Skip to contentSkip to conversation

media: atomisp: gc2235: fix UAF and memory leak

PATCH In review Yuho Choi 2 participants 4 messages Apr 3 – Apr 15

Subsystems (4)
lore
Share & tools
lore

Yuho Choi

author
#0 lore raw

PATCH v4 1/1 media: atomisp: gc2235: fix UAF and memory leak

gc2235_probe() handles its error paths incorrectly.

If media_entity_pads_init() fails, gc2235_remove() is called, which tears down the subdev and frees dev, but then still falls through to atomisp_register_i2c_module(). This results in use-after-free.

If atomisp_register_i2c_module() fails, the media entity and control handler are left initialized and dev is leaked.

gc2235_remove() unconditionally calls media_entity_cleanup() and v4l2_ctrl_handler_free(), but these are not initialized at every error path in gc2235_probe().

Replace gc2235_remove() calls in the probe error paths with explicit unwind labels that free only the resources initialized at each point of failure, in reverse order of initialization.

Fixes a49d25364dfb ("staging/atomisp: Add support for the Intel IPU v2")

---
Changes since v3:
- Replaced goto out_free/gc2235_remove() with explicit unwind labels to release only initialized resources at each failure point
- Replaced the "Fixes" tag with the original commit for the driver

Changes since v2:
- Replaced gc2235_remove() calls in remaining two error paths with goto labels to unwind only initialized resources
- Added Fixes tag

Changes since v1:
- Edited the commit message to be imperative mood
- Corrected the previous mangled patch

 .../media/atomisp/i2c/atomisp-gc2235.c        | 29 ++++++++++++-------
 1 file changed, 18 insertions(+), 11 deletions(-)
drivers/staging/media/atomisp/i2c/atomisp-gc2235.c +18 -11
index d3414312e1de2..998c9f46bd068 100644
@@ -809,7 +809,7 @@ static int gc2235_probe(struct i2c_client *client)
 
 	ret = gc2235_s_config(&dev->sd, client->irq, gcpdev);
 	if (ret)
-		goto out_free;
+		goto err_unregister_subdev;
 
 	dev->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE;
 	dev->pad.flags = MEDIA_PAD_FL_SOURCE;
@@ -818,18 +818,16 @@ static int gc2235_probe(struct i2c_client *client)
 	ret =
 	    v4l2_ctrl_handler_init(&dev->ctrl_handler,
 				   ARRAY_SIZE(gc2235_controls));
-	if (ret) {
-		gc2235_remove(client);
-		return ret;
-	}
+	if (ret)
+		goto err_csi_cfg;
 
 	for (i = 0; i < ARRAY_SIZE(gc2235_controls); i++)
 		v4l2_ctrl_new_custom(&dev->ctrl_handler, &gc2235_controls[i],
 				     NULL);
 
 	if (dev->ctrl_handler.error) {
-		gc2235_remove(client);
-		return dev->ctrl_handler.error;
+		ret = dev->ctrl_handler.error;
+		goto err_ctrl_handler;
 	}
 
 	/* Use same lock for controls as for everything else. */
@@ -838,14 +836,23 @@ static int gc2235_probe(struct i2c_client *client)
 
 	ret = media_entity_pads_init(&dev->sd.entity, 1, &dev->pad);
 	if (ret)
-		gc2235_remove(client);
+		goto err_ctrl_handler;
+
+	ret = atomisp_register_i2c_module(&dev->sd, gcpdev);
+	if (ret)
+		goto err_media_cleanup;
 
-	return atomisp_register_i2c_module(&dev->sd, gcpdev);
+	return 0;
 
-out_free:
+err_media_cleanup:
+	media_entity_cleanup(&dev->sd.entity);
+err_ctrl_handler:
+	v4l2_ctrl_handler_free(&dev->ctrl_handler);
+err_csi_cfg:
+	dev->platform_data->csi_cfg(&dev->sd, 0);
+err_unregister_subdev:
 	v4l2_device_unregister_subdev(&dev->sd);
 	kfree(dev);
-
 	return ret;
 }
 
Signature · 1 line
2.50.1 (Apple Git-155)
author
#1 lore raw

Dear Reviewers,

I am following up on my recent patch submission, "[PATCH v4] media:
atomisp: gc2235: fix
UAF and memory leak."

Please let me know if you have had a chance to review it or if any further changes are
required.

Best regards,
Yuho

Quoted 105 lines · Yuho Choi

gc2235_probe() handles its error paths incorrectly.

If media_entity_pads_init() fails, gc2235_remove() is called, which tears down the subdev and frees dev, but then still falls through to atomisp_register_i2c_module(). This results in use-after-free.

If atomisp_register_i2c_module() fails, the media entity and control handler are left initialized and dev is leaked.

gc2235_remove() unconditionally calls media_entity_cleanup() and v4l2_ctrl_handler_free(), but these are not initialized at every error path in gc2235_probe().

Replace gc2235_remove() calls in the probe error paths with explicit unwind labels that free only the resources initialized at each point of failure, in reverse order of initialization.

