[Bug #14658] Regression in efi.c

[Bug #14658] Regression in efi.c

[Rafael J. Wysocki]

This message has been generated automatically as a part of a report
of regressions introduced between 2.6.30 and 2.6.31.

The following bug entry is on the current list of known regressions
introduced between 2.6.30 and 2.6.31.  Please verify if it still should
be listed and let me know (either way).


Bug-Entry	: http://bugzilla.kernel.org/show_bug.cgi?id=14658
Subject		: Regression in efi.c
Submitter	: indexer <indexer@internode.on.net>
Date		: 2009-11-17 14:54 (5 days old)
References	: http://marc.info/?l=linux-kernel&m=125846988820120&w=4

Re: [Bug 14658] Regression in efi.c

[Thomas Gleixner]

On Sat, 21 Nov 2009, bugzilla-daemon@bugzilla.kernel.org wrote:

Switched to email. Please reply by mail (reply-to-all) and not in the
bugzilla.

> --- Comment #1 from Rafael J. Wysocki <rjw@sisk.pl>  2009-11-21 20:17:21 ---
> Caused by:
> 
> commit 7bd867dfb4e0357e06a3211ab2bd0e714110def3

That's completely bogus. That commit is not in 2.6.31. It was merged in the
2.6.32 merge window.

>From the original bug report:

http://marc.info/?l=linux-kernel&m=125846988820120&w=4

> I would like to report a possible regression in efi.c with kernels
> 2.6.31 , 2.6.32-rc5 and 2.6.32.rc7.
>
> Attempting to boot x86_64 with elilo succeeds using 2.6.30 . Using the
> same config cannot boot with any of the 3 afore mentioned kernels.

So the problem exists with 2.6.31 already which excludes the above
commit.

The above commit indeed caused problems on 64bit efi systems in
2.6.32-rc. The fix was merged between 32-rc5 and 32-rc6.

The bisect was done with:

git bisect start
# good: [59a3759d0fe8d969888c741bb33f4946e4d3750d] Linux 2.6.30-rc7
git bisect good 59a3759d0fe8d969888c741bb33f4946e4d3750d
# bad: [156171c71a0dc4bce12b4408bb1591f8fe32dc1a] Linux 2.6.32-rc7

while the bisect should have be done with

good 2.6.30
bad  2.6.31

William, could you please go through the bisect pain again?

Thanks,

	tglx

Re: [Bug 14658] Regression in efi.c

[indexer]

Thomas Gleixner wrote:
> On Sat, 21 Nov 2009, bugzilla-daemon@bugzilla.kernel.org wrote:
>
> Switched to email. Please reply by mail (reply-to-all) and not in the
> bugzilla.
>
>   
>> --- Comment #1 from Rafael J. Wysocki <rjw@sisk.pl>  2009-11-21 20:17:21 ---
>> Caused by:
>>
>> commit 7bd867dfb4e0357e06a3211ab2bd0e714110def3
>>     
>
> That's completely bogus. That commit is not in 2.6.31. It was merged in the
> 2.6.32 merge window.
>
> From the original bug report:
>
> http://marc.info/?l=linux-kernel&m=125846988820120&w=4
>
>   
>> I would like to report a possible regression in efi.c with kernels
>> 2.6.31 , 2.6.32-rc5 and 2.6.32.rc7.
>>
>> Attempting to boot x86_64 with elilo succeeds using 2.6.30 . Using the
>> same config cannot boot with any of the 3 afore mentioned kernels.
>>     
>
> So the problem exists with 2.6.31 already which excludes the above
> commit.
>
> The above commit indeed caused problems on 64bit efi systems in
> 2.6.32-rc. The fix was merged between 32-rc5 and 32-rc6.
>
> The bisect was done with:
>
> git bisect start
> # good: [59a3759d0fe8d969888c741bb33f4946e4d3750d] Linux 2.6.30-rc7
> git bisect good 59a3759d0fe8d969888c741bb33f4946e4d3750d
> # bad: [156171c71a0dc4bce12b4408bb1591f8fe32dc1a] Linux 2.6.32-rc7
>
> while the bisect should have be done with
>
> good 2.6.30
> bad  2.6.31
>
> William, could you please go through the bisect pain again?
>
> Thanks,
>
> 	tglx
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>   
Thomas

