Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
On Thu, May 16, 2019 at 03:00:01PM +0300, Laurent Pinchart wrote: > Hi Sean, > > On Mon, May 13, 2019 at 10:38:58AM -0400, Sean Paul wrote: > > On Sat, May 11, 2019 at 3:12 PM Laurent Pinchart wrote: > > > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > > >> From: Sean Paul > > >> > > >> Everyone who implements connector_helper_funcs->atomic_check reaches > > >> into the connector state to get the atomic state. Instead of continuing > > >> this pattern, change the callback signature to just give atomic state > > >> and let the driver determine what it does and does not need from it. > > >> > > >> Eventually all atomic functions should do this, but that's just too much > > >> busy work for me. > > > > > > Given that drivers also access the connector state, isn't this slightly > > > more inefficient ? > > > > Inefficient in terms of what? > > In terms of the operation having to lookup the connector state, when the > caller has it and can easily pass it. As Daniel commented, this may be > the price to pay for a cleaner API, but I wonder how much overhead all > the state tracking is costing. > It'd be interesting to quantify, but iirc the last time we benchmarked atomic check commits they were virtually free (~thousands/s). > > Agree that in isolation this patch might seem unnecessary, but it ties > > in with the encoder and bridge CLs which accept drm_atomic_state in > > CLs ? Googleism for patches, I usually catch these before sending... guess I missed one. Sean > > > their hooks. In general the idea is to convert all atomic functions to > > take overall atomic state instead of just their object state. Reality > > has proven to be more complicated and we need more access than what > > the current implementation provides. > > > > Sean > > > > >> Changes in v3: > > >> - Added to the set > > >> > > >> Cc: Daniel Vetter > > >> Cc: Ville Syrjälä > > >> Cc: Jani Nikula > > >> Cc: Joonas Lahtinen > > >> Cc: Rodrigo Vivi > > >> Cc: Ben Skeggs > > >> Cc: Laurent Pinchart > > >> Cc: Kieran Bingham > > >> Cc: Eric Anholt > > >> Signed-off-by: Sean Paul > > >> --- > > >> drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > > >> drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > > >> drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > > >> drivers/gpu/drm/i915/intel_drv.h | 2 +- > > >> drivers/gpu/drm/i915/intel_sdvo.c| 9 + > > >> drivers/gpu/drm/i915/intel_tv.c | 8 +--- > > >> drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > > >> drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > > >> drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > > >> include/drm/drm_modeset_helper_vtables.h | 2 +- > > >> 10 files changed, 37 insertions(+), 27 deletions(-) > > >> > > >> diff --git a/drivers/gpu/drm/drm_atomic_helper.c > > >> b/drivers/gpu/drm/drm_atomic_helper.c > > >> index 9d9e47276839..fa5a367507c1 100644 > > >> --- a/drivers/gpu/drm/drm_atomic_helper.c > > >> +++ b/drivers/gpu/drm/drm_atomic_helper.c > > >> @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device > > >> *dev, > > >> } > > >> > > >> if (funcs->atomic_check) > > >> - ret = funcs->atomic_check(connector, > > >> new_connector_state); > > >> + ret = funcs->atomic_check(connector, state); > > >> if (ret) > > >> return ret; > > >> > > >> @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device > > >> *dev, > > >> continue; > > >> > > >> if (funcs->atomic_check) > > >> - ret = funcs->atomic_check(connector, > > >> new_connector_state); > > >> + ret = funcs->atomic_check(connector, state); > > >> if (ret) > > >> return ret; > > >> } > > >> diff --git a/drivers/gpu/drm/i915/intel_atomic.c > > >> b/drivers/gpu/drm/i915/intel_atomic.c > > >> index b844e8840c6f..e8a5b82e9242 100644 > > >> --- a/drivers/gpu/drm/i915/intel_atomic.c > > >> +++ b/drivers/gpu/drm/i915/intel_atomic.c > > >> @@ -103,12 +103,14 @@ int > > >> intel_digital_connector_atomic_set_property(struct drm_connector > > >> *connector, > > >> } > > >> > > >> int intel_digital_connector_atomic_check(struct drm_connector *conn, > > >> - struct drm_connector_state > > >> *new_state) > > >> + struct drm_atomic_state *state) > > >> { > > >> + struct drm_connector_state *new_state = > > >> + drm_atomic_get_new_connector_state(state, conn); > > >> struct intel_digital_connector_state *new_conn_state = > > >> to_intel_digital_connector_state(new_state); > > >> struct drm_connector_state *old_state = > > >> - drm_atomic_get_old_connector_state(new_state->state, conn); > > >> + drm_atomic_get_old_connector_stat
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
