From: Arnaldo Carvalho de Melo <acme@redhat.com> Newt has widespread availability and provides a rather simple API as can be seen by the size of this patch. The work needed to support it will benefit other frontends too. In this initial patch it just checks if the output is a tty, if not it falls back to the previous behaviour, also if newt-devel/libnewt-dev is not installed the previous behaviour is maintaned. Pressing enter on a symbol will annotate it, ESC in the annotation window ...

From: Arnaldo Carvalho de Melo <acme@redhat.com> Just like we do for pr_debug, so that we can have a single point where to redirect to the currently used output system, be it stdio or newt. Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- tools/perf/util/debug.h | 2 -- ...

From: Arnaldo Carvalho de Melo <acme@redhat.com> Will be used by the newt code too. Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- tools/perf/builtin-top.c | 2 +- tools/perf/perf.h | 4 ++++ 2 files changed, 5 insertions(+), 1 deletions(-) diff --git a/tools/perf/builtin-top.c ...

Hi, We recently converted a few wireless drivers to use request_firmware_nowait() in order to be able to load firmware from probe. We would like to continue with that so we can load the firmware before registering with any other subsystems, as we had discussed at the wireless summit last year. However, in actually trying this today, I noticed that there's a problem with this. I made my drivers all built in, and then I ended up with it not working because the firmware agent that was in my ...

From: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- tools/perf/util/symbol.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/perf/util/symbol.c b/tools/perf/util/symbol.c index 323c0ae..75cd468 100644 --- a/tools/perf/util/symbol.c +++ ...

From: Arnaldo Carvalho de Melo <acme@redhat.com> Will be used by the newt code too. Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- tools/perf/builtin-top.c | 2 +- tools/perf/perf.h | 4 ++++ 2 files changed, 5 insertions(+), 1 deletions(-) diff --git a/tools/perf/builtin-top.c ...

Argh, please don't consider this last patch (6/6), I messed up, will repost just this one in a few moments. - Arnaldo --

From: Arnaldo Carvalho de Melo <acme@redhat.com> Newt has widespread availability and provides a rather simple API as can be seen by the size of this patch. The work needed to support it will benefit other frontends too. In this initial patch it just checks if the output is a tty, if not it falls back to the previous behaviour, also if newt-devel/libnewt-dev is not installed the previous behaviour is maintaned. Pressing enter on a symbol will annotate it, enter again and the report ...

Linus, please pull ALSA updates for v2.6.34-rc1 from: git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6.git for-linus containing the following small fixes. Thanks! Takashi === Ralf Gerbig (1): ALSA: hda - Sound MSI fallout on a Asus mobo NVIDIA MCP55 Randy Dunlap (1): sound: fix opti92x-ad1848 build Takashi Iwai (2): ALSA: hda - Fix wrong model range check for ALC268 ALSA: hda - Fix input source elements of secondary ADCs on ...

Hi. Well, it turns out that this is indeed a regression, but I don't know from which kernel version. I reverted (not cleanly) this patch: http://patchwork.kernel.org/patch/28591 and now echo 1 > /sys/devices/platform/eeepc/cpufv enables the powersave mode and echo 0 > /sys/devices/platform/eeepc/cpufv enables the performance mode. Tested with the non-benchmark glxgears (275 frames/sec in powersave mode and 405 in performance mode) and stellarium (14 ...

The new btrfs util was just added on the subvol branch of btrfs-progs-unstable, and man/Makefile wants to build btrfs.8.gz, but the source is missing. Presumably it should either be added or the makefile adjusted. --

Is that from Intel, and will they clarify the text in the next version --

In case of kvasprintf() failure, we can leak old kobject name. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> --- diff --git a/lib/kobject.c b/lib/kobject.c index 8115eb1..1247c57 100644 --- a/lib/kobject.c +++ b/lib/kobject.c @@ -222,8 +222,10 @@ int kobject_set_name_vargs(struct kobject *kobj, const char *fmt, return 0; kobj->name = kvasprintf(GFP_KERNEL, fmt, vargs); - if (!kobj->name) + if (!kobj->name) { + kobj->name = old_name; return -ENOMEM; + } /* ewww... ...

This patch adds a line to the output of /proc/drivers/rtc to show the value of the RTSR register. It will be used to demonstrate a nasty initialization bug that will be fixed in the sequence. Signed-off-by: Marcelo Roberto Jimenez <mroberto@cpti.cetuc.puc-rio.br> --- drivers/rtc/rtc-sa1100.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/drivers/rtc/rtc-sa1100.c b/drivers/rtc/rtc-sa1100.c index e19ed0f..b04c837 100644 --- a/drivers/rtc/rtc-sa1100.c +++ ...

This patch fixes a nasty initialization condition on the RTSR register. Sometimes, bit 1 will wake up set, sometimes not. This can be seen by checking the value of the RTSR by typing '$ cat /proc/driver/rtc', which has been provided by the previous patch. If this bit is set, the command '$ cat /dev/rtc0' will lock the system in an endless interrupt routine calling loop. This patch fixes the issue both at sa1100_rtc_probe(), where it avoids a spurious interrupt from happening, and at ...

Would that be... me? ;) I _think_ my cards are ray_cs. What I know for 199% certainty is that they _are_ FHSS, i.e. pre-DSSS 802.11, as such chances are moderately high ;) (these are two or possibly even three PCMCIA cards, with a PCI-PCMCIA adapter) I don't have them here right now (and didn't find any meaningful traces on the PC that I used to use them with, back in <= 2002 or so...), but I can fetch them very soon and possibly test them (no promises ATM though). Out of interest: how ...

Hi Andrew, Did you ever get this ancient (2006!) issue resolved? I didn't see any responses on the list, and Lara is having a sort of similar issue: http://bugzilla.kernel.org/show_bug.cgi?id=15436 I think your problem was that Linux saw the Radeon card but didn't assign resources correctly, while in Lara's case, Linux doesn't see the card at all. I suspect that your problem has been resolved in the meantime and that Lara's might be broken hardware, but if yours is still broken, I'd ...

The commit 38b7827fc removed cpu_local_xx macros that lead to a compile error: implicit declaration of function 'cpu_local_wrap', when compiled for the MIPS Cavium Octeon processors. Revert back to old code. CC: Christoph Lameter <cl@linux-foundation.org> CC: Tejun Heo <tj@kernel.org> CC: David Daney <ddaney@caviumnetworks.com> Signed-off-by: Philby John <pjohn@mvista.com> --- arch/mips/include/asm/local.h | 25 +++++++++++++++++++++++++ 1 files changed, 25 insertions(+), 0 ...

Document the circular buffering capabilities available in Linux. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> --- Documentation/circular-buffers.txt | 231 ++++++++++++++++++++++++++++++++++++ Documentation/memory-barriers.txt | 20 +++ include/linux/circ_buf.h | 4 + 3 files changed, 255 insertions(+), 0 deletions(-) create mode 100644 Documentation/circular-buffers.txt diff --git ...

Document the circular buffering capabilities available in Linux. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> --- Documentation/circular-buffers.txt | 231 ++++++++++++++++++++++++++++++++++++ Documentation/memory-barriers.txt | 20 +++ include/linux/circ_buf.h | 4 + 3 files changed, 255 insertions(+), 0 deletions(-) create mode 100644 Documentation/circular-buffers.txt diff --git ...

Linus, Please pull the latest perf-fixes-for-linus git tree from: git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git perf-fixes-for-linus Thanks, Ingo ------------------> Arnaldo Carvalho de Melo (1): perf archive: Don't try to collect files without a build-id Frederic Weisbecker (5): perf lock: Drop the buffers multiplexing dependency perf: Remove pointless breakpoint union x86/hw-breakpoints: Remove the name field ...

I need something like this to support full resolution of internal LCD panel on Kohjinsha subnotebook. (Note: I basically made up the numbers, if someone knows how to generate them properly, let me know. It somehow works even on external display, but last few columns are missing.) Signed-off-by: Pavel Machek <pavel@ucw.cz> index 889cbe3..2c04a01 100644 --- a/drivers/video/geode/lxfb_core.c +++ b/drivers/video/geode/lxfb_core.c @@ -37,6 +37,9 @@ static int vt_switch; */ static ...

From: Arnaldo Carvalho de Melo <acme@redhat.com> [acme@mica linux-2.6-tip]$ perf record -a -f Fatal: Permission error - are you root? Consider tweaking /proc/sys/kernel/perf_event_paranoid. [acme@mica linux-2.6-tip]$ Suggested-by: Ingo Molnar <mingo@elte.hu> Cc: Frédéric Weisbecker <fweisbec@gmail.com> Cc: Mike Galbraith <efault@gmx.de> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> --- ...

