I have 2 cases where statically built binaries have seg faulted with glibc-2.3.2-r1. These work fine in other distributions with the same glibc so I'm not sure what the problem is. One of them is an hsi program, and running with strace on a gentoo box with glibc-2.3.2-r1 produces the following output: close(3) = 0 open("/lib/ld-linux.so.2", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\260\t\0"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=89699, ...}) = 0 old_mmap(NULL, 80812, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4015b000 mprotect(0x4016e000, 2988, PROT_NONE) = 0 old_mmap(0x4016e000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x12000) = 0x4016e000 close(3) = 0 --- SIGSEGV (Segmentation fault) @ 0 (0) --- +++ killed by SIGSEGV +++ On a different gentoo machine, but still running glibc-2.3.2-r1, running strace with statically linked condor (http://www.cs.wisc.edu/condor) produces the following: close(3) = 0 open("/lib/ld-linux.so.2", O_RDONLY) = 3 read(3, "\177ELF\1\1\1\0\0\0\0\0\0\0\0\0\3\0\3\0\1\0\0\0\260y\221"..., 1024) = 1024 fstat64(3, {st_mode=S_IFREG|0755, st_size=89296, ...}) = 0 old_mmap(0x4a917000, 75404, PROT_READ|PROT_EXEC, MAP_PRIVATE, 3, 0) = 0x4a917000 mprotect(0x4a929000, 1676, PROT_NONE) = 0 old_mmap(0x4a929000, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_FIXED, 3, 0x12000) = 0x4a929000 close(3) = 0 brk(0x82a8000) = 0x82a8000 --- SIGSEGV (Segmentation fault) @ 0 (0) --- +++ killed by SIGSEGV +++ These both work on gentoo boxes with earlier versions of glibc. Please let me know if there is any more info that would be useful to this! Thanks
you sure they're statically linked ? if they were, they wouldnt have opened ld-linux.so ... `ldd <file>`
This is confirmed. Happens for me on QT static binaries and basically the same strace results.
The same problem here:I had to reemerge netcat which is static after upgrading to glibc 2.3.2 or it segfaulted. Another that does no longer work for me is p4v the perforce visual client which is binary only, compiled against earlier glibc. I saw on google that LFS folks had the same problem with the transition.
*** This bug has been marked as a duplicate of 25299 ***
