block: cover another queue enter recursion via BIO_QUEUE_ENTERED
PATCH 1/1 block: cover another queue enter recursion via BIO_QUEUE_ENTERED
Except for blk_queue_split(), bio_split() is used for splitting bio too, then the remained bio is often resubmit to queue via generic_make_request(). So the same queue enter recursion exits in this case too. Unfortunatley commit cd4a4ae4683dc2 doesn't help this case.
This patch covers the above case by setting BIO_QUEUE_ENTERED before calling q->make_request_fn.
In theory the per-bio flag is used to simulate one stack variable, it is just fine to clear it after q->make_request_fn is returned. Especially the same bio can't be submitted from another context.
--- block/blk-core.c | 11 +++++++++++ block/blk-merge.c | 10 ---------- 2 files changed, 11 insertions(+), 10 deletions(-)
block/blk-core.c +11 -0
@@ -1083,7 +1083,18 @@ blk_qc_t generic_make_request(struct bio *bio)/* Create a fresh bio_list for all subordinate requests */ bio_list_on_stack[1 ] = bio_list_on_stack[0 ];bio_list_init (&bio_list_on_stack[0 ]); + +/* +* Since we're recursing into make_request here, ensure +* that we mark this bio as already having entered the queue. +* If not, and the queue is going away, we can get stuck +* forever on waiting for the queue reference to drop. But +* that will never happen, as we're already holding a +* reference to it. +*/ +bio_set_flag (bio,BIO_QUEUE_ENTERED ); ret = q->make_request_fn (q, bio); +bio_clear_flag (bio,BIO_QUEUE_ENTERED );/* sort new bios into those for a lower level * and those for the same level
block/blk-merge.c +0 -10
@@ -339,16 +339,6 @@ void blk_queue_split(struct request_queue *q, struct bio **bio)/* there isn't chance to merge the splitted bio */ split->bi_opf |=REQ_NOMERGE ; -/* -* Since we're recursing into make_request here, ensure -* that we mark this bio as already having entered the queue. -* If not, and the queue is going away, we can get stuck -* forever on waiting for the queue reference to drop. But -* that will never happen, as we're already holding a -* reference to it. -*/ -bio_set_flag (*bio,BIO_QUEUE_ENTERED ); -bio_chain (split, *bio);trace_block_split (q, split, (*bio)->bi_iter.bi_sector);generic_make_request (*bio);
Signature · 1 line
2.9.5
Quoted 12 lines · Ming Lei
Except for
blk_queue_split(),bio_split()is used for splitting bio too, then the remained bio is often resubmit to queue viageneric_make_request(). So the same queue enter recursion exits in this case too. Unfortunatley commitcd4a4ae4683dc2doesn't help this case.This patch covers the above case by setting
BIO_QUEUE_ENTEREDbefore callingq->make_request_fn.In theory the per-bio flag is used to simulate one stack variable, it is just fine to clear it after
q->make_request_fnis returned. Especially the same bio can't be submitted from another context.
Quoted 5 lines
Fixes cd4a4ae4683dc2 ("block: don't use blocking queue entered for recursive bio submits")Cc Tetsuo Handa <[email protected]>Cc Mike Snitzer <[email protected]>Cc NeilBrown <[email protected]>Signed-off-by Ming Lei <[email protected]>
Looks good, I'll drop my DM patch that set/cleared BIO_QUEUE_ENTERED.
Jens, you OK with picking this up for 5.0-rc?
Jens Axboe
maintainerQuoted 21 lines · Mike Snitzer
Quoted 12 lines · Ming Lei
Except for
blk_queue_split(),bio_split()is used for splitting bio too, then the remained bio is often resubmit to queue viageneric_make_request(). So the same queue enter recursion exits in this case too. Unfortunatley commitcd4a4ae4683dc2doesn't help this case.This patch covers the above case by setting
BIO_QUEUE_ENTEREDbefore callingq->make_request_fn.In theory the per-bio flag is used to simulate one stack variable, it is just fine to clear it after
q->make_request_fnis returned. Especially the same bio can't be submitted from another context.Quoted 5 lines
Fixes cd4a4ae4683dc2 ("block: don't use blocking queue entered for recursive bio submits")Cc Tetsuo Handa <[email protected]>Cc Mike Snitzer <[email protected]>Cc NeilBrown <[email protected]>Signed-off-by Ming Lei <[email protected]>
Quoted 5 lines
Looks good, I'll drop my DM patch that set/cleared
BIO_QUEUE_ENTERED.Jens, you OK with picking this up for 5.0-rc?
Reviewed-by Mike Snitzer <[email protected]>
I am, thanks both of you for looking into this. I'm a bit behind on email and patches from over the long weekend.
Signature · 1 line
Jens Axboe