79333 2012-02-22 23:02:13 -0800 REGRESSION: Assertion failure in Document::setCompatibilityMode() (!m_styleSheets->length()) in fast/dynamic/crash-paint-no-documentElement-renderer.html 2012-02-24 10:25:47 -0800 1 1 1 Unclassified WebKit Tools / Tests 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED InRadar, LayoutTestFailure, Regression P1 Normal --- 1 mitz webkit-unassigned ap jchaffraix kling koivisto sam simon.fraser oldest_to_newest 562856 0 mitz 2012-02-22 23:02:13 -0800 <rdar://problem/10598598> To reproduce: DumpRenderTree LayoutTests/fast/dynamic/crash-paint-no-documentElement-renderer.html Result: ASSERTION FAILED: !m_styleSheets->length() /Projects/Safari/OpenSource/Source/WebCore/dom/Document.cpp(699) : void WebCore::Document::setCompatibilityMode(WebCore::Document::CompatibilityMode) 1 0x1026f3f45 WebCore::Document::setCompatibilityMode(WebCore::Document::CompatibilityMode) 2 0x102b7b370 WebCore::HTMLTreeBuilder::defaultForInitial() 3 0x102b7603f WebCore::HTMLTreeBuilder::processEndOfFile(WebCore::AtomicHTMLToken&) 4 0x102b71ff0 WebCore::HTMLTreeBuilder::processToken(WebCore::AtomicHTMLToken&) 5 0x102b7135b WebCore::HTMLTreeBuilder::constructTreeFromAtomicToken(WebCore::AtomicHTMLToken&) 6 0x102b7126c WebCore::HTMLTreeBuilder::constructTreeFromToken(WebCore::HTMLToken&) 7 0x102abd39c WebCore::HTMLDocumentParser::pumpTokenizer(WebCore::HTMLDocumentParser::SynchronousMode) 8 0x102abce55 WebCore::HTMLDocumentParser::pumpTokenizerIfPossible(WebCore::HTMLDocumentParser::SynchronousMode) 9 0x102abccf4 WebCore::HTMLDocumentParser::prepareToStopParsing() 10 0x102abdf13 WebCore::HTMLDocumentParser::attemptToEnd() 11 0x102abdf68 WebCore::HTMLDocumentParser::finish() 12 0x1026fbc64 WebCore::Document::explicitClose() 13 0x1026f6eea WebCore::Document::close() 14 0x102ea53e5 WebCore::jsHTMLDocumentPrototypeFunctionClose(JSC::ExecState*) […] 562870 1 mitz 2012-02-22 23:20:49 -0800 This is tracked by bug 68859. 562875 2 128406 mitz 2012-02-22 23:25:49 -0800 Created attachment 128406 Disable the test that triggers bug 68859 563204 3 ap 2012-02-23 10:23:47 -0800 (In reply to comment #1) > This is tracked by bug 68859. Also by bug 51793 and bug 75135, apparently. Could you please dupe as appropriate? 563227 4 128406 jchaffraix 2012-02-23 10:34:44 -0800 Comment on attachment 128406 Disable the test that triggers bug 68859 I don't think this is the way to go. This test doesn't cover bug 68859 (it does hit the ASSERT and I don't have a good solution for it), it covers bug 64284 which is a crash in release. I had disabled the test in debug or skipped it for all platforms to avoid hitting the ASSERT. I don't know what happened but you should add the test to your platform accordingly instead of losing coverage on all platforms. 