* remove mm_types.h inclusion and the rest of mm stuff from sched.h to make sched.h inclusion cheaper. * move stuff to linux/mm.h * move init_mm declaration to mm_types.h to overcome interdependencies on sparc64 * remove linux/aio.h inclusion, it wasn't needed in sched.h, no need to move it. Patch not exactly well compiled tested as I'd like, still survived many configs across 10 arches and even UML. Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> --- ...

From: Chihau Chau <chihau@gmail.com> This is a code style cleanup related with spaces, like a lot of trailing withspaces, spaces before a TAB, to use "foo * bar" instead of "foo *bar" and to put a else sentence below close brace '}'. Signed-off-by: Chihau Chau <chihau@gmail.com> --- drivers/pci/pci.c | 51 +++++++++++++++++++++++++-------------------------- 1 files changed, 25 insertions(+), 26 deletions(-) diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c index 77b493b..d81d02e ...

Hi Catalin On my i686 machine, I have CONFIG_DMI=y and 12 entries in /sys/firmware/memmap/ like first one named "0" : $ grep . /sys/firmware/memmap/0/* /sys/firmware/memmap/0/end:0x9f3ff /sys/firmware/memmap/0/start:0x0 /sys/firmware/memmap/0/type:System RAM All 12 kobjects are allocated in DMI zone between c1001000 b .brk.dmi_alloc c1011000 B __brk_limit) see arch/x86/kernel/setup.c:124 #ifdef CONFIG_DMI RESERVE_BRK(dmi_alloc, 65536); #endif Still kmemleak seems to thing ...

Here are three small (but important!) fixes to GFS2. Steve. --

From: Sachin Prabhu <sprabhu@redhat.com> gfs2_lock() will skip locks on file which have mode set to 02666. This is a problem in cases where the mode of the file is changed after a process has obtained a lock on the file. Such a lock will be skipped and will result in a BUG in locks_remove_flock(). gfs2_lock() should skip the check for mandatory locks when unlocking a file. Signed-off-by: Sachin Prabhu <sprabhu@redhat.com> Signed-off-by: Steven Whitehouse <swhiteho@redhat.com> --- ...

Hi Grant, ^^^^^^ You need to remove that. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

If CONFIG_OF_FLATTREE is not set, then don't process the body of linux/of_fdt.h Signed-off-by: Grant Likely <grant.likely@secretlab.ca> --- include/linux/of_fdt.h | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/include/linux/of_fdt.h b/include/linux/of_fdt.h index a1ca92c..f0fdd1f 100644 --- a/include/linux/of_fdt.h +++ b/include/linux/of_fdt.h @@ -57,6 +57,7 @@ struct boot_param_header { __be32 dt_struct_size; /* size of the DT structure block */ }; ...

Only process contents of of_platform.h and of_device.h if CONFIG_OF_DEVICE is set. --- include/linux/of_device.h | 2 ++ include/linux/of_platform.h | 2 ++ 2 files changed, 4 insertions(+), 0 deletions(-) diff --git a/include/linux/of_device.h b/include/linux/of_device.h index d3a74e0..e7904a9 100644 --- a/include/linux/of_device.h +++ b/include/linux/of_device.h @@ -1,6 +1,7 @@ #ifndef _LINUX_OF_DEVICE_H #define _LINUX_OF_DEVICE_H +#ifdef CONFIG_OF_DEVICE #include ...

Document the circular buffering capabilities available in Linux. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> --- Documentation/circular-buffers.txt | 231 ++++++++++++++++++++++++++++++++++++ Documentation/memory-barriers.txt | 20 +++ include/linux/circ_buf.h | 4 + 3 files changed, 255 insertions(+), 0 deletions(-) create mode 100644 Documentation/circular-buffers.txt diff --git ...

Good point. David --

Hello, I'm not subscribed to this list, sorry if it's the wrong place to send this. I've been having problems with 2.6.34-rc1, when the system tries to access a lot of files (eg, if I recompile the kernel) it freezes for about 10 seconds or so and then I get a message like this in the sys.log: ata2: lost interrupt (Status 0x50) ata2.01: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x0 frozen ata2.01: failed command: WRITE DMA ata2.01: cmd ca/00:08:4c:37:6f/00:00:00:00:00/fd tag 0 dma 4096 ...

From: Jani Nikula <ext-jani.1.nikula@nokia.com> Use IS_ERR() instead of comparing to NULL. Signed-off-by: Jani Nikula <ext-jani.1.nikula@nokia.com> --- NOTE: I'm afraid I'm unable to test this; please consider this more a bug report than a complete patch. --- drivers/misc/c2port/core.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/drivers/misc/c2port/core.c b/drivers/misc/c2port/core.c index b5346b4..3276216 100644 --- a/drivers/misc/c2port/core.c +++ ...

This looks fine to me, good catch. Acked-by: Sonny Rao <sonnyrao@us.ibm.com> --

From: Jani Nikula <ext-jani.1.nikula@nokia.com> A number of functions in the driver core return ERR_PTR() values on error. Document this in the kernel-doc of the functions. Signed-off-by: Jani Nikula <ext-jani.1.nikula@nokia.com> --- drivers/base/class.c | 2 ++ drivers/base/core.c | 6 ++++++ drivers/base/platform.c | 6 ++++++ 3 files changed, 14 insertions(+), 0 deletions(-) diff --git a/drivers/base/class.c b/drivers/base/class.c index 0147f47..9c6a0d6 100644 --- ...

You can include slab.h only for UP case. Since everyone tests on allmodconfig which has SMP=y, configuration will be more strict wrt headers, and compile breakages amount negligible. --

No, not really. Well, logical for people quickly prototyping code on their own machines, but perf has grown beyond that. However, I don't like my quick fix either, maybe I can come-up with something cleaner. John --

Commit-ID: 3d07467b7aa91623b31d7b5888a123a2c8c8e9cc Gitweb: http://git.kernel.org/tip/3d07467b7aa91623b31d7b5888a123a2c8c8e9cc Author: Peter Zijlstra <a.p.zijlstra@chello.nl> AuthorDate: Wed, 10 Mar 2010 17:07:24 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 15:21:50 +0100 sched: Fix pick_next_highest_task_rt() for cgroups Since pick_next_highest_task_rt() already iterates all the cgroups and is really only interested in tasks, skip over the !task ...

Commit-ID: ab3b3aa5dd01b3aaa6b15caee113b21b1b6520c4 Gitweb: http://git.kernel.org/tip/ab3b3aa5dd01b3aaa6b15caee113b21b1b6520c4 Author: Dan Carpenter <error27@gmail.com> AuthorDate: Sat, 6 Mar 2010 14:17:52 +0300 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 13:59:59 +0100 sched: Cleanup: remove unused variable in try_to_wake_up() We haven't used the "orig_rq" variable since 055a00865d "Fix/add missing update_rq_clock() calls" Signed-off-by: Dan Carpenter ...

--

This may or may not be connected to the ata-timeout issue but it is in the same place in my dmesg: kmemleak: 17 new suspected memory leaks (see /sys/kernel/debug/kmemleak) See attached file for all of the 17 issues. Let me know if you need any other information. kmemleak: 17 new suspected memory leaks (see /sys/kernel/debug/kmemleak) ata1: clearing spurious IRQ ata1: lost interrupt (Status 0x50) ata1.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen ata1.00: failed command: ...

The sched_switch trace event displays erroneous character codes of task states, after a new task state was added in the scheduler code but omitted to add in the trace event code. Define character codes of task states individually. In addition, define task state descriptions needed in /proc at the same place. This will help to keep the task state bits, characters and descriptions in sync should they ever need to be changed again. Signed-off-by: Carsten Emde <C.Emde@osadl.org> Index: ...

There are still outstanding comments. I've commented several times that this patch needs to be posted with the related patches so people can actually try it out. Regards, Tony --

The setup you describe would only work with SSDs. HDDs would seek themselves to death. The problem is the RAID-0 over the 4 partitions. At that point you would need, instead of the 4 partitions, something that is round-robin. So that the mapping of the (physical) blocks from the upper to the lower would be effectivly linear/unchanged. AFAIK something like that is (currently) not possible. Bis denn -- Real Programmers consider "what you see is what you get" to be just as ...

Nope. What i meant is: Let say you had a block-device which has 16 blocks: 0-15 With the OPs description the blocks would be distributed like this: Part 0: 00 01 02 03 Part 1: 04 05 06 07 Part 2: 08 09 10 11 Part 3: 12 13 14 15 What you need is a distribution like this: Device 0: 01 05 09 13 Device 1: 02 06 10 14 Device 2: 03 07 11 15 Device 3: 04 08 12 16 IOW: Blocks % 4 == 0 on device 0 Blocks % 4 == 1 on device 1 Blocks % 4 == 2 on device 2 Blocks % 4 == 3 on device 3 I ...

