drm: Fix memory leak at error path of drm_read()
Subsystems (2)
PATCH 1/2 drm: Fix memory leak at error path of drm_read()
--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)
drivers/gpu/drm/drm_fops.c +1 -0
@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);break ; }
Signature · 6 lines
2.1.3 _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
PATCH 2/2 drm: Implement nonblocking mode in drm_read()
drm_read() always blocks the read until an event comes up even if the file is opened in nonblocking mode. Also, there is a potential race between the wake_event_interruptible() call and the actual event retrieval in drm_dequeue_event(), and it may result in zero event read even in the blocking mode.
This patch fixes these issues. The fixes are two folds:
- Do drm_dequeue_event() at first, then call wait_event_interruptible() in a loop only if no event is pending. It's safe to call drm_dequeue_event() at first since it returns immediately if no event is found, and the function is protected properly by event_lock.
- Add a f_flags check when drm_dequeue_event() fails and returns immediately in nonblocking mode. This is evaluated only at the first read.
Note: originally this issue comes up while debugging the hangup of X at switching VT quickly. For example, running the below on KDE results in the stall of X:
for i in $(seq 1 100); do chvt 1; chvt 7; done
This was reproduced on various Intel machines. I took a stack trace of the hanging process, and it points to the blockage in drm_read().
This patch "fixes" such a bug. A drawback is that now it becomes more difficult to find out such a buggy code in the caller side.
--- drivers/gpu/drm/drm_fops.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-)
drivers/gpu/drm/drm_fops.c +13 -6
@@ -515,13 +515,20 @@ ssize_t drm_read(struct file *filp, char __user *buffer,size_t total;ssize_t ret; - ret =wait_event_interruptible (file_priv->event_wait, - !list_empty (&file_priv->event_list)); -if (ret <0 ) -return ret; - total =0 ; -while (drm_dequeue_event (file_priv, total, count, &e)) { +for (;;) { +if (!drm_dequeue_event (file_priv, total, count, &e)) { +if (total) +break ; +if (filp->f_flags &O_NONBLOCK ) +return -EAGAIN ; + ret =wait_event_interruptible (file_priv->event_wait, + !list_empty (&file_priv->event_list)); +if (ret <0 ) +return ret; +continue ; + } +if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ;
Signature · 6 lines
2.1.3 _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Daniel Vetter
AppliedQuoted 1 line · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>
Applied to my drm-misc branch. The second patch seems to be a dupe of
commit bd008e5b2953186fc0c6633a885ade95e7043800
Author: Chris Wilson <[email protected]> Date: Tue Oct 7 14:13:51 2014 +0100
drm: Implement O_NONBLOCK support on /dev/dri/cardN
so maybe we should backport that one once it's landed in 3.19? -Daniel
Quoted 23 lines
--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);break ; }Signature · 6 lines
2.1.3 _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
At Thu, 4 Dec 2014 12:46:16 +0100,
Quoted 6 lines · Daniel Vetter
Quoted 1 line · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>Applied to my drm-misc branch. The second patch seems to be a dupe of
Quoted 7 lines
commit
bd008e5b2953186fc0c6633a885ade95e7043800Author: Chris Wilson <[email protected]> Date: Tue Oct 7 14:13:51 2014 +0100drm: Implement
O_NONBLOCKsupport on/dev/dri/cardNso maybe we should backport that one once it's landed in 3.19?
Ah thanks, I misread Chris' comment as if it were for libdrm, and now understood correctly.
Takashi
Quoted 18 lines
-Daniel
Quoted 16 lines
--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);break ; }
Quoted 13 lines
Quoted 7 lines
Signature · 6 lines
2.1.3 _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-develSignature · 3 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch
_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel
At Thu, 04 Dec 2014 12:50:04 +0100,
Quoted 5 lines · Takashi Iwai
At Thu, 4 Dec 2014 12:46:16 +0100,
Quoted 2 lines · Daniel Vetter
so maybe we should backport that one once it's landed in 3.19?
Speaking about this: yes, it's worth, IMO, as this (unexpectedly) seems really fixing the issue on the current driver.
thanks,
Takashi
_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 10 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)drivers/gpu/drm/drm_fops.c +0
-0
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 4 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);
We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
At Thu, 4 Dec 2014 11:51:14 +0000,
Quoted 7 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)
Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
Yeah, I thought of it while writing this, but as a starter, I tried the simpler one. (And I didn't realize your comment referring to the already existing fix in kernel, sorry!)