On Thu, May 16, 2019 at 02:07:34PM +0200, Daniel Vetter wrote: > On Thu, May 16, 2019 at 2:02 PM Laurent Pinchart > wrote: > > > > Hi Daniel, > > > > On Mon, May 13, 2019 at 04:47:47PM +0200, Daniel Vetter wrote: > > > On Sat, May 11, 2019 at 10:12:02PM +0300, Laurent Pinchart wrote: > > > > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > > > >> From: Sean Paul > > > >> > > > >> Everyone who implements connector_helper_funcs->atomic_check reaches > > > >> into the connector state to get the atomic state. Instead of continuing > > > >> this pattern, change the callback signature to just give atomic state > > > >> and let the driver determine what it does and does not need from it. > > > >> > > > >> Eventually all atomic functions should do this, but that's just too > > > >> much > > > >> busy work for me. > > > > > > > > Given that drivers also access the connector state, isn't this slightly > > > > more inefficient ? > > > > > > It's atomic code, we're trying to optimize for clean code at the expense > > > of a bit of runtime overhead due to more pointer chasing. And I agree with > > > the general push, the pile of old/new_state pointers of various objects > > > we're passing around is confusing. Passing the overall drm_atomic_state > > > seems much more reasonable, and with that we can get everything else. Plus > > > it's much more obvious whether you have the old/new state (since that's > > > explicit when you look it up from the drm_atomic_state). > > > > Yes, I agree it's cleaner. I just hope the atomic state tracking cost > > can be kept under control :-) > > > > By the way, this is likely not going to happen as it would be way too > > intrusive, but it would be nice to rename drm_atomic_state to > > drm_atomic_transaction (or something similar). It doesn't model a state, > > but a change between an old state to a new state. This confused me in > > the past, and I'm sure it can still be confusing to newcomers. > > Why are you the first to suggest this, this is awesome! Can't quite tell if that's irony or not. Anyways, this has been suggested before but no volunteers stepped forward. > drm_atomic_state is indeed not a state, but a transaction representing > how we go from the old to the new state. On a semi-related topic, I've occasionally pondered about moving mode_changed & co. from the obj states to the top level state/transaction (maybe stored as a bitmask). But that would definitely not be a trivial sed job. -- Ville Syrjälä Intel ___ Intel-gfx mailing list Intel-gfx@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
On Thu, May 16, 2019 at 2:02 PM Laurent Pinchart wrote: > > Hi Daniel, > > On Mon, May 13, 2019 at 04:47:47PM +0200, Daniel Vetter wrote: > > On Sat, May 11, 2019 at 10:12:02PM +0300, Laurent Pinchart wrote: > > > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > > >> From: Sean Paul > > >> > > >> Everyone who implements connector_helper_funcs->atomic_check reaches > > >> into the connector state to get the atomic state. Instead of continuing > > >> this pattern, change the callback signature to just give atomic state > > >> and let the driver determine what it does and does not need from it. > > >> > > >> Eventually all atomic functions should do this, but that's just too much > > >> busy work for me. > > > > > > Given that drivers also access the connector state, isn't this slightly > > > more inefficient ? > > > > It's atomic code, we're trying to optimize for clean code at the expense > > of a bit of runtime overhead due to more pointer chasing. And I agree with > > the general push, the pile of old/new_state pointers of various objects > > we're passing around is confusing. Passing the overall drm_atomic_state > > seems much more reasonable, and with that we can get everything else. Plus > > it's much more obvious whether you have the old/new state (since that's > > explicit when you look it up from the drm_atomic_state). > > Yes, I agree it's cleaner. I just hope the atomic state tracking cost > can be kept under control :-) > > By the way, this is likely not going to happen as it would be way too > intrusive, but it would be nice to rename drm_atomic_state to > drm_atomic_transaction (or something similar). It doesn't model a state, > but a change between an old state to a new state. This confused me in > the past, and I'm sure it can still be confusing to newcomers. Why are you the first to suggest this, this is awesome! drm_atomic_state is indeed not a state, but a transaction representing how we go from the old to the new state. I think this is awesome enough that we should actually try to make it happen. Tree-wide cocci + bribing Dave on St. Patrick's day with lots of beer might be enough :-) -Daniel > > If we ever see this show up in profile, and it starts mattering, first > > thing we need is a hashtable I think (atm it's list walking, which is just > > terrible). But thus far no one cares. > > > > >> Changes in v3: > > >> - Added to the set > > >> > > >> Cc: Daniel Vetter > > >> Cc: Ville Syrjälä > > >> Cc: Jani Nikula > > >> Cc: Joonas Lahtinen > > >> Cc: Rodrigo Vivi > > >> Cc: Ben Skeggs > > >> Cc: Laurent Pinchart > > >> Cc: Kieran Bingham > > >> Cc: Eric Anholt > > >> Signed-off-by: Sean Paul > > >> --- > > >> drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > > >> drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > > >> drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > > >> drivers/gpu/drm/i915/intel_drv.h | 2 +- > > >> drivers/gpu/drm/i915/intel_sdvo.c| 9 + > > >> drivers/gpu/drm/i915/intel_tv.c | 8 +--- > > >> drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > > >> drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > > >> drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > > >> include/drm/drm_modeset_helper_vtables.h | 2 +- > > >> 10 files changed, 37 insertions(+), 27 deletions(-) > > >> > > >> diff --git a/drivers/gpu/drm/drm_atomic_helper.c > > >> b/drivers/gpu/drm/drm_atomic_helper.c > > >> index 9d9e47276839..fa5a367507c1 100644 > > >> --- a/drivers/gpu/drm/drm_atomic_helper.c > > >> +++ b/drivers/gpu/drm/drm_atomic_helper.c > > >> @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device > > >> *dev, > > >>} > > >> > > >>if (funcs->atomic_check) > > >> - ret = funcs->atomic_check(connector, > > >> new_connector_state); > > >> + ret = funcs->atomic_check(connector, state); > > >>if (ret) > > >>return ret; > > >> > > >> @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device > > >> *dev, > > >>continue; > > >> > > >>if (funcs->atomic_check) > > >> - ret = funcs->atomic_check(connector, > > >> new_connector_state); > > >> + ret = funcs->atomic_check(connector, state); > > >>if (ret) > > >>return ret; > > >>} > > >> diff --git a/drivers/gpu/drm/i915/intel_atomic.c > > >> b/drivers/gpu/drm/i915/intel_atomic.c > > >> index b844e8840c6f..e8a5b82e9242 100644 > > >> --- a/drivers/gpu/drm/i915/intel_atomic.c > > >> +++ b/drivers/gpu/drm/i915/intel_atomic.c > > >> @@ -103,12 +103,14 @@ int > > >> intel_digital_connector_atomic_set_property(struct drm_connector > > >> *connector, > > >> } > > >> > > >> int intel_digital_connector_atomic_check(struct drm_connector *conn, > > >> - struct drm_connector_state > > >> *new_state) > >
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
Hi Daniel, On Mon, May 13, 2019 at 04:47:47PM +0200, Daniel Vetter wrote: > On Sat, May 11, 2019 at 10:12:02PM +0300, Laurent Pinchart wrote: > > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > >> From: Sean Paul > >> > >> Everyone who implements connector_helper_funcs->atomic_check reaches > >> into the connector state to get the atomic state. Instead of continuing > >> this pattern, change the callback signature to just give atomic state > >> and let the driver determine what it does and does not need from it. > >> > >> Eventually all atomic functions should do this, but that's just too much > >> busy work for me. > > > > Given that drivers also access the connector state, isn't this slightly > > more inefficient ? > > It's atomic code, we're trying to optimize for clean code at the expense > of a bit of runtime overhead due to more pointer chasing. And I agree with > the general push, the pile of old/new_state pointers of various objects > we're passing around is confusing. Passing the overall drm_atomic_state > seems much more reasonable, and with that we can get everything else. Plus > it's much more obvious whether you have the old/new state (since that's > explicit when you look it up from the drm_atomic_state). Yes, I agree it's cleaner. I just hope the atomic state tracking cost can be kept under control :-) By the way, this is likely not going to happen as it would be way too intrusive, but it would be nice to rename drm_atomic_state to drm_atomic_transaction (or something similar). It doesn't model a state, but a change between an old state to a new state. This confused me in the past, and I'm sure it can still be confusing to newcomers. > If we ever see this show up in profile, and it starts mattering, first > thing we need is a hashtable I think (atm it's list walking, which is just > terrible). But thus far no one cares. > > >> Changes in v3: > >> - Added to the set > >> > >> Cc: Daniel Vetter > >> Cc: Ville Syrjälä > >> Cc: Jani Nikula > >> Cc: Joonas Lahtinen > >> Cc: Rodrigo Vivi > >> Cc: Ben Skeggs > >> Cc: Laurent Pinchart > >> Cc: Kieran Bingham > >> Cc: Eric Anholt > >> Signed-off-by: Sean Paul > >> --- > >> drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > >> drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > >> drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > >> drivers/gpu/drm/i915/intel_drv.h | 2 +- > >> drivers/gpu/drm/i915/intel_sdvo.c| 9 + > >> drivers/gpu/drm/i915/intel_tv.c | 8 +--- > >> drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > >> drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > >> drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > >> include/drm/drm_modeset_helper_vtables.h | 2 +- > >> 10 files changed, 37 insertions(+), 27 deletions(-) > >> > >> diff --git a/drivers/gpu/drm/drm_atomic_helper.c > >> b/drivers/gpu/drm/drm_atomic_helper.c > >> index 9d9e47276839..fa5a367507c1 100644 > >> --- a/drivers/gpu/drm/drm_atomic_helper.c > >> +++ b/drivers/gpu/drm/drm_atomic_helper.c > >> @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > >>} > >> > >>if (funcs->atomic_check) > >> - ret = funcs->atomic_check(connector, > >> new_connector_state); > >> + ret = funcs->atomic_check(connector, state); > >>if (ret) > >>return ret; > >> > >> @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > >>continue; > >> > >>if (funcs->atomic_check) > >> - ret = funcs->atomic_check(connector, > >> new_connector_state); > >> + ret = funcs->atomic_check(connector, state); > >>if (ret) > >>return ret; > >>} > >> diff --git a/drivers/gpu/drm/i915/intel_atomic.c > >> b/drivers/gpu/drm/i915/intel_atomic.c > >> index b844e8840c6f..e8a5b82e9242 100644 > >> --- a/drivers/gpu/drm/i915/intel_atomic.c > >> +++ b/drivers/gpu/drm/i915/intel_atomic.c > >> @@ -103,12 +103,14 @@ int > >> intel_digital_connector_atomic_set_property(struct drm_connector > >> *connector, > >> } > >> > >> int intel_digital_connector_atomic_check(struct drm_connector *conn, > >> - struct drm_connector_state *new_state) > >> + struct drm_atomic_state *state) > >> { > >> + struct drm_connector_state *new_state = > >> + drm_atomic_get_new_connector_state(state, conn); > >>struct intel_digital_connector_state *new_conn_state = > >>to_intel_digital_connector_state(new_state); > >>struct drm_connector_state *old_state = > >> - drm_atomic_get_old_connector_state(new_state->state, conn); > >> + drm_atomic_get_old_connector_state(state, conn); > >>struct intel_digital_connector_state *old_conn_state = > >>to_intel_digital_connector_sta
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
Hi Sean, On Mon, May 13, 2019 at 10:38:58AM -0400, Sean Paul wrote: > On Sat, May 11, 2019 at 3:12 PM Laurent Pinchart wrote: > > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > >> From: Sean Paul > >> > >> Everyone who implements connector_helper_funcs->atomic_check reaches > >> into the connector state to get the atomic state. Instead of continuing > >> this pattern, change the callback signature to just give atomic state > >> and let the driver determine what it does and does not need from it. > >> > >> Eventually all atomic functions should do this, but that's just too much > >> busy work for me. > > > > Given that drivers also access the connector state, isn't this slightly > > more inefficient ? > > Inefficient in terms of what? In terms of the operation having to lookup the connector state, when the caller has it and can easily pass it. As Daniel commented, this may be the price to pay for a cleaner API, but I wonder how much overhead all the state tracking is costing. > Agree that in isolation this patch might seem unnecessary, but it ties > in with the encoder and bridge CLs which accept drm_atomic_state in CLs ? > their hooks. In general the idea is to convert all atomic functions to > take overall atomic state instead of just their object state. Reality > has proven to be more complicated and we need more access than what > the current implementation provides. > > Sean > > >> Changes in v3: > >> - Added to the set > >> > >> Cc: Daniel Vetter > >> Cc: Ville Syrjälä > >> Cc: Jani Nikula > >> Cc: Joonas Lahtinen > >> Cc: Rodrigo Vivi > >> Cc: Ben Skeggs > >> Cc: Laurent Pinchart > >> Cc: Kieran Bingham > >> Cc: Eric Anholt > >> Signed-off-by: Sean Paul > >> --- > >> drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > >> drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > >> drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > >> drivers/gpu/drm/i915/intel_drv.h | 2 +- > >> drivers/gpu/drm/i915/intel_sdvo.c| 9 + > >> drivers/gpu/drm/i915/intel_tv.c | 8 +--- > >> drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > >> drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > >> drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > >> include/drm/drm_modeset_helper_vtables.h | 2 +- > >> 10 files changed, 37 insertions(+), 27 deletions(-) > >> > >> diff --git a/drivers/gpu/drm/drm_atomic_helper.c > >> b/drivers/gpu/drm/drm_atomic_helper.c > >> index 9d9e47276839..fa5a367507c1 100644 > >> --- a/drivers/gpu/drm/drm_atomic_helper.c > >> +++ b/drivers/gpu/drm/drm_atomic_helper.c > >> @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > >> } > >> > >> if (funcs->atomic_check) > >> - ret = funcs->atomic_check(connector, > >> new_connector_state); > >> + ret = funcs->atomic_check(connector, state); > >> if (ret) > >> return ret; > >> > >> @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > >> continue; > >> > >> if (funcs->atomic_check) > >> - ret = funcs->atomic_check(connector, > >> new_connector_state); > >> + ret = funcs->atomic_check(connector, state); > >> if (ret) > >> return ret; > >> } > >> diff --git a/drivers/gpu/drm/i915/intel_atomic.c > >> b/drivers/gpu/drm/i915/intel_atomic.c > >> index b844e8840c6f..e8a5b82e9242 100644 > >> --- a/drivers/gpu/drm/i915/intel_atomic.c > >> +++ b/drivers/gpu/drm/i915/intel_atomic.c > >> @@ -103,12 +103,14 @@ int > >> intel_digital_connector_atomic_set_property(struct drm_connector > >> *connector, > >> } > >> > >> int intel_digital_connector_atomic_check(struct drm_connector *conn, > >> - struct drm_connector_state > >> *new_state) > >> + struct drm_atomic_state *state) > >> { > >> + struct drm_connector_state *new_state = > >> + drm_atomic_get_new_connector_state(state, conn); > >> struct intel_digital_connector_state *new_conn_state = > >> to_intel_digital_connector_state(new_state); > >> struct drm_connector_state *old_state = > >> - drm_atomic_get_old_connector_state(new_state->state, conn); > >> + drm_atomic_get_old_connector_state(state, conn); > >> struct intel_digital_connector_state *old_conn_state = > >> to_intel_digital_connector_state(old_state); > >> struct drm_crtc_state *crtc_state; > >> @@ -118,7 +120,7 @@ int intel_digital_connector_atomic_check(struct > >> drm_connector *conn, > >> if (!new_state->crtc) > >> return 0; > >> > >> - crtc_state = drm_atomic_get_new_crtc_state(new_state->state, > >> new_state->crtc); > >> + crtc_state = drm_atomic_get_new_crtc_state(state, new_state->c
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
On Sat, May 11, 2019 at 10:12:02PM +0300, Laurent Pinchart wrote: > Hi Sean, > > Thank you for the patch. > > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > > From: Sean Paul > > > > Everyone who implements connector_helper_funcs->atomic_check reaches > > into the connector state to get the atomic state. Instead of continuing > > this pattern, change the callback signature to just give atomic state > > and let the driver determine what it does and does not need from it. > > > > Eventually all atomic functions should do this, but that's just too much > > busy work for me. > > Given that drivers also access the connector state, isn't this slightly > more inefficient ? It's atomic code, we're trying to optimize for clean code at the expense of a bit of runtime overhead due to more pointer chasing. And I agree with the general push, the pile of old/new_state pointers of various objects we're passing around is confusing. Passing the overall drm_atomic_state seems much more reasonable, and with that we can get everything else. Plus it's much more obvious whether you have the old/new state (since that's explicit when you look it up from the drm_atomic_state). If we ever see this show up in profile, and it starts mattering, first thing we need is a hashtable I think (atm it's list walking, which is just terrible). But thus far no one cares. -Daniel > > > Changes in v3: > > - Added to the set > > > > Cc: Daniel Vetter > > Cc: Ville Syrjälä > > Cc: Jani Nikula > > Cc: Joonas Lahtinen > > Cc: Rodrigo Vivi > > Cc: Ben Skeggs > > Cc: Laurent Pinchart > > Cc: Kieran Bingham > > Cc: Eric Anholt > > Signed-off-by: Sean Paul > > --- > > drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > > drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > > drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > > drivers/gpu/drm/i915/intel_drv.h | 2 +- > > drivers/gpu/drm/i915/intel_sdvo.c| 9 + > > drivers/gpu/drm/i915/intel_tv.c | 8 +--- > > drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > > drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > > drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > > include/drm/drm_modeset_helper_vtables.h | 2 +- > > 10 files changed, 37 insertions(+), 27 deletions(-) > > > > diff --git a/drivers/gpu/drm/drm_atomic_helper.c > > b/drivers/gpu/drm/drm_atomic_helper.c > > index 9d9e47276839..fa5a367507c1 100644 > > --- a/drivers/gpu/drm/drm_atomic_helper.c > > +++ b/drivers/gpu/drm/drm_atomic_helper.c > > @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > > } > > > > if (funcs->atomic_check) > > - ret = funcs->atomic_check(connector, > > new_connector_state); > > + ret = funcs->atomic_check(connector, state); > > if (ret) > > return ret; > > > > @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > > continue; > > > > if (funcs->atomic_check) > > - ret = funcs->atomic_check(connector, > > new_connector_state); > > + ret = funcs->atomic_check(connector, state); > > if (ret) > > return ret; > > } > > diff --git a/drivers/gpu/drm/i915/intel_atomic.c > > b/drivers/gpu/drm/i915/intel_atomic.c > > index