Call for help with Wine 1.0 testing
Hi Folks, One key goal for Wine 1.0 is that all of its conformance tests run successfully on nearly all systems. We would really like your help in figuring out how close we are to that goal. To that end, if you are comfortable with checking Wine out via git, could you please visit this page: http://wiki.winehq.org/MakeTestFailures and follow the instructions there? (It's really simple; build current git Wine, download + run a script). And, if you're a Wine developer, since Alexandre is away and the code freeze is on, why not look for one of those failures in your own make test results and see if you can fix it? Thanks! Jeremy
Jeremy White <jwhite <at> codeweavers.com> writes:
Hi Folks,
One key goal for Wine 1.0 is that all of its conformance tests run successfully on nearly all systems. We would really like your help in figuring out how close we are to that goal.
To that end, if you are comfortable with checking Wine out via git, could you please visit this page: http://wiki.winehq.org/MakeTestFailures and follow the instructions there? (It's really simple; build current git Wine, download + run a script).
And, if you're a Wine developer, since Alexandre is away and the code freeze is on, why not look for one of those failures in your own make test results
and see
if you can fix it?
Thanks!
Jeremy
Unfortunately the test crashes here with: Running: d3d8:volume (60 of 335) fixme:d3d:IWineD3DImpl_FillGLCaps OpenGL implementation supports 32 vertex s fixme:d3d:IWineD3DImpl_FillGLCaps Expected vertex samplers + MAX_TEXTURES(=8 fixme:win:EnumDisplayDevicesW ((null),0,0x33f7cc,0x00000000), stub! Running: d3d9:d3d9ex (61 of 335) fixme:d3d:IWineD3DImpl_FillGLCaps OpenGL implementation supports 32 vertex s amplers and 32 total samplers fixme:d3d:IWineD3DImpl_FillGLCaps Expected vertex samplers + MAX_TEXTURES(=8 ) > combined_samplers fixme:win:EnumDisplayDevicesW ((null),0,0x33f8cc,0x00000000), stub! X Error of failed request: GLXBadDrawable Major opcode of failed request: 143 (GLX) Minor opcode of failed request: 5 (X_GLXMakeCurrent) Serial number of failed request: 263 Current serial number in output stream: 263 Those tests in d3d8/d3d9 used to pass fine here a few weeks ago
On Thu, May 15, 2008 at 10:47 AM, Jeremy White <jwhite(a)codeweavers.com> wrote:
One key goal for Wine 1.0 is that all of its conformance tests run successfully on nearly all systems. We would really like your help in figuring out how close we are to that goal.
To that end, if you are comfortable with checking Wine out via git, could you please visit this page: http://wiki.winehq.org/MakeTestFailures
The results are pouring in at http://test.winehq.org/data/2470b0b31605133ec046330dd79fdccaa7ba33fe/#group_... Say, who maintains that web site? It'd be handy to have an option to suppress rows that have neither crashes nor failures; right now you have to scroll vertically a whole lot to see all the failures. - Dan
Say, who maintains that web site? It'd be handy to have an option to suppress rows that have neither crashes nor failures; right now you have to scroll vertically a whole lot to see all the failures.
I'm not sure. The source is in this git tree: http://source.winehq.org/git/tools.git Cheers, Jeremy
On Thu, May 15, 2008 at 4:52 PM, Jeremy White <jwhite(a)codeweavers.com> wrote:
Say, who maintains that web site? It'd be handy to have an option to suppress rows that have neither crashes nor failures; right now you have to scroll vertically a whole lot to see all the failures.
I'm not sure. The source is in this git tree: http://source.winehq.org/git/tools.git
I wrote a postprocessor to do it. It's at http://kegel.com/wine/skipgood.pl.txt An example of its output is at http://kegel.com/wine/failing.html It really does make it easier to see all the failures.
On Fri, May 16, 2008 at 12:33 AM, Dan Kegel <dank(a)kegel.com> wrote:
On Thu, May 15, 2008 at 4:52 PM, Jeremy White <jwhite(a)codeweavers.com> wrote: I wrote a postprocessor to do it. It's at http://kegel.com/wine/skipgood.pl.txt An example of its output is at http://kegel.com/wine/failing.html It really does make it easier to see all the failures.
