Don't ignore return values from ext4_ext_dirty, since the errors
indicate valid failures below Ext4. In all of the other instances of
ext4_ext_dirty calls, the error return value is handled in some
way. This patch makes those remaining couple of places to handle
ext4_ext_dirty errors as well. In the longer run, we probably should
make sure that errors from other mark_dirty routines are handled as
well.
Ran gce-xfstests smoke tests and verified that there were no
regressions.
Signed-off-by: Harshad Shirwadkar <[email protected]>
---
fs/ext4/extents.c | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
index f2b577b315a0..f62f55a16fe3 100644
--- a/fs/ext4/extents.c
+++ b/fs/ext4/extents.c
@@ -3244,8 +3244,7 @@ static int ext4_split_extent_at(handle_t *handle,
fix_extent_len:
ex->ee_len = orig_ex.ee_len;
- ext4_ext_dirty(handle, inode, path + path->p_depth);
- return err;
+ return ext4_ext_dirty(handle, inode, path + path->p_depth);
}
/*
@@ -3503,7 +3502,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
}
if (allocated) {
/* Mark the block containing both extents as dirty */
- ext4_ext_dirty(handle, inode, path + depth);
+ err = ext4_ext_dirty(handle, inode, path + depth);
/* Update path to point to the right extent */
path[depth].p_ext = abut_ex;
--
2.26.1.301.g55bc3eb7cb9-goog
If eh->eh_max is 0, EXT_MAX_EXTENT/INDEX would evaluate to unsigned
(-1) resulting in illegal memory accesses. Although there is no
consistent repro, we see that generic/019 sometimes crashes because of
this bug.
Ran gce-xfstests smoke and verified that there were no regressions.
Signed-off-by: Harshad Shirwadkar <[email protected]>
---
fs/ext4/ext4_extents.h | 9 ++++++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/fs/ext4/ext4_extents.h b/fs/ext4/ext4_extents.h
index 1c216fcc202a..44e59881a1f0 100644
--- a/fs/ext4/ext4_extents.h
+++ b/fs/ext4/ext4_extents.h
@@ -170,10 +170,13 @@ struct partial_cluster {
(EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_entries) - 1)
#define EXT_LAST_INDEX(__hdr__) \
(EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_entries) - 1)
-#define EXT_MAX_EXTENT(__hdr__) \
- (EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)
+#define EXT_MAX_EXTENT(__hdr__) \
+ ((le16_to_cpu((__hdr__)->eh_max)) ? \
+ ((EXT_FIRST_EXTENT((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)) \
+ : 0)
#define EXT_MAX_INDEX(__hdr__) \
- (EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)
+ ((le16_to_cpu((__hdr__)->eh_max)) ? \
+ ((EXT_FIRST_INDEX((__hdr__)) + le16_to_cpu((__hdr__)->eh_max) - 1)) : 0)
static inline struct ext4_extent_header *ext_inode_hdr(struct inode *inode)
{
--
2.26.1.301.g55bc3eb7cb9-goog
On Mon, Apr 20, 2020 at 7:40 PM Harshad Shirwadkar
<[email protected]> wrote:
>
> Don't ignore return values from ext4_ext_dirty, since the errors
> indicate valid failures below Ext4. In all of the other instances of
> ext4_ext_dirty calls, the error return value is handled in some
> way. This patch makes those remaining couple of places to handle
> ext4_ext_dirty errors as well. In the longer run, we probably should
> make sure that errors from other mark_dirty routines are handled as
> well.
>
> Ran gce-xfstests smoke tests and verified that there were no
> regressions.
>
> Signed-off-by: Harshad Shirwadkar <[email protected]>
> ---
> fs/ext4/extents.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
> index f2b577b315a0..f62f55a16fe3 100644
> --- a/fs/ext4/extents.c
> +++ b/fs/ext4/extents.c
> @@ -3244,8 +3244,7 @@ static int ext4_split_extent_at(handle_t *handle,
>
> fix_extent_len:
> ex->ee_len = orig_ex.ee_len;
> - ext4_ext_dirty(handle, inode, path + path->p_depth);
> - return err;
> + return ext4_ext_dirty(handle, inode, path + path->p_depth);
> }
I realized that this is not correct. That's because fix_extent_len is
an error path and this change would make ext4_split_extent_at() return
success if ext4_ext_dirty succeeds in this path. I think instead I
should be adding a comment here explaining why we are not handling
ext4_ext_dirty(). Sorry for that.
- Harshad
>
> /*
> @@ -3503,7 +3502,7 @@ static int ext4_ext_convert_to_initialized(handle_t *handle,
> }
> if (allocated) {
> /* Mark the block containing both extents as dirty */
> - ext4_ext_dirty(handle, inode, path + depth);
> + err = ext4_ext_dirty(handle, inode, path + depth);
>
> /* Update path to point to the right extent */
> path[depth].p_ext = abut_ex;
> --
> 2.26.1.301.g55bc3eb7cb9-goog
>
On Mon, Apr 20, 2020 at 07:39:59PM -0700, Harshad Shirwadkar wrote:
> If eh->eh_max is 0, EXT_MAX_EXTENT/INDEX would evaluate to unsigned
> (-1) resulting in illegal memory accesses. Although there is no
> consistent repro, we see that generic/019 sometimes crashes because of
> this bug.
>
> Ran gce-xfstests smoke and verified that there were no regressions.
>
> Signed-off-by: Harshad Shirwadkar <[email protected]>
Applied, thanks.
- Ted