September 2008 - Wine-Devel

Re: winhttp: Implement some more options.
by Jacek Caban Sept. 3, 2008

Sept. 3, 2008

Hi Hans, Hans Leidekker wrote: > > +static BOOL session_set_option( object_header_t *hdr, DWORD option, LPVOID buffer, DWORD buflen ) > +{ > + if (hdr->type != WINHTTP_HANDLE_TYPE_SESSION) > + { > + release_object( hdr ); > + set_last_error( ERROR_WINHTTP_INCORRECT_HANDLE_TYPE ); > + return FALSE; > + release_object call looks wrong here. Also, if you use vtbls properly, you shouldn't need to check the type here. Jacek

2 1

Re: [PATCH 4/5] wined3d: Remove glDrawBuffer calls in BltOverride.
by Henri Verbeet Sept. 3, 2008

Sept. 3, 2008

2008/8/31 Alexander Dorofeyev <alexd4(a)inbox.lv>: > > Handled in ActivateContext, must be remains of pre-context management code. > --- > dlls/wined3d/surface.c | 10 ---------- > 1 files changed, 0 insertions(+), 10 deletions(-) > If you're going to remove those glDrawBuffer calls, you should remove the buffer variable completely.

2 3

Re: [9/9] gdiplus: initial GdipFlattenPath implementation
by Alexandre Julliard Sept. 3, 2008

Sept. 3, 2008

Nikolay Sivov <bunglehead(a)gmail.com> writes: > +/* free all nodes including argument */ > +void free_path_list(path_list_node_t *node) > +{ > + path_list_node_t *n = node; > + > + while(!n){ > + GdipFree(node); > + node = n = n->next; > + } This won't do what you want. -- Alexandre Julliard julliard(a)winehq.org

2 1

Re: ntdll: Fix a few failing tests in Vista
by Alexandre Julliard Sept. 3, 2008

Sept. 3, 2008

"James Hawkins" <jhawkins(a)codeweavers.com> writes: > @@ -325,13 +327,21 @@ static void read_file_test(void) > iosb.Information = 0xdeadbeef; > offset.QuadPart = strlen(text) + 2; > status = pNtReadFile( handle, event, apc, &apc_count, &iosb, buffer, 2, &offset, NULL ); > - ok( status == STATUS_END_OF_FILE, "wrong status %x\n", status ); > - ok( U(iosb).Status == 0xdeadbabe, "wrong status %x\n", U(iosb).Status ); > - ok( iosb.Information == 0xdeadbeef, "wrong info %lu\n", iosb.Information ); > - ok( !is_signaled( event ), "event is signaled\n" ); > + ok( status == STATUS_END_OF_FILE || > + status == STATUS_PENDING, /* Vista */ > + "wrong status %x\n", status ); > + ok( U(iosb).Status == 0xdeadbabe || > + U(iosb).Status == STATUS_END_OF_FILE, /* Vista */ > + "wrong status %x\n", U(iosb).Status ); > + ok( iosb.Information == 0xdeadbeef || > + iosb.Information == 0, /* Vista */ > + "wrong info %lu\n", iosb.Information ); > ok( !apc_count, "apc was called\n" ); > SleepEx( 1, TRUE ); /* alertable sleep */ > - ok( !apc_count, "apc was called\n" ); > + if ( is_signaled( event ) ) > + ok( apc_count == 1, "apc was not called\n" ); > + else > + ok( !apc_count, "apc was called\n" ); The results are not independent, they should all depend on whether we got STATUS_PENDING or not. -- Alexandre Julliard julliard(a)winehq.org

1 0

Re: [msxml3/tests 1/2] Use CreateFileA to please win98
by James Hawkins Sept. 3, 2008

Sept. 3, 2008

On Wed, Sep 3, 2008 at 2:54 AM, Paul Vriens <paul.vriens.wine(a)gmail.com> wrote: > Hi, > > CreateFileW is not implemented on win98. This fixes the 2 remaining test > failures for win98. > > (This is a replacement for the patch James sent). > > Changelog > Use CreateFileA to please win98 > Just a nitpick: + static const CHAR testXmlA[] = "test.xml"; You don't really need this variable. Same thing for the next patch. -- James Hawkins

2 1

Re: msxml3: Fix a failing test in win98
by Paul Vriens Sept. 3, 2008

Sept. 3, 2008

James Hawkins wrote: > Hi, > > Changelog: > * Fix a failing test in win98. > > dlls/msxml3/tests/xmlelem.c | 4 +++- > 1 files changed, 3 insertions(+), 1 deletions(-) > > > > ------------------------------------------------------------------------ > > Two things: The comment says "Win98 returns ERROR_URL_NOT_FOUND" but the ok now has a different comment. I don't think this fix is correct as the failure comes most likely from the GetFullPathNameW just above that. (This probably means the first change, commit 558ae52cc464269457833cfb08879550a63a5d13, was wrong as well). -- Cheers, Paul.

