Hey guys, while snooping around some more trying to figure out what might be causing the MIGRATE_* value corruption and thus the crash, I found the following bug. In nid_range(), we can overshoot "end" if "start" begins on a non-page boundary. It may or may not fix the bug we're tracking down, I don't know, but it's worth testing :-) Anyways, the patch is below. I also look forward to the tracing results from that debugging patch I gave you guys earlier. Thanks. diff --git a/arch/sparc64/mm/init.c b/arch/sparc64/mm/init.c index 217de3e..6172951 100644 --- a/arch/sparc64/mm/init.c +++ b/arch/sparc64/mm/init.c @@ -796,6 +796,9 @@ static unsigned long nid_range(unsigned long start, unsigned long end, start += PAGE_SIZE; } + if (start > end) + start = end; + return start; } #else -- To unsubscribe from this list: send the line "unsubscribe sparclinux" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
