linux-toradex.git/kernel/debug/debug_core.c, branch v3.15-rc8 Linux kernel for Apalis and Colibri modules mm: per-thread vma caching 2014-04-07T23:35:53+00:00 Davidlohr Bueso davidlohr@hp.com 2014-04-07T22:37:25+00:00 615d6e8756c87149f2d4c1b93d471bca002bd849 This patch is a continuation of efforts trying to optimize find_vma(), avoiding potentially expensive rbtree walks to locate a vma upon faults. The original approach (https://lkml.org/lkml/2013/11/1/410), where the largest vma was also cached, ended up being too specific and random, thus further comparison with other approaches were needed. There are two things to consider when dealing with this, the cache hit rate and the latency of find_vma(). Improving the hit-rate does not necessarily translate in finding the vma any faster, as the overhead of any fancy caching schemes can be too high to consider. We currently cache the last used vma for the whole address space, which provides a nice optimization, reducing the total cycles in find_vma() by up to 250%, for workloads with good locality. On the other hand, this simple scheme is pretty much useless for workloads with poor locality. Analyzing ebizzy runs shows that, no matter how many threads are running, the mmap_cache hit rate is less than 2%, and in many situations below 1%. The proposed approach is to replace this scheme with a small per-thread cache, maximizing hit rates at a very low maintenance cost. Invalidations are performed by simply bumping up a 32-bit sequence number. The only expensive operation is in the rare case of a seq number overflow, where all caches that share the same address space are flushed. Upon a miss, the proposed replacement policy is based on the page number that contains the virtual address in question. Concretely, the following results are seen on an 80 core, 8 socket x86-64 box: 1) System bootup: Most programs are single threaded, so the per-thread scheme does improve ~50% hit rate by just adding a few more slots to the cache. +----------------+----------+------------------+ | caching scheme | hit-rate | cycles (billion) | +----------------+----------+------------------+ | baseline | 50.61% | 19.90 | | patched | 73.45% | 13.58 | +----------------+----------+------------------+ 2) Kernel build: This one is already pretty good with the current approach as we're dealing with good locality. +----------------+----------+------------------+ | caching scheme | hit-rate | cycles (billion) | +----------------+----------+------------------+ | baseline | 75.28% | 11.03 | | patched | 88.09% | 9.31 | +----------------+----------+------------------+ 3) Oracle 11g Data Mining (4k pages): Similar to the kernel build workload. +----------------+----------+------------------+ | caching scheme | hit-rate | cycles (billion) | +----------------+----------+------------------+ | baseline | 70.66% | 17.14 | | patched | 91.15% | 12.57 | +----------------+----------+------------------+ 4) Ebizzy: There's a fair amount of variation from run to run, but this approach always shows nearly perfect hit rates, while baseline is just about non-existent. The amounts of cycles can fluctuate between anywhere from ~60 to ~116 for the baseline scheme, but this approach reduces it considerably. For instance, with 80 threads: +----------------+----------+------------------+ | caching scheme | hit-rate | cycles (billion) | +----------------+----------+------------------+ | baseline | 1.06% | 91.54 | | patched | 99.97% | 14.18 | +----------------+----------+------------------+ [akpm@linux-foundation.org: fix nommu build, per Davidlohr] [akpm@linux-foundation.org: document vmacache_valid() logic] [akpm@linux-foundation.org: attempt to untangle header files] [akpm@linux-foundation.org: add vmacache_find() BUG_ON] [hughd@google.com: add vmacache_valid_mm() (from Oleg)] [akpm@linux-foundation.org: coding-style fixes] [akpm@linux-foundation.org: adjust and enhance comments] Signed-off-by: Davidlohr Bueso <davidlohr@hp.com> Reviewed-by: Rik van Riel <riel@redhat.com> Acked-by: Linus Torvalds <torvalds@linux-foundation.org> Reviewed-by: Michel Lespinasse <walken@google.com> Cc: Oleg Nesterov <oleg@redhat.com> Tested-by: Hugh Dickins <hughd@google.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> 
This patch is a continuation of efforts trying to optimize find_vma(),
avoiding potentially expensive rbtree walks to locate a vma upon faults.
The original approach (https://lkml.org/lkml/2013/11/1/410), where the
largest vma was also cached, ended up being too specific and random,
thus further comparison with other approaches were needed.  There are
two things to consider when dealing with this, the cache hit rate and
the latency of find_vma().  Improving the hit-rate does not necessarily
translate in finding the vma any faster, as the overhead of any fancy
caching schemes can be too high to consider.

