November 2009 - Wine-bugs - list.winehq.org

[Bug 20627] New: chromium's base_unittests's StackTrace.OutputToStream fails (RtlCaptureStackBackTrace unimplemented)
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20627 Summary: chromium's base_unittests's StackTrace.OutputToStream fails (RtlCaptureStackBackTrace unimplemented) Product: Wine Version: 1.1.32 Platform: PC OS/Version: Linux Status: NEW Keywords: download, source Severity: normal Priority: P2 Component: ntdll AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com $ wine base_unittests.exe --gtest_filter=StackTrace.OutputToStream fails with [ RUN ] StackTrace.OutputToStream fixme:seh:RtlCaptureStackBackTrace (0, 62, 0x32facc, (nil)) stub! .\debug_util_unittest.cc(22): error: Expected: (frames_found) >= (5u), actual: 0 vs 5 No stack frames found. Skipping rest of test. [ FAILED ] StackTrace.OutputToStream (1070 ms) Looks like RtlCaptureStackBackTrace needs implementing. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 2

[Bug 20626] New: Chromium's unit_tests.exe's DownloadManagerTest.TestDownloadFilename fails
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20626 Summary: Chromium's unit_tests.exe's DownloadManagerTest.TestDownloadFilename fails Product: Wine Version: 1.1.32 Platform: PC OS/Version: Linux Status: NEW Keywords: download, source Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com $ wine src/chrome/Debug/unit_tests.exe --gtest_filter=DownloadManagerTest.TestDownloadFilename fails with .\browser\download\download_manager_unittest.cc(471): error: Value of: file_name Actual: L"evil.download" Expected: kGeneratedFiles[i].expected_name Which is: L"evil.xml" Related source: http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/download/dow… http://src.chromium.org/viewvc/chrome/trunk/src/chrome/browser/download/dow… To repeat: $ wget http://kegel.com/wine/chromium/chromium-tests.tar.bz2 $ tar -xjvf chromium-tests.tar.bz2 $ wine src/chrome/Debug/unit_tests.exe --gtest_filter=DownloadManagerTest.TestDownloadFilename -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 2

[Bug 20619] New: chromium's unit_tests.exe's RenderViewTest.PrintLayoutTest fails with FATAL:image.cc(211)] Check failed: rect.x() == 0 (-1 vs. 0)
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20619 Summary: chromium's unit_tests.exe's RenderViewTest.PrintLayoutTest fails with FATAL:image.cc(211)] Check failed: rect.x() == 0 (-1 vs. 0) Product: Wine Version: 1.1.32 Platform: PC OS/Version: Linux Status: NEW Keywords: download, source Severity: normal Priority: P2 Component: gdi32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com The check that is failing in chromium's source is gfx::Rect rect(metafile.GetBounds()); DCHECK_EQ(rect.x(), 0); It looks like Chromium is creating an in-memory metafile to do a print preview, then loads that metafile, does a sanity check on its bounding box, and barfs. Related chromium sources seem to be http://src.chromium.org/viewvc/chrome/trunk/src/ chrome/renderer/render_view_unittest.cc chrome/renderer/print_web_view_helper_win.cc printing/image.cc To reproduce: wget http://kegel.com/wine/chromium/chromium-tests.tar.bz2 tar -xjvf chromium-tests.tar.bz2 wine src/chrome/Debug/unit_tests.exe --gtest_filter=RenderViewTest.PrintLayoutTest Also affects three other tests: RenderViewTest.OnPrintPages RenderViewTest.PrintWithIframe RenderViewTest.PrintWithJavascript I'll attach a +relay log, lightly trimmed. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 3