b844e8840c6f..e8a5b82e9242 100644 > > --- a/drivers/gpu/drm/i915/intel_atomic.c > > +++ b/drivers/gpu/drm/i915/intel_atomic.c > > @@ -103,12 +103,14 @@ int > > intel_digital_connector_atomic_set_property(struct drm_connector *connector, > > } > > > > int intel_digital_connector_atomic_check(struct drm_connector *conn, > > -struct drm_connector_state *new_state) > > +struct drm_atomic_state *state) > > { > > + struct drm_connector_state *new_state = > > + drm_atomic_get_new_connector_state(state, conn); > > struct intel_digital_connector_state *new_conn_state = > > to_intel_digital_connector_state(new_state); > > struct drm_connector_state *old_state = > > - drm_atomic_get_old_connector_state(new_state->state, conn); > > + drm_atomic_get_old_connector_state(state, conn); > > struct intel_digital_connector_state *old_conn_state = > > to_intel_digital_connector_state(old_state); > > struct drm_crtc_state *crtc_state; > > @@ -118,7 +120,7 @@ int intel_digital_connector_atomic_check(struct > > drm_connector *conn, > > if (!new_state->crtc) > > return 0; > > > > - crtc_state = drm_atomic_get_new_crtc_state(new_state->state, > > new_state->crtc); > > + crtc_state = drm_atomic_get_new_crtc_state(state, new_state->crtc); > > > > /* > > * These properties are handled by fastset, and might not end > > diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c > > b/drivers/gpu/drm/i915/inte
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
On Sat, May 11, 2019 at 3:12 PM Laurent Pinchart wrote: > > Hi Sean, > > Thank you for the patch. > Hey Laurent, Thanks for looking! > On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > > From: Sean Paul > > > > Everyone who implements connector_helper_funcs->atomic_check reaches > > into the connector state to get the atomic state. Instead of continuing > > this pattern, change the callback signature to just give atomic state > > and let the driver determine what it does and does not need from it. > > > > Eventually all atomic functions should do this, but that's just too much > > busy work for me. > > Given that drivers also access the connector state, isn't this slightly > more inefficient ? > Inefficient in terms of what? Agree that in isolation this patch might seem unnecessary, but it ties in with the encoder and bridge CLs which accept drm_atomic_state in their hooks. In general the idea is to convert all atomic functions to take overall atomic state instead of just their object state. Reality has proven to be more complicated and we need more access than what the current implementation provides. Sean > > Changes in v3: > > - Added to the set > > > > Cc: Daniel Vetter > > Cc: Ville Syrjälä > > Cc: Jani Nikula > > Cc: Joonas Lahtinen > > Cc: Rodrigo Vivi > > Cc: Ben Skeggs > > Cc: Laurent Pinchart > > Cc: Kieran Bingham > > Cc: Eric Anholt > > Signed-off-by: Sean Paul > > --- > > drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > > drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > > drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > > drivers/gpu/drm/i915/intel_drv.h | 2 +- > > drivers/gpu/drm/i915/intel_sdvo.c| 9 + > > drivers/gpu/drm/i915/intel_tv.c | 8 +--- > > drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > > drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > > drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > > include/drm/drm_modeset_helper_vtables.h | 2 +- > > 10 files changed, 37 insertions(+), 27 deletions(-) > > > > diff --git a/drivers/gpu/drm/drm_atomic_helper.c > > b/drivers/gpu/drm/drm_atomic_helper.c > > index 9d9e47276839..fa5a367507c1 100644 > > --- a/drivers/gpu/drm/drm_atomic_helper.c > > +++ b/drivers/gpu/drm/drm_atomic_helper.c > > @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > > } > > > > if (funcs->atomic_check) > > - ret = funcs->atomic_check(connector, > > new_connector_state); > > + ret = funcs->atomic_check(connector, state); > > if (ret) > > return ret; > > > > @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > > continue; > > > > if (funcs->atomic_check) > > - ret = funcs->atomic_check(connector, > > new_connector_state); > > + ret = funcs->atomic_check(connector, state); > > if (ret) > > return ret; > > } > > diff --git a/drivers/gpu/drm/i915/intel_atomic.c > > b/drivers/gpu/drm/i915/intel_atomic.c > > index b844e8840c6f..e8a5b82e9242 100644 > > --- a/drivers/gpu/drm/i915/intel_atomic.c > > +++ b/drivers/gpu/drm/i915/intel_atomic.c > > @@ -103,12 +103,14 @@ int > > intel_digital_connector_atomic_set_property(struct drm_connector *connector, > > } > > > > int intel_digital_connector_atomic_check(struct drm_connector *conn, > > - struct drm_connector_state > > *new_state) > > + struct drm_atomic_state *state) > > { > > + struct drm_connector_state *new_state = > > + drm_atomic_get_new_connector_state(state, conn); > > struct