The old behavior doesnt look very logical though, right? Ingo --

Without this change, the install path is relative to prefix/DESTDIR where prefix is automatically set to $HOME This can produce unexpected results. For example make -C tools/perf DESTDIR=/home/jkacur/tmp install-man creates the directory: /home/jkacur/home/jkacur/tmp/share/... instead of the expected: /home/jkacur/tmp/share/... Signed-off-by: John Kacur <jkacur@redhat.com> --- tools/perf/Documentation/Makefile | 4 +++- tools/perf/Makefile | 4 +++- 2 files ...

Your Email just won you ($500,000.00 US Dollars ATM CARD) You are to contact Mr. Abdul Razi With your,Names....Age...Country....Tell.... --

Hi, I Have following to scripts: a.sh #!/bin/sh echo "It's a.sh and b.sh: #! ./b.sh echo "It's b.sh" As per execve man page, script interpreter should not be script itself. When I run it on my Fedora 8 x86_64 box (with stock kernel, never updated) under strace I got following: strace -f -e execve setarch x86_64 ./b.sh execve("/usr/bin/setarch", ["setarch", "x86_64", "./b.sh"], [/* 23 vars */]) = 0 execve("./b.sh", ["./b.sh"], [/* 23 vars */]) = -1 ENOEXEC (Exec format ...

Commit 7d3cc8b tried to keep bounds.h and asm-offsets.h during make clean by filtering these out of $(clean-files), but they are listed in $(targets) and $(always) and thus removed automatically. Introduce a new $(no-clean-files) variable to really skip such files in Makefile.clean. Signed-off-by: Michal Marek <mmarek@suse.cz> --- Documentation/kbuild/makefiles.txt | 7 +++++++ Kbuild | 4 ++-- scripts/Makefile.clean | 2 ++ 3 files changed, ...

This sort of thing is caused by acquiring the same lock with softirq (AKA BH) blocked and not, which can result in self-deadlock. There was such a bug in the RCU lockdep stuff in -tip, but it has long since been fixed. If you were seeing that bug, rcu_do_batch() would be on the stack, which it does not appear to be. So does your patch involve the usbfs_mutex? Or attempt to manipulate vfs/fs state from withing networking softirq/BH context? --

sched: remove FAIR_SLEEPERS feature Our preemption model relies too heavily on sleeper fairness to disable it without dire consequences. Remove the feature, and save a branch or two. Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> --- kernel/sched_fair.c | 2 +- kernel/sched_features.h | 7 ------- 2 files changed, 1 insertion(+), 8 deletions(-) Index: ...

sched: remove AFFINE_WAKEUPS feature Disabling affine wakeups is too horrible to contemplate. Remove it. Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> --- kernel/sched_fair.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) Index: linux-2.6/kernel/sched_fair.c =================================================================== --- ...

sched: remove avg_wakeup. Testing the load which led to this heuristic (nfs4 kbuild) shows that it has outlived it's usefullness. With intervening load balancing changes, I cannot see any difference with/without, so recover there fastpath cycles. Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> --- include/linux/sched.h | 3 --- kernel/sched.c | 26 ...

sched: cleanup and optimize clock updates Now that we no longer depend on the clock being updated prior to enqueueing on migratory wakeup, we can clean up a bit, placing calls to update_rq_clock() exactly where they are needed, ie on enqueue, dequeue and schedule events. In the case of a freshly enqueued task immediately preempting, we can skip the update during preemption, as the clock was just updated by the enqueue event. We also save an unneeded call during a migratory wakeup by not ...

sched: fix select_idle_sibling() Don't bother with selection when the current cpu is idle. Recent load balancing changes also make it no longer necessary to check wake_affine() success before returning the selected sibling, so we now always use it. Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> --- kernel/sched_fair.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 ...

sched: remove WAKEUP_SYNC feature This feature never earned it's keep, remove it. Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> --- kernel/sched_fair.c | 4 ---- kernel/sched_features.h | 5 ----- 2 files changed, 9 deletions(-) Index: linux-2.6/kernel/sched_fair.c =================================================================== --- ...

sched: remove SYNC_WAKEUPS feature Sync wakeups are critical functionality with a long history. Remove it, we don't need the branch or icache footprint. Signed-off-by: Mike Galbraith <efault@gmx.de> Cc: Ingo Molnar <mingo@elte.hu> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> LKML-Reference: <new-submission> --- kernel/sched.c | 3 --- kernel/sched_features.h | 8 -------- 2 files changed, 11 deletions(-) Index: ...

Commit-ID: b42e0c41a422a212ddea0666d5a3a0e3c35206db Gitweb: http://git.kernel.org/tip/b42e0c41a422a212ddea0666d5a3a0e3c35206db Author: Mike Galbraith <efault@gmx.de> AuthorDate: Thu, 11 Mar 2010 17:15:38 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 18:32:50 +0100 sched: Remove avg_wakeup Testing the load which led to this heuristic (nfs4 kbuild) shows that it has outlived it's usefullness. With intervening load balancing changes, I cannot see any ...

Commit-ID: e12f31d3e5d36328c7fbd0fce40a95e70b59152c Gitweb: http://git.kernel.org/tip/e12f31d3e5d36328c7fbd0fce40a95e70b59152c Author: Mike Galbraith <efault@gmx.de> AuthorDate: Thu, 11 Mar 2010 17:15:51 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 18:32:50 +0100 sched: Remove avg_overlap Both avg_overlap and avg_wakeup had an inherent problem in that their accuracy was detrimentally affected by cross-cpu wakeups, this because we are missing the ...

Commit-ID: 8b911acdf08477c059d1c36c21113ab1696c612b Gitweb: http://git.kernel.org/tip/8b911acdf08477c059d1c36c21113ab1696c612b Author: Mike Galbraith <efault@gmx.de> AuthorDate: Thu, 11 Mar 2010 17:17:16 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 18:32:51 +0100 sched: Fix select_idle_sibling() Don't bother with selection when the current cpu is idle. Recent load balancing changes also make it no longer necessary to check wake_affine() success before ...

Commit-ID: 5ca9880c6f4ba4c84b517bc2fed5366adf63d191 Gitweb: http://git.kernel.org/tip/5ca9880c6f4ba4c84b517bc2fed5366adf63d191 Author: Mike Galbraith <efault@gmx.de> AuthorDate: Thu, 11 Mar 2010 17:17:17 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 18:32:52 +0100 sched: Remove FAIR_SLEEPERS feature Our preemption model relies too heavily on sleeper fairness to disable it without dire consequences. Remove the feature, and save a branch or ...

Commit-ID: c6ee36c423c3ed1fb86bb3eabba9fc256a300d16 Gitweb: http://git.kernel.org/tip/c6ee36c423c3ed1fb86bb3eabba9fc256a300d16 Author: Mike Galbraith <efault@gmx.de> AuthorDate: Thu, 11 Mar 2010 17:16:43 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 18:32:53 +0100 sched: Remove SYNC_WAKEUPS feature Sync wakeups are critical functionality with a long history. Remove it, we don't need the branch or icache footprint. Signed-off-by: Mike Galbraith ...

Commit-ID: beac4c7e4a1cc6d57801f690e5e82fa2c9c245c8 Gitweb: http://git.kernel.org/tip/beac4c7e4a1cc6d57801f690e5e82fa2c9c245c8 Author: Mike Galbraith <efault@gmx.de> AuthorDate: Thu, 11 Mar 2010 17:17:20 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 18:32:53 +0100 sched: Remove AFFINE_WAKEUPS feature Disabling affine wakeups is too horrible to contemplate. Remove the feature flag. Signed-off-by: Mike Galbraith <efault@gmx.de> Signed-off-by: Peter ...

This patch adds a new MFD driver for the RDC321x southbridge. This southbridge is always present in the RDC321x System-on-a-Chip and provides access to some GPIOs as well as a watchdog. Access to these two functions is done using the southbridge PCI device configuration space. Signed-off-by: Florian Fainelli <florian@openwrt.org> --- diff --git a/drivers/mfd/Kconfig b/drivers/mfd/Kconfig index 951fa9b..8cc8218 100644 --- a/drivers/mfd/Kconfig +++ b/drivers/mfd/Kconfig @@ -398,6 +398,15 @@ ...

