Commit Graph

143 Commits

Author SHA1 Message Date
Simon Zeni 42549a1c9a backend/drm: stop initializing backend renderer 2021-11-18 09:37:57 -05:00
Simon Ser e13f3f8608 backend/drm: handle per-connector hotplug events
When a connector ID is specified in a hotplug event, don't scan all
connectors. Only scan the connector that has changed.
2021-11-02 13:36:43 +01:00
Simon Ser 0817c52a21 backend/drm: get rid of BO handle table
The BO handle table exists to avoid double-closing a BO handle,
which aren't reference-counted by the kernel. But if we can
guarantee that there is only ever a single ref for each BO handle,
then we don't need the BO handle table anymore.

This is possible if we create the handle right before the ADDFB2
IOCTL, and close the handle right after. The handles are very
short-lived and we don't need to track their lifetime.

Because of multi-planar FBs, we need to be a bit careful: some
FB planes might share the same handle. But with a small check, it's
easy to avoid double-closing the same handle (which wouldn't be a
big deal anyways).

There's one gotcha though: drmModeSetCursor2 takes a BO handle as
input. Saving the handles until drmModeSetCursor2 time would require
us to track BO handle lifetimes, so we wouldn't be able to get rid
of the BO handle table. As a workaround, use drmModeGetFB to turn the
FB ID back to a BO handle, call drmModeSetCursor2 and then immediately
close the BO handle. The overhead should be minimal since these IOCTLs
are pretty cheap.

Closes: https://github.com/swaywm/wlroots/issues/3164
2021-10-29 11:38:37 +02:00
Simon Ser 3d0848daae backend: create renderer and allocator in wlr_backend_autocreate
Instead of ensuring the renderer and allocator are initialized in each
backend, do it in wlr_backend_autocreate. This allows compositors to
create backends without any renderer/allocator if they side-step
wlr_backend_autocreate.

Since the wlr_backend_get_renderer and backend_get_allocator end up
calling wlr_renderer_autocreate and wlr_allocator_autocreate, it sounds
like a good idea to centralize all of the opimionated bits in one place.
2021-09-30 08:50:43 -06:00
Simon Ser 274c8189d4 backend/drm: fix crash on VT switch
Don't set the MODE flag when disabling a CRTC. This fixes a NULL
pointer dereference in drm_connector_state_init.
2021-09-07 11:34:53 +02:00
Simon Ser 5dfaf5ea9c backend/drm: introduce wlr_drm_bo_handle_table
Using GBM to import DRM dumb buffers tends to not work well. By
using GBM we're calling some driver-specific functions in Mesa.
These functions check whether Mesa can work with the buffer.
Sometimes Mesa has requirements which differ from DRM dumb buffers
and the GBM import will fail (e.g. on amdgpu).

Instead, drop GBM and use drmPrimeFDToHandle directly. But there's
a twist: BO handles are not ref'counted by the kernel and need to
be ref'counted in user-space [1]. libdrm usually performs this
bookkeeping and is used under-the-hood by Mesa.

We can't re-use libdrm for this task without using driver-specific
APIs. So let's just re-implement the ref'counting logic in wlroots.
The wlroots implementation is inspired from amdgpu's in libdrm [2].

Closes: https://github.com/swaywm/wlroots/issues/2916

[1]: https://gitlab.freedesktop.org/mesa/drm/-/merge_requests/110
[2]: 1a4c0ec9ae/amdgpu/handle_table.c
2021-08-25 10:05:37 -04:00
Simon Ser eb0ce659cf backend/drm: add proper error handling to wlr_drm_backend_create
Some listeners weren't removed and caused a use-after-free with
e.g. vkms when used as a secondary GPU.
2021-08-10 15:42:36 +02:00
Simon Ser c55f70c8b7 backend/drm: stop initializing renderer for parent backend
Unless we're dealing with a multi-GPU setup and the backend being
initialized is secondary, we don't need a renderer nor an allocator.
Stop initializing these.
2021-07-28 22:52:35 +02:00
Simon Ser c74dc45bb6 backend/drm: drop get_renderer implementation
We can now just rely on the common code for this.
2021-07-28 22:52:35 +02:00
Simon Ser cc8bc0db20 backend/drm: stop restoring CRTCs on exit
This is the cause of the spurious "drmHandleEvent failed" messages
at exit. restore_drm_outputs calls handle_drm_event in a loop without
checking whether the FD is readable, so drmHandleEvent ends up with a
short read (0 bytes) and returns an error.

