Chromium Code Reviews
|
|
Chromium Code Reviews
Issue
10416036:
Roll Skia DEPS to 4015, with skia_test_expectations change (Closed)
|
Created:
8 years, 7 months ago by epoger Modified:
8 years, 7 months ago CC:
chromium-reviews, skia-dev_google.com, reed1 Base URL:
svn://svn.chromium.org/chrome/trunk/src/ Visibility:
Public. |
DescriptionRoll Skia DEPS to 4015, with skia_test_expectations change
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=138307
Patch Set 1 #Messages
Total messages: 11 (0 generated)
control group is https://chromiumcodereview.appspot.com/10332292 I will request review once the trybots come back.
The Windows trybots are useless right now, but everything else looks reasonably good. LGTY? linux : red in browser_tests, but test output looks fine linux_asan : green linux_chromeos : green linux_chromeos_asan : one new browser_test failure, doesn't look Skia-related linux_gpu : no new failures linux_heapcheck : deps roll looks about like the control linux_layout : failure, looks like trybot configuration problem You are missing /usr/share/fonts/truetype/msttcorefonts/Arial.ttf. linux_layout_rel : one new webkit_test failure, looks like a non-Skia flake fast/loader/loadInProgress.html = CRASH linux_shared : green mac : green mac_asan : failed compile in both control and deps roll mac_gpu : 4 new gpu_test failures, but they're a set that we see flake often mac_layout : green mac_layout_rel : green win : same compile error in both control and roll 628>E:\b\build\slave\win\build\src\ui\views\examples\content_client\examples_main_delegate.cc(113): error C2259: 'views::examples::ExamplesContentBrowserClient' : cannot instantiate abstract class 628> due to following members: 628> 'bool content::ContentBrowserClient::ShouldSwapProcessesForRedirect(content::ResourceContext *,const GURL &,const GURL &)' : is abstract 628> ..\..\content/public/browser/content_browser_client.h(169) : see declaration of 'content::ContentBrowserClient::ShouldSwapProcessesForRedirect' win_gpu : no new failures win_layout : builder failure in both control and roll MSVS 10 was requested but is not installed. win_layout_rel : builder failure in both control and roll MSVS 10 was requested but is not installed. android : green
LGTM
Is it possible that VectorCanvasTest.PathEffects in test_shell_tests started failing after this change? I don't think the trybots cover this test anymore (I'm moving the tests to other files so the trybots will cover it).
It looks like this roll broke some PathEffect tests: http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Win%20%28deps%2... On Tue, May 22, 2012 at 11:48 AM, <epoger@google.com> wrote: > The Windows trybots are useless right now, but everything else looks > reasonably > good. LGTY? > > linux : red in browser_tests, but test output looks fine > linux_asan : green > linux_chromeos : green > linux_chromeos_asan : one new browser_test failure, doesn't look > Skia-related > linux_gpu : no new failures > linux_heapcheck : deps roll looks about like the control > linux_layout : failure, looks like trybot configuration problem > You are missing /usr/share/fonts/truetype/**msttcorefonts/Arial.ttf. > linux_layout_rel : one new webkit_test failure, looks like a non-Skia flake > fast/loader/loadInProgress.**html = CRASH > linux_shared : green > > mac : green > mac_asan : failed compile in both control and deps roll > mac_gpu : 4 new gpu_test failures, but they're a set that we see flake > often > mac_layout : green > mac_layout_rel : green > > win : same compile error in both control and roll > > 628>E:\b\build\slave\win\**build\src\ui\views\examples\** > content_client\examples_main_**delegate.cc(113): > error C2259: 'views::examples::**ExamplesContentBrowserClient' : cannot > instantiate abstract class > 628> due to following members: > 628> 'bool > content::ContentBrowserClient:**:**ShouldSwapProcessesForRedirect** > (content::ResourceContext > *,const GURL &,const GURL &)' : is abstract > 628> ..\..\content/public/browser/**content_browser_client.h(169) > : see > declaration of 'content::**ContentBrowserClient::** > ShouldSwapProcessesForRedirect**' > > win_gpu : no new failures > win_layout : builder failure in both control and roll > MSVS 10 was requested but is not installed. > win_layout_rel : builder failure in both control and roll > MSVS 10 was requested but is not installed. > > android : green > > > https://chromiumcodereview.**appspot.com/10416036/<https://chromiumcodereview... > > > -- > You received this message because you are subscribed to the Google Groups > "skia-dev" group. > To post to this group, send email to skia-dev@google.com. > To unsubscribe from this group, send email to skia-dev+unsubscribe@google. > **com <skia-dev%2Bunsubscribe@google.com>. > For more options, visit this group at http://groups.google.com/a/** > google.com/group/skia-dev/?hl=**en<http://groups.google.com/a/google.com/group/skia-dev/?hl=en> > . > >