This patch adds a new GPIO driver for the RDC321x SoC GPIO controller. Signed-off-by: Florian Fainelli <florian@openwrt.org> --- diff --git a/drivers/gpio/Kconfig b/drivers/gpio/Kconfig index fee678f..dd10eb8 100644 --- a/drivers/gpio/Kconfig +++ b/drivers/gpio/Kconfig @@ -275,6 +275,14 @@ config GPIO_TIMBERDALE ---help--- Add support for the GPIO IP in the timberdale FPGA. +config GPIO_RDC321X + tristate "RDC R-321x GPIO support" + depends on PCI && GPIOLIB + select ...

This file is going to be replaced with a cleaner version with the adding of a MFD driver for the southbridge. Signed-off-by: Florian Fainelli <florian@openwrt.org> --- diff --git a/arch/x86/include/asm/rdc321x_defs.h b/arch/x86/include/asm/rdc321x_defs.h deleted file mode 100644 index c8e9c8b..0000000 --- a/arch/x86/include/asm/rdc321x_defs.h +++ /dev/null @@ -1,12 +0,0 @@ -#define PFX "rdc321x: " - -/* General purpose configuration and data registers */ -#define RDC3210_CFGREG_ADDR ...

Use DECLARE_EVENT_CLASS to remove duplicate code: text data bss dec hex filename 14655 2732 15 17402 43fa i915_trace_points.o.orig 11625 2732 10 14367 381f i915_trace_points.o 8 events are converted: i915_gem_object: i915_gem_object_{unbind, destroy} i915_gem_request: i915_gem_request_{complete, retire, wait_begin, wait_end} i915_ring: i915_ring_{wait_begin, wait_end} No functional change. Signed-off-by: Li Zefan ...

Make some comments consistent with the code. Signed-off-by: Li Zefan <lizf@cn.fujitsu.com> --- include/trace/ftrace.h | 33 +++++++++++++++++++-------------- 1 files changed, 19 insertions(+), 14 deletions(-) diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 25ab56f..9d05eab 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -154,9 +154,11 @@ * * field = (typeof(field))entry; * - * p = get_cpu_var(ftrace_event_seq); + * p = ...

On some platforms like DM355, the number of EDMA parameter slots available for EDMA_SLOT_ANY usage are few. In such cases, if MMC/SD uses 16 slots for each instance of MMC controller, then the number of slots available for other modules will be very few. By passing the number of EDMA slots to be used in MMC driver from platform data, EDMA slots available for other purposes can be controlled. Signed-off-by: Sudhakar Rajashekhara <sudhakar.raj@ti.com> --- ...

If a board doesn't setup pdata->nr_sg it will be zero, leaving Since host->nr_sg is unsigned, you get lucky and fix it up here, but for readability, this not too clean and should be more thorough. Wrapping the above in 'if (pdata->nr_sg)' is a more standard way of handling optional platform_data paramaters. Kevin --

From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> memcg's oom waitqueue is a system-wide wait_queue (for handling hierarchy.) So, it's better to add custom wake function and do flitering in wake up path. This patch adds a filtering feature for waking up oom-waiters. Hierarchy is properly handled. Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> --- mm/memcontrol.c | 61 ++++++++++++++++++++++++++++++++++++++++---------------- 1 file changed, 44 insertions(+), ...

From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> This adds a feature to disable oom-killer for memcg, if disabled, of course, tasks under memcg will stop. But now, we have oom-notifier for memcg. And the world around memcg is not under out-of-memory. memcg's out-of-memory just shows memcg hits limit. Then, administrator or management daemon can recover the situation by - kill some process - enlarge limit, add more swap. - migrate some tasks - remove file cache on tmps ...

Thank you. On Thu, 11 Mar 2010 16:47:00 +0200 Ah...sorry for garbages..I'll clean these up. Thanks, -Kame --

Export perf_trace_regs and perf_arch_fetch_caller_regs since module will use these Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com> --- arch/x86/kernel/cpu/perf_event.c | 1 + kernel/trace/trace_event_perf.c | 1 + 2 files changed, 2 insertions(+), 0 deletions(-) diff --git a/arch/x86/kernel/cpu/perf_event.c b/arch/x86/kernel/cpu/perf_event.c index 4a0514d..8700643 100644 --- a/arch/x86/kernel/cpu/perf_event.c +++ b/arch/x86/kernel/cpu/perf_event.c @@ -1697,3 +1697,4 ...

Pretty much all modular tracepoint users I think, I just got this from a pp64_defconfig build: ERROR: "perf_trace_regs" [fs/xfs/xfs.ko] undefined! ERROR: ".perf_arch_fetch_caller_regs" [fs/xfs/xfs.ko] undefined! ERROR: "perf_trace_regs" [arch/powerpc/platforms/cell/spufs/spufs.ko] undefined! ERROR: ".perf_arch_fetch_caller_regs" [arch/powerpc/platforms/cell/spufs/spufs.ko] undefined! --

Yeah, it's in -tip. Thanks guys for this fix. I think I'm unable to remind there are modules in the kernel, I'm not counting anymore the number of times I forget to export symbols for modules. May be I should start to stick posters with photos of modules entitled "I want to believe" everywhere in my flat. Or perhaps I'm going to buy electronic glasses that display modules advertizing in the street. I'm not sure yet but I'll find a way. --

Currently we can only get the cpu_stat of whole guest as one. This patch enhanced cpu_stat with more detail, has guest_system and guest_user cpu time statistics with a little overhead. Signed-off-by: Sheng Yang <sheng@linux.intel.com> --- This draft patch based on KVM upstream to show the idea. I would split it into more kernel friendly version later. The overhead is, the cost of get_cpl() after each exit from guest. Comments are welcome! arch/x86/kvm/x86.c | 10 ...

Yes, that's my concern too(though nested vmcs/vmcb read already too expensive, they should be optimized...). The other concern is, perf alike mechanism would Yanmin have already implement "perf kvm" to support this. We are just arguing if a normal top-alike mechanism is necessary. I am also considering to make it a feature that can be disabled. But seems it make things complicate and result in uncertain cpustat output. -- regards Yang, Sheng --

Another method is to add sysctl entry, such like /proc/sys/kernel/collect_guest_utilization, Here is the story why Sheng worked out the patch. Some guys work on KVM performance. They want us to extend top to show guest utilization info, such like guest kernel and guest userspace cpu utilization. With the new tool, they could find which VM (mapping with qemu process id) consumes too much cpu time in host space (including kernel and userspace), and compare them with guest kernel/userspace. That ...

You have won,confirm receipt (£891,934.00) by sending your name,address,age,phone number etc to cgclaims15@9.cn<mailto:cgclaims15@9.cn> --

From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Add support for hvm_op hypercall. Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com> Signed-off-by: Sheng Yang <sheng@linux.intel.com> --- arch/x86/include/asm/xen/hypercall.h | 6 ++ include/xen/hvm.h | 23 +++++++ include/xen/interface/hvm/hvm_op.h | 72 ++++++++++++++++++++++ include/xen/interface/hvm/params.h | 111 ++++++++++++++++++++++++++++++++++ 4 files changed, 212 ...

We mapped each IOAPIC pin to a VIRQ, so that we can deliver interrupt through these VIRQs. We used X86_PLATFORM_IPI_VECTOR as the notification vector for hypervisor to notify guest about the event. The patch also enabled SMP support, then we can support IPI through evtchn as well. When this feature is enabled, we would relay on Xen PV timer for clockevent, rather than other hardware emulated ones. Then we don't use IOAPIC/LAPIC, eliminated the overhead brought by unnecessary VMExit ...

The PV extended HVM(once known as Hybrid) is started from real mode like HVM guest, but also with a component based PV feature selection(e.g. PV halt, PV timer, event channel, then PV drivers). So guest can takes the advantages of both H/W virtualization and Para-Virtualization. This patch introduced the PV extension of HVM guest initialization. Guest would detect the capability using CPUID 0x40000002.edx, then call HVMOP_enable_pv hypercall to enable pv support in ...

xen_guest_init() would setup the environment. Signed-off-by: Sheng Yang <sheng@linux.intel.com> --- arch/x86/include/asm/xen/hypervisor.h | 6 ++++++ arch/x86/kernel/setup.c | 4 ++++ 2 files changed, 10 insertions(+), 0 deletions(-) diff --git a/arch/x86/include/asm/xen/hypervisor.h b/arch/x86/include/asm/xen/hypervisor.h index 396ff4c..13e089a 100644 --- a/arch/x86/include/asm/xen/hypervisor.h +++ b/arch/x86/include/asm/xen/hypervisor.h @@ -37,4 +37,10 @@ extern ...

Agreed, this makes perfect sense. --

This patch fixes a bug in 2.6.33 X86 event scheduling whereby all counts are bogus as soon as events need to be multiplexed because the PMU is overcommitted. The code in hw_perf_enable() was causing multiplexed events to accumulate collected counts twice causing bogus results. This is demonstrated on AMD Barcelona with the example below. First run, no conflict, you obtain the actual counts. Second run, PMU overcommitted, multiplexing, all events are over-counted. Third run, patch applied, ...

