this file should not be build with CONFIG_PCI=n. Please let me know how you were able to generate this illegal configuration. thanks, Len Brown, Intel Open Source Technology Center --

From: Randy Dunlap <randy.dunlap@oracle.com> Fix build error: drivers/media/IR/rc-map.c:51: error: implicit declaration of function 'msleep' Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: Mauro Carvalho Chehab <mchehab@infradead.org> --- drivers/media/IR/rc-map.c | 1 + 1 file changed, 1 insertion(+) --- linux-next-20100506.orig/drivers/media/IR/rc-map.c +++ linux-next-20100506/drivers/media/IR/rc-map.c @@ -13,6 +13,7 @@ */ #include <media/ir-core.h> +#include ...

When CONFIG_PM is not set: drivers/built-in.o: In function `acpi_init': bus.c:(.init.text+0x2d84): undefined reference to `pm_flags' bus.c:(.init.text+0x2d91): undefined reference to `pm_flags' --- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** --

You weren't ignored. Mimi responded and mentioned a subsequent patch, after which, there was no further discussion. I've reverted: b89e66e1e396f7b5436af154e58209320cc08aed "TPM: ACPI/PNP dependency removal" a674fa46c79ffa37995bd1c8e4daa2b3be5a95ae "ima: remove ACPI dependency" Note: any further ACPI-related changes here should have acks from linux-acpi folk. - James -- James Morris <jmorris@namei.org> --

Hi all, Changes since 20100505: Dropped trees: tmio-mmc (long unresolved conflicts and unable to fetch) hwlat (at the maintainer's request) The wireless tree lost its conflicts. The rr tree gained a conflict against the net tree. The security-testing tree gained a conflict against Linus' tree. The staging-next tree gained a conflict against the net tree. ---------------------------------------------------------------------------- I have created today's linux-next tree ...

-- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** --

From: Randy Dunlap <randy.dunlap@oracle.com> Fix build errors by including linux/semaphore.h: drivers/staging/cxt1e1/pmcc4_private.h:144: error: field 'sr_sem_busy' has incomplete type drivers/staging/cxt1e1/pmcc4_private.h:146: error: field 'sr_sem_wait' has incomplete type drivers/staging/cxt1e1/pmcc4_private.h:189: error: field 'sem_wdbusy' has incomplete type drivers/staging/cxt1e1/musycc.c:617: error: implicit declaration of function 'down' drivers/staging/cxt1e1/musycc.c:641: error: ...

> $ scripts/get_maintainer.pl --rolestats -f fs/eventpoll.c > Alexander Viro <viro@zeniv.linux.org.uk> (maintainer:FILESYSTEMS (VFS...) > Andrew Morton <akpm@linux-foundation.org> (commit_signer:2/6=33%) > "Eric W. Biederman" <ebiederm@xmission.com> (commit_signer:2/6=33%) > Davide Libenzi <davidel@xmailserver.org> (commit_signer:2/6=33%) > Roland Dreier <rolandd@cisco.com> (commit_signer:1/6=17%) > Stefan Richter <stefanr@s5r6.in-berlin.de> (commit_signer:1/6=17%) > ...

> I should add --nogit option. The script does get some useful information out of git (but obviously also some questionable). --roles and --rolestats give away some more data... $ scripts/get_maintainer.pl --rolestats -f fs/eventpoll.c Alexander Viro <viro@zeniv.linux.org.uk> (maintainer:FILESYSTEMS (VFS...) Andrew Morton <akpm@linux-foundation.org> (commit_signer:2/6=33%) "Eric W. Biederman" <ebiederm@xmission.com> (commit_signer:2/6=33%) Davide Libenzi <davidel@xmailserver.org> ...

use wrapper functions. epoll should not touch flags in wait_queue_t. This patch introduces a new function add_wait_queue_head_exclusive_locked(), for the users, who use wait queue as a LIFO queue. Signed-off-by: Changli Gao <xiaosuo@gmail.com> ---- fs/eventpoll.c | 5 ++--- include/linux/wait.h | 15 +++++++++++++-- 2 files changed, 15 insertions(+), 5 deletions(-) diff --git a/fs/eventpoll.c b/fs/eventpoll.c index bd056a5..8137f6e 100644 --- a/fs/eventpoll.c +++ ...

(not knowing the get_maintainers internals ...) Maybe if you weigh each contributor by the amount (in terms of diffstat modulo) of changes, you can get more accurate results. Part of the weight might be related to the time distance (relative to current time) since the changeset. - Davide --

That might tend to overweight cleanup against actual logic patches, Which is kind of what activity alone does. I think get_maintainer with --git (the default): Without a named MAINTAINER: For relatively active files: works moderately well. For relatively inactive files: might work less well. It tends to hide the original developer and show people that submit cleanup or fix patches. Maybe the --git portion could be reduce the number of commit signers added when files have named ...