I had already done a similar bisect to this in private conversation to 
Huang Ying, and is what prompted me to expand my bisect out. I redid it 
regardless and will send you the information.

commit 74fca6a42863ffacaf7ba6f1936a9f228950f657
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date:   Wed Sep 9 15:13:59 2009 -0700

    Linux 2.6.31

 Makefile |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)


git bisect start
# good: [07a2039b8eb0af4ff464efd3dfd95de5c02648c6] Linux 2.6.30
git bisect good 07a2039b8eb0af4ff464efd3dfd95de5c02648c6
# bad: [74fca6a42863ffacaf7ba6f1936a9f228950f657] Linux 2.6.31
git bisect bad 74fca6a42863ffacaf7ba6f1936a9f228950f657
# good: [925d74ae717c9a12d3618eb4b36b9fb632e2cef3] V4L/DVB (11736): 
videobuf: modify return value of VIDIOC_REQBUFS ioctl
git bisect good 925d74ae717c9a12d3618eb4b36b9fb632e2cef3
# good: [a380137900fca5c79e6daa9500bdb6ea5649188e] ixgbe: Fix device 
capabilities of 82599 single speed fiber NICs.
git bisect good a380137900fca5c79e6daa9500bdb6ea5649188e
# good: [b54c3835469c9548d470e7788cb22a2fd7e21133] Merge branch 
'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
git bisect good b54c3835469c9548d470e7788cb22a2fd7e21133
# good: [7b2aa037e878c939676675969983284a02958ae3] Merge 
git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6
git bisect good 7b2aa037e878c939676675969983284a02958ae3
# good: [8486a0f95c844b27ecc855cfec89b7e34f831cad] Merge 
git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
git bisect good 8486a0f95c844b27ecc855cfec89b7e34f831cad
# good: [f415c413f458837bd0c27086b79aca889f9435e4] Merge 
git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
git bisect good f415c413f458837bd0c27086b79aca889f9435e4
# good: [37d0892c5a94e208cf863e3b7bac014edee4346d] autofs4 - fix missed 
case when changing to use struct path
git bisect good 37d0892c5a94e208cf863e3b7bac014edee4346d
# good: [0edfa2b1b5a5e1475e76dd3c792447687d966de4] Merge branch 
'for-linus' of git://oss.sgi.com/xfs/xfs
git bisect good 0edfa2b1b5a5e1475e76dd3c792447687d966de4
# good: [5136a6c0fd5b26bbf39ad761cf7a4fc563ad83a3] Merge 
git://git.infradead.org/~dwmw2/mtd-2.6.31
git bisect good 5136a6c0fd5b26bbf39ad761cf7a4fc563ad83a3
# good: [f69fb9c39868463f6b0b8306824341bd5610250b] Merge branch 
'for-linus' of 
git://git.kernel.org/pub/scm/linux/kernel/git/anholt/drm-intel
git bisect good f69fb9c39868463f6b0b8306824341bd5610250b
# good: [755ae761c5519929a97567d61a379b87352c337c] Merge branch 
'for-linus' of 
git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6
git bisect good 755ae761c5519929a97567d61a379b87352c337c
# good: [7c8460db30dfd085ef3837c8fb02ecf2e718b983] drm/i915: fix mask 
bits setting
git bisect good 7c8460db30dfd085ef3837c8fb02ecf2e718b983
# good: [7135a71b19be1faf48b7148d77844d03bc0717d6] aoe: allocate unused 
request_queue for sysfs
git bisect good 7135a71b19be1faf48b7148d77844d03bc0717d6
 Sincerely

William

PS - The fact remains i still cant efi boot in 2.6.32-rc7, so this bug 
still exists, and I will keep doing what i can to help track this down.