Commit-ID: 45e16a6834b6af098702e5ea6c9a40de42ff77d8 Gitweb: http://git.kernel.org/tip/45e16a6834b6af098702e5ea6c9a40de42ff77d8 Author: Peter Zijlstra <a.p.zijlstra@chello.nl> AuthorDate: Thu, 11 Mar 2010 13:40:30 +0100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 15:21:28 +0100 perf, x86: Fix hw_perf_enable() event assignment What happens is that we schedule badly like: <...>-1987 [019] 280.252808: x86_pmu_start: event-46/1300c0: idx: 0 <...>-1987 ...

Dear John, Thanks for your answer. This is informative. With best regards, Yours sincerely, Srinivas Nayak My Home page: http://www.mathmeth.com/sn/ My Blog: http://srinivas-nayak.blogspot.com Your Mail works best with the New Yahoo Optimized IE8. Get it NOW! http://downloads.yahoo.com/in/internetexplorer/ --

The fix is in -tip, commit 99ee4ca746dda71326db7645463b4075ac1d665c. This is an initialization-time use of rcu_dereference(), so no other task has a reference to this data. Hence it is constant. Other uses of this code operate on shared data structures, which might change at any time. --

Hi Michal, After merging the kbuild tree, a build like "make modules modules_install" failed like this: cp: cannot stat `/home/sfr/kernels/ntest.obj/modules.builtin': No such file or directory Caused by commit a6c366324cacd5f71107dd01362b995a1c67b1ad ("kbuild: Do not unnecessarily regenerate modules.builtin"). I have reverted that commit for today. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

Hi Tony, OK, thanks. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

I'll split my patch in two parts, and let Tomi queue this fix. Regards, Tony --

drivers/xen/events.c:31: WARNING: Use #include <linux/ptrace.h> instead of <asm/ptrace.h> drivers/xen/events.c:76: ERROR: open brace '{' following struct go on the same line drivers/xen/events.c:509: WARNING: line over 80 characters drivers/xen/events.c:581: ERROR: space required before the open parenthesis '(' drivers/xen/events.c:585: ERROR: space required before the open parenthesis '(' drivers/xen/events.c:590: ERROR: space required before the open parenthesis '(' drivers/xen/events.c:595: ...

Hi Dominik, Thanks. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

Hey Stephen, oops, I'm very sorry, seems like I pushed the wrong tree to master... fixed it; so should be fine for tomorrow. Best, Dominik --

Now, a dedicated HEST tabling parsing code is used for PCIE AER firmware_first setup. It is rebased on general HEST tabling parsing code of APEI. The firmware_first setup code is moved from PCI core to AER driver too, because it is only AER related. Signed-off-by: Huang Ying <ying.huang@intel.com> --- drivers/acpi/Makefile | 1 drivers/acpi/hest.c | 135 ------------------------------------- drivers/pci/pcie/aer/aerdrv.h | 9 ++ ...

not sure what happened... but wireless-next-2.6.git 86415d43efd4f7093979cfa8a80232114266f1a4 is an older version of my patch. i have sent one which applies cleanly later on. your fix is correct, though. bruno --

Hi Bruno, The tree I call wireless-current usually only contains bug fixes, the Thanks for the confirmation. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> with the same comment about ideally wanting someone who worked on the driver previously to check stuff. --

It does not 'need' to be in .34 but if the fix is trivial enough then you could give it a try? Ingo --

If it needs to go in before .35, or it should go through Ingo's trees, I have no objection. thanks, greg k-h --

Thanks Greg! Ingo --

Yes, this works for most of the ZTE modems with exception for 628 IIRC, which needs 5553424312345678000000000000061b000000030000000000000000000000 instead. -- Jiri Kosina SUSE Labs, Novell Inc. --

[Empty message]

or snprintf --

Yes, please change it if you have no problem. Thanks. -- OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> --

Looks good to me. Acked-by: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> (From 1/4 with previous patch, I guess it's series of patches) Or I should take this patch into my tree? -- OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> --

On Wed, Mar 10, 2010 at 11:09 PM, Miguel Ojeda Cc'ing the original committers. --

You are right, I can choose between 128M/256M in BIOS too. The BIOS config is the following: IGD Aperture Size: 128 MB DVMT MODE: FIXED I see. I just guessed by seeing that gtt_total_size is the double than older num_entries and that "default" gtt_map_size is 256 instead of 128. Then I checked the webpage and lspci, I wrote the fix and it It also occurs in other applications/ways. Maybe you could try to do some scrolling (I recall it also triggered it) or try to open other browsers ...

I vote for LMB too, it's already used across multiple architectures and has proven to be quite versatile. If x86 is concerned about consolidation, then it's a good a place to start as any, particularly since it's not even that conceptually different from how the e820 maps are used. This may come as a surprise, but if this had actually been brought up on linux-arch instead of buried in -tip and magically showing up in -next everyone involved could have saved a lot of time. --

Thanks for catching and fixing this. -- Three Cheers, Balbir --

On Wed, 10 Mar 2010 20:34:12 +0200 The extra performance was just a "bug": http://www.mail-archive.com/pgsql-performance@postgresql.org/msg34841.html "[This change] is required for safe behavior with volatile write caches on drives. You could mount with -o nobarrier and [the performance drop] would go away, but a sequence like write->fsync->lose power->reboot may well find your file without the data that you synced, if the drive had write caches enabled. If you know you have no write cache, ...

commit ca03770(perf, x86: Add PEBS infrastructure) introduces a new function validate_event that calls x86_pmu.get_event_constraints. static int validate_event(struct perf_event *event) { ... c = x86_pmu.get_event_constraints(fake_cpuc, event); ... } So we need to add .get_event_constraints to p4_pmu. diff --git a/arch/x86/kernel/cpu/perf_event_p4.c b/arch/x86/kernel/cpu/perf_event_p4.c index 4eb79b1..99a2a7c 100644 --- a/arch/x86/kernel/cpu/perf_event_p4.c +++ ...

Patch is updated (and latest -tip/master is taken into account as well). Robert, introducing additional logic instead of function pointers would lead to a valuable code duplication. But I'll think about it anyway. Perhaps there is a way (still) to implement it with a minimal former code intersection. Also note that the patch squash really a small change into former code, which make pretty easy to step back if we will need it (for some reason). And naming issue of header file is ...

yes, it's expected since it's not implemented yet, but there yeah, wide testing would be great. Thanks! -- Cyrill --

Perhaps something like the patch below (tested with kvm)? With this patch we will actually waste ~4/8 bytes per PMU (intel,amd,p6) since this call hits on p4 only, so I think perhaps better to use one x86 scheduler hook instead of empty schedule_events() in PMU, hmm? --- x86,perf: Fix NULL deref on not assigned x86_pmu In case of not assigned x86_pmu and software events NULL dereference may being hit via x86_pmu::schedule_events method. Fix it by calling x86_pmu::schedule_events only if ...

As far as I see it'll not update tstamp_running then (in x86_event_sched_in). Or I miss somethig? -- Cyrill --

OK, thanks, I see what you mean. Will cook patch shortly. -- Cyrill --

Dave, this modification sounds promising. Will do them during the weekend. Also Christoph mentioned some pending patches for fdatasync and NFS metadata updates in his XFS status report from February, that sounded _really_ exciting. Happily awaiting these bits in the stable universe ;-) Thanks, Pete --

Yes, already done. Hopefully my users will notice.. As I've upgraded this server and the clients only two weeks ago, calming things down has highest priority. Switching kernel versions in production systems is always painful, thus I try to avoid that, but this time I already needed to roll my own kernel for the clients due to some aufs2 vs. apparmor disharmony. That led to the loss of the latter - I can live without apparmor, but certainly not without a reliable layered filesystem¹. ...

Is this 64-bit specific like Sheng's version? I also asked him to move the setup code out of enlighten.c into a new There's no point testing this; if we don't use hypercalls beyond the I think you can combine these into the same if(). Also put it in a --

Please see my comments on Sheng's version of this patch and his updates, specifically regarding the xen_hvm_domain() tests in xen_evtchn_do_upcall(). Thanks, --

Hello Ogawa, We could perhaps reduce the disadvantage by moving the logic (test first with the trailing dots, and then without if needed) into fat_search_long instead of putting it in vfat_find as my patch proposal does. This way the name to find would only be compared once to unrelated entries, instead of possibly twice as my patch does. But therefore I need your help : fat_search_long isn't easy to read/modify. The logic could be modified as such : - if the searched name contains ...

