Skip to contentSkip to conversation

scsi: sg: fix slab-use-after-free Read in sg_release

PATCH Applied Suraj Sonawane 3 participants 7 messages Nov 20, 2024 – Dec 6, 2024

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 scsi: sg: fix slab-use-after-free Read in sg_release

Fix a use-after-free bug in sg_release,
detected by syzbot with KASAN:

BUG: KASAN: slab-use-after-free in lock_release+0x151/0xa30
kernel/locking/lockdep.c:5838
__mutex_unlock_slowpath+0xe2/0x750 kernel/locking/mutex.c:912
sg_release+0x1f4/0x2e0 drivers/scsi/sg.c:407

Root Cause:
In sg_release, the function kref_put(&sfp->f_ref, sg_remove_sfp) is called before releasing the open_rel_lock mutex. The kref_put call may decrement the reference count of sfp to zero, triggering its cleanup through sg_remove_sfp. This cleanup includes scheduling deferred work via sg_remove_sfp_usercontext, which ultimately frees sfp.

After kref_put, sg_release continues to unlock open_rel_lock and may reference sfp or sdp. If sfp has already been freed, this results in a slab-use-after-free error.

Fix:
The kref_put(&sfp->f_ref, sg_remove_sfp) call is moved after unlocking the open_rel_lock mutex. This ensures:
- No references to sfp or sdp occur after the reference count is decremented.
- Cleanup functions such as sg_remove_sfp and sg_remove_sfp_usercontext can safely execute without impacting the mutex handling in sg_release.

The fix has been tested and validated by syzbot. This patch closes the bug reported at the following syzkaller link and ensures proper sequencing of resource cleanup and mutex operations, eliminating the risk of use-after-free errors in sg_release.

Reported-by [email protected]
Fixes cc833acbee9d ("sg: O_EXCL and other lock handling ")
---
 drivers/scsi/sg.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/sg.c +1 -1
index f86be197f..457d54171 100644
@@ -393,7 +393,6 @@ sg_release(struct inode *inode, struct file *filp)
 
 	mutex_lock(&sdp->open_rel_lock);
 	scsi_autopm_put_device(sdp->device);
-	kref_put(&sfp->f_ref, sg_remove_sfp);
 	sdp->open_cnt--;
 
 	/* possibly many open()s waiting on exlude clearing, start many;
@@ -405,6 +404,7 @@ sg_release(struct inode *inode, struct file *filp)
 		wake_up_interruptible(&sdp->open_wait);
 	}
 	mutex_unlock(&sdp->open_rel_lock);
+	kref_put(&sfp->f_ref, sg_remove_sfp);
 	return 0;
 }
 
Signature · 1 line
2.34.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 13 lines · Suraj Sonawane
drivers/scsi/sg.c +0 -1
index f86be197f..457d54171 100644
@@ -393,7 +393,6 @@ sg_release(struct inode *inode, struct file *filp)
 
  	mutex_lock(&sdp->open_rel_lock);
  	scsi_autopm_put_device(sdp->device);
-	kref_put(&sfp->f_ref, sg_remove_sfp);
  	sdp->open_cnt--;
 
  	/* possibly many open()s waiting on exlude clearing, start many;

drivers/scsi/sg.c sg_release(struct inode *inode, struct file *filp) · line 409

Quoted 6 lines
drivers/scsi/sg.c +1 -0
@@ -405,6 +404,7 @@ sg_release(struct inode *inode, struct file *filp)
  		wake_up_interruptible(&sdp->open_wait);
  	}
  	mutex_unlock(&sdp->open_rel_lock);
+	kref_put(&sfp->f_ref, sg_remove_sfp);
  	return 0;
  }
#2 lore raw
In reply to Suraj Sonawane
Quoted 55 lines · Suraj Sonawane

Fix a use-after-free bug in sg_release,
detected by syzbot with KASAN:

BUG: KASAN: slab-use-after-free in lock_release+0x151/0xa30
kernel/locking/lockdep.c:5838
__mutex_unlock_slowpath+0xe2/0x750 kernel/locking/mutex.c:912
sg_release+0x1f4/0x2e0 drivers/scsi/sg.c:407

Root Cause:
In sg_release, the function kref_put(&sfp->f_ref, sg_remove_sfp) is called before releasing the open_rel_lock mutex. The kref_put call may decrement the reference count of sfp to zero, triggering its cleanup through sg_remove_sfp. This cleanup includes scheduling deferred work via sg_remove_sfp_usercontext, which ultimately frees sfp.

After kref_put, sg_release continues to unlock open_rel_lock and may reference sfp or sdp. If sfp has already been freed, this results in a slab-use-after-free error.

