diff options
Diffstat (limited to 'Documentation/gpu/todo.rst')
| -rw-r--r-- | Documentation/gpu/todo.rst | 298 |
1 files changed, 298 insertions, 0 deletions
diff --git a/Documentation/gpu/todo.rst b/Documentation/gpu/todo.rst new file mode 100644 index 000000000000..3813a2525ada --- /dev/null +++ b/Documentation/gpu/todo.rst @@ -0,0 +1,298 @@ +.. _todo: + +========= +TODO list +========= + +This section contains a list of smaller janitorial tasks in the kernel DRM +graphics subsystem useful as newbie projects. Or for slow rainy days. + +Subsystem-wide refactorings +=========================== + +De-midlayer drivers +------------------- + +With the recent ``drm_bus`` cleanup patches for 3.17 it is no longer required +to have a ``drm_bus`` structure set up. Drivers can directly set up the +``drm_device`` structure instead of relying on bus methods in ``drm_usb.c`` +and ``drm_platform.c``. The goal is to get rid of the driver's ``->load`` / +``->unload`` callbacks and open-code the load/unload sequence properly, using +the new two-stage ``drm_device`` setup/teardown. + +Once all existing drivers are converted we can also remove those bus support +files for USB and platform devices. + +All you need is a GPU for a non-converted driver (currently almost all of +them, but also all the virtual ones used by KVM, so everyone qualifies). + +Contact: Daniel Vetter, Thierry Reding, respective driver maintainers + +Switch from reference/unreference to get/put +-------------------------------------------- + +For some reason DRM core uses ``reference``/``unreference`` suffixes for +refcounting functions, but kernel uses ``get``/``put`` (e.g. +``kref_get``/``put()``). It would be good to switch over for consistency, and +it's shorter. Needs to be done in 3 steps for each pair of functions: + +* Create new ``get``/``put`` functions, define the old names as compatibility + wrappers +* Switch over each file/driver using a cocci-generated spatch. +* Once all users of the old names are gone, remove them. + +This way drivers/patches in the progress of getting merged won't break. + +Contact: Daniel Vetter + +Convert existing KMS drivers to atomic modesetting +-------------------------------------------------- + +3.19 has the atomic modeset interfaces and helpers, so drivers can now be +converted over. Modern compositors like Wayland or Surfaceflinger on Android +really want an atomic modeset interface, so this is all about the bright +future. + +There is a conversion guide for atomic and all you need is a GPU for a +non-converted driver (again virtual HW drivers for KVM are still all +suitable). + +As part of this drivers also need to convert to universal plane (which means +exposing primary & cursor as proper plane objects). But that's much easier to +do by directly using the new atomic helper driver callbacks. + +Contact: Daniel Vetter, respective driver maintainers + +Convert early atomic drivers to async commit helpers +---------------------------------------------------- + +For the first year the atomic modeset helpers didn't support asynchronous / +nonblocking commits, and every driver had to hand-roll them. This is fixed +now, but there's still a pile of existing drivers that easily could be +converted over to the new infrastructure. + +One issue with the helpers is that they require that drivers handle completion +events for atomic commits correctly. But fixing these bugs is good anyway. + +Contact: Daniel Vetter, respective driver maintainers + +Fallout from atomic KMS +----------------------- + +``drm_atomic_helper.c`` provides a batch of functions which implement legacy +IOCTLs on top of the new atomic driver interface. Which is really nice for +gradual conversion of drivers, but unfortunately the semantic mismatches are +a bit too severe. So there's some follow-up work to adjust the function +interfaces to fix these issues: + +* atomic needs the lock acquire context. At the moment that's passed around + implicitly with some horrible hacks, and it's also allocate with + ``GFP_NOFAIL`` behind the scenes. All legacy paths need to start allocating + the acquire context explicitly on stack and then also pass it down into + drivers explicitly so that the legacy-on-atomic functions can use them. + +* A bunch of the vtable hooks are now in the wrong place: DRM has a split + between core vfunc tables (named ``drm_foo_funcs``), which are used to + implement the userspace ABI. And