driver core: Fix use-after-free and double free on glue directory
Subsystems (2)
PATCH 1/1 driver core: Fix use-after-free and double free on glue directory
There is a race condition between removing glue directory and adding a new device under the glue directory. It can be reproduced in following test:
path 1: Add the child device under glue dirdevice_add()
get_device_parent()
mutex_lock(&gdp_mutex);
....
/*find parent from glue_dirs.list*/
list_for_each_entry(k, &dev->class->p->glue_dirs.list, entry)
if (k->parent == parent_kobj) {
kobj = kobject_get(k);
break;
}
....
mutex_unlock(&gdp_mutex);
....
....
kobject_add()
kobject_add_internal()
create_dir()
sysfs_create_dir_ns()
if (kobj->parent)
parent = kobj->parent->sd;
....
kernfs_create_dir_ns(parent)
kernfs_new_node()
kernfs_get(parent)
....
/* link in */
rc = kernfs_add_one(kn);
if (!rc)
return kn;
kernfs_put(kn)
....
repeat:
kmem_cache_free(kn)
kn = parent;
if (kn) {
if (atomic_dec_and_test(&kn->count))
goto repeat;
}
....path2: Remove last child device under glue dirdevice_del()
cleanup_device_parent()
cleanup_glue_dir()
mutex_lock(&gdp_mutex);
if (!kobject_has_children(glue_dir))
kobject_del(glue_dir);
kobject_put(glue_dir);
mutex_unlock(&gdp_mutex);Before path2 remove last child device under glue dir, If path1 add a new device under glue dir, the glue_dir kobject reference count will be increase to 2 via kobject_get(k) in get_device_parent(). And path1 has been called kernfs_new_node(), but not call kernfs_get(parent). Meanwhile, path2 call kobject_del(glue_dir) beacause 0 is returned by kobject_has_children(). This result in glue_dir->sd is freed and it's reference count will be 0. Then path1 call kernfs_get(parent) will trigger a warning in kernfs_get()(WARN_ON(!atomic_read(&kn->count))) and increase it's reference count to 1. Because glue_dir->sd is freed by path2, the next call kernfs_add_one() by path1 will fail(This is also use-after-free) and call atomic_dec_and_test() to decrease reference count. Because the reference count is decremented to 0, it will also call kmem_cache_free() to free glue_dir->sd again. This will result in double free.
In order to avoid this happening, we we should not call kobject_del() on path2 when the reference count of glue_dir is greater than 1. So we add a conditional statement to fix it.
The following calltrace is captured in kernel 4.14 with the following patch applied:
commit 726e41097920 ("drivers: core: Remove glue dirs from sysfs earlier")
--------------------------------------------------------------------------
[ 3.633703] WARNING: CPU: 4 PID: 513 at .../fs/kernfs/dir.c:494
Here is WARN_ON(!atomic_read(&kn->count) in kernfs_get().
....
[ 3.633986] Call trace:
[ 3.633991] kernfs_create_dir_ns+0xa8/0xb0
[ 3.633994] sysfs_create_dir_ns+0x54/0xe8
[ 3.634001] kobject_add_internal+0x22c/0x3f0
[ 3.634005] kobject_add+0xe4/0x118
[ 3.634011] device_add+0x200/0x870
[ 3.634017] _request_firmware+0x958/0xc38
[ 3.634020] request_firmware_into_buf+0x4c/0x70
....
[ 3.634064] kernel BUG at .../mm/slub.c:294!
Hrer is BUG_ON(object == fp) in set_freepointer().
....
[ 3.634346] Call trace:
[ 3.634351] kmem_cache_free+0x504/0x6b8
[ 3.634355] kernfs_put+0x14c/0x1d8
[ 3.634359] kernfs_create_dir_ns+0x88/0xb0
[ 3.634362] sysfs_create_dir_ns+0x54/0xe8
[ 3.634366] kobject_add_internal+0x22c/0x3f0
[ 3.634370] kobject_add+0xe4/0x118
[ 3.634374] device_add+0x200/0x870
[ 3.634378] _request_firmware+0x958/0xc38
[ 3.634381] request_firmware_into_buf+0x4c/0x70--------------------------------------------------------------------------
--- drivers/base/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/base/core.c +1 -1
@@ -1825,7 +1825,7 @@ static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)return ;mutex_lock (&gdp_mutex); -if (!kobject_has_children (glue_dir)) +if (!kobject_has_children (glue_dir) &&kref_read (&glue_dir->kref) ==1 )kobject_del (glue_dir);kobject_put (glue_dir);mutex_unlock (&gdp_mutex);
Signature · 1 line
2.17.1
Quoted 3 lines · Muchun Song
There is a race condition between removing glue directory and adding a new device under the glue directory. It can be reproduced in following test:
.../...
