Skip to contentSkip to conversation

FAILED: patch "[PATCH] Btrfs: fix early ENOSPC due to delalloc" failed to apply to 4.12-stable tree

PATCH [email protected] 3 participants 3 messages Aug 1, 2017 – Aug 11, 2017

Subsystems (2)
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 4.12-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

Quoted 1 line

From 17024ad0a0fdfcfe53043afb969b813d3e020c21 Mon Sep 17 00:00:00 2001

From: Omar Sandoval <[email protected]>
Date: Thu, 20 Jul 2017 15:10:35 -0700
Subject: [PATCH] Btrfs: fix early ENOSPC due to delalloc

If a lot of metadata is reserved for outstanding delayed allocations, we rely on shrink_delalloc() to reclaim metadata space in order to fulfill reservation tickets. However, shrink_delalloc() has a shortcut where if it determines that space can be overcommitted, it will stop early. This made sense before the ticketed enospc system, but now it means that shrink_delalloc() will often not reclaim enough space to fulfill any tickets, leading to an early ENOSPC. (Reservation tickets don't care about being able to overcommit, they need every byte accounted for.)

Fix it by getting rid of the shortcut so that shrink_delalloc() reclaims all of the metadata it is supposed to. This fixes early ENOSPCs we were seeing when doing a btrfs receive to populate a new filesystem, as well as early ENOSPCs Christoph saw when doing a big cp -r onto Btrfs.

Fixes 957780eb2788 ("Btrfs: introduce ticketed enospc infrastructure")
fs/btrfs/extent-tree.c +0 -4
index a6635f07b8f1..e3b0b4196d3d 100644
@@ -4825,10 +4825,6 @@ static void shrink_delalloc(struct btrfs_fs_info *fs_info, u64 to_reclaim,
 		else
 			flush = BTRFS_RESERVE_NO_FLUSH;
 		spin_lock(&space_info->lock);
-		if (can_overcommit(fs_info, space_info, orig, flush, false)) {
-			spin_unlock(&space_info->lock);
-			break;
-		}
 		if (list_empty(&space_info->tickets) &&
 		    list_empty(&space_info->priority_tickets)) {
 			spin_unlock(&space_info->lock);
#1 lore raw

Hey.

Could someone of the devs put some attention on this...?

Thanks,
Chris :-)

Quoted 66 lines · On Mon, 2017-07-31 at 18:06 -0700, [email protected] wrote:

The patch below does not apply to the 4.12-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit
id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 17024ad0a0fdfcfe53043afb969b813d3e020c21 Mon Sep 17 00:00:00 2001
From: Omar Sandoval <[email protected]>
Date: Thu, 20 Jul 2017 15:10:35 -0700
Subject: [PATCH] Btrfs: fix early ENOSPC due to delalloc

If a lot of metadata is reserved for outstanding delayed allocations, we
rely on shrink_delalloc() to reclaim metadata space in order to fulfill
reservation tickets. However, shrink_delalloc() has a shortcut where if
it determines that space can be overcommitted, it will stop early. This
made sense before the ticketed enospc system, but now it means that shrink_delalloc() will often not reclaim enough space to fulfill any tickets, leading to an early ENOSPC. (Reservation tickets don't care about being able to overcommit, they need every byte accounted for.)

Fix it by getting rid of the shortcut so that shrink_delalloc() reclaims
all of the metadata it is supposed to. This fixes early ENOSPCs we were
seeing when doing a btrfs receive to populate a new filesystem, as well
as early ENOSPCs Christoph saw when doing a big cp -r onto Btrfs.

Fixes 957780eb2788 ("Btrfs: introduce ticketed enospc

infrastructure")

me>

fs/btrfs/extent-tree.c +0 -0
index a6635f07b8f1..e3b0b4196d3d 100644
@@ -4825,10 +4825,6 @@ static void shrink_delalloc(struct
 btrfs_fs_info *fs_info, u64 to_reclaim,

  else
  flush = BTRFS_RESERVE_NO_FLUSH; spin_lock(&space_info->lock);

fs/btrfs/extent-tree.c +0 -1
-		if (can_overcommit(fs_info, space_info, orig, flush,

false)) {

fs/btrfs/extent-tree.c +0 -3
-			spin_unlock(&space_info->lock);
-			break;
-		}

  if (list_empty(&space_info->tickets) && list_empty(&space_info->priority_tickets)) { spin_unlock(&space_info->lock);

#2 lore raw
Quoted 6 lines · Christoph Anton Mitterer

Hey.

Could someone of the devs put some attention on this...?

Thanks,
Chris :-)

Done, you can also grab it here:

https://git.kernel.org/pub/scm/linux/kernel/git/mason/linux-btrfs.git/commit/?h=for-stable-4.12&id=354850ad1948af13248031e5180d495044d05aa5

-chris

Quoted 68 lines · On Mon, 2017-07-31 at 18:06 -0700, [email protected] wrote:
Quoted 65 lines

The patch below does not apply to the 4.12-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit
id to <[email protected]>.

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From 17024ad0a0fdfcfe53043afb969b813d3e020c21 Mon Sep 17 00:00:00
2001
From: Omar Sandoval <[email protected]>
Date: Thu, 20 Jul 2017 15:10:35 -0700
Subject: [PATCH] Btrfs: fix early ENOSPC due to delalloc

If a lot of metadata is reserved for outstanding delayed allocations, we
rely on shrink_delalloc() to reclaim metadata space in order to fulfill
reservation tickets. However, shrink_delalloc() has a shortcut where if
it determines that space can be overcommitted, it will stop early. This
made sense before the ticketed enospc system, but now it means that shrink_delalloc() will often not reclaim enough space to fulfill any tickets, leading to an early ENOSPC. (Reservation tickets don't care about being able to overcommit, they need every byte accounted for.)

Fix it by getting rid of the shortcut so that shrink_delalloc() reclaims
all of the metadata it is supposed to. This fixes early ENOSPCs we were
seeing when doing a btrfs receive to populate a new filesystem, as well
as early ENOSPCs Christoph saw when doing a big cp -r onto Btrfs.

Fixes 957780eb2788 ("Btrfs: introduce ticketed enospc

infrastructure")

me>

fs/btrfs/extent-tree.c +0 -4
index a6635f07b8f1..e3b0b4196d3d 100644
@@ -4825,10 +4825,6 @@ static void shrink_delalloc(struct
 btrfs_fs_info *fs_info, u64 to_reclaim,
  		else
  			flush = BTRFS_RESERVE_NO_FLUSH;
  		spin_lock(&space_info->lock);
-		if (can_overcommit(fs_info, space_info, orig, flush,
 false)) {
-			spin_unlock(&space_info->lock);
-			break;
-		}
  		if (list_empty(&space_info->tickets) &&
  		    list_empty(&space_info->priority_tickets)) {
  			spin_unlock(&space_info->lock);