Re: [Bug 14658] Regression in efi.c

[indexer]

Tang, Feng wrote:
> Hi William,
>
> One stupid question, when you did the bisect, did you really try to boot every kernel that you built out?
>
> - Feng
>   
>> -----Original Message-----
>> Thomas
>>
>> I had already done a similar bisect to this in private conversation to
>> Huang Ying, and is what prompted me to expand my bisect out. I redid it
>> regardless and will send you the information.
>>
>> commit 74fca6a42863ffacaf7ba6f1936a9f228950f657
>> Author: Linus Torvalds <torvalds@linux-foundation.org>
>> Date:   Wed Sep 9 15:13:59 2009 -0700
>>
>>    Linux 2.6.31
>>
>> Makefile |    2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>
>>
>> git bisect start
>> # good: [07a2039b8eb0af4ff464efd3dfd95de5c02648c6] Linux 2.6.30
>> git bisect good 07a2039b8eb0af4ff464efd3dfd95de5c02648c6
>> # bad: [74fca6a42863ffacaf7ba6f1936a9f228950f657] Linux 2.6.31
>> git bisect bad 74fca6a42863ffacaf7ba6f1936a9f228950f657
>> # good: [925d74ae717c9a12d3618eb4b36b9fb632e2cef3] V4L/DVB (11736):
>> videobuf: modify return value of VIDIOC_REQBUFS ioctl
>> git bisect good 925d74ae717c9a12d3618eb4b36b9fb632e2cef3
>> # good: [a380137900fca5c79e6daa9500bdb6ea5649188e] ixgbe: Fix device
>> capabilities of 82599 single speed fiber NICs.
>> git bisect good a380137900fca5c79e6daa9500bdb6ea5649188e
>> # good: [b54c3835469c9548d470e7788cb22a2fd7e21133] Merge branch
>> 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound-2.6
>> git bisect good b54c3835469c9548d470e7788cb22a2fd7e21133
>> # good: [7b2aa037e878c939676675969983284a02958ae3] Merge
>> git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6
>> git bisect good 7b2aa037e878c939676675969983284a02958ae3
>> # good: [8486a0f95c844b27ecc855cfec89b7e34f831cad] Merge
>> git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
>> git bisect good 8486a0f95c844b27ecc855cfec89b7e34f831cad
>> # good: [f415c413f458837bd0c27086b79aca889f9435e4] Merge
>> git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
>> git bisect good f415c413f458837bd0c27086b79aca889f9435e4
>> # good: [37d0892c5a94e208cf863e3b7bac014edee4346d] autofs4 - fix missed
>> case when changing to use struct path
>> git bisect good 37d0892c5a94e208cf863e3b7bac014edee4346d
>> # good: [0edfa2b1b5a5e1475e76dd3c792447687d966de4] Merge branch
>> 'for-linus' of git://oss.sgi.com/xfs/xfs
>> git bisect good 0edfa2b1b5a5e1475e76dd3c792447687d966de4
>> # good: [5136a6c0fd5b26bbf39ad761cf7a4fc563ad83a3] Merge
>> git://git.infradead.org/~dwmw2/mtd-2.6.31
>> git bisect good 5136a6c0fd5b26bbf39ad761cf7a4fc563ad83a3
>> # good: [f69fb9c39868463f6b0b8306824341bd5610250b] Merge branch
>> 'for-linus' of
>> git://git.kernel.org/pub/scm/linux/kernel/git/anholt/drm-intel
>> git bisect good f69fb9c39868463f6b0b8306824341bd5610250b
>> # good: [755ae761c5519929a97567d61a379b87352c337c] Merge branch
>> 'for-linus' of
>> git://git.kernel.org/pub/scm/linux/kernel/git/jmorris/security-testing-2.6
>> git bisect good 755ae761c5519929a97567d61a379b87352c337c
>> # good: [7c8460db30dfd085ef3837c8fb02ecf2e718b983] drm/i915: fix mask
>> bits setting
>> git bisect good 7c8460db30dfd085ef3837c8fb02ecf2e718b983
>> # good: [7135a71b19be1faf48b7148d77844d03bc0717d6] aoe: allocate unused
>> request_queue for sysfs
>> git bisect good 7135a71b19be1faf48b7148d77844d03bc0717d6
>> Sincerely
>>
>> William
>>
>> PS - The fact remains i still cant efi boot in 2.6.32-rc7, so this bug
>> still exists, and I will keep doing what i can to help track this down.
>>     
Feng

Yes, i did boot every single on. We had this problem where it was the 
makefile when i did a similar bisect with Huang Ying, and it is what 
prompted me to expand the bisect parameters to v2.6.32. I believe the 
email about that is in the  mailing list still, and im more than happy 
to do any other tests you can think of to test and isolate this bug as 
it currently causes a significant issue.