Hi Greg, Today's linux-next merge of the staging-next tree got a conflict in drivers/staging/arlan/arlan-main.c between commit 22bedad3ce112d5ca1eaf043d4990fa2ed698c87 ("net: convert multicast list to list_head") from the net tree and commit dd730b627cf8ff0b9d20df94fd31b6192b188710 ("Staging: arlan: fixed unnecessary whitespace style issue in arlan-main.c") from the staging-next tree. I fixed it up (see below - this is bigger than is would otherwise be because there is a conflict with ...

Hi Greg, Ah, yes, that would be easier. Thanks. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

Hi, I was looking at the commits in http://www.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.20.14. Seems all the commits are not in the current git. For example, if I run 'git show 0b7da0500bbf612fc532358c90aad4497309d61a', I got 'fatal: bad object 0b7da0500bbf612fc532358c90aad4497309d61a'. Why is this the case? I thought all these commits are committed into the git repository. Thank you! Thanks, Tammy --

The current message, 'not registered' is confusing as it implies it was not registered with something, whereas printing 'failed to register' implies it was the gpiochip_add() call that did not work correctly. Signed-off-by: Ben Dooks <ben-linux@fluff.org> --- drivers/gpio/gpiolib.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c index eb0c3fe..384d993 100644 --- a/drivers/gpio/gpiolib.c +++ ...

hi Liam, I just found my mail client make the first patch not able to be cleanly applied by git-am. The resend one fix it, and should be cleanly applied by git-am command. Regards, --

In the case of "min_uV == max_uV == mc13783_regulators[id].voltages[0]", mc13783_fixed_regulator_set_voltage should return 0 instead of -EINVAL. This patch also adds a missing ">" character for MODULE_AUTHOR, a trivial fix. Signed-off-by: Axel Lin <axel.lin@gmail.com> Cc: Mark Brown <broonie@opensource.wolfsonmicro.com> Cc: Liam Girdwood <lrg@slimlogic.co.uk> --- drivers/regulator/mc13783-regulator.c | 6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git ...

-- Pengutronix e.K. | | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 | --

--- drivers/staging/comedi/comedi_fops.c | 56 +++++++++++++++++----------------- 1 files changed, 28 insertions(+), 28 deletions(-) diff --git a/drivers/staging/comedi/comedi_fops.c b/drivers/staging/comedi/comedi_fops.c index aca9674..999cbb6 100644 --- a/drivers/staging/comedi/comedi_fops.c +++ b/drivers/staging/comedi/comedi_fops.c @@ -598,20 +598,20 @@ copyback: static int parse_insn(struct comedi_device *dev, struct comedi_insn *insn, unsigned int *data, void *file); ...

It doesn't. All the cpu_power stuff is for SMP load-balancing, and basically means work-capacity. We normalize the per cpu runqueue weights with the cpu_power to get an even (fair) distribution. Say one cpu only has half the capacity of another cpu, then its not fair to given them equal weight, because the tasks on the 'slow' cpu would only progress at half the speed of those on the other. The cpufreq hooks are there if someone were to peg one cpu at a lower frequency than others. ...

I see, I think I've misinterpreted "cpu_power" here. So cpu_power in this context is the capability (speed / processing power) of the cpu, With cpufreq, we know how much time is spent at each cpu frequency. With specific board files in mach-* (at least in the ARM world) we can provide power measurements for frequency X. We can also currently provide power numbers for cpu idle states. This means we should be able to calculate the overall power consumed by the cpu with proper frequency ...

If we can't initialise the IOMMU, can we even get a panic message out to the user? -- Matthew Wilcox Intel Open Source Technology Centre "Bill, look, we understand that you're interested in selling us this operating system, but compare it to ours. We can't possibly take such a retrograde step." --

Check kmalloc return value before use the buffer. Signed-off-by: Zhitong Wang <zhitong.wangzt@alibaba-inc.com> --- drivers/parisc/ccio-dma.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/drivers/parisc/ccio-dma.c b/drivers/parisc/ccio-dma.c index f511e70..f465417 100644 --- a/drivers/parisc/ccio-dma.c +++ b/drivers/parisc/ccio-dma.c @@ -1407,6 +1407,11 @@ static void __init ccio_init_resources(struct ioc *ioc) struct resource *res = ioc->mmio_region; ...

Okay. However, I'd like to NAK some bits of c5b60b5e67af8be4c58d3ffcc36894f69c4fbdc1 now that I see it... @@ -1282,16 +1282,16 @@ long keyctl_session_to_parent(void) /* the parent must have the same effective ownership and mustn't be * SUID/SGID */ - if (pcred-> uid != mycred->euid || + if (pcred->uid != mycred->euid || pcred->euid != mycred->euid || pcred->suid != mycred->euid || - pcred-> gid != mycred->egid ...

Will a patch -R of below be correct enough for you? Otherwise, please send me a new patch. (This is one of the reasons why I'm not so keen on whitespace fixes, also -- James Morris <jmorris@namei.org> --

Are there any issues with injecting the GPL-licensed plug-in into the Windows vmxnet3 NDIS driver? -scott --

--

When we tested the series of jim owens' direct-IO patch, a bug occured. It could be reproduced on a 2.6.34-rc6 kernel with direct-IO patch by the following steps... # mkfs.btrfs /dev/sdb # mount /dev/sdb /home/btrfsdisk # cd /home/btrfsdisk # create test.c # dd if=/dev/zero of=file bs=7K count=1 # gcc test.c -o test # ./test =================================== test.c: =================================== #include <stdio.h> #include <unistd.h> #include <sys/types.h> #include ...

