Avoid panic with recordsize > 128k, raw sending and no large_blocks

The current codebase does not support raw sending buffers with block
size > 128kB when large_blocks is not active. This can happen in the
codepath dsl_dataset_sync()->dmu_objset_sync()->zio_nowait() which
calls back dmu_objset_write_done()->dsl_dataset_block_born(). If
dsl_dataset_sync() completes its run before dsl_dataset_block_born() is
called, we will end up not activating some of the necessary flags, while
having blocks based on those flags written in the filesystem. A
subsequent send will then panic.

Fix this by directly deciding in dmu_objset_sync() whether these flags
need to be activated later by dsl_dataset_sync(). Instead of panicking
due to a NULL pointer dereference in dmu_dump_write() in case of a send,
print out an error message. Also during scrub verify there are no
contradicting filesystem flags.

Reviewed-by: Paul Dagnelie <pcd@delphix.com>
Signed-off-by: George Amanakis <gamanakis@gmail.com>
Closes #12275
Closes #12438
This commit is contained in:
George Amanakis
2022-06-27 23:17:25 +02:00
committed by GitHub
parent 1cd72b9c13
commit 80a650b7bb
6 changed files with 66 additions and 20 deletions
+15
View File
@@ -2004,6 +2004,21 @@ dsl_scan_visitbp(blkptr_t *bp, const zbookmark_phys_t *zb,
return;
}
/*
* Check if this block contradicts any filesystem flags.
*/
spa_feature_t f = SPA_FEATURE_LARGE_BLOCKS;
if (BP_GET_LSIZE(bp) > SPA_OLD_MAXBLOCKSIZE)
ASSERT3B(dsl_dataset_feature_is_active(ds, f), ==, B_TRUE);
f = zio_checksum_to_feature(BP_GET_CHECKSUM(bp));
if (f != SPA_FEATURE_NONE)
ASSERT3B(dsl_dataset_feature_is_active(ds, f), ==, B_TRUE);
f = zio_compress_to_feature(BP_GET_COMPRESS(bp));
if (f != SPA_FEATURE_NONE)
ASSERT3B(dsl_dataset_feature_is_active(ds, f), ==, B_TRUE);
if (bp->blk_birth <= scn->scn_phys.scn_cur_min_txg) {
scn->scn_lt_min_this_txg++;
return;