[Bug 20567] New: Uninitialised memory reference in RSAENH_CPImportKey
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20567 Summary: Uninitialised memory reference in RSAENH_CPImportKey Product: Wine Version: 1.1.32 Platform: PC OS/Version: Linux Status: NEW Keywords: download, source, testcase Severity: normal Priority: P2 Component: rsaenh AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com This shows up in Chromium's base_unittests.exe in RSAPrivateKeyUnitTest.ShortIntegers test, and also in the wine conformance tests, http://kegel.com/wine/valgrind/logs/2009-10-30-09.22/vg-rsaenh_rsaenh.txt Use of uninitialised value of size 4 at desfunc (des.c:1366) by des3_ecb_encrypt (des.c:1478) by encrypt_block_impl (implglue.c:282) by RSAENH_CPEncrypt (rsaenh.c:2173) by CryptEncrypt (crypt.c:1083) by CryptProtectData (protectdata.c:906) by store_key_pair (rsaenh.c:962) by store_key_container_keys (rsaenh.c:1147) by release_and_install_key (rsaenh.c:2644) by import_public_key (rsaenh.c:2790) by import_key (rsaenh.c:2973) by RSAENH_CPImportKey (rsaenh.c:3020) by CryptImportKey (crypt.c:1827) by test_import_export (rsaenh.c:1654) Uninitialised value was created by a client request at mark_block_uninitialized (heap.c:187) by RtlAllocateHeap (heap.c:1429) by store_key_pair (rsaenh.c:953) by store_key_container_keys (rsaenh.c:1147) by release_and_install_key (rsaenh.c:2644) by import_public_key (rsaenh.c:2790) by import_key (rsaenh.c:2973) by RSAENH_CPImportKey (rsaenh.c:3020) by CryptImportKey (crypt.c:1827) by test_import_export (rsaenh.c:1654) BTW someone should check whether we want to refresh the rsaenh source that was derived from libtomcrypt; there is a two years' newer version at http://libtomcrypt.com/download.html -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 10

[Bug 20523] New: Installing Internet Explorer fails due to problem in setupapi
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20523 Summary: Installing Internet Explorer fails due to problem in setupapi Product: Wine Version: 1.1.31 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: setupapi AssignedTo: wine-bugs(a)winehq.org ReportedBy: spaschukat(a)users.sourceforge.net Created an attachment (id=24450) --> (http://bugs.winehq.org/attachment.cgi?id=24450) +setupapi output of IE setup Since 1.1.31 installing Internet Explorer 6 fails at beginning of setup. Regression: b70bbe3d29c5a4f151d1a9987f15fdb93490f4f6 is first bad commit commit b70bbe3d29c5a4f151d1a9987f15fdb93490f4f6 Author: Alexandre Julliard <julliard(a)winehq.org> Date: Thu Oct 1 19:44:43 2009 +0200 setupapi: Duplicate the string substitution code to implement GenFormStrWithoutPlaceHolders16. :040000 040000 540654246bcf3220ec0277858817f74af9205ea5 3ff9d15a6a87a364cb24826e0db67aa988f86496 M dlls Attached a +setupapi output, last line is: trace:setupapi:GenFormStrWithoutPlaceHolders16 "%IESetup_CIF%" -> "IESetup_CIF" Output result of previous version is: trace:setupapi:GenFormStrWithoutPlaceHolders16 "%IESetup_CIF%" -> "iesetup.cif" -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 8

[Bug 20503] New: ref counting bug in crypt32
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20503 Summary: ref counting bug in crypt32 Product: Wine Version: 1.1.32 Platform: PC OS/Version: Linux Status: NEW Severity: normal Priority: P2 Component: crypt32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: juan_lang(a)yahoo.com Created an attachment (id=24412) --> (http://bugs.winehq.org/attachment.cgi?id=24412) Show presence of ref counting bug crypt32 uses ref counting to keep track of certificates, CRLs, and CTLs. The ref counting isn't done correctly, which can be seen by running 'WINEDEBUG=crypt make store.ok' in the tests. Looking through the output you'll find instances where the ref count becomes negative. I'm attaching a patch which adds an assert that the ref count is positive before decrementing it. Applying the patch and running 'make store.ok' produces an assertion failure: context.c:150: Context_Release: Assertion `base->ref > 0' failed. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 7

