xref: /linux/Documentation/gpu/todo.rst (revision 8a79db5e83a5d52c74e6f3c40d6f312cf899213e)
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
10Difficulty
11----------
12
13To make it easier task are categorized into different levels:
14
15Starter: Good tasks to get started with the DRM subsystem.
16
17Intermediate: Tasks which need some experience with working in the DRM
18subsystem, or some specific GPU/display graphics knowledge. For debugging issue
19it's good to have the relevant hardware (or a virtual driver set up) available
20for testing.
21
22Advanced: Tricky tasks that need fairly good understanding of the DRM subsystem
23and graphics topics. Generally need the relevant hardware for development and
24testing.
25
26Subsystem-wide refactorings
27===========================
28
29Remove custom dumb_map_offset implementations
30---------------------------------------------
31
32All GEM based drivers should be using drm_gem_create_mmap_offset() instead.
33Audit each individual driver, make sure it'll work with the generic
34implementation (there's lots of outdated locking leftovers in various
35implementations), and then remove it.
36
37Contact: Daniel Vetter, respective driver maintainers
38
39Level: Intermediate
40
41Convert existing KMS drivers to atomic modesetting
42--------------------------------------------------
43
443.19 has the atomic modeset interfaces and helpers, so drivers can now be
45converted over. Modern compositors like Wayland or Surfaceflinger on Android
46really want an atomic modeset interface, so this is all about the bright
47future.
48
49There is a conversion guide for atomic and all you need is a GPU for a
50non-converted driver (again virtual HW drivers for KVM are still all
51suitable).
52
53As part of this drivers also need to convert to universal plane (which means
54exposing primary & cursor as proper plane objects). But that's much easier to
55do by directly using the new atomic helper driver callbacks.
56
57Contact: Daniel Vetter, respective driver maintainers
58
59Level: Advanced
60
61Clean up the clipped coordination confusion around planes
62---------------------------------------------------------
63
64We have a helper to get this right with drm_plane_helper_check_update(), but
65it's not consistently used. This should be fixed, preferrably in the atomic
66helpers (and drivers then moved over to clipped coordinates). Probably the
67helper should also be moved from drm_plane_helper.c to the atomic helpers, to
68avoid confusion - the other helpers in that file are all deprecated legacy
69helpers.
70
71Contact: Ville Syrjälä, Daniel Vetter, driver maintainers
72
73Level: Advanced
74
75Convert early atomic drivers to async commit helpers
76----------------------------------------------------
77
78For the first year the atomic modeset helpers didn't support asynchronous /
79nonblocking commits, and every driver had to hand-roll them. This is fixed
80now, but there's still a pile of existing drivers that easily could be
81converted over to the new infrastructure.
82
83One issue with the helpers is that they require that drivers handle completion
84events for atomic commits correctly. But fixing these bugs is good anyway.
85
86Contact: Daniel Vetter, respective driver maintainers
87
88Level: Advanced
89
90Fallout from atomic KMS
91-----------------------
92
93``drm_atomic_helper.c`` provides a batch of functions which implement legacy
94IOCTLs on top of the new atomic driver interface. Which is really nice for
95gradual conversion of drivers, but unfortunately the semantic mismatches are
96a bit too severe. So there's some follow-up work to adjust the function
97interfaces to fix these issues:
98
99* atomic needs the lock acquire context. At the moment that's passed around
100  implicitly with some horrible hacks, and it's also allocate with
101  ``GFP_NOFAIL`` behind the scenes. All legacy paths need to start allocating
102  the acquire context explicitly on stack and then also pass it down into
103  drivers explicitly so that the legacy-on-atomic functions can use them.
104
105  Except for some driver code this is done. This task should be finished by
106  adding WARN_ON(!drm_drv_uses_atomic_modeset) in drm_modeset_lock_all().
107
108* A bunch of the vtable hooks are now in the wrong place: DRM has a split
109  between core vfunc tables (named ``drm_foo_funcs``), which are used to
110  implement the userspace ABI. And then there's the optional hooks for the
111  helper libraries (name ``drm_foo_helper_funcs``), which are purely for
112  internal use. Some of these hooks should be move from ``_funcs`` to
113  ``_helper_funcs`` since they are not part of the core ABI. There's a
114  ``FIXME`` comment in the kerneldoc for each such case in ``drm_crtc.h``.