Your Webmail Quota Has Exceeded The Set Quota/Limit Which Is 20GB. You Are Currently Running On 23GB Due To Hidden Files And Folder On Your Mailbox. Please Click the Link Below To Validate Your Mailbox And Increase Your Quota. http://www.emailmeform.com/fid.php?formid=690283 Failure To Click This Link And Validate Your Quota May Result In Loss Of Important Information In Your Mailbox/Or Cause Limited Access To It. Thanks Help Desk--

kvmalloc() will try to allocate physically contiguous memory first, and try vmalloc to allocate virtually contiguous memory when the former allocation fails. kvfree() is used to free the memory allocated by kvmalloc(). It can't be used in atomic context. If the callers are in atomic contex, they can use kvfree_inatomic() instead. There is much duplicate code to do such things in kernel, so I generate the above APIs. Thank Eric Dumazet for the "kv" prefix. :) #include ...

I wonder why "struct kvfree_work_struct" is needed. According to http://kernel.ubuntu.com/git?p=jj/ubuntu-lucid.git;a=blobdiff;f=security/apparmor/matc... , static void kvfree_work(struct work_struct *work) { vfree(work); } void kvfree_inatomic(void *ptr, size_t size) { if (size < PAGE_SIZE) { kfree(ptr); } ...

Note that converting users from vmalloc() to kvmalloc() may increase fragmentation problems for other parts of the kernel, because it will tend to use up more of the available large blocks. Especially users who allocate large blocks and often. That's worth a mention somewhere. On the other hand, this API could make it easier to convert some kmalloc() calls to kvmalloc(), reducing fragmentation problems. :-) Since the caller is indicating they don't mind which happens, ...

Hello, Ingo, Here are three more RCU-lockdep fixes for 2.6.34: git://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-2.6-rcu.git rcu/urgent Thanx, Paul ------------------> Vivek Goyal <vgoyal@redhat.com> (1) blk-cgroup: Fix RCU correctness warning in cfq_init_queue() Paul E. McKenney <paulmck@linux.vnet.ibm.com> (2) vfs: fix RCU-lockdep false positive due to /proc access memcg: css_id() must be called under rcu_read_lock() block/cfq-iosched.c | 2 ++ ...

Sure, I did think it was a bit odd to funnel it through that route, but it's no big deal. -- Jens Axboe --

Yeah, please apply it to the block tree (i have just zapped it from my local tree), we really want subsystem patches to go via their respective trees. I've attached the latest patch below. Thanks! Ingo -------------------> From: Vivek Goyal <vgoyal@redhat.com> Date: Thu, 22 Apr 2010 11:54:52 -0400 Subject: [PATCH] blk-cgroup: Fix RCU correctness warning in cfq_init_queue() It is necessary to be in an RCU read-side critical section when invoking css_id(), so this patch adds one ...

Hi, Yes, that is true :-) I was planning to send an announcement to LKML this weekend. Actual coding doesn't start until the last week of May. So hopefully we can soon get rid of select altogether :-) Vegard --

I think Geert is asking "If the arch says CONFIG_HAS_IOMEM=n, but some driver does a 'select CONFIG_FOO' which then (under your proposal) forces the value CONFIG_BAR=y, which eventually ends up with CONFIG_HAS_IOMEM=y, what should the behavior be?" (I suspect the right answer here is "one of the symbols is buggy and its 'select' should be a 'depends' instead", but somebody else better double-check that conclusion - I'm hardly a Kconfig expert).

I don't quite understand the question. If a selected symbol has a critical dependency which is config'd to N then the build usually breaks ... that's what I'm calling the select problem. I thought CONFIG_HAS_IOMEM was usually selected by the architecture, though. In the new proposal, we wouldn't be able to generate the invalid configuration in the first place. James --

We should check if some Kconfig file doesn't rely on this "first hit" behavior and fix it to explicitly list the condition for a given default. Another option would be to add default SYM1 || SYM2 Other than the above, right now I don't see any issues with such approach. On a related note, I see Vegard's GSoC project to use a sat solver for kconfig got accepted [1]. Vegard, how is the project ...

No, the CONFIG_FOO will inherit the dependency on CONFIG_HAS_IOMEM. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds --

Actually, it's not any step. It's just validation of the contention that I can alter the default parser from taking the first valid default to calculating the or of the defaults without changing the current kbuild behaviour. The patch to fix kbuild goes on top of this change minus the warnings (when I actually code it up). James --

I used a command line option for our boards which I find far more convenient. Have a look at arch/arch/mach-mx2/mach-pca100.c how this is done. As a side effect you get rid of the 'defined but not used' warnings this patch introduces and you don't have to put platform related kconfig options to a totally unrelated place. -- Pengutronix e.K. | | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, ...

