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

[PATCH] vbscript: Implement IsDate().

by Robert Wilhelm

Signed-off-by: Robert Wilhelm <robert.wilhelm(a)gmx.net> --- dlls/vbscript/global.c | 31 +++++++++++++++++++++++++++++-- dlls/vbscript/tests/api.vbs | 13 +++++++++++++ 2 files changed, 42 insertions(+), 2 deletions(-) diff --git a/dlls/vbscript/global.c b/dlls/vbscript/global.c index 90cfd557038..222e0534665 100644 --- a/dlls/vbscript/global.c +++ b/dlls/vbscript/global.c @@ -425,6 +425,25 @@ HRESULT to_int(VARIANT *v, int *ret) return S_OK; } +static HRESULT to_date(VARIANT *v, DATE *ret) +{ + VARIANT r; + HRESULT hres; + + V_VT(&r) = VT_EMPTY; + hres = VariantChangeType(&r, v, 0, VT_DATE); + if(FAILED(hres)) { + hres = VariantChangeType(&r, v, 0, VT_R8); + if(FAILED(hres)) + return hres; + hres = VariantChangeType(&r, &r, 0, VT_DATE); + if(FAILED(hres)) + return hres; + } + *ret = V_DATE(&r); + return S_OK; +} + static HRESULT to_double(VARIANT *v, double *ret) { VARIANT dst; @@ -935,8 +954,16 @@ static HRESULT Global_VarType(BuiltinDisp *This, VARIANT *arg, unsigned args_cnt static HRESULT Global_IsDate(BuiltinDisp *This, VARIANT *arg, unsigned args_cnt, VARIANT *res) { - FIXME("\n"); - return E_NOTIMPL; + HRESULT hres; + DATE d; + + TRACE("%s\n", debugstr_variant(arg)); + + assert(args_cnt == 1); + + hres = to_date(arg, &d); + + return return_bool(res, SUCCEEDED(hres)); } static HRESULT Global_IsEmpty(BuiltinDisp *This, VARIANT *arg, unsigned args_cnt, VARIANT *res) diff --git a/dlls/vbscript/tests/api.vbs b/dlls/vbscript/tests/api.vbs index be1d4a806c0..d70f95ce69b 100644 --- a/dlls/vbscript/tests/api.vbs +++ b/dlls/vbscript/tests/api.vbs @@ -181,6 +181,19 @@ Call ok(not isObject(4), "isObject(4) is true?") Call ok(not isObject("x"), "isObject(""x"") is true?") Call ok(not isObject(Null), "isObject(Null) is true?") +Call ok(not isDate(new EmptyClass), "isDate(new EmptyClass) is true?") +Set x = new EmptyClass +Call ok(not isDate(x), "isDate(x) is true?") +x = CDate(1) +Call ok(isDate(x), "isDate(x) is not true?") +Call ok(isDate(empty), "isDate(empty) is not true?") +Call ok(not isDate(Nothing), "isDate(Nothing) is true?") +Call ok(isDate(true), "isDate(true) is not true?") +Call ok(isDate(4), "isDate(4) is not true?") +Call ok(isDate("4"), "isDate(""4"") is not true?") +Call ok(not isDate("x"), "isDate(""x"") is true?") +Call ok(not isDate(Null), "isDate(Null) is true?") + Call ok(not isEmpty(new EmptyClass), "isEmpty(new EmptyClass) is true?") Set x = new EmptyClass Call ok(not isEmpty(x), "isEmpty(x) is true?") -- 2.36.1

June 3, 2022

Re: [PATCH v2 0/5] MR131: wined3d: Introduce wined3d_device_check_format_support().

by Chip Davis (＠cdavis5e)

On Fri Jun 3 16:30:23 2022 +0000, Jan Sikorski wrote: > For the record, I agree on the first point, to me it's unclear what the > motivation is behind this move. Mostly, it's because my change to implement logic op is on top of these changes, because `D3D11_FORMAT_SUPPORT2_OM_LOGIC_OP` is in the second format support group. This also felt cleaner, but I concede your point. But in order to do this from `d3d11`, we would need to make the format attribute and capability flags part of the API anyway. Do you have a better idea? -- https://gitlab.winehq.org/wine/wine/-/merge_requests/131#note_1566

June 3, 2022

Re: [PATCH 1/1] mscoree: Update Wine Mono to 7.3.0.

