[Bug 54169] New: ntoskrnl.exe:ntoskrnl - test_pnp_devices() sometimes fails on Windows and Linux
https://bugs.winehq.org/show_bug.cgi?id=54169
Bug ID: 54169 Summary: ntoskrnl.exe:ntoskrnl - test_pnp_devices() sometimes fails on Windows and Linux Product: Wine Version: unspecified Hardware: x86-64 OS: Windows Status: NEW Severity: normal Priority: P2 Component: ntoskrnl Assignee: wine-bugs@winehq.org Reporter: fgouget@codeweavers.com
ntoskrnl.exe:ntoskrnl - test_pnp_devices() fails on Windows 10 with test signing turned on:
ntoskrnl.c:1801: Test failed: failed to create device, error 0xe0000207 ntoskrnl.c:1805: Test failed: failed to create set hardware ID, error 0x57 ntoskrnl.c:1808: Test failed: failed to register device, error 0x57 ntoskrnl.c:1813: Test failed: got error 0x57 ntoskrnl.c:1817: Test failed: failed to install device, error 0xe000020b ntoskrnl.c:1822: Test failed: got error 0x57 ntoskrnl.c:1823: Test failed: got flags 0x10 ntoskrnl.c:1824: Test failed: got type 0 ntoskrnl.c:1546: Test failed: failed to get interface, error 0x103 ntoskrnl.c:1547: Test failed: wrong class {deadbeef-29ef-4538-a5fd-b69573a362c0} ntoskrnl.c:1549: Test failed: got flags 0x1 ntoskrnl.c:1563: Test failed: got 0 bus arrival messages ntoskrnl.c:1569: Test failed: failed to get interface, error 0x103 ntoskrnl.c:1570: Test failed: wrong class {deadbeef-29ef-4538-a5fd-b69573a362c0} ntoskrnl.c:1579: Test failed: got 0 bus arrival messages ntoskrnl.c:1580: Test failed: got 0 bus removal messages ntoskrnl.c:1585: Test failed: failed to get interface, error 0x103 ntoskrnl.c:1586: Test failed: wrong class {deadbeef-29ef-4538-a5fd-b69573a362c0} ntoskrnl.c:1588: Test failed: got flags 0x1 ntoskrnl.c:1605: Test failed: got 0 child arrival messages ntoskrnl.c:1612: Test failed: failed to get device, error 0x103 ntoskrnl.c:1613: Test failed: wrong class {4d36e97d-e325-11ce-bfc1-08002be10318} ntoskrnl.c:1616: Test failed: failed to get device ID, error 0x57 ntoskrnl.c:1617: Test failed: got ID "\x08" ntoskrnl.c:1621: Test failed: got error 0x57 ntoskrnl.c:1637: Test failed: got error 0x57 ntoskrnl.c:1639: Test failed: got flags 0x69f1b90b ntoskrnl.c:1640: Test failed: got type 0 ntoskrnl.c:1645: Test failed: got error 0x57 ntoskrnl.c:1650: Test failed: got error 0x57 ntoskrnl.c:1654: Test failed: got error 0x57 ntoskrnl.c:1655: Test failed: got type 0 ntoskrnl.c:1656: Test failed: got size 0 ntoskrnl.c:1662: Test failed: got error 0x57 ntoskrnl.c:1663: Test failed: got type 0 ntoskrnl.c:1664: Test failed: got size 0 ntoskrnl.c:1669: Test failed: got error 0x57 ntoskrnl.c:1670: Test failed: got type 0 ntoskrnl.c:1671: Test failed: got size 0 ntoskrnl.c:1676: Test failed: got error 0x57 ntoskrnl.c:1688: Test failed: failed to open child: 0xc0000034 ntoskrnl.c:1692: Test failed: got error 6 ntoskrnl.c:1693: Test failed: got id -559038737 ntoskrnl.c:1694: Test failed: got size 0 ntoskrnl.c:1699: Test failed: failed to open child: 0xc0000034 ntoskrnl.c:1703: Test failed: got error 6 ntoskrnl.c:1711: Test failed: got 0 child arrival messages ntoskrnl.c:1712: Test failed: got 0 child removal messages ntoskrnl.c:1715: Test failed: got error 6 ntoskrnl.c:1718: Test failed: got 0xc0000034 ntoskrnl.c:1722: Test failed: got error 6 ntoskrnl.c:1728: Test failed: got 0 child arrival messages ntoskrnl.c:1729: Test failed: got 0 child removal messages ntoskrnl.c:1833: Test failed: failed to remove device, error 0x57 ntoskrnl.c:1836: Test failed: expected failure ntoskrnl.c:1837: Test failed: got error 0 ntoskrnl.c:1857: Test failed: failed to open service, error 1060 ntoskrnl.c:291: Test failed: expected SERVICE_STOPPED, got 0
See https://test.winehq.org/data/patterns.html#ntoskrnl.exe:ntoskrnl
This happens a couple of times per month, mostly in the 64-bit w1064-tsign tests. However this also happened on: * 2022-07-04 w7u-pt-PT despite it not having test signing turned on. * 2022-10-24 master: this is the first known instance of these failures in Wine. * 2022-12-01 lastestmaster: these failures seem to have been systematic on this Wine test configuration.
Unlike on Windows, none of the official Wine test configurations seem to be having this failure.
https://bugs.winehq.org/show_bug.cgi?id=54169
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |source, testcase
https://bugs.winehq.org/show_bug.cgi?id=54169
Zeb Figura z.figura12@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |z.figura12@gmail.com
--- Comment #1 from Zeb Figura z.figura12@gmail.com --- ntoskrnl.c:1801: Test failed: failed to create device, error 0xe0000207
This is the basic problem, and everything that follows is almost certainly a consequence. This should only happen if the test image either had ntoskrnl tests already run, or had dinput tests already run, and either one failed to clean up properly. I'm guessing it was dinput, but unfortunately there aren't currently any examples on the test pages, so I can't figure out why.
https://bugs.winehq.org/show_bug.cgi?id=54169
--- Comment #2 from François Gouget fgouget@codeweavers.com --- Created attachment 74082 --> https://bugs.winehq.org/attachment.cgi?id=74082 WineTest w1064-tsign and debian11 logs
I attached a pair of WineTest logs in case it helps debug this. They correspond to the following instances (truncated after the ntoskrnl traces):
2022-12-15 on w1064-tsign (64-bit), no dinput crash 2023-01-25 on debian11-win32-ar-MA, which does have a dinput:force_feed crash
In both cases the tests were started from a clean slate so ntoskrnl was not run before these failures.
https://bugs.winehq.org/show_bug.cgi?id=54169
--- Comment #3 from Zeb Figura z.figura12@gmail.com --- The Windows log includes what looks like a hang in dinput:joystick (and another in dinput:joystick8, but that might be a consequence of the first hang.)
The Linux log shows a crash in dinput:force_feedback.
In both cases we can't clean up the device properly. We could use a different device ID in ntoskrnl, I guess, but the dinput bugs should be fixed anyway.
-
WineHQ Bugzilla