xref: /linux/Documentation/gpu/todo.rst (revision 99a97a8ba9881fc47901ff36b057e5cd0bf06af0)
1.. _todo:
2
3=========
4TODO list
5=========
6
7This section contains a list of smaller janitorial tasks in the kernel DRM
8graphics subsystem useful as newbie projects. Or for slow rainy days.
9
10Subsystem-wide refactorings
11===========================
12
13De-midlayer drivers
14-------------------
15
16With the recent ``drm_bus`` cleanup patches for 3.17 it is no longer required
17to have a ``drm_bus`` structure set up. Drivers can directly set up the
18``drm_device`` structure instead of relying on bus methods in ``drm_usb.c``
19and ``drm_platform.c``. The goal is to get rid of the driver's ``->load`` /
20``->unload`` callbacks and open-code the load/unload sequence properly, using
21the new two-stage ``drm_device`` setup/teardown.
22
23Once all existing drivers are converted we can also remove those bus support
24files for USB and platform devices.
25
26All you need is a GPU for a non-converted driver (currently almost all of
27them, but also all the virtual ones used by KVM, so everyone qualifies).
28
29Contact: Daniel Vetter, Thierry Reding, respective driver maintainers
30
31Switch from reference/unreference to get/put
32--------------------------------------------
33
34For some reason DRM core uses ``reference``/``unreference`` suffixes for
35refcounting functions, but kernel uses ``get``/``put`` (e.g.
36``kref_get``/``put()``). It would be good to switch over for consistency, and
37it's shorter. Needs to be done in 3 steps for each pair of functions:
38
39* Create new ``get``/``put`` functions, define the old names as compatibility
40  wrappers
41* Switch over each file/driver using a cocci-generated spatch.
42* Once all users of the old names are gone, remove them.
43
44This way drivers/patches in the progress of getting merged won't break.
45
46Contact: Daniel Vetter
47
48Convert existing KMS drivers to atomic modesetting
49--------------------------------------------------
50
513.19 has the atomic modeset interfaces and helpers, so drivers can now be
52converted over. Modern compositors like Wayland or Surfaceflinger on Android
53really want an atomic modeset interface, so this is all about the bright
54future.
55
56There is a conversion guide for atomic and all you need is a GPU for a
57non-converted driver (again virtual HW drivers for KVM are still all
58suitable).
59
60As part of this drivers also need to convert to universal plane (which means
61exposing primary & cursor as proper plane objects). But that's much easier to
62do by directly using the new atomic helper driver callbacks.
63
64Contact: Daniel Vetter, respective driver maintainers
65
66Clean up the clipped coordination confusion around planes
67---------------------------------------------------------
68
69We have a helper to get this right with drm_plane_helper_check_update(), but
70it's not consistently used. This should be fixed, preferrably in the atomic
71helpers (and drivers then moved over to clipped coordinates). Probably the
72helper should also be moved from drm_plane_helper.c to the atomic helpers, to
73avoid confusion - the other helpers in that file are all deprecated legacy
74helpers.
75
76Contact: Ville Syrjälä, Daniel Vetter, driver maintainers
77
78Implement deferred fbdev setup in the helper
79--------------------------------------------
80
81Many (especially embedded drivers) want to delay fbdev setup until there's a
82real screen plugged in. This is to avoid the dreaded fallback to the low-res
83fbdev default. Many drivers have a hacked-up (and often broken) version of this,
84better to do it once in the shared helpers. Thierry has a patch series, but that
85one needs to be rebased and final polish applied.
86
87Contact: Thierry Reding, Daniel Vetter, driver maintainers
88
89Convert early atomic drivers to async commit helpers
90----------------------------------------------------
91
92For the first year the atomic modeset helpers didn't support asynchronous /
93nonblocking commits, and every driver had to hand-roll them. This is fixed
94now, but there's still a pile of existing drivers that easily could be
95converted over to the new infrastructure.
96
97One issue with the helpers is that they require that drivers handle completion
98events for atomic commits correctly. But fixing these bugs is good anyway.
99
100Contact: Daniel Vetter, respective driver maintainers
101
102Better manual-upload support for atomic
103---------------------------------------
104
105This would be especially useful for tinydrm:
106
107- Add a struct drm_rect dirty_clip to drm_crtc_state. When duplicating the
108  crtc state, clear that to the max values, x/y = 0 and w/h = MAX_INT, in
109  __drm_atomic_helper_crtc_duplicate_state().
110
111- Move tinydrm_merge_clips into drm_framebuffer.c, dropping the tinydrm_
112  prefix ofc and using drm_fb_. drm_framebuffer.c makes sense since this
113  is a function useful to implement the fb->dirty function.