then there's the optional hooks for the + helper libraries (name ``drm_foo_helper_funcs``), which are purely for + internal use. Some of these hooks should be move from ``_funcs`` to + ``_helper_funcs`` since they are not part of the core ABI. There's a + ``FIXME`` comment in the kerneldoc for each such case in ``drm_crtc.h``. + +* There's a new helper ``drm_atomic_helper_best_encoder()`` which could be + used by all atomic drivers which don't select the encoder for a given + connector at runtime. That's almost all of them, and would allow us to get + rid of a lot of ``best_encoder`` boilerplate in drivers. + +Contact: Daniel Vetter + +Get rid of dev->struct_mutex from GEM drivers +--------------------------------------------- + +``dev->struct_mutex`` is the Big DRM Lock from legacy days and infested +everything. Nowadays in modern drivers the only bit where it's mandatory is +serializing GEM buffer object destruction. Which unfortunately means drivers +have to keep track of that lock and either call ``unreference`` or +``unreference_locked`` depending upon context. + +Core GEM doesn't have a need for ``struct_mutex`` any more since kernel 4.8, +and there's a ``gem_free_object_unlocked`` callback for any drivers which are +entirely ``struct_mutex`` free. + +For drivers that need ``struct_mutex`` it should be replaced with a driver- +private lock. The tricky part is the BO free functions, since those can't +reliably take that lock any more. Instead state needs to be protected with +suitable subordinate locks or some cleanup work pushed to a worker thread. For +performance-critical drivers it might also be better to go with a more +fine-grained per-buffer object and per-context lockings scheme. Currently the +following drivers still use ``struct_mutex``: ``msm``, ``omapdrm`` and +``udl``. + +Contact: Daniel Vetter + +Core refactorings +================= + +Use new IDR deletion interface to clean up drm_gem_handle_delete() +------------------------------------------------------------------ + +See the "This is gross" comment -- apparently the IDR system now can return an +error code instead of oopsing. + +Clean up the DRM header mess +---------------------------- + +Currently the DRM subsystem has only one global header, ``drmP.h``. This is +used both for functions exported to helper libraries and drivers and functions +only used internally in the ``drm.ko`` module. The goal would be to move all +header declarations not needed outside of ``drm.ko`` into +``drivers/gpu/drm/drm_*_internal.h`` header files. ``EXPORT_SYMBOL`` also +needs to be dropped for these functions. + +This would nicely tie in with the below task to create kerneldoc after the API +is cleaned up. Or with the "hide legacy cruft better" task. + +Note that this is well in progress, but ``drmP.h`` is still huge. The updated +plan is to switch to per-file driver API headers, which will also structure +the kerneldoc better. This should also allow more fine-grained ``#include`` +directives. + +Contact: Daniel Vetter + +Add missing kerneldoc for exported functions +-------------------------------------------- + +The DRM reference documentation is still lacking kerneldoc in a few areas. The +task would be to clean up interfaces like moving functions around between +files to better group them and improving the interfaces like dropping return +values for functions that never fail. Then write kerneldoc for all exported +functions and an overview section and integrate it all into the drm DocBook. + +See https://dri.freedesktop.org/docs/drm/ for what's there already. + +Contact: Daniel Vetter + +Hide legacy cruft better +------------------------ + +Way back DRM supported only drivers which shadow-attached to PCI devices with +userspace or fbdev drivers setting up outputs. Modern DRM drivers take charge +of the entire device, you can spot them with the DRIVER_MODESET flag. + +Unfortunately there's still large piles of legacy code around which needs to +be hidden so that driver writers don't accidentally end up using it. And to +prevent security issues in those legacy IOCTLs from being exploited on modern +drivers. This has multiple possible subtasks: + +* Make sure legacy IOCTLs can't be used on modern drivers. +* Extract support code for legacy features into a ``drm-legacy.ko`` kernel + module and compile it only when one of the legacy drivers is enabled. +* Extract legacy functions into their own headers