On 6/10/20 10:48 PM, Hugh Dickins wrote: > While stressing compaction, one run oopsed on NULL capc->cc in > __free_one_page()'s task_capc(zone): compact_zone_order() had been > interrupted, and a page was being freed in the return from interrupt. > > Though you would not expect it from the source, both gccs I was using > (a 4.8.1 and a 7.5.0) had chosen to compile compact_zone_order() with > the ".cc = &cc" implemented by mov %rbx,-0xb0(%rbp) immediately before > callq compact_zone - long after the "current->capture_control = &capc". > An interrupt in between those finds capc->cc NULL (zeroed by an earlier > rep stos). Ugh, nasty. Same here with gcc 10. > This could presumably be fixed by a barrier() before setting > current->capture_control in compact_zone_order(); but would also need > more care on return from compact_zone(), in order not to risk leaking > a page captured by interrupt just before capture_control is reset. I was hoping a WRITE_ONCE(current->capture_control) would be enough, but apparently it's not (I tried). > Maybe that is the preferable fix, but I felt safer for task_capc() to > exclude the rather surprising possibility of capture at interrupt time. > Fixes: 5e1f0f098b46 ("mm, compaction: capture a page under direct compaction") > Cc: stable@xxxxxxxxxxxxxxx # 5.1+ > Signed-off-by: Hugh Dickins <hughd@xxxxxxxxxx> Acked-by: Vlastimil Babka <vbabka@xxxxxxx> But perhaps I would also make sure that we don't expose the half initialized capture_control and run into this problem again later. It's not like this is a fast path where barriers hurt. Something like this then? (with added comments) diff --git a/mm/compaction.c b/mm/compaction.c index fd988b7e5f2b..c89e26817278 100644 --- a/mm/compaction.c +++ b/mm/compaction.c @@ -2316,15 +2316,17 @@ static enum compact_result compact_zone_order(struct zone *zone, int order, .page = NULL, }; - current->capture_control = &capc; + barrier(); + + WRITE_ONCE(current->capture_control, &capc); ret = compact_zone(&cc, &capc); VM_BUG_ON(!list_empty(&cc.freepages)); VM_BUG_ON(!list_empty(&cc.migratepages)); - *capture = capc.page; - current->capture_control = NULL; + WRITE_ONCE(current->capture_control, NULL); + *capture = READ_ONCE(capc.page); return ret; }