No, it's not one of the cancel_work_sync() that hangs but it's the del_timer_sync() right before them that hangs! (del_timer_sync() also hangs if I put it last, so the cancel_work_sync() don't hang anything) static void hid_cancel_delayed_stuff(struct usbhid_device *usbhid) { del_timer_sync(&usbhid->io_retry); /* this one never returns */ cancel_work_sync(&usbhid->restart_work); cancel_work_sync(&usbhid->reset_work); } Thanks, Bruno --

This is not a rare event. For example, the matrix keypad driver blocks If user space has to initiate every suspend attempt, then you are forcing it to poll whenever a driver needs to block suspend. -- Arve Hjønnevåg --

Because every attempted suspend requires freezing userspace, suspending devices until you hit one that refuses to suspend, resuming the devies that did suspend and then unfreezing userspace. That's not an attractive option. -- Matthew Garrett | mjg59@srcf.ucam.org --

Hmm I don't follow you. If the userspace policy daemon timer times out, the device suspends. If the device does not suspend because of a blocking driver, then the timers get reset and you try again based on some event such as when the screen blanks. Regards, Tony --

But then the userspace knows you're getting data from the modem, and it can kick some inactivity timer that determines when to try to suspend next. Why would you need to constantly try to suspend in that case? Regards, Tony --

If your userspace is suspended, how does it process the events that generated a system wakeup? If we had a good answer to that then suspend blockers would be much less necessary. -- Matthew Garrett | mjg59@srcf.ucam.org --

The alternative logic I'm suggesting is get the GUI into idle mode as soon as possible, then limp along with off-while-idle or retention-while-idle until some timer expires, then suspend the whole device. Tony --

Could you elaborate on "off-while-idle" and "retention-while-idle" ? I'm not sure I follow what you mean. Daniel --

Fascinating. So you use this to guess if the guest is still processing? I haven't thought about it hard, but is that racy? Obviously happy to apply this when you finalize it. Thanks! Rusty. --

Shouldn't this be on its own cache line? One way to achieve this is to allocate it at the end. -- error compiling committee.c: too many arguments to function --

It might be better to do something like this so the ns_to_timespec isn't performe when unnecessary. static void settimeofday(winfiletime_t hosttime) { s64 host_tns = (hosttime - WLTIMEDELTA) * 100; struct timespec host_ts = ns_to_timespec(host_tns); do_settimeofday(&host_ts); } static void adj_guesttime(winfiletime_t hosttime, u8 flags) { static s32 scnt = 50; if ((flags & ICTIMESYNCFLAG_SYNC) != 0) { settimeofday(hosttime); return; } if ((flags & ...

> Why a maximum of 50 samples? After reboot the flag ICTIMESYNCFLAG_SYNC is included in the first time message after the timesync channel is opened. Since the hv_utils module is loaded after hv_vmbus, the first message is usually missed. The other thing is, systime is automatically set to emulated hardware clock which may not be UTC time or the same time zone. So, to override these effects, we use the first 50 time samples Thanks for the optimization, I will put it into the code. - ...

I agree that this appears to be the best approach for the time being, although I can only speak for myself. Rafael --

I cannot comment on the input stuff, but the sound stuff looks good Why is this feature needed? Does the device report key releases incorrectly? These three sysfs controls could also be implemented as mixer controls. This would allow them to be automatically saved and restored when the Your mailer wrapped long lines and killed the tabs. Please see <Documentation/email-clients.txt>. And your code looks as if it does not use eight-column tabs for If you're using the atomic ...

Alan, This would be better, but before entering phy low power mode, the HW need the exact wakeup bits to be set, that's why we have: /* only enable appropriate wake bits, otherwise the * hardware can not go phy low power mode. If a race * condition happens here(connection change during bits * set), the port change detection will finally fix it. */ In the code. When port is connected, we must only enable PORT_WKOC_E | PORT_WKDISC_E, and when port is disconnected, we must ...

Yeah, my point was either disable preemption or keep the checks. In other words, if you don't disable preemption, do not use raw_smp_procesor_id(), because then we will not catch it if it changes Send another patch, I'll test it again ;-) -- Steve --

the thing are that p4 is only snippet here which is sensible to preemtion, and hw_perf_event_init is executing with preemtion off (but i could miss the details here, dont have code under my hands at moment, so PeterZ help is needed ;) but more important reason why i've saved get/put here is that otherwise i would not have rights to put tested-by tag, since it would not be the patch Steven has tested. We could make a patch on top of this one, or we could drop this one, make new with explicit ...

ok, Ingo, dont apply this patch then for a while. --

Steve, while I'm diving through call graph could you give this patch a shot? If preemtion happens -- it'll trigger it fast. -- Cyrill --- arch/x86/kernel/cpu/perf_event_p4.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) Index: linux-2.6.git/arch/x86/kernel/cpu/perf_event_p4.c ===================================================================== --- linux-2.6.git.orig/arch/x86/kernel/cpu/perf_event_p4.c +++ linux-2.6.git/arch/x86/kernel/cpu/perf_event_p4.c @@ -421,7 ...

