Open Bug 660355 Opened 15 years ago Updated 3 years ago

Random red on Windows Opt builds during packaging: Failed to get ScriptSecurityManager service

Categories

(Firefox Build System :: General, defect)

Product:
Firefox Build System
Component:
General
Platform:
x86
Windows 7
Type:
defect

Tracking

(Not tracked)

People

(Reporter: mwu, Unassigned)

Details

http://tinderbox.mozilla.org/showlog.cgi?log=Firefox/1306527711.1306533048.1876.gz This occurs during packaging, presumably when xpcshell is precompiling the js. Manifest file: ../../dist/firefox//components/interfaces.manifestLinking .xpt files completed. make[3]: Leaving directory `/e/builds/moz2_slave/cen-w32/build/obj-firefox/browser/installer' make[2]: Leaving directory `/e/builds/moz2_slave/cen-w32/build/obj-firefox/browser/installer' make[1]: Leaving directory `/e/builds/moz2_slave/cen-w32/build/obj-firefox' +++ Failed to get ScriptSecurityManager service, running without principalsmake[3]: *** [make-package] Error 5 make[2]: *** [all] Error 2 make[1]: *** [package] Error 2 make: *** [profiledbuild] Error 2
I also get the same error when I build firefox on top of xulrunner on Linux x86_64. I don't know GENERATE_CACHE should be enabled or not when LIBXUL_SDK is set. +++ Failed to get ScriptSecurityManager service, running without principals/bin/sh: line 1: 21824 Segmentation fault (core dumped) /media/bkup/work/rpmbuild/BUILD/firefox-8.0/mozilla-central/browser/installer/../../dist/bin/run-mozilla.sh /usr/lib64/xulrunner-sdk-8/bin/xpcshell -g "$PWD" -a "$PWD" -f ../../toolkit/mozapps/installer/precompile_cache.js -e 'populate_startupcache("omni.jar", "startupCache.zip");' make[1]: *** [install] Error 139
(In reply to comment #1) > I also get the same error when I build firefox on top of xulrunner on Linux > x86_64. > I don't know GENERATE_CACHE should be enabled or not when LIBXUL_SDK is set. > > +++ Failed to get ScriptSecurityManager service, running without > principals/bin/sh: line 1: 21824 Segmentation fault (core dumped) > /media/bkup/work/rpmbuild/BUILD/firefox-8.0/mozilla-central/browser/ > installer/../../dist/bin/run-mozilla.sh > /usr/lib64/xulrunner-sdk-8/bin/xpcshell -g "$PWD" -a "$PWD" -f > ../../toolkit/mozapps/installer/precompile_cache.js -e > 'populate_startupcache("omni.jar", "startupCache.zip");' > make[1]: *** [install] Error 139 Does this happen every time?
(In reply to comment #2) > Does this happen every time? Yes. Red Hat added the patch to remove GENERATE_CACHE section for their firefox 5 RPM package.
(In reply to comment #2) > (In reply to comment #1) > > I also get the same error when I build firefox on top of xulrunner on Linux > > x86_64. > > I don't know GENERATE_CACHE should be enabled or not when LIBXUL_SDK is set. > > > > +++ Failed to get ScriptSecurityManager service, running without > > principals/bin/sh: line 1: 21824 Segmentation fault (core dumped) > > /media/bkup/work/rpmbuild/BUILD/firefox-8.0/mozilla-central/browser/ > > installer/../../dist/bin/run-mozilla.sh > > /usr/lib64/xulrunner-sdk-8/bin/xpcshell -g "$PWD" -a "$PWD" -f > > ../../toolkit/mozapps/installer/precompile_cache.js -e > > 'populate_startupcache("omni.jar", "startupCache.zip");' > > make[1]: *** [install] Error 139 I think this is a separate issue, after the landing of bug 655678. Can you file a new bug and assign it to me? > Does this happen every time? Probably.
(In reply to comment #3) > (In reply to comment #2) > > Does this happen every time? > > Yes. > Red Hat added the patch to remove GENERATE_CACHE section for their firefox 5 > RPM package. I hope they apply the patches from bug 620931, then, because they are going to have other problems without...
(In reply to comment #4) > I think this is a separate issue, after the landing of bug 655678. Can you > file a new bug and assign it to me? I cannot agree with you because Red Hat generate a patch for firefox 5. This means the issue exists before the bug 655678 is fixed. Anyway I'll file another bug.
(In reply to comment #6) > (In reply to comment #4) > > I think this is a separate issue, after the landing of bug 655678. Can you > > file a new bug and assign it to me? > > I cannot agree with you because Red Hat generate a patch for firefox 5. This > means the issue exists before the bug 655678 is fixed. Yes, I figured that after writing it. Bug bug 655678 will actually help fix it. But I doubt the non ff-on-xr failure here is related, because the problem here is running xpcshell with the gre directory (-g) not pointing to the xulrunner directory. > Anyway I'll file another bug. Thanks