I cannot tell from reading the src how to see the differences. Are the images written to files, so we can inspect them? How do we trigger the tests? On Tue, May 22, 2012 at 6:07 PM, Steve VanDeBogart <vandebo@google.com> wrote: > It looks like this roll broke some PathEffect > tests: http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Win%20%28deps%29/builds/18286 > > On Tue, May 22, 2012 at 11:48 AM, <epoger@google.com> wrote: >> >> The Windows trybots are useless right now, but everything else looks >> reasonably >> good. LGTY? >> >> linux : red in browser_tests, but test output looks fine >> linux_asan : green >> linux_chromeos : green >> linux_chromeos_asan : one new browser_test failure, doesn't look >> Skia-related >> linux_gpu : no new failures >> linux_heapcheck : deps roll looks about like the control >> linux_layout : failure, looks like trybot configuration problem >> You are missing /usr/share/fonts/truetype/msttcorefonts/Arial.ttf. >> linux_layout_rel : one new webkit_test failure, looks like a non-Skia >> flake >> fast/loader/loadInProgress.html = CRASH >> linux_shared : green >> >> mac : green >> mac_asan : failed compile in both control and deps roll >> mac_gpu : 4 new gpu_test failures, but they're a set that we see flake >> often >> mac_layout : green >> mac_layout_rel : green >> >> win : same compile error in both control and roll >> >> >> 628>E:\b\build\slave\win\build\src\ui\views\examples\content_client\examples_main_delegate.cc(113): >> error C2259: 'views::examples::ExamplesContentBrowserClient' : cannot >> instantiate abstract class >> 628> due to following members: >> 628> 'bool >> >> content::ContentBrowserClient::ShouldSwapProcessesForRedirect(content::ResourceContext >> *,const GURL &,const GURL &)' : is abstract >> 628> ..\..\content/public/browser/content_browser_client.h(169) >> : see >> declaration of >> 'content::ContentBrowserClient::ShouldSwapProcessesForRedirect' >> >> win_gpu : no new failures >> win_layout : builder failure in both control and roll >> MSVS 10 was requested but is not installed. >> win_layout_rel : builder failure in both control and roll >> MSVS 10 was requested but is not installed. >> >> android : green >> >> >> https://chromiumcodereview.appspot.com/10416036/ >> >> >> -- >> You received this message because you are subscribed to the Google Groups >> "skia-dev" group. >> To post to this group, send email to skia-dev@google.com. >> To unsubscribe from this group, send email to >> skia-dev+unsubscribe@google.com. >> For more options, visit this group at >> http://groups.google.com/a/google.com/group/skia-dev/?hl=en. >> > > -- > You received this message because you are subscribed to the Google Groups > "skia-dev" group. > To post to this group, send email to skia-dev@google.com. > To unsubscribe from this group, send email to > skia-dev+unsubscribe@google.com. > For more options, visit this group at > http://groups.google.com/a/google.com/group/skia-dev/?hl=en.
steveblock filed http://code.google.com/p/chromium/issues/detail?id=129331 and disabled the test... I will follow up on that bug. It really cheeses me that our trybots do not cover the same set of tests that people care about in the waterfall bots. There is so much flaky redness in the waterfall bots, and other people's CLs come in so quickly, that it is really hard to tell when a mysterious red is a result of my CL.
On 2012/05/22 21:17:57, tony wrote: > Is it possible that VectorCanvasTest.PathEffects in test_shell_tests started > failing after this change? I don't think the trybots cover this test anymore > (I'm moving the tests to other files so the trybots will cover it). Tony- I think the "win" trybots *do* cover this step. Here's a recent one, showing the failure: http://build.chromium.org/p/tryserver.chromium/builders/win/builds/12816 The problem in my case was that the "win" trybot failed to compile in both control and deps-roll CLs, so I ignored it. I guess the lessons are: 1. Don't ignore trybot compile failures, because they can mask regressions later in the trybot process. Instead, re-run trybots that fail to compile, patching against a later revision. 2. People should validate that their code compiles on all trybots before submitting it, so that other developers are not blocked by such failures.