Quoted 3 lines
In order to avoid this happening, we we should not call
kobject_del()on path2 when the reference count ofglue_diris greater than 1. So we add a conditional statement to fix it.
Good catch ! However I'm not completely happy about the fix you propose.
I find relying on the object count for such decisions rather fragile as it could be taken temporarily for other reasons, couldn't it ? In which case we would just fail...
Ideally, the looking up of the glue dir and creation of its child should be protected by the same lock instance (the gdp_mutex in that case).
That might require a bit of shuffling around though.
Greg, thoughts ? This whole gluedir business is annoyingly racy still.
My gut feeling is that the "right fix" is to ensure the lookup of the glue dir and creation of the child object(s) are done under a single instance of gdp_mutex so we never see a stale "empty" but still poentially used glue dir around.
This should also be true when creating such gluedir in the first place in fact, though that race is a lot harder to hit.
Cheers,
Ben.
Hi Cheers,
Quoted 12 lines · Benjamin Herrenschmidt <[email protected]> 于2019年4月25日周四 下午5:24写道:
Quoted 3 lines · Muchun Song
There is a race condition between removing glue directory and adding a new device under the glue directory. It can be reproduced in following test:
.../...
Quoted 3 lines
In order to avoid this happening, we we should not call
kobject_del()on path2 when the reference count ofglue_diris greater than 1. So we add a conditional statement to fix it.
Quoted 6 lines
Good catch ! However I'm not completely happy about the fix you propose.
I find relying on the object count for such decisions rather fragile as it could be taken temporarily for other reasons, couldn't it ? In which case we would just fail...
It could be taken temporarily for other reasons, what reasons? I also can not figure out which case could result in this.
Quoted 5 lines
Ideally, the looking up of the glue dir and creation of its child should be protected by the same lock instance (the
gdp_mutexin that case).
Quoted 8 lines
That might require a bit of shuffling around though.
Greg, thoughts ? This whole gluedir business is annoyingly racy still.
My gut feeling is that the "right fix" is to ensure the lookup of the glue dir and creation of the child object(s) are done under a single instance of
gdp_mutexso we never see a stale "empty" but still poentially used glue dir around.
I agree with you that the looking up of the glue dir and creation of its child should be protected by the same lock of gdp_mutex. So, do you agree with the fix of the following code snippet?
drivers/base/core.c +0 -0
@@ -1740,8 +1740,11 @@ class_dir_create_and_add(struct class *class,
struct kobject *parent_kobj)
static DEFINE_MUTEX(gdp_mutex);
static struct kobject *get_device_parent(struct device *dev,
- struct device *parent)
+ struct device *parent,
+ bool *locked)
{
+ *locked = false;
+
if (dev->class) {
struct kobject *kobj = NULL;
struct kobject *parent_kobj;
@@ -1779,7 +1782,7 @@ static struct kobject *get_device_parent(struct
device *dev,
}
spin_unlock(&dev->class->p->glue_dirs.list_lock);
if (kobj) {
- mutex_unlock(&gdp_mutex);
+ *locked = true;
return kobj;
}@@ -2007,6 +2010,7 @@ int device_add(struct device *dev)
struct class_interface *class_intf;
int error = -EINVAL;
struct kobject *glue_dir = NULL;+ bool locked;
dev = get_device(dev);
if (!dev)
@@ -2040,7 +2044,7 @@ int device_add(struct device *dev)
pr_debug("device: '%s': %s\n", dev_name(dev), __func__);
parent = get_device(dev->parent);
- kobj = get_device_parent(dev, parent);
+ kobj = get_device_parent(dev, parent, &locked);
if (IS_ERR(kobj)) {
error = PTR_ERR(kobj);
goto parent_error;
@@ -2057,9 +2061,14 @@ int device_add(struct device *dev)
error = kobject_add(&dev->kobj, dev->kobj.parent, NULL);
if (error) {
glue_dir = get_glue_dir(dev);
+ if (locked)
+ mutex_unlock(&gdp_mutex);
goto Error;
}+ if (locked)
+ mutex_unlock(&gdp_mutex);
+
/* notify platform of device entry */
error = device_platform_notify(dev, KOBJ_ADD);
if (error)Yours
Muchun
Quoted 3 lines · Muchun Song
I agree with you that the looking up of the glue dir and creation of its child should be protected by the same lock of
gdp_mutex. So, do you agree with the fix of the following code snippet?
