Skip to contentSkip to conversation

Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs

PATCH In review Zdenek Kabelac 3 participants 7 messages Sep 21, 2009 – Sep 25, 2009

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

PATCH 1/1 Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs

Subject: [PATCH] Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs
From: Zdenek Kabelac <[email protected]>

Adds missing blk_trace_remove_sysfs() to be in pair with blk_trace_init_sysfs() introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during device removal:

unreferenced object 0xffff88001aa76640 (size 80):
comm "lvcreate", pid 2120, jiffies 4294885144
hex dump (first 32 bytes):

    01 00 00 00 00 00 00 00 f0 65 a7 1a 00 88 ff ff  .........e......
    90 66 a7 1a 00 88 ff ff 86 1d 53 81 ff ff ff ff  .f........S.....
  backtrace:
    [<ffffffff813f9cc6>] kmemleak_alloc+0x26/0x60
    [<ffffffff8111d693>] kmem_cache_alloc+0x133/0x1c0
    [<ffffffff81195891>] sysfs_new_dirent+0x41/0x120
    [<ffffffff81194b0c>] sysfs_add_file_mode+0x3c/0xb0
    [<ffffffff81197c81>] internal_create_group+0xc1/0x1a0
    [<ffffffff81197d93>] sysfs_create_group+0x13/0x20
    [<ffffffff810d8004>] blk_trace_init_sysfs+0x14/0x20
    [<ffffffff8123f45c>] blk_register_queue+0x3c/0xf0
    [<ffffffff812447e4>] add_disk+0x94/0x160
    [<ffffffffa00d8b08>] dm_create+0x598/0x6e0 [dm_mod]
    [<ffffffffa00de951>] dev_create+0x51/0x350 [dm_mod]
    [<ffffffffa00de823>] ctl_ioctl+0x1a3/0x240 [dm_mod]
    [<ffffffffa00de8f2>] dm_compat_ctl_ioctl+0x12/0x20 [dm_mod]
    [<ffffffff81177bfd>] compat_sys_ioctl+0xcd/0x4f0
    [<ffffffff81036ed8>] sysenter_dispatch+0x7/0x2c
    [<ffffffffffffffff>] 0xffffffffffffffff
---
 block/blk-sysfs.c            |    2 ++
 include/linux/blktrace_api.h |    2 ++
 kernel/trace/blktrace.c      |    5 +++++
 3 files changed, 9 insertions(+), 0 deletions(-)
block/blk-sysfs.c +2 -0
index d3aa2aa..130b10c 100644
@@ -447,6 +447,7 @@ int blk_register_queue(struct gendisk *disk)
 	if (ret) {
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
+		blk_trace_remove_sysfs(disk_to_dev(disk));
 		return ret;
 	}
 
@@ -465,6 +466,7 @@ void blk_unregister_queue(struct gendisk *disk)
 
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
+		blk_trace_remove_sysfs(disk_to_dev(disk));
 		kobject_put(&disk_to_dev(disk)->kobj);
 	}
 }
include/linux/blktrace_api.h +0 -0
index 7e4350e..622939a 100644
@@ -198,6 +198,7 @@ extern int blk_trace_setup(struct request_queue

*q, char *name, dev_t dev,
char __user *arg);
extern int blk_trace_startstop(struct request_queue *q, int start); extern int blk_trace_remove(struct request_queue *q);
+extern void blk_trace_remove_sysfs(struct device *dev);
extern int blk_trace_init_sysfs(struct device *dev);

extern struct attribute_group blk_trace_attr_group;
@@ -211,6 +212,7 @@ extern struct attribute_group blk_trace_attr_group;

 # define blk_trace_startstop(q, start)			(-ENOTTY)
 # define blk_trace_remove(q)				(-ENOTTY)
 # define blk_add_trace_msg(q, fmt, ...)			do { } while (0)
+# define blk_trace_remove_sysfs(struct device *dev)	do { } while (0)
 static inline int blk_trace_init_sysfs(struct device *dev)
 {
 	return 0;
kernel/trace/blktrace.c +5 -0
index 7a34cb5..b5b8e35 100644
@@ -1653,6 +1653,11 @@ int blk_trace_init_sysfs(struct device *dev)
 	return sysfs_create_group(&dev->kobj, &blk_trace_attr_group);
 }
 
+void blk_trace_remove_sysfs(struct device *dev)
+{
+	sysfs_remove_group(&dev->kobj, &blk_trace_attr_group);
+}
+
 #endif /* CONFIG_BLK_DEV_IO_TRACE */
 