intel_digital_connector_state *new_conn_state = > > to_intel_digital_connector_state(new_state); > > struct drm_connector_state *old_state = > > - drm_atomic_get_old_connector_state(new_state->state, conn); > > + drm_atomic_get_old_connector_state(state, conn); > > struct intel_digital_connector_state *old_conn_state = > > to_intel_digital_connector_state(old_state); > > struct drm_crtc_state *crtc_state; > > @@ -118,7 +120,7 @@ int intel_digital_connector_atomic_check(struct > > drm_connector *conn, > > if (!new_state->crtc) > > return 0; > > > > - crtc_state = drm_atomic_get_new_crtc_state(new_state->state, > > new_state->crtc); > > + crtc_state = drm_atomic_get_new_crtc_state(state, new_state->crtc); > > > > /* > >* These properties are handled by fastset, and might not end > > diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c > > b/drivers/gpu/drm/i915/intel_dp_mst.c > > index 19d81cef2ab6..89cfec128ba0 100644 > > --- a/drivers/gpu/drm/i915/intel_dp_mst.c > > +++ b/drivers/gpu/drm/i915/intel_dp_mst.c > > @@ -143,9 +143,10 @@ static int intel_dp
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
Hi Sean, Thank you for the patch. On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > From: Sean Paul > > Everyone who implements connector_helper_funcs->atomic_check reaches > into the connector state to get the atomic state. Instead of continuing > this pattern, change the callback signature to just give atomic state > and let the driver determine what it does and does not need from it. > > Eventually all atomic functions should do this, but that's just too much > busy work for me. Given that drivers also access the connector state, isn't this slightly more inefficient ? > Changes in v3: > - Added to the set > > Cc: Daniel Vetter > Cc: Ville Syrjälä > Cc: Jani Nikula > Cc: Joonas Lahtinen > Cc: Rodrigo Vivi > Cc: Ben Skeggs > Cc: Laurent Pinchart > Cc: Kieran Bingham > Cc: Eric Anholt > Signed-off-by: Sean Paul > --- > drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > drivers/gpu/drm/i915/intel_drv.h | 2 +- > drivers/gpu/drm/i915/intel_sdvo.c| 9 + > drivers/gpu/drm/i915/intel_tv.c | 8 +--- > drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > include/drm/drm_modeset_helper_vtables.h | 2 +- > 10 files changed, 37 insertions(+), 27 deletions(-) > > diff --git a/drivers/gpu/drm/drm_atomic_helper.c > b/drivers/gpu/drm/drm_atomic_helper.c > index 9d9e47276839..fa5a367507c1 100644 > --- a/drivers/gpu/drm/drm_atomic_helper.c > +++ b/drivers/gpu/drm/drm_atomic_helper.c > @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > } > > if (funcs->atomic_check) > - ret = funcs->atomic_check(connector, > new_connector_state); > + ret = funcs->atomic_check(connector, state); > if (ret) > return ret; > > @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > continue; > > if (funcs->atomic_check) > - ret = funcs->atomic_check(connector, > new_connector_state); > + ret = funcs->atomic_check(connector, state); > if (ret) > return ret; > } > diff --git a/drivers/gpu/drm/i915/intel_atomic.c > b/drivers/gpu/drm/i915/intel_atomic.c > index b844e8840c6f..e8a5b82e9242 100644 > --- a/drivers/gpu/drm/i915/intel_atomic.c > +++ b/drivers/gpu/drm/i915/intel_atomic.c > @@ -103,12 +103,14 @@ int intel_digital_connector_atomic_set_property(struct > drm_connector *connector, > } > > int intel_digital_connector_atomic_check(struct drm_connector *conn, > - struct drm_connector_state *new_state) > + struct drm_atomic_state *state) > { > + struct drm_connector_state *new_state = > + drm_atomic_get_new_connector_state(state, conn); > struct intel_digital_connector_state *new_conn_state = > to_intel_digital_connector_state(new_state); > struct drm_connector_state *old_state = > - drm_atomic_get_old_connector_state(new_state->state, conn); > + drm_atomic_get_old_connector_state(state, conn); > struct intel_digital_connector_state *old_conn_state = > to_intel_digital_connector_state(old_state); > struct drm_crtc_state *crtc_state; > @@ -118,7 +120,7 @@ int intel_digital_connector_atomic_check(struct > drm_connector *conn, > if (!new_state->crtc) > return 0; > > - crtc_state = drm_atomic_get_new_crtc_state(new_state->state, > new_state->crtc); > + crtc_state = drm_atomic_get_new_crtc_state(state, new_state->crtc); > > /* >* These properties are handled by fastset, and might not end > diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c > b/drivers/gpu/drm/i915/intel_dp_mst.c > index 19d81cef2ab6..89cfec128ba0 100644 > --- a/drivers/gpu/drm/i915/intel_dp_mst.c > +++ b/drivers/gpu/drm/i915/intel_dp_mst.c > @@ -143,9 +143,10 @@ static int intel_dp_mst_compute_config(struct > intel_encoder *encoder, > > static int > intel_dp_mst_atomic_check(struct drm_connector *connector, > - struct drm_connector_state *new_conn_state) > + struct drm_atomic_state *state) > { > - struct drm_atomic_state *state = new_conn_state->state; > + struct drm_connector_state *new_conn_state = > + drm_atomic_get_new_connector_state(state, connector); > struct drm_connector_state *old_conn_state = > drm_atomic_get_old_connector_state(state, connector); > struct intel_connector *intel_connector = > @@ -155,7 +156,7 @@ intel_dp_mst_atomic_check(struct drm_connector *connector, > struct
