lightnvm: pblk: verify that cache read is still valid

When a read is directed to the cache, we risk that the lba has been
updated during the time we made the L2P table lookup and the time we are
actually reading form the cache. We intentionally not hold the L2P lock
not to block other threads.

While strict ordering is not a guarantee at this level (unless REQ_FLUSH
has been previously issued), we have experience that some databases that
have recently implemented direct I/O support, issue metadata reads very
close to the writes, without issuing a fsync in the middle. An easy way
to support them while they is to make an extra effort and check the L2P
map right before reading the cache.

Signed-off-by: Javier González <javier@cnexlabs.com>
Signed-off-by: Matias Bjørling <matias@cnexlabs.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>

1 files changed, 12 insertions, 3 deletions

diff --git a/drivers/lightnvm/pblk-rb.c b/drivers/lightnvm/pblk-rb.c
index 7300be98e831..5ecc154f6831 100644
--- a/drivers/lightnvm/pblk-rb.c
+++ b/drivers/lightnvm/pblk-rb.c
@@ -150,6 +150,7 @@ try:
 	/* Release flags on context. Protect from writes and reads */
 	smp_store_release(&w_ctx->flags, PBLK_WRITABLE_ENTRY);
 	pblk_ppa_set_empty(&w_ctx->ppa);
+	w_ctx->lba = ADDR_EMPTY;
 }
 
 #define pblk_rb_ring_count(head, tail, size) CIRC_CNT(head, tail, size)
@@ -656,15 +657,17 @@ try:
  * be directed to disk.
  */
 int pblk_rb_copy_to_bio(struct pblk_rb *rb, struct bio *bio, sector_t lba,
-			u64 pos, int bio_iter)
+			struct ppa_addr ppa, int bio_iter)
 {
+	struct pblk *pblk = container_of(rb, struct pblk, rwb);
 	struct pblk_rb_entry *entry;
 	struct pblk_w_ctx *w_ctx;
+	struct ppa_addr l2p_ppa;
+	u64 pos = pblk_addr_to_cacheline(ppa);
 	void *data;
 	int flags;
 	int ret = 1;
 
-	spin_lock(&rb->w_lock);
 
 #ifdef CONFIG_NVM_DEBUG
 	/* Caller must ensure that the access will not cause an overflow */
@@ -674,8 +677,14 @@ int pblk_rb_copy_to_bio(struct pblk_rb *rb, struct bio *bio, sector_t lba,
 	w_ctx = &entry->w_ctx;
 	flags = READ_ONCE(w_ctx->flags);
 
+	spin_lock(&rb->w_lock);
+	spin_lock(&pblk->trans_lock);
+	l2p_ppa = pblk_trans_map_get(pblk, lba);
+	spin_unlock(&pblk->trans_lock);
+
 	/* Check if the entry has been overwritten or is scheduled to be */
-	if (w_ctx->lba != lba || flags & PBLK_WRITABLE_ENTRY) {
+	if (!pblk_ppa_comp(l2p_ppa, ppa) || w_ctx->lba != lba ||
+						flags & PBLK_WRITABLE_ENTRY) {
 		ret = 0;
 		goto out;
 	}