563242 5 mitz 2012-02-23 10:42:12 -0800 (In reply to comment #4) > (From update of attachment 128406 [details]) > I had disabled the test in debug or skipped it for all platforms to avoid hitting the ASSERT Where and how did you do that? > I don't know what happened but you should add the test to your platform accordingly instead of losing coverage on all platforms. Are there platforms that don’t fail the assertion when running this test? Skipped lists are platform-specific and are meant to address platform-specific issues. As far as I can tell, this test would cause the assertion to fail on any platform, so it should be disabled. 563252 6 jchaffraix 2012-02-23 10:52:55 -0800 (In reply to comment #5) > (In reply to comment #4) > > (From update of attachment 128406 [details] [details]) > > I had disabled the test in debug or skipped it for all platforms to avoid hitting the ASSERT > > Where and how did you do that? By adding the test to the appropriate test_expectations.txt & Skipped files. > > > I don't know what happened but you should add the test to your platform accordingly instead of losing coverage on all platforms. > > Are there platforms that don’t fail the assertion when running this test? No, all platform should hit the assertion in debug and this is expected. > Skipped lists are platform-specific and are meant to address platform-specific issues. As far as I can tell, this test would cause the assertion to fail on any platform, so it should be disabled. I understand your point and agree that it is unfortunate to have this test hit the assertion on all platforms. If the test were providing no coverage and be a maintenance burden, then yes disabling would be the way to go. However here it checks against regressing the crash from bug 64284 in release so the situation is not clear cut to me and I would rather keep this coverage. I don't have the context of your wanting to disable this test so maybe there is something I am missing here. 563761 7 128406 sam 2012-02-23 17:46:08 -0800 Comment on attachment 128406 Disable the test that triggers bug 68859 The correct thing to do here is disable the test until the assertion can be dealt with. 563762 8 sam 2012-02-23 17:47:05 -0800 > I don't have the context of your wanting to disable this test so maybe there is something I am missing here. The context is that it is not acceptable to have a test that asserts. 