Re: [Intel-gfx] [PATCH v3 04/10] drm: Convert connector_helper_funcs->atomic_check to accept drm_atomic_state
On Thu, May 02, 2019 at 03:49:46PM -0400, Sean Paul wrote: > From: Sean Paul > > Everyone who implements connector_helper_funcs->atomic_check reaches > into the connector state to get the atomic state. Instead of continuing > this pattern, change the callback signature to just give atomic state > and let the driver determine what it does and does not need from it. > > Eventually all atomic functions should do this, but that's just too much > busy work for me. > > Changes in v3: > - Added to the set > > Cc: Daniel Vetter > Cc: Ville Syrjälä > Cc: Jani Nikula > Cc: Joonas Lahtinen > Cc: Rodrigo Vivi > Cc: Ben Skeggs > Cc: Laurent Pinchart > Cc: Kieran Bingham > Cc: Eric Anholt > Signed-off-by: Sean Paul Assuming it compiles everywhere and intel-gfx-ci approves too Acked-by: Daniel Vetter > --- > drivers/gpu/drm/drm_atomic_helper.c | 4 ++-- > drivers/gpu/drm/i915/intel_atomic.c | 8 +--- > drivers/gpu/drm/i915/intel_dp_mst.c | 7 --- > drivers/gpu/drm/i915/intel_drv.h | 2 +- > drivers/gpu/drm/i915/intel_sdvo.c| 9 + > drivers/gpu/drm/i915/intel_tv.c | 8 +--- > drivers/gpu/drm/nouveau/dispnv50/disp.c | 5 +++-- > drivers/gpu/drm/rcar-du/rcar_lvds.c | 12 +++- > drivers/gpu/drm/vc4/vc4_txp.c| 7 --- > include/drm/drm_modeset_helper_vtables.h | 2 +- > 10 files changed, 37 insertions(+), 27 deletions(-) > > diff --git a/drivers/gpu/drm/drm_atomic_helper.c > b/drivers/gpu/drm/drm_atomic_helper.c > index 9d9e47276839..fa5a367507c1 100644 > --- a/drivers/gpu/drm/drm_atomic_helper.c > +++ b/drivers/gpu/drm/drm_atomic_helper.c > @@ -683,7 +683,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > } > > if (funcs->atomic_check) > - ret = funcs->atomic_check(connector, > new_connector_state); > + ret = funcs->atomic_check(connector, state); > if (ret) > return ret; > > @@ -725,7 +725,7 @@ drm_atomic_helper_check_modeset(struct drm_device *dev, > continue; > > if (funcs->atomic_check) > - ret = funcs->atomic_check(connector, > new_connector_state); > + ret = funcs->atomic_check(connector, state); > if (ret) > return ret; > } > diff --git a/drivers/gpu/drm/i915/intel_atomic.c > b/drivers/gpu/drm/i915/intel_atomic.c > index b844e8840c6f..e8a5b82e9242 100644 > --- a/drivers/gpu/drm/i915/intel_atomic.c > +++ b/drivers/gpu/drm/i915/intel_atomic.c > @@ -103,12 +103,14 @@ int intel_digital_connector_atomic_set_property(struct > drm_connector *connector, > } > > int intel_digital_connector_atomic_check(struct drm_connector *conn, > - struct drm_connector_state *new_state) > + struct drm_atomic_state *state) > { > + struct drm_connector_state *new_state = > + drm_atomic_get_new_connector_state(state, conn); > struct intel_digital_connector_state *new_conn_state = > to_intel_digital_connector_state(new_state); > struct drm_connector_state *old_state = > - drm_atomic_get_old_connector_state(new_state->state, conn); > + drm_atomic_get_old_connector_state(state, conn); > struct intel_digital_connector_state *old_conn_state = > to_intel_digital_connector_state(old_state); > struct drm_crtc_state *crtc_state; > @@ -118,7 +120,7 @@ int intel_digital_connector_atomic_check(struct > drm_connector *conn, > if (!new_state->crtc) > return 0; > > - crtc_state = drm_atomic_get_new_crtc_state(new_state->state, > new_state->crtc); > + crtc_state = drm_atomic_get_new_crtc_state(state, new_state->crtc); > > /* >* These properties are handled by fastset, and might not end > diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c > b/drivers/gpu/drm/i915/intel_dp_mst.c > index 19d81cef2ab6..89cfec128ba0 100644 > --- a/drivers/gpu/drm/i915/intel_dp_mst.c > +++ b/drivers/gpu/drm/i915/intel_dp_mst.c > @@ -143,9 +143,10 @@ static int intel_dp_mst_compute_config(struct > intel_encoder *encoder, > > static int > intel_dp_mst_atomic_check(struct drm_connector *connector, > - struct drm_connector_state *new_conn_state) > + struct drm_atomic_state *state) > { > - struct drm_atomic_state *state = new_conn_state->state; > + struct drm_connector_state *new_conn_state = > + drm_atomic_get_new_connector_state(state, connector); > struct drm_connector_state *old_conn_state = > drm_atomic_get_old_connector_state(state, connector); > struct intel_connector *intel_connector = > @@ -155,7 +156,7 @@ intel_dp_mst_atomic_check(struct drm_connector *connector, > struct drm_dp_mst_topology_mgr *mgr; >