I mean, when readahead windows A and B are submitted in one IO -- let's call it AB -- commit 65a80b4c61 will explicitly unplug on doing readahead C. While in your trace, the unplug appears on AB. The 68% improvement is very impressive. Wondering if commit 65a80b4c61 They are reasonable assumptions. However I'm not sure if this unconditional unplug will defeat CFQ's anticipatory logic -- if there are any. You know commit 65a80b4c61 is more about a *defensive* protection against the rare ...

Andrew/Greg, shall we push the patch to .32 stable? You are welcome~ Thanks, Fengguang --

On Wed, 10 Mar 2010 12:03:37 +0100 Why can't use dma_mapping_error()? --

Applied, thanks. -- error compiling committee.c: too many arguments to function --

Commit-ID: a12b51c478899fe0b7e874a559b05ba35f1128ee Gitweb: http://git.kernel.org/tip/a12b51c478899fe0b7e874a559b05ba35f1128ee Author: Paul Mackerras <paulus@samba.org> AuthorDate: Wed, 10 Mar 2010 20:36:09 +1100 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 13:36:53 +0100 perf tools: Fix sparse CPU numbering related bugs At present, the perf subcommands that do system-wide monitoring (perf stat, perf record and perf top) don't work properly unless the online ...

Sorry, I can't trigger it today, either with SLAB or SLUB. --

No, I did a git pull, but it looks like only some btrfs updates... --

Thanks! --

Put all statistic fields of sched_entity in one struct, sched_statistics, and embed it into sched_entity. This change allows to memset the sched_statistics to 0 when needed (for instance when forking), avoiding bugs of non initialized fields. Signed-off-by: Lucas De Marchi <lucas.de.marchi@gmail.com> --- include/linux/sched.h | 54 ++++++++++++-------------- kernel/sched.c | 47 ++++------------------ kernel/sched_debug.c | 101 ++++++++++++++++++------------------------------ ...

this is for the working setup: $ ls /sys/block/ fd0 hdb loop0 loop2 loop4 loop6 md0 md2 md4 hda hdc loop1 loop3 loop5 loop7 md1 md3 md5 I don't have one for the notworking side of things as that'd require taking the server down, opening it, cursing heavily as I tried to manipulate a tiny jumper in the worst location possible (thank god From /var/log/messages I'm assuming you want the dmesg output. That's attached. I'll provide both the libata and ide subsystem ones as ...

On Wed, 10 Mar 2010 00:00:31 +0100 maybe. But in this 2 years, one of our biggest concerns was the performance. So, we do something complex in memcg. But complex-locking is , yes, complex. Hmm..I don't want to bet we can fix locking scheme without something complex. Thanks, -Kame --

Not sure though, while this does the accounting the actual writeout is still !cgroup aware and can definately impact performance negatively by shrinking too much. --

On Thu, 11 Mar 2010 18:25:00 +0900 Hmm..like this ? The bad point of this patch is that this will corrupt FILE_MAPPED status in root cgroup. This kind of change is not very good. So, one way is to use this kind of function only for new parameters. Hmm. == From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Now, file-mapped is maintaiend. But more generic update function will be needed for dirty page accounting. For accountig page status, we have to guarantee lock_page_cgroup() will ...

Hi Andrea, I am doing a simple dd test of writting a 4G file. This machine has got 64G of memory and I have created one cgroup with 100M as limit_in_bytes. I run following dd program both in root cgroup as well as test1/ cgroup(100M limit) one after the other. In root cgroup ============== dd if=/dev/zero of=/root/zerofile bs=4K count=1000000 1000000+0 records in 1000000+0 records out 4096000000 bytes (4.1 GB) copied, 59.5571 s, 68.8 MB/s In test1/ cgroup =============== dd ...

Consider that I'm not running the kernel build on tmpfs, but on a fs defined on /dev/sda. So the additional overhead should be normal compared to the mmotm vanilla, where there's only FILE_MAPPED accounting. -Andrea --

Correct. More exactly: bdi_thresh = memcg dirty memory limit * BDI's share of the global dirty memory Before: There'll be the opposite problem I think, the number of dirty pages (system-wide) will increase, because in this way we'll consider BDI shares of memcg dirty memory. So I think we need both: per memcg BDI stats and system-wide BDI stats, then we need to take the min of the two when evaluating bdi_thresh. Maybe... I'm not really sure about this, and need to figure better this ...

On Fri, 12 Mar 2010 00:27:09 +0100 Ah, sorry. I misunderstood something. But it's depends on dirty_ratio param. If background_dirty_ratio = 5 dirty_ratio = 100 under 100M cgroup, I think background write-out will be a help. (nonsense ? ;) And I wonder make -j can get better number....Hmm. Thanks, -Kame --

Amen ! Cheers, Ben. --

Wait. We need to fix the CPU_DEAD case anyway? Hmm. 6ad4c18884e864cf4c77f9074d3d1816063f99cd "sched: Fix balance vs hotplug race" did s/CPU_DEAD/CPU_DOWN_PREPARE/ in cpuset_track_online_cpus(). This doesn't look exactly right to me, we shouldn't do remove_tasks_in_empty_cpuset() at CPU_DOWN_PREPARE stage, it can fail. Otoh. This means that move_task_of_dead_cpu() can never see the task without active cpus in ->cpus_allowed, it is called later by CPU_DEAD. So, cpuset_lock() is not needed at ...

Right,.. so the whole problem is cpumask ops are terribly expensive since we got this CONFIG_CPUMASK_OFFSTACK muck, so we try to reduce these ops in the regular scheduling paths, in the patch you referenced above the tradeof was between fixing the sched_domains up too often vs adding a cpumask_and in a hot-path, guess who won ;-) --

Ah, have that re-validate the p->cpus_allowed for all cpuset tasks, ok Confusion^2 ;-), I failed to grasp your fixup idea and got confused, Ah indeed. One would almost construct a cpumask_assign that uses RCU Thanks! --

Good idea. Thanks! Regards, Wu Zhangjin --

Commit-ID: 83ff56f46a8532488ee364bb93a9cb2a59490d33 Gitweb: http://git.kernel.org/tip/83ff56f46a8532488ee364bb93a9cb2a59490d33 Author: Masami Hiramatsu <mhiramat@redhat.com> AuthorDate: Tue, 9 Mar 2010 10:22:19 -0500 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 14:06:16 +0100 kprobes: Calculate the index correctly when freeing the out-of-line execution slot From : Ananth N Mavinakayanahalli <ananth@in.ibm.com> When freeing the instruction slot, the ...

I must say this patch isn't based on your previous one in which mask_ack_irq() is modified to flag IRQ_MASKED. I let IRQ_MASKED serialise interrupt-handler and irq-thread in oneshot I have thought of that kind of reentering you point above, if IRQ_MASKED is cleared by irq_finalize_oneshot(), the thread_fn() is done as well as the hardware opration. If another irq comes in, the reentering does happen, but I think it's harmless because at this time we just set IRQTF_RUNTHREAD and IRQ_MASKED and ...

Yeah, I see. Thanks for your point. It will smooth the concerns. Thanks, Yong --

Well, since this is a prime time optimization i'd rather see the Kconfig value eliminated and the optimization to become unconditional. (if it doesnt work somewhere then it should be fixed) Ingo --

Jeff Moyer wrote: It appears to work here so far. I did not run massive i/o test suite but it now passes basic multiple kernel unpacking in parallel tests (before it didnt' even recognize the partition table). So the patch - at least basically - works. Except that we now have code duplication in kernel. Look at fs/compat.c:compat_do_readv_writev() - it does the same as compat_rw_copy_check_uvector() plus a bit more. I guess it is a good idea to use compat_rw_copy_check_uvector() ...

It looks like the conclusion was preliminary a bit. I tested the wrong kvm binary which does not have aio support... And when running real thing it crashes. I tested it on 2.6.32 (trivial corrections to the patch required). BUG: unable to handle kernel NULL pointer dereference at 0000000000000008 IP: [<ffffffff810c0e57>] generic_segment_checks+0x17/0xd0 PGD 18afe0067 PUD 18ace1067 PMD 0 Oops: 0000 [#1] SMP last sysfs file: /sys/devices/virtual/input/mice/uevent CPU 0 Modules linked ...

I think there's no need since in 2.6.33 (where your patch applies without any offsets) it crashes in exactly the same way: ------ BUG: unable to handle kernel NULL pointer dereference at 0000000000000008 IP: [<ffffffff810bbf37>] generic_segment_checks+0x17/0xd0 PGD 184f1c067 PUD 184cd1067 PMD 0 Oops: 0000 [#1] SMP last sysfs file: /sys/devices/virtual/input/mice/uevent CPU 1 Pid: 5182, comm: kvm Not tainted 2.6.33-amd64 #2.6.33.0 M3A78-EM/System Product Name RIP: 0010:[<ffffffff810bbf37>] ...

