[MIPS]: Use mcontext_t structure for MIPS

[MIPS]: Use mcontext_t structure for MIPS

This change removes user_regs_struct and
user_fpregs_struct structures for mips
and uses mcontext_t instead.

[fdegans]

non-owner lgtm, just one comment.

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
File common/android/include/sys/user.h (right):

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
common/android/include/sys/user.h:37: //   in NDK but exist in glibc.
Is there some bug tracking this on the Android side?

[rmcilroy]

Thanks for the fixes, some comments below.

https://breakpad.appspot.com/3744002/diff/30001/client/linux/dump_writer_comm...
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/client/linux/dump_writer_comm...
client/linux/dump_writer_common/thread_info.cc:246: out->mdhi =  mcontext.mdhi;
remove extra whitespace after '=' (here and below)

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...
client/linux/minidump_writer/linux_ptrace_dumper.cc:195: #else
I don't like all this #ifdefing. Please add some getters to thread_info.h which
allow getting the regs and fregs fields from the context and use them here to
avoid the #ifdefing.

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...
client/linux/minidump_writer/linux_ptrace_dumper.cc:275: 
nit - remove the newline.

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
File common/android/include/sys/user.h (right):

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
common/android/include/sys/user.h:43: #ifdef __mips__
Could you move this whole block down into the same if/else block as __i386__
below (i.e. make this an #else if block)

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
common/android/include/sys/user.h:102: #endif
add "  // __mips__"

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
File tools/linux/md2core/minidump-2-core.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:214: #endif
please add comments describing the 'if' which each endif corresponds too (here
and below

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:389: for (int i = 0; i <
MD_FLOATINGSAVEAREA_MIPS_FPR_COUNT; ++i)
use '{' since this is now more than two lines (codestyle)

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:391: rawregs->float_save.regs[i];
indent by four spaces for line continuation.

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:759: memcpy(&pr.pr_reg,
&thread.mcontext.gregs, sizeof(user_regs_struct));
nit - use the getters added to thread_info for this too

[gordanac]

Hi Ross,

Please take a look into new patchsets:

Patchset #3 includes code rebase as well.
Patchset #4 is using register definitions from asm/reg.h from NDK.

And some comments:


https://breakpad.appspot.com/3744002/diff/30001/client/linux/dump_writer_comm...>
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/client/linux/dump_writer_comm...>
client/linux/dump_writer_common/thread_info.cc:246: out->mdhi =  mcontext.mdhi;
On 2015/04/08 08:02:07, rmcilroy wrote:
> remove extra whitespace after '=' (here and below)

Done.

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...>
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:195: #else
On 2015/04/08 08:02:07, rmcilroy wrote:
> I don't like all this #ifdefing. Please add some getters to thread_info.h
which
> allow getting the regs and fregs fields from the context and use them here to
> avoid the #ifdefing.

Done.

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:275:
On 2015/04/08 08:02:07, rmcilroy wrote:
> nit - remove the newline.

Done.

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...>
File common/android/include/sys/user.h (right):

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...>
common/android/include/sys/user.h:37: //   in NDK but exist in glibc.
On 2015/04/07 09:46:40, fdegans wrote:
> Is there some bug tracking this on the Android side?

I took a double look in bionic and register definitions are there in asm/reg.h,
though slightly different from those in user.h in glibc. Take a look into
patchset 4.

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...>
common/android/include/sys/user.h:43: #ifdef __mips__
On 2015/04/08 08:02:07, rmcilroy wrote:
> Could you move this whole block down into the same if/else block as __i386__
> below (i.e. make this an #else if block)

Done.

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...>
common/android/include/sys/user.h:102: #endif
On 2015/04/08 08:02:07, rmcilroy wrote:
> add "  // __mips__"

Done.

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...>
File tools/linux/md2core/minidump-2-core.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...>
tools/linux/md2core/minidump-2-core.cc:214: #endif
On 2015/04/08 08:02:07, rmcilroy wrote:
> please add comments describing the 'if' which each endif corresponds too
(here
> and below

Done.

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...>
tools/linux/md2core/minidump-2-core.cc:389: for (int i = 0; i <
MD_FLOATINGSAVEAREA_MIPS_FPR_COUNT; ++i)
On 2015/04/08 08:02:07, rmcilroy wrote:
> use '{' since this is now more than two lines (codestyle)

Done.

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...>
tools/linux/md2core/minidump-2-core.cc:391: rawregs->float_save.regs[i];
On 2015/04/08 08:02:07, rmcilroy wrote:
> indent by four spaces for line continuation.

Done.

