| Message ID | 20241002131639.3425022-2-matthew.brost@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Use user-defined workqueue lockdep map for drm sched | expand |
On 10/2/2024 3:16 PM, Matthew Brost wrote: > Avoid leaking a lockdep map on each drm sched creation and destruction > by using a single lockdep map for all drm sched allocated submit_wq. > > v2: > - Use alloc_ordered_workqueue_lockdep_map (Tejun) > > Cc: Luben Tuikov <ltuikov89@gmail.com> > Cc: Christian König <christian.koenig@amd.com> > Signed-off-by: Matthew Brost <matthew.brost@intel.com> LGTM Reviewed-by: Nirmoy Das <nirmoy.das@intel.com> > --- > drivers/gpu/drm/scheduler/sched_main.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c > index 36db5c7736fc..e32b0f7d7e94 100644 > --- a/drivers/gpu/drm/scheduler/sched_main.c > +++ b/drivers/gpu/drm/scheduler/sched_main.c > @@ -87,6 +87,12 @@ > #define CREATE_TRACE_POINTS > #include "gpu_scheduler_trace.h" > > +#ifdef CONFIG_LOCKDEP > +static struct lockdep_map drm_sched_lockdep_map = { > + .name = "drm_sched_lockdep_map" > +}; > +#endif > + > #define to_drm_sched_job(sched_job) \ > container_of((sched_job), struct drm_sched_job, queue_node) > > @@ -1270,7 +1276,12 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, > sched->submit_wq = submit_wq; > sched->own_submit_wq = false; > } else { > +#ifdef CONFIG_LOCKDEP > + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, > + &drm_sched_lockdep_map); > +#else > sched->submit_wq = alloc_ordered_workqueue(name, 0); > +#endif > if (!sched->submit_wq) > return -ENOMEM; >
On Wed, Oct 02, 2024 at 06:16:38AM -0700, Matthew Brost wrote: > Avoid leaking a lockdep map on each drm sched creation and destruction > by using a single lockdep map for all drm sched allocated submit_wq. > > v2: > - Use alloc_ordered_workqueue_lockdep_map (Tejun) > > Cc: Luben Tuikov <ltuikov89@gmail.com> > Cc: Christian König <christian.koenig@amd.com> > Signed-off-by: Matthew Brost <matthew.brost@intel.com> Good catch, Acked-by: Danilo Krummrich <dakr@kernel.org> > --- > drivers/gpu/drm/scheduler/sched_main.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c > index 36db5c7736fc..e32b0f7d7e94 100644 > --- a/drivers/gpu/drm/scheduler/sched_main.c > +++ b/drivers/gpu/drm/scheduler/sched_main.c > @@ -87,6 +87,12 @@ > #define CREATE_TRACE_POINTS > #include "gpu_scheduler_trace.h" > > +#ifdef CONFIG_LOCKDEP > +static struct lockdep_map drm_sched_lockdep_map = { > + .name = "drm_sched_lockdep_map" > +}; > +#endif > + > #define to_drm_sched_job(sched_job) \ > container_of((sched_job), struct drm_sched_job, queue_node) > > @@ -1270,7 +1276,12 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, > sched->submit_wq = submit_wq; > sched->own_submit_wq = false; > } else { > +#ifdef CONFIG_LOCKDEP > + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, > + &drm_sched_lockdep_map); > +#else > sched->submit_wq = alloc_ordered_workqueue(name, 0); > +#endif > if (!sched->submit_wq) > return -ENOMEM; > > -- > 2.34.1 >
On Wed, Oct 02, 2024 at 04:16:00PM +0200, Danilo Krummrich wrote: > On Wed, Oct 02, 2024 at 06:16:38AM -0700, Matthew Brost wrote: > > Avoid leaking a lockdep map on each drm sched creation and destruction > > by using a single lockdep map for all drm sched allocated submit_wq. > > > > v2: > > - Use alloc_ordered_workqueue_lockdep_map (Tejun) > > > > Cc: Luben Tuikov <ltuikov89@gmail.com> > > Cc: Christian König <christian.koenig@amd.com> > > Signed-off-by: Matthew Brost <matthew.brost@intel.com> > > Good catch, > Thanks. I'm likely to merge this patch through drm-xe-next rather than drm-misc assuming it is ok with the drm-misc maintainers. I've pinged Maarten about this. Any objection on your end if I do this? Matt > Acked-by: Danilo Krummrich <dakr@kernel.org> > > > --- > > drivers/gpu/drm/scheduler/sched_main.c | 11 +++++++++++ > > 1 file changed, 11 insertions(+) > > > > diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c > > index 36db5c7736fc..e32b0f7d7e94 100644 > > --- a/drivers/gpu/drm/scheduler/sched_main.c > > +++ b/drivers/gpu/drm/scheduler/sched_main.c > > @@ -87,6 +87,12 @@ > > #define CREATE_TRACE_POINTS > > #include "gpu_scheduler_trace.h" > > > > +#ifdef CONFIG_LOCKDEP > > +static struct lockdep_map drm_sched_lockdep_map = { > > + .name = "drm_sched_lockdep_map" > > +}; > > +#endif > > + > > #define to_drm_sched_job(sched_job) \ > > container_of((sched_job), struct drm_sched_job, queue_node) > > > > @@ -1270,7 +1276,12 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, > > sched->submit_wq = submit_wq; > > sched->own_submit_wq = false; > > } else { > > +#ifdef CONFIG_LOCKDEP > > + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, > > + &drm_sched_lockdep_map); > > +#else > > sched->submit_wq = alloc_ordered_workqueue(name, 0); > > +#endif > > if (!sched->submit_wq) > > return -ENOMEM; > > > > -- > > 2.34.1 > >
