June 2022 - Wine-Devel - list.winehq.org

Re: Test system for the Widl compiler

by Zebediah Figura

On 6/15/22 12:49, Nikolay Sivov wrote: > > > On 6/15/22 18:20, Bernhard Kölbl wrote: >> Hey everyone, >> >> during the course of implementing WinRT / UWP stuff, I regularly come >> across missing IDL 3.0 features in our Widl compiler. >> Unfortunately, Widl's code seems to have grown in a rater unpleasant >> direction and is sometimes a real burden to understand and maintain. I >> think some refactoring to its code would be very well fitting and make >> things a lot easier for everyone, but refactoring can also bring >> breakage and makes tracking down errors with Git harder. Additionally, >> as the code is already not that small, errors have an easy way to slip >> in without noticing. >> So to me, the ideal solution seems to add a testing system to the >> compiler, like most of Wine's components already have. >> >> Rémi and I already outlined some ideas: >> - Port the Winetest system to Unix programs >> - Make a libwidl or Widl.exe and test against that with our current >> test system >> - Add the tests into Widl itself as callable parameter >> >> What do you think is the best solution? >> Getting some suggestions would be great. :) > > There are options. First it depends whether you want to validate against > midl or not. If you do then you'll need something like widl.exe to run > on Windows. > > Next question is what your test data would be, and how to compare the > output reliably with expected output. For example we don't really need > exact matching for generated headers, or even marshaling format strings, > as long as they achieve their purpose. For typelibs however, it might be > useful to go for exact binary match of selected sections. We already have tests for typelibs and marshalling, in oleaut32:typelib and oleaut32:tmarshal respectively. I see no reason that's not sufficient already. As far as headers are concerned, we could probably take a similar approach in some ways. I don't know what exactly needs testing, but we could do things like ok(offsetof(mystruct.field) == 16, "wrong offset %u\n", offsetof(mystruct.field)); or (for a compile-time assertion) extern void test(ITestInterface *obj) { HRESULT (*myfunc)(type1 arg1, type2 *arg2, ...); myfunc = obj->lpVtbl->myfunc; } Ultimately we may not need to mess with widl, or write any extra test infrastructure.

June 15, 2022

[PATCH] include: Remove duplicate ICodecAPI::GetParameterValues function

by Andrew Eikum

Signed-off-by: Andrew Eikum <aeikum(a)codeweavers.com> --- include/icodecapi.idl | 1 - 1 file changed, 1 deletion(-) diff --git a/include/icodecapi.idl b/include/icodecapi.idl index 7954b3fc2f6..dc31daf2e2c 100644 --- a/include/icodecapi.idl +++ b/include/icodecapi.idl @@ -42,7 +42,6 @@ interface ICodecAPI : IUnknown HRESULT GetParameterRange ([in] const GUID *api, [out] VARIANT *min, [out] VARIANT *max, [out] VARIANT *delta); HRESULT GetParameterValues ([in] const GUID *api, [out, size_is(,*count)] VARIANT **values, [out] ULONG *count); - HRESULT GetParameterValues ([in] const GUID *api, [out, size_is(,*count)] VARIANT **values, [out] ULONG *count); HRESULT GetDefaultValue ([in] const GUID *api, [out] VARIANT *value); HRESULT GetValue ([in] const GUID *api, [out] VARIANT *value); -- 2.36.1