The problem to hold the spinlock for the whole is that you can't it with copy_to_user(). So it'd be a bit tricky.
And, why not using event_wait.lock instead of the extra event_lock? Then we can use wait_event_lock_*() variant that covers more race between dequeuing and wait_event.
Takashi
_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 24 lines · Takashi Iwai
At Thu, 4 Dec 2014 11:51:14 +0000,
Quoted 7 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
Quoted 6 lines
Yeah, I thought of it while writing this, but as a starter, I tried the simpler one. (And I didn't realize your comment referring to the already existing fix in kernel, sorry!)
The problem to hold the spinlock for the whole is that you can't it with
copy_to_user(). So it'd be a bit tricky.
We can use access_ok(VERFIY_WRITE) and then copy_to_user_inatomic(). Doable with a bit of code rearrangement.
Quoted 3 lines
And, why not using event_wait.lock instead of the extra
event_lock? Then we can use wait_event_lock_*() variant that covers more race between dequeuing andwait_event.
The tricky part would appear to be then protecting dev->vblank_event_list. That looks like it could be an RCU list instead? -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
At Thu, 4 Dec 2014 12:13:46 +0000,
Quoted 30 lines · Chris Wilson
Quoted 28 lines · Takashi Iwai
At Thu, 4 Dec 2014 11:51:14 +0000,
Quoted 7 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
Yeah, I thought of it while writing this, but as a starter, I tried the simpler one. (And I didn't realize your comment referring to the already existing fix in kernel, sorry!)
Quoted 5 lines
Quoted 2 lines
The problem to hold the spinlock for the whole is that you can't it with
copy_to_user(). So it'd be a bit tricky.We can use
access_ok(VERFIY_WRITE)and thencopy_to_user_inatomic(). Doable with a bit of code rearrangement.
Yes, but this is really not necessarily atomic. We can simply prepend the event back in the error case, instead, too.
Quoted 6 lines
Quoted 3 lines
And, why not using event_wait.lock instead of the extra
event_lock? Then we can use wait_event_lock_*() variant that covers more race between dequeuing andwait_event.The tricky part would appear to be then protecting
dev->vblank_event_list. That looks like it could be an RCU list instead?
Sounds interesting.
Takashi
_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 6 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)
Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
I've read the manpage for read and it explicitly states that when you get an error it's undefined what happens to the read position. Since -EFAULT is really just a userspace bug I think we can happily drop the event on the floor, no reason to bend over in the kernel.
I'll add a note to the commit message.
-Daniel
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 22 lines · Daniel Vetter
Quoted 6 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
Quoted 4 lines
I've read the manpage for read and it explicitly states that when you get an error it's undefined what happens to the read position. Since
-EFAULTis really just a userspace bug I think we can happily drop the event on the floor, no reason to bend over in the kernel.
Hmm. Actually the code is buggy is the provided buffer is too short for the first event in O_NONBLOCK mode.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 23 lines · Chris Wilson
Quoted 22 lines · Daniel Vetter
Quoted 20 lines · Chris Wilson
Quoted 10 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)drivers/gpu/drm/drm_fops.c +0
-0drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 4 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
Quoted 7 lines
Quoted 4 lines
I've read the manpage for read and it explicitly states that when you get an error it's undefined what happens to the read position. Since
-EFAULTis really just a userspace bug I think we can happily drop the event on the floor, no reason to bend over in the kernel.Hmm. Actually the code is buggy is the provided buffer is too short for the first event in
O_NONBLOCKmode.
