On 5/20/2024 9:19 PM, Christoph Hellwig wrote: >> diff --git a/block/bio-integrity.c b/block/bio-integrity.c >> index 2e3e8e04961e..8b528e12136f 100644 >> --- a/block/bio-integrity.c >> +++ b/block/bio-integrity.c >> @@ -144,16 +144,38 @@ void bio_integrity_free(struct bio *bio) >> struct bio_integrity_payload *bip = bio_integrity(bio); >> struct bio_set *bs = bio->bi_pool; >> >> + if (bip->bip_flags & BIP_INTEGRITY_USER) >> + return; >> if (bip->bip_flags & BIP_BLOCK_INTEGRITY) >> kfree(bvec_virt(bip->bip_vec)); >> - else if (bip->bip_flags & BIP_INTEGRITY_USER) >> - bio_integrity_unmap_user(bip); >> >> __bio_integrity_free(bs, bip); >> bio->bi_integrity = NULL; >> bio->bi_opf &= ~REQ_INTEGRITY; > > This looks correct. I wish we could go one step further (maybe > in a separate patch/series) to also move freeing the bio integrity > data to the callers. In fact I wonder if there is any point in > doing this early separate free vs just doing it as part of the > final bio put. Tried to think in this direction in past. But bio_put will not be called if bio is statically allocated. Currently that gets freed implicitly because bio_endio is called and it tries to free the integrity.