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it work to have a get_device_parent_locked that always returns with the mutex held, or just move the mutex to the caller or something simpler like this ?
Ben.
Hi Greg and Rafael:
Quoted 5 lines · Benjamin Herrenschmidt <[email protected]> 于2019年4月28日周日 下午6:10写道:
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it work to have a
get_device_parent_lockedthat always returns with the mutex held, or just move the mutex to the caller or something simpler like this ?
Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Yours,
Muchun
Quoted 14 lines · Muchun Song
Hi Greg and Rafael:
Benjamin Herrenschmidt <[email protected]> 于2019年4月28日周日
Quoted 8 lines · 下午6:10写道:
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?
Quoted 2 lines
Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
Ben.
Quoted 8 lines · Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四 下午2:25写道:
Quoted 8 lines
Quoted 7 lines
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?
Quoted 4 lines
Quoted 2 lines
Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
From your suggestions, I think introduce get_device_parent_locked() may easy to fix. So, do you agree with the fix of the following code snippet (You can also
view attachments)?
I introduce a new function named get_device_parent_locked_if_glue_dir() which always returns with the mutex held only when we live in glue dir. We should call unlock_if_glue_dir() to release the mutex. The get_device_parent_locked_if_glue_dir()
and unlock_if_glue_dir() should be called in pairs.
---drivers/base/core.c | 44 ++++++++++++++++++++++++++++++++++++--------
1 file changed, 36 insertions(+), 8 deletions(-)
drivers/base/core.c +0 -0
@@ -1739,8 +1739,9 @@ class_dir_create_and_add(struct class *class,
struct kobject *parent_kobj)
static DEFINE_MUTEX(gdp_mutex);
-static struct kobject *get_device_parent(struct device *dev,
- struct device *parent)
+static struct kobject *__get_device_parent(struct device *dev,
+ struct device *parent,
+ bool lock)
{
if (dev->class) {
struct kobject *kobj = NULL;
@@ -1779,14 +1780,16 @@ static struct kobject
*get_device_parent(struct device *dev,
}
spin_unlock(&dev->class->p->glue_dirs.list_lock);
if (kobj) {
- mutex_unlock(&gdp_mutex);
+ if (!lock)
+ mutex_unlock(&gdp_mutex);
return kobj;
}
/* or create a new class-directory at the parent device */
k = class_dir_create_and_add(dev->class, parent_kobj);
/* do not emit an uevent for this simple "glue" directory */
- mutex_unlock(&gdp_mutex);
+ if (!lock)
+ mutex_unlock(&gdp_mutex);
return k; }
@@ -1799,6 +1802,19 @@ static struct kobject *get_device_parent(struct device *dev,
return NULL;
}
+static inline struct kobject *get_device_parent(struct device *dev,
+ struct device *parent)
+{
+ return __get_device_parent(dev, parent, false);
+}
+
+static inline struct kobject * +get_device_parent_locked_if_glue_dir(struct device *dev,
+ struct device *parent)
+{
+ return __get_device_parent(dev, parent, true);
+}
+
static inline bool live_in_glue_dir(struct kobject *kobj,
struct device *dev)
{
@@ -1831,6 +1847,16 @@ static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)
mutex_unlock(&gdp_mutex);
}
+static inline void unlock_if_glue_dir(struct device *dev,
+ struct kobject *glue_dir)
+{
+ /* see if we live in a "glue" directory */
+ if (!live_in_glue_dir(glue_dir, dev))
+ return;
+
+ mutex_unlock(&gdp_mutex);
+}
+
static int device_add_class_symlinks(struct device *dev)
{
struct device_node *of_node = dev_of_node(dev);