We currently cache the last used vma for the whole address space, which
provides a nice optimization, reducing the total cycles in find_vma() by
up to 250%, for workloads with good locality.  On the other hand, this
simple scheme is pretty much useless for workloads with poor locality.
Analyzing ebizzy runs shows that, no matter how many threads are
running, the mmap_cache hit rate is less than 2%, and in many situations
below 1%.

The proposed approach is to replace this scheme with a small per-thread
cache, maximizing hit rates at a very low maintenance cost.
Invalidations are performed by simply bumping up a 32-bit sequence
number.  The only expensive operation is in the rare case of a seq
number overflow, where all caches that share the same address space are
flushed.  Upon a miss, the proposed replacement policy is based on the
page number that contains the virtual address in question.  Concretely,
the following results are seen on an 80 core, 8 socket x86-64 box:

1) System bootup: Most programs are single threaded, so the per-thread
   scheme does improve ~50% hit rate by just adding a few more slots to
   the cache.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 50.61%   | 19.90            |
| patched        | 73.45%   | 13.58            |
+----------------+----------+------------------+

2) Kernel build: This one is already pretty good with the current
   approach as we're dealing with good locality.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 75.28%   | 11.03            |
| patched        | 88.09%   | 9.31             |
+----------------+----------+------------------+

3) Oracle 11g Data Mining (4k pages): Similar to the kernel build workload.

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 70.66%   | 17.14            |
| patched        | 91.15%   | 12.57            |
+----------------+----------+------------------+

4) Ebizzy: There's a fair amount of variation from run to run, but this
   approach always shows nearly perfect hit rates, while baseline is just
   about non-existent.  The amounts of cycles can fluctuate between
   anywhere from ~60 to ~116 for the baseline scheme, but this approach
   reduces it considerably.  For instance, with 80 threads:

+----------------+----------+------------------+
| caching scheme | hit-rate | cycles (billion) |
+----------------+----------+------------------+
| baseline       | 1.06%    | 91.54            |
| patched        | 99.97%   | 14.18            |
+----------------+----------+------------------+

[akpm@linux-foundation.org: fix nommu build, per Davidlohr]
[akpm@linux-foundation.org: document vmacache_valid() logic]
[akpm@linux-foundation.org: attempt to untangle header files]
[akpm@linux-foundation.org: add vmacache_find() BUG_ON]
[hughd@google.com: add vmacache_valid_mm() (from Oleg)]
[akpm@linux-foundation.org: coding-style fixes]
[akpm@linux-foundation.org: adjust and enhance comments]
Signed-off-by: Davidlohr Bueso <davidlohr@hp.com>
Reviewed-by: Rik van Riel <riel@redhat.com>
Acked-by: Linus Torvalds <torvalds@linux-foundation.org>
Reviewed-by: Michel Lespinasse <walken@google.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Tested-by: Hugh Dickins <hughd@google.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
KGDB: make kgdb_breakpoint() as noinline 2014-02-26T11:16:25+00:00 Vijaya Kumar K Vijaya.Kumar@caviumnetworks.com 2014-01-28T11:20:20+00:00 d498d4b47fb3050f2f7840cc49251f87f04d1ca9 The function kgdb_breakpoint() sets up break point at compile time by calling arch_kgdb_breakpoint(); Though this call is surrounded by wmb() barrier, the compile can still re-order the break point, because this scheduling barrier is not a code motion barrier in gcc. Making kgdb_breakpoint() as noinline solves this problem of code reording around break point instruction and also avoids problem of being called as inline function from other places More details about discussion on this can be found here http://comments.gmane.org/gmane.linux.ports.arm.kernel/269732 Signed-off-by: Vijaya Kumar K <Vijaya.Kumar@caviumnetworks.com> Acked-by: Will Deacon <will.deacon@arm.com> Acked-by: Jason Wessel <jason.wessel@windriver.com> Signed-off-by: Catalin Marinas <catalin.marinas@arm.com> 
The function kgdb_breakpoint() sets up break point at
compile time by calling arch_kgdb_breakpoint();
Though this call is surrounded by wmb() barrier,
the compile can still re-order the break point,
because this scheduling barrier is not a code motion
barrier in gcc.

