Closed Bug 834845 Opened 13 years ago Closed 13 years ago

Angle build using pymake fails with No rule to make target 'C:/Program\' needed by ['<command-line>', 'C:/Program\\']

Categories

(Core :: Graphics: CanvasWebGL, defect)

x86_64
Windows 7
defect
Not set
normal

Tracking

()

RESOLVED FIXED

People

(Reporter: jimm, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: regression)

jim@CAGE /F/Mozilla/MC-DBG/gfx/angle $ pymake make.py[0]: Entering directory 'f:\Mozilla\MC-DBG\gfx\angle' make.py[1]: Entering directory 'f:\Mozilla\MC-DBG\gfx\angle' make.py[2]: Entering directory 'f:\Mozilla\MC-DBG\gfx\angle\src/libGLESv2' make.py[2]: Leaving directory 'f:\Mozilla\MC-DBG\gfx\angle\src/libGLESv2' make.py[2]: Entering directory 'f:\Mozilla\MC-DBG\gfx\angle\src/libEGL' make.py[2]: Leaving directory 'f:\Mozilla\MC-DBG\gfx\angle\src/libEGL' make.py[1]: Leaving directory 'f:\Mozilla\MC-DBG\gfx\angle' make.py[1]: Entering directory 'f:\Mozilla\MC-DBG\gfx\angle' No rule to make target 'C:/Program\' needed by ['<command-line>', 'C:/Program\\'] f:\Mozilla\mc\config\rules.mk:582:0: command 'f:/mozilla-build/python/python.exe f:/Mozilla/mc/build/pymake/pymake/../make.py libs' failed, return code 2 The new rev of angle changed the make logic here which seems to have broken local windows builds. adding $(warning MOZ_D3DX9_CAB - $(MOZ_D3DX9_CAB)) $(warning $(subst $(space),\$(space),$(MOZ_D3DX9_CAB))) before the target is defined shows: WARNING:pymake.data:MOZ_D3DX9_CAB - C:/Program Files (x86)/Microsoft DirectX SDK (June 2010)/Redist/Jun2010_d3dx9_43_x86.cab WARNING:pymake.data:C:/Program\ Files\ (x86)/Microsoft\ DirectX\ SDK\ (June\ 2010)/Redist/Jun2010_d3dx9_43_x86.cab pymake does not like this target one bit.
Switch back to the older more simplified logic as a work around - http://hg.mozilla.org/mozilla-central/diff/65a26453f9a6/gfx/angle/Makefile.in
Looks like the build slaves don't die because they don't have spacing in their paths: e:\builds\moz2_slave\m-cen-w32\build\obj-firefox\gfx\angle\Makefile:145:0$ expand "C:/Tools/sdks/dx10/Redist/Jun2010_d3dx9_43_x86.cab" -F:d3dx9_43.dll "../../dist/bin" e:\builds\moz2_slave\m-cen-w32\build\obj-firefox\gfx\angle\Makefile:151:0$ expand "C:/Tools/sdks/dx10/Redist/Jun2010_D3DCompiler_43_x86.cab" -F:D3DCompiler_43.dll "../../dist/bin"
The Makefile changes from bug 823109 were backed out on inbound, but the backout has not yet been merged to mozilla-central. Another workaround is to revert to using make, rather than pymake. But then you can't do parallel makes.
Blocks: 823109
Keywords: regression
Status: NEW → RESOLVED
Closed: 13 years ago
Resolution: --- → DUPLICATE
as a workaround you can use --disable-webgl
No longer blocks: 823109
Summary: Angle build fails with No rule to make target 'C:/Program\' needed by ['<command-line>', 'C:/Program\\'] → Angle build using pymake fails with No rule to make target 'C:/Program\' needed by ['<command-line>', 'C:/Program\\']
(In reply to Marco Bonardo [:mak] from comment #5) > as a workaround you can use --disable-webgl Well, that is fine unless Windows WebGL issues are what you are trying to debug and/or work on. A better workaround is to just backout the code form you tree using this patch. http://www.wg9s.com/mozilla/firefox/patches/backout_bug823109.diff
(In reply to Marco Bonardo [:mak] from comment #4) > > *** This bug has been marked as a duplicate of bug 823109 *** I am not sure how this can be a duplicate of the bug for which the check-in caused the issue. I do not think that is how we normally handle this. My Intention was to keep this open as a regression bug caused by bug 823109 until such time as the backout landed on mozilla-central and then resolve this bug as fixed by the backout.
Status: RESOLVED → REOPENED
Resolution: DUPLICATE → ---
Blocks: 823109
No longer blocks: 801158
Status: REOPENED → NEW
The backout is on m-c now and I can confirm that my builds succeed again.
Status: NEW → RESOLVED
Closed: 13 years ago13 years ago
Resolution: --- → FIXED
Oh hey, thanks for looking into this but - please cc authors and reviewers of existing code, as shown by hg log for the files you're modifying - please paste changeset links here So what is it that got backed out here (comment 8) ?
(In reply to Benoit Jacob [:bjacob] from comment #9) > Oh hey, thanks for looking into this but > - please cc authors and reviewers of existing code, as shown by hg log for > the files you're modifying > - please paste changeset links here > > So what is it that got backed out here (comment 8) ? This code from bug 823109: https://hg.mozilla.org/mozilla-central/rev/ed1d3c2794a8
(In reply to Bill Gianopoulos [:WG9s] from comment #10) > (In reply to Benoit Jacob [:bjacob] from comment #9) > > Oh hey, thanks for looking into this but > > - please cc authors and reviewers of existing code, as shown by hg log for > > the files you're modifying > > - please paste changeset links here > > > > So what is it that got backed out here (comment 8) ? > > This code from bug 823109: > > https://hg.mozilla.org/mozilla-central/rev/ed1d3c2794a8 That changeset was the actual backout.
Ah OK. CC'ing Jeff M (!= Jeff G) here.
You need to log in before you can comment on or make changes to this bug.