Comment on attachment 758666 [details] [diff] [review] move HOST_CSRCS to moz.build (logic) Add HOST_CSRCS as a passthrough variable.

Comment on attachment 758666 [details] [diff] [review] move HOST_CSRCS to moz.build (logic) Review of attachment 758666 [details] [diff] [review]: ----------------------------------------------------------------- ::: python/mozbuild/mozbuild/frontend/sandbox_symbols.py @@ +92,5 @@ > delimiters. > """), > > + 'HOST_CSRCS': (StrictOrderingOnAppendList, list, [], > + """C code source files to include when cross-compiling. I would probably say "C source files to compile with the host compiler."

Changed variable description per review, rebased, r=ted carried forward.

Comment on attachment 759373 [details] [diff] [review] move HOST_CSRCS to moz.build (logic) Inbound push: http://hg.mozilla.org/integration/mozilla-inbound/rev/ebae7298e381

Comment on attachment 759953 [details] [diff] [review] move HOST_CSRCS to mozbuild (file batch #1). First batch of HOST_CSRCS files converted. Try results pending.

Comment on attachment 759953 [details] [diff] [review] move HOST_CSRCS to mozbuild (file batch #1). >diff --git a/toolkit/crashreporter/google-breakpad/src/common/moz.build b/toolkit/crashreporter/google-breakpad/src/common/moz.build >--- a/toolkit/crashreporter/google-breakpad/src/common/moz.build >+++ b/toolkit/crashreporter/google-breakpad/src/common/moz.build >@@ -42,8 +42,11 @@ if CONFIG['OS_ARCH'] == 'Linux': > if CONFIG['OS_TARGET'] == 'Android': > pass > else: > if CONFIG['OS_TARGET'] != 'WINNT': > CPP_SOURCES += [ > 'stabs_to_module.cc', > 'stabs_reader.cc', > ] >+ >+if 'WINNT' != CONFIG['OS_TARGET'] and CONFIG['MOZ_CRASHREPORTER']: >+ HOST_CSRCS += CSRCS nit: For consistency with other moz.build files, please put the CONFIG variable before the string it is being compared to: if CONFIG['OS_TARGET'] != 'WINNT' and CONFIG['MOZ_CRASHREPORTER']:

Comment on attachment 760476 [details] [diff] [review] move HOST_CSRCS to mozbuild (file batch #1). Nit fix, r=mshal carried forward

Comment on attachment 760476 [details] [diff] [review] move HOST_CSRCS to mozbuild (file batch #1). try job: https://tbpl.mozilla.org/?tree=Try&rev=3cf081964557

Comment on attachment 760476 [details] [diff] [review] move HOST_CSRCS to mozbuild (file batch #1). push to inbound: https://hg.mozilla.org/integration/mozilla-inbound/rev/3e73c00a7db1

https://hg.mozilla.org/mozilla-central/rev/3e73c00a7db1#l16.11 looks silly.

(In reply to :Ms2ger from comment #15) > https://hg.mozilla.org/mozilla-central/rev/3e73c00a7db1#l16.11 looks silly. Feel free to grab and work on more of the conversions then.

Conversion patch has stuck for a week so should hold. Remove DISABLED_ variables and add HOST_CSRCS to the restricted variable list.

Comment on attachment 763709 [details] [diff] [review] Cleanup for mozbuild HOST_CSRCS conversion. Review of attachment 763709 [details] [diff] [review]: ----------------------------------------------------------------- Good except for the one CSRCS issue (which isn't fatal, but might as well clean it up now). ::: modules/libmar/src/Makefile.in @@ +23,5 @@ > +CSRCS = \ > + mar_create.c \ > + mar_extract.c \ > + mar_read.c \ > + $(NULL) What's up with this file? Hasn't CSRCS moved to moz.build files? Should this list be in a moz.build file?

(In reply to Gregory Szorc [:gps] from comment #19) > Comment on attachment 763709 [details] [diff] [review] > Cleanup for mozbuild HOST_CSRCS conversion. > > Review of attachment 763709 [details] [diff] [review]: > ----------------------------------------------------------------- > > Good except for the one CSRCS issue (which isn't fatal, but might as well > clean it up now). > > ::: modules/libmar/src/Makefile.in > @@ +23,5 @@ > > +CSRCS = \ > > + mar_create.c \ > > + mar_extract.c \ > > + mar_read.c \ > > + $(NULL) > > What's up with this file? Hasn't CSRCS moved to moz.build files? Should this > list be in a moz.build file? There are a boatload of variables/values that need to move and there has only been traction for landing passthrough variable/conversion patches recently. The nomakefiles bug https://bugzilla.mozilla.org/show_bug.cgi?id=847009 bug and wiki page still itemize all the gory details for conversion status. Most of the CSRCS assignments have already moved to mozbuild. Any lingering (like libmar/) were included in a try patch with failures and need triage/diagnosis. Temporarily expanding CSRCS=$(HOST_CSRCS) in this case helps to prevent patches from becoming inter-dependent. If CSRCS= is still failing that can block this patch which is ready to land on researching a try failure.

CSRCS= will be handled in the conversion bug 870406

Inbound push for HOST_CSRCS conversion cleanup: https://hg.mozilla.org/integration/mozilla-inbound/rev/045aa030111e