115
116Contact: Daniel Vetter
117
118Level: Intermediate
119
120Get rid of dev->struct_mutex from GEM drivers
121---------------------------------------------
122
123``dev->struct_mutex`` is the Big DRM Lock from legacy days and infested
124everything. Nowadays in modern drivers the only bit where it's mandatory is
125serializing GEM buffer object destruction. Which unfortunately means drivers
126have to keep track of that lock and either call ``unreference`` or
127``unreference_locked`` depending upon context.
128
129Core GEM doesn't have a need for ``struct_mutex`` any more since kernel 4.8,
130and there's a ``gem_free_object_unlocked`` callback for any drivers which are
131entirely ``struct_mutex`` free.
132
133For drivers that need ``struct_mutex`` it should be replaced with a driver-
134private lock. The tricky part is the BO free functions, since those can't
135reliably take that lock any more. Instead state needs to be protected with
136suitable subordinate locks or some cleanup work pushed to a worker thread. For
137performance-critical drivers it might also be better to go with a more
138fine-grained per-buffer object and per-context lockings scheme. Currently only the
139``msm`` driver still use ``struct_mutex``.
140
141Contact: Daniel Vetter, respective driver maintainers
142
143Level: Advanced
144
145Convert instances of dev_info/dev_err/dev_warn to their DRM_DEV_* equivalent
146----------------------------------------------------------------------------
147
148For drivers which could have multiple instances, it is necessary to
149differentiate between which is which in the logs. Since DRM_INFO/WARN/ERROR
150don't do this, drivers used dev_info/warn/err to make this differentiation. We
151now have DRM_DEV_* variants of the drm print macros, so we can start to convert
152those drivers back to using drm-formwatted specific log messages.
153
154Before you start this conversion please contact the relevant maintainers to make
155sure your work will be merged - not everyone agrees that the DRM dmesg macros
156are better.
157
158Contact: Sean Paul, Maintainer of the driver you plan to convert
159
160Level: Starter
161
162Convert drivers to use simple modeset suspend/resume
163----------------------------------------------------
164
165Most drivers (except i915 and nouveau) that use
166drm_atomic_helper_suspend/resume() can probably be converted to use
167drm_mode_config_helper_suspend/resume(). Also there's still open-coded version
168of the atomic suspend/resume code in older atomic modeset drivers.
169
170Contact: Maintainer of the driver you plan to convert
171
172Level: Intermediate
173
174Convert drivers to use drm_fb_helper_fbdev_setup/teardown()
175-----------------------------------------------------------
176
177Most drivers can use drm_fb_helper_fbdev_setup() except maybe:
178
179- amdgpu which has special logic to decide whether to call
180  drm_helper_disable_unused_functions()
181
182- armada which isn't atomic and doesn't call
183  drm_helper_disable_unused_functions()
184
185- i915 which calls drm_fb_helper_initial_config() in a worker
186
187Drivers that use drm_framebuffer_remove() to clean up the fbdev framebuffer can
188probably use drm_fb_helper_fbdev_teardown().
189
190Contact: Maintainer of the driver you plan to convert
191
192Level: Intermediate
193
194Clean up mmap forwarding
195------------------------
196
197A lot of drivers forward gem mmap calls to dma-buf mmap for imported buffers.
198And also a lot of them forward dma-buf mmap to the gem mmap implementations.
199There's drm_gem_prime_mmap() for this now, but still needs to be rolled out.
200
201Contact: Daniel Vetter
202
203Level: Intermediate
204
205Generic fbdev defio support
206---------------------------
207
208The defio support code in the fbdev core has some very specific requirements,
209which means drivers need to have a special framebuffer for fbdev. The main
210issue is that it uses some fields in struct page itself, which breaks shmem
211gem objects (and other things). To support defio, affected drivers require
212the use of a shadow buffer, which may add CPU and memory overhead.