by Marvin

Hi, While running your changed tests, I think I found new failures. Being a bot and all I'm not very good at pattern recognition, so I might be wrong, but could you please double-check? Full results can be found at: https://testbot.winehq.org/JobDetails.pl?Key=116193 Your paranoid android. === debian11 (build log) === WineRunWineTest.pl:error: The task timed out === debian11 (build log) === Task: Could not create the wow32 wineprefix: Failed to disable the crash dialogs: Task: WineTest did not produce the wow32 report

June 3, 2022

Re: [PATCH 0/9] MR180: misc: Fixes and workarounds for UAF warnings. (GCC12)

by Hans Leidekker (＠hans)

On Fri Jun 3 15:57:12 2022 +0000, Rémi Bernon wrote: > I suppose you mean moved below the problematic `IHTMLStorage_getItem` > test maybe? I wasn't sure if that second test intended to use key on > purpose, but on a second look it seems to intend to test `NULL` out parameter. You're right, 'key' should be freed after that test. -- https://gitlab.winehq.org/wine/wine/-/merge_requests/180#note_1565

June 3, 2022

[PATCH 1/1] mscoree: Update Wine Mono to 7.3.0.

by Esme Povirk

From: Esme Povirk <esme(a)codeweavers.com> --- dlls/appwiz.cpl/addons.c | 4 ++-- dlls/mscoree/mscoree_private.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/dlls/appwiz.cpl/addons.c b/dlls/appwiz.cpl/addons.c index 73f44a80f79..021f5484af9 100644 --- a/dlls/appwiz.cpl/addons.c +++ b/dlls/appwiz.cpl/addons.c @@ -58,10 +58,10 @@ WINE_DEFAULT_DEBUG_CHANNEL(appwizcpl); #define GECKO_SHA "???" #endif -#define MONO_VERSION "7.2.0" +#define MONO_VERSION "7.3.0" #if defined(__i386__) || defined(__x86_64__) #define MONO_ARCH "x86" -#define MONO_SHA "5f06eafbae3a49ecc31dfcea777075bd44f0cc126e4fb4e28ecefa06eb436323" +#define MONO_SHA "939e2f5669720d0d0fc7e305426622a11a33b7ae385c4d7e581e29ea264e22ff" #else #define MONO_ARCH "" #define MONO_SHA "???" diff --git a/dlls/mscoree/mscoree_private.h b/dlls/mscoree/mscoree_private.h index 44b2fd88d2a..0e51e0b3d7f 100644 --- a/dlls/mscoree/mscoree_private.h +++ b/dlls/mscoree/mscoree_private.h @@ -45,7 +45,7 @@ extern HRESULT assembly_get_runtime_version(ASSEMBLY *assembly, LPSTR *version) extern HRESULT assembly_get_vtable_fixups(ASSEMBLY *assembly, VTableFixup **fixups, DWORD *count) DECLSPEC_HIDDEN; extern HRESULT assembly_get_native_entrypoint(ASSEMBLY *assembly, NativeEntryPointFunc *func) DECLSPEC_HIDDEN; -#define WINE_MONO_VERSION "7.2.0" +#define WINE_MONO_VERSION "7.3.0" /* Mono embedding */ typedef struct _MonoDomain MonoDomain; -- GitLab https://gitlab.winehq.org/wine/wine/-/merge_requests/184

June 3, 2022

[PATCH 0/1] MR184: mscoree: Update Wine Mono to 7.3.0.

by Esme (＠madewokherd)

-- https://gitlab.winehq.org/wine/wine/-/merge_requests/184

June 3, 2022

[tools] testbot/orm: Fix multi-column key support in Item::Compare().

