On Sun, 23 Feb 2014, H. Peter Anvin wrote: > So we do a write to the buffer rather immediately before this happens, > and in particular that will update the head: > > rb->user_page->data_head = head; > > However, that doesn't explain what is going on and in particular the > write to whatever address was in %rbp. The rest pretty much seems to be > the page fault logic.
It turns out you don't even have to over-write rb->user_page->data_head. Just touching the mmap page with a write of a single byte (it doesn't matter where) is enough to trigger the bug. This is a pain to track down, it would be easier if I could get a replayable syscall trace, but even though the segfault is very reproducible with my fuzzer, it's very sensitive to extra syscalls in the trace path and the fuzzer logger/replayer path has a different number of write syscalls and won't trigger the problem. > Incidentally, I doubt that this is x32-related in any way; there seems > to be absolutely no difference between x86-64 perf and x32 perf; more > likely it just makes the error more reproducible because the address > space is so much smaller. quite possibly. I only began chasing the problem because when compiled for x32 this bug apparently will reboot the machine now and then (not just segfault the program). I never saw that failure mode with x86_64, but again maybe it's just easier to hit with the reduced address space as you say. Vince -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/