The loop's goal is to wait for all queued page-flip events to complete,
to allow drmModeSetCrtc calls to succeed without EBUSY. The
drmModeSetCrtc calls are supposed to restore whatever KMS state we were
started with. But it's not clear from my PoV that restoring the KMS
state on exit is desirable.

KMS clients are supposed to save and restore the (full) KMS state on VT
switch, but not on exit. Leaving our KMS state on exit avoids unnecessary
modesets and allows flicker-free transitions between clients. See [1]
for more details, and note that with Pekka we've concluded that a new
flag to reset some KMS props to their default value on compositor
start-up is the best way forward. As a side note, Weston doesn't restore
the CRTC by does disable the cursor plane on exit (see
drm_output_deinit_planes, I still think disabling the cursor plane
shouldn't be necessary on exit).

Additionally, restore_drm_outputs only a subset of the KMS state.
Gamma and other atomic properties aren't accounted for. If the previous
KMS client had some outputs disabled, restore_drm_outputs would restore
a garbage mode.

[1]: https://blog.ffwll.ch/2016/01/vt-switching-with-atomic-modeset.html
2021-07-20 15:33:26 +02:00
Simon Ser f94eb174c7 backend/drm: fix NULL data in handle_drm_event
wl_event_loop_add_fd was called with a NULL data argument, but the
function expects the data argument to be set to the wlr_drm_backend.

Fixes: 053ebe7c27 ("backend/drm: terminate display on drmHandleEvent failure")
2021-07-19 09:01:21 -04:00
Simon Ser d71ed635b9 backend/drm: force linear layout for multi-GPU buffers
Some buffers need to be copied across GPUs. Such buffers need to be
allocated with a format and modifier suitable for both the source
and the destination.

When explicit modifiers aren't supported, we were forcing the buffers
to be allocated with a linear layout, because implicit modifiers
aren't portable across GPUs. All is well with this case.

When explicit modifiers are supported, we were advertising the whole
list of destination modifiers, in the hope that the source might
have some in common and might be able to allocate a buffer with a
more optimized layout. This works well if the source supports explicit
modifiers. However, if the source doesn't, then wlr_drm_format_intersect
will fallback to implicit modifiers, and everything goes boom: the
source uses a GPU-specific tiling and the destination interprets it
as linear.

To avoid this, just force linear unconditionally. We'll be able to
revert this once we have a good way to indicate that an implicit modifier
isn't supported in wlr_drm_format_set, see [1].

[1]: https://github.com/swaywm/wlroots/pull/2815

Closes: https://github.com/swaywm/wlroots/issues/3030
2021-07-08 17:06:20 +02:00
Simon Ser a38baec1f8 buffer: make enum wlr_buffer_cap public
Custom backends and renderers need to implement
wlr_backend_impl.get_buffer_caps and
wlr_renderer_impl.get_render_buffer_caps. They can't if enum
wlr_buffer_cap isn't made public.
2021-07-01 16:40:19 -04:00
Simon Ser dbb0e2f75b Remove unused wlr_list.h includes 2021-07-01 10:35:39 -04:00
Simon Ser 9e9be83a58 backend/drm: implement get_cursor_formats and get_cursor_size 2021-06-02 11:08:52 +02:00
Simon Ser 7ec5bf6b10 backend: introduce wlr_backend_finish
This new functions cleans up the common backend state. While this
currently only emits the destroy signal, this will also clean up
the renderer and allocator in upcoming patches.
2021-05-21 22:13:54 +02:00
Simon Ser a1e8a639b3 backend/drm: introduce drm_connector_commit_state
Backend-initiated mode changes can use this function instead of
going through drm_connector_set_mode. drm_connector_set_mode becomes
a mere drm_connector_commit_state helper.
2021-04-29 19:51:57 +02:00
Simon Ser 218955ce95 backend/drm: remove mode arg from drm_connector_set_mode
All of the information is in wlr_output_state.
2021-04-29 19:51:57 +02:00
Simon Ser c9c2d7539c backend/drm: fill scratch wlr_output_state for set_mode
Populate the wlr_output_state when setting a mode. This will allow
drm_connector_set_mode to stop relying on ephemeral fields in
wlr_drm_crtc. Also drm_connector_set_mode will be able to apply
both a new buffer and a new mode atomically.
2021-04-29 19:51:57 +02:00
Simon Ser 1a9701cd7c backend/drm: take wlr_output_state as arg in commit callers
Instead of relying on wlr_output.pending to be empty when performing
backend-initiated CRTC commits, use a zero wlr_output_state.
2021-04-29 19:51:57 +02:00
Simon Zeni 144189674e backend: introduce backend_get_buffer_caps 2021-04-28 20:55:57 +02:00
Simon Ser 9b0e0970f9 backend/drm: destroy backend on udev remove event
Any use of the DRM FD after the remove event results in a "Permission
denied" error.
2021-04-27 09:11:44 +02:00
Simon Ser c49ea9ef4f backend/drm: destroy when parent is destroyed 2021-04-27 09:11:44 +02:00
Manuel Stoeckl a290d7a78d Make implementation function lists static const
This requires a change to the type of `struct wlr_tablet` and
`wlr_tablet_init` signature, both of which are part of the unstable API.
2021-02-05 10:04:20 +01:00
Simon Ser b7c95d483a backend/drm: implement get_drm_fd 2021-01-16 08:57:42 +01:00
Simon Ser d9bbc416a6 backend/drm: re-use FBs
Instead of importing buffers to GBM and KMS at each frame, cache them
and re-use them while the wlr_buffer is alive.