by Francois Gouget

Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- I think that so far this issue has been masked by the MasterColumns. --- testbot/lib/ObjectModel/Item.pm | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) diff --git a/testbot/lib/ObjectModel/Item.pm b/testbot/lib/ObjectModel/Item.pm index 7d36ffa38..0b1da92a2 100644 --- a/testbot/lib/ObjectModel/Item.pm +++ b/testbot/lib/ObjectModel/Item.pm @@ -422,22 +422,27 @@ sub Compare($$) { my $ColName = $ColNames->[0]; my $ColType = $PropertyDescriptor->GetType(); - return ($ColType eq "N" or $ColType eq "S" or $ColType eq "DT") ? - $self->{ColValues}{$ColName} <=> $B->{ColValues}{$ColName} : - $self->{ColValues}{$ColName} cmp $B->{ColValues}{$ColName}; + my $Cmp = ($ColType eq "N" or $ColType eq "S" or $ColType eq "DT") ? + $self->{ColValues}{$ColName} <=> $B->{ColValues}{$ColName} : + $self->{ColValues}{$ColName} cmp $B->{ColValues}{$ColName}; + return $Cmp if ($Cmp); } - if ($PropertyDescriptor->GetClass() eq "Enum") + elsif ($PropertyDescriptor->GetClass() eq "Enum") { my $ColName = $ColNames->[0]; - return $self->{ColValues}{$ColName} cmp $B->{ColValues}{$ColName}; - } - - # A Detailref cannot be a key so this is an Itemref - foreach my $ColName (@$ColNames) - { my $Cmp = $self->{ColValues}{$ColName} cmp $B->{ColValues}{$ColName}; return $Cmp if ($Cmp); } + else + { + # A Detailref cannot be a key so this is an Itemref. Note that its + # underlying column types are unknown so treat them as strings. + foreach my $ColName (@$ColNames) + { + my $Cmp = $self->{ColValues}{$ColName} cmp $B->{ColValues}{$ColName}; + return $Cmp if ($Cmp); + } + } } return 0; } -- 2.30.2

June 3, 2022

[tools] testbot/orm: Protect DBIBackend::LoadItem() from invalid keys.

by Francois Gouget

Check that the requested key has the right number of parts to avoid a crash during the SQL execution. Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/lib/ObjectModel/DBIBackEnd.pm | 29 +++++++++++++-------------- 1 file changed, 14 insertions(+), 15 deletions(-) diff --git a/testbot/lib/ObjectModel/DBIBackEnd.pm b/testbot/lib/ObjectModel/DBIBackEnd.pm index 07b838d74..62332c1f6 100644 --- a/testbot/lib/ObjectModel/DBIBackEnd.pm +++ b/testbot/lib/ObjectModel/DBIBackEnd.pm @@ -169,22 +169,19 @@ sub BuildKeyWhere($$$) { my ($self, $PropertyDescriptors, $Where) = @_; + # Faster than join+map+grep + my $PartCount; foreach my $PropertyDescriptor (@{$PropertyDescriptors}) { - if ($PropertyDescriptor->GetIsKey()) + next if (!$PropertyDescriptor->GetIsKey()); + foreach my $ColName (@{$PropertyDescriptor->GetColNames()}) { - foreach my $ColName (@{$PropertyDescriptor->GetColNames()}) - { - if ($Where ne "") - { - $Where .= " AND "; - } - $Where .= "$ColName = ?"; - } + $Where .= " AND " if ($Where ne ""); + $Where .= "$ColName = ?"; + $PartCount++; } } - - return $Where; + return ($PartCount, $Where); } =pod @@ -422,8 +419,10 @@ sub LoadItem($$$) $Where = join(" = ? AND ", @{$MasterColNames}) . " = ?"; push @Data, @{$MasterColValues}; } - $Where = $self->BuildKeyWhere($Collection->GetPropertyDescriptors(), $Where); - push @Data, $Collection->SplitKey($RequestedKey); + (my $PartCount, $Where) = $self->BuildKeyWhere($Collection->GetPropertyDescriptors(), $Where); + my @KeyParts = $Collection->SplitKey($RequestedKey); + return undef if ($PartCount != @KeyParts); + push @Data, @KeyParts; my $Query = "SELECT $Fields FROM " . $Collection->GetTableName(); if ($Where ne "") @@ -572,7 +571,7 @@ sub BuildUpdateStatement($$$$) { $Where = join(" = ? AND ", @{$MasterColNames}) . " = ?"; } - $Where = $self->BuildKeyWhere($PropertyDescriptors, $Where); + (my $_PartCount, $Where) = $self->BuildKeyWhere($PropertyDescriptors, $Where); return "UPDATE $TableName SET $Fields WHERE $Where"; } @@ -713,7 +712,7 @@ sub DeleteItem($$) $Where = join(" = ? AND ", @{$MasterColNames}) . " = ?"; push @Data, @{$MasterColValues}; } - $Where = $self->BuildKeyWhere($Item->GetPropertyDescriptors(), $Where); + (my $_PartCount, $Where) = $self->BuildKeyWhere($Item->GetPropertyDescriptors(), $Where); push @Data, $Item->GetKeyComponents(); my $Statement = $self->GetDb()->prepare("DELETE FROM " . -- 2.30.2