Just inlining drm_dequeue_event is simple enough to fix the issue:
drivers/gpu/drm/drm_fops.c +31 -41
@@ -478,43 +478,17 @@ int drm_release(struct inode *inode, struct file *filp) }EXPORT_SYMBOL (drm_release); -static bool -drm_dequeue_event (struct drm_file *file_priv, -size_t total,size_t max,struct drm_pending_event **out) -{ -struct drm_device *dev = file_priv->minor->dev; -struct drm_pending_event *e; -unsigned long flags; -bool ret = false; - -spin_lock_irqsave (&dev->event_lock, flags); - - *out =NULL ; -if (list_empty (&file_priv->event_list)) - goto out; - e =list_first_entry (&file_priv->event_list, - struct drm_pending_event, link); -if (e->event->length + total > max) - goto out; - - file_priv->event_space += e->event->length; -list_del (&e->link); - *out = e; - ret = true; - -out: -spin_unlock_irqrestore (&dev->event_lock, flags); -return ret; -} -ssize_t drm_read (struct file *filp,char __user *buffer,size_t count,loff_t *offset) {struct drm_file *file_priv = filp->private_data; -struct drm_pending_event *e; -size_t total; +struct drm_device *dev = file_priv->minor->dev; +unsigned long flags;ssize_t ret; +if (!access_ok (VERIFY_WRITE , buffer, count)) +return -EFAULT ; +if ((filp->f_flags &O_NONBLOCK ) ==0 ) { ret =wait_event_interruptible (file_priv->event_wait, !list_empty (&file_priv->event_list));
@@ -522,19 +496,35 @@ ssize_t drm_read(struct file *filp, char __user *buffer,return ret; } - total =0 ; -while (drm_dequeue_event (file_priv, total, count, &e)) { -if (copy_to_user (buffer + total, - e->event, e->event->length)) { - total = -EFAULT ; -break ; - } +spin_lock_irqsave (&dev->event_lock, flags); +if (list_empty (&file_priv->event_list)) { + ret = -EAGAIN ; + }else { + ret =0 ; +do { +struct drm_pending_event *e; + + e =list_first_entry (&file_priv->event_list, + struct drm_pending_event, link); +if (e->event->length + ret > count) +break ; + +if (__copy_to_user_inatomic (buffer + ret, + e->event, e->event->length)) { +if (ret ==0 ) + ret = -EFAULT ; +break ; + } - total += e->event->length; - e->destroy (e); + file_priv->event_space += e->event->length; + ret += e->event->length; +list_del (&e->link); + e->destroy (e); + }while (!list_empty (&file_priv->event_list)); } +spin_unlock_irqrestore (&dev->event_lock, flags); -return total ?: -EAGAIN ; +return ret; }EXPORT_SYMBOL (drm_read);
which looks familar to the previous attempt to get O_NONBLOCK fixed. -Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
At Thu, 4 Dec 2014 16:38:57 +0000,
Quoted 30 lines · Chris Wilson
Quoted 28 lines · Chris Wilson
Quoted 26 lines · Daniel Vetter
Quoted 6 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
I've read the manpage for read and it explicitly states that when you get an error it's undefined what happens to the read position. Since
-EFAULTis really just a userspace bug I think we can happily drop the event on the floor, no reason to bend over in the kernel.
Quoted 4 lines
Quoted 2 lines
Hmm. Actually the code is buggy is the provided buffer is too short for the first event in
O_NONBLOCKmode.Just inlining
drm_dequeue_eventis simple enough to fix the issue:
Right, this avoids the dequeuing at error.
But still there is a race between wait_event_interruptible() and dequeuing. So, the more comprehensive code should look like:
if (!access_ok(VERIFY_WRITE, buffer, count))
return -EFAULT;
spin_lock_irq(&dev->event_lock);
for (;;) {
if (list_empty(&file_priv->event_list)) {
if (ret)
break;
if (!(filp->f_flags & O_NONBLOCK)) {
ret = -EAGAIN;
break;
}
spin_unlock_irq(&dev->event_lock);
ret = wait_event_interruptible(....);
spin_lock_irq(&dev->event_lock);