2 1

Re: msxml3: Fix two failing tests in win98
by Paul Vriens Sept. 3, 2008

Sept. 3, 2008

James Hawkins wrote: > Hi, > > Changelog: > * Fix two failing tests in win98. > > dlls/msxml3/tests/saxreader.c | 6 ++++-- > 1 files changed, 4 insertions(+), 2 deletions(-) > > > > ------------------------------------------------------------------------ > > I don't think this is the correct fix. It all fails because of the CreateFileW slightly above that. -- Cheers, Paul.

1 0

Common conformance test errors
by Dan Kegel Sept. 3, 2008

Sept. 3, 2008

The file http://kegel.com/wine/common-errors.txt is a list of the most common errors in patchwatcher log files, computed by running egrep ": Test failed: |: Test succeeded inside todo block: |^make\[.*\]: \*\*\* \["| sort | uniq -c | sort -rn on about 500 "make test" runs on a dual core box with a new nvidia card. Here are the top ten: 2597 urlmon:protocol.c:1597: Test failed: Read failed: 8000000a (0 bytes) 927 wininet:http.c:115: Test failed: unexpected status 51 (INTERNET_STATUS_CONNECTION_CLOSED) 927 wininet:http.c:115: Test failed: unexpected status 50 (INTERNET_STATUS_CLOSING_CONNECTION) 927 wininet:http.c:115: Test failed: unexpected status 11 (INTERNET_STATUS_NAME_RESOLVED) 927 wininet:http.c:115: Test failed: unexpected status 10 (INTERNET_STATUS_RESOLVING_NAME) 824 d3d9:device.c:829: Test failed: Screen height is 1024 824 d3d9:device.c:828: Test failed: Screen width is 1280 824 d3d9:device.c:782: Test failed: Screen height is 1024 824 d3d9:device.c:781: Test failed: Screen width is 1280 519 make[1]: *** [user32/tests/__test__] Error 2 So many of the tests fail always that I will probably have to start running the baseline tests fifty times after each git change, to catch all the ones that only happen a small percentage of the time. - Dan

1 0

Re: rpcrt4: Fix a failing test in Vista
by Rob Shearman Sept. 2, 2008

Sept. 2, 2008

2008/9/2 James Hawkins <jhawkins(a)codeweavers.com>: > Hi, > > Changelog: > * Fix a failing test in Vista. > > dlls/rpcrt4/tests/rpc.c | 4 +++- > dlls/schannel/tests/main.c | 4 +++- > 2 files changed, 6 insertions(+), 2 deletions(-) It looks like another fix got mixed in with this commit. Please resend without the schannel change. -- Rob Shearman

1 0

Re: [PATCH 3/5] wined3d: Make the condition to enter p8 conversion shader path more strict.
by Henri Verbeet Sept. 2, 2008

Sept. 2, 2008

2008/8/31 Alexander Dorofeyev <alexd4(a)inbox.lv>: > > This prevents shader path from being entered for an offscreen surface when > there is p8 render target and fixes failures in ddraw visual test (with > opengl rendering and RTL_READDRAW mode) and visual glitches in Red Alert. > --- > dlls/wined3d/surface.c | 6 +++++- > 1 files changed, 5 insertions(+), 1 deletions(-) > This is probably ok for this specific case, but in general I think primary_render_target_is_p8() is somewhat broken. It seems that in most places where it's used it is assumed that "This" is the primary render target. In some cases that's actually always true, but in those cases primary_render_target_is_p8() wouldn't need to check if there's a primary render target, it could just use a surface passed in as parameter. In the other cases, you don't actually know the surface is the primary render target, and you'd need to verify that first, after which you've essentially got the first case again.

2 2