When reclaiming a suballocator block group, first reduce the on-disk
cluster count by cl_cpg.  The current code then subtracts that new
count (fe->i_clusters) from the old cached count
(OCFS2_I(alloc_inode)->ip_clusters).

For an allocator with N block groups, that leaves the cache at

    N * cl_cpg - (N * cl_cpg - cl_cpg) = cl_cpg

i.e. ip_clusters -= (fe->i_clusters - cl_cpg) leaves ip_clusters equal
to cl_cpg regardless of N.  This happens to be correct when reclaiming
from two block groups, but undercounts the clusters from three block
groups onwards.  The incorrect cache value is also used immediately to
update i_blocks.

Assign the updated on-disk count to the cache, matching the allocation
and inode refresh paths.

In a QEMU test using a clean 256 MiB OCFS2 image and a 10,000-file
create/delete workload, the first buggy reclaim left the on-disk
(fe->i_clusters) and cached (ip_clusters) counts at 2048 and 512
clusters respectively; later reclaims underflowed the cache.  With this
change, the cache matched the on-disk count across all four reclaims:
2048, 1536, 1024, and 512 clusters.

Fixes: 4a54331616b3 ("ocfs2: give ocfs2 the ability to reclaim suballocator free bg")
Cc: stable@vger.kernel.org
Signed-off-by: Matthias Goergens <matthias.goergens@gmail.com>
---

v2: commit-log wording revisions suggested by Heming Zhao; code unchanged.
 fs/ocfs2/suballoc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
index a4a2b87a45fe3..20c3aec6b9873 100644
--- a/fs/ocfs2/suballoc.c
+++ b/fs/ocfs2/suballoc.c
@@ -2759,7 +2759,7 @@ static int _ocfs2_reclaim_suballoc_to_main(handle_t *handle,
 	fe->i_clusters = cpu_to_le32(tmp_used - le16_to_cpu(cl->cl_cpg));
 
 	spin_lock(&OCFS2_I(alloc_inode)->ip_lock);
-	OCFS2_I(alloc_inode)->ip_clusters -= le32_to_cpu(fe->i_clusters);
+	OCFS2_I(alloc_inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
 	fe->i_size = cpu_to_le64(ocfs2_clusters_to_bytes(alloc_inode->i_sb,
 					     le32_to_cpu(fe->i_clusters)));
 	spin_unlock(&OCFS2_I(alloc_inode)->ip_lock);
-- 
2.55.0

On 8/5/26 7:39 PM, Matthias Goergens wrote:
> When reclaiming a suballocator block group, first reduce the on-disk
> cluster count by cl_cpg.  The current code then subtracts that new
> count (fe->i_clusters) from the old cached count
> (OCFS2_I(alloc_inode)->ip_clusters).
> 
> For an allocator with N block groups, that leaves the cache at
> 
>     N * cl_cpg - (N * cl_cpg - cl_cpg) = cl_cpg
> 
> i.e. ip_clusters -= (fe->i_clusters - cl_cpg) leaves ip_clusters equal
> to cl_cpg regardless of N.  This happens to be correct when reclaiming
> from two block groups, but undercounts the clusters from three block
> groups onwards.  The incorrect cache value is also used immediately to
> update i_blocks.
> 
> Assign the updated on-disk count to the cache, matching the allocation
> and inode refresh paths.
> 
> In a QEMU test using a clean 256 MiB OCFS2 image and a 10,000-file
> create/delete workload, the first buggy reclaim left the on-disk
> (fe->i_clusters) and cached (ip_clusters) counts at 2048 and 512
> clusters respectively; later reclaims underflowed the cache.  With this
> change, the cache matched the on-disk count across all four reclaims:
> 2048, 1536, 1024, and 512 clusters.
> 
> Fixes: 4a54331616b3 ("ocfs2: give ocfs2 the ability to reclaim suballocator free bg")
> Cc: stable@vger.kernel.org
> Signed-off-by: Matthias Goergens <matthias.goergens@gmail.com>

Reviewed-by: Joseph Qi <joseph.qi@linux.alibaba.com>

> ---
> 
> v2: commit-log wording revisions suggested by Heming Zhao; code unchanged.
>  fs/ocfs2/suballoc.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c
> index a4a2b87a45fe3..20c3aec6b9873 100644
> --- a/fs/ocfs2/suballoc.c
> +++ b/fs/ocfs2/suballoc.c
> @@ -2759,7 +2759,7 @@ static int _ocfs2_reclaim_suballoc_to_main(handle_t *handle,
>  	fe->i_clusters = cpu_to_le32(tmp_used - le16_to_cpu(cl->cl_cpg));
>  
>  	spin_lock(&OCFS2_I(alloc_inode)->ip_lock);
> -	OCFS2_I(alloc_inode)->ip_clusters -= le32_to_cpu(fe->i_clusters);
> +	OCFS2_I(alloc_inode)->ip_clusters = le32_to_cpu(fe->i_clusters);
>  	fe->i_size = cpu_to_le64(ocfs2_clusters_to_bytes(alloc_inode->i_sb,
>  					     le32_to_cpu(fe->i_clusters)));
>  	spin_unlock(&OCFS2_I(alloc_inode)->ip_lock);