if (ret < 0)
break;
ret = 0;
continue;
}
e = list_first_entry(&file_priv->event_list,
struct drm_pending_event, link);
if (e->event->length + ret > count)
break;
if (__copy_to_user_inatomic(buffer + ret,
e->event, e->event->length)) {
if (!ret)
ret = -EFAULT;
break;
}
file_priv->event_space += e->event->length;
ret += e->event->length;
list_del(&e->link);
e->destroy(e);
}
spin_unlock_irq(&dev->event_lock);
return ret;thanks,
Takashi
Quoted 101 lines
drivers/gpu/drm/drm_fops.c +31
-41@@ -478,43 +478,17 @@ int drm_release(struct inode *inode, struct file *filp) }EXPORT_SYMBOL (drm_release); -static bool -drm_dequeue_event (struct drm_file *file_priv, -size_t total,size_t max,struct drm_pending_event **out) -{ -struct drm_device *dev = file_priv->minor->dev; -struct drm_pending_event *e; -unsigned long flags; -bool ret = false; - -spin_lock_irqsave (&dev->event_lock, flags); - - *out =NULL ; -if (list_empty (&file_priv->event_list)) - goto out; - e =list_first_entry (&file_priv->event_list, - struct drm_pending_event, link); -if (e->event->length + total > max) - goto out; - - file_priv->event_space += e->event->length; -list_del (&e->link); - *out = e; - ret = true; - -out: -spin_unlock_irqrestore (&dev->event_lock, flags); -return ret; -} -ssize_t drm_read (struct file *filp,char __user *buffer,size_t count,loff_t *offset) {struct drm_file *file_priv = filp->private_data; -struct drm_pending_event *e; -size_t total; +struct drm_device *dev = file_priv->minor->dev; +unsigned long flags;ssize_t ret; +if (!access_ok (VERIFY_WRITE , buffer, count)) +return -EFAULT ; +if ((filp->f_flags &O_NONBLOCK ) ==0 ) { ret =wait_event_interruptible (file_priv->event_wait, !list_empty (&file_priv->event_list));@@ -522,19 +496,35 @@ ssize_t drm_read(struct file *filp, char __user *buffer,return ret; } - total =0 ; -while (drm_dequeue_event (file_priv, total, count, &e)) { -if (copy_to_user (buffer + total, - e->event, e->event->length)) { - total = -EFAULT ; -break ; - } +spin_lock_irqsave (&dev->event_lock, flags); +if (list_empty (&file_priv->event_list)) { + ret = -EAGAIN ; + }else { + ret =0 ; +do { +struct drm_pending_event *e; + + e =list_first_entry (&file_priv->event_list, + struct drm_pending_event, link); +if (e->event->length + ret > count) +break ; + +if (__copy_to_user_inatomic (buffer + ret, + e->event, e->event->length)) { +if (ret ==0 ) + ret = -EFAULT ; +break ; + } - total += e->event->length; - e->destroy (e); + file_priv->event_space += e->event->length; + ret += e->event->length; +list_del (&e->link); + e->destroy (e); + }while (!list_empty (&file_priv->event_list)); } +spin_unlock_irqrestore (&dev->event_lock, flags); -return total ?: -EAGAIN ; +return ret; }EXPORT_SYMBOL (drm_read);
Quoted 6 lines
which looks familar to the previous attempt to get
O_NONBLOCKfixed. -ChrisSignature · 1 line
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
dri-devel mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/dri-devel
PATCH 1/1 drm: Make drm_read() more robust against multithreaded races
The current implementation of drm_read() faces a number of issues:
1. Upon an error, it consumes the event which may lead to the client blocking.
2. Upon an error, it forgets about events already copied
3. If it fails to copy a single event with O_NONBLOCK it falls into a infinite loop of reporting EAGAIN.
3. There is a race between multiple waiters and blocking reads of the events list.
Here, we inline drm_dequeue_event() into drm_read() so that we can take the spinlock around the list walking and event copying, and importantly reorder the error handling to avoid the issues above.
