----- Original Message ----- > Greetings, > > I'm trying to debug a crash dump with the 'crash' > utility but I'm running into the stack pointer and > frame pointer being absent from the stack backtrace. What you see is what you get -- it's not "missing" anything. That's just the output format of x86_64 backtraces. > > A snippet of the dump is: > > .. > > #4 [ffff88084c969ef0] vfs_write at ffffffff8116d098 > ffff88084c969ef8: ffff88084c969f78 ffffffff810d42b2 > ffff88084c969f08: ffff88085b280660 ffff880858d5f240 > ffff88084c969f18: 00007f43f3d9d000 0000000000000002 > ffff88084c969f28: ffff88084c969f78 ffffffff8116dad1 > #5 [ffff88084c969f30] sys_write at ffffffff8116dad1 > ffff88084c969f38: ffff88084c969f78 00007f43f3d9d000 > ffff88084c969f48: 0000000000000000 0000000000000002 > ffff88084c969f58: 000000324b779780 0000000000000002 > ffff88084c969f68: 000000324b779780 0000000000000002 > ffff88084c969f78: 00007f43f3d9d000 ffffffff81013172 If you're comparing it to the older x86 "bt -f" format, then the "SP" and "FP" values would simply be the stack addresses of the first and last words in the dumped stack frame. So for frame #4 above, "SP" would be ffff88084c969ef8, and "FP" would be ffff88084c969f30. And for frame #5, "SP" would be ffff88084c969f38, and "FP" ffff88084c969f80. Dave -- Crash-utility mailing list Crash-utility@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/crash-utility
