wine-bugs June 2023

wine-bugs@winehq.org

1 participants
2705 discussions

[Bug 55157] New: Net6 x64 app: unable to allocate more than ~1GB of contiguous memory, OutOfMemory exception
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=55157 Bug ID: 55157 Summary: Net6 x64 app: unable to allocate more than ~1GB of contiguous memory, OutOfMemory exception Product: Wine Version: 8.11 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: korshunov(a)macroscop.com Distribution: --- A "big" net6 app may throw an OutOfMemory exception at some point in time. It was found that the exception occurs when the size of byte array is increased and exception is repeatable 100% - it always fails when byte array resized from 26mb to 53mb. Since big app is hard to debug, a simple test app was written, all it tries to do is to resize array x2 in a loop. Test app fails with dotnet OutOfMemory exception 100% of the time, but in the test app the "limit" at which it fails is different, about ~1gb or "Failed at resizing from 534773760 to 1069547520". Here's the code of the test app: var array = new byte[1023]; while (true) try { Array.Resize(ref array, array.Length * 2); } catch (Exception ex) { Console.WriteLine($"Failed at resizing from {array.Length} to {array.Length * 2}\r\n" + ex); return; } Test app is built for "AnyCPU, x64 preferred". Net6 runtime (6.0.16) is installed via winetricks. For the tests, x86 version of net6 runtime has been uninstalled, so the test app is guaranteed to run in x64. Test app is available via the link below. What I've already tried: - wine 8.11 staging - fails with OutOfMemory exception, "Failed at resizing from 534773760 to 1069547520". WINEDEBUG output available through the link below. - wine 8.0 with patch from https://bugs.winehq.org/show_bug.cgi?id=52396 - fails with same error. WINEDEBUG output available through the link below. - running app with wine64, used clean prefix - same error. - wine 7.13 - fails with same error. - wine 8.11, but test app built for netFramework46 - no OutOfMemory, fails when overflowing integer, "Failed at resizing from 2139095040 to -16777216" which is expected, but it shows that it gets to 2gb mark for the array. I consider this as "working fine". WINEDEBUG output available through the link below. - WineGE with multiple versions from 6.14 to 8.8, consisting of Proton and WineGE patches - net6 app working fine with every wineGE version tested, just as netFramework Test app - created prefix with wine 8.11, run test app, then run test app with WineGE in the same exact prefix - wine fails with OutOfMemory while wineGE executes app normally. - made different net6 app, which just allocates a single array of ~1.5 GB (new byte[1_500_000_000]) - app fails with OutOfMemory every time. It fails at about 1GB (950mb works, while 1050mb does not) Tests were done on three different virtual machines (ubuntu22, centos7, debian9) with different amount of RAM (4-8 GB, swap is 2.5GB). Wine 8.11 is installed on Ubuntu 22 with 4GB RAM. What caught my attention in the logs was the message "find_free_block created new sub-heap xxx of 0x400000 bytes" (4GB, all of the RAM), which mentioned in "net6" logs 1 time, and not mentioned in "netFramework46" logs. What else can I try to fix this or to provide more information? Links to a WINEDEBUG output (wine8.11-net6, wine8.11-netFramework46, "wine 8.0 with patch"-net6), logs taken with command WINEDEBUG=+timestamp,+seh,+heap,+virtual,+pid wine Tests_autoresize/Tests_net6.exe >>wine811_tshvp_net6_resize.log 2>&1 https://drive.google.com/drive/folders/1fgr0J6dKBvBOHWrfYC0UCrflxiAYOjZ2?us… -- 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 5

