[GIT PULL] kmemcheck updates for tip/kmemcheck

[GIT PULL] kmemcheck updates for tip/kmemcheck

[Vegard Nossum]

Ingo,

Here follows a small update to the REP instruction handling and removal
of some obsolete code (if you worry that this is useful code, git will
save it for us and it can be revived if needed. I believe that the code
is simply a burden as it is today).

I couldn't figure out how to merge v2.6.27-rc5 without littering the
pull request with all the commits that were made in Linus's repository
since v2.6.27-rc3, but feel free to do it yourself. There should be no
conflicts.

You may also delete the tip/kmemcheck-dma branch as this has now been
integrated since the last pull you did.


The following changes since commit db4b0f1e8ae7c9ffde8bef2eb3dcd4fad5988ac1:
  Vegard Nossum (1):
        kmemcheck: implement REP MOVS/STOS emulation

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/vegard/kmemcheck.git for-tip

Vegard Nossum (3):
      kmemcheck: hide/show pages in each iteration of a REP instruction
      kmemcheck: rip out the optimized memset()
      kmemcheck: rip out SMP code

 arch/x86/Kconfig.debug            |   22 --------
 arch/x86/mm/kmemcheck/Makefile    |    4 +-
 arch/x86/mm/kmemcheck/kmemcheck.c |   27 ++--------
 arch/x86/mm/kmemcheck/shadow.c    |    2 +-
 arch/x86/mm/kmemcheck/smp.c       |   80 ------------------------------
 arch/x86/mm/kmemcheck/smp.h       |   23 ---------
 arch/x86/mm/kmemcheck/string.c    |   99 -------------------------------------
 include/asm-x86/string_32.h       |    8 ---
 include/asm-x86/string_64.h       |    1 -
 9 files changed, 7 insertions(+), 259 deletions(-)
 delete mode 100644 arch/x86/mm/kmemcheck/smp.c
 delete mode 100644 arch/x86/mm/kmemcheck/smp.h
 delete mode 100644 arch/x86/mm/kmemcheck/string.c

Thanks,


Vegard

Re: [GIT PULL] kmemcheck updates for tip/kmemcheck

[Ingo Molnar]

* Vegard Nossum <vegard.nossum@gmail.com> wrote:

> Ingo,
> 
> Here follows a small update to the REP instruction handling and 
> removal of some obsolete code (if you worry that this is useful code, 
> git will save it for us and it can be revived if needed. I believe 
> that the code is simply a burden as it is today).

pulled into tip/kmemcheck, thanks Vegard!

> I couldn't figure out how to merge v2.6.27-rc5 without littering the 
> pull request with all the commits that were made in Linus's repository 
> since v2.6.27-rc3, but feel free to do it yourself. There should be no 
> conflicts.

yeah, merges should generally be done in the upstream repo. Obviously 
git-request-pull must list all the commits that happened - it doesnt 
know which commits are 'interesting' or not. I've done an upstream 
merge.

btw., a small build fix (see below) was needed.

	Ingo

---------------->
>From 92a9abb2c571ab8fedaee3537c5667b4e61188ce Mon Sep 17 00:00:00 2001
From: Ingo Molnar <mingo@elte.hu>
Date: Mon, 8 Sep 2008 16:50:43 +0200
Subject: [PATCH] kmemcheck: build fix
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 8bit

fix:

 arch/x86/mm/kmemcheck/kmemcheck.c: In function ‘kmemcheck_init’:
 arch/x86/mm/kmemcheck/kmemcheck.c:42: error: ‘setup_max_cpus’ undeclared (first use in this function)
 arch/x86/mm/kmemcheck/kmemcheck.c:42: error: (Each undeclared identifier is reported only once
 arch/x86/mm/kmemcheck/kmemcheck.c:42: error: for each function it appears in.)