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...>
tools/linux/md2core/minidump-2-core.cc:759: memcpy(&pr.pr_reg,
&thread.mcontext.gregs, sizeof(user_regs_struct));
On 2015/04/08 08:02:07, rmcilroy wrote:
> nit - use the getters added to thread_info for this too

This is Thread class, defined in this file (above), not thread_info.
In this case, I think, leaving it as is is more simple than adding getters for
this class as well.

Regards
Gordana

________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Wednesday, April 08, 2015 10:02 AM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

Thanks for the fixes, some comments below.


https://breakpad.appspot.com/3744002/diff/30001/client/linux/dump_writer_comm...
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/client/linux/dump_writer_comm...
client/linux/dump_writer_common/thread_info.cc:246: out->mdhi =
mcontext.mdhi;
remove extra whitespace after '=' (here and below)

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...
client/linux/minidump_writer/linux_ptrace_dumper.cc:195: #else
I don't like all this #ifdefing. Please add some getters to
thread_info.h which allow getting the regs and fregs fields from the
context and use them here to avoid the #ifdefing.

https://breakpad.appspot.com/3744002/diff/30001/client/linux/minidump_writer/...
client/linux/minidump_writer/linux_ptrace_dumper.cc:275:
nit - remove the newline.

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
File common/android/include/sys/user.h (right):

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
common/android/include/sys/user.h:43: #ifdef __mips__
Could you move this whole block down into the same if/else block as
__i386__ below (i.e. make this an #else if block)

https://breakpad.appspot.com/3744002/diff/30001/common/android/include/sys/us...
common/android/include/sys/user.h:102: #endif
add " // __mips__"

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
File tools/linux/md2core/minidump-2-core.cc (right):

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:214: #endif
please add comments describing the 'if' which each endif corresponds too
(here and below

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:389: for (int i = 0; i <
MD_FLOATINGSAVEAREA_MIPS_FPR_COUNT; ++i)
use '{' since this is now more than two lines (codestyle)

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:391: rawregs->float_save.regs[i];
indent by four spaces for line continuation.

https://breakpad.appspot.com/3744002/diff/30001/tools/linux/md2core/minidump-...
tools/linux/md2core/minidump-2-core.cc:759: memcpy(&pr.pr_reg,
&thread.mcontext.gregs, sizeof(user_regs_struct));
nit - use the getters added to thread_info for this too

https://breakpad.appspot.com/3744002/

[rmcilroy]

Looking a lot better, a few more comments.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:266: // Returns the pointer to
general purpose register area.
Don't need the comments here, only in the header (here and below)

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:271: return &regs;
Fix indentation

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:280: return sizeof(regs);
indentation

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:193: io.iov_base =
const_cast<void*>(info->GetGeneralPurposeRegisters());
Don't cast away const - make the function return void* instead of const void* to
make this explicit.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:194: io.iov_len =
info->GetGeneralPurposeRegistersSize();
I think this would be cleaner if you had a single function:
 GetGeneralPurposeRegisters(void** regs, size_t* size) {
which could be called as:
 GetGeneralPurposeRegisters(&io.iov_base, &io.iov_len);

Same for Float registers below.

[gordanac]

Hi Ross,
New patchset is uploaded with comments:


https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
client/linux/dump_writer_common/thread_info.cc:266: // Returns the pointer to
general purpose register area.
On 2015/04/09 14:07:14, rmcilroy wrote:
> Don't need the comments here, only in the header (here and below)

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
client/linux/dump_writer_common/thread_info.cc:271: return &regs;
On 2015/04/09 14:07:14, rmcilroy wrote:
> Fix indentation

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
client/linux/dump_writer_common/thread_info.cc:280: return sizeof(regs);
On 2015/04/09 14:07:14, rmcilroy wrote:
> indentation

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...>
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:193: io.iov_base =
const_cast<void*>(info->GetGeneralPurposeRegisters());
On 2015/04/09 14:07:14, rmcilroy wrote:
> Don't cast away const - make the function return void* instead of const void*
to
> make this explicit.

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:194: io.iov_len =
info->GetGeneralPurposeRegistersSize();
On 2015/04/09 14:07:14, rmcilroy wrote:
> I think this would be cleaner if you had a single function:
>  GetGeneralPurposeRegisters(void** regs, size_t* size) {
> which could be called as:
>  GetGeneralPurposeRegisters(&io.iov_base, &io.iov_len);
>
> Same for Float registers below.

Done.

I didn't make both arguments mandatory. That would require one more (unused)
variable in linux_ptrace_dumper.cc.