 #ifdef CONFIG_EVENT_TRACING
Signature · 1 line
1.6.4.4
#1 lore raw
Quoted 11 lines · Zdenek Kabelac

Subject: [PATCH] Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs
From: Zdenek Kabelac <[email protected]>

Adds missing blk_trace_remove_sysfs() to be in pair with blk_trace_init_sysfs() introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during device removal:

Thanks for reporting and fixing this!

block/blk-sysfs.c void blk_unregister_queue(struct gendisk *disk) · line 469

Quoted 5 lines
block/blk-sysfs.c +1 -0
@@ -465,6 +466,7 @@ void blk_unregister_queue(struct gendisk *disk)
 
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
+		blk_trace_remove_sysfs(disk_to_dev(disk));

This should be moved outside of 'if'.

Quoted 3 lines
block/blk-sysfs.c +0 -0
 		kobject_put(&disk_to_dev(disk)->kobj);
 	}
 }
#2 lore raw

2009/9/22 Li Zefan <[email protected]>:

Quoted 15 lines
Quoted 11 lines · Zdenek Kabelac

Subject: [PATCH] Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs
From: Zdenek Kabelac <[email protected]>

Adds missing blk_trace_remove_sysfs() to be in pair with blk_trace_init_sysfs() introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during  device removal:

Thanks for reporting and fixing this!

Quoted 8 lines

block/blk-sysfs.c void blk_unregister_queue(struct gendisk *disk) · line 469

Quoted 5 lines
block/blk-sysfs.c +1 -0
@@ -465,6 +466,7 @@ void blk_unregister_queue(struct gendisk *disk)
 
               kobject_uevent(&q->kobj, KOBJ_REMOVE);
               kobject_del(&q->kobj);
+             blk_trace_remove_sysfs(disk_to_dev(disk));

This should be moved outside of 'if'.

I was not really sure about the proper place - if it could be placed before if() or after the if(){} - as I've not checked in depth connection between kobj and sysfs. It's somewhat unclear why all the kobject operation are only within this if(){} block - so I've thought there is some reason...
IMHO only elv_unregister_queue() should be probably in the if(){} block.

Feel free to update/fix.

Zdenek

#3 lore raw
Quoted 15 lines
Quoted 15 lines
Quoted 11 lines

Subject: [PATCH] Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs
From: Zdenek Kabelac <[email protected]>

Adds missing blk_trace_remove_sysfs() to be in pair with blk_trace_init_sysfs() introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during device removal:

Thanks for reporting and fixing this!

Quoted 2 lines
(quoted diff) +0 -0
@@ -465,6 +466,7 @@ void blk_unregister_queue(struct gendisk *disk)
 
Quoted 8 lines
Quoted 5 lines

block/blk-sysfs.c int blk_register_queue(struct gendisk *disk) · line 450

Quoted 3 lines
block/blk-sysfs.c +1 -0
              kobject_uevent(&q->kobj, KOBJ_REMOVE);
              kobject_del(&q->kobj);
+             blk_trace_remove_sysfs(disk_to_dev(disk));

This should be moved outside of 'if'.

I was not really sure about the proper place - if it could be placed before if() or after the if(){} - as I've not checked in depth

Just use the reverse order against blk_register_queue() should be fine.

Quoted 5 lines

connection between kobj and sysfs. It's somewhat unclear why all the kobject operation are only within this if(){} block - so I've thought there is some reason...
IMHO only elv_unregister_queue() should be probably in the if(){} block.

Seems it's a bug to put kobject_put(dev->kobj) in the if block.

I created a stacked device (mdadm) and kmemleak still reported leaks even after I fixed the blktrace issue. And then I moved kobejct_put() outside the if, no more leaks reports.

Quoted 2 lines

Feel free to update/fix.

#4 lore raw

2009/9/24 Li Zefan <[email protected]>:

Quoted 21 lines
Quoted 21 lines
Quoted 20 lines
Quoted 11 lines

Subject: [PATCH] Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs
From: Zdenek Kabelac <[email protected]>

Adds missing blk_trace_remove_sysfs() to be in pair with blk_trace_init_sysfs() introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during  device removal:

Thanks for reporting and fixing this!

Review void blk_unregister_queue(struct gendisk *disk) · line 469

Quoted 5 lines
(quoted diff) +1 -0
@@ -465,6 +466,7 @@ void blk_unregister_queue(struct gendisk *disk)
 
               kobject_uevent(&q->kobj, KOBJ_REMOVE);
               kobject_del(&q->kobj);
+             blk_trace_remove_sysfs(disk_to_dev(disk));

This should be moved outside of 'if'.