You can run the tests by building the tests_shell_tests target and running it with --gtest-filter=VectorCanvasTest.PathEffects. It looks like the failing result is not written to disk, but you could add a line in ProcessImage to store the output. -- Steve On Tue, May 22, 2012 at 7:19 PM, Mike Reed <reed@google.com> wrote: > I cannot tell from reading the src how to see the differences. Are the > images written to files, so we can inspect them? How do we trigger the > tests? > > On Tue, May 22, 2012 at 6:07 PM, Steve VanDeBogart <vandebo@google.com> > wrote: > > It looks like this roll broke some PathEffect > > tests: > http://build.chromium.org/p/chromium.webkit/builders/Webkit%20Win%20%28deps%2... > > > > On Tue, May 22, 2012 at 11:48 AM, <epoger@google.com> wrote: > >> > >> The Windows trybots are useless right now, but everything else looks > >> reasonably > >> good. LGTY? > >> > >> linux : red in browser_tests, but test output looks fine > >> linux_asan : green > >> linux_chromeos : green > >> linux_chromeos_asan : one new browser_test failure, doesn't look > >> Skia-related > >> linux_gpu : no new failures > >> linux_heapcheck : deps roll looks about like the control > >> linux_layout : failure, looks like trybot configuration problem > >> You are missing /usr/share/fonts/truetype/msttcorefonts/Arial.ttf. > >> linux_layout_rel : one new webkit_test failure, looks like a non-Skia > >> flake > >> fast/loader/loadInProgress.html = CRASH > >> linux_shared : green > >> > >> mac : green > >> mac_asan : failed compile in both control and deps roll > >> mac_gpu : 4 new gpu_test failures, but they're a set that we see flake > >> often > >> mac_layout : green > >> mac_layout_rel : green > >> > >> win : same compile error in both control and roll > >> > >> > >> > 628>E:\b\build\slave\win\build\src\ui\views\examples\content_client\examples_main_delegate.cc(113): > >> error C2259: 'views::examples::ExamplesContentBrowserClient' : cannot > >> instantiate abstract class > >> 628> due to following members: > >> 628> 'bool > >> > >> > content::ContentBrowserClient::ShouldSwapProcessesForRedirect(content::ResourceContext > >> *,const GURL &,const GURL &)' : is abstract > >> 628> > ..\..\content/public/browser/content_browser_client.h(169) > >> : see > >> declaration of > >> 'content::ContentBrowserClient::ShouldSwapProcessesForRedirect' > >> > >> win_gpu : no new failures > >> win_layout : builder failure in both control and roll > >> MSVS 10 was requested but is not installed. > >> win_layout_rel : builder failure in both control and roll > >> MSVS 10 was requested but is not installed. > >> > >> android : green > >> > >> > >> https://chromiumcodereview.appspot.com/10416036/ > >> > >> > >> -- > >> You received this message because you are subscribed to the Google > Groups > >> "skia-dev" group. > >> To post to this group, send email to skia-dev@google.com. > >> To unsubscribe from this group, send email to > >> skia-dev+unsubscribe@google.com. > >> For more options, visit this group at > >> http://groups.google.com/a/google.com/group/skia-dev/?hl=en. > >> > > > > -- > > You received this message because you are subscribed to the Google Groups > > "skia-dev" group. > > To post to this group, send email to skia-dev@google.com. > > To unsubscribe from this group, send email to > > skia-dev+unsubscribe@google.com. > > For more options, visit this group at > > http://groups.google.com/a/google.com/group/skia-dev/?hl=en. >