Regards
Gordana
________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Thursday, April 09, 2015 4:07 PM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

Looking a lot better, a few more comments.


https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:266: // Returns the
pointer to general purpose register area.
Don't need the comments here, only in the header (here and below)

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:271: return &regs;
Fix indentation

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:280: return sizeof(regs);
indentation

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:193: io.iov_base =
const_cast<void*>(info->GetGeneralPurposeRegisters());
Don't cast away const - make the function return void* instead of const
void* to make this explicit.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:194: io.iov_len =
info->GetGeneralPurposeRegistersSize();
I think this would be cleaner if you had a single function:
GetGeneralPurposeRegisters(void** regs, size_t* size) {
which could be called as:
GetGeneralPurposeRegisters(&io.iov_base, &io.iov_len);

Same for Float registers below.

https://breakpad.appspot.com/3744002/

[gordanac]

Hi Ross,

In case  https://breakpad.appspot.com/6824002/ gets merged I will have to rebase
this one since they modify the same parts of code.

Please have in mind that I will be on holidays until Wednesday and cannot make
additional changes till then.

Regards,
Gordana

________________________________
From: Gordana Cmiljanovic
Sent: Thursday, April 09, 2015 10:33 PM
To: fdegans@chromium.org; mark@chromium.org; primiano@chromium.org;
rmcilroy@chromium.org; Petar Jovanovic; Paul Lind;
reply@breakpad-hr.appspotmail.com
Subject: RE: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

Hi Ross,
New patchset is uploaded with comments:


https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
client/linux/dump_writer_common/thread_info.cc:266: // Returns the pointer to
general purpose register area.
On 2015/04/09 14:07:14, rmcilroy wrote:
> Don't need the comments here, only in the header (here and below)

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
client/linux/dump_writer_common/thread_info.cc:271: return &regs;
On 2015/04/09 14:07:14, rmcilroy wrote:
> Fix indentation

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...>
client/linux/dump_writer_common/thread_info.cc:280: return sizeof(regs);
On 2015/04/09 14:07:14, rmcilroy wrote:
> indentation

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...>
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:193: io.iov_base =
const_cast<void*>(info->GetGeneralPurposeRegisters());
On 2015/04/09 14:07:14, rmcilroy wrote:
> Don't cast away const - make the function return void* instead of const void*
to
> make this explicit.

Done.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:194: io.iov_len =
info->GetGeneralPurposeRegistersSize();
On 2015/04/09 14:07:14, rmcilroy wrote:
> I think this would be cleaner if you had a single function:
>  GetGeneralPurposeRegisters(void** regs, size_t* size) {
> which could be called as:
>  GetGeneralPurposeRegisters(&io.iov_base, &io.iov_len);
>
> Same for Float registers below.

Done.

I didn't make both arguments mandatory. That would require one more (unused)
variable in linux_ptrace_dumper.cc.

Regards
Gordana
________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Thursday, April 09, 2015 4:07 PM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

Looking a lot better, a few more comments.


https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
File client/linux/dump_writer_common/thread_info.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:266: // Returns the
pointer to general purpose register area.
Don't need the comments here, only in the header (here and below)

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:271: return &regs;
Fix indentation

https://breakpad.appspot.com/3744002/diff/240001/client/linux/dump_writer_com...
client/linux/dump_writer_common/thread_info.cc:280: return sizeof(regs);
indentation

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:193: io.iov_base =
const_cast<void*>(info->GetGeneralPurposeRegisters());
Don't cast away const - make the function return void* instead of const
void* to make this explicit.

https://breakpad.appspot.com/3744002/diff/240001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:194: io.iov_len =
info->GetGeneralPurposeRegistersSize();
I think this would be cleaner if you had a single function:
GetGeneralPurposeRegisters(void** regs, size_t* size) {
which could be called as:
GetGeneralPurposeRegisters(&io.iov_base, &io.iov_len);

Same for Float registers below.

https://breakpad.appspot.com/3744002/

[rmcilroy]

One last comment but then lgtm.

https://breakpad.appspot.com/3744002/diff/350001/client/linux/minidump_writer...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/350001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:203: void* addr;
have a seperate variable for both regs and fregs here please

[gordanac]

https://breakpad.appspot.com/3744002/diff/350001/client/linux/minidump_writer...>
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/350001/client/linux/minidump_writer...>
client/linux/minidump_writer/linux_ptrace_dumper.cc:203: void* addr;
On 2015/04/10 16:56:30, rmcilroy wrote:
> have a seperate variable for both regs and fregs here please

Done.