June 3, 2022

[tools] testbot/web: Only link to the VM details page for administrators.

by Francois Gouget

Signed-off-by: Francois Gouget <fgouget(a)codeweavers.com> --- testbot/web/index.pl | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/testbot/web/index.pl b/testbot/web/index.pl index b07b39d22..21ee5fea4 100644 --- a/testbot/web/index.pl +++ b/testbot/web/index.pl @@ -176,6 +176,15 @@ use ObjectModel::CGI::CollectionBlock; our @ISA = qw(ObjectModel::CGI::CollectionBlock); +sub GetDetailsPage($) +{ + my ($self) = @_; + my $Session = $self->{EnclosingPage}->GetCurrentSession(); + my $CurrentUser = $Session->User if (defined $Session); + return (!$CurrentUser or !$CurrentUser->HasRole("admin")) ? undef : + "/admin/VMDetails.pl"; +} + sub DisplayProperty($$) { my ($self, $PropertyDescriptor) = @_; -- 2.30.2

June 3, 2022

Re: Gitlab experiment status

by Alex Henrie

On Fri, Jun 3, 2022 at 8:28 AM Gabriel Ivăncescu <gabrielopcode(a)gmail.com> wrote: > > On 03/06/2022 02:50, Alex Henrie wrote: > > On Fri, May 27, 2022 at 12:03 PM Zebediah Figura > > <zfigura(a)codeweavers.com> wrote: > >> > >> On 5/27/22 10:46, Rémi Bernon wrote: > >>> On 5/27/22 17:30, Zebediah Figura (she/her) wrote: > >>>> On 5/27/22 07:34, Jacek Caban wrote: > >>>>> * I think that a mailing list dominated by a bot does not give an > >>>>> impression welcoming for general discussion. If we're moving forward > >>>>> with this, I think we should move Gitlab bridge to a separated > >>>>> mailing list. We had wine-patches in the past (although in this case > >>>>> it would contain review comments as well). > >>>> > >>>> If our intent is to switch to gitlab because mailing lists are too > >>>> difficult for people to work with, then it seems we shouldn't have any > >>>> discussion at all take place on a mailing list, otherwise it'll be > >>>> missed by the people who are only using gitlab. Not that I know of a > >>>> better way to have such discussion... > >>>> > >>> > >>> > >>> Imho the mailing list still makes sense, for general Wine development > >>> discussion, though the volume may be much smaller than what we currently > >>> have. > >>> > >>> Maybe we don't currently have many discussion either because the mailing > >>> list is felt more as a place for patches, and that discussions are > >>> believed to happen elsewhere, or because discussions go more often > >>> unnoticed under the volume of patches. > >>> > >> > >> I've seen this kind of thought floated several times, and I'm not sure I > >> understand the reasoning. What kind of discussions or replies are you > >> expecting to see that don't currently happen? > > > > This thread itself is an example: I didn't see it until today. GitLab > > allows me to subscribe to the patch discussions that I care about > > rather than getting the deafening torrent of every single message. So, > > for me, resurrecting the wine-patches list as the GitLab bridge while > > leaving process and architecture discussion on wine-devel would be a > > welcome change. Mailing lists also aren't very friendly when a > > newcomer wants to reply to a message that was sent before they were > > subscribed to the list, but GitLab allows newcomers to jump right in. > > > > -Alex > > > > FWIW, I see you use gmail, so you can just use a filter to send all the > mails coming to this list that have [PATCH] in Subject line to a > different tag. Which is the same as having two different mailing lists > and sending them to different tags (or even remove the mails with > [PATCH] which is akin to not subscribing to wine-patches). Thanks for the tip. I have altered my filter for wine-devel to separate mail that includes "patch" in the subject line from mail that does not. However, it's still not ideal because Gmail ignores brackets in the search keyword, and there's no way to make the search case-sensitive either. Perhaps those are good reasons to switch to a different email client or provider, but even so, I don't want to have to write a new filter every time I want to subscribe to a particular patch thread. -Alex

June 3, 2022