Skip to contentSkip to conversation

nilfs2: fix potential kernel bug due to lack of writeback flag waiting

PATCH In review Ryusuke Konishi 2 participants 2 messages Jun 16, 2024 – Jun 19, 2024 Stable branch

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 nilfs2: fix potential kernel bug due to lack of writeback flag waiting

commit a4ca369ca221bb7e06c725792ac107f0e48e82e7 upstream.

Destructive writes to a block device on which nilfs2 is mounted can cause a kernel bug in the folio/page writeback start routine or writeback end routine (__folio_start_writeback in the log below):

kernel BUG at mm/page-writeback.c:3070!

 Oops: invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI
 ...
 RIP: 0010:__folio_start_writeback+0xbaa/0x10e0
 Code: 25 ff 0f 00 00 0f 84 18 01 00 00 e8 40 ca c6 ff e9 17 f6 ff ff
  e8 36 ca c6 ff 4c 89 f7 48 c7 c6 80 c0 12 84 e8 e7 b3 0f 00 90 <0f>
  0b e8 1f ca c6 ff 4c 89 f7 48 c7 c6 a0 c6 12 84 e8 d0 b3 0f 00
 ...
 Call Trace:
  <TASK>
  nilfs_segctor_do_construct+0x4654/0x69d0 [nilfs2]
  nilfs_segctor_construct+0x181/0x6b0 [nilfs2]
  nilfs_segctor_thread+0x548/0x11c0 [nilfs2]
  kthread+0x2f0/0x390
  ret_from_fork+0x4b/0x80
  ret_from_fork_asm+0x1a/0x30
  </TASK>

This is because when the log writer starts a writeback for segment summary blocks or a super root block that use the backing device's page cache, it does not wait for the ongoing folio/page writeback, resulting in an inconsistent writeback state.

Fix this issue by waiting for ongoing writebacks when putting folios/pages on the backing device into writeback state.

---
Please apply this patch to the stable trees indicated by the subject prefix instead of the patch that failed.

This patch is tailored to account for page/folio conversion and can be applied to v6.7 and earlier.

Also, all the builds and tests I did on each stable tree passed.

Thanks,
Ryusuke Konishi

 fs/nilfs2/segment.c | 3 +++
 1 file changed, 3 insertions(+)
fs/nilfs2/segment.c +3 -0
index 2d74fb229799..5783efafbabd 100644
@@ -1694,6 +1694,7 @@ static void nilfs_segctor_prepare_write(struct nilfs_sc_info *sci)
 			if (bh->b_page != bd_page) {
 				if (bd_page) {
 					lock_page(bd_page);
+					wait_on_page_writeback(bd_page);
 					clear_page_dirty_for_io(bd_page);
 					set_page_writeback(bd_page);
 					unlock_page(bd_page);
@@ -1707,6 +1708,7 @@ static void nilfs_segctor_prepare_write(struct nilfs_sc_info *sci)
 			if (bh == segbuf->sb_super_root) {
 				if (bh->b_page != bd_page) {
 					lock_page(bd_page);
+					wait_on_page_writeback(bd_page);
 					clear_page_dirty_for_io(bd_page);
 					set_page_writeback(bd_page);
 					unlock_page(bd_page);
@@ -1723,6 +1725,7 @@ static void nilfs_segctor_prepare_write(struct nilfs_sc_info *sci)
 	}
 	if (bd_page) {
 		lock_page(bd_page);
+		wait_on_page_writeback(bd_page);
 		clear_page_dirty_for_io(bd_page);
 		set_page_writeback(bd_page);
 		unlock_page(bd_page);
Signature · 1 line
2.43.0
#1 lore raw
Quoted 38 lines · Ryusuke Konishi

commit a4ca369ca221bb7e06c725792ac107f0e48e82e7 upstream.

Destructive writes to a block device on which nilfs2 is mounted can cause a kernel bug in the folio/page writeback start routine or writeback end routine (__folio_start_writeback in the log below):

kernel BUG at mm/page-writeback.c:3070!

 Oops: invalid opcode: 0000 [#1] PREEMPT SMP KASAN PTI
 ...
 RIP: 0010:__folio_start_writeback+0xbaa/0x10e0
 Code: 25 ff 0f 00 00 0f 84 18 01 00 00 e8 40 ca c6 ff e9 17 f6 ff ff
  e8 36 ca c6 ff 4c 89 f7 48 c7 c6 80 c0 12 84 e8 e7 b3 0f 00 90 <0f>
  0b e8 1f ca c6 ff 4c 89 f7 48 c7 c6 a0 c6 12 84 e8 d0 b3 0f 00
 ...
 Call Trace:
  <TASK>
  nilfs_segctor_do_construct+0x4654/0x69d0 [nilfs2]
  nilfs_segctor_construct+0x181/0x6b0 [nilfs2]
  nilfs_segctor_thread+0x548/0x11c0 [nilfs2]
  kthread+0x2f0/0x390
  ret_from_fork+0x4b/0x80
  ret_from_fork_asm+0x1a/0x30
  </TASK>

This is because when the log writer starts a writeback for segment summary blocks or a super root block that use the backing device's page cache, it does not wait for the ongoing folio/page writeback, resulting in an inconsistent writeback state.

Fix this issue by waiting for ongoing writebacks when putting folios/pages on the backing device into writeback state.

Quoted 8 lines

---
Please apply this patch to the stable trees indicated by the subject prefix instead of the patch that failed.

This patch is tailored to account for page/folio conversion and can be applied to v6.7 and earlier.

Also, all the builds and tests I did on each stable tree passed.

Now queued up, thanks.

greg k-h