Cheers

William

Re: [Bug 14658] Regression in efi.c

[Thomas Gleixner]

On Mon, 23 Nov 2009, indexer wrote:
> > while the bisect should have be done with
> > 
> > good 2.6.30
> > bad  2.6.31
> > 
> > William, could you please go through the bisect pain again?
>> 
> I had already done a similar bisect to this in private conversation to Huang
> Ying, and is what prompted me to expand my bisect out. I redid it regardless
> and will send you the information.
> 
> commit 74fca6a42863ffacaf7ba6f1936a9f228950f657
> Author: Linus Torvalds <torvalds@linux-foundation.org>
> Date:   Wed Sep 9 15:13:59 2009 -0700
> 
>    Linux 2.6.31
> 
> Makefile |    2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)

So you are saying that 
 
> # good: [7135a71b19be1faf48b7148d77844d03bc0717d6] aoe: allocate unused
> request_queue for sysfs
> git bisect good 7135a71b19be1faf48b7148d77844d03bc0717d6

which is the last code changing commit before the 2.6.31 release is
booting fine, but with the release commit which merily changes the
Makefile it is not ?

> PS - The fact remains i still cant efi boot in 2.6.32-rc7, so this bug still
> exists, and I will keep doing what i can to help track this down.

Sure, but I hope we agree that it does not make much sense to search
2.6.31..now when we already know that 2.6.31 is not booting and the
change which causes the problem is between 2.6.30 and 2.6.31.

Thanks,

	tglx

Re: [Bug 14658] Regression in efi.c

[indexer]

Thomas Gleixner wrote:
> On Mon, 23 Nov 2009, indexer wrote:
>   
>>> while the bisect should have be done with
>>>
>>> good 2.6.30
>>> bad  2.6.31
>>>
>>> William, could you please go through the bisect pain again?
>>>
>>>       
>> I had already done a similar bisect to this in private conversation to Huang
>> Ying, and is what prompted me to expand my bisect out. I redid it regardless
>> and will send you the information.
>>
>> commit 74fca6a42863ffacaf7ba6f1936a9f228950f657
>> Author: Linus Torvalds <torvalds@linux-foundation.org>
>> Date:   Wed Sep 9 15:13:59 2009 -0700
>>
>>    Linux 2.6.31
>>
>> Makefile |    2 +-
>> 1 files changed, 1 insertions(+), 1 deletions(-)
>>     
>
> So you are saying that 
>  
>   
>> # good: [7135a71b19be1faf48b7148d77844d03bc0717d6] aoe: allocate unused
>> request_queue for sysfs
>> git bisect good 7135a71b19be1faf48b7148d77844d03bc0717d6
>>     
>
> which is the last code changing commit before the 2.6.31 release is
> booting fine, but with the release commit which merily changes the
> Makefile it is not ?
>   
I dont know, I will test this as soon as possible. I might change the 
bisect parameters a bit, but really im quite new to all this. It may 
also be a possibility it is gen-patches breaking it as i reported 
2.6.30-gentoo was the broken kernel. Either way, the fact remains that 
on 2.6.31 and higher vanilla it is broken.
>   
>> PS - The fact remains i still cant efi boot in 2.6.32-rc7, so this bug still
>> exists, and I will keep doing what i can to help track this down.
>>     
>
> Sure, but I hope we agree that it does not make much sense to search
> 2.6.31..now when we already know that 2.6.31 is not booting and the
> change which causes the problem is between 2.6.30 and 2.6.31.
>
> Thanks,
>
> 	tglx
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>   
 Oh i have no issues agreeing with you, but its really confusing me, and 
i of course may have made an error in what i reported. I am after all 
only human, and i might have made a mistake, and this is the first time 
i have ever submitted a bug, let alone found one like this. The end 
product is that 2.6.32 doesn't boot, but maybe the last working version 
was not 2.6.30-rc8, maybe it was a 2.6.31 kernel. I might redo the 
bisect and try this out again, as i'm quite intent on finding the source 
of this. I will also see if i can get my hands on another x86_64 linux 
install with efi to test it on a different machine to see if it is 
hardware specific to my computer. For the record i am running a late 
2009 macbook pro gen 5 rev 3, efi version 1.7 with refit and elilo.