Making kgdb_breakpoint() as noinline solves this problem
of code reording around break point instruction and also
avoids problem of being called as inline function from
other places

More details about discussion on this can be found here
http://comments.gmane.org/gmane.linux.ports.arm.kernel/269732

Signed-off-by: Vijaya Kumar K <Vijaya.Kumar@caviumnetworks.com>
Acked-by: Will Deacon <will.deacon@arm.com>
Acked-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
kgdb/kdb: Fix no KDB config problem 2014-01-25T07:55:09+00:00 Mike Travis travis@sgi.com 2014-01-14T16:25:52+00:00 fc8b13740b2978b34872650cc8e928392e3758aa Some code added to the debug_core module had KDB dependencies that it shouldn't have. Move the KDB dependent REASON back to the caller to remove the dependency in the debug core code. Update the call from the UV NMI handler to conform to the new interface. Signed-off-by: Mike Travis <travis@sgi.com> Reviewed-by: Hedi Berriche <hedi@sgi.com> Cc: Russ Anderson <rja@sgi.com> Cc: Jason Wessel <jason.wessel@windriver.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Link: http://lkml.kernel.org/r/20140114162551.318251993@asylum.americas.sgi.com Signed-off-by: Ingo Molnar <mingo@kernel.org> 
Some code added to the debug_core module had KDB dependencies
that it shouldn't have.  Move the KDB dependent REASON back to
the caller to remove the dependency in the debug core code.

Update the call from the UV NMI handler to conform to the new
interface.

Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Russ Anderson <rja@sgi.com>
Cc: Jason Wessel <jason.wessel@windriver.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Link: http://lkml.kernel.org/r/20140114162551.318251993@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kdb: Add support for external NMI handler to call KGDB/KDB 2013-10-03T16:47:54+00:00 Mike Travis travis@sgi.com 2013-10-02T15:14:18+00:00 8daaa5f8261bffd2f6217a960f9182d0503a5c44 This patch adds a kgdb_nmicallin() interface that can be used by external NMI handlers to call the KGDB/KDB handler. The primary need for this is for those types of NMI interrupts where all the CPUs have already received the NMI signal. Therefore no send_IPI(NMI) is required, and in fact it will cause a 2nd unhandled NMI to occur. This generates the "Dazed and Confuzed" messages. Since all the CPUs are getting the NMI at roughly the same time, it's not guaranteed that the first CPU that hits the NMI handler will manage to enter KGDB and set the dbg_master_lock before the slaves start entering. The new argument "send_ready" was added for KGDB to signal the NMI handler to release the slave CPUs for entry into KGDB. Signed-off-by: Mike Travis <travis@sgi.com> Acked-by: Jason Wessel <jason.wessel@windriver.com> Reviewed-by: Dimitri Sivanich <sivanich@sgi.com> Reviewed-by: Hedi Berriche <hedi@sgi.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Link: http://lkml.kernel.org/r/20131002151417.928886849@asylum.americas.sgi.com Signed-off-by: Ingo Molnar <mingo@kernel.org> 
This patch adds a kgdb_nmicallin() interface that can be used by
external NMI handlers to call the KGDB/KDB handler.  The primary
need for this is for those types of NMI interrupts where all the
CPUs have already received the NMI signal.  Therefore no
send_IPI(NMI) is required, and in fact it will cause a 2nd
unhandled NMI to occur. This generates the "Dazed and Confuzed"
messages.

Since all the CPUs are getting the NMI at roughly the same time,
it's not guaranteed that the first CPU that hits the NMI handler
will manage to enter KGDB and set the dbg_master_lock before the
slaves start entering. The new argument "send_ready" was added
for KGDB to signal the NMI handler to release the slave CPUs for
entry into KGDB.