and remove it that from the + monolithic ``drmP.h`` header. +* Remove any lingering cruft from the OS abstraction layer from modern + drivers. + +This is mostly done, the only thing left is to split up ``drm_irq.c`` into +legacy cruft and the parts needed by modern KMS drivers. + +Contact: Daniel Vetter + +Make panic handling work +------------------------ + +This is a really varied tasks with lots of little bits and pieces: + +* The panic path can't be tested currently, leading to constant breaking. The + main issue here is that panics can be triggered from hardirq contexts and + hence all panic related callback can run in hardirq context. It would be + awesome if we could test at least the fbdev helper code and driver code by + e.g. trigger calls through drm debugfs files. hardirq context could be + achieved by using an IPI to the local processor. + +* There's a massive confusion of different panic handlers. DRM fbdev emulation + helpers have one, but on top of that the fbcon code itself also has one. We + need to make sure that they stop fighting over each another. + +* ``drm_can_sleep()`` is a mess. It hides real bugs in normal operations and + isn't a full solution for panic paths. We need to make sure that it only + returns true if there's a panic going on for real, and fix up all the + fallout. + +* The panic handler must never sleep, which also means it can't ever + ``mutex_lock()``. Also it can't grab any other lock unconditionally, not + even spinlocks (because NMI and hardirq can panic too). We need to either + make sure to not call such paths, or trylock everything. Really tricky. + +* For the above locking troubles reasons it's pretty much impossible to + attempt a synchronous modeset from panic handlers. The only thing we could + try to achive is an atomic ``set_base`` of the primary plane, and hope that + it shows up. Everything else probably needs to be delayed to some worker or + something else which happens later on. Otherwise it just kills the box + harder, prevent the panic from going out on e.g. netconsole. + +* There's also proposal for a simplied DRM console instead of the full-blown + fbcon and DRM fbdev emulation. Any kind of panic handling tricks should + obviously work for both console, in case we ever get kmslog merged. + +Contact: Daniel Vetter + +Better Testing +============== + +Enable trinity for DRM +---------------------- + +And fix up the fallout. Should be really interesting ... + +Make KMS tests in i-g-t generic +------------------------------- + +The i915 driver team maintains an extensive testsuite for the i915 DRM driver, +including tons of testcases for corner-cases in the modesetting API. It would +be awesome if those tests (at least the ones not relying on Intel-specific GEM +features) could be made to run on any KMS driver. + +Basic work to run i-g-t tests on non-i915 is done, what's now missing is mass- +converting things over. For modeset tests we also first need a bit of +infrastructure to use dumb buffers for untiled buffers, to be able to run all +the non-i915 specific modeset tests. + +Contact: Daniel Vetter + +Create a virtual KMS driver for testing (vkms) +---------------------------------------------- + +With all the latest helpers it should be fairly simple to create a virtual KMS +driver useful for testing, or for running X or similar on headless machines +(to be able to still use the GPU). This would be similar to vgem, but aimed at +the modeset side. + +Once the basics are there there's tons of possibilities to extend it. + +Contact: Daniel Vetter + +Driver Specific +=============== + +Outside DRM +=========== + +Better kerneldoc +---------------- + +This is pretty much done, but there's some advanced topics: + +Come up with a way to hyperlink to struct members. Currently you can hyperlink +to the struct using ``#struct_name``, but not to a member within. Would need +buy-in from kerneldoc maintainers, and the big question is how to make it work +without totally unsightly +``drm_foo_bar_really_long_structure->even_longer_memeber`` all over the text +which breaks text flow. + +Figure out how to integrate the asciidoc support for ascii-diagrams. We have a +few of those (e.g. to describe mode timings), and asciidoc supports converting +some ascii-art dialect into pngs. Would be really pretty to make that work. + +Contact: Daniel Vetter, Jani Nikula + +Jani is working on this already, hopefully lands in 4.8. |