--- drivers/gpu/drm/drm_fops.c | 90 ++++++++++++++++++++++------------------------ 1 file changed, 43 insertions(+), 47 deletions(-)
drivers/gpu/drm/drm_fops.c +43 -47
@@ -478,63 +478,59 @@ int drm_release(struct inode *inode, struct file *filp) }EXPORT_SYMBOL (drm_release); -static bool -drm_dequeue_event (struct drm_file *file_priv, -size_t total,size_t max,struct drm_pending_event **out) +ssize_t drm_read (struct file *filp,char __user *buffer, +size_t count,loff_t *offset) { +struct drm_file *file_priv = filp->private_data;struct drm_device *dev = file_priv->minor->dev; -struct drm_pending_event *e; -unsigned long flags; -bool ret = false; - -spin_lock_irqsave (&dev->event_lock, flags); +ssize_t ret =0 ; - *out =NULL ; -if (list_empty (&file_priv->event_list)) - goto out; - e =list_first_entry (&file_priv->event_list, - struct drm_pending_event, link); -if (e->event->length + total > max) - goto out; +if (!access_ok (VERIFY_WRITE , buffer, count)) +return -EFAULT ; - file_priv->event_space += e->event->length; -list_del (&e->link); - *out = e; - ret = true; +spin_lock_irq (&dev->event_lock); +for (;;) { +if (list_empty (&file_priv->event_list)) { +if (ret) +break ; -out: -spin_unlock_irqrestore (&dev->event_lock, flags); -return ret; -} - -ssize_t drm_read (struct file *filp,char __user *buffer, -size_t count,loff_t *offset) -{ -struct drm_file *file_priv = filp->private_data; -struct drm_pending_event *e; -size_t total; -ssize_t ret; +if (filp->f_flags &O_NONBLOCK ) { + ret = -EAGAIN ; +break ; + } -if ((filp->f_flags &O_NONBLOCK ) ==0 ) { - ret =wait_event_interruptible (file_priv->event_wait, - !list_empty (&file_priv->event_list)); -if (ret <0 ) -return ret; - } +spin_unlock_irq (&dev->event_lock); + ret =wait_event_interruptible (file_priv->event_wait, + !list_empty (&file_priv->event_list)); +spin_lock_irq (&dev->event_lock); +if (ret <0 ) +break ; + + ret =0 ; + }else { +struct drm_pending_event *e; + + e =list_first_entry (&file_priv->event_list, + struct drm_pending_event, link); +if (e->event->length + ret > count) +break ; + +if (__copy_to_user_inatomic (buffer + ret, + e->event, e->event->length)) { +if (ret ==0 ) + ret = -EFAULT ; +break ; + } - total =0 ; -while (drm_dequeue_event (file_priv, total, count, &e)) { -if (copy_to_user (buffer + total, - e->event, e->event->length)) { - total = -EFAULT ; -break ; + file_priv->event_space += e->event->length; + ret += e->event->length; +list_del (&e->link); + e->destroy (e); } - - total += e->event->length; - e->destroy (e); } +spin_unlock_irq (&dev->event_lock); -return total ?: -EAGAIN ; +return ret; }EXPORT_SYMBOL (drm_read);
Signature · 6 lines
2.1.3 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH] drm: Make drm_read() more robust against multithreaded races
-------------------------------------Summary-------------------------------------
Platform Delta drm-intel-nightly Series Applied
PNV 364/364 364/364 ILK 366/366 366/366 SNB 450/450 450/450 IVB +17 481/498 498/498 BYT 289/289 289/289 HSW 564/564 564/564 BDW 417/417 417/417 -------------------------------------Detailed------------------------------------- Platform Test drm-intel-nightly Series Applied IVB igt_kms_3d DMESG_WARN(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-128x128-onscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-128x128-random NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-128x128-sliding NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-256x256-offscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-256x256-onscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-256x256-sliding NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-offscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-onscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-random NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-sliding NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-size-change NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_fence_pin_leak NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_mmio_vs_cs_flip_setcrtc_vs_cs_flip NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_mmio_vs_cs_flip_setplane_vs_cs_flip NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_rotation_crc_primary-rotation NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_rotation_crc_sprite-rotation NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34)
Note: You need to pay more attention to line start with '*'
_______________________________________________
Intel-gfx mailing list [email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH] drm: Make drm_read() more robust against multithreaded races
Quoted 7 lines · On Thu, Dec 04, 2014 at 06:19:54PM -0800, [email protected] wrote:
-------------------------------------Summary-------------------------------------
Platform Delta drm-intel-nightly Series AppliedPNV 364/364 364/364 ILK 366/366 366/366 SNB 450/450 450/450 IVB +17 481/498 498/498
Today there have been a lot of prts test results with +17 for ivb. Is something wrong with the baseline results that there's always the same regressions/improvements?
In general there's an awful lot of noise in prts results, and off-by-one in comparing results would explain a lot ...