This is the same as [1] and [2] but for the DRM backend.

[1]: https://github.com/swaywm/wlroots/pull/2538
[2]: https://github.com/swaywm/wlroots/pull/2539
2021-01-10 11:24:35 +01:00
Simon Zeni 9192c0480a remove unnecessary egl includes 2021-01-07 17:11:22 +01:00
Simon Ser 576ff57db0 backend/drm: ignore hotplug events while inactive
When the session is inactive, we can't change the KMS state. Ignore
hotplug events so that compositors don't try to perform a modeset when
a connector is plugged in. We already re-scan connectors when the
session becomes active.

To test, run a wlroots compositor on VT 1, switch to VT 2, unplug a
connector, re-plug it, switch back to VT 1. Without this patch the
screen is black on VT 1.

References: https://github.com/swaywm/wlroots/issues/2370
2021-01-03 19:34:54 +01:00
Simon Ser 4b03bdc3ab Remove wlr_create_renderer_func_t
This callback allowed compositors to customize the EGL config used by
the renderer. However with renderer v6 EGL configs aren't used anymore.
Instead, buffers are allocated via GBM and GL FBOs are rendered to. So
customizing the EGL config is a no-op.
2020-12-30 17:09:40 +01:00
Simon Ser c5f239f411
backend/drm: make listener names more idiomatic
Use the "<object>_<event>" notation for listeners, use
"handle_<listener>" for handlers.
2020-12-25 14:45:00 +01:00
Simon Ser 248c7787c7 backend/drm: refactor wlr_output destruction
Instead of hand-rolling our own manual wlr_output cleanup function, rely
on wlr_output_destroy to remove an output from the compositor's state.
2020-12-18 10:48:44 +01:00
Simon Ser 9cd3f03f65
backend/drm: add wlr_drm_backend.name
Save the DRM device name in a wlr_drm_backend field, so that we can
easily use it for logging purposes.
2020-12-17 20:48:47 +01:00
Simon Ser 0e76f92de7 backend/session: replace session_signal with events.active
This is more idiomatic wlroots API. The new name makes it clear that the
signal is emitted when wlr_session.active changes.
2020-11-19 22:47:49 +01:00
Simon Ser 44a4792fd8 backend/session: operate on wlr_device
Instead of operating on FDs in {open,close}_device, operate on
wlr_devices. This avoids the device lookup in wlr_session and allows
callers to have access to wlr_device fields.

For now, we use it to remove wlr_session_signal_add and replace it with
a more idiomatic wlr_session.events.change field. In the future, other
events will be added.
2020-11-19 22:47:49 +01:00
Simon Ser 61095f4a12 backend/drm: make drm_connector_set_mode take a wlr_drm_connector
Since this is an internal DRM backend function, there's no reason we
need to take a generic wlr_output.
2020-05-26 23:34:37 +02:00
Simon Ser 445750aa9a backend/drm: remove enable_drm_connector
Merge enable_drm_connector into drm_connector_set_mode. This allows us
to de-duplicate logic since enabling an output performs a modeset.
2020-05-26 23:34:37 +02:00
Simon Ser f72686c0b6 backend/drm: fix crash in session_signal
conn->crtc is NULL in case the output is disabled.

