Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(31)

Issue 149503002: Enable stack tool and ASan symbolization of logcat on Android bots. (Closed)

Created:
6 years, 10 months ago by eugenis
Modified:
6 years, 10 months ago
Reviewers:
craigdh, navabi1, navabi, frankf
CC:
chromium-reviews, craigdh+watch_chromium.org, bulach+watch_chromium.org, yfriedman+watch_chromium.org, ilevy-cc_chromium.org, klundberg+watch_chromium.org, frankf+watch_chromium.org
Visibility:
Public.

Description

Enable stack tool and ASan symbolization of logcat on Android bots. BUG=325685 R=frankf@chromium.org, craigdh@chromium.org, navabi@chromium.org Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=248694

Patch Set 1 #

Unified diffs Side-by-side diffs Delta from patch set Stats (+24 lines, -2 lines) Patch
M build/android/buildbot/bb_device_steps.py View 3 chunks +23 lines, -1 line 0 comments Download
M build/android/buildbot/bb_run_bot.py View 1 chunk +1 line, -1 line 0 comments Download

Messages

Total messages: 10 (0 generated)
eugenis
6 years, 10 months ago (2014-01-29 12:25:58 UTC) #1
frankf
lgtm but please wait for navabi. Is downstream script going to reuse this?
6 years, 10 months ago (2014-02-03 21:01:38 UTC) #2
navabi
On 2014/02/03 21:01:38, frankf wrote: > lgtm but please wait for navabi. > > Is ...
6 years, 10 months ago (2014-02-03 22:13:16 UTC) #3
eugenis
On 2014/02/03 22:13:16, navabi wrote: > On 2014/02/03 21:01:38, frankf wrote: > > lgtm but ...
6 years, 10 months ago (2014-02-04 07:56:19 UTC) #4
eugenis
Committed patchset #1 manually as r248694 (presubmit successful).
6 years, 10 months ago (2014-02-04 08:04:47 UTC) #5
navabi
On 2014/02/04 08:04:47, eugenis wrote: > Committed patchset #1 manually as r248694 (presubmit successful). Thanks ...
6 years, 10 months ago (2014-02-04 17:39:37 UTC) #6
navabi
A revert of this CL has been created in https://codereview.chromium.org/154403002/ by navabi@google.com. The reason for ...
6 years, 10 months ago (2014-02-04 21:16:11 UTC) #7
eugenis
On 2014/02/04 21:16:11, navabi wrote: > A revert of this CL has been created in ...
6 years, 10 months ago (2014-02-05 08:31:26 UTC) #8
pasko
On 2014/02/05 08:31:26, eugenis wrote: > On 2014/02/04 21:16:11, navabi wrote: > > A revert ...
6 years, 10 months ago (2014-02-05 12:11:20 UTC) #9
eugenis
6 years, 10 months ago (2014-02-05 12:14:33 UTC) #10
Message was sent while issue was closed.
On 2014/02/05 12:11:20, pasko wrote:
> On 2014/02/05 08:31:26, eugenis wrote:
> > On 2014/02/04 21:16:11, navabi wrote:
> > > A revert of this CL has been created in
> > > https://codereview.chromium.org/154403002/ by mailto:navabi@google.com.
> > > 
> > > The reason for reverting is: Sorry, we can not commit this without the
> > > accompanying downstream change, because without that there are two
> > > --disable-stack-tool option strings (one in the downstream script and the
> > newly
> > > added one here).
> > > 
> > > This causes the following error on the clank tot bot:
> > > optparse.OptionConflictError: option --disable-stack-tool: conflicting
> option
> > > string(s): --disable-stack-tool
> > > 
> > > Link:
> > >
> >
>
https://chromegw.corp.google.com/i/clank.tot/builders/instrumentation-yakju-c...
> > > 
> > > Reverting because this is keeping clank from picking up changes, and I'm
not
> > > sure how to do a quick fix right now. Removing the option from clank
> requires
> > > the public script to have it, which requires clank to roll. Slightly
tricky
> > and
> > > I don't have the time to do it today..
> > 
> > Ouch. Are you saying that the automerger won't pick up a change that breaks
> > clank bots?
> 
> yes, automerger bot referenced above rolls only if some tests pass,
compilation
> is a necessary step :)
> 
> > Sounds like we must disable symbolization downstream first, land this
change,
> > and re-enable symbolization later.
> 
> I don't quite understand the goal. The breakage is coming from the downstream
> code picking up the flag "--disable-stack-tool" which is introduced in this
> script upstream. Is the ultimate goal to move the symbolization flag upstream
or
> have it both upstream and downstream somehow? In the former case we'd need to
> remove the flag from downstream first (and disable the code that depends on
it),
> then autoroller picks up the change, then re-enable.
> 
> I am looking at another ToT breakage, autoroller may be slow to pick things up
> today.

The goal is to have this guy:
http://build.chromium.org/p/chromium.fyi/builders/Android%20Asan%20Builder%20...
symbolize reports in logcat dump.
So, yes, both upstream and downstream.

Powered by Google App Engine
This is Rietveld 408576698