William

Re: [Bug 14658] Regression in efi.c

[indexer]

[-- Attachment #1: Type: text/plain, Size: 1465 bytes --]

To whom it may concern

I have redone the bisect and i believe the versions i originally stated 
were wrong. I apologise for this mistake

commit 7bd867dfb4e0357e06a3211ab2bd0e714110def3
Author: Feng Tang <feng.tang@intel.com>
Date:   Thu Sep 10 10:48:56 2009 +0800

    x86: Move get/set_wallclock to x86_platform_ops
   
    get/set_wallclock() have already a set of platform dependent
    implementations (default, EFI, paravirt). MRST will add another
    variant.
   
    Moving them to platform ops simplifies the existing code and minimizes
    the effort to integrate new variants.
   
    Signed-off-by: Feng Tang <feng.tang@intel.com>
    LKML-Reference: <new-submission>
    Signed-off-by: Thomas Gleixner <tglx@linutronix.de>

 arch/x86/include/asm/paravirt.h       |   10 ------
 arch/x86/include/asm/paravirt_types.h |    4 --
 arch/x86/include/asm/time.h           |   50 
---------------------------------
 arch/x86/include/asm/x86_init.h       |    4 ++
 arch/x86/kernel/efi.c                 |    4 ++
 arch/x86/kernel/kvmclock.c            |    4 +-
 arch/x86/kernel/paravirt.c            |    2 -
 arch/x86/kernel/rtc.c                 |   12 ++-----
 arch/x86/kernel/vmi_32.c              |    4 +-
 arch/x86/kernel/x86_init.c            |    2 +
 arch/x86/lguest/boot.c                |    4 +--
 arch/x86/xen/enlighten.c              |    4 +-
 12 files changed, 21 insertions(+), 83 deletions(-)


Also find my bisect log attached.

William

[-- Attachment #2: bisect-log --]
[-- Type: text/plain, Size: 2556 bytes --]

git bisect start
# good: [28d0325ce6e0a52f53d8af687e6427fee59004d3] Linux 2.6.31-rc1
git bisect good 28d0325ce6e0a52f53d8af687e6427fee59004d3
# bad: [17d857be649a21ca90008c6dc425d849fa83db5c] Linux 2.6.32-rc1
git bisect bad 17d857be649a21ca90008c6dc425d849fa83db5c
# good: [99bc47067910f7070e65ee318a6dd79a2371f1e5] Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc-next-2.6
git bisect good 99bc47067910f7070e65ee318a6dd79a2371f1e5
# good: [515b696b282f856c3ad1679ccd658120faa387d0] Merge git://git.kernel.org/pub/scm/linux/kernel/git/lethal/sh-2.6
git bisect good 515b696b282f856c3ad1679ccd658120faa387d0
# bad: [be90a49ca22a95f184d9f32d35b5247b44032849] Merge git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb-2.6
git bisect bad be90a49ca22a95f184d9f32d35b5247b44032849
# bad: [8e4bc3dd2ca07d77882eba73cea240aba95a1854] Merge branch 'sched-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
git bisect bad 8e4bc3dd2ca07d77882eba73cea240aba95a1854
# bad: [047efc7db9fea32703e8ba04629562e52cccdf88] V4L/DVB (13026): s2250-board: Implement brightness and contrast controls
git bisect bad 047efc7db9fea32703e8ba04629562e52cccdf88
# bad: [650901c0b6917505e81f6593d230ea3cdcf6518a] V4L/DVB (12868): tda18271: improve error log in function tda18271_write_regs
git bisect bad 650901c0b6917505e81f6593d230ea3cdcf6518a
# good: [3530c1886291df061e3972c55590777ef1cb67f8] Merge branch 'for_linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4
git bisect good 3530c1886291df061e3972c55590777ef1cb67f8
# good: [54e2603f1a85b9725aa13518d69148b6e7061aa9] x86: platform: Fix section annotations
git bisect good 54e2603f1a85b9725aa13518d69148b6e7061aa9
# bad: [19bf00384a6d5bbe5d7b8afbcc25772e3675d423] V4L/DVB (12748): em28xx: implement g_std v4l call
git bisect bad 19bf00384a6d5bbe5d7b8afbcc25772e3675d423
# bad: [53e712d0844e99b8d8720327470b86ef401fb727] V4L/DVB (12734): cx25821: Fix some compilation troubles
git bisect bad 53e712d0844e99b8d8720327470b86ef401fb727
# bad: [78f28b7c555359c67c2a0d23f7436e915329421e] Merge branch 'x86-platform-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
git bisect bad 78f28b7c555359c67c2a0d23f7436e915329421e
# good: [9c88b25bcfde51eccf2cca13d933a2948d0ca4eb] frv: clean up linker script using new linker script macros.
git bisect good 9c88b25bcfde51eccf2cca13d933a2948d0ca4eb
# bad: [7bd867dfb4e0357e06a3211ab2bd0e714110def3] x86: Move get/set_wallclock to x86_platform_ops
git bisect bad 7bd867dfb4e0357e06a3211ab2bd0e714110def3

