Prev: [PATCH 4/4] msi-laptop: Add i8042 filter to sync sw state with BIOS when function key pressed
Next: ibft: Use IBFT_SIGN instead of open-coding the search string.
From: Taras Glek on 12 May 2010 15:40 Hi, From the docs it seems that perf may be useful for logging page faults. I tried perf record -g -d -f -e page-faults <mycommand> but all I got were PERF_RECORD_MMAP events. I'm using 2.6.32. Ideally I'd like to collect fault addresses and userspace stacks that caused them. Isn't this what the page-fault event is for? Taras -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Andi Kleen on 12 May 2010 16:10 Taras Glek <tglek(a)mozilla.com> writes: > Hi, > From the docs it seems that perf may be useful for logging page > faults. I tried > perf record -g -d -f -e page-faults <mycommand> > but all I got were PERF_RECORD_MMAP events. > > I'm using 2.6.32. Ideally I'd like to collect fault addresses and > userspace stacks that caused them. Isn't this what the page-fault > event is for? Some time ago I used systemtap to do this for my pbitmaps works. This was the old systemtap script. -Andi probe vm.pagefault { if (task_execname(task_current()) == "executable" && address < 0x100000000) { printf("%u\n", address); } } -- ak(a)linux.intel.com -- Speaking for myself only. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Frederic Weisbecker on 12 May 2010 16:10 On Wed, May 12, 2010 at 12:38:30PM -0700, Taras Glek wrote: > Hi, > From the docs it seems that perf may be useful for logging page faults. > I tried > perf record -g -d -f -e page-faults <mycommand> > but all I got were PERF_RECORD_MMAP events. > > I'm using 2.6.32. Ideally I'd like to collect fault addresses and > userspace stacks that caused them. Isn't this what the page-fault event > is for? > > Taras Could you try with -c 1 ? Otherwise, does that work on the latest kernel? Thanks. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Taras Glek on 12 May 2010 17:20 On 05/12/2010 12:59 PM, Frederic Weisbecker wrote: > On Wed, May 12, 2010 at 12:38:30PM -0700, Taras Glek wrote: > >> Hi, >> From the docs it seems that perf may be useful for logging page faults. >> I tried >> perf record -g -d -f -e page-faults<mycommand> >> but all I got were PERF_RECORD_MMAP events. >> >> I'm using 2.6.32. Ideally I'd like to collect fault addresses and >> userspace stacks that caused them. Isn't this what the page-fault event >> is for? >> >> Taras >> > > Could you try with -c 1 ? > > Otherwise, does that work on the latest kernel? > this just gets me PERF_RECORD_SAMPLE events which don't seem to be related to page-faults(ie they show up with/without -e page-faults). I don't have the latest kernel installed here. Can you get perf page-fault events on your system? Taras -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
From: Frederic Weisbecker on 12 May 2010 17:30
On Wed, May 12, 2010 at 02:18:08PM -0700, Taras Glek wrote: > On 05/12/2010 12:59 PM, Frederic Weisbecker wrote: >> On Wed, May 12, 2010 at 12:38:30PM -0700, Taras Glek wrote: >> >>> Hi, >>> From the docs it seems that perf may be useful for logging page faults. >>> I tried >>> perf record -g -d -f -e page-faults<mycommand> >>> but all I got were PERF_RECORD_MMAP events. >>> >>> I'm using 2.6.32. Ideally I'd like to collect fault addresses and >>> userspace stacks that caused them. Isn't this what the page-fault event >>> is for? >>> >>> Taras >>> >> >> Could you try with -c 1 ? >> >> Otherwise, does that work on the latest kernel? >> > this just gets me PERF_RECORD_SAMPLE events which don't seem to be > related to page-faults(ie they show up with/without -e page-faults). > > I don't have the latest kernel installed here. Can you get perf > page-fault events on your system? Yeah, no problem on 2.6.34-rc7. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo(a)vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/ |