114
115- Create a new drm_fb_dirty function which does essentially what e.g.
116  mipi_dbi_fb_dirty does. You can use e.g. drm_atomic_helper_update_plane as the
117  template. But instead of doing a simple full-screen plane update, this new
118  helper also sets crtc_state->dirty_clip to the right coordinates. And of
119  course it needs to check whether the fb is actually active (and maybe where),
120  so there's some book-keeping involved. There's also some good fun involved in
121  scaling things appropriately. For that case we might simply give up and
122  declare the entire area covered by the plane as dirty.
123
124Contact: Noralf Trønnes, Daniel Vetter
125
126Fallout from atomic KMS
127-----------------------
128
129``drm_atomic_helper.c`` provides a batch of functions which implement legacy
130IOCTLs on top of the new atomic driver interface. Which is really nice for
131gradual conversion of drivers, but unfortunately the semantic mismatches are
132a bit too severe. So there's some follow-up work to adjust the function
133interfaces to fix these issues:
134
135* atomic needs the lock acquire context. At the moment that's passed around
136  implicitly with some horrible hacks, and it's also allocate with
137  ``GFP_NOFAIL`` behind the scenes. All legacy paths need to start allocating
138  the acquire context explicitly on stack and then also pass it down into
139  drivers explicitly so that the legacy-on-atomic functions can use them.
140
141* A bunch of the vtable hooks are now in the wrong place: DRM has a split
142  between core vfunc tables (named ``drm_foo_funcs``), which are used to
143  implement the userspace ABI. And then there's the optional hooks for the
144  helper libraries (name ``drm_foo_helper_funcs``), which are purely for
145  internal use. Some of these hooks should be move from ``_funcs`` to
146  ``_helper_funcs`` since they are not part of the core ABI. There's a
147  ``FIXME`` comment in the kerneldoc for each such case in ``drm_crtc.h``.
148
149* There's a new helper ``drm_atomic_helper_best_encoder()`` which could be
150  used by all atomic drivers which don't select the encoder for a given
151  connector at runtime. That's almost all of them, and would allow us to get
152  rid of a lot of ``best_encoder`` boilerplate in drivers.
153
154Contact: Daniel Vetter
155
156Get rid of dev->struct_mutex from GEM drivers
157---------------------------------------------
158
159``dev->struct_mutex`` is the Big DRM Lock from legacy days and infested
160everything. Nowadays in modern drivers the only bit where it's mandatory is
161serializing GEM buffer object destruction. Which unfortunately means drivers
162have to keep track of that lock and either call ``unreference`` or
163``unreference_locked`` depending upon context.
164
165Core GEM doesn't have a need for ``struct_mutex`` any more since kernel 4.8,
166and there's a ``gem_free_object_unlocked`` callback for any drivers which are
167entirely ``struct_mutex`` free.
168
169For drivers that need ``struct_mutex`` it should be replaced with a driver-
170private lock. The tricky part is the BO free functions, since those can't
171reliably take that lock any more. Instead state needs to be protected with
172suitable subordinate locks or some cleanup work pushed to a worker thread. For
173performance-critical drivers it might also be better to go with a more
174fine-grained per-buffer object and per-context lockings scheme. Currently the
175following drivers still use ``struct_mutex``: ``msm``, ``omapdrm`` and
176``udl``.
177
178Contact: Daniel Vetter
179
180Switch to drm_connector_list_iter for any connector_list walking
181----------------------------------------------------------------
182
183Connectors can be hotplugged, and we now have a special list of helpers to walk
184the connector_list in a race-free fashion, without incurring deadlocks on
185mutexes and other fun stuff.
186
187Unfortunately most drivers are not converted yet. At least all those supporting
188DP MST hotplug should be converted, since for those drivers the difference
189matters. See drm_for_each_connector_iter() vs. drm_for_each_connector().
190
191Contact: Daniel Vetter
192
193Core refactorings
194=================
195
196Use new IDR deletion interface to clean up drm_gem_handle_delete()
197------------------------------------------------------------------
198
199See the "This is gross" comment -- apparently the IDR system now can return an
200error code instead of oopsing.
201
202Clean up the DRM header mess
203----------------------------
204
205Currently the DRM subsystem has only one global header, ``drmP.h``. This is
206used both for functions exported to helper libraries and drivers and functions
207only used internally in the ``drm.ko`` module. The goal would be to move all
208header declarations not needed outside of ``drm.ko`` into
209``drivers/gpu/drm/drm_*_internal.h`` header files. ``EXPORT_SYMBOL`` also
210needs to be dropped for these functions.