On Thu, May 06, 2010 at 02:32:33PM -0400, Steven Rostedt wrote: Thanks a huge, Steven! I'll tinkle a bit more with p4 pmu and will send a series of patches with this one included. -- Cyrill --

Ok, I looked at this again and reproduced the traces I already deleted and fetched the Xen crash message and found something I missed before. The relevant part of the KVM trace is: qemu-system-x86-7364 [012] 790.715351: kvm_exit: reason msr rip 0xffff82c4801b5c93 qemu-system-x86-7364 [012] 790.715352: kvm_msr: msr_write c0000080 = 0x3d01 qemu-system-x86-7364 [012] 790.715354: kvm_inj_exception: #GP (0x0) And the Xen-Crash message is: (XEN) Xen call trace: (XEN) ...

Are you sure? I thought this as well but considered a situation something like root anon_vma <--- rmap_walk starts here anon_vma a anon_vma b anon_vma c <--- an munmap/mmap/mprotect/etc here anon_vma d anon_vma e The rmap_walk takes the root lock and then locks a, b, c, d and e as it walks along. The mSomething event happens on c and takes the lock if rmap_walk gets there first, it takes the lock and the mSomething event waits until the ...

No, that can't happen. If you find an avc struct, it _will_ have a anon_vma pointer. So there's no point in testing for NULL. If some bug What makes this ok is the fact that it must be running under the RCU read lock, and anon_vma's thus cannot be released. My version of the code made that explicit. Yours does not, and doesn't even have comments about the fact that it needs to be called RCU read-locked. Tssk, tssk. Please don't just assume locking. Either lock it, or say "this must ...

On Wed, 5 May 2010 14:14:40 +0100 I'm sorry I couldn't catch all details but can I make a question ? Why seq_counter is bad finally ? I can't understand why we have to lock anon_vma with risks of costs, which is mysterious struct now. Adding a new to mm_struct is too bad ? Thanks, -Kame == From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> At treating rmap, there is no guarantee that "rmap is always correct" because vma->vm_start, vma->vm_pgoff are modified without any ...

Dumb question. I can't understand why we should use list_first_entry. I looked over the code. anon_vma_chain_link uses list_add_tail so I think that's right. But anon_vma_prepare uses list_add. So it's not consistent. How do we make sure list_first_entry returns deepest vma? Sorry if I am missing. -- Kind regards, Minchan Kim --

Good. If this returns NULL, it should oops when spin_lock(NULL->lock) This is very subtle in itself. RCU guarantees that the anon_vma exists but does it guarantee that it's the same one we expect and that it Sure. It was an oversight when merging what I had with what you posted up. -- Mel Gorman Part-time Phd Student Linux Technology Center University of Limerick IBM Dublin Software Lab --

The hatchet-job mutex is off the table so it's down to start-with-root-anon_vma-and-lock-in-order-when-walking-list (what I last posted) take-all-anon_vma-locks-when-changing-vmas (Rik's) use-seq-counter-to-spot-changes-to-VMAs-when-walking-list (Kamezawa-san's approach) Any strong preference? I still haven't read the other comments Linus made so I don't have a strong preference yet. Either Rik's or the patch I posted should be enough for -- Mel Gorman Part-time Phd Student ...

It should work, but only if we always take the deepest anon_vma lock. Not just in the migration code, but also in mmap, munmap, mprotect (for split_vma), expand_stack, etc... Otherwise we will still not provide exclusion of migrate vs. those events. I'm guessing that means changing both anon_vma_lock and page_lock_anon_vma to always take the deepest anon_vma lock - not introducing a new function that is only called by the migration code. -- All rights reversed --

Nothing. And we shouldn't care. If it's been freed and re-used, then all the anon_vma's (and vma's) associated with the original anon_vma (and page) have been free'd. And that, in turn, means that we don't really need to lock anything at all. The fact that we end up locking an anon_vma that _used_ to be the root anon_vma is immaterial - the lock won't _help_, but it shouldn't hurt either, since it's still a valid spinlock. Now, the above is only true as far as the anon_vma itself ...

On Thu, May 6, 2010 at 11:06 PM, Linus Torvalds Thanks, Linus and Mel. You understood my question correctly. :) My concern was following case. Child process does mmap new VMA but anon_vma is reused nearer child's VMA which is linked parent's VMA by fork. In that case, anon_vma_prepare calls list_add not list_add_tail. ex) list_add(&avc->same_anon_vma, &anon_vma->head); It means list_first_entry is the new VMA not old VMA and new VMA's root_avc isn't linked at parent's one. It means we ...

On Wed, 5 May 2010 14:21:48 +0300 Acked-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> --

On Wed, 5 May 2010 14:21:49 +0300 afacit the return value of cftype.unregister_event() is always ignored anyway. Perhaps it should be changed to void-returning, or fixed. --

Those four functions shouldn't grab port->lock because the caller in Either the interrupt routine should grab port->lock (see equivalent function in altera_jtaguart.c for an example), or this could be done lower down in altera_uart_rx_chars() and altera_uart_tx_chars(). (It's I'd suggest calling udelay(1) in each iteration of the loop. Perhaps the second loop is overkill. I was also going to suggest using the port->lock spin lock, but I'm not sure about the policy regarding that for ...