Signed-off-by: Ingo Molnar <mingo@elte.hu>
---
 arch/x86/mm/kmemcheck/kmemcheck.c |    2 ++
 1 files changed, 2 insertions(+), 0 deletions(-)

diff --git a/arch/x86/mm/kmemcheck/kmemcheck.c b/arch/x86/mm/kmemcheck/kmemcheck.c
index eef8c6a..75831f0 100644
--- a/arch/x86/mm/kmemcheck/kmemcheck.c
+++ b/arch/x86/mm/kmemcheck/kmemcheck.c
@@ -35,6 +35,7 @@ void __init kmemcheck_init(void)
 {
 	printk(KERN_INFO "kmemcheck: \"Bugs, beware!\"\n");
 
+#if defined(CONFIG_SMP) && !defined(CONFIG_KMEMCHECK_USE_SMP)
 	/*
 	 * Limit SMP to use a single CPU. We rely on the fact that this code
 	 * runs before SMP is set up.
@@ -44,6 +45,7 @@ void __init kmemcheck_init(void)
 			"kmemcheck: Limiting number of CPUs to 1.\n");
 		setup_max_cpus = 1;
 	}
+#endif
 }
 
 #ifdef CONFIG_KMEMCHECK_DISABLED_BY_DEFAULT

Re: [GIT PULL] kmemcheck updates for tip/kmemcheck

[Vegard Nossum]

On Mon, Sep 8, 2008 at 5:17 PM, Ingo Molnar <mingo@elte.hu> wrote:
> yeah, merges should generally be done in the upstream repo. Obviously
> git-request-pull must list all the commits that happened - it doesnt
> know which commits are 'interesting' or not. I've done an upstream
> merge.
>
> btw., a small build fix (see below) was needed.

Oops. I was being enthusiastic and removed too much.

>
> ---------------->
> From 92a9abb2c571ab8fedaee3537c5667b4e61188ce Mon Sep 17 00:00:00 2001
> From: Ingo Molnar <mingo@elte.hu>
> Date: Mon, 8 Sep 2008 16:50:43 +0200
> Subject: [PATCH] kmemcheck: build fix
> MIME-Version: 1.0
> Content-Type: text/plain; charset=utf-8
> Content-Transfer-Encoding: 8bit
>
> fix:
>
>  arch/x86/mm/kmemcheck/kmemcheck.c: In function 'kmemcheck_init':
>  arch/x86/mm/kmemcheck/kmemcheck.c:42: error: 'setup_max_cpus' undeclared (first use in this function)
>  arch/x86/mm/kmemcheck/kmemcheck.c:42: error: (Each undeclared identifier is reported only once
>  arch/x86/mm/kmemcheck/kmemcheck.c:42: error: for each function it appears in.)
>
> Signed-off-by: Ingo Molnar <mingo@elte.hu>
> ---
>  arch/x86/mm/kmemcheck/kmemcheck.c |    2 ++
>  1 files changed, 2 insertions(+), 0 deletions(-)
>
> diff --git a/arch/x86/mm/kmemcheck/kmemcheck.c b/arch/x86/mm/kmemcheck/kmemcheck.c
> index eef8c6a..75831f0 100644
> --- a/arch/x86/mm/kmemcheck/kmemcheck.c
> +++ b/arch/x86/mm/kmemcheck/kmemcheck.c
> @@ -35,6 +35,7 @@ void __init kmemcheck_init(void)
>  {
>        printk(KERN_INFO "kmemcheck: \"Bugs, beware!\"\n");
>
> +#if defined(CONFIG_SMP) && !defined(CONFIG_KMEMCHECK_USE_SMP)

Should really be just:

+#if defined(CONFIG_SMP)

Can you amend?

Thanks a lot,


Vegard

-- 
"The animistic metaphor of the bug that maliciously sneaked in while
the programmer was not looking is intellectually dishonest as it
disguises that the error is the programmer's own creation."
	-- E. W. Dijkstra, EWD1036

Re: [GIT PULL] kmemcheck updates for tip/kmemcheck

[Vegard Nossum]

On Mon, Sep 8, 2008 at 5:24 PM, Vegard Nossum <vegard.nossum@gmail.com> wrote:
> On Mon, Sep 8, 2008 at 5:17 PM, Ingo Molnar <mingo@elte.hu> wrote:
>> +#if defined(CONFIG_SMP) && !defined(CONFIG_KMEMCHECK_USE_SMP)
>
> Should really be just:
>
> +#if defined(CONFIG_SMP)
>
> Can you amend?

