Properly handle microdump processing, when the system_log file contains an incomplete microdump sec…

Handle multiple microdumps in system log.

Properly handle microdump processing, when the system_log file contains an incomplete microdump section at the top. The processor will process the first complete microdump section.

R=primiano@chromium.org

Committed: https://chromium.googlesource.com/breakpad/breakpad/+/8f7e09cf6dda7a2d346258078334003d83533222

[mmandlis, 2016-02-26 21:54:24]

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

[mmandlis, 2016-02-26 21:54:24]

Yet another dejavu review. I hope this is not too annoying. I just want to
release the fix sooner than we do breakpad import.

[Primiano Tucci (use gerrit), 2016-02-26 22:14:28]

No prob, I'm going to review in some order anyways.
Maybe just fix the commit message to fit the chromium style:

Short description
<blank linke>
Longer description and BUG= link if you have any

[Primiano Tucci (use gerrit), 2016-02-26 22:19:27]

ehm forgot... LGTM :)

[vapier, 2016-02-26 22:41:40]

vapier@chromium.org changed reviewers:
+ vapier@chromium.org

[vapier, 2016-02-26 22:41:41]

that dump file is ~270KB.  does it really need to be that big ?  can you shrink
it down to like <50KB while maintaining its usefulness ?

[mmandlis, 2016-02-26 22:44:27]

On 2016/02/26 22:41:41, vapier wrote:
> that dump file is ~270KB.  does it really need to be that big ?  can you
shrink
> it down to like <50KB while maintaining its usefulness ?

This is a dump from a real report whose processing fails now. Are you concerned
about the size of the test file, or the dump itself?

[vapier, 2016-02-26 22:48:00]

purely its size.  any file that goes into git stays in its history forever and
is pulled down in every `git clone`.  for the sake of unittests, the majority of
the content is rarely useful and is entirely overhead.

[mmandlis, 2016-02-26 23:01:39]

Description was changed from

==========
Properly handle microdump processing, when the system_log file contains an
incomplete microdump section at the top. The processor will process the first
complete microdump section.

BUG=
==========

to

==========
Handle multiple microdumps in system log.

Properly handle microdump processing, when the system_log file contains an
incomplete microdump section at the top. The processor will process the first
complete microdump section.
==========

[mmandlis, 2016-02-26 23:01:45]

On 2016/02/26 22:48:00, vapier wrote:
> purely its size.  any file that goes into git stays in its history forever and
> is pulled down in every `git clone`.  for the sake of unittests, the majority
of
> the content is rarely useful and is entirely overhead.

[mmandlis, 2016-02-26 23:02:35]

Done, removed the unneeded contents and reduced file size.

Primiano, I don't have a corresponding bug.

Submitting shortly.

Thanks!

[mmandlis, 2016-02-27 00:40:46]

Description was changed from

==========
Handle multiple microdumps in system log.

Properly handle microdump processing, when the system_log file contains an
incomplete microdump section at the top. The processor will process the first
complete microdump section.
==========

to

==========
Handle multiple microdumps in system log.

Properly handle microdump processing, when the system_log file contains an
incomplete microdump section at the top. The processor will process the first
complete microdump section.

R=primiano@chromium.org

Committed:
https://chromium.googlesource.com/breakpad/breakpad/+/8f7e09cf6dda7a2d3462580...
==========

[mmandlis, 2016-02-27 00:40:47]

Committed patchset #2 (id:20001) manually as
8f7e09cf6dda7a2d346258078334003d83533222 (presubmit successful).

[Primiano Tucci (use gerrit), 2016-02-27 03:14:41]

On 2016/02/27 00:40:47, mmandlis wrote:
> Committed patchset #2 (id:20001) manually as
> 8f7e09cf6dda7a2d346258078334003d83533222 (presubmit successful).

P.S: can we keep in sync the test change downstream vs upstream? Thanks :)

[vapier, 2016-02-27 07:33:44]

what do you mean by downstream? this is upstream.

[Primiano Tucci (use gerrit), 2016-02-27 08:33:37]

Was referring to mmandlis@, should have been more explicit sorry.

On 2016/02/27 07:33:44, vapier wrote:
> what do you mean by downstream? this is upstream.
I know, in fact I said "can we keep un sync the test change downstream" ? 
But there is a downstream part of this :)  See internal codereview 115703046,
but now the test diverged and we should converge it.

[vapier, 2016-02-27 21:05:51]

keeping downstream repos (like google3) in sync w/upstream is up to the owners
of the respective downstreams.  so you'll want to talk to the owners of that
repo.

Chromium pins the git repo directly, and Chromium OS/Android mirror the git repo
directly now.

[Primiano Tucci (use gerrit), 2016-02-28 04:36:47]

On 2016/02/27 21:05:51, vapier wrote:
> keeping downstream repos (like google3) in sync w/upstream is up to the owners
> of the respective downstreams.  so you'll want to talk to the owners of that
> repo.
Yes I know. That would be mmandlis@. That's why I posted here my original
message. It was a reminder for her to backport the change downstream.

[vapier, 2016-02-28 05:22:17]

gotcha.  sorry for the noise.

[mmandlis, 2016-02-29 17:52:53]

On 2016/02/28 05:22:17, vapier wrote:
> gotcha.  sorry for the noise.

Primiano, our periodically scheduled breakpad import into google3 is scheduled
for this week, so it will handle it.

Thanks!