Quoted 4 lines
Quoted 2 lines

I was not really sure about the proper place - if it could be placed before if() or after the if(){} - as I've not checked in depth

Just use the reverse order against blk_register_queue() should be fine.

Yes - I think the order of release is correct.

Quoted 7 lines
Quoted 6 lines

connection between kobj and sysfs. It's somewhat unclear why all the kobject operation are only within this if(){} block - so I've thought there is some reason...
IMHO only elv_unregister_queue() should be probably in the if(){} block.

Quoted 6 lines

Seems it's a bug to put kobject_put(dev->kobj) in the if block.

I created a stacked device (mdadm) and kmemleak still reported leaks even after I fixed the blktrace issue. And then I moved kobejct_put() outside the if, no more leaks reports.

Ok - I didn't have a testcase where the request_fn would be NULL. So in this case I propose this patch:

---
Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82. Release kobject also in case the request_fn is NULL.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during device removal:

unreferenced object 0xffff88001aa76640 (size 80):
comm "lvcreate", pid 2120, jiffies 4294885144
hex dump (first 32 bytes):

    01 00 00 00 00 00 00 00 f0 65 a7 1a 00 88 ff ff  .........e......
    90 66 a7 1a 00 88 ff ff 86 1d 53 81 ff ff ff ff  .f........S.....
  backtrace:
    [<ffffffff813f9cc6>] kmemleak_alloc+0x26/0x60
    [<ffffffff8111d693>] kmem_cache_alloc+0x133/0x1c0
    [<ffffffff81195891>] sysfs_new_dirent+0x41/0x120
    [<ffffffff81194b0c>] sysfs_add_file_mode+0x3c/0xb0
    [<ffffffff81197c81>] internal_create_group+0xc1/0x1a0
    [<ffffffff81197d93>] sysfs_create_group+0x13/0x20
    [<ffffffff810d8004>] blk_trace_init_sysfs+0x14/0x20
    [<ffffffff8123f45c>] blk_register_queue+0x3c/0xf0
    [<ffffffff812447e4>] add_disk+0x94/0x160
    [<ffffffffa00d8b08>] dm_create+0x598/0x6e0 [dm_mod]
    [<ffffffffa00de951>] dev_create+0x51/0x350 [dm_mod]
    [<ffffffffa00de823>] ctl_ioctl+0x1a3/0x240 [dm_mod]
    [<ffffffffa00de8f2>] dm_compat_ctl_ioctl+0x12/0x20 [dm_mod]
    [<ffffffff81177bfd>] compat_sys_ioctl+0xcd/0x4f0
    [<ffffffff81036ed8>] sysenter_dispatch+0x7/0x2c
    [<ffffffffffffffff>] 0xffffffffffffffff
---
 block/blk-sysfs.c            |   11 ++++++-----
 include/linux/blktrace_api.h |    2 ++
 kernel/trace/blktrace.c      |    5 +++++
 3 files changed, 13 insertions(+), 5 deletions(-)
block/blk-sysfs.c +6 -5
index b78c9c3..8a6d81a 100644
@@ -452,6 +452,7 @@ int blk_register_queue(struct gendisk *disk)
 	if (ret) {
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
+		blk_trace_remove_sysfs(disk_to_dev(disk));
 		return ret;
 	}
 
@@ -465,11 +466,11 @@ void blk_unregister_queue(struct gendisk *disk)
 	if (WARN_ON(!q))
 		return;
 
-	if (q->request_fn) {
+	if (q->request_fn)
 		elv_unregister_queue(q);
 
-		kobject_uevent(&q->kobj, KOBJ_REMOVE);
-		kobject_del(&q->kobj);
-		kobject_put(&disk_to_dev(disk)->kobj);
-	}
+	kobject_uevent(&q->kobj, KOBJ_REMOVE);
+	kobject_del(&q->kobj);
+	blk_trace_remove_sysfs(disk_to_dev(disk));
+	kobject_put(&disk_to_dev(disk)->kobj);
 }
include/linux/blktrace_api.h +0 -0
index 7e4350e..622939a 100644
@@ -198,6 +198,7 @@ extern int blk_trace_setup(struct request_queue

*q, char *name, dev_t dev,
char __user *arg);
extern int blk_trace_startstop(struct request_queue *q, int start); extern int blk_trace_remove(struct request_queue *q);
+extern void blk_trace_remove_sysfs(struct device *dev);
extern int blk_trace_init_sysfs(struct device *dev);

extern struct attribute_group blk_trace_attr_group;
@@ -211,6 +212,7 @@ extern struct attribute_group blk_trace_attr_group;

