BUG=10737,10737,16129,16153,16156,16161,16324,16326
TEST=watch linux ui valgrind test bot go greener
Review URL: http://codereview.chromium.org/155196
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20262 0039d316-1c4b-4281-b951-d872f2087c98
The problem is scoped_ptr<ResourceRequestDetails> is used to delete an instance of ResourceRedirectDetails, however the base class's destructor is non-virtual, so ResourceRedirectDetails extra field (GURL new_url_) does not get torn down properly.
Note that the blame callstacks in the bug report appear unrelated, because std::string is doing some fancy refcounting under the hood. So as GURLs get passed around, they take over the GURL::spec's memory that got allocated earlier by the network stack.
BUG=http://crbug.com/10873
TEST=The "RedirectTest.*" ui tests when run on linux valgrind bots should show no leaks
Review URL: http://codereview.chromium.org/155204
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20179 0039d316-1c4b-4281-b951-d872f2087c98
BUG=none
TEST=Fewer leaks on the Mac valgrind UI test bots
Review URL: http://codereview.chromium.org/155190
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20156 0039d316-1c4b-4281-b951-d872f2087c98
from a web page, and always run it through valgrind once no matter
how perfect it is visually.
Review URL: http://codereview.chromium.org/149271
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20065 0039d316-1c4b-4281-b951-d872f2087c98
BUG=16096, 10737, 10738
TEST=watch the linux ui buildbots for valgrind warnings; this should make several go away
Review URL: http://codereview.chromium.org/155152
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20050 0039d316-1c4b-4281-b951-d872f2087c98
so fork() doesn't corrupt log files; lets --generate_suppressions
work even with the hacky valgrind fix that makes anything but --log-file
crash on child of fork().
Also avoid killing user desktop session if zygote dies!
BUG=none, but related to fix for http://crbug.com/15771
TEST=patch valgrind with fork workaround; sh tools/valgrind/chrome_tests.sh --generate_suppressions -t ui logs you out on linux without this.
Review URL: http://codereview.chromium.org/155130
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@20024 0039d316-1c4b-4281-b951-d872f2087c98
Times out too often under Valgrind on Mac
Tracking bug: crbug.com/15817
Review URL: http://codereview.chromium.org/150218
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19846 0039d316-1c4b-4281-b951-d872f2087c98
enabled the ImageOperations test.
This should be fixed eventually, but suppressing it now
to get the build green since it's not a regression.
TEST=none
BUG=Issue tracker is down
Review URL: http://codereview.chromium.org/150172
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19748 0039d316-1c4b-4281-b951-d872f2087c98
These common files are used by page cyclers to provide a
standard JS test harness and report base.
Add acid3 page cycler to chromium checkout.
Review URL: http://codereview.chromium.org/151053
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19564 0039d316-1c4b-4281-b951-d872f2087c98
BUG=none
TEST="make -r base_unittests; sh tools/valgrind/chrome-tests.sh -t base_unittests" on 64 bit ubuntu and verify no valgrind warnings from ld-*.so
Review URL: http://codereview.chromium.org/147254
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19497 0039d316-1c4b-4281-b951-d872f2087c98
BUG=http://crbug.com/15454
TEST=export GYP_GENERATORS=make; rm -rf sconsbuild; gclient runhooks --force; make -r base_unittests layout_tests; sh tools/valgrind/chrome_tests.sh -t base_unittests; verify that it finds and runs base_unittests; same thing with -t layout_tests
Review URL: http://codereview.chromium.org/147209
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19453 0039d316-1c4b-4281-b951-d872f2087c98
Already filed with Apple.
BUG=none
TEST=fewer leaks reported from the UI tests under valgrind on Mac
Review URL: http://codereview.chromium.org/146115
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@19195 0039d316-1c4b-4281-b951-d872f2087c98
BUG=none
TEST=Most/all of the UI tests should pass on the Mac when run under valgrind.
Review URL: http://codereview.chromium.org/140008
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@18849 0039d316-1c4b-4281-b951-d872f2087c98
try:-finally: so builds will work without forced upgrades to Python 2.5.
BUG=12818
TEST=successful build
Review URL: http://codereview.chromium.org/140005
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@18848 0039d316-1c4b-4281-b951-d872f2087c98
BUG=none
TEST=Run ui_tests under valgrind; see fewer warning that we can't fix.
Review URL: http://codereview.chromium.org/131081
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@18824 0039d316-1c4b-4281-b951-d872f2087c98
This will be used by the buildbots after Xvfb is executated to make sure that Xvfb is alive.
Review URL: http://codereview.chromium.org/126165
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@18549 0039d316-1c4b-4281-b951-d872f2087c98
the magic value in buildbot changed.
TEST=Buildbot will turn orange if there is a warning
BUG=None
Review URL: http://codereview.chromium.org/126048
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@18270 0039d316-1c4b-4281-b951-d872f2087c98
SVG UI:
- Fix some TODOs about clipping SVG objects to the scene. This fixes bugs
where a rectangle is too large and not displayed at all.
- Fix the calculation of the thread endms done time. This fixes the thread
gray background not being drawn for the full length of the thread.
Traceline tracer:
- Enable PatchCreateThread on XP. This fixes not having the thread creation
information output in the JSON files on XP. Still doesn't work on Vista.
- Print the manual-quit message to stderr, so it doesn't end up in the JSON.
- Cast a symbol API callback to compile with both older and newer SDKs.
- Add two small scripts for breaking apart large JSON files into smaller
pieces, so that the Python tools can handle processing the smaller files.
Review URL: http://codereview.chromium.org/118377
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@17859 0039d316-1c4b-4281-b951-d872f2087c98
Revert DCHECKs to what they used to be - turns out a bunch of stuff in the code depended on input.empty() == true.
---
I believe the code (with your width() * bbp fix) is correct.
The UMR errors occur when the source image contain alpha. I believe the issue
comes from webkit glue image decoder, and the reason this only started
triggering valgrind errors is that the old ones were masked by the change
detailed in bug 12640
Also adds valgrind suppression for the new code.
TBR=Nick
BUG=12891,12640
TEST=none
Review URL: http://codereview.chromium.org/119271
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@17811 0039d316-1c4b-4281-b951-d872f2087c98
The UMR errors occur when the source image contain alpha. I believe the issue comes from webkit glue image decoder, and the reason this only started triggering valgrind errors is that the old ones were masked by the change detailed in bug 12640
Also adds valgrind suppression for the new code.
BUG=12891,12640
TEST=none
Review URL: http://codereview.chromium.org/118297
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@17792 0039d316-1c4b-4281-b951-d872f2087c98
Minor fixes to coverage related scripts.
Review URL: http://codereview.chromium.org/118298
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@17785 0039d316-1c4b-4281-b951-d872f2087c98
Re-enable zip unit tests on mac.
Review URL: http://codereview.chromium.org/118087
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@17369 0039d316-1c4b-4281-b951-d872f2087c98
nirnimesh@chromium.org