Er, also written like:

+#ifdef CONFIG_SMP

Bonk.


Vegard

-- 
"The animistic metaphor of the bug that maliciously sneaked in while
the programmer was not looking is intellectually dishonest as it
disguises that the error is the programmer's own creation."
	-- E. W. Dijkstra, EWD1036

Re: [GIT PULL] kmemcheck updates for tip/kmemcheck

[Ingo Molnar]

* Vegard Nossum <vegard.nossum@gmail.com> wrote:

> On Mon, Sep 8, 2008 at 5:24 PM, Vegard Nossum <vegard.nossum@gmail.com> wrote:
> > On Mon, Sep 8, 2008 at 5:17 PM, Ingo Molnar <mingo@elte.hu> wrote:
> >> +#if defined(CONFIG_SMP) && !defined(CONFIG_KMEMCHECK_USE_SMP)
> >
> > Should really be just:
> >
> > +#if defined(CONFIG_SMP)
> >
> > Can you amend?
> 
> Er, also written like:
> 
> +#ifdef CONFIG_SMP
> 
> Bonk.

thx, fixed it.

	Ingo

[GIT PULL] kmemcheck updates for tip/kmemcheck

[Vegard Nossum]

Hi Ingo,

A rather large update, most of it is implementing the correct REP MOVS/STOS
handling on P4s. I've tested it on both 32-bit and 64-bit and everything
seems good! Also included is the DMA API hooks which prevents false-positive
reports from reads in DMA-able memory.

--

The following changes since commit d99de804af81f223445368b8cb40d44ff12ec89a:
  Ingo Molnar (1):
        Merge commit 'v2.6.27-rc3' into kmemcheck

are available in the git repository at:

  git://git.kernel.org/pub/scm/linux/kernel/git/vegard/kmemcheck.git for-tip2

Vegard Nossum (7):
      kmemcheck: add DMA hooks
      kmemcheck: work with sizes in terms of bytes instead of bits
      kmemcheck: allow memory accesses that cross page boundaries
      kmemcheck: add some more documentation
      kmemcheck: add some comments
      kmemcheck: save memory contents on use of uninitialized memory
      kmemcheck: implement REP MOVS/STOS emulation

 arch/x86/Kconfig.debug            |    2 +-
 arch/x86/kernel/traps_32.c        |    8 +-
 arch/x86/kernel/traps_64.c        |    8 +-
 arch/x86/mm/kmemcheck/error.c     |   25 +++-
 arch/x86/mm/kmemcheck/kmemcheck.c |  322 +++++++++++++++++++++++++++----------
 arch/x86/mm/kmemcheck/opcode.c    |   38 ++++--
 arch/x86/mm/kmemcheck/opcode.h    |    3 +-
 arch/x86/mm/kmemcheck/shadow.c    |   76 ++--------
 arch/x86/mm/kmemcheck/string.c    |    4 +
 include/asm-x86/dma-mapping.h     |    2 +
 include/asm-x86/kmemcheck.h       |    6 +
 include/linux/kmemcheck.h         |   16 ++
 mm/kmemcheck.c                    |    6 +
 13 files changed, 339 insertions(+), 177 deletions(-)

Thanks,


Vegard

Re: [GIT PULL] kmemcheck updates for tip/kmemcheck

[Ingo Molnar]

* Vegard Nossum <vegard.nossum@gmail.com> wrote:

> Hi Ingo,
> 
> A rather large update, most of it is implementing the correct REP 
> MOVS/STOS handling on P4s. I've tested it on both 32-bit and 64-bit 
> and everything seems good! Also included is the DMA API hooks which 
> prevents false-positive reports from reads in DMA-able memory.

pulled, thanks Vegard. Great progress on a rather difficult aspect of 
kmemcheck!

	Ingo