@@ -2040,7 +2066,7 @@ int device_add(struct device *dev)
pr_debug("device: '%s': %s\n", dev_name(dev), __func__); parent = get_device(dev->parent);
- kobj = get_device_parent(dev, parent);
+ kobj = get_device_parent_locked_if_glue_dir(dev, parent);
if (IS_ERR(kobj)) {
error = PTR_ERR(kobj);
goto parent_error;@@ -2055,10 +2081,12 @@ int device_add(struct device *dev)
/* first, register with generic layer. */
/* we require the name to be set before, and pass NULL */ error = kobject_add(&dev->kobj, dev->kobj.parent, NULL);
- if (error) {
- glue_dir = get_glue_dir(dev);
+
+ glue_dir = get_glue_dir(dev);
+ unlock_if_glue_dir(dev, glue_dir);
+
+ if (error)
goto Error;
- }
/* notify platform of device entry */
error = device_platform_notify(dev, KOBJ_ADD);
--
From 091a0800664201c1719b2aa1bcd965f9578ee13e Mon Sep 17 00:00:00 2001
From: Muchun Song <[email protected]>
Date: Sun, 28 Apr 2019 23:31:06 +0800
Subject: [PATCH] driver core: Fix use-after-free and double free on glue
directory
There is a race condition between removing glue directory and adding a new device under the glue directory. It can be reproduced in following test:
path 1: Add the child device under glue dirdevice_add()
get_device_parent()
mutex_lock(&gdp_mutex);
....
/*find parent from glue_dirs.list*/
list_for_each_entry(k, &dev->class->p->glue_dirs.list, entry)
if (k->parent == parent_kobj) {
kobj = kobject_get(k);
break;
}
....
mutex_unlock(&gdp_mutex);
....
....
kobject_add()
kobject_add_internal()
create_dir()
sysfs_create_dir_ns()
if (kobj->parent)
parent = kobj->parent->sd;
....
kernfs_create_dir_ns(parent)
kernfs_new_node()
kernfs_get(parent)
....
/* link in */
rc = kernfs_add_one(kn);
if (!rc)
return kn;
kernfs_put(kn)
....
repeat:
kmem_cache_free(kn)
kn = parent;
if (kn) {
if (atomic_dec_and_test(&kn->count))
goto repeat;
}
....path2: Remove last child device under glue dirdevice_del()
cleanup_device_parent()
cleanup_glue_dir()
mutex_lock(&gdp_mutex);
if (!kobject_has_children(glue_dir))
kobject_del(glue_dir);
kobject_put(glue_dir);
mutex_unlock(&gdp_mutex);Before path2 remove last child device under glue dir, If path1 add a new device under glue dir, the glue_dir kobject reference count will be increase to 2 via kobject_get(k) in get_device_parent(). And path1 has been called kernfs_new_node(), but not call kernfs_get(parent). Meanwhile, path2 call kobject_del(glue_dir) beacause 0 is returned by kobject_has_children(). This result in glue_dir->sd is freed and it's reference count will be 0. Then path1 call kernfs_get(parent) will trigger a warning in kernfs_get()(WARN_ON(!atomic_read(&kn->count))) and increase it's reference count to 1. Because glue_dir->sd is freed by path2, the next call kernfs_add_one() by path1 will fail(This is also use-after-free) and call atomic_dec_and_test() to decrease reference count. Because the reference count is decremented to 0, it will also call kmem_cache_free() to free glue_dir->sd again. This will result in double free.
In order to avoid this happening, we we should not call kobject_del() on path2 when the reference count of glue_dir is greater than 1. So we add a conditional statement to fix it.
The following calltrace is captured in kernel 4.14 with the following patch applied:
commit 726e41097920 ("drivers: core: Remove glue dirs from sysfs earlier")
--------------------------------------------------------------------------
[ 3.633703] WARNING: CPU: 4 PID: 513 at .../fs/kernfs/dir.c:494
Here is WARN_ON(!atomic_read(&kn->count) in kernfs_get().
....