Your webserver doesn't like serving your skipgood.pl.txt file =/ --John
On Thu, May 15, 2008 at 10:43 PM, John Klehm <xixsimplicityxix(a)gmail.com> wrote:
I wrote a postprocessor to do it. It's at http://kegel.com/wine/skipgood.pl.txt An example of its output is at http://kegel.com/wine/failing.html It really does make it easier to see all the failures.
Your webserver doesn't like serving your skipgood.pl.txt file =/
Bleah. I added a symlink, you can see it now at http://kegel.com/wine/skipgood.txt I also updated http://kegel.com/wine/failing.html It's very interesting to see what's failing! The d3d9/visual tests seemed to pass for Jeremy White, but only because a skip() was missing. Patch sent, http://winehq.org/pipermail/wine-patches/2008-May/054930.html The only tests that are failing for every system are shell32:shelllink and d3d9/visual. shelllink doesn't fail when run via make test, so it might be a winetest artifact. Do the d3d9/visual tests really pass for *anybody*? The user32/msg tests pass if you have a good window manager, as noted on http://wiki.winehq.org/MakeTestFailures As Jeremy suggested, several of the remaining failures might be missing packages at build time, and we could deal with some of those by making them mandatory. - Dan
2008/5/16 Dan Kegel <dank(a)kegel.com>:
On Thu, May 15, 2008 at 4:52 PM, Jeremy White <jwhite(a)codeweavers.com> wrote:
Say, who maintains that web site? It'd be handy to have an option to suppress rows that have neither crashes nor failures; right now you have to scroll vertically a whole lot to see all the failures.
I'm not sure. The source is in this git tree: http://source.winehq.org/git/tools.git
I wrote a postprocessor to do it. It's at http://kegel.com/wine/skipgood.pl.txt An example of its output is at http://kegel.com/wine/failing.html It really does make it easier to see all the failures.
Looks nice, but the original one is useful too as it shows the todo and skipped results in there as well. Is there a way to show how many tests are failing because they are todo block successes. For example, the ntdll results can be ignored as the error is a todo block success that seems to be triggered only on certain machines (3/4 of the 17 results thus far); this would then help focus the effort on the failures that are more important. - Reece
On Fri, May 16, 2008 at 12:37 AM, Reece Dunn <msclrhd(a)googlemail.com> wrote:
I wrote a postprocessor to do it. It's at http://kegel.com/wine/skipgood.pl.txt An example of its output is at http://kegel.com/wine/failing.html
Looks nice, but the original one is useful too as it shows the todo and skipped results in there as well.
Yeah. How 'bout this: provide both the original and the postprocessed file. People who want the postprocessed version will just have to know to add /failing.html for now. That will get us up and running quick.
Is there a way to show how many tests are failing because they are todo block successes. For example, the ntdll results can be ignored as the error is a todo block success that seems to be triggered only on certain machines (3/4 of the 17 results thus far); this would then help focus the effort on the failures that are more important.
Not easy for the postprocessor given the current data, I think. I wouldn't worry about that detail too much.
On Thu, May 15, 2008 at 12:47 PM, Jeremy White <jwhite(a)codeweavers.com> wrote:
Hi Folks,
One key goal for Wine 1.0 is that all of its conformance tests run successfully on nearly all systems. We would really like your help in figuring out how close we are to that goal.
To that end, if you are comfortable with checking Wine out via git, could you please visit this page: http://wiki.winehq.org/MakeTestFailures and follow the instructions there? (It's really simple; build current git Wine, download + run a script).
And, if you're a Wine developer, since Alexandre is away and the code freeze is on, why not look for one of those failures in your own make test results and see if you can fix it?
Thanks!
Jeremy
I've updated http://wiki.winehq.org/MakeTestFailures a bit. I've added more bug reports for test failures, and split them into generic and specific bug reports. @wine-devel: I've tagged quite a few bugs that only occur on PC-BSD and/or OpenSolaris, but not in Ubuntu Feisty (on a vm with the same settings). Anyone with spare time might take a look at those bugs, and see what is blocking our porting efforts. I'll work on getting those two on real hardware sometime in the next week or two for better testing. @wine-users Try testing this under as much hardware as you can. Just make sure wine is compiling properly (check ./configure --verbose for any warnings). -Austin
participants (6)
-
Austin English -
Dan Kegel -
Jeremy White -
John Klehm -
Louis Lenders -
Reece Dunn