On 2012/05/23 19:48:02, vandebo1 wrote: > You can run the tests by building the tests_shell_tests target and running > it with --gtest-filter=VectorCanvasTest.PathEffects. It looks like the > failing result is not written to disk, but you could add a line > in ProcessImage to store the output. Thanks, Steve. How do you build that target? Do you open a particular solution file in Visual Studio, find that target, and build it? Or is there a command-line approach? > > -- > Steve > > On Tue, May 22, 2012 at 7:19 PM, Mike Reed <mailto:reed@google.com> wrote: > > > I cannot tell from reading the src how to see the differences. Are the > > images written to files, so we can inspect them? How do we trigger the > > tests? > > > > On Tue, May 22, 2012 at 6:07 PM, Steve VanDeBogart <mailto:vandebo@google.com> > > wrote: > > > It looks like this roll broke some PathEffect > > > tests: > > > http://build.chromium.org/p/chromium.webkit/builders/Webkit%2520Win%2520%2528... > > > > > > On Tue, May 22, 2012 at 11:48 AM, <mailto:epoger@google.com> wrote: > > >> > > >> The Windows trybots are useless right now, but everything else looks > > >> reasonably > > >> good. LGTY? > > >> > > >> linux : red in browser_tests, but test output looks fine > > >> linux_asan : green > > >> linux_chromeos : green > > >> linux_chromeos_asan : one new browser_test failure, doesn't look > > >> Skia-related > > >> linux_gpu : no new failures > > >> linux_heapcheck : deps roll looks about like the control > > >> linux_layout : failure, looks like trybot configuration problem > > >> You are missing /usr/share/fonts/truetype/msttcorefonts/Arial.ttf. > > >> linux_layout_rel : one new webkit_test failure, looks like a non-Skia > > >> flake > > >> fast/loader/loadInProgress.html = CRASH > > >> linux_shared : green > > >> > > >> mac : green > > >> mac_asan : failed compile in both control and deps roll > > >> mac_gpu : 4 new gpu_test failures, but they're a set that we see flake > > >> often > > >> mac_layout : green > > >> mac_layout_rel : green > > >> > > >> win : same compile error in both control and roll > > >> > > >> > > >> > > > 628>E:\b\build\slave\win\build\src\ui\views\examples\content_client\examples_main_delegate.cc(113): > > >> error C2259: 'views::examples::ExamplesContentBrowserClient' : cannot > > >> instantiate abstract class > > >> 628> due to following members: > > >> 628> 'bool > > >> > > >> > > > content::ContentBrowserClient::ShouldSwapProcessesForRedirect(content::ResourceContext > > >> *,const GURL &,const GURL &)' : is abstract > > >> 628> > > ..\..\content/public/browser/content_browser_client.h(169) > > >> : see > > >> declaration of > > >> 'content::ContentBrowserClient::ShouldSwapProcessesForRedirect' > > >> > > >> win_gpu : no new failures > > >> win_layout : builder failure in both control and roll > > >> MSVS 10 was requested but is not installed. > > >> win_layout_rel : builder failure in both control and roll > > >> MSVS 10 was requested but is not installed. > > >> > > >> android : green > > >> > > >> > > >> https://chromiumcodereview.appspot.com/10416036/ > > >> > > >> > > >> -- > > >> You received this message because you are subscribed to the Google > > Groups > > >> "skia-dev" group. > > >> To post to this group, send email to mailto:skia-dev@google.com. > > >> To unsubscribe from this group, send email to > > >> mailto:skia-dev+unsubscribe@google.com. > > >> For more options, visit this group at > > >> http://groups.google.com/a/google.com/group/skia-dev/?hl=en. > > >> > > > > > > -- > > > You received this message because you are subscribed to the Google Groups > > > "skia-dev" group. > > > To post to this group, send email to mailto:skia-dev@google.com. > > > To unsubscribe from this group, send email to > > > mailto:skia-dev+unsubscribe@google.com. > > > For more options, visit this group at > > > http://groups.google.com/a/google.com/group/skia-dev/?hl=en. > >
On Wed, May 23, 2012 at 12:54 PM, <epoger@google.com> wrote: > On 2012/05/23 19:48:02, vandebo1 wrote: > >> You can run the tests by building the tests_shell_tests target and running >> it with --gtest-filter=**VectorCanvasTest.PathEffects. It looks like the >> failing result is not written to disk, but you could add a line >> in ProcessImage to store the output. >> > > Thanks, Steve. How do you build that target? Do you open a particular > solution > file in Visual Studio, find that target, and build it? Or is there a > command-line approach? > The project is in src/webkit/test_shell_tests.vcproj. You should be able to open in VS, or use the command line build utils as described for the Chrome building