 # define blk_trace_startstop(q, start)			(-ENOTTY)
 # define blk_trace_remove(q)				(-ENOTTY)
 # define blk_add_trace_msg(q, fmt, ...)			do { } while (0)
+# define blk_trace_remove_sysfs(struct device *dev)	do { } while (0)
 static inline int blk_trace_init_sysfs(struct device *dev)
 {
 	return 0;
kernel/trace/blktrace.c +5 -0
index 3eb159c..60b5c5a 100644
@@ -1657,6 +1657,11 @@ int blk_trace_init_sysfs(struct device *dev)
 	return sysfs_create_group(&dev->kobj, &blk_trace_attr_group);
 }
 
+void blk_trace_remove_sysfs(struct device *dev)
+{
+	sysfs_remove_group(&dev->kobj, &blk_trace_attr_group);
+}
+
 #endif /* CONFIG_BLK_DEV_IO_TRACE */
 
 #ifdef CONFIG_EVENT_TRACING
Signature · 1 line
1.6.4.4
Reviewed-by
#5 lore raw
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

connection between kobj and sysfs. It's somewhat unclear why all the kobject operation are only within this if(){} block - so I've thought there is some reason...
IMHO only elv_unregister_queue() should be probably in the if(){} block.

Quoted 8 lines
Quoted 6 lines

Seems it's a bug to put kobject_put(dev->kobj) in the if block.

I created a stacked device (mdadm) and kmemleak still reported leaks even after I fixed the blktrace issue. And then I moved kobejct_put() outside the if, no more leaks reports.

Ok - I didn't have a testcase where the request_fn would be NULL.

Here you are ;)

mdadm --create --level=linear --raid-devices=1 --force /dev/md0 /dev/sda6

Quoted 2 lines

So in this case I propose this patch:

Looks good.

Quoted 104 lines

---
Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82. Release kobject also in case the request_fn is NULL.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during device removal:

unreferenced object 0xffff88001aa76640 (size 80):
comm "lvcreate", pid 2120, jiffies 4294885144
hex dump (first 32 bytes):

    01 00 00 00 00 00 00 00 f0 65 a7 1a 00 88 ff ff  .........e......
    90 66 a7 1a 00 88 ff ff 86 1d 53 81 ff ff ff ff  .f........S.....
  backtrace:
    [<ffffffff813f9cc6>] kmemleak_alloc+0x26/0x60
    [<ffffffff8111d693>] kmem_cache_alloc+0x133/0x1c0
    [<ffffffff81195891>] sysfs_new_dirent+0x41/0x120
    [<ffffffff81194b0c>] sysfs_add_file_mode+0x3c/0xb0
    [<ffffffff81197c81>] internal_create_group+0xc1/0x1a0
    [<ffffffff81197d93>] sysfs_create_group+0x13/0x20
    [<ffffffff810d8004>] blk_trace_init_sysfs+0x14/0x20
    [<ffffffff8123f45c>] blk_register_queue+0x3c/0xf0
    [<ffffffff812447e4>] add_disk+0x94/0x160
    [<ffffffffa00d8b08>] dm_create+0x598/0x6e0 [dm_mod]
    [<ffffffffa00de951>] dev_create+0x51/0x350 [dm_mod]
    [<ffffffffa00de823>] ctl_ioctl+0x1a3/0x240 [dm_mod]
    [<ffffffffa00de8f2>] dm_compat_ctl_ioctl+0x12/0x20 [dm_mod]
    [<ffffffff81177bfd>] compat_sys_ioctl+0xcd/0x4f0
    [<ffffffff81036ed8>] sysenter_dispatch+0x7/0x2c
    [<ffffffffffffffff>] 0xffffffffffffffff
---
 block/blk-sysfs.c            |   11 ++++++-----
 include/linux/blktrace_api.h |    2 ++
 kernel/trace/blktrace.c      |    5 +++++
 3 files changed, 13 insertions(+), 5 deletions(-)
block/blk-sysfs.c +6 -5
index b78c9c3..8a6d81a 100644
@@ -452,6 +452,7 @@ int blk_register_queue(struct gendisk *disk)
 	if (ret) {
 		kobject_uevent(&q->kobj, KOBJ_REMOVE);
 		kobject_del(&q->kobj);
+		blk_trace_remove_sysfs(disk_to_dev(disk));
 		return ret;
 	}
 
@@ -465,11 +466,11 @@ void blk_unregister_queue(struct gendisk *disk)
 	if (WARN_ON(!q))
 		return;
 
