472611 - set up try-mac-slave03

Status: RESOLVED FIXED

(Release Engineering :: General, defect, P2)

Description

[Nick Thomas [:nthomas] (UTC+12)]

Builds on this machine have been failing recently with errors like
g++-4.0 -arch ppc  -fno-rtti -fno-exceptions -Wall -Wpointer-arith -Woverloaded-virtual -Wsynth -Wno-ctor-dtor-privacy -Wno-non-virtual-dtor -Wcast-align -Wno-invalid-offsetof -Wno-long-long -gdwarf-2 -isysroot /Developer/SDKs/MacOSX10.4u.sdk -fno-strict-aliasing -fpascal-strings -fno-common -fshort-wchar  -DNDEBUG -DTRIMMED -Os -fstrict-aliasing  -fPIC  -o libmozjs.dylib  jsapi.o jsarena.o jsarray.o jsatom.o jsbool.o jscntxt.o jsdate.o jsdbgapi.o jsdhash.o jsdtoa.o jsemit.o jsexn.o jsfun.o jsgc.o jshash.o jsinterp.o jsinvoke.o jsiter.o jslock.o jslog2.o jslong.o jsmath.o jsnum.o jsobj.o json.o jsopcode.o jsparse.o jsprf.o jsregexp.o jsscan.o jsscope.o jsscript.o jsstr.o jsutil.o jsxdrapi.o jsxml.o prmjtime.o     -framework Cocoa  -Wl,-executable_path,/bin -Wl,-dead_strip -L/builds/buildbot/sendchange-slave/sendchange-mac-hg/mozilla/objdir/ppc/dist/lib -lplds4 -lplc4 -lnspr4 -dynamiclib -install_name @executable_path/libmozjs.dylib -compatibility_version 1 -current_version 1 -single_module  -lm  
ld: jslong.o r_address (0x10b393) field of relocation entry 3 in section (__DWARFA,__debug_info) out of range
/usr/bin/libtool: internal link edit command failed

The other two slaves (try-mac-slave01/02) are not failing on this, so I'm going to take a _guess_ that this from enabling DWARF symbols. I see that bm-xserve15 has XCode 2.2.1 on OSX 10.4.8, compared to XCode 3.1 on 10.5.2 for try-mac-slave01. So really this is "make sure all the mac try slaves are the same". 

This box was also flailing away with several process trees for "make -C browser/installer", with 15 hdiutil chewing up CPU time (since Jan08 one of them). I've rebooted it.

Buildbot is currently stopped on sm-xserve15 to prevent more false bustage.

Comment 1

[John O'Duinn [:joduinn] (please use "needinfo?" flag)]

Thanks Nick for taking "mac1" out of the pool for now while we investigate+fix. 

Earlier today, I asked bz submit the same job 3 times in a row. The jobs handled by "mac1" failed, while the job handled by "mac3" passed. This experiment did two things:

- got bz unstuck
- confirmed something is busted on this "mac1" slave.

Comment 2

[Boris Zbarsky [:bzbarsky]]

Didn't quite get me unstuck, since I needed the Mac talos data for this patch, and the Mac talos box on tryserver seems to be AWOL.  :(

Comment 3

[[e-mail] (:bhearsum)]

I think I've finished setting this up now. I'm doing a test build in staging before I move it into production.

Comment 4

[[e-mail] (:bhearsum)]

This machine is back in the production pool and successfully producing builds.

Another one to come in 475748.