on window page. > > > -- >> Steve >> > > On Tue, May 22, 2012 at 7:19 PM, Mike Reed <mailto:reed@google.com> >> wrote: >> > > > I cannot tell from reading the src how to see the differences. Are the >> > images written to files, so we can inspect them? How do we trigger the >> > tests? >> > >> > On Tue, May 22, 2012 at 6:07 PM, Steve VanDeBogart >> > <mailto:vandebo@google.com> > > > wrote: >> > > It looks like this roll broke some PathEffect >> > > tests: >> > >> > > http://build.chromium.org/p/**chromium.webkit/builders/** > Webkit%2520Win%2520%2528deps%**2529/builds/18286<http://build.chromium.org/p/chromium.webkit/builders/Webkit%2520Win%2520%2528deps%2529/builds/18286> > > > > >> > > On Tue, May 22, 2012 at 11:48 AM, <mailto:epoger@google.com> wrote: >> > >> >> > >> The Windows trybots are useless right now, but everything else looks >> > >> reasonably >> > >> good. LGTY? >> > >> >> > >> linux : red in browser_tests, but test output looks fine >> > >> linux_asan : green >> > >> linux_chromeos : green >> > >> linux_chromeos_asan : one new browser_test failure, doesn't look >> > >> Skia-related >> > >> linux_gpu : no new failures >> > >> linux_heapcheck : deps roll looks about like the control >> > >> linux_layout : failure, looks like trybot configuration problem >> > >> You are missing /usr/share/fonts/truetype/** >> msttcorefonts/Arial.ttf. >> > >> linux_layout_rel : one new webkit_test failure, looks like a non-Skia >> > >> flake >> > >> fast/loader/loadInProgress.**html = CRASH >> > >> linux_shared : green >> > >> >> > >> mac : green >> > >> mac_asan : failed compile in both control and deps roll >> > >> mac_gpu : 4 new gpu_test failures, but they're a set that we see >> flake >> > >> often >> > >> mac_layout : green >> > >> mac_layout_rel : green >> > >> >> > >> win : same compile error in both control and roll >> > >> >> > >> >> > >> >> > >> > > 628>E:\b\build\slave\win\**build\src\ui\views\examples\** > content_client\examples_main_**delegate.cc(113): > >> > >> error C2259: 'views::examples::**ExamplesContentBrowserClient' : >> cannot >> > >> instantiate abstract class >> > >> 628> due to following members: >> > >> 628> 'bool >> > >> >> > >> >> > >> > > content::ContentBrowserClient:**:**ShouldSwapProcessesForRedirect** > (content::ResourceContext > >> > >> *,const GURL &,const GURL &)' : is abstract >> > >> 628> >> > ..\..\content/public/browser/**content_browser_client.h(169) >> > >> : see >> > >> declaration of >> > >> 'content::**ContentBrowserClient::**ShouldSwapProcessesForRedirect** >> ' >> > >> >> > >> win_gpu : no new failures >> > >> win_layout : builder failure in both control and roll >> > >> MSVS 10 was requested but is not installed. >> > >> win_layout_rel : builder failure in both control and roll >> > >> MSVS 10 was requested but is not installed. >> > >> >> > >> android : green >> > >> >> > >> >> > >> https://chromiumcodereview.**appspot.com/10416036/<https://chromiumcodereview... >> > >> >> > >> >> > >> -- >> > >> You received this message because you are subscribed to the Google >> > Groups >> > >> "skia-dev" group. >> > >> To post to this group, send email to mailto:skia-dev@google.com. >> >> > >> To unsubscribe from this group, send email to >> > >> mailto:skia-dev+unsubscribe@**google.com<skia-dev%2Bunsubscribe@google.com> >> . >> >> > >> For more options, visit this group at >> > >> http://groups.google.com/a/**google.com/group/skia-dev/?hl=**en<http://groups... >> . >> > >> >> > > >> > > -- >> > > You received this message because you are subscribed to the Google >> Groups >> > > "skia-dev" group. >> > > To post to this group, send email to mailto:skia-dev@google.com. >> >> > > To unsubscribe from this group, send email to >> > > mailto:skia-dev+unsubscribe@**google.com<skia-dev%2Bunsubscribe@google.com> >> . >> >> > > For more options, visit this group at >> > > http://groups.google.com/a/**google.com/group/skia-dev/?hl=**en<http://groups... >> . >> > >> > > > > https://chromiumcodereview.**appspot.com/10416036/<https://chromiumcodereview... > |