213
214Possible solution would be to write our own defio mmap code in the drm fbdev
215emulation. It would need to fully wrap the existing mmap ops, forwarding
216everything after it has done the write-protect/mkwrite trickery:
217
218- In the drm_fbdev_fb_mmap helper, if we need defio, change the
219  default page prots to write-protected with something like this::
220
221      vma->vm_page_prot = pgprot_wrprotect(vma->vm_page_prot);
222
223- Set the mkwrite and fsync callbacks with similar implementions to the core
224  fbdev defio stuff. These should all work on plain ptes, they don't actually
225  require a struct page.  uff. These should all work on plain ptes, they don't
226  actually require a struct page.
227
228- Track the dirty pages in a separate structure (bitfield with one bit per page
229  should work) to avoid clobbering struct page.
230
231Might be good to also have some igt testcases for this.
232
233Contact: Daniel Vetter, Noralf Tronnes
234
235Level: Advanced
236
237idr_init_base()
238---------------
239
240DRM core&drivers uses a lot of idr (integer lookup directories) for mapping
241userspace IDs to internal objects, and in most places ID=0 means NULL and hence
242is never used. Switching to idr_init_base() for these would make the idr more
243efficient.
244
245Contact: Daniel Vetter
246
247Level: Starter
248
249struct drm_gem_object_funcs
250---------------------------
251
252GEM objects can now have a function table instead of having the callbacks on the
253DRM driver struct. This is now the preferred way and drivers can be moved over.
254
255We also need a 2nd version of the CMA define that doesn't require the
256vmapping to be present (different hook for prime importing). Plus this needs to
257be rolled out to all drivers using their own implementations, too.
258
259Level: Intermediate
260
261Use DRM_MODESET_LOCK_ALL_* helpers instead of boilerplate
262---------------------------------------------------------
263
264For cases where drivers are attempting to grab the modeset locks with a local
265acquire context. Replace the boilerplate code surrounding
266drm_modeset_lock_all_ctx() with DRM_MODESET_LOCK_ALL_BEGIN() and
267DRM_MODESET_LOCK_ALL_END() instead.
268
269This should also be done for all places where drm_modest_lock_all() is still
270used.
271
272As a reference, take a look at the conversions already completed in drm core.
273
274Contact: Sean Paul, respective driver maintainers
275
276Level: Starter
277
278Rename CMA helpers to DMA helpers
279---------------------------------
280
281CMA (standing for contiguous memory allocator) is really a bit an accident of
282what these were used for first, a much better name would be DMA helpers. In the
283text these should even be called coherent DMA memory helpers (so maybe CDM, but
284no one knows what that means) since underneath they just use dma_alloc_coherent.
285
286Contact: Laurent Pinchart, Daniel Vetter
287
288Level: Intermediate (mostly because it is a huge tasks without good partial
289milestones, not technically itself that challenging)
290
291Convert direct mode.vrefresh accesses to use drm_mode_vrefresh()
292----------------------------------------------------------------
293
294drm_display_mode.vrefresh isn't guaranteed to be populated. As such, using it
295is risky and has been known to cause div-by-zero bugs. Fortunately, drm core
296has helper which will use mode.vrefresh if it's !0 and will calculate it from
297the timings when it's 0.
298
299Use simple search/replace, or (more fun) cocci to replace instances of direct
300vrefresh access with a call to the helper. Check out
301https://lists.freedesktop.org/archives/dri-devel/2019-January/205186.html for
302inspiration.
303
304Once all instances of vrefresh have been converted, remove vrefresh from
305drm_display_mode to avoid future use.
306
307Contact: Sean Paul
308
309Level: Starter
310
311Remove drm_display_mode.hsync
312-----------------------------
313
314We have drm_mode_hsync() to calculate this from hsync_start/end, since drivers
315shouldn't/don't use this, remove this member to avoid any temptations to use it
316in the future. If there is any debug code using drm_display_mode.hsync, convert
317it to use drm_mode_hsync() instead.
318
319Contact: Sean Paul
320
321Level: Starter
322
323drm_fb_helper tasks
324-------------------
325
326- drm_fb_helper_restore_fbdev_mode_unlocked() should call restore_fbdev_mode()
327  not the _force variant so it can bail out if there is a master. But first
328  these igt tests need to be fixed: kms_fbcon_fbt@psr and
329  kms_fbcon_fbt@psr-suspend.