[ 3.633986] Call trace:
[ 3.633991] kernfs_create_dir_ns+0xa8/0xb0
[ 3.633994] sysfs_create_dir_ns+0x54/0xe8
[ 3.634001] kobject_add_internal+0x22c/0x3f0
[ 3.634005] kobject_add+0xe4/0x118
[ 3.634011] device_add+0x200/0x870
[ 3.634017] _request_firmware+0x958/0xc38
[ 3.634020] request_firmware_into_buf+0x4c/0x70
....
[ 3.634064] kernel BUG at .../mm/slub.c:294!
Here is BUG_ON(object == fp) in set_freepointer().
....
[ 3.634346] Call trace:
[ 3.634351] kmem_cache_free+0x504/0x6b8
[ 3.634355] kernfs_put+0x14c/0x1d8
[ 3.634359] kernfs_create_dir_ns+0x88/0xb0
[ 3.634362] sysfs_create_dir_ns+0x54/0xe8
[ 3.634366] kobject_add_internal+0x22c/0x3f0
[ 3.634370] kobject_add+0xe4/0x118
[ 3.634374] device_add+0x200/0x870
[ 3.634378] _request_firmware+0x958/0xc38
[ 3.634381] request_firmware_into_buf+0x4c/0x70--------------------------------------------------------------------------
--- drivers/base/core.c | 44 ++++++++++++++++++++++++++++++++++++-------- 1 file changed, 36 insertions(+), 8 deletions(-)
drivers/base/core.c +36 -8
87 lines not on this page · raw message
Signature · 1 line
2.17.1
Greg KH
maintainerQuoted 10 lines · Muchun Song
Quoted 8 lines · Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四 下午2:25写道:
Quoted 8 lines
Quoted 7 lines
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?
Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
It's in my list of patches to get to, sorry, hopefully will dig out of that next week with the buffer that the merge window provides me.
thanks,
greg k-h
Hi ,
Last patch will serialize the addition of child to parent directory, won't it affect performance.
Regards
Gaurav
Quoted 25 lines · Greg KH
Quoted 10 lines · Muchun Song
Quoted 8 lines · Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四 下午2:25写道:
Quoted 8 lines
Quoted 7 lines
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?Quoted 6 lines
Quoted 4 lines
Quoted 2 lines
Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
It's in my list of patches to get to, sorry, hopefully will dig out of that next week with the buffer that the merge window provides me.
thanks,
greg k-h
Signature · 3 lines
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.
Quoted 4 lines · Gaurav Kohli
Hi ,
Last patch will serialize the addition of child to parent directory, won't it affect performance.
I doubt this is a significant issue, and there's already a global lock taken once or twice in that path, the fix is purely to make sure that the some locked section is used both for the lookup and the addition as the bug comes from the window in between those two operations allowing the object to be removed after it was "found".
Cheers,
Ben.
Quoted 39 lines
Regards
GauravQuoted 32 lines · Greg KH
Quoted 14 lines · Muchun Song
Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四
Quoted 11 lines · 下午2:25写道:
Quoted 11 lines
Quoted 10 lines
The basic idea yes, the whole bool *locked is horrid though.
Wouldn't it
work to have aget_device_parent_lockedthat always returns with
the mutex held,
or just move the mutex to the caller or something simpler like this
?Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Greg and Rafael, do you have any suggestions for this? Or you also
agree with Ben?Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
It's in my list of patches to get to, sorry, hopefully will dig out of
that next week with the buffer that the merge window provides me.thanks,
greg k-h
Thanks for the comment, will check the patch and update.
Regards
Gaurav
Quoted 55 lines · Benjamin Herrenschmidt
Quoted 4 lines · Gaurav Kohli
Hi ,
Last patch will serialize the addition of child to parent directory, won't it affect performance.
I doubt this is a significant issue, and there's already a global lock taken once or twice in that path, the fix is purely to make sure that the some locked section is used both for the lookup and the addition as the bug comes from the window in between those two operations allowing the object to be removed after it was "found".
Cheers,
Ben.Quoted 39 lines
Regards
GauravQuoted 32 lines · Greg KH
Quoted 14 lines · Muchun Song
Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四
Quoted 11 lines · 下午2:25写道:
Quoted 11 lines
Quoted 10 lines
The basic idea yes, the whole bool *locked is horrid though.
