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

Re: [PATCH v5 1/7] mfplat/buffer: Fix size for IMC2/IMC4 buffers with odd height.

by Nikolay Sivov

Signed-off-by: Nikolay Sivov <nsivov(a)codeweavers.com>

June 27, 2022

[tools] testbot/web: Use GetErrMessage() instead of direct field access.

by Francois Gouget

Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/web/Stats.pl | 2 +- testbot/web/admin/Log.pl | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/testbot/web/Stats.pl b/testbot/web/Stats.pl index 773fe3bb99..ec2eaba7e2 100644 --- a/testbot/web/Stats.pl +++ b/testbot/web/Stats.pl @@ -49,7 +49,7 @@ sub _initialize($$$) if (!$self->GetParam("Days") or !$self->Validate() or !int($self->GetParam("Days"))) # 00 case! { - $self->SetParam("Days", $DAYS_DEFAULT) if (!defined $self->{ErrMessage}); + $self->SetParam("Days", $DAYS_DEFAULT) if (!defined $self->GetErrMessage()); $self->{NoStats} = 1; } # The action is a no-op so unset it to not need to redefine OnAction() diff --git a/testbot/web/admin/Log.pl b/testbot/web/admin/Log.pl index 9c18c375e2..74711df5f3 100644 --- a/testbot/web/admin/Log.pl +++ b/testbot/web/admin/Log.pl @@ -43,7 +43,7 @@ sub _initialize($$$) if (!$self->GetParam("Hours") or !$self->Validate() or !int($self->GetParam("Hours"))) # 00 case! { - $self->SetParam("Hours", $HOURS_DEFAULT) if (!defined $self->{ErrMessage}); + $self->SetParam("Hours", $HOURS_DEFAULT) if (!defined $self->GetErrMessage()); $self->SetParam("Action", undef); } } -- 2.30.2

June 27, 2022

[tools] testbot/web: Show the VM name for failure matches.

by Francois Gouget

Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/web/FailureDetails.pl | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/testbot/web/FailureDetails.pl b/testbot/web/FailureDetails.pl index ae4ccb4041..02803ed27a 100644 --- a/testbot/web/FailureDetails.pl +++ b/testbot/web/FailureDetails.pl @@ -45,7 +45,9 @@ sub GetPropertyDescriptors($) push @PropertyDescriptors, $PropertyDescriptor; if ($PropertyName eq "TaskLog") { - push @PropertyDescriptors, CreateBasicPropertyDescriptor("Date", "Date", !1, !1, "DT", 19); + push @PropertyDescriptors, + CreateBasicPropertyDescriptor("VM", "VM", !1, !1, "A", 20), + CreateBasicPropertyDescriptor("Date", "Date", !1, !1, "DT", 19); } } push @PropertyDescriptors, CreateBasicPropertyDescriptor("Remarks", "Remarks", !1, !1, "A", 128); @@ -89,6 +91,10 @@ sub GenerateDataView($$$) print "<a href='", $self->escapeHTML($URL), "'>", $self->escapeHTML($Row->{Item}->TaskLog), "</a>"; } + elsif ($PropertyName eq "VM") + { + print $self->escapeHTML($Row->{Item}->Task->VMName); + } elsif ($PropertyName eq "Date") { # The enclosing page already pulled in the datetime JavaScript code -- 2.30.2

June 27, 2022

[tools] testbot: Better track VM status changes.