[Bug 20477] New: Regression -- Dragon NaturallySpeaking MSI installer fails
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20477 Summary: Regression -- Dragon NaturallySpeaking MSI installer fails Product: Wine Version: unspecified Platform: PC URL: http://nuance.com OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: msi AssignedTo: wine-bugs(a)winehq.org ReportedBy: susancragin(a)earthlink.net Created an attachment (id=24373) --> (http://bugs.winehq.org/attachment.cgi?id=24373) terminal output from dump git version: wine-1.1.32-55-gd1472e2 Today's git will not install Dragon NaturallySpeaking. I tried installing on a clean wineprefix. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 21

[Bug 20371] New: SolidWorks 2008: Installation fails
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20371 Summary: SolidWorks 2008: Installation fails Product: Wine Version: 1.1.31 Platform: PC OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: msi AssignedTo: wine-bugs(a)winehq.org ReportedBy: lukasz.wojnilowicz(a)gmail.com Created an attachment (id=24138) --> (http://bugs.winehq.org/attachment.cgi?id=24138) wine-1.1.31 with dotnet installed through winetricks I'm using Fedora 11 32 bit with Wine 1.1.31 The installation of SolidWorks 2008 fails with msiexec error. Last installer step was something like adding Windows registry keys. The application needs .NET Framework 3.0. If i install it through winetricks then whole wineprefix is 2.9 GB big after installation error. If I use clean wineprefix then whole wineprefix is 2.4 GB big after installation error. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 29

[Bug 20358] New: chromium's base_unittests.exe fails on RSAPrivateKeyUnitTest.*
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20358 Summary: chromium's base_unittests.exe fails on RSAPrivateKeyUnitTest.* Product: Wine Version: 1.1.31 Platform: PC OS/Version: Linux Status: NEW Keywords: download, source, testcase Severity: normal Priority: P2 Component: crypt32 AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com The three tests, RSAPrivateKeyUnitTest.InitRandomTest, RSAPrivateKeyUnitTest.PublicKeyTest, and RSAPrivateKeyUnitTest.ShortIntegers all fail. Source involved is http://src.chromium.org/viewvc/chrome/trunk/src/base/crypto/rsa_private_key… http://src.chromium.org/viewvc/chrome/trunk/src/base/crypto/rsa_private_key… http://src.chromium.org/viewvc/chrome/trunk/src/base/crypto/rsa_private_key… etc. I'll attach +relay,+crypt logs of the three tests failing. -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 24

[Bug 20353] New: chromium ipc_tests.exe fails when KERNEL32.GetQueuedCompletionStatus() hangs on peer disconnect of named pipe?
by wine-bugs＠winehq.org 13 Nov '09