Wouldn't it
work to have aget_device_parent_lockedthat always returns with
the mutex held,
or just move the mutex to the caller or something simpler like this
?Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Greg and Rafael, do you have any suggestions for this? Or you also
agree with Ben?Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
It's in my list of patches to get to, sorry, hopefully will dig out of
that next week with the buffer that the merge window provides me.thanks,
greg k-h
Signature · 3 lines
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.
++
Quoted 123 lines · Muchun Song
Quoted 10 lines · Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四 下午2:25写道:
Quoted 9 lines
Quoted 7 lines
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
From your suggestions, I think introduce
get_device_parent_locked()may easy
to fix. So, do you agree with the fix of the following code snippet (You can also
view attachments)?I introduce a new function named
get_device_parent_locked_if_glue_dir()which always returns with the mutex held only when we live in glue dir. We should callunlock_if_glue_dir()to release the mutex. Theget_device_parent_locked_if_glue_dir()
andunlock_if_glue_dir()should be called in pairs.---
drivers/base/core.c| 44 ++++++++++++++++++++++++++++++++++++--------1 file changed, 36 insertions(+), 8 deletions(-)drivers/base/core.c +36
-9@@ -1739,8 +1739,9 @@ class_dir_create_and_add(struct class *class, struct kobject *parent_kobj)static DEFINE_MUTEX (gdp_mutex); -static struct kobject *get_device_parent (struct device *dev, -struct device *parent) +static struct kobject *__get_device_parent (struct device *dev, +struct device *parent, +bool lock) {if (dev->class) {struct kobject *kobj =NULL ;@@ -1779,14 +1780,16 @@ static struct kobject *get_device_parent(struct device *dev, } spin_unlock(&dev->class->p->glue_dirs.list_lock);if (kobj) { -mutex_unlock (&gdp_mutex); +if (!lock) +mutex_unlock (&gdp_mutex);return kobj; }/* or create a new class-directory at the parent device */ k =class_dir_create_and_add (dev->class, parent_kobj);/* do not emit an uevent for this simple "glue" directory */ -mutex_unlock (&gdp_mutex); +if (!lock) +mutex_unlock (&gdp_mutex);return k; }@@ -1799,6 +1802,19 @@ static struct kobject *get_device_parent(structdevice *dev, return NULL; } +static inline struct kobject *get_device_parent (struct device *dev, +struct device *parent) +{ +return __get_device_parent (dev, parent, false); +} + +static inline struct kobject * +get_device_parent_locked_if_glue_dir (struct device *dev, +struct device *parent) +{ +return __get_device_parent (dev, parent, true); +} +static inline bool live_in_glue_dir (struct kobject *kobj,struct device *dev) {@@ -1831,6 +1847,16 @@ static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)mutex_unlock (&gdp_mutex); } +static inline void unlock_if_glue_dir (struct device *dev, +struct kobject *glue_dir) +{ +/* see if we live in a "glue" directory */ +if (!live_in_glue_dir (glue_dir, dev)) +return ; + +mutex_unlock (&gdp_mutex); +} +static int device_add_class_symlinks (struct device *dev) {struct device_node *of_node =dev_of_node (dev);@@ -2040,7 +2066,7 @@ int device_add(struct device *dev)pr_debug ("device: '%s': %s\n" ,dev_name (dev), __func__); parent =get_device (dev->parent); - kobj =get_device_parent (dev, parent); + kobj =get_device_parent_locked_if_glue_dir (dev, parent);if (IS_ERR (kobj)) { error =PTR_ERR (kobj); goto parent_error;@@ -2055,10 +2081,12 @@ int device_add(struct device *dev)/* first, register with generic layer. */ /* we require the name to be set before, and pass NULL */ error =kobject_add (&dev->kobj, dev->kobj.parent,NULL ); -if (error) { - glue_dir =get_glue_dir (dev); + + glue_dir =get_glue_dir (dev); +unlock_if_glue_dir (dev, glue_dir); + +if (error) goto Error; - }/* notify platform of device entry */ error =device_platform_notify (dev,KOBJ_ADD ); --
Quoted 118 lines · Mukesh Ojha
++
Quoted 115 lines · Muchun Song
Quoted 10 lines · Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四 下午2:25写道:
Quoted 9 lines
Quoted 7 lines
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
