Skip to contentSkip to conversation

null_blk: fix kmemleak by releasing references to fault configfs items

PATCH In review Nilay Shroff 3 participants 4 messages Jan 12 – Jan 13 Block layer

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 null_blk: fix kmemleak by releasing references to fault configfs items

When CONFIG_BLK_DEV_NULL_BLK_FAULT_INJECTION is enabled, the null-blk driver sets up fault injection support by creating the timeout_inject, requeue_inject, and init_hctx_fault_inject configfs items as children of the top-level nullb configfs group.

However, when the nullb device is removed, the references taken to these fault-config configfs items are not released. As a result, kmemleak reports a memory leak, for example:

unreferenced object 0xc00000021ff25c40 (size 32):
comm "mkdir", pid 10665, jiffies 4322121578
hex dump (first 32 bytes):

    69 6e 69 74 5f 68 63 74 78 5f 66 61 75 6c 74 5f  init_hctx_fault_
    69 6e 6a 65 63 74 00 88 00 00 00 00 00 00 00 00  inject..........
  backtrace (crc 1a018c86):
    __kmalloc_node_track_caller_noprof+0x494/0xbd8
    kvasprintf+0x74/0xf4
    config_item_set_name+0xf0/0x104
    config_group_init_type_name+0x48/0xfc
    fault_config_init+0x48/0xf0
    0xc0080000180559e4
    configfs_mkdir+0x304/0x814
    vfs_mkdir+0x49c/0x604
    do_mkdirat+0x314/0x3d0
    sys_mkdir+0xa0/0xd8
    system_call_exception+0x1b0/0x4f0
    system_call_vectored_common+0x15c/0x2ec

Fix this by explicitly releasing the references to the fault-config configfs items when dropping the reference to the top-level nullb configfs group.

---
 drivers/block/null_blk/main.c | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
drivers/block/null_blk/main.c +11 -1
index c7c0fb79a6bf..4c0632ab4e1b 100644
@@ -665,12 +665,22 @@ static void nullb_add_fault_config(struct nullb_device *dev)
 	configfs_add_default_group(&dev->init_hctx_fault_config.group, &dev->group);
 }
 
+static void nullb_del_fault_config(struct nullb_device *dev)
+{
+	config_item_put(&dev->init_hctx_fault_config.group.cg_item);
+	config_item_put(&dev->requeue_config.group.cg_item);
+	config_item_put(&dev->timeout_config.group.cg_item);
+}
+
 #else
 
 static void nullb_add_fault_config(struct nullb_device *dev)
 {
 }
 
+static void nullb_del_fault_config(struct nullb_device *dev)
+{
+}
 #endif
 
 static struct
@@ -702,7 +712,7 @@ nullb_group_drop_item(struct config_group *group, struct config_item *item)
 		null_del_dev(dev->nullb);
 		mutex_unlock(&lock);
 	}
-
+	nullb_del_fault_config(dev);
 	config_item_put(item);
 }
 
Signature · 1 line
2.52.0
maintainer
#1 lore raw
Quoted 28 lines · Nilay Shroff

When CONFIG_BLK_DEV_NULL_BLK_FAULT_INJECTION is enabled, the null-blk driver sets up fault injection support by creating the timeout_inject, requeue_inject, and init_hctx_fault_inject configfs items as children of the top-level nullb configfs group.

However, when the nullb device is removed, the references taken to these fault-config configfs items are not released. As a result, kmemleak reports a memory leak, for example:

unreferenced object 0xc00000021ff25c40 (size 32):
comm "mkdir", pid 10665, jiffies 4322121578
hex dump (first 32 bytes):

    69 6e 69 74 5f 68 63 74 78 5f 66 61 75 6c 74 5f  init_hctx_fault_
    69 6e 6a 65 63 74 00 88 00 00 00 00 00 00 00 00  inject..........
  backtrace (crc 1a018c86):
    __kmalloc_node_track_caller_noprof+0x494/0xbd8
    kvasprintf+0x74/0xf4
    config_item_set_name+0xf0/0x104
    config_group_init_type_name+0x48/0xfc
    fault_config_init+0x48/0xf0
    0xc0080000180559e4
    configfs_mkdir+0x304/0x814
    vfs_mkdir+0x49c/0x604
    do_mkdirat+0x314/0x3d0
    sys_mkdir+0xa0/0xd8
    system_call_exception+0x1b0/0x4f0
    system_call_vectored_common+0x15c/0x2ec
Quoted 3 lines

Fix this by explicitly releasing the references to the fault-config configfs items when dropping the reference to the top-level nullb configfs group.

