Commit graph

64 commits

Author SHA1 Message Date
Simon Ser
1ee86c6290
Add logging functions
This allows turning on and off debug logs, as well as defining a logging
callback.
2019-10-11 12:03:14 +03:00
Simon Ser
3671a64ade
Add opaque structs forward declaration in libliftoff.h 2019-09-30 14:19:27 -04:00
Simon Ser
fe1ee8563d
example/compositor: add flag to specify number of layers 2019-09-30 12:25:38 +03:00
Simon Ser
dd0e550a99
readme: add missing variable decl in example 2019-09-26 10:43:05 +03:00
Simon Ser
de7a7fe5f9
example/compositor: new example 2019-09-25 15:58:38 +03:00
Simon Ser
03679c9303
example/simple: don't hardcode number of layers, use white background 2019-09-25 15:16:14 +03:00
Simon Ser
a30668b06e
example: refactor dumb FB helper
This is necessary for the upcoming composer example.
2019-09-25 15:06:46 +03:00
Simon Ser
a7128585fe
example: extract helpers into separate file 2019-09-25 14:08:07 +03:00
Simon Ser
42b58e8802
example/simple: disable CRTCs we don't use 2019-09-25 12:26:06 +03:00
Simon Ser
35da32683f
Unset CRTC_ID on planes before plane allocation
If a plane was previously enabled on another CRTC, we need to set its CRTC_ID to
zero. Otherwise atomic commits fail with:

    [drm:drm_atomic_check_only [drm]] [PLANE:31:primary A] CRTC set but no FB
2019-09-25 12:23:42 +03:00
Simon Ser
3999474b32
Move example to its own directory 2019-09-16 23:41:19 +03:00
Simon Ser
67549c6c9f
readme: add basic usage 2019-09-16 11:38:54 +03:00
Simon Ser
74040a817b
Rename plane_alloc and plane_data
Rename plane_alloc to alloc_result and plane_data to alloc_step. This is more
accurate and less confusing.
2019-09-16 00:13:04 +03:00
Simon Ser
0596d6f66e
Rename on_top to over
Rename has_composited_layer_on_top to has_composited_layer_over. This is more
consistent with has_allocated_layer_over.

While at it, add a comment to explain why has_allocated_layer_over is named like
so.
2019-09-16 00:00:03 +03:00
Simon Ser
7382dfd346
Add intersection check for layers over the current one
Make sure we don't use planes with lower zpos for layers over the previous ones,
unless there is no collision.

This adds the last intersection check.

Closes: https://github.com/emersion/libliftoff/issues/3
2019-09-15 23:54:29 +03:00
Simon Ser
c63c8c0d15
test: add zpos-4x-disjoint-alt
And restore the original zpos-4x-disjoint test. Depending on the constraints on
the last two layers, one branch of the layer zpos check or the other is taken.

Fixes: 8617ddfd05 ("Add intersection check for planes under the current one")
2019-09-15 23:48:10 +03:00
Simon Ser
e898ce0631
Fix format string in liftoff_display_apply 2019-09-15 23:28:08 +03:00
Simon Ser
00fb6fce21
ci: add .build.yml 2019-09-15 22:44:18 +03:00
Simon Ser
cf861bae35
Remove last_plane_zpos from plane allocation state
Comparing with the last plane zpos isn't enough, we now perform intersection
checks too. The last plane zpos isn't used anymore, remove it.
2019-09-15 22:32:42 +03:00
Simon Ser
8617ddfd05
Add intersection check for planes under the current one
References: https://github.com/emersion/libliftoff/issues/3
2019-09-15 22:31:10 +03:00
Simon Ser
c5d22e1310
Add intersection checks for composited layer on top check
has_composited_layer_on_top is now less strict and performs plane intersection
checks before returning true.

This allows for more layers to be assigned to planes, the relevant tests have
been enabled.

References: https://github.com/emersion/libliftoff/issues/3
2019-09-15 17:08:32 +03:00
Simon Ser
e85a1a5ecf
Add layer_intersects
This will be useful for implementing optimizations in case planes don't collide.
In these cases we can ignore some zpos constraints.

References: https://github.com/emersion/libliftoff/issues/3
2019-09-15 17:00:48 +03:00
Simon Ser
b447a3bc34
test: prevent layers from being assigned to primary plane
In these tests, we always want to use the same layer for the primary plane.
Prevent othr layers from ending up on the primary plane.
2019-09-15 16:56:24 +03:00
Simon Ser
aa2478cda8
Extract plane_data init logic to function
This allows to unclutter output_choose_layers a little.
2019-09-15 15:11:29 +03:00
Simon Ser
364be1c5a6
Document how plane allocation works
Describe the core of the idea in a big comment right before the algorithm
implementation.
2019-09-15 14:31:00 +03:00
Simon Ser
dcf6bc750b
Move composited layer on top constraint up
Move it before applying properties. There's no point in applying properties if
this constraint fails.
2019-09-14 21:24:20 +03:00
Simon Ser
8c93a6fc88
test: add alloc@zpos-4x-domino-partial
This is a variant of alloc@zpos-4x-domino-fail, with plane ordering changed so
that one layer can be mapped to a plane.
2019-09-14 21:18:20 +03:00
Simon Ser
a93ea4a300
test: add alloc@zpos-4x-domino-fail
This test makes sure composited layers make layers underneath fall back to
composition too, in a transitive manner. That is, if plane A is under plane B
which is itself under plane C, plane A falling back to composition means that
both planes B and C fall back to composition too (even if C isn't directly
underneath A).
2019-09-14 21:05:14 +03:00
Simon Ser
07e04a50b6
Make composited layers prevent plane allocation underneath
If a layer cannot be mapped to a plane, make it so all layers under it also fail
plane allocation so that they can be properly composited on the primary plane.