My use of "here" is highly ambiguous. (In reply to comment #7) > But I doubt the non ff-on-xr failure here is related, here = this bug > because the problem here (...) here = the ff-on-xr problem
(In reply to comment #6) > Anyway I'll file another bug. File as bug bug 672756.
I see this failure consistently on x86_64 with ff 6.0.1, doing a profiled build using client.mk *without* xulrunner in the loop. build flags are --enable-application=browser --disable-tests --disable-crashreporter --disable-necko-wifi --enable-official-branding (I don't have the necessary wifi libraries. With --enable-crashreporter the build fails much, much sooner.) The backtrace is only one level deep: spindle 166 /usr/src/mozilla/x86_64-spindle/shai-build/browser/installer% /usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin/run-mozilla.sh -g -d gdb /usr/src/mozilla/x86_64-spindle/shai-build/dist/bin/xpcshell -g /usr/src/mozilla/x86_64-spindle -a /usr/src/mozilla/x86_64-spindle MOZILLA_FIVE_HOME=/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin LD_LIBRARY_PATH=/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin:/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin/plugins:/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin DISPLAY=localhost:10.0 DYLD_LIBRARY_PATH=/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin:/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin LIBRARY_PATH= SHLIB_PATH=/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin:/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin LIBPATH=/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin:/usr/src/mozilla/x86_64-spindle/shai-build/browser/installer/../../dist/bin ADDON_PATH= MOZ_PROGRAM=/usr/src/mozilla/x86_64-spindle/shai-build/dist/bin/xpcshell MOZ_TOOLKIT= moz_debug=1 moz_debugger=gdb moz_debugger_args= /usr/bin/gdb --args /usr/src/mozilla/x86_64-spindle/shai-build/dist/bin/xpcshell -g /usr/src/mozilla/x86_64-spindle -a /usr/src/mozilla/x86_64-spindle GNU gdb (GDB) 7.3.0.20110830-cvs Copyright (C) 2011 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-unknown-linux-gnu". For bug reporting instructions, please see: <http://www.gnu.org/software/gdb/bugs/>... Reading symbols from /usr/src/mozilla/x86_64-spindle/shai-build/dist/bin/xpcshell...done. (gdb) run Starting program: /usr/src/mozilla/x86_64-spindle/shai-build/dist/bin/xpcshell -g /usr/src/mozilla/x86_64-spindle -a /usr/src/mozilla/x86_64-spindle [Thread debugging using libthread_db enabled] warning: the debug information found in "/usr/lib/debug//usr/lib/libexpat.so.1.5.2.dbg" does not match "/usr/lib/libexpat.so.1" (CRC mismatch). warning: the debug information found in "/usr/lib/debug/usr/lib/libexpat.so.1.5.2.dbg" does not match "/usr/lib/libexpat.so.1" (CRC mismatch). [New Thread 0x7fffeb07f700 (LWP 19234)] [New Thread 0x7fffe8a72700 (LWP 19235)] [New Thread 0x7fffe6a71700 (LWP 19236)] [New Thread 0x7fffe4a70700 (LWP 19237)] +++ Failed to get ScriptSecurityManager service, running without principals Program received signal SIGSEGV, Segmentation fault. main (argc=1, argv=0x7fffffffda38, envp=0x7fffffffda48) at /usr/src/mozilla/x86_64-spindle/js/src/xpconnect/shell/xpcshell.cpp:1916 1916 cb->findObjectPrincipals = FindObjectPrincipals;
(In reply to Nix from comment #10) > I see this failure consistently on x86_64 with ff 6.0.1, doing a profiled > build using client.mk *without* xulrunner in the loop. build flags are That most likely isn't related. Are you building in a chroot? If yes, is /proc mounted in that chroot?
I'm not building in a chroot. /proc is mounted. I'll do some more hunting soon, but this dives immediately into XPCOM code, about which I know less than nothing. straces looking for -ENOENT returns from open() tell me nothing useful. (I also see a report of a similar-looking failure from Arch Linux, at http://aur.archlinux.org/packages.php?ID=14169. That seems to have been resolved but it is not at all clear how.) What exactly is xpcshell failing to find here? A better error message would be nice... assuming that the 'failed to get ScriptSecurityManager service' message is related to the coredump, which seems extremely likely from looking at the code.
https://tbpl.mozilla.org/php/getParsedLog.php?id=6564211&tree=Firefox
Summary: Random red on Windows Opt builds during packaging → Random red on Windows Opt builds during packaging: Failed to get ScriptSecurityManager service
Product: Core → Firefox Build System
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.