don't use glibc-specific execinfo.h on uclibc builds

don't use glibc-specific execinfo.h on uclibc builds

The execinfo.h header is glibc-specific, as is the backtrace function.
Let's skip them for uclibc builds.

BUG=361130
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=262841

[Mostyn Bramley-Moore, 2014-04-08 15:20:41]

@mark, earthdok: here's another attempt at CL 226623008.  The first patchset is
the original patch, the second patchset is my fixup, which removes additional
defined(__GLIBCXX__) checks.

Is there trybot config I can use to test the earlier failure?

[earthdok, 2014-04-08 15:28:30]

lgtm with nit

https://codereview.chromium.org/228753003/diff/2/base/debug/stack_trace_posix.cc
File base/debug/stack_trace_posix.cc (right):

https://codereview.chromium.org/228753003/diff/2/base/debug/stack_trace_posix...
base/debug/stack_trace_posix.cc:751: #endif
#else
count_ = 0;
#endif

[Mostyn Bramley-Moore, 2014-04-08 15:40:46]

https://codereview.chromium.org/228753003/diff/2/base/debug/stack_trace_posix.cc
File base/debug/stack_trace_posix.cc (right):

https://codereview.chromium.org/228753003/diff/2/base/debug/stack_trace_posix...
base/debug/stack_trace_posix.cc:751: #endif
On 2014/04/08 15:28:30, earthdok wrote:
> #else
> count_ = 0;
> #endif

Done.

[earthdok, 2014-04-08 15:47:44]

On 2014/04/08 15:20:41, Mostyn Bramley-Moore wrote:
> Is there trybot config I can use to test the earlier failure?

Sorry, forgot to answer this. There isn't one yet.

[Mark Mentovai, 2014-04-08 15:56:27]

LGTM otherwise

https://codereview.chromium.org/228753003/diff/30001/base/debug/stack_trace_p...
File base/debug/stack_trace_posix.cc (right):

https://codereview.chromium.org/228753003/diff/30001/base/debug/stack_trace_p...
base/debug/stack_trace_posix.cc:9: #if !defined(__UCLIBC__)
Why is this still up here? Can you move it down with the other conditional
#includes?

[Mostyn Bramley-Moore, 2014-04-08 16:16:38]

https://codereview.chromium.org/228753003/diff/30001/base/debug/stack_trace_p...
File base/debug/stack_trace_posix.cc (right):

https://codereview.chromium.org/228753003/diff/30001/base/debug/stack_trace_p...
base/debug/stack_trace_posix.cc:9: #if !defined(__UCLIBC__)
On 2014/04/08 15:56:28, Mark Mentovai wrote:
> Why is this still up here? Can you move it down with the other conditional
> #includes?

I wasn't sure if libc++ builds need this, and if they define __GLIBCXX__ or not.
 I suspect no on both counts.  I'll move it back down.

[Mark Mentovai, 2014-04-08 16:23:32]

https://codereview.chromium.org/228753003/diff/50001/base/debug/stack_trace_p...
File base/debug/stack_trace_posix.cc (right):

https://codereview.chromium.org/228753003/diff/50001/base/debug/stack_trace_p...
base/debug/stack_trace_posix.cc:24: #include <execinfo.h>
Well, you could

#if defined(__GLIBCXX__)
#include <cxxabi.h>
#endif
#if !defined(__UCLIBC__)
#include <execinfo.h>
#endif

We still want to #include <execinfo.h> on Mac, for example, it’s just µclibc
that doesn’t have it. <execinfo.h> is a C standard library header, so it
shouldn’t have anything to do with the __GLIBCXX__ check.

[Mostyn Bramley-Moore, 2014-04-08 17:11:50]

https://codereview.chromium.org/228753003/diff/50001/base/debug/stack_trace_p...
File base/debug/stack_trace_posix.cc (right):

https://codereview.chromium.org/228753003/diff/50001/base/debug/stack_trace_p...
base/debug/stack_trace_posix.cc:24: #include <execinfo.h>
On 2014/04/08 16:23:32, Mark Mentovai wrote:
> Well, you could
> 
> #if defined(__GLIBCXX__)
> #include <cxxabi.h>
> #endif
> #if !defined(__UCLIBC__)
> #include <execinfo.h>
> #endif
> 
> We still want to #include <execinfo.h> on Mac, for example, it’s just µclibc
> that doesn’t have it. <execinfo.h> is a C standard library header, so it
> shouldn’t have anything to do with the __GLIBCXX__ check.

Done.

[Mark Mentovai, 2014-04-08 18:50:33]

LGTM

[Mostyn Bramley-Moore, 2014-04-09 18:49:21]

The CQ bit was checked by mostynb@opera.com

[commit-bot: I haz the power, 2014-04-09 18:49:23]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mostynb@opera.com/228753003/100001

[commit-bot: I haz the power, 2014-04-09 22:29:42]

Change committed as 262841

[bedupontcisco_gmail.com, 2014-06-16 08:41:48]

On 2014/04/09 18:49:21, Mostyn Bramley-Moore wrote:
> The CQ bit was checked by mailto:mostynb@opera.com

I apologize if this isn't the good place to raise my issue (may be it's better
to create a new bug entry or something else).

BTW, there is still a call to the Print() method
(base::debug::StackTrace().Print()) in content/public/test/browser_test_base.cc.
Thus, when you build content_shell with __UCLIBC__ defined, there is a
compilation error.
Shouldn't we stub the Print() method for uClibc build or surrounded all Print()
method calls with #if !defined(__UCLIBC__)?

[Mostyn Bramley-Moore, 2014-06-25 21:28:38]

On 2014/06/16 08:41:48, Benjamin Dupont wrote:
> I apologize if this isn't the good place to raise my issue (may be it's better
> to create a new bug entry or something else).

Sorry I didn't respond earlier.  Here's a possible fix:
https://codereview.chromium.org/354873002/