-Daniel
Quoted 27 lines
BYT 289/289 289/289 HSW 564/564 564/564 BDW 417/417 417/417 -------------------------------------Detailed------------------------------------- Platform Test drm-intel-nightly Series Applied IVB igt_kms_3d DMESG_WARN(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-128x128-onscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-128x128-random NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-128x128-sliding NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-256x256-offscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-256x256-onscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-256x256-sliding NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-offscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-onscreen NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-random NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-64x64-sliding NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_cursor_crc_cursor-size-change NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_fence_pin_leak NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_mmio_vs_cs_flip_setcrtc_vs_cs_flip NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_mmio_vs_cs_flip_setplane_vs_cs_flip NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_rotation_crc_primary-rotation NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34) IVB igt_kms_rotation_crc_sprite-rotation NSPT(1, M34)PASS(10, M4M34M21) PASS(1, M34)Note: You need to pay more attention to line start with '*'
_______________________________________________
Intel-gfx mailing list [email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH] drm: Make drm_read() more robust against multithreaded races
In reply to Chris WilsonAt Thu, 4 Dec 2014 21:03:25 +0000,
Quoted 11 lines · Chris Wilson
The current implementation of
drm_read()faces a number of issues:1. Upon an error, it consumes the event which may lead to the client blocking.
2. Upon an error, it forgets about events already copied
3. If it fails to copy a single event withO_NONBLOCKit falls into a infinite loop of reporting EAGAIN.
3. There is a race between multiple waiters and blocking reads of the events list.
Quoted 5 lines
Here, we inline
drm_dequeue_event()intodrm_read()so that we can take the spinlock around the list walking and event copying, and importantly reorder the error handling to avoid the issues above.Cc Takashi Iwai <[email protected]>
Takashi
Quoted 119 lines
Signed-off-by Chris Wilson <[email protected]>--- drivers/gpu/drm/drm_fops.c | 90 ++++++++++++++++++++++------------------------ 1 file changed, 43 insertions(+), 47 deletions(-)drivers/gpu/drm/drm_fops.c +43
-47@@ -478,63 +478,59 @@ int drm_release(struct inode *inode, struct file *filp) }EXPORT_SYMBOL (drm_release); -static bool -drm_dequeue_event (struct drm_file *file_priv, -size_t total,size_t max,struct drm_pending_event **out) +ssize_t drm_read (struct file *filp,char __user *buffer, +size_t count,loff_t *offset) { +struct drm_file *file_priv = filp->private_data;struct drm_device *dev = file_priv->minor->dev; -struct drm_pending_event *e; -unsigned long flags; -bool ret = false; - -spin_lock_irqsave (&dev->event_lock, flags); +ssize_t ret =0 ; - *out =NULL ; -if (list_empty (&file_priv->event_list)) - goto out; - e =list_first_entry (&file_priv->event_list, - struct drm_pending_event, link); -if (e->event->length + total > max) - goto out; +if (!access_ok (VERIFY_WRITE , buffer, count)) +return -EFAULT ; - file_priv->event_space += e->event->length; -list_del (&e->link); - *out = e; - ret = true; +spin_lock_irq (&dev->event_lock); +for (;;) { +if (list_empty (&file_priv->event_list)) { +if (ret) +break ; -out: -spin_unlock_irqrestore (&dev->event_lock, flags); -return ret; -} - -ssize_t drm_read (struct file *filp,char __user *buffer, -size_t count,loff_t *offset) -{ -struct drm_file *file_priv = filp->private_data; -struct drm_pending_event *e; -size_t total; -ssize_t ret; +if (filp->f_flags &O_NONBLOCK ) { + ret = -EAGAIN ; +break ; + } -if ((filp->f_flags &O_NONBLOCK ) ==0 ) { - ret =wait_event_interruptible (file_priv->event_wait, - !list_empty (&file_priv->event_list)); -if (ret <0 ) -return ret; - } +spin_unlock_irq (&dev->event_lock); + ret =wait_event_interruptible (file_priv->event_wait, + !list_empty (&file_priv->event_list)); +spin_lock_irq (&dev->event_lock); +if (ret <0 ) +break ; + + ret =0 ; + }else { +struct drm_pending_event *e; + + e =list_first_entry (&file_priv->event_list, + struct drm_pending_event, link); +if (e->event->length + ret > count) +break ; + +if (__copy_to_user_inatomic (buffer + ret, + e->event, e->event->length)) { +if (ret ==0 ) + ret = -EFAULT ; +break ; + } - total =0 ; -while (drm_dequeue_event (file_priv, total, count, &e)) { -if (copy_to_user (buffer + total, - e->event, e->event->length)) { - total = -EFAULT ; -break ; + file_priv->event_space += e->event->length; + ret += e->event->length; +list_del (&e->link); + e->destroy (e); } - - total += e->event->length; - e->destroy (e); } +spin_unlock_irq (&dev->event_lock); -return total ?: -EAGAIN ; +return ret; }EXPORT_SYMBOL (drm_read);Signature · 1 line
2.1.3
_______________________________________________
Intel-gfx mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [Intel-gfx] [PATCH] drm: Make drm_read() more robust against multithreaded races
Quoted 13 lines · Takashi Iwai
At Thu, 4 Dec 2014 21:03:25 +0000,
Quoted 11 lines · Chris Wilson
The current implementation of
drm_read()faces a number of issues:1. Upon an error, it consumes the event which may lead to the client blocking.
