/var/folders/58/lf0k_v8n6jlg2h5gw3_crs3m0000gn/T/abtmp-ad6589ed742c-R0Upyn/compilePath/js/src/vm/PosixNSPR.cpp:267:13: note: candidate constructor (the implicit move constructor) not viable: cannot convert argument of incomplete type 'PRLock *' to 'nspr::CondVar' class nspr::CondVar ^ /var/folders/58/lf0k_v8n6jlg2h5gw3_crs3m0000gn/T/abtmp-ad6589ed742c-R0Upyn/compilePath/js/src/vm/PosixNSPR.cpp:273:5: note: candidate constructor not viable: cannot convert argument of incomplete type 'PRLock *' to 'nspr::Lock *' CondVar(nspr::Lock *lock) : lock_(lock) {} fatal error: too many errors emitted, stopping now [-ferror-limit=] 20 errors generated. In the middle of the attached build log are these errors. Perhaps this is causing the build to fail somewhat? autoBisect shows this is probably related to the following changeset: The first bad revision is: changeset: http://hg.mozilla.org/mozilla-central/rev/3b9e118ded0f user: Ehsan Akhgari date: Fri Nov 22 00:16:31 2013 -0500 summary: Bug 941424 - Build more of the JS engine in unified mode; r=djvj

My configure arguments were: CC="clang -Qunused-arguments" AR=ar CXX="clang++ -Qunused-arguments" sh ./configure --target=x86_64-apple-darwin12.5.0 --enable-optimize --enable-debug --enable-profiling --enable-gczeal --enable-debug-symbols --enable-methodjit --enable-type-inference --disable-tests --enable-more-deterministic --with-ccache --disable-threadsafe Probably the key flag here is "--disable-threadsafe". I think threadsafe builds worked for me (and TBPL produces threadsafe builds too).

jslock.h doesn't try to use PosixNSPR.h if JS_THREADSAFE is not defined. However, we still try to build PosixNSPR.cpp in non-threadsafe mode, and when that file tries to include PosixNSPR.h, its definition clashes with the PRThread etc typedefs in jslock.h. Not sure what the desired semantics are here, but it seems to me like we should not be building PosixNSPR.cpp in non-threadsafe builds at all.

Comment on attachment 8337396 [details] [diff] [review] Don't build PosixNSPR.cpp in non-threadsafe mode; r=billm Like this? This seems to fix the build for me. Gary, can you please check this patch to see if it fixes your build as well?

I think I'd prefer this. JS_POSIX_NSPR shouldn't be enabled without JS_THREADSAFE.

Comment on attachment 8337396 [details] [diff] [review] Don't build PosixNSPR.cpp in non-threadsafe mode; r=billm Yes, this does fix the issue I had.

Comment on attachment 8337402 [details] [diff] [review] posix-nspr-fix Review of attachment 8337402 [details] [diff] [review]: ----------------------------------------------------------------- Looks good!

I helped land this to fix this breakage for Monday. Thanks everyone! https://hg.mozilla.org/integration/mozilla-inbound/rev/48161187ac9a

Comment on attachment 8337396 [details] [diff] [review] Don't build PosixNSPR.cpp in non-threadsafe mode; r=billm Thanks Gary.

Gary, can you confirm this is no longer an issue for you with today's builds?

This particular issue was fixed then. More recent build breakage comprised of bug 947683 and bug 948301.