However, the DRM backend will set the GAMMA_LUT property anyway. On each
commit the whole state is sent to KMS. Adding WLR_DRM_CRTC_GAMMA_LUT to
the pending state would just make the backend re-create the blob
containing the gamma LUT.
2020-05-10 12:07:05 +02:00
Simon Ser da63d11d34 backend/drm: remove crtc_set_cursor from interface 2020-05-09 16:42:25 +02:00
Simon Ser 70883fd10b backend/drm: apply gamma LUT on page-flip 2020-05-09 16:42:25 +02:00
Simon Ser c9d6b18eef backend/drm: remove wlr_drm_interface.crtc_move_cursor
Instead, make the legacy backend call drmModeMoveCursor on page-flip.
2020-05-09 16:42:25 +02:00
Scott Anderson 8da9d9679e backend/drm: introduce wlr_drm_fb
This is a type which manages gbm_surfaces and imported dmabufs in the
same place, and makes the lifetime management between the two shared. It
should lead to easier to understand code, and fewer special cases.

This also contains a fair bit of refactoring to start using this new
type.

Co-authored-by: Simon Ser <contact@emersion.fr>
2020-04-28 09:54:52 +02:00
Simon Ser 21e1953b61 backend/drm: don't modeset with a NULL mode after TTY switch
This fixes a segfault in drm_connector_set_mode (mode = NULL).

This happens because we set wlr_output.enabled to true if the connector
is attached to the CRTC. When the user disables an output in the
wlroots-based compositor, switches to another VT (enabling the output),
then switches back, wlroots sets wlr_output.enabled to true but
wlr_output.current_mode is NULL.

We should consider not reading properties from KMS after a TTY switch,
disabling all connectors. However this may result in flickering (outputs
being disabled then re-enabled).

Closes: https://github.com/swaywm/wlroots/issues/1874
2019-12-30 14:32:37 -07:00
Simon Ser 8681e4ab8a backend/drm, backend/libinput: listen to session destroy
This fixes a heap-use-after-free when the session is destroyed before
the backend during wl_display_destroy:

    ==1085==ERROR: AddressSanitizer: heap-use-after-free on address 0x614000000180 at pc 0x7f88e3590c2d bp 0x7ffdc4e33f90 sp 0x7ffdc4e33f80
    READ of size 8 at 0x614000000180 thread T0
        #0 0x7f88e3590c2c in find_device ../subprojects/wlroots/backend/session/session.c:192
        #1 0x7f88e3590e85 in wlr_session_close_file ../subprojects/wlroots/backend/session/session.c:204
        #2 0x7f88e357b80c in libinput_close_restricted ../subprojects/wlroots/backend/libinput/backend.c:24
        #3 0x7f88e21af274  (/lib64/libinput.so.10+0x28274)
        #4 0x7f88e21aff1d  (/lib64/libinput.so.10+0x28f1d)
        #5 0x7f88e219ddac  (/lib64/libinput.so.10+0x16dac)
        #6 0x7f88e21b415d in libinput_unref (/lib64/libinput.so.10+0x2d15d)
        #7 0x7f88e357c9d6 in backend_destroy ../subprojects/wlroots/backend/libinput/backend.c:130
        #8 0x7f88e3545a09 in wlr_backend_destroy ../subprojects/wlroots/backend/backend.c:50
        #9 0x7f88e358981a in multi_backend_destroy ../subprojects/wlroots/backend/multi/backend.c:54
        #10 0x7f88e358a059 in handle_display_destroy ../subprojects/wlroots/backend/multi/backend.c:107
        #11 0x7f88e314acde  (/lib64/libwayland-server.so.0+0x8cde)
        #12 0x7f88e314b466 in wl_display_destroy (/lib64/libwayland-server.so.0+0x9466)
        #13 0x559fefb52385 in main ../main.c:67
        #14 0x7f88e2639152 in __libc_start_main (/lib64/libc.so.6+0x27152)
        #15 0x559fefb4297d in _start (/home/simon/src/glider/build/glider+0x2297d)

    0x614000000180 is located 320 bytes inside of 416-byte region [0x614000000040,0x6140000001e0)
    freed by thread T0 here:
        #0 0x7f88e3d0a6b0 in __interceptor_free /build/gcc/src/gcc/libsanitizer/asan/asan_malloc_linux.cc:122
        #1 0x7f88e35b51fb in logind_session_destroy ../subprojects/wlroots/backend/session/logind.c:270
        #2 0x7f88e35905a4 in wlr_session_destroy ../subprojects/wlroots/backend/session/session.c:156
        #3 0x7f88e358f440 in handle_display_destroy ../subprojects/wlroots/backend/session/session.c:65
        #4 0x7f88e314acde  (/lib64/libwayland-server.so.0+0x8cde)

    previously allocated by thread T0 here:
        #0 0x7f88e3d0acd8 in __interceptor_calloc /build/gcc/src/gcc/libsanitizer/asan/asan_malloc_linux.cc:153
        #1 0x7f88e35b911c in logind_session_create ../subprojects/wlroots/backend/session/logind.c:746
        #2 0x7f88e358f6b4 in wlr_session_create ../subprojects/wlroots/backend/session/session.c:91
        #3 0x559fefb51ea6 in main ../main.c:20
        #4 0x7f88e2639152 in __libc_start_main (/lib64/libc.so.6+0x27152)
2019-12-01 10:26:12 -05:00
Simon Ser ca45f4490c Remove all wayland-server.h includes
The documentation for wayland-server.h says:

> Use of this header file is discouraged. Prefer including
> wayland-server-core.h instead, which does not include the server protocol
> header and as such only defines the library PI, excluding the deprecated API
> below.