211
212This would nicely tie in with the below task to create kerneldoc after the API
213is cleaned up. Or with the "hide legacy cruft better" task.
214
215Note that this is well in progress, but ``drmP.h`` is still huge. The updated
216plan is to switch to per-file driver API headers, which will also structure
217the kerneldoc better. This should also allow more fine-grained ``#include``
218directives.
219
220Contact: Daniel Vetter
221
222Add missing kerneldoc for exported functions
223--------------------------------------------
224
225The DRM reference documentation is still lacking kerneldoc in a few areas. The
226task would be to clean up interfaces like moving functions around between
227files to better group them and improving the interfaces like dropping return
228values for functions that never fail. Then write kerneldoc for all exported
229functions and an overview section and integrate it all into the drm DocBook.
230
231See https://dri.freedesktop.org/docs/drm/ for what's there already.
232
233Contact: Daniel Vetter
234
235Hide legacy cruft better
236------------------------
237
238Way back DRM supported only drivers which shadow-attached to PCI devices with
239userspace or fbdev drivers setting up outputs. Modern DRM drivers take charge
240of the entire device, you can spot them with the DRIVER_MODESET flag.
241
242Unfortunately there's still large piles of legacy code around which needs to
243be hidden so that driver writers don't accidentally end up using it. And to
244prevent security issues in those legacy IOCTLs from being exploited on modern
245drivers. This has multiple possible subtasks:
246
247* Make sure legacy IOCTLs can't be used on modern drivers.
248* Extract support code for legacy features into a ``drm-legacy.ko`` kernel
249  module and compile it only when one of the legacy drivers is enabled.
250* Extract legacy functions into their own headers and remove it that from the
251  monolithic ``drmP.h`` header.
252* Remove any lingering cruft from the OS abstraction layer from modern
253  drivers.
254
255This is mostly done, the only thing left is to split up ``drm_irq.c`` into
256legacy cruft and the parts needed by modern KMS drivers.
257
258Contact: Daniel Vetter
259
260Make panic handling work
261------------------------
262
263This is a really varied tasks with lots of little bits and pieces:
264
265* The panic path can't be tested currently, leading to constant breaking. The
266  main issue here is that panics can be triggered from hardirq contexts and
267  hence all panic related callback can run in hardirq context. It would be
268  awesome if we could test at least the fbdev helper code and driver code by
269  e.g. trigger calls through drm debugfs files. hardirq context could be
270  achieved by using an IPI to the local processor.
271
272* There's a massive confusion of different panic handlers. DRM fbdev emulation
273  helpers have one, but on top of that the fbcon code itself also has one. We
274  need to make sure that they stop fighting over each another.
275
276* ``drm_can_sleep()`` is a mess. It hides real bugs in normal operations and
277  isn't a full solution for panic paths. We need to make sure that it only
278  returns true if there's a panic going on for real, and fix up all the
279  fallout.
280
281* The panic handler must never sleep, which also means it can't ever
282  ``mutex_lock()``. Also it can't grab any other lock unconditionally, not
283  even spinlocks (because NMI and hardirq can panic too). We need to either
284  make sure to not call such paths, or trylock everything. Really tricky.
285
286* For the above locking troubles reasons it's pretty much impossible to
287  attempt a synchronous modeset from panic handlers. The only thing we could
288  try to achive is an atomic ``set_base`` of the primary plane, and hope that
289  it shows up. Everything else probably needs to be delayed to some worker or
290  something else which happens later on. Otherwise it just kills the box
291  harder, prevent the panic from going out on e.g. netconsole.
292
293* There's also proposal for a simplied DRM console instead of the full-blown
294  fbcon and DRM fbdev emulation. Any kind of panic handling tricks should
295  obviously work for both console, in case we ever get kmslog merged.
296
297Contact: Daniel Vetter
298
299Clean up the debugfs support
300----------------------------
301
302There's a bunch of issues with it:
303
304- The drm_info_list ->show() function doesn't even bother to cast to the drm
305  structure for you. This is lazy.
306
307- We probably want to have some support for debugfs files on crtc/connectors and
308  maybe other kms objects directly in core. There's even drm_print support in
309  the funcs for these objects to dump kms state, so it's all there. And then the
310  ->show() functions should obviously give you a pointer to the right object.
311
312- The drm_info_list stuff is centered on drm_minor instead of drm_device. For
313  anything we want to print drm_device (or maybe drm_file) is the right thing.