by Francois Gouget

Only set OldStatus once to avoid issues in case the status is changed multiple times before the VM is saved. Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- This matches the recent job status tracking change. --- testbot/lib/WineTestBot/VMs.pm | 21 ++++----------------- 1 file changed, 4 insertions(+), 17 deletions(-) diff --git a/testbot/lib/WineTestBot/VMs.pm b/testbot/lib/WineTestBot/VMs.pm index f0c16d19ec..2f736c636e 100644 --- a/testbot/lib/WineTestBot/VMs.pm +++ b/testbot/lib/WineTestBot/VMs.pm @@ -157,15 +157,6 @@ use WineTestBot::TestAgent; use WineTestBot::Utils; -sub _initialize($$) -{ - my ($self, $VMs) = @_; - - $self->SUPER::_initialize($VMs); - - $self->{OldStatus} = undef; -} - sub InitializeNew($$) { my ($self, $Collection) = @_; @@ -270,8 +261,8 @@ sub Status($;$) if (!defined $CurrentStatus or $NewStatus ne $CurrentStatus) { + $self->{OldStatus} ||= $CurrentStatus; $self->SUPER::Status($NewStatus); - $self->{OldStatus} = $CurrentStatus; } return $NewStatus; @@ -424,15 +415,11 @@ sub OnSaved($) my ($self) = @_; $self->SUPER::OnSaved(); - - if (defined($self->{OldStatus})) + if (defined $self->{OldStatus} and $self->{OldStatus} ne $self->Status) { - my $NewStatus = $self->Status; - if ($NewStatus ne $self->{OldStatus}) - { - VMStatusChange($self->GetKey(), $self->{OldStatus}, $NewStatus); - } + VMStatusChange($self->GetKey(), $self->{OldStatus}, $self->Status); } + delete $self->{OldStatus}; } sub Run($$$$$$) -- 2.30.2

June 27, 2022

[tools] testbot/Janitor: Clean up obsolete failures.

by Francois Gouget

Delete failures that have been marked for deletion once they are no longer referenced. Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/bin/Janitor.pl | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/testbot/bin/Janitor.pl b/testbot/bin/Janitor.pl index 618c31d022..839e1a942c 100755 --- a/testbot/bin/Janitor.pl +++ b/testbot/bin/Janitor.pl @@ -43,6 +43,7 @@ $Name0 =~ s+^.*/++; use File::Path; use WineTestBot::Config; +use WineTestBot::Failures; use WineTestBot::Jobs; use WineTestBot::Log; use WineTestBot::Patches; @@ -193,6 +194,22 @@ if ($JobPurgeDays != 0) } } +# And also failures marked for deletion +my $Failures = CreateFailures(); +$Failures->AddFilter("BugStatus", ["deleted"]); +foreach my $Failure (@{$Failures->GetItems()}) +{ + if ($Failure->TaskFailures->IsEmpty()) + { + $Failures->DeleteItem($Failure); + } + else + { + my $TaskFailure = @{$Failure->TaskFailures->GetItems()}[0]; + Trace "Keeping failure ", $Failure->Id, " for task log ", join("/", $TaskFailure->JobId, $TaskFailure->StepNo, $TaskFailure->TaskNo, $TaskFailure->TaskLog), "\n"; + } +} + # Purge the deleted VMs if they are not referenced anymore my $VMs = CreateVMs(); $VMs->AddFilter("Role", ["deleted"]); -- 2.30.2

June 27, 2022

Re: [PATCH 0/1] MR315: windowscodecs: Fix non-zero alpha detection in ImagingFactory_CreateBitmapFromHICON.

by Esme Povirk (＠madewokherd)

This didn't initially look right to me because the loop is written in a confusing way. We should always change x in the inner loop. I think it's probably correct but I'd like to see it cleaned up a bit more. Since we don't actually care about the x and y coordinates, I would suggest a single loop over `width * height` pixels. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/315#note_2801

June 27, 2022

Re: [PATCH v6 0/5] MR310: mf: Fix several reference leaks.

by Rémi Bernon

v6: Restore temporary todo_wine in PATCH 2, lost in the split. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/310#note_2800

