Skip to contentSkip to conversation

gfs2: Fix error path kobject memory leak

PATCH In review Tobin C. Harding 4 participants 5 messages May 13, 2019 GFS2 file system

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 gfs2: Fix error path kobject memory leak

If a call to kobject_init_and_add() fails we must call kobject_put() otherwise we leak memory.

Function always calls kobject_init_and_add() which always calls kobject_init().

It is safe to leave object destruction up to the kobject release function and never free it manually.

Remove call to kfree() and always call kobject_put() in the error path.

---

Is it ok to send patches during the merge window?

Applies on top of Linus' mainline tag: v5.1

Happy to rebase if there are conflicts.

thanks,
Tobin.

 fs/gfs2/sys.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
fs/gfs2/sys.c +1 -6
index 1787d295834e..98586b139386 100644
@@ -661,8 +661,6 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 	if (error)
 		goto fail_reg;
 
-	sysfs_frees_sdp = 1; /* Freeing sdp is now done by sysfs calling
-				function gfs2_sbd_release. */
 	error = sysfs_create_group(&sdp->sd_kobj, &tune_group);
 	if (error)
 		goto fail_reg;
@@ -687,10 +685,7 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 fail_reg:
 	free_percpu(sdp->sd_lkstats);
 	fs_err(sdp, "error %d adding sysfs files\n", error);
-	if (sysfs_frees_sdp)
-		kobject_put(&sdp->sd_kobj);
-	else
-		kfree(sdp);
+	kobject_put(&sdp->sd_kobj);
 	sb->s_fs_info = NULL;
 	return error;
 }
Signature · 1 line
2.21.0
Reviewed-by on v1 1/1
#1 lore raw
Quoted 31 lines · Tobin C. Harding

If a call to kobject_init_and_add() fails we must call kobject_put() otherwise we leak memory.

Function always calls kobject_init_and_add() which always calls kobject_init().

It is safe to leave object destruction up to the kobject release function and never free it manually.

Remove call to kfree() and always call kobject_put() in the error path.

---

Is it ok to send patches during the merge window?

Applies on top of Linus' mainline tag: v5.1

Happy to rebase if there are conflicts.

thanks,
Tobin.

 fs/gfs2/sys.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
fs/gfs2/sys.c +0 -0
index 1787d295834e..98586b139386 100644

fs/gfs2/sys.c int gfs2_sys_fs_add(struct gfs2_sbd *sdp) · line 665

Quoted 5 lines
fs/gfs2/sys.c +0 -2
@@ -661,8 +661,6 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 	if (error)
 		goto fail_reg;
 
-	sysfs_frees_sdp = 1; /* Freeing sdp is now done by sysfs calling
-				function gfs2_sbd_release. */

You should also delete this variable at the top of the function, as it is now only set once there and never used.

With that:

#2 lore raw
Quoted 35 lines · Greg Kroah-Hartman
Quoted 34 lines · Tobin C. Harding

If a call to kobject_init_and_add() fails we must call kobject_put() otherwise we leak memory.

Function always calls kobject_init_and_add() which always calls kobject_init().

It is safe to leave object destruction up to the kobject release function and never free it manually.

Remove call to kfree() and always call kobject_put() in the error path.

---

Is it ok to send patches during the merge window?

Applies on top of Linus' mainline tag: v5.1

Happy to rebase if there are conflicts.

thanks,
Tobin.

 fs/gfs2/sys.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
fs/gfs2/sys.c +0 -0
index 1787d295834e..98586b139386 100644
@@ -661,8 +661,6 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
 	if (error)
 		goto fail_reg;
 
Quoted 5 lines

fs/gfs2/sys.c int gfs2_sys_fs_add(struct gfs2_sbd *sdp) · line 665

Quoted 2 lines
fs/gfs2/sys.c +0 -2
-	sysfs_frees_sdp = 1; /* Freeing sdp is now done by sysfs calling
-				function gfs2_sbd_release. */

You should also delete this variable at the top of the function, as it is now only set once there and never used.

Thanks, I should have gotten a compiler warning for that. I was feeling so confident with my builds this morning ... pays not to get too cocky I suppose.

Quoted 3 lines

With that:

Thanks, will re-spin.

Tobin.

#3 lore raw
Quoted 47 lines · Tobin C. Harding
Quoted 35 lines · Greg Kroah-Hartman
Quoted 34 lines · Tobin C. Harding

If a call to kobject_init_and_add() fails we must call kobject_put() otherwise we leak memory.

Function always calls kobject_init_and_add() which always calls kobject_init().

It is safe to leave object destruction up to the kobject release function and never free it manually.

Remove call to kfree() and always call kobject_put() in the error path.

---

Is it ok to send patches during the merge window?

Applies on top of Linus' mainline tag: v5.1

Happy to rebase if there are conflicts.

thanks,
Tobin.

 fs/gfs2/sys.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
fs/gfs2/sys.c +0 -0
index 1787d295834e..98586b139386 100644
@@ -661,8 +661,6 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
    if (error)
            goto fail_reg;
 
Quoted 5 lines

fs/gfs2/sys.c int gfs2_sys_fs_add(struct gfs2_sbd *sdp) · line 665

Quoted 2 lines
fs/gfs2/sys.c +0 -2
-   sysfs_frees_sdp = 1; /* Freeing sdp is now done by sysfs calling
-                           function gfs2_sbd_release. */

You should also delete this variable at the top of the function, as it is now only set once there and never used.

Thanks, I should have gotten a compiler warning for that. I was feeling so confident with my builds this morning ... pays not to get too cocky I suppose.

Quoted 5 lines
Quoted 3 lines

With that:

Thanks, will re-spin.

Don't bother, I'll fix that up.

Thanks,
Andreas

#4 lore raw
Quoted 48 lines · Andreas Gruenbacher
Quoted 47 lines · Tobin C. Harding
Quoted 35 lines · Greg Kroah-Hartman
Quoted 34 lines · Tobin C. Harding

If a call to kobject_init_and_add() fails we must call kobject_put() otherwise we leak memory.

Function always calls kobject_init_and_add() which always calls kobject_init().

It is safe to leave object destruction up to the kobject release function and never free it manually.

Remove call to kfree() and always call kobject_put() in the error path.

---

Is it ok to send patches during the merge window?

Applies on top of Linus' mainline tag: v5.1

Happy to rebase if there are conflicts.

thanks,
Tobin.

 fs/gfs2/sys.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)
fs/gfs2/sys.c +0 -0
index 1787d295834e..98586b139386 100644
@@ -661,8 +661,6 @@ int gfs2_sys_fs_add(struct gfs2_sbd *sdp)
    if (error)
            goto fail_reg;
 
Quoted 5 lines

fs/gfs2/sys.c int gfs2_sys_fs_add(struct gfs2_sbd *sdp) · line 665

Quoted 2 lines
fs/gfs2/sys.c +0 -2
-   sysfs_frees_sdp = 1; /* Freeing sdp is now done by sysfs calling
-                           function gfs2_sbd_release. */

You should also delete this variable at the top of the function, as it is now only set once there and never used.

Thanks, I should have gotten a compiler warning for that. I was feeling so confident with my builds this morning ... pays not to get too cocky I suppose.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

With that:

Thanks, will re-spin.

Don't bother, I'll fix that up.

Thanks man!