Now I'm confused, because I can't remember SIIG related patches for 2.6.28. --

From: Ingo Molnar <mingo@elte.hu> Ah, yep, this patch removes the dependency on K8_NB and it could be that your config has CONFIG_K8_NB disabled but with CONFIG_AGP_AMD64 enabled. Can you send me your .config pls? -- Regards/Gruss, Boris. -- Advanced Micro Devices, Inc. Operating Systems Research Center --

Hi Michal, Great, thanks. -- Cheers, Stephen Rothwell sfr@canb.auug.org.au http://www.canb.auug.org.au/~sfr/

Hi Stephen, Andrew reported this yesterday, too. The followin patch should fix it (added to the for-next branch): From 73d1393eb8507ed5fd7f8e696f6b1ecc18035ebe Mon Sep 17 00:00:00 2001 From: Michal Marek <mmarek@suse.cz> Date: Wed, 10 Mar 2010 12:28:58 +0100 Subject: [PATCH] kbuild: Generate modules.builtin in make modules_install The previous approach didn't work if one did make modules && make modules_install Add modules.builtin as dependency of _modinst_, which is the target ...

Ok, I try to make a new patch by using seqlock. Miao --

I'm going thru the printing routines now to get them usable by TUI/GUIs, starting with a libnewt based browser integrating initially report and annotate, after I get the first patch in shape for merging I'll revisit this table class of yours :-) - Arnaldo --

I think my first 6 patches can be used without modification. And I've already spent lots of efforts on them with Jeremy. We can work on evtchn and platform pci later. Thanks. -- regards Yang, Sheng --

I guess, what he meant was, to keep filesystem blocks aligned, even if the partition is not. Say if the partition is mis-aligned by 512-bytes, let the filesystem waste 4k-512bytes and keep it's blocks aligned. But it might be a case of over-engineering, possibly requiring disk format change. Thanks Nikanth --

Actually, it's a layering violation. The filesystem shouldn't need to probe the device layout ... particularly when there are complexities like is it logical 512 or physical, and if logical 512 on 4k does it have an offset exponent or not. We can transmit certain abstractions of information up the stack (like stripe width for RAID arrays which should be the fs optimal write size), but for this type of alignment, which can be completely solved at the partition layer, the information should ...

Yes. May be, just a simple but transparent device-mapper like mapping on top of the mis-aligned partition, to do the alignment. Then the file-system code need not change much. But Linux already has device-mapper and Linux will not be affected with mis- aligned partitions, when we use LVM. But the actual problem here is that partitioning tools might create partitions that wont allow other operating-systems to boot. So it might be enough, if the partitioning tools just create partitions ...

Hello, Turns out XP is generally OK. The reported problem was only on specific configurations (some BIOS stuff). Windows 2000 reportedly would be hurt but I really think we don't have to care about that too much. So, it seems like we wouldn't have to worry too much about it and just go ahead with new alignment schemes. I'll update the doc this weekend with new information from this now rather large thread. Thanks. -- tejun --

Well, device-mapper and LVM needed to be updated to make them "just I'm not following... Anyway, 4K drives that are 512b logical and 4K physical may or may not also have "DOS partition compensation" that use LBA -1 as the first naturally (4K) aligned start. This means that the partition tools need to shift the start of the first primary partition to be offset by 3584 bytes (7 512b sectors) for use with Linux. But for windows, AFAIK windows XP and windows 7 create all partitions aligned on ...

Ted, Apparently the real issue is Win2K, not XP. It seems to require the boot partition and possibly all partitions start on a cylinder boundary. And may have 255/63 hard-coded in to define what a cylinder is. I agree with the apparent consensus that a 2010 era linux partitioner does not need to be Win2K compatible. If someone wants to install Win2K they will need to either use an older generation partitioner to create the partitions or use specific command-line args to force a ...

I also have some older patches for btrfs that I need to get back out to the list. There was some talk of major changes to the organization of the tools so I held it back for a while longer. --

Hello, Miles, This one is a false positive that is addressed by a set of boot-time fixes that have recently gone into mainline, this series of commits: c598a070..71da8132 --

I made a patch(attached) using both local_irq_disable/enable and local_irq_save/restore. local_irq_save/restore is used only in mem_cgroup_update_file_mapped. And I attached a histogram graph of 30 times kernel build in root cgroup for each. before_root: no irq operation(original) after_root: local_irq_disable/enable for all after2_root: local_irq_save/restore for all after3_root: mixed version(attached) hmm, there seems to be a tendency that before < after < after3 < after2 ...

My test machine have 8CPUs, and I run all the test with "make -j8". me too. Thanks, Daisuke Nishimura. --

On Thu, 11 Mar 2010 14:13:00 +0900 How this looks ? == From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> Now, file-mapped is maintaiend. But more generic update function will be needed for dirty page accounting. For accountig page status, we have to guarantee lock_page_cgroup() will be never called under tree_lock held. To guarantee that, we use trylock at updating status. By this, we do fuzyy accounting, but in almost all case, it's correct. Signed-off-by: KAMEZAWA Hiroyuki ...

On Thu, 11 Mar 2010 16:50:20 +0900 If necessary, I'd like to prepare fixed one as independent patch for mmotm. Thanks, -Kame --

cgroup disable is at boot time "cgroup_disable=...", so root cgroup accounting should be enabled. The same for the local_irq_save/restore() overhead, lock/unlock_page_cgroup() is called during each charge. Many thanks for testing! -Andrea --

why not do it like this: @@ -2434,7 +2434,7 @@ static struct dentry *proc_base_instantiate(struct inode *dir, const struct pid_entry *p = ptr; struct inode *inode; struct proc_inode *ei; - struct dentry *error = ERR_PTR(-EINVAL); + struct dentry *error = ERR_PTR(-ENOMEM); /* Allocate the inode */ -- Zhenwen Xu - Seven Helight Home Page: http://zhwen.org --

I think most of the points have already been convered, but just for completeness, What is the history of the hardware you're running these builds on? Has it proven itself on previous kernel versions running much the same tests? Another point to consider: how are you running the compiler - is it over NFS from a PC? The reason I ask is that you can suffer from very weird corruption issues - I have a nice illustration of one which takes a copy of 1GB worth of data each day, and every once ...

Pulled, thanks Steve! Ingo --

Commit-ID: 007b09243b099811124f69d492adeebe9e439f96 Gitweb: http://git.kernel.org/tip/007b09243b099811124f69d492adeebe9e439f96 Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com> AuthorDate: Fri, 5 Mar 2010 15:03:26 -0800 Committer: Ingo Molnar <mingo@elte.hu> CommitDate: Thu, 11 Mar 2010 13:38:01 +0100 rcu: Increase RCU CPU stall timeouts if PROVE_RCU CONFIG_PROVE_RCU imposes additional overhead on the kernel, so increase the RCU CPU stall timeouts in an attempt to allow ...

I saw very similar failures for a long time on our iMX31 based device. Eventually I found a Freescale errata where the RAM inside the USB2 macrocell started to make single bit errors below 1.38V Vcore; ours was 1.4V at that time but dipped on CPU load. I cranked up the Vcore to 1.6V and that solved it, we also added some ceramic caps to Vcore to help with the dips. So it might be worth looking at PMU arrangements for Vcore level / look for dips with a 'scope (despite this isn't an ...

Unless there's more to it in the way the zaurus using it that regulator isn't programmable digitally. Reading about your CF Card WLAN related issues they suck down a good amount of power when their radio is up, I would definitely suggest monitoring with a 'scope the various rails (Vcore, RAM and whatever it is the CF Card is powered by) while putting it under load. -Andy --

Thank you for the pointer, I will take a look. Thanx, Paul --

This is one of the various solutions I disliked already, isn't it? To me this is just a more convoluted and obscure variant of the block_device_operations swap_slot_free_notify patch you had in mmotm, which Linus has rejected. I admit, I did not understand at all what Linus was proposing with readpage, writepage and address_space_operations: I kept quiet in the hope that you'd understand where I didn't! Hugh --

-- js --

Sorry I wrote that in haste. It is SLES 10 in the end. That means it is 2.6.16, not 2.6.27. Hence no sparsemem whatsoever. With SLES11 it should be OK, we are using flatmem only for i386. Whatever, it should be no issue now, as flatmem currently (as of 2.6.25) depends on i386. On the other hand I still considered the patch as applicable to contemporary kernels since there might be weird bios e820 maps and huge (and sparse) bootmem allocations/reservations (memory cgroups, initrd) so ...