From your suggestions, I think introduce
get_device_parent_locked()may easy to fix. So, do you agree with the fix of the following code snippet (You can also
view attachments)?I introduce a new function named
get_device_parent_locked_if_glue_dir()which always returns with the mutex held only when we live in glue dir. We should callunlock_if_glue_dir()to release the mutex. Theget_device_parent_locked_if_glue_dir()
andunlock_if_glue_dir()should be called in pairs.---
drivers/base/core.c| 44 ++++++++++++++++++++++++++++++++++++--------1 file changed, 36 insertions(+), 8 deletions(-)drivers/base/core.c +33
-7@@ -1739,8 +1739,9 @@ class_dir_create_and_add(struct class *class, struct kobject *parent_kobj)static DEFINE_MUTEX (gdp_mutex); -static struct kobject *get_device_parent (struct device *dev, - struct device *parent) +static struct kobject *__get_device_parent (struct device *dev, + struct device *parent, + bool lock) { if (dev->class) { struct kobject *kobj = NULL;@@ -1779,14 +1780,16 @@ static struct kobject *get_device_parent(struct device *dev, } spin_unlock(&dev->class->p->glue_dirs.list_lock); if (kobj) { - mutex_unlock(&gdp_mutex); + if (!lock) + mutex_unlock(&gdp_mutex); return kobj; } /* or create a new class-directory at the parent device */ k = class_dir_create_and_add(dev->class, parent_kobj); /* do not emit an uevent for this simple "glue" directory */ - mutex_unlock(&gdp_mutex); + if (!lock) + mutex_unlock(&gdp_mutex); return k; }@@ -1799,6 +1802,19 @@ static struct kobject *get_device_parent(struct device *dev, return NULL; } +static inline struct kobject *get_device_parent (struct device *dev, + struct device *parent) +{ + return __get_device_parent(dev, parent, false); +} + +static inline struct kobject * +get_device_parent_locked_if_glue_dir (struct device *dev, + struct device *parent) +{ + return __get_device_parent(dev, parent, true); +} +static inline bool live_in_glue_dir (struct kobject *kobj, struct device *dev) {@@ -1831,6 +1847,16 @@ static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir) mutex_unlock(&gdp_mutex); } +static inline void unlock_if_glue_dir (struct device *dev, + struct kobject *glue_dir) +{ + /* see if we live in a "glue" directory */ + if (!live_in_glue_dir(glue_dir, dev)) + return; + + mutex_unlock(&gdp_mutex); +} +static int device_add_class_symlinks (struct device *dev) { struct device_node *of_node = dev_of_node(dev);@@ -2040,7 +2066,7 @@ int device_add(struct device *dev) pr_debug("device: '%s': %s\n", dev_name(dev), __func__); parent = get_device(dev->parent); - kobj = get_device_parent(dev, parent); + kobj = get_device_parent_locked_if_glue_dir(dev, parent); if (IS_ERR(kobj)) { error = PTR_ERR(kobj); goto parent_error;@@ -2055,10 +2081,12 @@ int device_add(struct device *dev) /* first, register with generic layer. */ /* we require the name to be set before, and pass NULL */ error = kobject_add(&dev->kobj, dev->kobj.parent, NULL); - if (error) { - glue_dir = get_glue_dir(dev); + + glue_dir = get_glue_dir(dev);
Quoted 8 lines
drivers/base/core.c int device_add(struct device *dev) · line 2092
Quoted 8 lines
drivers/base/core.c +3
-1+ unlock_if_glue_dir(dev, glue_dir); + + if (error) goto Error; - } /* notify platform of device entry */ error = device_platform_notify(dev, KOBJ_ADD);
This change has been done in device_add(). AFAICT, locked version of get_device_parent should be used in device_move() also.
Thanks
Signature · 3 lines
Qualcomm India Private Limited, on behalf of Qualcomm Innovation Center, Inc., is a member of Code Aurora Forum, a Linux Foundation Collaborative Project
Quoted 124 lines · Prateek Sood <[email protected]> 于2019年5月14日周二 下午7:00写道:
Quoted 122 lines · Mukesh Ojha
++
Quoted 119 lines · Muchun Song
Quoted 10 lines · Benjamin Herrenschmidt <[email protected]> 于2019年5月2日周四 下午2:25写道:
Quoted 9 lines
Quoted 7 lines
The basic idea yes, the whole bool *locked is horrid though. Wouldn't it
work to have aget_device_parent_lockedthat always returns with the mutex held,
or just move the mutex to the caller or something simpler like this ?Greg and Rafael, do you have any suggestions for this? Or you also agree with Ben?