June 15, 2022

Re: Test system for the Widl compiler

by Nikolay Sivov

On 6/15/22 18:20, Bernhard Kölbl wrote: > Hey everyone, > > during the course of implementing WinRT / UWP stuff, I regularly come > across missing IDL 3.0 features in our Widl compiler. > Unfortunately, Widl's code seems to have grown in a rater unpleasant > direction and is sometimes a real burden to understand and maintain. I > think some refactoring to its code would be very well fitting and make > things a lot easier for everyone, but refactoring can also bring > breakage and makes tracking down errors with Git harder. Additionally, > as the code is already not that small, errors have an easy way to slip > in without noticing. > So to me, the ideal solution seems to add a testing system to the > compiler, like most of Wine's components already have. > > Rémi and I already outlined some ideas: > - Port the Winetest system to Unix programs > - Make a libwidl or Widl.exe and test against that with our current test system > - Add the tests into Widl itself as callable parameter > > What do you think is the best solution? > Getting some suggestions would be great. :) There are options. First it depends whether you want to validate against midl or not. If you do then you'll need something like widl.exe to run on Windows. Next question is what your test data would be, and how to compare the output reliably with expected output. For example we don't really need exact matching for generated headers, or even marshaling format strings, as long as they achieve their purpose. For typelibs however, it might be useful to go for exact binary match of selected sections. If testing against midl is not a goal, we could have tests for widl itself, even in form of scripts, to run on host system, and not through wine. This way you'll need expected test data that you trust, and maintain in a good shape without validating it on Windows. Even if it doesn't match midl output exactly, it's still potentially useful to catch regressions. > > Thanks, > Bernhard >

June 15, 2022

Re: [PATCH v3 0/1] MR134: ntdll: Don't hard-code the battery and AC adapter names on Linux - approved

by Huw Davies (＠huw)

This merge request was approved by Huw Davies. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/134

June 15, 2022

Re: [PATCH 0/1] MR249: include: Emit CFI directives when building on Mac with gcc.

by Stefan Dösinger

Maybe, but I am sure this has been like that since forever. The || APPLE branch here is old. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/249#note_2127

June 15, 2022

Re: [PATCH] comctl32: Add partial support for LVM_SETBKIMAGE.

by Dmitry Timoshkov

Zhiyi Zhang <zzhang(a)codeweavers.com> wrote: > >>>>>>>>> Also, I found LVM_SETBKIMAGE is acting a bit weird on Windows. For example, > >>>>>>>>> if a LVM_SETBKIMAGE message trying to set a bitmap > >>>>>>>>> is sent the second time. The second message returns FALSE and the background > >>>>>>>>> image gets reset to none(from LVM_GETBKIMAGE). I don't know if we should > >>>>>>>>> implement such a behavior but I would be more comfortable with some tests. > >>>>>>>> According to my testing LVM_SETBKIMAGE always returns 0. > >>>>>>> If LVM_SETBKIMAGE for a bitmap succeeds, it does return TRUE. > >>>>>> Where do you see this? In my tests LVM_SETBKIMAGE always returns 0: > >>>>>> https://testbot.winehq.org/JobDetails.pl?Key=116659 > >>>>> You're only testing with comctl32 v5. Please see https://testbot.winehq.org/JobDetails.pl?Key=116666&f204=exe32.report#k204 > >>>> Thanks. Do you think it's really necessary to change return value for the v6 case? > >>>> > >>> Yes, I think it's necessary. I don't think comctl32 v5 support LVM_SETBKIMAGE at all. The FALSE is probably from DefWindowProc(). > >> I don't think that explains destroying the bitmap on the 2nd call. It looks > >> like v5 really supports LVM_SETBKIMAGE, and the only difference in v6 is > >> the return value. > > Is there anything else that should be addressed in this patch? > > > I see. No. I sent a v4 https://source.winehq.org/patches/data/238807. Thanks! -- Dmitry.

June 15, 2022

Re: Gitlab transition roadmap

by Esme Povirk (she/they)

On Wed, Jun 15, 2022 at 7:08 AM Alexandre Julliard <julliard(a)winehq.org> wrote: > - wine-mono I've been thinking about this and my primary requirement is CI. I want a test run for MRs and pushes to the develop branch, at least on Linux, ideally on Windows too. I'd probably convert my github repos to mirrors of winehq once we're ported over, and I'd use my own personal repos on winehq gitlab for WIP stuff and my own MRs. I'd still want CI on the github mirror as an extra check and to make sure macOS continues to work. I would probably continue to accept github PRs. There are also a lot of submodules that need to be ported over. Some of them point to upstreams, should we mirror those on winehq just in case?