314
315- The drm_driver->debugfs_init hooks we have is just an artifact of the old
316  midlayered load sequence. DRM debugfs should work more like sysfs, where you
317  can create properties/files for an object anytime you want, and the core
318  takes care of publishing/unpuplishing all the files at register/unregister
319  time. Drivers shouldn't need to worry about these technicalities, and fixing
320  this (together with the drm_minor->drm_device move) would allow us to remove
321  debugfs_init.
322
323Contact: Daniel Vetter
324
325Better Testing
326==============
327
328Enable trinity for DRM
329----------------------
330
331And fix up the fallout. Should be really interesting ...
332
333Make KMS tests in i-g-t generic
334-------------------------------
335
336The i915 driver team maintains an extensive testsuite for the i915 DRM driver,
337including tons of testcases for corner-cases in the modesetting API. It would
338be awesome if those tests (at least the ones not relying on Intel-specific GEM
339features) could be made to run on any KMS driver.
340
341Basic work to run i-g-t tests on non-i915 is done, what's now missing is mass-
342converting things over. For modeset tests we also first need a bit of
343infrastructure to use dumb buffers for untiled buffers, to be able to run all
344the non-i915 specific modeset tests.
345
346Contact: Daniel Vetter
347
348Create a virtual KMS driver for testing (vkms)
349----------------------------------------------
350
351With all the latest helpers it should be fairly simple to create a virtual KMS
352driver useful for testing, or for running X or similar on headless machines
353(to be able to still use the GPU). This would be similar to vgem, but aimed at
354the modeset side.
355
356Once the basics are there there's tons of possibilities to extend it.
357
358Contact: Daniel Vetter
359
360Driver Specific
361===============
362
363tinydrm
364-------
365
366Tinydrm is the helper driver for really simple fb drivers. The goal is to make
367those drivers as simple as possible, so lots of room for refactoring:
368
369- backlight helpers, probably best to put them into a new drm_backlight.c.
370  This is because drivers/video is de-facto unmaintained. We could also
371  move drivers/video/backlight to drivers/gpu/backlight and take it all
372  over within drm-misc, but that's more work.
373
374- spi helpers, probably best put into spi core/helper code. Thierry said
375  the spi maintainer is fast&reactive, so shouldn't be a big issue.
376
377- extract the mipi-dbi helper (well, the non-tinydrm specific parts at
378  least) into a separate helper, like we have for mipi-dsi already. Or follow
379  one of the ideas for having a shared dsi/dbi helper, abstracting away the
380  transport details more.
381
382- tinydrm_lastclose could be drm_fb_helper_lastclose. Only thing we need
383  for that is to store the drm_fb_helper pointer somewhere in
384  drm_device->mode_config. And then we could roll that out to all the
385  drivers.
386
387- tinydrm_gem_cma_prime_import_sg_table should probably go into the cma
388  helpers, as a _vmapped variant (since not every driver needs the vmap).
389  And tinydrm_gem_cma_free_object could the be merged into
390  drm_gem_cma_free_object().
391
392- tinydrm_fb_create we could move into drm_simple_pipe, only need to add
393  the fb_create hook to drm_simple_pipe_funcs, which would again simplify a
394  bunch of things (since it gives you a one-stop vfunc for simple drivers).
395
396- Quick aside: The unregister devm stuff is kinda getting the lifetimes of
397  a drm_device wrong. Doesn't matter, since everyone else gets it wrong
398  too :-)
399
400- With the fbdev pointer in dev->mode_config we could also make
401  suspend/resume helpers entirely generic, at least if we add a
402  dev->mode_config.suspend_state. We could even provide a generic pm_ops
403  structure with those.
404
405- also rework the drm_framebuffer_funcs->dirty hook wire-up, see above.
406
407Contact: Noralf Trønnes, Daniel Vetter
408
409Outside DRM
410===========
411
412Better kerneldoc
413----------------
414
415This is pretty much done, but there's some advanced topics:
416
417Come up with a way to hyperlink to struct members. Currently you can hyperlink
418to the struct using ``#struct_name``, but not to a member within. Would need
419buy-in from kerneldoc maintainers, and the big question is how to make it work
420without totally unsightly
421``drm_foo_bar_really_long_structure->even_longer_memeber`` all over the text
422which breaks text flow.
423
424Figure out how to integrate the asciidoc support for ascii-diagrams. We have a
425few of those (e.g. to describe mode timings), and asciidoc supports converting
426some ascii-art dialect into pngs. Would be really pretty to make that work.
427
428Contact: Daniel Vetter, Jani Nikula
429
430Jani is working on this already, hopefully lands in 4.8.
431