563774 9 jchaffraix 2012-02-23 17:59:58 -0800 (In reply to comment #8) > > I don't have the context of your wanting to disable this test so maybe there is something I am missing here. > > The context is that it is not acceptable to have a test that asserts. So be it, though you should at least remove the test from the test_expectations.txt & Skipped files (I wonder if this will break platforms that have it). Also adding a comment to bug 68859 that this test should be re-enabled once we fix the assert. 563902 10 mitz 2012-02-23 21:25:06 -0800 (In reply to comment #9) > (In reply to comment #8) > > > I don't have the context of your wanting to disable this test so maybe there is something I am missing here. > > > > The context is that it is not acceptable to have a test that asserts. > > So be it, though you should at least remove the test from the test_expectations.txt & Skipped files Done. > Also adding a comment to bug 68859 that this test should be re-enabled once we fix the assert. Will do. Landed in <http://trac.webkit.org/r108725>. 563911 11 ap 2012-02-23 21:37:23 -0800 > Also by bug 51793 and bug 75135, apparently. Could you please dupe as appropriate? Could someone who understands this bug please clean up the duplicates? 564438 12 jchaffraix 2012-02-24 10:25:47 -0800 (In reply to comment #11) > > Also by bug 51793 and bug 75135, apparently. Could you please dupe as appropriate? > > Could someone who understands this bug please clean up the duplicates? Done, those were indeed duplicates. Thanks Mitz for addressing the comments. 128406 2012-02-22 23:25:49 -0800 2012-02-23 17:46:08 -0800 Disable the test that triggers bug 68859 79333_r1.diff text/plain 2983 mitz SW5kZXg6IExheW91dFRlc3RzL0NoYW5nZUxvZwo9PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09Ci0tLSBMYXlvdXRUZXN0cy9D aGFuZ2VMb2cJKHJldmlzaW9uIDEwODYxNCkKKysrIExheW91dFRlc3RzL0NoYW5nZUxvZwkod29y a2luZyBjb3B5KQpAQCAtMSwzICsxLDE1IEBACisyMDEyLTAyLTIyICBEYW4gQmVybnN0ZWluICA8 bWl0ekBhcHBsZS5jb20+CisKKyAgICAgICAgUkVHUkVTU0lPTjogQXNzZXJ0aW9uIGZhaWx1cmUg aW4gRG9jdW1lbnQ6OnNldENvbXBhdGliaWxpdHlNb2RlKCkgKCFtX3N0eWxlU2hlZXRzLT5sZW5n dGgoKSkgaW4gZmFzdC9keW5hbWljL2NyYXNoLXBhaW50LW5vLWRvY3VtZW50RWxlbWVudC1yZW5k ZXJlci5odG1sCisgICAgICAgIGh0dHBzOi8vYnVncy53ZWJraXQub3JnL3Nob3dfYnVnLmNnaT9p ZD03OTMzMworCisgICAgICAgIFJldmlld2VkIGJ5IE5PQk9EWSAoT09QUyEpLgorCisgICAgICAg IERpc2FibGVkIHRoZSB0ZXN0LiBUaGUgYXNzZXJ0aW9uIGZhaWx1cmUgaXMgdHJhY2tlZCBieSA8 aHR0cDovL3dlYmtpdC5vcmcvYi82ODg1OT4uCisKKyAgICAgICAgKiBmYXN0L2R5bmFtaWMvY3Jh c2gtcGFpbnQtbm8tZG9jdW1lbnRFbGVtZW50LXJlbmRlcmVyLmh0bWw6IFJlbW92ZWQuCisgICAg ICAgICogZmFzdC9keW5hbWljL2NyYXNoLXBhaW50LW5vLWRvY3VtZW50RWxlbWVudC1yZW5kZXJl ci5odG1sLWRpc2FibGVkOiBDb3BpZWQgZnJvbSBMYXlvdXRUZXN0cy9mYXN0L2R5bmFtaWMvY3Jh c2gtcGFpbnQtbm8tZG9jdW1lbnRFbGVtZW50LXJlbmRlcmVyLmh0bWwuCisKIDIwMTItMDItMjIg IERtaXRyeSBMb21vdiAgPGRzbG9tb3ZAZ29vZ2xlLmNvbT4KIAogICAgICAgICBbQ2hyb21pdW1d