Re: [Bug 14658] Regression in efi.c

[Huang Ying]

Hi, William,

On Wed, 2009-11-25 at 08:35 +0800, indexer wrote: 
> To whom it may concern
> 
> I have redone the bisect and i believe the versions i originally stated 
> were wrong. I apologise for this mistake
> 
> commit 7bd867dfb4e0357e06a3211ab2bd0e714110def3
> Author: Feng Tang <feng.tang@intel.com>
> Date:   Thu Sep 10 10:48:56 2009 +0800
> 
>     x86: Move get/set_wallclock to x86_platform_ops
>    
>     get/set_wallclock() have already a set of platform dependent
>     implementations (default, EFI, paravirt). MRST will add another
>     variant.
>    
>     Moving them to platform ops simplifies the existing code and minimizes
>     the effort to integrate new variants.

Now Legacy RTC code instead of EFI get/set_wallclock is used on x86_64.
Your test shows that before this patch, kernel can boot on your system,
but can not boot after that?

Best Regards,
Huang Ying

Re: [Bug 14658] Regression in efi.c

[indexer]

Huang Ying wrote:
> Hi, William,
>
> On Wed, 2009-11-25 at 08:35 +0800, indexer wrote: 
>   
>> To whom it may concern
>>
>> I have redone the bisect and i believe the versions i originally stated 
>> were wrong. I apologise for this mistake
>>
>> commit 7bd867dfb4e0357e06a3211ab2bd0e714110def3
>> Author: Feng Tang <feng.tang@intel.com>
>> Date:   Thu Sep 10 10:48:56 2009 +0800
>>
>>     x86: Move get/set_wallclock to x86_platform_ops
>>    
>>     get/set_wallclock() have already a set of platform dependent
>>     implementations (default, EFI, paravirt). MRST will add another
>>     variant.
>>    
>>     Moving them to platform ops simplifies the existing code and minimizes
>>     the effort to integrate new variants.
>>     
>
> Now Legacy RTC code instead of EFI get/set_wallclock is used on x86_64.
> Your test shows that before this patch, kernel can boot on your system,
> but can not boot after that?
>
> Best Regards,
> Huang Ying
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>   
Huang Ying

Yes that is correct, i cant boot after this patch was applied, but 
former to this patch, my system would boot.

William

Re: [Bug 14658] Regression in efi.c

[indexer]

Feng tang

I have just tested and applied that correction, and it resulted in a 
working system. I dont really know where to go from here so any advice 
would be appreciated.

William

RE: [Bug 14658] Regression in efi.c

[Tang, Feng]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="gb2312", Size: 883 bytes --]

William,

You've confirmed that apply commit 772be899b "x86: Make EFI RTC function depend
on 32bit again" right after 7bd867dfb "x86: Move get/set_wallclock to 
x86_platform_ops" will get a bootable kernel.

