Created attachment 8484595 [details] Compressed build.log User Agent: Mozilla/5.0 (X11; Linux ia64; rv:30.0) Gecko/20100101 Firefox/30.0 Build ID: 20140618024221 Steps to reproduce: Compiling brand new Firefox 32 on ia64 platform (Itanium). Actual results: Compilation failed with a lot of errors related to JSVal. Expected results: Successful compilation ;-)
Severity: normal → major
Version: 30 Branch → 32 Branch
I have the strong impression that the JS engine is once again completely broken on ia64, as these JSValue-related errors remind me a similar (unfixed) issue in WebKit . A summary of the situation can be found there. Émeric  https://bugs.webkit.org/show_bug.cgi?id=129992
This is a regression from Firefox 30 and 31. While Firefox 31 (and some other versions) final link is failing because of ld: libxul.so: short data segment overflowed (see  if interested), I'm filling in this bug report from a successfully built Firefox 30 instance. Émeric  https://bugs.gentoo.org/show_bug.cgi?id=497514
https://software.intel.com/en-us/articles/short-data-segment-overflow-error-on-linux-64-on-itaniumr-architecture The error basically means that there are too many global variables and exported/imported functions.
(In reply to Masatoshi Kimura [:emk] from comment #3) > https://software.intel.com/en-us/articles/short-data-segment-overflow-error- > on-linux-64-on-itaniumr-architecture The error basically means that there > are too many global variables and exported/imported functions. Just to be sure that I correctly understand, you're talking about the libxul short data segment overflow that I was hitting during the final link of Firefox 31, not the jsval-related compilation issue that I'm now having with Firefox 32, right? Émeric
(In reply to Émeric Maschino from comment #4) > Just to be > sure that I correctly understand, you're talking about the libxul short data > segment overflow that I was hitting during the final link of Firefox 31, not > the jsval-related compilation issue that I'm now having with Firefox 32, > right? In case my previous question might not appear clear, is the link to the Intel website only for the short data segment overflow, or are you also seeing in something in the attached build.log (https://bugzilla.mozilla.org/attachment.cgi?id=8484595) a similar issue that would trigger compilation failure of Firefox 32? Émeric
You need to log in before you can comment on or make changes to this bug.