Fixes a49d25364dfb ("staging/atomisp: Add support for the Intel IPU v2")

---
Changes since v3:
- Replaced goto out_free/gc2235_remove() with explicit unwind labels to release only initialized resources at each failure point
- Replaced the "Fixes" tag with the original commit for the driver

Changes since v2:
- Replaced gc2235_remove() calls in remaining two error paths with goto labels to unwind only initialized resources
- Added Fixes tag

Changes since v1:
- Edited the commit message to be imperative mood
- Corrected the previous mangled patch

 .../media/atomisp/i2c/atomisp-gc2235.c        | 29 ++++++++++++-------
 1 file changed, 18 insertions(+), 11 deletions(-)
drivers/staging/media/atomisp/i2c/atomisp-gc2235.c +18 -12
index d3414312e1de2..998c9f46bd068 100644
@@ -809,7 +809,7 @@ static int gc2235_probe(struct i2c_client *client)
 
        ret = gc2235_s_config(&dev->sd, client->irq, gcpdev);
        if (ret)
-               goto out_free;
+               goto err_unregister_subdev;
 
        dev->sd.flags |= V4L2_SUBDEV_FL_HAS_DEVNODE;
        dev->pad.flags = MEDIA_PAD_FL_SOURCE;
@@ -818,18 +818,16 @@ static int gc2235_probe(struct i2c_client *client)
        ret =
            v4l2_ctrl_handler_init(&dev->ctrl_handler,
                                   ARRAY_SIZE(gc2235_controls));
-       if (ret) {
-               gc2235_remove(client);
-               return ret;
-       }
+       if (ret)
+               goto err_csi_cfg;
 
        for (i = 0; i < ARRAY_SIZE(gc2235_controls); i++)
                v4l2_ctrl_new_custom(&dev->ctrl_handler, &gc2235_controls[i],
                                     NULL);
 
        if (dev->ctrl_handler.error) {
-               gc2235_remove(client);
-               return dev->ctrl_handler.error;
+               ret = dev->ctrl_handler.error;
+               goto err_ctrl_handler;
        }
 
        /* Use same lock for controls as for everything else. */
@@ -838,14 +836,23 @@ static int gc2235_probe(struct i2c_client *client)
 
        ret = media_entity_pads_init(&dev->sd.entity, 1, &dev->pad);
        if (ret)
-               gc2235_remove(client);
+               goto err_ctrl_handler;
+
+       ret = atomisp_register_i2c_module(&dev->sd, gcpdev);
+       if (ret)
+               goto err_media_cleanup;
 
-       return atomisp_register_i2c_module(&dev->sd, gcpdev);
+       return 0;
 
-out_free:
+err_media_cleanup:
+       media_entity_cleanup(&dev->sd.entity);
+err_ctrl_handler:
+       v4l2_ctrl_handler_free(&dev->ctrl_handler);
+err_csi_cfg:
+       dev->platform_data->csi_cfg(&dev->sd, 0);
+err_unregister_subdev:
        v4l2_device_unregister_subdev(&dev->sd);
        kfree(dev);
-
        return ret;
 }
 
--

2.50.1 (Apple Git-155)

Reviewed-by on v4 1/1
#2 lore raw
In reply to Yuho Choi
Quoted 13 lines · Yuho Choi

gc2235_probe() handles its error paths incorrectly.

If media_entity_pads_init() fails, gc2235_remove() is called, which tears down the subdev and frees dev, but then still falls through to atomisp_register_i2c_module(). This results in use-after-free.

If atomisp_register_i2c_module() fails, the media entity and control handler are left initialized and dev is leaked.

gc2235_remove() unconditionally calls media_entity_cleanup() and v4l2_ctrl_handler_free(), but these are not initialized at every error path in gc2235_probe().

Quoted 7 lines

Replace gc2235_remove() calls in the probe error paths with explicit unwind labels that free only the resources initialized at each point of failure, in reverse order of initialization.

Fixes a49d25364dfb ("staging/atomisp: Add support for the Intel IPU v2")

---

Thanks. LGTM!

regards,
dan carpenter

author
#3 lore raw

Dear Dan,

Thanks for the review. I do appreciate the LGTM.

Best regards,
Yuho Choi

Quoted 30 lines · Dan Carpenter
Quoted 13 lines · Yuho Choi

gc2235_probe() handles its error paths incorrectly.

If media_entity_pads_init() fails, gc2235_remove() is called, which tears down the subdev and frees dev, but then still falls through to atomisp_register_i2c_module(). This results in use-after-free.

If atomisp_register_i2c_module() fails, the media entity and control handler are left initialized and dev is leaked.

gc2235_remove() unconditionally calls media_entity_cleanup() and v4l2_ctrl_handler_free(), but these are not initialized at every error path in gc2235_probe().

Quoted 7 lines

Replace gc2235_remove() calls in the probe error paths with explicit unwind labels that free only the resources initialized at each point of failure, in reverse order of initialization.

Fixes a49d25364dfb ("staging/atomisp: Add support for the Intel IPU v2")

---

Thanks. LGTM!

regards,
dan carpenter