Signed-off-by: Mike Travis <travis@sgi.com>
Acked-by: Jason Wessel <jason.wessel@windriver.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net>
Link: http://lkml.kernel.org/r/20131002151417.928886849@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kgdb/sysrq: fix inconstistent help message of sysrq key 2013-05-01T00:04:10+00:00 zhangwei(Jovi) jovi.zhangwei@huawei.com 2013-04-30T22:28:51+00:00 f345650964936395307d9d6bdee0168cf7d926e7 Currently help message of /proc/sysrq-trigger highlight its upper-case characters, like below: SysRq : HELP : loglevel(0-9) reBoot Crash terminate-all-tasks(E) memory-full-oom-kill(F) kill-all-tasks(I) ... this would confuse user trigger sysrq by upper-case character, which is inconsistent with the real lower-case character registed key. This inconsistent help message will also lead more confused when 26 upper-case letters put into use in future. This patch fix kgdb sysrq key: "debug(g)" Signed-off-by: zhangwei(Jovi) <jovi.zhangwei@huawei.com> Cc: Jason Wessel <jason.wessel@windriver.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> 
Currently help message of /proc/sysrq-trigger highlight its upper-case
characters, like below:

      SysRq : HELP : loglevel(0-9) reBoot Crash terminate-all-tasks(E)
      memory-full-oom-kill(F) kill-all-tasks(I) ...

this would confuse user trigger sysrq by upper-case character, which is
inconsistent with the real lower-case character registed key.

This inconsistent help message will also lead more confused when
26 upper-case letters put into use in future.

This patch fix kgdb sysrq key: "debug(g)"

Signed-off-by: zhangwei(Jovi) <jovi.zhangwei@huawei.com>
Cc: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
kgdb: remove #include <linux/serial_8250.h> from kgdb.h 2013-02-04T23:35:26+00:00 Greg Kroah-Hartman gregkh@linuxfoundation.org 2013-02-04T23:35:26+00:00 16559ae48c76f1ceb970b9719dea62b77eb5d06b There's no reason kgdb.h itself needs to include the 8250 serial port header file. So push it down to the _very_ limited number of individual drivers that need the values in that file, and fix up the places where people really wanted serial_core.h and platform_device.h. Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
There's no reason kgdb.h itself needs to include the 8250 serial port
header file.  So push it down to the _very_ limited number of individual
drivers that need the values in that file, and fix up the places where
people really wanted serial_core.h and platform_device.h.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Merge tag 'for_linus-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb 2012-10-13T02:16:58+00:00 Linus Torvalds torvalds@linux-foundation.org 2012-10-13T02:16:58+00:00 6c536a17fa049d0fb690c1a947b97dbfd304a916 Pull KGDB/KDB fixes and cleanups from Jason Wessel: "Cleanups - Clean up compile warnings in kgdboc.c and x86/kernel/kgdb.c - Add module event hooks for simplified debugging with gdb Fixes - Fix kdb to stop paging with 'q' on bta and dmesg - Fix for data that scrolls off the vga console due to line wrapping when using the kdb pager New - The debug core registers for kernel module events which allows a kernel aware gdb to automatically load symbols and break on entry to a kernel module - Allow kgdboc=kdb to setup kdb on the vga console" * tag 'for_linus-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb: tty/console: fix warnings in drivers/tty/serial/kgdboc.c kdb,vt_console: Fix missed data due to pager overruns kdb: Fix dmesg/bta scroll to quit with 'q' kgdboc: Accept either kbd or kdb to activate the vga + keyboard kdb shell kgdb,x86: fix warning about unused variable mips,kgdb: fix recursive page fault with CONFIG_KPROBES kgdb: Add module event hooks 
Pull KGDB/KDB fixes and cleanups from Jason Wessel:
 "Cleanups
   - Clean up compile warnings in kgdboc.c and x86/kernel/kgdb.c
   - Add module event hooks for simplified debugging with gdb
 Fixes
   - Fix kdb to stop paging with 'q' on bta and dmesg
   - Fix for data that scrolls off the vga console due to line wrapping
     when using the kdb pager
 New
   - The debug core registers for kernel module events which allows a
     kernel aware gdb to automatically load symbols and break on entry
     to a kernel module
   - Allow kgdboc=kdb to setup kdb on the vga console"