13 Nov '09

http://bugs.winehq.org/show_bug.cgi?id=20353 Summary: chromium ipc_tests.exe fails when KERNEL32.GetQueuedCompletionStatus() hangs on peer disconnect of named pipe? Product: Wine Version: unspecified Platform: PC OS/Version: Linux Status: NEW Keywords: download, source, testcase Severity: normal Priority: P2 Component: ntdll AssignedTo: wine-bugs(a)winehq.org ReportedBy: dank(a)kegel.com The script wget http://kegel.com/wine/chromium/ipc_tests.exe.rz rzip -d ipc_tests.exe.rz WINEPREFIX=/data/dkegel/.winetest /data/dkegel/wine-git/wine notepad & i=1 while test $i -lt 100 do WINEDEBUG=+process,+relay WINEPREFIX=/data/dkegel/.winetest /data/dkegel/wine-git/wine ipc_tests.exe --gtest_filter=IPCChannelTest.SendMessageInChannelConnected > ipc$i.log 2>&1 i=`expr $i + 1` done grep FAIL ipc*.log shows about 10 failures of IPCChannelTest.SendMessageInChannelConnected here (with or without logging). Looking at the logs with grep shows that successful runs end with a call to ExitProcess(0) in both parent and child, but failed runs end with just one ExitProcess(1) and a call to GetQueuedCompletionStatus() that does not return. Searching through the chromium codebase seems to indicate that this function is called only in MessagePumpForIO::GetIOItem() in http://src.chromium.org/viewvc/chrome/trunk/src/base/message_pump_win.cc The IO completion events in this test seem mainly to come from reads of a named pipe. Grepping for the handle of that named pipe plus a few other interesting words, egrep 'FAIL|Exit|error|PASS|RUN|CreateProcess|GetQueuedCompletionStatus|0x68[^0-9a-f]|00068[^0-9a-f]' shows lots of normal operation (an NtReadFile() followed by a few GetQueuedCompletionStatus()'s until one succeeds, then repeat; both parent and child process do this), but towards the end when the test is shutting down, something funny happens: [parent process shuts its end down] 0024:Call KERNEL32.GetQueuedCompletionStatus(00000064,0033f9e8,0033f920,0033f914,00000000) ret=004b9d69 trace:sync:GetQueuedCompletionStatus (0x64,0x33f9e8,0x33f920,0x33f914,0) 0024:Ret KERNEL32.GetQueuedCompletionStatus() retval=00000000 ret=004b9d69 0024:Call KERNEL32.CancelIo(00000068) ret=00520112 trace:ntdll:NtCancelIoFile 0x68 0x33fa10 0024:Call KERNEL32.CloseHandle(00000068) ret=00520135 0024:Call KERNEL32.GetQueuedCompletionStatus(00000064,0033fa68,0033f9a0,0033f994,ffffffff) ret=004b9d69 trace:sync:GetQueuedCompletionStatus (0x64,0x33fa68,0x33f9a0,0x33f994,-1) 0024:Ret KERNEL32.GetQueuedCompletionStatus() retval=00000001 ret=004b9d69 0024:Call KERNEL32.GetQueuedCompletionStatus(00000064,0033fa68,0033f9a0,0033f994,ffffffff) ret=004b9d69 trace:sync:GetQueuedCompletionStatus (0x64,0x33fa68,0x33f9a0,0x33f994,-1) 0024:Ret KERNEL32.GetQueuedCompletionStatus() retval=00000001 ret=004b9d69 [child process wakes up, does some I/O, checks queue, hangs] 0037:Call KERNEL32.ReadFile(00000068,009ba620,00001000,0033f780,009ba5c4) ret=00520f00 trace:ntdll:NtReadFile (0x68,(nil),(nil),0x9ba5c4,0x9ba5c4,0x9ba620,0x00001000,0x33f2e0,(nil)),partial stub! 0037:Call KERNEL32.GetQueuedCompletionStatus(00000064,0033fa04,0033f93c,0033f930,00000000) ret=004b9d69 trace:sync:GetQueuedCompletionStatus (0x64,0x33fa04,0x33f93c,0x33f930,0) 0037:Ret KERNEL32.GetQueuedCompletionStatus() retval=00000000 ret=004b9d69 0037:Call KERNEL32.GetQueuedCompletionStatus(00000064,0033f95c,0033f894,0033f888,ffffffff) ret=004b9d69 trace:sync:GetQueuedCompletionStatus (0x64,0x33f95c,0x33f894,0x33f888,-1) [parent process gives up and declares failure] .\ipc_tests.cc(352): error: Value of: base::WaitForSingleProcess(process_handle, 5000) 0024:Call KERNEL32.OutputDebugStringA(009d5f90 ".\\ipc_tests.cc(352): error: Value of: base::WaitForSingleProcess(process_handle, 5000)\n Actual: false\nExpected: true") ret=00453839 This is all just extrapolation from one log, but it does seem like wine might not be doing the right thing with GetQueuedCompletionStatus() and named pipes. (Dang, I should have covered that case when I wrote my named pipe tests in 2003 :-) -- Configure bugmail: http://bugs.winehq.org/userprefs.cgi?tab=email Do not reply to this email, post in Bugzilla using the above URL to reply. ------- You are receiving this mail because: ------- You are watching all bug changes.

1 3