330
331- The max connector argument for drm_fb_helper_init() and
332  drm_fb_helper_fbdev_setup() isn't used anymore and can be removed.
333
334- The helper doesn't keep an array of connectors anymore so these can be
335  removed: drm_fb_helper_single_add_all_connectors(),
336  drm_fb_helper_add_one_connector() and drm_fb_helper_remove_one_connector().
337
338Level: Intermediate
339
340connector register/unregister fixes
341-----------------------------------
342
343- For most connectors it's a no-op to call drm_connector_register/unregister
344  directly from driver code, drm_dev_register/unregister take care of this
345  already. We can remove all of them.
346
347- For dp drivers it's a bit more a mess, since we need the connector to be
348  registered when calling drm_dp_aux_register. Fix this by instead calling
349  drm_dp_aux_init, and moving the actual registering into a late_register
350  callback as recommended in the kerneldoc.
351
352Level: Intermediate
353
354Core refactorings
355=================
356
357Make panic handling work
358------------------------
359
360This is a really varied tasks with lots of little bits and pieces:
361
362* The panic path can't be tested currently, leading to constant breaking. The
363  main issue here is that panics can be triggered from hardirq contexts and
364  hence all panic related callback can run in hardirq context. It would be
365  awesome if we could test at least the fbdev helper code and driver code by
366  e.g. trigger calls through drm debugfs files. hardirq context could be
367  achieved by using an IPI to the local processor.
368
369* There's a massive confusion of different panic handlers. DRM fbdev emulation
370  helpers have one, but on top of that the fbcon code itself also has one. We
371  need to make sure that they stop fighting over each another.
372
373* ``drm_can_sleep()`` is a mess. It hides real bugs in normal operations and
374  isn't a full solution for panic paths. We need to make sure that it only
375  returns true if there's a panic going on for real, and fix up all the
376  fallout.
377
378* The panic handler must never sleep, which also means it can't ever
379  ``mutex_lock()``. Also it can't grab any other lock unconditionally, not
380  even spinlocks (because NMI and hardirq can panic too). We need to either
381  make sure to not call such paths, or trylock everything. Really tricky.
382
383* For the above locking troubles reasons it's pretty much impossible to
384  attempt a synchronous modeset from panic handlers. The only thing we could
385  try to achive is an atomic ``set_base`` of the primary plane, and hope that
386  it shows up. Everything else probably needs to be delayed to some worker or
387  something else which happens later on. Otherwise it just kills the box
388  harder, prevent the panic from going out on e.g. netconsole.
389
390* There's also proposal for a simplied DRM console instead of the full-blown
391  fbcon and DRM fbdev emulation. Any kind of panic handling tricks should
392  obviously work for both console, in case we ever get kmslog merged.
393
394Contact: Daniel Vetter
395
396Level: Advanced
397
398Clean up the debugfs support
399----------------------------
400
401There's a bunch of issues with it:
402
403- The drm_info_list ->show() function doesn't even bother to cast to the drm
404  structure for you. This is lazy.
405
406- We probably want to have some support for debugfs files on crtc/connectors and
407  maybe other kms objects directly in core. There's even drm_print support in
408  the funcs for these objects to dump kms state, so it's all there. And then the
409  ->show() functions should obviously give you a pointer to the right object.
410
411- The drm_info_list stuff is centered on drm_minor instead of drm_device. For
412  anything we want to print drm_device (or maybe drm_file) is the right thing.
413
414- The drm_driver->debugfs_init hooks we have is just an artifact of the old
415  midlayered load sequence. DRM debugfs should work more like sysfs, where you
416  can create properties/files for an object anytime you want, and the core
417  takes care of publishing/unpuplishing all the files at register/unregister
418  time. Drivers shouldn't need to worry about these technicalities, and fixing
419  this (together with the drm_minor->drm_device move) would allow us to remove
420  debugfs_init.
421
422- Drop the return code and error checking from all debugfs functions. Greg KH is
423  working on this already.
424
425Contact: Daniel Vetter
426
427Level: Intermediate
428
429KMS cleanups
430------------
431
432Some of these date from the very introduction of KMS in 2008 ...
433
434- Make ->funcs and ->helper_private vtables optional. There's a bunch of empty
435  function tables in drivers, but before we can remove them we need to make sure
436  that all the users in helpers and drivers do correctly check for a NULL
437  vtable.