June 15, 2022

Re: Test system for the Widl compiler

by Connor McAdams

On Wed, Jun 15, 2022 at 05:20:00PM +0200, Bernhard Kölbl wrote: > Hey everyone, > > during the course of implementing WinRT / UWP stuff, I regularly come > across missing IDL 3.0 features in our Widl compiler. > Unfortunately, Widl's code seems to have grown in a rater unpleasant > direction and is sometimes a real burden to understand and maintain. I > think some refactoring to its code would be very well fitting and make > things a lot easier for everyone, but refactoring can also bring > breakage and makes tracking down errors with Git harder. Additionally, > as the code is already not that small, errors have an easy way to slip > in without noticing. > So to me, the ideal solution seems to add a testing system to the > compiler, like most of Wine's components already have. > > Rémi and I already outlined some ideas: > - Port the Winetest system to Unix programs > - Make a libwidl or Widl.exe and test against that with our current test system > - Add the tests into Widl itself as callable parameter > > What do you think is the best solution? > Getting some suggestions would be great. :) > > Thanks, > Bernhard > I don't have any suggestions or anything, just wanted to say I think this is a good idea as well and something I've thought about in the past. There have been a few times where I've wanted to touch widl and been a bit scared off of it due to fear of breaking a bunch of things by accident. Any sort of test system would be great to prevent these kinds of issues. It'd also help when adding support for new features. Sorry for not having any suggestions, just wanted to voice my support for something like this. :)

June 15, 2022

[tools 6/6] testbot/UpdateFailures: Add a script to refresh the failures' bug information.