And altera_uart_get_mctrl(). (And for completeness, altera_uart_enable_ms() but the current implementation doesn't grab it anyway as it's just an empty function.) -- -=( Ian Abbott @ MEV Ltd. E-mail: <abbotti@mev.co.uk> )=- -=( Tel: +44 (0)161 477 1898 FAX: +44 (0)161 718 3587 )=- --

Hi Tobias, It's not enough. The spin lock is needed to protect modification of pp->imr and writing this modified value to the control register. Not all paths that modify pp->imr and the control register grab port->mutex beforehand (for example, the interrupt routine doesn't and can't). -- -=( Ian Abbott @ MEV Ltd. E-mail: <abbotti@mev.co.uk> )=- -=( Tel: +44 (0)161 477 1898 FAX: +44 (0)161 718 3587 )=- --

FYI, -tip testing started triggering this boot crash today (x86, 64-bit): bus: 'mdio_bus': add driver STe101p initcall ste10Xp_init+0x0/0x22 returned 0 after 52 usecs calling ksphy_init+0x0/0x5e @ 1 BUG: unable to handle kernel NULL pointer dereference at (null) IP: [<ffffffff813b1e98>] strcmp+0x6/0x21 PGD 0 Oops: 0000 [#1] SMP DEBUG_PAGEALLOC last sysfs file: CPU 1 Pid: 1, comm: swapper Not tainted 2.6.34-rc5 #3328 A8N-E/System Product Name RIP: 0010:[<ffffffff813b1e98>] ...

From: David Miller <davem@davemloft.net> Nevermind, Ingo please test this fix: diff --git a/drivers/net/phy/micrel.c b/drivers/net/phy/micrel.c index 0cd80e4..e67691d 100644 --- a/drivers/net/phy/micrel.c +++ b/drivers/net/phy/micrel.c @@ -32,6 +32,7 @@ static int kszphy_config_init(struct phy_device *phydev) static struct phy_driver ks8001_driver = { .phy_id = PHY_ID_KS8001, + .name = "Micrel KS8001", .phy_id_mask = 0x00fffff0, .features = PHY_BASIC_FEATURES, .flags = ...

We should put the MTD device even if we continue, because we did not I've added a comment and fixed your patch, and pushed to ubi-2.6.git. From: Marc Kleine-Budde <mkl@pengutronix.de> Date: Wed, 5 May 2010 10:17:25 +0200 Subject: [PATCH] UBI: init even if MTD device cannot be attached, if built into kernel UBI can be built into the kernel or be compiled as a kernel module. Further on the command line one can specify MTD devices to be attach to UBI while loading. In the current implementation ...

(cc dri-devel) On Wed, 5 May 2010 00:22:16 +0200 --

From: Matt Mackall <mpm@selenic.com> Yes the partial support handling is a thorny issue. But this patch set makes things better than they were before, because support over such devices didn't work at all previously. So I'll toss these patches into net-next-2.6, thanks everyone! --

memrar currently selects RAR_REGISTER, but I'll be happy to change that to depends on... -- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** --

That's fine, if I get David's ack on these. thanks, greg k-h --

Ah, true. Given the relative obscurity of using NUMA emulation to being with, it would probably benefit from being even more verbose as well. I'll rewrite the changelog and reply to this message with it. Thanks. --

Applied to tip:x86/urgent, thanks David! Ingo --

Overhead of interpreting bytecode plugin is written in. Or are you saying plugin is x86 assembly (32bit or 64bit btw?) and other arches will have to have in kernel x86 emulator to use the plugin (like some of them had for vgabios)? -- Gleb. --

Plugin is x86 or x64 machine code. You write the plugin in C and compile it using gcc/ld to get the object file, we map the relevant sections only to the OS space. NPA is a way of enabling passthrough of SR-IOV NICs with live migration support on ESX Hypervisor which runs only on x86/x64 hardware. It only supports x86/x64 guest OS. So we don't have to worry about other architectures. If NPA approach needs to be extended and adopted by other hypervisors then we have to take care of that. Today we ...

Which is simply not supportable for a cross-platform operating system like Linux. --

Hello, .34-rc6-git5 kernel: [ 1176.647885] pktgen 2.72: Packet Generator for packet performance testing. kernel: [ 1210.999903] ------------[ cut here ]------------ kernel: [ 1210.999914] WARNING: at net/sched/sch_generic.c:256 dev_watchdog+0xc1/0x129() kernel: [ 1210.999918] Hardware name: F3JC kernel: [ 1210.999921] NETDEV WATCHDOG: eth0 (r8169): transmit queue 0 timed out kernel: [ 1210.999924] Modules linked in: pktgen ppp_async crc_ccitt ipv6 ppp_generic slhc ...