* tag 'for_linus-3.7' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb:
  tty/console: fix warnings in drivers/tty/serial/kgdboc.c
  kdb,vt_console: Fix missed data due to pager overruns
  kdb: Fix dmesg/bta scroll to quit with 'q'
  kgdboc: Accept either kbd or kdb to activate the vga + keyboard kdb shell
  kgdb,x86: fix warning about unused variable
  mips,kgdb: fix recursive page fault with CONFIG_KPROBES
  kgdb: Add module event hooks
kgdb: Add module event hooks 2012-10-12T11:37:33+00:00 Jason Wessel jason.wessel@windriver.com 2012-10-12T11:37:33+00:00 f30fed10c440a25937e509860fa207399b26efe5 Allow gdb to auto load kernel modules when it is attached, which makes it trivially easy to debug module init functions or pre-set breakpoints in a kernel module that has not loaded yet. Signed-off-by: Jason Wessel <jason.wessel@windriver.com> 
Allow gdb to auto load kernel modules when it is attached,
which makes it trivially easy to debug module init functions
or pre-set breakpoints in a kernel module that has not loaded yet.

Signed-off-by: Jason Wessel <jason.wessel@windriver.com>
kernel/debug: Mask KGDB NMI upon entry 2012-09-26T20:42:25+00:00 Anton Vorontsov anton.vorontsov@linaro.org 2012-09-24T21:27:50+00:00 5a14fead07bcf4e0acc877a8d9e1d1f40a441153 The new arch callback should manage NMIs that usually cause KGDB to enter. That is, not all NMIs should be enabled/disabled, but only those that issue kgdb_handle_exception(). We must mask it as serial-line interrupt can be used as an NMI, so if the original KGDB-entry cause was say a breakpoint, then every input to KDB console will cause KGDB to reenter, which we don't want. Signed-off-by: Anton Vorontsov <anton.vorontsov@linaro.org> Acked-by: Jason Wessel <jason.wessel@windriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> 
The new arch callback should manage NMIs that usually cause KGDB to
enter. That is, not all NMIs should be enabled/disabled, but only
those that issue kgdb_handle_exception().

We must mask it as serial-line interrupt can be used as an NMI, so
if the original KGDB-entry cause was say a breakpoint, then every
input to KDB console will cause KGDB to reenter, which we don't want.

Signed-off-by: Anton Vorontsov <anton.vorontsov@linaro.org>
Acked-by: Jason Wessel <jason.wessel@windriver.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Merge tag 'for_linus-3.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb 2012-04-05T00:26:08+00:00 Linus Torvalds torvalds@linux-foundation.org 2012-04-05T00:26:08+00:00 6c216ec636f75d834461be15f83ec41a6759bd2b Pull KGDB/KDB regression fixes from Jason Wessel: - Fix a Smatch warning that appeared in the 3.4 merge window - Fix kgdb test suite with SMP for all archs without HW single stepping - Fix kgdb sw breakpoints with CONFIG_DEBUG_RODATA=y limitations on x86 - Fix oops on kgdb test suite with CONFIG_DEBUG_RODATA - Fix kgdb test suite with SMP for all archs with HW single stepping * tag 'for_linus-3.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb: x86,kgdb: Fix DEBUG_RODATA limitation using text_poke() kgdb,debug_core: pass the breakpoint struct instead of address and memory kgdbts: (2 of 2) fix single step awareness to work correctly with SMP kgdbts: (1 of 2) fix single step awareness to work correctly with SMP kgdbts: Fix kernel oops with CONFIG_DEBUG_RODATA kdb: Fix smatch warning on dbg_io_ops->is_console 
Pull KGDB/KDB regression fixes from Jason Wessel:
 - Fix a Smatch warning that appeared in the 3.4 merge window
 - Fix kgdb test suite with SMP for all archs without HW single stepping
 - Fix kgdb sw breakpoints with CONFIG_DEBUG_RODATA=y limitations on x86
 - Fix oops on kgdb test suite with CONFIG_DEBUG_RODATA
 - Fix kgdb test suite with SMP for all archs with HW single stepping

* tag 'for_linus-3.4-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/jwessel/kgdb:
  x86,kgdb: Fix DEBUG_RODATA limitation using text_poke()
  kgdb,debug_core: pass the breakpoint struct instead of address and memory
  kgdbts: (2 of 2) fix single step awareness to work correctly with SMP
  kgdbts: (1 of 2) fix single step awareness to work correctly with SMP
  kgdbts: Fix kernel oops with CONFIG_DEBUG_RODATA
  kdb: Fix smatch warning on dbg_io_ops->is_console