* compositing broken in -next (idr bug)
@ 2013-02-12 21:55 Jiri Slaby
2013-02-12 21:58 ` Tejun Heo
2013-02-12 22:03 ` Jiri Slaby
0 siblings, 2 replies; 6+ messages in thread
From: Jiri Slaby @ 2013-02-12 21:55 UTC (permalink / raw)
To: Tejun Heo; +Cc: Andrew Morton, David Airlie, dri-devel, LKML
Hi,
this commit in -next causes my KDE to get stuck while starting. I see
only the splash screen. If I disable effects by alt-shift-f12, it continues.
I bisected it to this commit:
commit 430440fce7da4ad52c2af06a04a5132e5d19faaf
Author: Tejun Heo <tj@kernel.org>
Date: Thu Feb 7 12:31:37 2013 +1100
drm: convert to idr_alloc()
More concretely to the change in drm_gem_flink_ioctl. If I revert solely
that one, it works again. Maybe I'm blind, but I do not see why...
Well I see a bug in there, but that is not the root cause -- there is a
missing call to idr_preload_end. Of course, that one only leaves
preemption disabled forever.
What I tried with no effect is this:
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -453,7 +453,8 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
spin_lock(&dev->object_name_lock);
if (!obj->name) {
ret = idr_alloc(&dev->object_name_idr, obj, 1, 0,
GFP_NOWAIT);
- obj->name = ret;
+ if (ret >= 0)
+ obj->name = ret;
args->name = (uint64_t) obj->name;
spin_unlock(&dev->object_name_lock);
@@ -469,6 +470,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
}
err:
+ idr_preload_end();
drm_gem_object_unreference_unlocked(obj);
return ret;
}
thanks,
--
js
suse labs
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: compositing broken in -next (idr bug)
2013-02-12 21:55 compositing broken in -next (idr bug) Jiri Slaby
@ 2013-02-12 21:58 ` Tejun Heo
2013-02-12 21:59 ` Jiri Slaby
2013-02-12 22:03 ` Jiri Slaby
1 sibling, 1 reply; 6+ messages in thread
From: Tejun Heo @ 2013-02-12 21:58 UTC (permalink / raw)
To: Jiri Slaby; +Cc: Andrew Morton, David Airlie, dri-devel, LKML
Hello, Jiri.
On Tue, Feb 12, 2013 at 10:55:37PM +0100, Jiri Slaby wrote:
> this commit in -next causes my KDE to get stuck while starting. I see
> only the splash screen. If I disable effects by alt-shift-f12, it continues.
>
> I bisected it to this commit:
> commit 430440fce7da4ad52c2af06a04a5132e5d19faaf
> Author: Tejun Heo <tj@kernel.org>
> Date: Thu Feb 7 12:31:37 2013 +1100
>
> drm: convert to idr_alloc()
>
>
> More concretely to the change in drm_gem_flink_ioctl. If I revert solely
> that one, it works again. Maybe I'm blind, but I do not see why...
>
> Well I see a bug in there, but that is not the root cause -- there is a
> missing call to idr_preload_end. Of course, that one only leaves
> preemption disabled forever.
Fix for this bug is already queued in -mm.
http://article.gmane.org/gmane.linux.kernel/1439101/raw
But you're saying that adding the missing idr_preload_end() doesn't
make the problem go away, right?
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: compositing broken in -next (idr bug)
2013-02-12 21:58 ` Tejun Heo
@ 2013-02-12 21:59 ` Jiri Slaby
0 siblings, 0 replies; 6+ messages in thread
From: Jiri Slaby @ 2013-02-12 21:59 UTC (permalink / raw)
To: Tejun Heo; +Cc: Andrew Morton, David Airlie, dri-devel, LKML
On 02/12/2013 10:58 PM, Tejun Heo wrote:
> Hello, Jiri.
>
> On Tue, Feb 12, 2013 at 10:55:37PM +0100, Jiri Slaby wrote:
>> this commit in -next causes my KDE to get stuck while starting. I see
>> only the splash screen. If I disable effects by alt-shift-f12, it continues.
>>
>> I bisected it to this commit:
>> commit 430440fce7da4ad52c2af06a04a5132e5d19faaf
>> Author: Tejun Heo <tj@kernel.org>
>> Date: Thu Feb 7 12:31:37 2013 +1100
>>
>> drm: convert to idr_alloc()
>>
>>
>> More concretely to the change in drm_gem_flink_ioctl. If I revert solely
>> that one, it works again. Maybe I'm blind, but I do not see why...
>>
>> Well I see a bug in there, but that is not the root cause -- there is a
>> missing call to idr_preload_end. Of course, that one only leaves
>> preemption disabled forever.
>
> Fix for this bug is already queued in -mm.
>
> http://article.gmane.org/gmane.linux.kernel/1439101/raw
>
> But you're saying that adding the missing idr_preload_end() doesn't
> make the problem go away, right?