________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Friday, April 10, 2015 6:56 PM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

One last comment but then lgtm.


https://breakpad.appspot.com/3744002/diff/350001/client/linux/minidump_writer...
File client/linux/minidump_writer/linux_ptrace_dumper.cc (right):

https://breakpad.appspot.com/3744002/diff/350001/client/linux/minidump_writer...
client/linux/minidump_writer/linux_ptrace_dumper.cc:203: void* addr;
have a seperate variable for both regs and fregs here please

https://breakpad.appspot.com/3744002/

[rmcilroy]

lgtm, thanks. 

Did you test this doens't break anything on the other architectures?

[gordanac]

Tests run (beside mipsel): Breakpads's make check for x86-64 Linux and
Chromium's breakpad_unittests for ARM Android. 
________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Wednesday, April 15, 2015 2:31 PM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

lgtm, thanks.

Did you test this doens't break anything on the other architectures?

https://breakpad.appspot.com/3744002/

[rmcilroy]

On 2015/04/15 13:58:37, gordanac wrote:
> Tests run (beside mipsel): Breakpads's make check for x86-64 Linux and
> Chromium's breakpad_unittests for ARM Android. 
> ________________________________________
> From: mailto:rmcilroy@chromium.org [rmcilroy@chromium.org]
> Sent: Wednesday, April 15, 2015 2:31 PM
> To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
> mailto:primiano@chromium.org
> Cc: Petar Jovanovic; Paul Lind; mailto:reply@breakpad-hr.appspotmail.com
> Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)
> 
> lgtm, thanks.
> 
> Did you test this doens't break anything on the other architectures?
> 
> https://breakpad.appspot.com/3744002/

Great. Would you like me to land this or are you waiting for another change to
be merged first?

[gordanac]

Lets wait till tomorrow.
________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Wednesday, April 15, 2015 4:33 PM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

On 2015/04/15 13:58:37, gordanac wrote:
> Tests run (beside mipsel): Breakpads's make check for x86-64 Linux and
> Chromium's breakpad_unittests for ARM Android.
> ________________________________________
> From: mailto:rmcilroy@chromium.org [rmcilroy@chromium.org]
> Sent: Wednesday, April 15, 2015 2:31 PM
> To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
> mailto:primiano@chromium.org
> Cc: Petar Jovanovic; Paul Lind;
mailto:reply@breakpad-hr.appspotmail.com
> Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

> lgtm, thanks.

> Did you test this doens't break anything on the other architectures?

> https://breakpad.appspot.com/3744002/

Great. Would you like me to land this or are you waiting for another
change to be merged first?

https://breakpad.appspot.com/3744002/

[gordanac]

Rebased. New patch set uploaded.
________________________________________
From: Gordana Cmiljanovic
Sent: Wednesday, April 15, 2015 6:34 PM
To: fdegans@chromium.org; mark@chromium.org; primiano@chromium.org;
rmcilroy@chromium.org; Petar Jovanovic; Paul Lind;
reply@breakpad-hr.appspotmail.com
Subject: RE: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

Lets wait till tomorrow.
________________________________________
From: rmcilroy@chromium.org [rmcilroy@chromium.org]
Sent: Wednesday, April 15, 2015 4:33 PM
To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
primiano@chromium.org
Cc: Petar Jovanovic; Paul Lind; reply@breakpad-hr.appspotmail.com
Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

On 2015/04/15 13:58:37, gordanac wrote:
> Tests run (beside mipsel): Breakpads's make check for x86-64 Linux and
> Chromium's breakpad_unittests for ARM Android.
> ________________________________________
> From: mailto:rmcilroy@chromium.org [rmcilroy@chromium.org]
> Sent: Wednesday, April 15, 2015 2:31 PM
> To: Gordana Cmiljanovic; fdegans@chromium.org; mark@chromium.org;
> mailto:primiano@chromium.org
> Cc: Petar Jovanovic; Paul Lind;
mailto:reply@breakpad-hr.appspotmail.com
> Subject: Re: [MIPS]: Use mcontext_t structure for MIPS (issue 3744002)

> lgtm, thanks.

> Did you test this doens't break anything on the other architectures?

> https://breakpad.appspot.com/3744002/

Great. Would you like me to land this or are you waiting for another
change to be merged first?

https://breakpad.appspot.com/3744002/

[rmcilroy]

lgtm, thanks.

Mark: do you want a look before I land this?

[Mark Mentovai]

LGTM

[rmcilroy]

On 2015/04/21 13:25:02, Mark Mentovai wrote:
> LGTM

Landed as r1452 [1].  I can't close this CL myself so you will need to do that
manually. Thanks.

[1] https://code.google.com/p/google-breakpad/source/detail?r=1452