Nope, I don't think so. Although it's probably worth putting a comment in lockdep_rcu_dereference() to state that lockdep can be used by multiple concurrent instances here, just in case someone ever consider adding code to this splat handler thinking lockdep is always only used by a single "splat" at a time. Thanks, -- Mathieu Desnoyers Operating System Efficiency R&D Consultant EfficiOS Inc. http://www.efficios.com --

Done! Thanx, Paul --

In theory, but not in practice. We don't have many rings, so the I think everyone is settled on 128 byte cache lines for the forseeable future, so it's not really an issue. Cheers, Rusty. --

To support this multiple device thingy and avoid the single device limit, I plan to do something like this, The ST driver would be platform_device - as it is already. ST's probe would do the ldisc registration, and also do the dev_set_drvdata of all the internal data (Tx queues, locks, list of protocols) on this device. Apart from this the BT, FM and GPS would also further be platform devices, and then, In the dev.platform_data of each of these BT, FM and GPS devices, I will enter the "parent ...

As a contrast, at least on some embedded SoCs without firmware limitations the low level sleep code for Runtime PM / CPU Idle is shared with system wide suspend. So the RAM is put into self refresh regardless of entering a half-deep CPU Idle state or entering suspend-to-memory. I've heard that newer SH-Mobile ARM hardware monitors the memory bus activity behind the scenes and can put the system RAM into self refresh automatically. On older parts we had to manage that from CPU idle ...

Ah, right. What was I thinking? I'll include your patch and push it to Ingo. Thanks. -- tejun --

On Tue, 4 May 2010 14:49:50 +0200 end_page_writeback() never got run against a page. My guess would be that NBD got its knickers in a knot. --

Direct calls into code were fine 10 years ago, but since then we got: - preempt notifiers - sw events - tracepoints Which add up to a lot more than just a plain call into code. Also, with the jump-optimizations we will have tracepoints that are _cheaper_ we already have an ABIs in place here - this would just properly unify and enumerate it. Ingo --

Yep. Also, most of the default distro kernels will have 3 sets of facilities: - preempt notifiers - tracepoints - sw events which is crazy. We can just standardize on using the tracepoint interface definition methods - they are properly typed, widespread and well-known enough to be perfect for this. ( They are also under intense optimization - the jump-tracepoints patch makes them probably even cheaper than preempt notifiers, in the off case. ) So lets get over this and ...

Hello, But if you put all those calls on top of TPs and apply static optimization to TPs (yeah, that can be applied directly to perf calls too but it's simpler to consolidate, IMHO), perf becomes zero-cost when disabled. Thanks. -- tejun --

What about the on case? kvm has preempt notifiers enabled all the time, and relies on their being fast. -- error compiling committee.c: too many arguments to function --

Hello, Ingo. Okay. Well, cmwq's usage of it isn't that hairy. It's used for two things. 1. to count the number of currently running workers which I think is fine for such hooks. 2. to wake up another worker. I don't think this falls in the category of "behavior/semantics" change. It just does try_to_wake_up() with the only caveat being it's called under the rq lock so requires a different API. There's nothing which alters the scheduler behavior in any pervasive way. ...

Pulled, thanks Robert! Ingo --

On Mon, 3 May 2010 22:30:01 -0700 These traces look odd - the kernel shouldn't be calling schedule() from below rmqueue_bulk()! If possible, please try a more recent kernel. If the problem occurs there and if we manage to fix it, the fix can be backported into whatever-kernel-version-you're-using. Can you get a better trace? The sysrq-T output would be good. That's known to work sufficiently well. Please avoid wordwrapping it when sending. --

I applied your patch to 2.6.33.3. It applied with a few offsets, but no errors: % patch -p1 < ~/pz_patch.diff patching file include/linux/perf_event.h Hunk #1 succeeded at 517 (offset -31 lines). patching file kernel/perf_event.c Hunk #1 succeeded at 349 (offset 7 lines). Hunk #2 succeeded at 1777 (offset -87 lines). The patch works with the test case I provided, but for some reason it breaks the normal operation of the libpfm4 "task" utility. If I put more than one event in a group, ...

Hrm, I see how that could be more elegant, but I'd rather avoid yet another layer of function pointer passing around. This code is already hard enough to review... --

I also see a hard lockup some time after connection to my companies wireless network. My private network does not seem to trigger that bug. Unfortunately the kernel is not able to switch back graphics, so I cannot tell if I see the same BUG - even if the problem description is the same. For reference, the patch above does not help on my T61p. It started soon after 2.6.34-rc4. Before and with rc4 I had to apply this http://patchwork.ozlabs.org/patch/49850/mbox/ patch to avoid the other ...