by Francois Gouget

This queries Bugzilla to update the failure bug status and description fields. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=48912 Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- This requires updating the TestBot crontab. --- testbot/bin/UpdateFailures.pl | 192 ++++++++++++++++++++++++++++++++++ testbot/doc/INSTALL.txt | 5 + 2 files changed, 197 insertions(+) create mode 100755 testbot/bin/UpdateFailures.pl diff --git a/testbot/bin/UpdateFailures.pl b/testbot/bin/UpdateFailures.pl new file mode 100755 index 000000000..7c89b7b99 --- /dev/null +++ b/testbot/bin/UpdateFailures.pl @@ -0,0 +1,192 @@ +#!/usr/bin/perl -Tw +# -*- Mode: Perl; perl-indent-level: 2; indent-tabs-mode: nil -*- +# +# This script updates the failures bug information. +# +# Copyright 2022 Francois Gouget +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, write to the Free Software +# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301, USA + +use strict; + +sub BEGIN +{ + if ($0 !~ m=^/=) + { + # Turn $0 into an absolute path so it can safely be used in @INC + require Cwd; + $0 = Cwd::cwd() . "/$0"; + } + if ($0 =~ m=^(/.*)/[^/]+/[^/]+$=) + { + $::RootDir = $1; + unshift @INC, "$::RootDir/lib"; + } +} +my $Name0 = $0; +$Name0 =~ s+^.*/++; + +my $WineBugUrl = "https://bugs.winehq.org/"; + +use Text::CSV::Encoded; + +use ObjectModel::Collection; +use WineTestBot::Failures; +use WineTestBot::Log; + + +# +# Logging and error handling helpers +# + +my $Debug; +sub Debug(@) +{ + print STDERR @_ if ($Debug); +} + +my $LogOnly; +sub Error(@) +{ + print STDERR "$Name0:error: ", @_ if (!$LogOnly); + LogMsg @_; +} + + +# +# Bugzilla interface +# + +sub CollectBugInfo($) +{ + my ($BugIds) = @_; + + my $Bugs = {}; + my $Cmd = "wget -qO- '$WineBugUrl/buglist.cgi?bug_id=". join(",", @$BugIds) + ."&columnlist=bug_status,resolution,short_desc&ctype=csv'"; + if (open(my $Fh, "-|", $Cmd)) + { + <$Fh>; # skip the header line + + my $Csv = Text::CSV::Encoded->new({ binary => 1 }); + while (my $Line = <$Fh>) + { + if ($Csv->parse($Line)) + { + my ($BugId, $Status, $Resolution, $Description) = $Csv->fields(); + $Bugs->{$BugId} = { + Status => $Status, + Resolution => $Resolution, + Description => $Description, + }; + Debug("Bug $BugId - $Status - $Resolution - $Description\n"); + } + else + { + Error("could not parse line: ", $Csv->error_input, "\n"); + } + } + close($Fh); + } + return $Bugs; +} + + +# +# Main +# + +my (@FailureIds, $Usage); +while (@ARGV) +{ + my $Arg = shift @ARGV; + if ($Arg eq "--log-only") + { + $LogOnly = 1; + } + elsif ($Arg eq "--debug") + { + $Debug = 1; + } + elsif ($Arg =~ /^(?:-\?|-h|--help)$/) + { + $Usage = 0; + last; + } + elsif ($Arg =~ /^([0-9]+)$/) + { + push @FailureIds, $1; + } + else + { + Error "unexpected argument '$Arg'\n"; + $Usage = 2; + last; + } +} +# Check parameters +if (defined $Usage) +{ + print "Usage: $Name0 [--log-only] [--debug] [--help] [FID...]\n"; + exit $Usage; +} + +my $Failures = CreateFailures(); +(a)FailureIds = @{$Failures->GetKeys()} if (!@FailureIds); +(a)FailureIds = sort { $a <=> $b } @FailureIds; + +# Deduplicate the bug ids +my $BugIds; +foreach my $FailureId (@FailureIds) +{ + my $Failure = $Failures->GetItem($FailureId); + $BugIds->{$Failure->BugId} = 1; +} + +# Update the failures bug information +my $Bugs = CollectBugInfo([keys %$BugIds]); +foreach my $FailureId (@FailureIds) +{ + my $Failure = $Failures->GetItem($FailureId); + my $Bug = $Bugs->{$Failure->BugId}; + if ($Bug) + { + if ($Failure->BugStatus ne "deleted") + { + my $Resolution = $Bug->{Resolution} || " ---"; + $Resolution = $Resolution eq " ---" ? "" : " $Resolution"; + $Failure->BugStatus("$Bug->{Status}$Resolution"); + } + $Failure->BugDescription($Bug->{Description}); + } + else + { + $Failure->BugStatus("Does not exist"); + $Failure->BugDescription("The WineHQ bug ". $Failure->BugId ." does not exist!"); + } + if ($Debug and $Failure->GetIsModified()) + { + Debug("Failure $FailureId - ", $Failure->BugStatus, " - ", $Failure->BugDescription, "\n"); + } +} + +my ($ErrKey, $ErrProperty, $ErrMessage) = $Failures->Save(); +if ($ErrMessage) +{ + Error("Could not save $ErrKey because of the $ErrProperty field: $ErrMessage\n"); + exit 1; +} + +exit 0; diff --git a/testbot/doc/INSTALL.txt b/testbot/doc/INSTALL.txt index 46edeea52..fac4580b2 100644 --- a/testbot/doc/INSTALL.txt +++ b/testbot/doc/INSTALL.txt @@ -81,6 +81,11 @@ Setup for Winetest updates: instance: */10 * * * * $HOME/tools/testbot/bin/CheckForWinetestUpdate.pl --log-only +Setup for failure bug information updates: +- Use a cron job to run UpdateFailures.pl periodically. For + instance: + 0 */4 * * * $HOME/tools/testbot/bin/UpdateFailures.pl --log-only + Setup for wine-devel: - A proper WineTestBot system should integrate with Wine's patches site. An alternative is to interface directly with wine-devel as described in -- 2.30.2