Ping guys ? This is worth fixing...
I also agree with you. But Greg and Rafael seem to be high latency right now.
From your suggestions, I think introduce
get_device_parent_locked()may easy
to fix. So, do you agree with the fix of the following code snippet (You can also
view attachments)?I introduce a new function named
get_device_parent_locked_if_glue_dir()which always returns with the mutex held only when we live in glue dir. We should callunlock_if_glue_dir()to release the mutex. Theget_device_parent_locked_if_glue_dir()
andunlock_if_glue_dir()should be called in pairs.---
drivers/base/core.c| 44 ++++++++++++++++++++++++++++++++++++--------1 file changed, 36 insertions(+), 8 deletions(-)drivers/base/core.c +36
-7@@ -1739,8 +1739,9 @@ class_dir_create_and_add(struct class *class, struct kobject *parent_kobj)static DEFINE_MUTEX (gdp_mutex); -static struct kobject *get_device_parent (struct device *dev, -struct device *parent) +static struct kobject *__get_device_parent (struct device *dev, +struct device *parent, +bool lock) {if (dev->class) {struct kobject *kobj =NULL ;@@ -1779,14 +1780,16 @@ static struct kobject *get_device_parent(struct device *dev, } spin_unlock(&dev->class->p->glue_dirs.list_lock);if (kobj) { -mutex_unlock (&gdp_mutex); +if (!lock) +mutex_unlock (&gdp_mutex);return kobj; }/* or create a new class-directory at the parent device */ k =class_dir_create_and_add (dev->class, parent_kobj);/* do not emit an uevent for this simple "glue" directory */ -mutex_unlock (&gdp_mutex); +if (!lock) +mutex_unlock (&gdp_mutex);return k; }@@ -1799,6 +1802,19 @@ static struct kobject *get_device_parent(structdevice *dev, return NULL; } +static inline struct kobject *get_device_parent (struct device *dev, +struct device *parent) +{ +return __get_device_parent (dev, parent, false); +} + +static inline struct kobject * +get_device_parent_locked_if_glue_dir (struct device *dev, +struct device *parent) +{ +return __get_device_parent (dev, parent, true); +} +static inline bool live_in_glue_dir (struct kobject *kobj,struct device *dev) {@@ -1831,6 +1847,16 @@ static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)mutex_unlock (&gdp_mutex); } +static inline void unlock_if_glue_dir (struct device *dev, +struct kobject *glue_dir) +{ +/* see if we live in a "glue" directory */ +if (!live_in_glue_dir (glue_dir, dev)) +return ; + +mutex_unlock (&gdp_mutex); +} +static int device_add_class_symlinks (struct device *dev) {struct device_node *of_node =dev_of_node (dev);@@ -2040,7 +2066,7 @@ int device_add(struct device *dev)pr_debug ("device: '%s': %s\n" ,dev_name (dev), __func__); parent =get_device (dev->parent); - kobj =get_device_parent (dev, parent); + kobj =get_device_parent_locked_if_glue_dir (dev, parent);if (IS_ERR (kobj)) { error =PTR_ERR (kobj); goto parent_error;@@ -2055,10 +2081,12 @@ int device_add(struct device *dev)/* first, register with generic layer. */ /* we require the name to be set before, and pass NULL */ error =kobject_add (&dev->kobj, dev->kobj.parent,NULL ); -if (error) { - glue_dir =get_glue_dir (dev); + + glue_dir =get_glue_dir (dev); +unlock_if_glue_dir (dev, glue_dir); + +if (error) goto Error;
Quoted 9 lines
Quoted 4 lines
drivers/base/core.c int device_add(struct device *dev)
Quoted 4 lines
drivers/base/core.c +0
-2- }/* notify platform of device entry */ error =device_platform_notify (dev,KOBJ_ADD ); --This change has been done in
device_add(). AFAICT, locked version ofget_device_parentshould be used indevice_move()also.
Yeah, I agree with you. I will send the v2 patch later to fix it also. Thanks.
Yours,
Muchun