[Bug 55044] New: dnSpy 6.2.0 crashes on start with a stackoverflow
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=55044 Bug ID: 55044 Summary: dnSpy 6.2.0 crashes on start with a stackoverflow Product: Wine Version: 8.9 Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: dark.shadow4(a)web.de Distribution: --- This is introduced in https://github.com/dnSpyEx/dnSpy/releases/download/v6.2.0-rc1/dnSpy-net-win… Version 6.1.9 works: https://github.com/dnSpyEx/dnSpy/releases/download/v6.1.9/dnSpy-net-win64.z… -- 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 55098] New: wine 8.10 32-bit build fails with clang-16
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=55098 Bug ID: 55098 Summary: wine 8.10 32-bit build fails with clang-16 Product: Wine Version: 8.10 Hardware: x86 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: randrik(a)mail.ru Distribution: --- I tried to build new wine with llvm/clang 16.0.6 and gcc 11.2 (on Slackware 15.0 i586 upgraded manually from earlier vesrions). Kernel is x86_64 but al userland is i586/i686 Build via slackbuild fails like this: (will attach end of build log as attach) -- 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 54739] New: oleaut32:typelib - The 64-bit test_register_typelib_64() crashes in Wine
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=54739 Bug ID: 54739 Summary: oleaut32:typelib - The 64-bit test_register_typelib_64() crashes in Wine Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: oleaut32 Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- oleaut32:typelib - The 64-bit test_register_typelib_64() crashes in Wine: typelib.c:7057: Test failed: got 00000002 typelib.c:8195: Test failed: got wrong return code: 2, side: 0200 typelib.c:8231: Test failed: got: 80040155, side: 0200 typelib.c:8235: Test failed: got: 80040154, side: 0200 Unhandled exception: page fault on read access to 0x0000000000000000 in 64-bit code (0x00000000428dc1). [...] Backtrace: =>0 0x00000000428dc1 in oleaut32_test (+0x28dc1) (0x0000000050e538) 1 0x000000004460c8 in oleaut32_test (+0x460c8) (0x000000621080c0) 2 0x0000000050dbeb in oleaut32_test (+0x10dbeb) (0x00000000ab4415) 3 0x0000000050d447 in oleaut32_test (+0x10d447) (0000000000000000) 4 0x0000007b629409 BaseThreadInitThunk+0x9(unknown=<internal error>, entry=<internal error>, arg=<internal error>) [Z:\home\winetest\tools\testbot\var\wine\dlls\kernel32\thread.c:61] in kernel32 (0000000000000000) 5 0x00000170061213 __wine_pop_frame(entry=000000000050D3E0, arg=00000000003F0000) [Z:\home\winetest\tools\testbot\var\wine\include\wine\exception.h:277] in ntdll (0000000000000000) 6 0x00000170061213 RtlUserThreadStart+0x83(entry=[<register RSP not accessible in this frame>, arg=[<register RSP not accessible in this frame>) [Z:\home\winetest\tools\testbot\var\wine\dlls\ntdll\thread.c:295] in ntdll (0000000000000000) 0x00000000428dc1 oleaut32_test+0x28dc1: mov (%rcx), %rax See https://test.winehq.org/data/patterns.html#oleaut32:typelib Where: 2 == ERROR_FILE_NOT_FOUND 0x80040154 == REGDB_E_CLASSNOTREG 0x80040155 == REGDB_E_IIDNOTREG -- 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 5

[Bug 50938] New: Stop alt key from getting stuck on alt-tab
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=50938 Bug ID: 50938 Summary: Stop alt key from getting stuck on alt-tab Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: sojuz151(a)gmail.com Distribution: --- When you use alt-tab shortcut to switch to another window and then switch back in, application will at no point receive alt key up signal. User needs to press alt again to make this key unstuck. This behevior is observed in Plasma and in lxde, both of them use alt-tab for windows switching. Behaviour on windows: alt key is not stuck in down position after alt-tabing out of any application How to test: this can be observed in any keyboard testing program. I used switch hitter for testing. Is it really a bug? Maybe not. Still very annoying in full-screen games. -- 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 5