June 15, 2022

[tools 5/6] testbot/WineSendLog: Take into account the known failures.

by Francois Gouget

Don't count known failures as new failures, in particular when updating the patches site. Add links to the WineHQ bugs for known failures found in the logs. And identify the bug each known failure line corresponds to. Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=48912 Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/bin/WineSendLog.pl | 78 ++++++++++++++++++++++++++------------ 1 file changed, 54 insertions(+), 24 deletions(-) diff --git a/testbot/bin/WineSendLog.pl b/testbot/bin/WineSendLog.pl index 3897ac034..a2f2369b6 100755 --- a/testbot/bin/WineSendLog.pl +++ b/testbot/bin/WineSendLog.pl @@ -44,6 +44,7 @@ $Name0 =~ s+^.*/++; use Algorithm::Diff; use WineTestBot::Config; +use WineTestBot::Failures; use WineTestBot::Jobs; use WineTestBot::Log; use WineTestBot::LogUtils; @@ -177,10 +178,11 @@ sub SendLog($) # Collect and count the new and old failures, etc. # - my (@New, $JobInfo); - my ($NewCount, $OldCount, $BotCount, $NotRun) = (0, 0, 0, 0); + my ($JobInfo, @New, %BugDescriptions); + my ($NewCount, $FailCount, $OldCount, $BotCount, $NotRun) = (0, 0, 0, 0, 0); my $FileType = "test executable"; + my $Failures = CreateFailures($Job); my $StepsTasks = CreateStepsTasks(undef, $Job); my $SortedStepsTasks = $StepsTasks->GetSortedItems(); foreach my $StepTask (@$SortedStepsTasks) @@ -193,12 +195,13 @@ sub SendLog($) my $TaskDir = $StepTask->GetTaskDir(); my $LogNames = GetLogFileNames($TaskDir); - $JobInfo->{$StepTask->Id}->{LogNames} = $LogNames; - $JobInfo->{$StepTask->Id}->{NewCount} = 0; - $JobInfo->{$StepTask->Id}->{OldCount} = 0; + my $TaskInfo = $JobInfo->{$StepTask->Id} = {}; + $TaskInfo->{LogNames} = $LogNames; + $TaskInfo->{NewCount} = 0; + $TaskInfo->{OldCount} = 0; if ($StepTask->Status eq "boterror") { - $JobInfo->{$StepTask->Id}->{BotCount} = 1; + $TaskInfo->{BotCount} = 1; $BotCount++; } $FileType = "patch" if ($StepTask->FileType eq "patch"); @@ -207,7 +210,7 @@ sub SendLog($) { my $LogInfo = LoadLogErrors("$TaskDir/$LogName"); next if (!defined $LogInfo->{BadLog} and !$LogInfo->{ErrCount}); - $JobInfo->{$StepTask->Id}->{$LogName} = $LogInfo; + $TaskInfo->{$LogName} = $LogInfo; my $HasLogHeader; foreach my $GroupName (@{$LogInfo->{ErrGroupNames}}) @@ -216,7 +219,21 @@ sub SendLog($) my $Group = $LogInfo->{ErrGroups}->{$GroupName}; foreach my $ErrIndex (0..$#{$Group->{Errors}}) { - if ($Group->{IsNew}->[$ErrIndex]) + my $LineNo = $Group->{LineNos}->[$ErrIndex]; + my $LineFailureIds = $Group->{Failures}->{$ErrIndex}; + if ($LineFailureIds) + { + foreach my $FailureId (@$LineFailureIds) + { + my $Failure = $Failures->GetItem($FailureId); + next if (!$Failure); + $BugDescriptions{$Failure->BugId} = $Failure->BugDescription; + # Add bug information to $LogInfo + push @{$LogInfo->{Bugs}->{$LineNo}}, $Failure->BugId; + } + $TaskInfo->{OldCount}++; + } + elsif ($Group->{IsNew}->[$ErrIndex]) { if (!