-	if (q->request_fn) {
+	if (q->request_fn)
 		elv_unregister_queue(q);
 
-		kobject_uevent(&q->kobj, KOBJ_REMOVE);
-		kobject_del(&q->kobj);
-		kobject_put(&disk_to_dev(disk)->kobj);
-	}
+	kobject_uevent(&q->kobj, KOBJ_REMOVE);
+	kobject_del(&q->kobj);
+	blk_trace_remove_sysfs(disk_to_dev(disk));
+	kobject_put(&disk_to_dev(disk)->kobj);
 }
include/linux/blktrace_api.h +1 -0
index 7e4350e..622939a 100644
@@ -198,6 +198,7 @@ extern int blk_trace_setup(struct request_queue
 *q, char *name, dev_t dev,
 			   char __user *arg);
 extern int blk_trace_startstop(struct request_queue *q, int start);
 extern int blk_trace_remove(struct request_queue *q);
+extern void blk_trace_remove_sysfs(struct device *dev);
 extern int blk_trace_init_sysfs(struct device *dev);
 

extern struct attribute_group blk_trace_attr_group;

include/linux/blktrace_api.h +1 -0
@@ -211,6 +212,7 @@ extern struct attribute_group blk_trace_attr_group;
 # define blk_trace_startstop(q, start)			(-ENOTTY)
 # define blk_trace_remove(q)				(-ENOTTY)
 # define blk_add_trace_msg(q, fmt, ...)			do { } while (0)
+# define blk_trace_remove_sysfs(struct device *dev)	do { } while (0)
 static inline int blk_trace_init_sysfs(struct device *dev)
 {
 	return 0;
kernel/trace/blktrace.c +5 -0
index 3eb159c..60b5c5a 100644
@@ -1657,6 +1657,11 @@ int blk_trace_init_sysfs(struct device *dev)
 	return sysfs_create_group(&dev->kobj, &blk_trace_attr_group);
 }
 
+void blk_trace_remove_sysfs(struct device *dev)
+{
+	sysfs_remove_group(&dev->kobj, &blk_trace_attr_group);
+}
+
 #endif /* CONFIG_BLK_DEV_IO_TRACE */
 
 #ifdef CONFIG_EVENT_TRACING
#6 lore raw
In reply to Zdenek Kabelac
Quoted 46 lines · Zdenek Kabelac

2009/9/24 Li Zefan <[email protected]>:

Quoted 21 lines
Quoted 21 lines
Quoted 20 lines
Quoted 11 lines

Subject: [PATCH] Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs
From: Zdenek Kabelac <[email protected]>

Adds missing blk_trace_remove_sysfs() to be in pair with blk_trace_init_sysfs() introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during  device removal:

Thanks for reporting and fixing this!

block/blk-sysfs.c void blk_unregister_queue(struct gendisk *disk) · line 469

Quoted 5 lines
block/blk-sysfs.c +1 -0
@@ -465,6 +466,7 @@ void blk_unregister_queue(struct gendisk *disk)
 
               kobject_uevent(&q->kobj, KOBJ_REMOVE);
               kobject_del(&q->kobj);
+             blk_trace_remove_sysfs(disk_to_dev(disk));

This should be moved outside of 'if'.

Quoted 4 lines
Quoted 2 lines

I was not really sure about the proper place - if it could be placed before if() or after the if(){} - as I've not checked in depth

Just use the reverse order against blk_register_queue() should be fine.

Yes - I think the order of release is correct.

Quoted 7 lines
Quoted 6 lines

connection between kobj and sysfs. It's somewhat unclear why all the kobject operation are only within this if(){} block - so I've thought there is some reason...
IMHO only elv_unregister_queue() should be probably in the if(){} block.

Quoted 6 lines

Seems it's a bug to put kobject_put(dev->kobj) in the if block.

I created a stacked device (mdadm) and kmemleak still reported leaks even after I fixed the blktrace issue. And then I moved kobejct_put() outside the if, no more leaks reports.

Ok - I didn't have a testcase where the request_fn would be NULL. So in this case I propose this patch:

Quoted 7 lines

---
Add missing blk_trace_remove_sysfs to be in pair with blk_trace_init_sysfs introduced in commit 1d54ad6da9192fed5dd3b60224d9f2dfea0dcd82. Release kobject also in case the request_fn is NULL.

Problem was noticed via kmemleak backtrace when some sysfs entries were note properly destroyed during device removal:

Thanks, this looks good now, applied.

Signature · 1 line
Jens Axboe