From: Kent Overstreet <kent.overstreet@linux.dev>
Date: Tue, 19 Dec 2023 22:16:34 +0000 (-0500)
Subject: bcachefs: Fix bch2_alloc_sectors_start_trans() error handling
X-Git-Url: http://git.maquefel.me/?a=commitdiff_plain;h=247ce5f1bb3ea90879e8552b8edf4885b9a9f849;p=linux.git

bcachefs: Fix bch2_alloc_sectors_start_trans() error handling

When we fail to allocate because of insufficient open buckets, we don't
want to retry from the full set of devices - we just want to retry in
blocking mode.

But if the retry in blocking mode fails with a different error code, we
end up squashing the -BCH_ERR_open_buckets_empty error with an error
that makes us thing we won't be able to allocate (insufficient_devices)
- which is incorrect when we didn't try to allocate from the full set of
devices, and causes the write to fail.

Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>
---

diff --git a/fs/bcachefs/alloc_foreground.c b/fs/bcachefs/alloc_foreground.c
index 1ba0eeb7552a2..0e61579826072 100644
--- a/fs/bcachefs/alloc_foreground.c
+++ b/fs/bcachefs/alloc_foreground.c
@@ -1374,8 +1374,17 @@ retry:
 			goto alloc_done;
 
 		/* Don't retry from all devices if we're out of open buckets: */
-		if (bch2_err_matches(ret, BCH_ERR_open_buckets_empty))
-			goto allocate_blocking;
+		if (bch2_err_matches(ret, BCH_ERR_open_buckets_empty)) {
+			int ret = open_bucket_add_buckets(trans, &ptrs, wp, devs_have,
+					      target, erasure_code,
+					      nr_replicas, &nr_effective,
+					      &have_cache, watermark,
+					      flags, cl);
+			if (!ret ||
+			    bch2_err_matches(ret, BCH_ERR_transaction_restart) ||
+			    bch2_err_matches(ret, BCH_ERR_open_buckets_empty))
+				goto alloc_done;
+		}
 
 		/*
 		 * Only try to allocate cache (durability = 0 devices) from the
@@ -1389,7 +1398,6 @@ retry:
 					      &have_cache, watermark,
 					      flags, cl);
 	} else {
-allocate_blocking:
 		ret = open_bucket_add_buckets(trans, &ptrs, wp, devs_have,
 					      target, erasure_code,
 					      nr_replicas, &nr_effective,