W1Y4XSBTdXBwb3J0IFVpbnQ4Q2xhbXBlZEFycmF5IGluIHBvc3RNZXNzYWdlCkluZGV4OiBMYXlv dXRUZXN0cy9mYXN0L2R5bmFtaWMvY3Jhc2gtcGFpbnQtbm8tZG9jdW1lbnRFbGVtZW50LXJlbmRl cmVyLmh0bWwKPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PQotLS0gTGF5b3V0VGVzdHMvZmFzdC9keW5hbWljL2NyYXNoLXBh aW50LW5vLWRvY3VtZW50RWxlbWVudC1yZW5kZXJlci5odG1sCShyZXZpc2lvbiAxMDg2MDkpCisr KyBMYXlvdXRUZXN0cy9mYXN0L2R5bmFtaWMvY3Jhc2gtcGFpbnQtbm8tZG9jdW1lbnRFbGVtZW50 LXJlbmRlcmVyLmh0bWwJKHdvcmtpbmcgY29weSkKQEAgLTEsMjAgKzAsMCBAQAotPCFET0NUWVBF IGh0bWw+Ci08aHRtbD4KLTxib2R5IG9ubG9hZD0iZ28oKSI+Ci08cD5CdWcgNjQyODQ6IENyYXNo IGluIFJlbmRlckJveDo6cGFpbnRCb3hEZWNvcmF0aW9ucyB3aGVuIGRvY3VtZW50RWxlbWVudCBo YXMgbm8gcmVuZGVyZXI8L3A+Ci08cD5Gb3IgdGhpcyB0ZXN0IHRvIFBBU1MsIGl0IHNob3VsZCBu b3QgQVNTRVJUIG9yIENSQVNILjwvcD4KLTxzY3JpcHQ+Ci0gICAgaWYgKHdpbmRvdy5sYXlvdXRU ZXN0Q29udHJvbGxlcikKLSAgICAgICAgbGF5b3V0VGVzdENvbnRyb2xsZXIuZHVtcEFzVGV4dCgp OwotCi0gICAgZnVuY3Rpb24gZ28oKSB7Ci0gICAgICAgIGRvY3VtZW50Lm9wZW4oKTsKLSAgICAg ICAgdmFyIG9VRWxlbWVudCA9IGRvY3VtZW50LmNyZWF0ZUVsZW1lbnQoIlUiKTsKLSAgICAgICAg b1VFbGVtZW50LmhpZGRlbj10cnVlOwotICAgICAgICBvVUVsZW1lbnQuaW5uZXJIVE1MPSI8c3R5 bGU+Knstd2Via2l0LWJvcmRlci1iZWZvcmUtc3R5bGU6Z3Jvb3ZlfSI7Ci0gICAgICAgIGRvY3Vt ZW50LmFwcGVuZENoaWxkKG9VRWxlbWVudCk7Ci0gICAgICAgIGRvY3VtZW50LmNsb3NlKCk7Ci0g ICAgfQotPC9zY3JpcHQ+Ci08L2JvZHk+Ci08L2h0bWw+CkluZGV4OiBMYXlvdXRUZXN0cy9mYXN0 L2R5bmFtaWMvY3Jhc2gtcGFpbnQtbm8tZG9jdW1lbnRFbGVtZW50LXJlbmRlcmVyLmh0bWwtZGlz YWJsZWQKPT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PQotLS0gTGF5b3V0VGVzdHMvZmFzdC9keW5hbWljL2NyYXNoLXBhaW50 LW5vLWRvY3VtZW50RWxlbWVudC1yZW5kZXJlci5odG1sLWRpc2FibGVkCShyZXZpc2lvbiAxMDg2 MDkpCShmcm9tIExheW91dFRlc3RzL2Zhc3QvZHluYW1pYy9jcmFzaC1wYWludC1uby1kb2N1bWVu dEVsZW1lbnQtcmVuZGVyZXIuaHRtbDoxMDg2MDkpCisrKyBMYXlvdXRUZXN0cy9mYXN0L2R5bmFt aWMvY3Jhc2gtcGFpbnQtbm8tZG9jdW1lbnRFbGVtZW50LXJlbmRlcmVyLmh0bWwtZGlzYWJsZWQJ KHdvcmtpbmcgY29weSkKQEAgLTAsMCArMSwyMCBAQAorPCFET0NUWVBFIGh0bWw+Cis8aHRtbD4K Kzxib2R5IG9ubG9hZD0iZ28oKSI+Cis8cD5CdWcgNjQyODQ6IENyYXNoIGluIFJlbmRlckJveDo6 cGFpbnRCb3hEZWNvcmF0aW9ucyB3aGVuIGRvY3VtZW50RWxlbWVudCBoYXMgbm8gcmVuZGVyZXI8 L3A+Cis8cD5Gb3IgdGhpcyB0ZXN0IHRvIFBBU1MsIGl0IHNob3VsZCBub3QgQVNTRVJUIG9yIENS QVNILjwvcD4KKzxzY3JpcHQ+CisgICAgaWYgKHdpbmRvdy5sYXlvdXRUZXN0Q29udHJvbGxlcikK KyAgICAgICAgbGF5b3V0VGVzdENvbnRyb2xsZXIuZHVtcEFzVGV4dCgpOworCisgICAgZnVuY3Rp b24gZ28oKSB7CisgICAgICAgIGRvY3VtZW50Lm9wZW4oKTsKKyAgICAgICAgdmFyIG9VRWxlbWVu dCA9IGRvY3VtZW50LmNyZWF0ZUVsZW1lbnQoIlUiKTsKKyAgICAgICAgb1VFbGVtZW50LmhpZGRl bj10cnVlOworICAgICAgICBvVUVsZW1lbnQuaW5uZXJIVE1MPSI8c3R5bGU+Knstd2Via2l0LWJv cmRlci1iZWZvcmUtc3R5bGU6Z3Jvb3ZlfSI7CisgICAgICAgIGRvY3VtZW50LmFwcGVuZENoaWxk KG9VRWxlbWVudCk7CisgICAgICAgIGRvY3VtZW50LmNsb3NlKCk7CisgICAgfQorPC9zY3JpcHQ+ Cis8L2JvZHk+Cis8L2h0bWw+Cg==