Description of problem: Java segfaults when building subversion [1]. This is just short snippet: # # A fatal error has been detected by the Java Runtime Environment: # # [thread -1584399552 also had an error] SIGSEGV (0xb) at pc=0xf6c83483, pid=14367, tid=2714762048 # [thread -1590691008 also had an error]# JRE version: OpenJDK Runtime Environment (8.0_40-b12) (build 1.8.0_40-b12) # Java VM: OpenJDK Server VM (25.40-b16 mixed mode linux-x86 ) # Problematic frame: # [thread -1582302400 also had an error] V [libjvm.so+0x701483][thread -1578108096 also had an error] Version-Release number of selected component (if applicable): java-1.8.0-openjdk.i686 How reproducible: It fails 4 times out of 5 tries, so quite often. Steps to Reproduce: 1. fedpkg co subversion 2. cd subversion 3. fedpkg scratch-build Actual results: Expected results: Additional info: From subversion.spec "make javahl" might be probably shortcut. [1] http://koji.fedoraproject.org/koji/taskinfo?taskID=9428327
Looking at root.log, the complete version is: java-1.8.0-openjdk-devel-1.8.0.40-19.b12.fc22.i686
root.log shows glibc v2.21.90-9.fc23. It might be a duplicate of bug 1209451. Can you reproduce with glibc v2.21.90-10.fc23 being used?
(In reply to Severin Gehwolf from comment #2) > root.log shows glibc v2.21.90-9.fc23. It might be a duplicate of bug > 1209451. Can you reproduce with glibc v2.21.90-10.fc23 being used? That seems to be the case, since the build can now get past the java build: http://koji.fedoraproject.org/koji/taskinfo?taskID=9443450 Thank you. *** This bug has been marked as a duplicate of bug 1209451 ***