Yeah, right. I had to revert the whole hunk...
--
js
suse labs
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: compositing broken in -next (idr bug)
2013-02-12 21:55 compositing broken in -next (idr bug) Jiri Slaby
2013-02-12 21:58 ` Tejun Heo
@ 2013-02-12 22:03 ` Jiri Slaby
2013-02-12 22:08 ` Jiri Slaby
1 sibling, 1 reply; 6+ messages in thread
From: Jiri Slaby @ 2013-02-12 22:03 UTC (permalink / raw)
To: Tejun Heo; +Cc: Andrew Morton, David Airlie, dri-devel, LKML
On 02/12/2013 10:55 PM, Jiri Slaby wrote:
> --- a/drivers/gpu/drm/drm_gem.c
> +++ b/drivers/gpu/drm/drm_gem.c
> @@ -453,7 +453,8 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
> spin_lock(&dev->object_name_lock);
> if (!obj->name) {
> ret = idr_alloc(&dev->object_name_idr, obj, 1, 0,
> GFP_NOWAIT);
> - obj->name = ret;
> + if (ret >= 0)
> + obj->name = ret;
> args->name = (uint64_t) obj->name;
> spin_unlock(&dev->object_name_lock);
>
> @@ -469,6 +470,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
> }
>
> err:
> + idr_preload_end();
> drm_gem_object_unreference_unlocked(obj);
> return ret;
Oh my, maybe: return ret < 0 ? ret : 0... Let's try.
--
js
suse labs
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: compositing broken in -next (idr bug)
2013-02-12 22:03 ` Jiri Slaby
@ 2013-02-12 22:08 ` Jiri Slaby
2013-02-12 22:13 ` Tejun Heo
0 siblings, 1 reply; 6+ messages in thread
From: Jiri Slaby @ 2013-02-12 22:08 UTC (permalink / raw)
To: Tejun Heo; +Cc: Andrew Morton, David Airlie, dri-devel, LKML
On 02/12/2013 11:03 PM, Jiri Slaby wrote:
> On 02/12/2013 10:55 PM, Jiri Slaby wrote:
>> --- a/drivers/gpu/drm/drm_gem.c
>> +++ b/drivers/gpu/drm/drm_gem.c
>> @@ -453,7 +453,8 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
>> spin_lock(&dev->object_name_lock);
>> if (!obj->name) {
>> ret = idr_alloc(&dev->object_name_idr, obj, 1, 0,
>> GFP_NOWAIT);
>> - obj->name = ret;
>> + if (ret >= 0)
>> + obj->name = ret;
>> args->name = (uint64_t) obj->name;
>> spin_unlock(&dev->object_name_lock);
>>
>> @@ -469,6 +470,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
>> }
>>
>> err:
>> + idr_preload_end();
>> drm_gem_object_unreference_unlocked(obj);
>> return ret;
>
> Oh my, maybe: return ret < 0 ? ret : 0... Let's try.
Bull's eye.
--- a/drivers/gpu/drm/drm_gem.c
+++ b/drivers/gpu/drm/drm_gem.c
@@ -459,6 +459,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
if (ret < 0)
goto err;
+ ret = 0;
/* Allocate a reference for the name table. */
drm_gem_object_reference(obj);
--
js
suse labs
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: compositing broken in -next (idr bug)
2013-02-12 22:08 ` Jiri Slaby
@ 2013-02-12 22:13 ` Tejun Heo
0 siblings, 0 replies; 6+ messages in thread
From: Tejun Heo @ 2013-02-12 22:13 UTC (permalink / raw)
To: Jiri Slaby; +Cc: Andrew Morton, David Airlie, dri-devel, LKML
Hello, Jiri.
On Tue, Feb 12, 2013 at 11:08:41PM +0100, Jiri Slaby wrote:
> > Oh my, maybe: return ret < 0 ? ret : 0... Let's try.
>
> Bull's eye.
Aieee....
> --- a/drivers/gpu/drm/drm_gem.c
> +++ b/drivers/gpu/drm/drm_gem.c
> @@ -459,6 +459,7 @@ drm_gem_flink_ioctl(struct drm_device *dev, void *data,
>
> if (ret < 0)
> goto err;
> + ret = 0;
>
> /* Allocate a reference for the name table. */
> drm_gem_object_reference(obj);
Care to write up a proper patch w/ SOB?
Thanks!
--
tejun
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2013-02-12 22:13 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-02-12 21:55 compositing broken in -next (idr bug) Jiri Slaby
2013-02-12 21:58 ` Tejun Heo
2013-02-12 21:59 ` Jiri Slaby
2013-02-12 22:03 ` Jiri Slaby
2013-02-12 22:08 ` Jiri Slaby
2013-02-12 22:13 ` Tejun Heo
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).