[Bug 49787] New: Java fails to start with "Could not create the Java Virtual Machine" - "Could not reserve enough space for object heap"
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=49787 Bug ID: 49787 Summary: Java fails to start with "Could not create the Java Virtual Machine" - "Could not reserve enough space for object heap" Product: Wine Version: 5.16 Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: dark.shadow4(a)web.de Distribution: --- You can download java from https://jdk.java.net/java-se-ri/7 (just unpack) To reproduce, just go into "bin" and run "wine java -Xmx1024M -version". -- 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 55105] New: libHttpClient requires WINHTTP_OPTION_CLIENT_CERT_ISSUER_LIST support
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=55105 Bug ID: 55105 Summary: libHttpClient requires WINHTTP_OPTION_CLIENT_CERT_ISSUER_LIST support Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: winhttp Assignee: wine-bugs(a)winehq.org Reporter: flibitijibibo(a)gmail.com Distribution: --- libHttpClient is the web socket abstraction layer for Xbox Live titles on various platforms; on Windows it uses winhttp for http sources. Two example games using this library are Gears 5 and Halo: The Master Chief collection. One part of the implementation handles certificate authentication; in particular it handles ERROR_WINHTTP_CLIENT_AUTH_CERT_NEEDED, which according to MSDN should be handled by calling WinHttpQueryOption with WINHTTP_OPTION_CLIENT_CERT_ISSUER_LIST, so that it can attempt to find a relevant certificate chain and re-attempt the connection if a cert is found. Luckily, libHttpClient is open source, so the exact block that does this can be found here: https://github.com/microsoft/libHttpClient/blob/1c8386866fb180dd31890186291… As of the latest Git revision, Wine does not handle this parameter, so the default case is tripped: https://gitlab.winehq.org/wine/wine/-/blob/81859c9af70aa5ed6da3060488d1815a… -- 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 6

[Bug 49551] New: the installer tries to run Java 2 version 1.3.1 but failed
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=49551 Bug ID: 49551 Summary: the installer tries to run Java 2 version 1.3.1 but failed Product: Wine Version: 1.8.7 Hardware: x86-64 OS: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown Assignee: wine-bugs(a)winehq.org Reporter: pritishparihar95(a)gmail.com Distribution: --- Created attachment 67691 --> https://bugs.winehq.org/attachment.cgi?id=67691 Installanywhere windows for reference Installing Software Parashar Light Profession 7.0.1 After InstallAnywhere window no window comes up, I tried to find error code fron winedbg Error code: WineDbg starting on pid 001c 0x7b85f870: subl $8,%esp Invalid address (0x7b85f87a) for breakpoint 0, disabling it Process of pid=001c has terminated -- 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 7767] Zero G Software 'InstallAnywhere' installers crash or silently exit (Rosetta Stone, Reading Counts, World Religions, Parashara's Light Professional 7.0.1)
by WineHQ Bugzilla 28 Jun '23

28 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=7767 Damjan Jovanovic <damjan.jov(a)gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Fixed by SHA1| |f70fc9cbfc1b91864d99c6b3a52 | |7a77ad0106ad7 Resolution|--- |FIXED --- Comment #17 from Damjan Jovanovic <damjan.jov(a)gmail.com> --- My merge request was merged: https://gitlab.winehq.org/wine/wine/-/merge_requests/3111 Resolving fixed :-). Thank you for your bug report and tests. -- 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 0

[Bug 54860] New: mf:mf - test_media_session_events() sometimes gets an unexpected refcount in Wine
by WineHQ Bugzilla 27 Jun '23

27 Jun '23

https://bugs.winehq.org/show_bug.cgi?id=54860 Bug ID: 54860 Summary: mf:mf - test_media_session_events() sometimes gets an unexpected refcount in Wine Product: Wine Version: unspecified Hardware: x86-64 OS: Linux Status: NEW Severity: normal Priority: P2 Component: mfplat Assignee: wine-bugs(a)winehq.org Reporter: fgouget(a)codeweavers.com Distribution: --- mf:mf - test_media_session_events() sometimes gets an unexpected refcount: * The most recent Windows failure happened on 2023-04-06 w10pro64v2004: mf.c:2382: Test failed: Release returned 2 mf.c:2384: Test failed: Release returned 3 mf.c:2386: Test failed: Release returned 2 mf.c:2388: Test failed: Release returned 1 * The oldest Wine failure happened on 2023-04-11 and the refcount is always 1 in that case. mf.c:2558: Test failed: Release returned 1 See https://test.winehq.org/data/patterns.html#mf:mf Despite the line number changes these refcount failures happen in the same location. There have been some commits on 2023-04-10 and 11 so it looks like they may have moved the failure(s) from Windows to Wine (which would be progress), which is why I'm only mentioning Wine in the summary. -- 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 4

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

wine-bugs June 2023