Replacing wayland-server.h with wayland-server-core.h allows us to drop the
WL_HIDE_DEPRECATED declaration.
2019-07-27 15:49:32 -04:00
Brian Ashworth 50011e7170 backend/drm: fix modeset on drm fd resume
On DRM resume, such as switching back to a TTY, the output needs to be
modeset to the current mode. However, wlr_output_set_mode will return
early when attempting to set the mode to the current mode. This just
steps around wlr_output_set_mode and calls drm_connector_set_mode
directly.
2019-02-19 16:49:51 +01:00
Vincent Vanlaer 7bc43413ed Allow cursor render surface to be used as fb
In order for a surface to be used as a cursor plane framebuffer, it
appears that requiring the buffer to be linear is sufficient.

GBM_BO_USE_SCANOUT is added in case GBM_BO_USE_LINEAR isn't sufficient
on untested hardware.

Fixes #1323

Removed wlr_drm_plane.cursor_bo as it does not serve any purpose
anymore.

Relevant analysis (taken from the PR description):

While trying to implement a fix for #1323, I found that when exporting
the rendered surface into a DMA-BUF and reimporting it with
`GBM_BO_USE_CURSOR`, the resulting object does not appear to be valid.
After some digging (turning on drm-kms debugging and switching to legacy
mode), I managed to extract the following error: ```
[drm:__setplane_check.isra.1 [drm]] Invalid pixel format AR24
little-endian (0x34325241), modifier 0x100000000000001 ``` The format
itself refers to ARGB8888 which is the same format as
`renderer->gbm_format` used in master to create the cursor bo. However,
using `gbm_bo_create` with `GBM_BO_USE_CURSOR` results in a modifier of
0. A modifier of zero represents a linear buffer while the modifier of
the surface that is rendered to is  `I915_FORMAT_MOD_X_TILED` (see
https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/include/uapi/drm/drm_fourcc.h?h=v4.20.6#n263).
In order to fix this mismatch in modifier, I added the
`GBM_BO_USE_LINEAR` to the render surface and everything started to work
just fine. I wondered however, whether the export and import is really
necessary. I then decided to test if the back buffer of the render
surface works as well, and at least on my hardware (Intel HD 530 and
Intel UHD 620) it does. This is the patch in this PR and this requires
no exporting and importing.

I have to note that I cheated in order to import DMA_BUFs into a cursor
bo when doing the first tests, since on import the Intel drivers check
that the cursor is 64x64. This is strange since cursor sizes other than
64x64 have been around for quite some time now
(https://lists.freedesktop.org/archives/mesa-commit/2014-June/050268.html).
Removing this check made everything work fine. I later (while writing
this PR) found out that `__DRI_IMAGE_USE_CURSOR` (to which
`GBM_BO_USE_CURSOR` translates) has been deprecated in mesa
(https://gitlab.freedesktop.org/mesa/mesa/blob/master/include/GL/internal/dri_interface.h#L1296),
which makes me wonder what the usecase of `GBM_BO_USE_CURSOR` is. The
reason we never encountered this is that when specifying
`GBM_BO_USE_WRITE`, a dumb buffer is created trough DRM and the usage
flag never reaches the Intel driver directly. The relevant code is in
https://gitlab.freedesktop.org/mesa/mesa/blob/master/src/gbm/backends/dri/gbm_dri.c#L1011-1089
. From this it seems that as long as the size, format and modifiers are
right, any surface can be used as a cursor.
2019-02-04 20:47:07 +01:00
emersion b0635bf3e7 Rename get_present_clock to get_presentation clock, use it 2018-10-04 22:00:22 +02:00
emersion 54e1287f30 backend: add get_present_clock 2018-10-04 21:58:17 +02:00
Drew DeVault 841d04db99
Merge pull request #1262 from nyorain/session_fix
Improve session handling
2018-10-04 17:36:39 +02:00
emersion e88db9a3fb backend/drm: reset gamma table on VT switch 2018-10-03 10:53:35 +02:00