A test was incorrect and has been fixed. Another test had two possible solutions
and has been updated with the one libliftoff yields.
2019-09-14 21:03:30 +03:00
Simon Ser
d9024483c8
test: make our mock drmModeAtomicReq support more properties
We are hitting the 64 properties limit in some new tests.
2019-09-14 20:59:11 +03:00
Simon Ser
a3b8bc33ea
test: add various zpos tests
Some of these are failing, so they are disabled for now.
2019-09-14 12:59:37 +03:00
Simon Ser
3200016033
test: add basic zpos tests
zpos-3x maps three layers to primary, cursor and one of the overlays planes.

zpos-4x-partial sets up 4 layers but is only able to find a plane for 3 layers,
because ordering between the two overlay planes is undefined.
2019-09-13 22:53:15 +03:00
Simon Ser
de8175c998
test: add an alloc test with 3 layers 2019-09-13 22:35:23 +03:00
Simon Ser
060023c6f1
test: fix atomic commit layer compatibility
Used wrong index while iterating.
2019-09-13 22:34:34 +03:00
Simon Ser
fccdebfd6e
test: introduce structured alloc tests
This is a new kind of tests defined by a static structure. The structure
contains the layers and planes to setup as well as the expected result.
2019-09-13 19:47:00 +03:00
Simon Ser
54bb501cae
test: add liftoff_mock_drm_get_plane
Allows to get a mock plane from a plane ID.
2019-09-13 19:46:21 +03:00
Simon Ser
1cd18f9156
test: mock atomic commit improvements
Don't fail the commit if FB_ID and CRTC_ID are set to zero. Print the reason
why the commit fails.
2019-09-13 19:44:12 +03:00
Simon Ser
a1e3798f6f
Use PRIu32 for printing plane IDs 2019-09-13 19:43:01 +03:00
Simon Ser
5afeeb531a
Add test framework
When testing the library, the libdrm dependency is swapped with a mock libdrm.
The test sets up the mock libdrm's internal state and then runs the test.
2019-09-13 10:33:49 +03:00
Simon Ser
bb887aa7d0
Fix segfault when dup fails in liftoff_display_create 2019-09-12 20:15:50 +03:00
Simon Ser
caac9bc87f
Rename library to liftoff
Let's make these planes useful!
2019-09-12 11:40:21 +03:00
Simon Ser
c4a8afb39b
Add basic support for zpos
This commit adds special handling for the layer zpos property. Setting it won't
forward it to KMS, instead it will be handled internally.

The plane zpos (either retrieved from KMS if the property is supported, or
guessed from the plane type and ID) is checked during plane allocation to have
the same ordering as the layer zpos. This adds more constraints to the
allocation.

Note that the new constraints are too strict. In case two layers don't overlap,
there's no need for one plane to be on top of the other. See [1].

Also, the new constraints don't account for underlay planes [2].

[1]: https://github.com/emersion/libhwc/issues/3
[2]: https://github.com/emersion/libhwc/issues/7

Closes: https://github.com/emersion/libhwc/issues/4
2019-09-12 11:27:00 +03:00
Simon Ser
fc901d0e26
Fix early bailout when we already have a better allocation
In case we can't do better than the current best allocation even if we manage
to allocate all remaining planes, we bail out early. However the bailout
condition was incorrect and didn't account for the current score.
2019-09-12 10:17:03 +03:00
Simon Ser
1509e50fd5
Extract allocation arguments to a struct
Just trying to keep the number of arguments under a reasonable limit. More
arguments will be added in the future.
2019-09-11 19:23:35 +03:00
Simon Ser
4d641041bb
Don't assign layer to plane if missing props
Sometimes the properties applied on a layer are missing on a plane. Instead of
making the whole allocation fail, just mark the plane as incompatible with the
layer.
2019-09-10 22:58:29 +03:00
Simon Ser
e2df5e73d0
List planes by allocation order
Given a primary plane with zpos=0 and overlay planes with zpos 1 2, we want
first to allocate the primary plane, then the furthest overlay plane, then the
remaining one: 0 2 1.

References: https://github.com/emersion/libhwc/issues/3
2019-09-09 20:55:43 +03:00
Simon Ser
dac885222b
Add support for reading immutable plane zpos
If zpos is present, instead of guessing the zpos from the plane type, just use
it. If two planes have the same zpos, we consider their relative position as
undefined.

We don't handle the case where some planes have zpos set but some others don't.

References: https://github.com/emersion/libhwc/issues/4
2019-09-09 19:46:44 +03:00
Simon Ser
42db49c6fe
Guess zpos from plane type
Many drivers don't expose the zpos plane property. We can guess it from the
plane type though: primary, overlay, cursor. Except some drivers expose
underlays too: underlays, primary, overlay, cursor. Underlay planes don't have a
special type, they are just marked as overlays. To detect them, check whether
their plane ID is less than the primary plane ID.

References: https://github.com/emersion/libhwc/issues/3
2019-09-08 18:33:09 +03:00
Simon Ser
d90d7bdc36
Check plane is compatible with CRTC
This allows us to rule out some potential solutions.

Closes: https://github.com/emersion/libhwc/issues/2
2019-09-08 17:41:25 +03:00
Simon Ser
66bedbc854
Reset best allocation to NULL
Without this, if no allocation has been found, garbage will be read from
alloc.best.
2019-09-08 17:39:55 +03:00