2. Upon an error, it forgets about events already copied
3. If it fails to copy a single event withO_NONBLOCKit falls into a infinite loop of reporting EAGAIN.
3. There is a race between multiple waiters and blocking reads of the events list.
Quoted 7 lines
Quoted 5 lines
Here, we inline
drm_dequeue_event()intodrm_read()so that we can take the spinlock around the list walking and event copying, and importantly reorder the error handling to avoid the issues above.Cc Takashi Iwai <[email protected]>Reviewed-by Takashi Iwai <[email protected]>
Merged to drm-misc with the tag for the drm_read testcase added. Thanks for the patch&reivew.
-Daniel
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: [Intel-gfx] [PATCH] drm: Make drm_read() more robust against multithreaded races
In reply to Chris WilsonQuoted 10 lines · Chris Wilson
The current implementation of
drm_read()faces a number of issues:1. Upon an error, it consumes the event which may lead to the client blocking.
2. Upon an error, it forgets about events already copied
3. If it fails to copy a single event withO_NONBLOCKit falls into a infinite loop of reporting EAGAIN.
3. There is a race between multiple waiters and blocking reads of the events list.
Quoted 6 lines
Here, we inline
drm_dequeue_event()intodrm_read()so that we can take the spinlock around the list walking and event copying, and importantly reorder the error handling to avoid the issues above.Cc Takashi Iwai <[email protected]>Signed-off-by Chris Wilson <[email protected]>
Imo if you go through all the trouble of fixing the corner-cases then we should also have a testcase to exercise them. Otherwise I expect this to fall apart again. So a little igt would be great.
-Daniel
Quoted 122 lines
--- drivers/gpu/drm/drm_fops.c | 90 ++++++++++++++++++++++------------------------ 1 file changed, 43 insertions(+), 47 deletions(-)drivers/gpu/drm/drm_fops.c +43
-47@@ -478,63 +478,59 @@ int drm_release(struct inode *inode, struct file *filp) }EXPORT_SYMBOL (drm_release); -static bool -drm_dequeue_event (struct drm_file *file_priv, -size_t total,size_t max,struct drm_pending_event **out) +ssize_t drm_read (struct file *filp,char __user *buffer, +size_t count,loff_t *offset) { +struct drm_file *file_priv = filp->private_data;struct drm_device *dev = file_priv->minor->dev; -struct drm_pending_event *e; -unsigned long flags; -bool ret = false; - -spin_lock_irqsave (&dev->event_lock, flags); +ssize_t ret =0 ; - *out =NULL ; -if (list_empty (&file_priv->event_list)) - goto out; - e =list_first_entry (&file_priv->event_list, - struct drm_pending_event, link); -if (e->event->length + total > max) - goto out; +if (!access_ok (VERIFY_WRITE , buffer, count)) +return -EFAULT ; - file_priv->event_space += e->event->length; -list_del (&e->link); - *out = e; - ret = true; +spin_lock_irq (&dev->event_lock); +for (;;) { +if (list_empty (&file_priv->event_list)) { +if (ret) +break ; -out: -spin_unlock_irqrestore (&dev->event_lock, flags); -return ret; -} - -ssize_t drm_read (struct file *filp,char __user *buffer, -size_t count,loff_t *offset) -{ -struct drm_file *file_priv = filp->private_data; -struct drm_pending_event *e; -size_t total; -ssize_t ret; +if (filp->f_flags &O_NONBLOCK ) { + ret = -EAGAIN ; +break ; + } -if ((filp->f_flags &O_NONBLOCK ) ==0 ) { - ret =wait_event_interruptible (file_priv->event_wait, - !list_empty (&file_priv->event_list)); -if (ret <0 ) -return ret; - } +spin_unlock_irq (&dev->event_lock); + ret =wait_event_interruptible (file_priv->event_wait, + !list_empty (&file_priv->event_list)); +spin_lock_irq (&dev->event_lock); +if (ret <0 ) +break ; + + ret =0 ; + }else { +struct drm_pending_event *e; + + e =list_first_entry (&file_priv->event_list, + struct drm_pending_event, link); +if (e->event->length + ret > count) +break ; + +if (__copy_to_user_inatomic (buffer + ret, + e->event, e->event->length)) { +if (ret ==0 ) + ret = -EFAULT ; +break ; + } - total =0 ; -while (drm_dequeue_event (file_priv, total, count, &e)) { -if (copy_to_user (buffer + total, - e->event, e->event->length)) { - total = -EFAULT ; -break ; + file_priv->event_space += e->event->length; + ret += e->event->length; +list_del (&e->link); + e->destroy (e); } - - total += e->event->length; - e->destroy (e); } +spin_unlock_irq (&dev->event_lock); -return total ?: -EAGAIN ; +return ret; }EXPORT_SYMBOL (drm_read);Signature · 6 lines
2.1.3 _______________________________________________ Intel-gfx mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/intel-gfx
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 23 lines · Chris Wilson
Quoted 22 lines · Daniel Vetter
Quoted 20 lines · Chris Wilson
Quoted 10 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)drivers/gpu/drm/drm_fops.c +0
-0drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 4 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
Quoted 7 lines
Quoted 4 lines
I've read the manpage for read and it explicitly states that when you get an error it's undefined what happens to the read position. Since
-EFAULTis really just a userspace bug I think we can happily drop the event on the floor, no reason to bend over in the kernel.Hmm. Actually the code is buggy is the provided buffer is too short for the first event in
O_NONBLOCKmode.
Well we essentially send out datagrams instead of a bytestream. If we look at recvmsg and friends then discarding the additional bytes is something that's already being done. So I'm not terribly concerned about that either. It's a bit non-pretty that we use read and not reicvmsg but since we use read already not something we can ever fix.
-Daniel
Signature · 7 lines
Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel
Quoted 29 lines · Daniel Vetter
Quoted 28 lines · Chris Wilson
Quoted 26 lines · Daniel Vetter
Quoted 6 lines · Chris Wilson
Quoted 5 lines · Takashi Iwai
Signed-off-by Takashi Iwai <[email protected]>--- drivers/gpu/drm/drm_fops.c | 1 + 1 file changed, 1 insertion(+)Quoted 14 lines
drivers/gpu/drm/drm_fops.c ssize_t drm_read(struct file *filp, char __user *buffer, · line 528
Quoted 9 lines
drivers/gpu/drm/drm_fops.c +1
-0@@ -525,6 +525,7 @@ ssize_t drm_read(struct file *filp, char __user *buffer,if (copy_to_user (buffer + total, e->event, e->event->length)) { total = -EFAULT ; + e->destroy (e);We shouldn't just be throwing away the event here, but put the event back at the front of the queue. Poses an interesting race issue. Seems like we want to hold the spinlock until the copy is complete so that we can fix up the failure correctly.
I've read the manpage for read and it explicitly states that when you get an error it's undefined what happens to the read position. Since
-EFAULTis really just a userspace bug I think we can happily drop the event on the floor, no reason to bend over in the kernel.
Quoted 8 lines
Quoted 2 lines
Hmm. Actually the code is buggy is the provided buffer is too short for the first event in
O_NONBLOCKmode.Well we essentially send out datagrams instead of a bytestream. If we look at recvmsg and friends then discarding the additional bytes is something that's already being done. So I'm not terribly concerned about that either. It's a bit non-pretty that we use read and not reicvmsg but since we use read already not something we can ever fix.
In all of this consider what the impact of dropping an event is: system lockup.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ dri-devel mailing list [email protected] http://lists.freedesktop.org/mailman/listinfo/dri-devel