On Wed, Oct 02, 2024 at 03:00:08PM +0000, Matthew Brost wrote: > On Wed, Oct 02, 2024 at 04:16:00PM +0200, Danilo Krummrich wrote: > > On Wed, Oct 02, 2024 at 06:16:38AM -0700, Matthew Brost wrote: > > > Avoid leaking a lockdep map on each drm sched creation and destruction > > > by using a single lockdep map for all drm sched allocated submit_wq. > > > > > > v2: > > > - Use alloc_ordered_workqueue_lockdep_map (Tejun) > > > > > > Cc: Luben Tuikov <ltuikov89@gmail.com> > > > Cc: Christian König <christian.koenig@amd.com> > > > Signed-off-by: Matthew Brost <matthew.brost@intel.com> > > > > Good catch, > > > > Thanks. I'm likely to merge this patch through drm-xe-next rather than > drm-misc assuming it is ok with the drm-misc maintainers. I've pinged > Maarten about this. > > Any objection on your end if I do this? No, I think that makes sense. > > Matt > > > Acked-by: Danilo Krummrich <dakr@kernel.org> > > > > > --- > > > drivers/gpu/drm/scheduler/sched_main.c | 11 +++++++++++ > > > 1 file changed, 11 insertions(+) > > > > > > diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c > > > index 36db5c7736fc..e32b0f7d7e94 100644 > > > --- a/drivers/gpu/drm/scheduler/sched_main.c > > > +++ b/drivers/gpu/drm/scheduler/sched_main.c > > > @@ -87,6 +87,12 @@ > > > #define CREATE_TRACE_POINTS > > > #include "gpu_scheduler_trace.h" > > > > > > +#ifdef CONFIG_LOCKDEP > > > +static struct lockdep_map drm_sched_lockdep_map = { > > > + .name = "drm_sched_lockdep_map" > > > +}; > > > +#endif > > > + > > > #define to_drm_sched_job(sched_job) \ > > > container_of((sched_job), struct drm_sched_job, queue_node) > > > > > > @@ -1270,7 +1276,12 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, > > > sched->submit_wq = submit_wq; > > > sched->own_submit_wq = false; > > > } else { > > > +#ifdef CONFIG_LOCKDEP > > > + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, > > > + &drm_sched_lockdep_map); > > > +#else > > > sched->submit_wq = alloc_ordered_workqueue(name, 0); > > > +#endif > > > if (!sched->submit_wq) > > > return -ENOMEM; > > > > > > -- > > > 2.34.1 > > > >
On 02-10-2024 18:46, Matthew Brost wrote: > Avoid leaking a lockdep map on each drm sched creation and destruction > by using a single lockdep map for all drm sched allocated submit_wq. > > v2: > - Use alloc_ordered_workqueue_lockdep_map (Tejun) > > Cc: Luben Tuikov <ltuikov89@gmail.com> > Cc: Christian König <christian.koenig@amd.com> > Signed-off-by: Matthew Brost <matthew.brost@intel.com> > --- > drivers/gpu/drm/scheduler/sched_main.c | 11 +++++++++++ > 1 file changed, 11 insertions(+) > > diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c > index 36db5c7736fc..e32b0f7d7e94 100644 > --- a/drivers/gpu/drm/scheduler/sched_main.c > +++ b/drivers/gpu/drm/scheduler/sched_main.c > @@ -87,6 +87,12 @@ > #define CREATE_TRACE_POINTS > #include "gpu_scheduler_trace.h" > > +#ifdef CONFIG_LOCKDEP > +static struct lockdep_map drm_sched_lockdep_map = { > + .name = "drm_sched_lockdep_map" > +}; > +#endif > + > #define to_drm_sched_job(sched_job) \ > container_of((sched_job), struct drm_sched_job, queue_node) > > @@ -1270,7 +1276,12 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, > sched->submit_wq = submit_wq; > sched->own_submit_wq = false; > } else { > +#ifdef CONFIG_LOCKDEP > + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, > + &drm_sched_lockdep_map); > +#else > sched->submit_wq = alloc_ordered_workqueue(name, 0); > +#endif LGTM. Reviewed-by: Himal Prasad Ghimiray <himal.prasad.ghimiray@intel.com> > if (!sched->submit_wq) > return -ENOMEM; >
diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c index 36db5c7736fc..e32b0f7d7e94 100644 --- a/drivers/gpu/drm/scheduler/sched_main.c +++ b/drivers/gpu/drm/scheduler/sched_main.c @@ -87,6 +87,12 @@ #define CREATE_TRACE_POINTS #include "gpu_scheduler_trace.h" +#ifdef CONFIG_LOCKDEP +static struct lockdep_map drm_sched_lockdep_map = { + .name = "drm_sched_lockdep_map" +}; +#endif + #define to_drm_sched_job(sched_job) \ container_of((sched_job), struct drm_sched_job, queue_node) @@ -1270,7 +1276,12 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, sched->submit_wq = submit_wq; sched->own_submit_wq = false; } else { +#ifdef CONFIG_LOCKDEP + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, + &drm_sched_lockdep_map); +#else sched->submit_wq = alloc_ordered_workqueue(name, 0); +#endif if (!sched->submit_wq) return -ENOMEM;
Avoid leaking a lockdep map on each drm sched creation and destruction by using a single lockdep map for all drm sched allocated submit_wq. v2: - Use alloc_ordered_workqueue_lockdep_map (Tejun) Cc: Luben Tuikov <ltuikov89@gmail.com> Cc: Christian König <christian.koenig@amd.com> Signed-off-by: Matthew Brost <matthew.brost@intel.com> --- drivers/gpu/drm/scheduler/sched_main.c | 11 +++++++++++ 1 file changed, 11 insertions(+)