Bug 131216 – [WK2][GTK] Multipart tests ASSERT in !m_multipartSubresourceLoaders.contains(loader->identifier())

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKitGTK

[Sergio Villar Senin]

At least 2 tests are crashing:
http/tests/multipart/invalid-image-data.html
http/tests/multipart/multipart-html.php

Crash log for http/tests/multipart/invalid-image-data.html

STDERR: ASSERTION FAILED: !m_multipartSubresourceLoaders.contains(loader->identifier())
STDERR: ../../Source/WebCore/loader/DocumentLoader.cpp(1499) : void WebCore::DocumentLoader::subresourceLoaderFinishedLoadingOnePart(WebCore::ResourceLoader*)
STDERR: 1   0x7f5bdc499e2d /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libjavascriptcoregtk-3.0.so.0(WTFCrash+0x1e) [0x7f5bdc499e2d]
STDERR: 2   0x7f5bd6e0b6de /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libwebkit2gtk-3.0.so.25(_ZN7WebCore14DocumentLoader39subresourceLoaderFinishedLoadingOnePartEPNS_14ResourceLoaderE+0x94) [0x7f5bd6e0b6de]
STDERR: 3   0x7f5bd6e5fe06 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libwebkit2gtk-3.0.so.25(_ZN7WebCore17SubresourceLoader18didReceiveResponseERKNS_16ResourceResponseE+0x37e) [0x7f5bd6e5fe06]
STDERR: 4   0x7f5bd6e5c759 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libwebkit2gtk-3.0.so.25(_ZN7WebCore14ResourceLoader18didReceiveResponseEPNS_14ResourceHandleERKNS_16ResourceResponseE+0x61) [0x7f5bd6e5c759]
STDERR: 5   0x7f5bd7705c7f /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libwebkit2gtk-3.0.so.25(+0x523bc7f) [0x7f5bd7705c7f]
STDERR: 6   0x7f5bd0555591 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libgio-2.0.so.0(+0x95591) [0x7f5bd0555591]
STDERR: 7   0x7f5bd05555bb /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libgio-2.0.so.0(+0x955bb) [0x7f5bd05555bb]
STDERR: 8   0x7f5bd0e25a91 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libglib-2.0.so.0(+0x55a91) [0x7f5bd0e25a91]
STDERR: 9   0x7f5bd0e23295 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libglib-2.0.so.0(+0x53295) [0x7f5bd0e23295]
STDERR: 10  0x7f5bd0e23fec /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libglib-2.0.so.0(g_main_context_dispatch+0x33) [0x7f5bd0e23fec]
STDERR: 11  0x7f5bd0e241de /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libglib-2.0.so.0(+0x541de) [0x7f5bd0e241de]
STDERR: 12  0x7f5bd0e24607 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Dependencies/Root/lib64/libglib-2.0.so.0(g_main_loop_run+0x1d9) [0x7f5bd0e24607]
STDERR: 13  0x7f5bd65d89b8 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libwebkit2gtk-3.0.so.25(_ZN3WTF7RunLoop3runEv+0x42) [0x7f5bd65d89b8]
STDERR: 14  0x7f5bd652020a /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/lib/libwebkit2gtk-3.0.so.25(WebProcessMainGtk+0xed) [0x7f5bd652020a]
STDERR: 15  0x400b4d /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/bin/WebKitWebProcess(main+0x20) [0x400b4d]
STDERR: 16  0x7f5bcd8b6995 /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xf5) [0x7f5bcd8b6995]
STDERR: 17  0x400a69 /home/slave/webkitgtk/gtk-linux-64-debug/build/WebKitBuild/Debug/bin/WebKitWebProcess() [0x400a69]

[Charlie Turner]

None of these are crashing anymore, the multipart-html.php keeps failing with reason "failed unexpectedly (test was not run)" for reasons I don't understand. It's very similar to multipart-replace-non-html-content.php test in same directory that does run fine.

It's not crashing however, so I think this bug should be closed.

[Michael Catanzaro]

(In reply to Charlie Turner from comment #1)
> None of these are crashing anymore, the multipart-html.php keeps failing
> with reason "failed unexpectedly (test was not run)" for reasons I don't
> understand. It's very similar to multipart-replace-non-html-content.php test
> in same directory that does run fine.

I've never seen a PHP layout test before. It looks like a resource. We don't even have any expectations for this test. I think Sergio's post was probably just a typo?

> It's not crashing however, so I think this bug should be closed.

I agree, as long as you remove the expectation and leave bug #54251 open, this bug can be closed. If you post a patch to remove the expectation, commit-queue will close it.

[Michael Catanzaro]

Expectations updated in bug #173160.