Roll breakpad r1454:r1456 and add product info in microdumps

Roll breakpad r1454:r1456 and add product info in microdumps

The change complements Breakpad's crrev.com/1125153008 and introduces
the ability of attaching production information and build id to
microdumps.

The DEPS roll pulls in the following change:
r1456 d7e4030  [microdump] Add build fingerprint and product info metadata.

Summary of all changes available at:
https://chromium.googlesource.com/external/google-breakpad/src/+log/9fcbe25..d7e4030

BUG=456494, 410294
Committed: https://crrev.com/b2da245aa3a437766d78a2c8eb26b9f819809efd
Cr-Commit-Position: refs/heads/master@{#330339}

[Primiano Tucci (use gerrit), 2015-05-14 17:39:33]

primiano@chromium.org changed reviewers:
+ thestig@chromium.org

[Primiano Tucci (use gerrit), 2015-05-14 17:40:28]

Lei, this is matching crrev.com/1125153008
I will obviously rename the TODO in the commit message and do the roll once the
breakpad side lands.

Thanks!

[Lei Zhang, 2015-05-14 19:14:27]

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
File components/crash/app/breakpad_linux.cc (right):

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:92: std::string
g_microdump_build_fingerprint;
Mark these Android only?

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:92: std::string
g_microdump_build_fingerprint;
Won't this and |g_microdump_product_info| add static initializers, and thus fail
the "size" step on the buildbot?

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:93: const char*
g_c_microdump_build_fingerprint = NULL;
Let's start using nullptr in new code.

[Primiano Tucci (use gerrit), 2015-05-15 10:20:28]

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
File components/crash/app/breakpad_linux.cc (right):

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:92: std::string
g_microdump_build_fingerprint;
On 2015/05/14 19:14:26, Lei Zhang wrote:
> Mark these Android only?
Facepalm, right. switched to lazy_instance with no static initializers (linker
initialized in the bss).

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:92: std::string
g_microdump_build_fingerprint;
On 2015/05/14 19:14:26, Lei Zhang wrote:
> Mark these Android only?

Done.

https://codereview.chromium.org/1019663003/diff/20001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:93: const char*
g_c_microdump_build_fingerprint = NULL;
On 2015/05/14 19:14:27, Lei Zhang wrote:
> Let's start using nullptr in new code.
If you are fine with that, I did mass s/NULL/nullptr/ in this file. Felt a bit
weird to me to have a mixture.
In the contrary case, I'll just delete PatchSet 4 (which has just the s///).

[Lei Zhang, 2015-05-15 10:37:18]

lgtm

https://codereview.chromium.org/1019663003/diff/60001/components/crash/app/br...
File components/crash/app/breakpad_linux.cc (right):

https://codereview.chromium.org/1019663003/diff/60001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:752: g_microdump_build_fingerprint.Get()
=
or maybe get rid of the LazyInstance, and just do:
g_c_microdump_build_fingerprint = strdup(foo_c_string) ?

[Primiano Tucci (use gerrit), 2015-05-15 11:41:07]

Patchset #5 (id:80001) has been deleted

[Primiano Tucci (use gerrit), 2015-05-15 11:41:44]

https://codereview.chromium.org/1019663003/diff/60001/components/crash/app/br...
File components/crash/app/breakpad_linux.cc (right):

https://codereview.chromium.org/1019663003/diff/60001/components/crash/app/br...
components/crash/app/breakpad_linux.cc:752: g_microdump_build_fingerprint.Get()
=
On 2015/05/15 10:37:18, Lei Zhang wrote:
> or maybe get rid of the LazyInstance, and just do:
> g_c_microdump_build_fingerprint = strdup(foo_c_string) ?
Yeah, I guess it's easier to read if I leak strings with annotations, rather
than using LazyInstance (we don't actually need their thread safety here)
How about now?

[Lei Zhang, 2015-05-15 19:31:15]

On 2015/05/15 11:41:44, Primiano Tucci wrote:
> Yeah, I guess it's easier to read if I leak strings with annotations, rather
> than using LazyInstance (we don't actually need their thread safety here)
> How about now?

Sounds good.

[Lei Zhang, 2015-05-15 19:34:47]

On 2015/05/15 19:31:15, Lei Zhang wrote:
> On 2015/05/15 11:41:44, Primiano Tucci wrote:
> > Yeah, I guess it's easier to read if I leak strings with annotations, rather
> > than using LazyInstance (we don't actually need their thread safety here)
> > How about now?
> 
> Sounds good.

But some Android bots are unhappy and crashing in
InitMicrodumpCrashHandlerIfNecessary(). Did we strdup(nullptr) or something?

[Primiano Tucci (use gerrit), 2015-05-18 08:18:04]

> But some Android bots are unhappy and crashing in
> InitMicrodumpCrashHandlerIfNecessary(). Did we strdup(nullptr) or something?
Ah, of course, WebView is still returning NULL because the CL which will do that
depends on this one (crrev.com/1131083003).
Which in practice reminds me that, strdup or not, we should check for null
string before setting them.
Done!

(Also filed crbug.com/489160 as the stack trace produced by the bots tombstones
is layout horizontally and unreadable for humans ^__^)

[Primiano Tucci (use gerrit), 2015-05-18 09:29:59]

The CQ bit was checked by primiano@chromium.org

[Primiano Tucci (use gerrit), 2015-05-18 09:30:01]

The patchset sent to the CQ was uploaded after l-g-t-m from thestig@chromium.org
Link to the patchset: https://codereview.chromium.org/1019663003/#ps120001
(title: "null string check")

[commit-bot: I haz the power, 2015-05-18 09:30:08]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1019663003/120001

[commit-bot: I haz the power, 2015-05-18 09:34:01]

Committed patchset #6 (id:120001)

[commit-bot: I haz the power, 2015-05-18 11:34:08]

Patchset 6 (id:??) landed as
https://crrev.com/b2da245aa3a437766d78a2c8eb26b9f819809efd
Cr-Commit-Position: refs/heads/master@{#330339}

[Lei Zhang, 2015-05-18 20:26:11]

lgtm++