2019-09-13 22:30:28

by Lyude Paul

[permalink] [raw]
Subject: [PATCH 0/3] drm/encoder: Various doc fixes

Some random issues with documentation that I noticed while working on
nouveau the other day. There are no functional changes in this series.

Lyude Paul (3):
drm/encoder: Fix possible_clones documentation
drm/encoder: Fix possible_crtcs documentation
drm/encoder: Don't raise voice in drm_encoder_mask() documentation

include/drm/drm_encoder.h | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)

--
2.21.0


2019-09-14 13:45:32

by Lyude Paul

[permalink] [raw]
Subject: [PATCH 1/3] drm/encoder: Fix possible_clones documentation

We say that all of the bits in possible_clones must be set before
calling drm_encoder_init(). This isn't true though, since:

* The driver may not even have all of the encoder objects that could be
used as clones initialized at that point
* possible_crtcs isn't used at all outside of userspace, so it's not
actually needed to initialize it until drm_dev_register()

So, fix it.

Signed-off-by: Lyude Paul <[email protected]>
---
include/drm/drm_encoder.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/drm/drm_encoder.h b/include/drm/drm_encoder.h
index 70cfca03d812..81273b50b3f6 100644
--- a/include/drm/drm_encoder.h
+++ b/include/drm/drm_encoder.h
@@ -154,7 +154,7 @@ struct drm_encoder {
* using drm_encoder_index() as the index into the bitfield. The driver
* must set the bits for all &drm_encoder objects which can clone a
* &drm_crtc together with this encoder before calling
- * drm_encoder_init(). Drivers should set the bit representing the
+ * drm_dev_register(). Drivers should set the bit representing the
* encoder itself, too. Cloning bits should be set such that when two
* encoders can be used in a cloned configuration, they both should have
* each another bits set.
--
2.21.0

2019-09-14 13:50:53

by Lyude Paul

[permalink] [raw]
Subject: [PATCH 2/3] drm/encoder: Fix possible_crtcs documentation

Similar to possible_clones, we don't actually use possible_crtcs until
the driver is registered with userspace. So, fix the documentation to
indicate this.

Signed-off-by: Lyude Paul <[email protected]>
---
include/drm/drm_encoder.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/include/drm/drm_encoder.h b/include/drm/drm_encoder.h
index 81273b50b3f6..d65173d413b7 100644
--- a/include/drm/drm_encoder.h
+++ b/include/drm/drm_encoder.h
@@ -140,7 +140,7 @@ struct drm_encoder {
* @possible_crtcs: Bitmask of potential CRTC bindings, using
* drm_crtc_index() as the index into the bitfield. The driver must set
* the bits for all &drm_crtc objects this encoder can be connected to
- * before calling drm_encoder_init().
+ * before calling drm_dev_register().
*
* In reality almost every driver gets this wrong.
*
--
2.21.0

2019-09-17 13:31:44

by Daniel Vetter

[permalink] [raw]
Subject: Re: [PATCH 0/3] drm/encoder: Various doc fixes

On Fri, Sep 13, 2019 at 06:27:00PM -0400, Lyude Paul wrote:
> Some random issues with documentation that I noticed while working on
> nouveau the other day. There are no functional changes in this series.

Nice! On all three:

Reviewed-by: Daniel Vetter <[email protected]>

>

> Lyude Paul (3):
> drm/encoder: Fix possible_clones documentation
> drm/encoder: Fix possible_crtcs documentation
> drm/encoder: Don't raise voice in drm_encoder_mask() documentation
>
> include/drm/drm_encoder.h | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> --
> 2.21.0
>

--
Daniel Vetter
Software Engineer, Intel Corporation
http://blog.ffwll.ch