$HasLogHeader) { @@ -229,16 +246,16 @@ sub SendLog($) $HasGroupHeader = 1; } push @New, "$Group->{Errors}->[$ErrIndex]\n"; - $JobInfo->{$StepTask->Id}->{NewCount}++; + $TaskInfo->{NewCount}++; } else { - $JobInfo->{$StepTask->Id}->{OldCount}++; + $TaskInfo->{OldCount}++; } } } - $NewCount += $JobInfo->{$StepTask->Id}->{NewCount}; - $OldCount += $JobInfo->{$StepTask->Id}->{OldCount}; + $NewCount += $TaskInfo->{NewCount}; + $OldCount += $TaskInfo->{OldCount}; } } @@ -281,10 +298,9 @@ sub SendLog($) { print $Sendmail <<"EOF"; It looks like your $FileType introduces some new failures. Please -investigate and fix them if they are indeed new. Note that rare -failures and failures with always changing text (e.g. because of memory -addresses) can cause false positives. If this is what happened, then -fixing those would really help. +investigate and fix them if they are indeed new. If they are not new, +fixing them anyway would help a lot. Otherwise please ask for the known +failures list to be updated. EOF } @@ -309,10 +325,21 @@ EOF if ($OldCount) { print $Sendmail <<"EOF"; -Some preexisting failures (not caused by your $FileType) happened. +There are some preexisting failures (not caused by your $FileType). If you know how to fix them that would be helpful. - EOF + if (%BugDescriptions) + { + print $Sendmail <<"EOF"; +In particular some failures are tracked in the bug(s) below: +EOF + foreach my $BugId (sort { $a <=> $b } keys %BugDescriptions) + { + print $Sendmail "* $BugId - $BugDescriptions{$BugId}\n"; + print $Sendmail " $WineBugUrl$BugId\n"; + } + print $Sendmail "\n"; + } } print $Sendmail " Failures\n"; @@ -360,7 +387,7 @@ EOF my $TaskDir = $StepTask->GetTaskDir(); foreach my $LogName (@{$TaskInfo->{LogNames}}) { - my $LogInfo = LoadLogErrors("$TaskDir/$LogName"); + my $LogInfo = $TaskInfo->{$LogName}; next if (!defined $LogInfo->{BadLog} and !$LogInfo->{ErrCount}); print $Sendmail "\n=== ", GetTitle($StepTask, $LogName), " ===\n"; @@ -372,7 +399,11 @@ EOF my $Group = $LogInfo->{ErrGroups}->{$GroupName}; foreach my $ErrIndex (0..$#{$Group->{Errors}}) { - my $Prefix = $Group->{IsNew}->[$ErrIndex] ? "new" : "old"; + my $LineNo = $Group->{LineNos}->[$ErrIndex]; + my $Bugs = $LogInfo->{Bugs}->{$LineNo}; + my $Prefix = $Bugs ? join(" ", sort { $a <=> $b } @$Bugs) : + $Group->{IsNew}->[$ErrIndex] ? "new" : + "old"; print $Sendmail "[$Prefix] $Group->{Errors}->[$ErrIndex]\n"; } } @@ -433,10 +464,9 @@ EOF { print $Sendmail <<"EOF"; It looks like your $FileType introduced the new failures shown below. -Please investigate and fix them if they are indeed new. Note that rare -failures and failures with always changing text (e.g. because of memory -addresses) can cause false positives. If this is what happened, then -fixing those would really help. +Please investigate and fix them before resubmitting your $FileType. +If they are not new, fixing them anyway would help a lot. Otherwise +please ask for the known failures list to be updated. EOF } -- 2.30.2

June 15, 2022