Now queued up, thanks. greg k-h --

I don't believe this patch is valid. The discovery order in the patch This is nice, I'll rework it and resend. Thanks a bunch. --

This is a fix for memcg-move-charge-of-file-pages.patch. After this patch, it passed the build test for all the configs with possible combinations of SWAP, SHMEM, MEM_RES_CTLR and MEM_RES_CTLR_SWAP. === From: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> Build fix for !CONFIG_SWAP case. mm/built-in.o: In function `mem_cgroup_get_shmem_target': (.text+0xde06): undefined reference to `swapper_space' Reported-by: Randy Dunlap <rdunlap@xenotime.net> Signed-off-by: Daisuke Nishimura ...

about @flags: Above says both of these: a) @flags will have future meaning (must be 0 for now) b) @flags may be one or more of: returns -EINVAL. ... --- ~Randy *** Remember to use Documentation/SubmitChecklist when testing your code *** --

I guess these three patches are for me. Do you want me to handle them? --

Hi Rafael, I think you can take 15 as it's an update to the comments. 16 and 17 are specific to checkpoint/restart. There are no other in-kernel uses, that we've identified, for them. So I was under the impression that it's inappropriate to add them without also adding the checkpoint/restart patches that use them. Thanks! Cheers, --

At Thu, 29 Apr 2010 23:00:42 +0200, Could you give alsa-info.sh output (run with --no-upload option)? I would revert in the worst case, but before that, I'd like to analyze the real issue. thanks, Takashi --

Ok, I forgot to pull from Linus' tree before merging locally, looks fine now. -- James Morris <jmorris@namei.org> --

Your tree is not up to date with respect to Linus's: you're missing the three patches he pulled in yesterday. David --

Yes, please wait a bit, I really don't know what is going on here, sorry. greg k-h --

mount -t hugetlbfs none /mnt/ echo xyz > /proc/sys/vm/nr_hugepages FC12 guest. --

On 05/06/10 06:10, Frederic Weisbecker wrote: > On Mon, May 03, 2010 at 02:12:00PM +0900, Hitoshi Mitake wrote: >> Hi Frederic, >> >> I added "info" subcommand to perf lock, >> this can be used as dumping metadata like thread or address of lock instances. >> "map" was removed because info should do the work of it. >> >> This will be useful not only for debugging but also for ordinary analyzing. >> >> I made this patch on perf/core of your tree, could you queue this? >> >> v2: ...

Well, we have #ifdef CONFIG_K8_NB static inline struct pci_dev *node_to_k8_nb_misc(int node) { return (node < num_k8_northbridges) ? k8_northbridges[node] : NULL; } #else static inline struct pci_dev *node_to_k8_nb_misc(int node) { return NULL; } #endif So it legitimately returns NULL in two cases: 1) if someone passes too large node 2) if CONFIG_K8_NB is unset 1) Either we assume that node will always be in the range (i.e. amd_get_nb_id() will never go crazy ...

From: Jiri Kosina <jkosina@suse.cz> Take a look at http://git.kernel.org/?p=linux/kernel/git/x86/linux-2.6-tip.git;a=commitdiff;h=f2b20e4... This does + if (num_k8_northbridges == 0) + return; in amd_check_l3_disable(). So if K8_NB has failed initializing for some reason, we never go near the pci devs and the node_to_k8_nb_misc() calls since we effectively disable the L3 functionality. Thus the NULL pointer checks you remove ...

Sorry. I've been caught up in work in the past few days. I can certainly help with the soft lockup if you are able to supply either a dump that includes all threads stuck in the NFS, or a (binary) wireshark dump that shows the NFSv4 traffic between the client and server around the time of the hang. Cheers Trond --

Hi Kyungmin, The original compiler error message was: mm/sparse.c: In function '__section_nr': mm/sparse.c:135: warning: 'root' is used uninitialized in this function Regards, Marcelo. --

Hello, I hadn't noticed that one (it's operating several layers above the piece of code I modified). So it seems we could do way with the #else clause completely (requiring the user to set 'nohlt' when debugging with JTAG)? Cheers Anders -- happy 'ptxdist' user --

Please do - i havent seen any other review feedback. Thanks, Ingo --

Commit-ID: 4f47b4c9f0b711bf84adb8c27774ae80d346b628 Gitweb: http://git.kernel.org/tip/4f47b4c9f0b711bf84adb8c27774ae80d346b628 Author: Eric W. Biederman <ebiederm@xmission.com> AuthorDate: Wed, 5 May 2010 13:22:25 -0700 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 6 May 2010 08:17:51 +0200 x86, acpi/irq: Define gsi_end when X86_IO_APIC is undefined My recent changes introducing a global gsi_end variable failed to take into account the case of using acpi on a system not ...

I just notice that this patch was never applied. Anyone against applying it? -- Steve --

Umm. Why? Adding an INF field will make it essentially impossible to use that thing as a compatibility function, and/or together with the legacy rlimit() system call. And that is going to be 99.999% of all uses. I don't think you understand how system calls work. People don't use Linux-only features. It's been shown over and over and over again. People use the standard interfaces, and they don't _have_ that INF field. We can extend ranges by just updating structures (see "stat()" ...

Thanks Artem, that's exactly the kind of feedback we need. For most people, module parameters are rare, and module removal is rare. So the amount of leak is less than the size of the code we would add to fix it. If this is hitting you, it clearly changes the priorities. I will include the patches now. Thanks! Rusty. --