Hi M, > > > > What you missed is that the native aio system calls require O_DIRECT. > > > > > > > > > > Thanks, that made it work. It seems without O_DIRECT it's just like > > > aio_* but without the separate thread. But I now get the "benefits" of > > > O_DIRECT for free... > > > > That is awesome news; I was worried. I saw that about O_DIRECT in the > > doc but assumed you were doing it. > > > > > Where did you see that? I reverted to the kernel source where ...

From include/asm-generic/uaccess.h static inline long copy_from_user(void *to, const void __user * from, unsigned long n) { might_sleep(); if (access_ok(VERIFY_READ, from, n)) return __copy_from_user(to, from, n); else return n; } static inline long copy_to_user(void __user *to, const void *from, unsigned long n) { might_sleep(); if (access_ok(VERIFY_WRITE, to, n)) return __copy_to_user(to, from, n); else return n; This cast is needed, and common, since 'arg' ...

Alternatively, could we maybe further simplify this reservation into: int reserve_pmu(void); void release_pmu(void); And not bother with anything finer grained. --

It does not solve the current problem that some parts of the pmu *are* used simultaneously by different subsystems. But, even if only perf would be used in the kernel you still can't be sure that all parts of the pmu are available to be used, you simply don't have it under your control. So why such limitations as an 'int reserve_pmu(int index)' is almost the same but provides much more flexibility? The question already arose if the watchdog would use perf permanently and thus would block ...

On 3/11/2010 4:46 AM, Ingo Molnar wrote: Is TRACE_EVENT an appropriate way to add hardware-specific counter events? I We can enumerate events this way, but there are other aspects to events than I'm not sure how that would work. The issue I am trying to solve here is that Power arch chips have a large number of very hardware-specific events that are not generalizable. Many of these events not only have names, but other user-configurable bits as well that select or narrow the scope of ...

Here are a few examples for the Power A2 processor. I've distorted the names because PMU architecture isn't publicly released yet. PM_DE_PMC_9:hrd_mask=0xff:hrd=0x22:pma_mask=0x3fff:pma=0x1b2d:culling_mode=3 PM_EX_0x03:lane=2:vlane=1 PM_OWE_ENG_MAC_FULL:usu=3 Note that the attribute fields shown above are fitted into the config field of Regards, - Corey --

I like the mmu.direct_map name. Its a good term too, I will change it in the patch. Joerg --

OK, makes sense now, I was missing the fact that the l1-nested-ptable needs to be shadowed and l1 translations to it must be write protected. You should disable out of sync shadow so that l1 guest writes to l1-nested-ptables always trap. And in the trap case, you'd have to invalidate l2 shadow pagetable entries that used the (now obsolete) --

Any problem with using a seqlock? The other thing you could do is store a pointer to the nodemask, and allocate a new nodemask when changing it, issue a smp_wmb(), and then store the new pointer. Read side only needs a smp_read_barrier_depends() --

On a closely related topic, I've just recently made a change to one of my branches that exposes the blkio.time and blkio.sectors information for the root cgroup. These stats would not show because the major and minor information for the root blkio_croup structures is zero. This information is not available at the when the root blkio_cgroup structures are instantiated, so they are left without major and minor information. I have a simple fix that updates the major and minor information for the ...

It'll also give people heart-attacks because they'll think it adds a new set of deprecated binary sysctl() interfaces to the kernel. Best drop the 'sysctl' part of the name as well. --

I don't understand one bit of this. What happens if the RESET_TO_ENABLE flag is set for the clock? The exact same bit is &-masked and then immediately |:ed to 1 again. Then it is written to the register. Practical effect: absolutely none. Is there a writel(val, clk->en_reg); missing from the unlikely() execution Isnt if (++clk.>usage_count == 1) easier to understand, or is it just me? BTW doing this: clk->usage_count++; if (clk->usage_count == 1) will not use more memory, the compiler ...

Hello Linus, Can we use clockevents directly? I though they are abstracted by the kernel framework and using them directly is not advisable/feasible. regards Shiraz --

Not that I know, but as tglx says it might soon be possible to do so, and if you have a patch for using some surplus clockevents for this along with some example of it actually being used too, people will probably react, atleast they will have opinions. Yours, Linus Walleij --

Hi Viresh, I'm working my way through the patch set, be patient :) As mentioned I prefer that you merge this file into the timer.c file and remove the extra .h interface totally. Also of course refactor to I usually just have raw writew() and skip all this extra abstraction in I forgot to mention that if .reserved can only have the values 0 and 1, the chances are big that you should turn that member of struct spear_timer I don't see the need for this functions. Are not all timers the ...

Err... drivers/base/Makefile:obj-$(CONFIG_HAVE_GENERIC_DMA_COHERENT) += dma-coherent.o include/asm-generic/dma-coherent.h:#ifdef CONFIG_HAVE_GENERIC_DMA_COHERENT --

-EPIPE means underrun condition. Try 'echo 101 > /proc/asound/card0/pcm0p/xrun_debug' and send dmesg output to check if the ring buffer pointer in the hda-intel driver does not return wrong values. Replace card0 and pcm0p with right identifiers (look to 'aplay -l' for right numbers). More info: http://www.alsa-project.org/main/index.php/XRUN_Debug Jaroslav ----- Jaroslav Kysela <perex@perex.cz> Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc. --

All these parameters should be set: CONFIG_SND_PCM_XRUN_DEBUG, CONFIG_SND_VERBOSE_PROCFS, CONFIG_SND_DEBUG . I updated wiki. Jaroslav ----- Jaroslav Kysela <perex@perex.cz> Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc. --

Ping. can we get this added to linux-next, please? -- ~Randy --

From: Randy Dunlap <randy.dunlap@oracle.com> r852 fails to build when CONFIG_PCI is not enabled since it uses pci_*() calls and is a PCI driver, so it should depend on PCI to prevent build errors. It should also #include <linux/pci.h>. drivers/mtd/nand/r852.c:1053: error: implicit declaration of function 'pci_prepare_to_sleep' drivers/mtd/nand/r852.c:1062: error: implicit declaration of function 'pci_back_from_sleep' Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com> Cc: David ...

On Mon, 01 Mar 2010 14:14:36 +0800 ^^ these two lines don't assemble with gas 2.16.1: arch/x86/crypto/aesni-intel_asm.S: Assembler messages: arch/x86/crypto/aesni-intel_asm.S:752: Error: suffix or operands invalid for `movq' --

Commit-ID: 0e95017355dcf43031da6d0e360a748717e56df1 Gitweb: http://git.kernel.org/tip/0e95017355dcf43031da6d0e360a748717e56df1 Author: Tim Bird <tim.bird@am.sony.com> AuthorDate: Thu, 25 Feb 2010 15:36:43 -0800 Committer: Steven Rostedt <rostedt@goodmis.org> CommitDate: Fri, 5 Mar 2010 21:20:57 -0500 function-graph: Add tracing_thresh support to function_graph tracer Add support for tracing_thresh to the function_graph tracer. This version of this feature isolates the checks into ...

Of course, if we can fix, it's better. However, probably, users of this patch would be only boot loader, because this is a first sector on extended-partition itself, not logical-partitions in extended-partition. So, maybe this wouldn't be so major problem for normal users, and what is needed actually is not sure to me for now (I guess it might be depending to BIOS, if boot loader is using BIOS call.). Thanks. -- OGAWA Hirofumi <hirofumi@mail.parknet.co.jp> --

Wrong. A lot of BIOSes will attempt to boot from USB storage. Worse, a fair number of BIOSes will hang during startup if a USB storage device I would consider this a bugfix. As such, it should be pushed outside the merge window. -hpa --

Just in case I wasn't clear enough -- this patch is a bugfix and should be pushed upstream as long as it is technically ready. -hpa --

The 2/2 patch seems to be a bug fix too; it implements the intent of the original code as far as I can see. -hpa --

Great! it may be a good idea to start accumulating pointers, identified This is indeed the problem I currently see with Ftrace, suitability for huge live/realtime traces. For this you need an extremely compact format and a good way to pass and update metadata along with the trace. Otherwise, Ftrace and Perf offer a large number of exciting features. In LTTng, following some feedback from Google among others, quite a bit of information is implicit: per cpu files and scheduling events ...

Did anyone look at the problem described and checked the patch? I still received no reply. -- Robert Schoene Technische Universitaet Dresden Zentrum fuer Informationsdienste und Hochleistungsrechnen 01062 Dresden Tel.: (0351) 463-42483, Fax: (0351) 463-37773 E-Mail: Robert.Schoene@tu-dresden.de --

These look reasonable. Do you have a GIT tree I can pull? David --

I guess it doesn't fit for gnome-session. because gtop or similar system monitoring process assume it can see all processes in the system. thanks. --