June 27, 2022

Re: HLSL offsetting

by Francisco Casas

Hello, On 25-06-22 13:29, Zebediah Figura wrote: > On 6/24/22 23:46, Francisco Casas wrote: >> Hello, >> >> I attach the third version of the previous patch, applying the >> suggestions made by Zeb, and some other changes that arose from them. >> >> The patch is based on the master branch, but I guess I will have to >> update it if the recent patches from Giovanni are accepted, before >> sending it for review. >> > > Is it not possible to split up this patch, as I had asked earlier? > I think I can separate some small parts, like hlsl_free_deref() as you suggested, but we would still have a large patch that's hard to split. > There's also a couple things I notice from skimming: > > * I think add_load() should also be split into deref / component > versions. Most of its users only need to count components. > add_record_load() and add_array_load() will require indexes, but yes, add_matrix_scalar_load(), intrinsic_mul(), add_expr(), and add_cast(), in particular the parts where they deal with matrices, could be simplified if we use component offsets. I will split the function then. > * Actually, should we be passing a deref instead of just a var to > hlsl_new_component_load()? See e.g. add_cast(). That would make it a > simple wrapper around hlsl_new_load_from_deref() [and the names would > also need to be changed...] I'm not sure why I thought it made sense > like this, frankly. > Allowing to pass a deref instead of just a var to hlsl_new_component_load() seems like a good idea. It may save us from creating synthetic variables e.g. in initialize_var_components(). I don't think it can be just made a simple wrapper around hlsl_new_load_from_deref() though, since a single component within a struct may require to be accessed with path of larger length than 2. It requires a little more complex implementation, but it can be done. > * Do we need to always split up matrices into scalars? Should we make > non-contiguous loads a problem for parse time? As long as we're trying > to keep backend-specific details out of the HLSL IR, that seems > potentially reasonable; it'd just require emitting multiple copy > instructions for some cases. That would allow us to simplify > split_copy() and add_matrix_scalar_load(), and I believe would thus also > get rid of the need to pass multiple offsets to hlsl_new_load_from_deref(). If we go back to make non-contiguous loads a problem for parse time, then I think it would make sense for row_major matrices to retrieve a row, and column_matrices to retrieve a column when indexing with a path. I will see if I can do it while keeping the component offsets consistent with the order in the initializers. Best regards, Francisco.

June 27, 2022

Re: [PATCH v5 3/5] mf: Avoid leaking nodes in IMFTopoLoader_Load.

by Marvin

Hi, It looks like your patch introduced the new failures shown below. Please investigate and fix them before resubmitting your patch. If they are not new, fixing them anyway would help a lot. Otherwise please ask for the known failures list to be updated. The tests also ran into some preexisting test failures. If you know how to fix them that would be helpful. See the TestBot job for the details: The full results can be found at: https://testbot.winehq.org/JobDetails.pl?Key=117719 Your paranoid android. === debian11 (32 bit report) === Report validation errors: mf:mf prints too much data (39971 bytes) === debian11 (build log) === 0100:err:winediag:wma_decoder_create GStreamer doesn't support WMA decoding, please install appropriate plugins 0100:err:winediag:h264_decoder_create GStreamer doesn't support H.264 decoding, please install appropriate plugins 0100:err:winediag:color_convert_create GStreamer doesn't support video conversion, please install appropriate plugins. === debian11 (32 bit WoW report) === Report validation errors: mf:mf prints too much data (40401 bytes)

June 27, 2022

Re: [PATCH v5 2/5] mf/tests: Add some missing IMFActivate_Shutdown calls.

by Marvin

Hi, It looks like your patch introduced the new failures shown below. Please investigate and fix them before resubmitting your patch. If they are not new, fixing them anyway would help a lot. Otherwise please ask for the known failures list to be updated. The tests also ran into some preexisting test failures. If you know how to fix them that would be helpful. See the TestBot job for the details: The full results can be found at: https://testbot.winehq.org/JobDetails.pl?Key=117718 Your paranoid android. === w864 (64 bit report) === mf: mf.c:4767: Test failed: Unexpected hr 0x800401f0. === debian11 (32 bit report) === mf: mf.c:2559: Test failed: Release returned 2 Report validation errors: mf:mf prints too much data (40357 bytes) === debian11 (32 bit Arabic:Morocco report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit German report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit French report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit Hebrew:Israel report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit Hindi:India report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit Japanese:Japan report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit Chinese:China report) === mf: mf.c:2559: Test failed: Release returned 2 === debian11 (32 bit WoW report) === mf: mf.c:2559: Test failed: Release returned 2 Report validation errors: mf:mf prints too much data (40875 bytes) === debian11 (64 bit WoW report) === mf: mf.c:2559: Test failed: Release returned 2

June 27, 2022