Seems like this should have a fixes and stable tag, too?

Signature · 1 line
Jens Axboe
#2 lore raw
Quoted 29 lines · Jens Axboe
Quoted 28 lines · Nilay Shroff

When CONFIG_BLK_DEV_NULL_BLK_FAULT_INJECTION is enabled, the null-blk driver sets up fault injection support by creating the timeout_inject, requeue_inject, and init_hctx_fault_inject configfs items as children of the top-level nullb configfs group.

However, when the nullb device is removed, the references taken to these fault-config configfs items are not released. As a result, kmemleak reports a memory leak, for example:

unreferenced object 0xc00000021ff25c40 (size 32):
comm "mkdir", pid 10665, jiffies 4322121578
hex dump (first 32 bytes):

    69 6e 69 74 5f 68 63 74 78 5f 66 61 75 6c 74 5f  init_hctx_fault_
    69 6e 6a 65 63 74 00 88 00 00 00 00 00 00 00 00  inject..........
  backtrace (crc 1a018c86):
    __kmalloc_node_track_caller_noprof+0x494/0xbd8
    kvasprintf+0x74/0xf4
    config_item_set_name+0xf0/0x104
    config_group_init_type_name+0x48/0xfc
    fault_config_init+0x48/0xf0
    0xc0080000180559e4
    configfs_mkdir+0x304/0x814
    vfs_mkdir+0x49c/0x604
    do_mkdirat+0x314/0x3d0
    sys_mkdir+0xa0/0xd8
    system_call_exception+0x1b0/0x4f0
    system_call_vectored_common+0x15c/0x2ec
Quoted 6 lines
Quoted 3 lines

Fix this by explicitly releasing the references to the fault-config configfs items when dropping the reference to the top-level nullb configfs group.

Seems like this should have a fixes and stable tag, too?

Yeah, I will send out v2 with these changes.

Thanks,
--Nilay

Reviewed-by on v1 1/1
#3 lore raw
In reply to Nilay Shroff
Quoted 28 lines · Nilay Shroff

When CONFIG_BLK_DEV_NULL_BLK_FAULT_INJECTION is enabled, the null-blk driver sets up fault injection support by creating the timeout_inject, requeue_inject, and init_hctx_fault_inject configfs items as children of the top-level nullb configfs group.

However, when the nullb device is removed, the references taken to these fault-config configfs items are not released. As a result, kmemleak reports a memory leak, for example:

unreferenced object 0xc00000021ff25c40 (size 32):

   comm "mkdir", pid 10665, jiffies 4322121578
   hex dump (first 32 bytes):
     69 6e 69 74 5f 68 63 74 78 5f 66 61 75 6c 74 5f  init_hctx_fault_
     69 6e 6a 65 63 74 00 88 00 00 00 00 00 00 00 00  inject..........
   backtrace (crc 1a018c86):
     __kmalloc_node_track_caller_noprof+0x494/0xbd8
     kvasprintf+0x74/0xf4
     config_item_set_name+0xf0/0x104
     config_group_init_type_name+0x48/0xfc
     fault_config_init+0x48/0xf0
     0xc0080000180559e4
     configfs_mkdir+0x304/0x814
     vfs_mkdir+0x49c/0x604
     do_mkdirat+0x314/0x3d0
     sys_mkdir+0xa0/0xd8
     system_call_exception+0x1b0/0x4f0
     system_call_vectored_common+0x15c/0x2ec
Quoted 5 lines

Fix this by explicitly releasing the references to the fault-config configfs items when dropping the reference to the top-level nullb configfs group.

with fixes tag added :-

Looks good.

-ck