Fix:
The kref_put(&sfp->f_ref, sg_remove_sfp) call is moved after unlocking the open_rel_lock mutex. This ensures:
- No references to sfp or sdp occur after the reference count is decremented.
- Cleanup functions such as sg_remove_sfp and sg_remove_sfp_usercontext can safely execute without impacting the mutex handling in sg_release.

The fix has been tested and validated by syzbot. This patch closes the bug reported at the following syzkaller link and ensures proper sequencing of resource cleanup and mutex operations, eliminating the risk of use-after-free errors in sg_release.

Reported-by [email protected]
Fixes cc833acbee9d ("sg: O_EXCL and other lock handling ")
---
  drivers/scsi/sg.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/sg.c +0 -1
index f86be197f..457d54171 100644
@@ -393,7 +393,6 @@ sg_release(struct inode *inode, struct file *filp)
 
  	mutex_lock(&sdp->open_rel_lock);
  	scsi_autopm_put_device(sdp->device);
-	kref_put(&sfp->f_ref, sg_remove_sfp);
  	sdp->open_cnt--;
 
  	/* possibly many open()s waiting on exlude clearing, start many;

drivers/scsi/sg.c sg_release(struct inode *inode, struct file *filp) · line 410

Quoted 7 lines
drivers/scsi/sg.c +1 -0
@@ -405,6 +404,7 @@ sg_release(struct inode *inode, struct file *filp)
  		wake_up_interruptible(&sdp->open_wait);
  	}
  	mutex_unlock(&sdp->open_rel_lock);
+	kref_put(&sfp->f_ref, sg_remove_sfp);
  	return 0;
  }
 

Hello!

I wanted to follow up on the patch I submitted. I was wondering if you had a chance to review it and if there are any comments or feedback.

Thank you for your time and consideration. I look forward to your response.

Best regards,
Suraj Sonawane

#3 lore raw
Quoted 1 line · Suraj Sonawane

Hello!

Which person are you addressing with this email?

Quoted 1 line

I wanted to follow up on the patch I submitted. I was wondering if you had a chance to review it and if there are any comments or feedback.

Sending a ping after 5 days is too quick. I think that you should wait at least a week before sending a ping.

Bart.

#4 lore raw
Quoted 5 lines · Bart Van Assche
Quoted 1 line · Suraj Sonawane

Hello!

Which person are you addressing with this email?

I apologize for not clarifying earlier—my email was intended for everyone who might be reviewing the patch.

Quoted 5 lines
Quoted 1 line

I wanted to follow up on the patch I submitted. I was wondering if you had a chance to review it and if there are any comments or feedback.

Sending a ping after 5 days is too quick. I think that you should wait at least a week before sending a ping.

Thank you for pointing out that sending a follow-up after five days might be too soon. I truly appreciate your guidance on this. The reason for my follow-up is that, as a mentee in the Linux Kernel Bug Fixing Program, I need to include updates on my contributions in my progress report. I’ll make sure to wait at least a week before sending any future follow-ups.

Quoted 2 lines

Bart.

#5 lore raw
In reply to Suraj Sonawane
Quoted 9 lines · Suraj Sonawane

Fix a use-after-free bug in sg_release,
detected by syzbot with KASAN:

BUG: KASAN: slab-use-after-free in lock_release+0x151/0xa30
kernel/locking/lockdep.c:5838
__mutex_unlock_slowpath+0xe2/0x750 kernel/locking/mutex.c:912
sg_release+0x1f4/0x2e0 drivers/scsi/sg.c:407

[...]

Applied to 6.13/scsi-fixes, thanks!

[1/1] scsi: sg: fix slab-use-after-free Read in sg_release

      https://git.kernel.org/mkp/scsi/c/f10593ad9bc3
Signature · 1 line
Martin K. Petersen	Oracle Linux Engineering
#6 lore raw
Quoted 10 lines · Martin K. Petersen
Quoted 8 lines · Suraj Sonawane

Fix a use-after-free bug in sg_release,
detected by syzbot with KASAN:

BUG: KASAN: slab-use-after-free in lock_release+0x151/0xa30
kernel/locking/lockdep.c:5838
__mutex_unlock_slowpath+0xe2/0x750 kernel/locking/mutex.c:912
sg_release+0x1f4/0x2e0 drivers/scsi/sg.c:407
Quoted 7 lines
Quoted 1 line

[...]

Applied to 6.13/scsi-fixes, thanks!

[1/1] scsi: sg: fix slab-use-after-free Read in sg_release

       https://git.kernel.org/mkp/scsi/c/f10593ad9bc3

Thank you for applying the patch and for the update.

Best regards,
Suraj Sonawane