438
439- Cleanup up the various ->destroy callbacks. A lot of them just wrapt the
440  drm_*_cleanup implementations and can be removed. Some tack a kfree() at the
441  end, for which we could add drm_*_cleanup_kfree(). And then there's the (for
442  historical reasons) misnamed drm_primary_helper_destroy() function.
443
444Level: Intermediate
445
446Better Testing
447==============
448
449Enable trinity for DRM
450----------------------
451
452And fix up the fallout. Should be really interesting ...
453
454Level: Advanced
455
456Make KMS tests in i-g-t generic
457-------------------------------
458
459The i915 driver team maintains an extensive testsuite for the i915 DRM driver,
460including tons of testcases for corner-cases in the modesetting API. It would
461be awesome if those tests (at least the ones not relying on Intel-specific GEM
462features) could be made to run on any KMS driver.
463
464Basic work to run i-g-t tests on non-i915 is done, what's now missing is mass-
465converting things over. For modeset tests we also first need a bit of
466infrastructure to use dumb buffers for untiled buffers, to be able to run all
467the non-i915 specific modeset tests.
468
469Level: Advanced
470
471Extend virtual test driver (VKMS)
472---------------------------------
473
474See the documentation of :ref:`VKMS <vkms>` for more details. This is an ideal
475internship task, since it only requires a virtual machine and can be sized to
476fit the available time.
477
478Contact: Daniel Vetter
479
480Level: See details
481
482Backlight Refactoring
483---------------------
484
485Backlight drivers have a triple enable/disable state, which is a bit overkill.
486Plan to fix this:
487
4881. Roll out backlight_enable() and backlight_disable() helpers everywhere. This
489   has started already.
4902. In all, only look at one of the three status bits set by the above helpers.
4913. Remove the other two status bits.
492
493Contact: Daniel Vetter
494
495Level: Intermediate
496
497Driver Specific
498===============
499
500AMD DC Display Driver
501---------------------
502
503AMD DC is the display driver for AMD devices starting with Vega. There has been
504a bunch of progress cleaning it up but there's still plenty of work to be done.
505
506See drivers/gpu/drm/amd/display/TODO for tasks.
507
508Contact: Harry Wentland, Alex Deucher
509
510Bootsplash
511==========
512
513There is support in place now for writing internal DRM clients making it
514possible to pick up the bootsplash work that was rejected because it was written
515for fbdev.
516
517- [v6,8/8] drm/client: Hack: Add bootsplash example
518  https://patchwork.freedesktop.org/patch/306579/
519
520- [RFC PATCH v2 00/13] Kernel based bootsplash
521  https://lkml.org/lkml/2017/12/13/764
522
523Contact: Sam Ravnborg
524
525Level: Advanced
526
527Outside DRM
528===========
529
530Convert fbdev drivers to DRM
531----------------------------
532
533There are plenty of fbdev drivers for older hardware. Some hwardware has
534become obsolete, but some still provides good(-enough) framebuffers. The
535drivers that are still useful should be converted to DRM and afterwards
536removed from fbdev.
537
538Very simple fbdev drivers can best be converted by starting with a new
539DRM driver. Simple KMS helpers and SHMEM should be able to handle any
540existing hardware. The new driver's call-back functions are filled from
541existing fbdev code.
542
543More complex fbdev drivers can be refactored step-by-step into a DRM
544driver with the help of the DRM fbconv helpers. [1] These helpers provide
545the transition layer between the DRM core infrastructure and the fbdev
546driver interface. Create a new DRM driver on top of the fbconv helpers,
547copy over the fbdev driver, and hook it up to the DRM code. Examples for
548several fbdev drivers are available at [1] and a tutorial of this process
549available at [2]. The result is a primitive DRM driver that can run X11
550and Weston.
551
552 - [1] https://gitlab.freedesktop.org/tzimmermann/linux/tree/fbconv
553 - [2] https://gitlab.freedesktop.org/tzimmermann/linux/blob/fbconv/drivers/gpu/drm/drm_fbconv_helper.c
554
555Contact: Thomas Zimmermann <tzimmermann@suse.de>
556
557Level: Advanced
558