Then, one debug method may be use "git-rebase -i" to move commit 772be899b right after 
7bd867dfb, and do the bisect from there on. 

Thanks,
Feng
>-----Original Message-----
>From: indexer [mailto:indexer@internode.on.net]
>Sent: 2009Äê11ÔÂ25ÈÕ 17:43
>To: Tang, Feng
>Cc: LKML
>Subject: Re: [Bug 14658] Regression in efi.c
>
>Feng tang
>
>I have just tested and applied that correction, and it resulted in a
>working system. I dont really know where to go from here so any advice
>would be appreciated.
>
>William
ÿôèº{.nÇ+‰·Ÿ®‰­†+%ŠËÿ±éݶ\x17¥Šwÿº{.nÇ+‰·¥Š{±þG«éÿŠ{ayº\x1dʇڙë,j\a­¢f£¢·hšïêÿ‘êçz_è®\x03(­éšŽŠÝ¢j"ú\x1a¶^[m§ÿÿ¾\a«þG«éÿ¢¸?™¨è­Ú&£ø§~á¶iO•æ¬z·švØ^\x14\x04\x1a¶^[m§ÿÿÃ\fÿ¶ìÿ¢¸?–I¥

Re: [Bug 14658] Regression in efi.c

[indexer]

Feng, Peter

I decided to go back a few versions and check out if there any previous 
bugs, and upon bisecting i found this one

commit 37ba7ab5e33cebc25c68fffe33e9f21e7c2014e8
Author: H. Peter Anvin <hpa@zytor.com>
Date: Mon May 11 15:56:08 2009 -0700

x86, boot: make kernel_alignment adjustable; new bzImage fields

Make the kernel_alignment field adjustable; this allows us to set it
to a large value (intended to be 16 MB to avoid ZONE_DMA contention,
memory holes and other weirdness) while a smart bootloader can still
force a loading at a lesser alignment if absolutely necessary.

Also export pref_address (preferred loading address, corresponding to
the link-time address) and init_size, the total amount of linear
memory the kernel will require during initialization.

[ Impact: allows better kernel placement, gives bootloader more info ]

Signed-off-by: H. Peter Anvin <hpa@zytor.com>

arch/x86/boot/compressed/head_32.S | 7 +++++--
arch/x86/boot/compressed/head_64.S | 14 ++++++++++----
arch/x86/boot/header.S | 15 +++++++++++++--
arch/x86/include/asm/boot.h | 15 +++++++++++++++
arch/x86/kernel/asm-offsets_32.c | 1 +
arch/x86/kernel/asm-offsets_64.c | 1 +
6 files changed, 45 insertions(+), 8 deletions(-)


This seems to stop efi booting as well on x86_64. I have also been able 
to recently test this on a gen 4 macbook pro as well as my own gen 5 to 
make sure it is not an issue specific to this model.

William

Tang, Feng wrote:
> William,
>
> You've confirmed that apply commit 772be899b "x86: Make EFI RTC function depend
> on 32bit again" right after 7bd867dfb "x86: Move get/set_wallclock to 
> x86_platform_ops" will get a bootable kernel.
>
> Then, one debug method may be use "git-rebase -i" to move commit 772be899b right after 
> 7bd867dfb, and do the bisect from there on. 
>
> Thanks,
> Feng
>   
>> -----Original Message-----
>> From: indexer [mailto:indexer@internode.on.net]
>> Sent: 2009年11月25日 17:43
>> To: Tang, Feng
>> Cc: LKML
>> Subject: Re: [Bug 14658] Regression in efi.c
>>
>> Feng tang
>>
>> I have just tested and applied that correction, and it resulted in a
>> working system. I dont really know where to go from here so any advice
>> would be appreciated.
>>
>> William
>>     
> N嫥叉靣笡y氊b瞂千v豝�)藓{.n�+壏{睉赙zXФ\x17洝塄}财爖�&j:+v墾�\a珣赙zZ+€�+zf＂穐殘啳嗃i�z�\x1e畐ア�?櫒璀�&�)撷^[f旟^j谦y呩@A玜囤\x7f�\f0鹅h�\x0f鍜i\x7f