Add the infrastructure to support the idea of multiple ringbuffers.
Assign each ringbuffer an id and use that as an index for the various
ring specific operations.
The biggest delta is to support legacy fences. Each fence gets its own
sequence number but the legacy functions expect to use a unique integer.
To handle this we return a unique identifier for each submission but
map it to a specific ring/sequence under the covers. Newer users use
a dma_fence pointer anyway so they don't care about the actual sequence
ID or ring.
The actual mechanics for multiple ringbuffers are very target specific
so this code just allows for the possibility but still only defines
one ringbuffer for each target family.
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Currently the behavior of a command stream is provided by the user
application during submission and the application is expected to internally
maintain the settings for each 'context' or 'rendering queue' and specify
the correct ones.
This works okay for simple cases but as applications become more
complex we will want to set context specific flags and do various
permission checks to allow certain contexts to enable additional
privileges.
Add kernel-side submit queues to be analogous to 'contexts' or
'rendering queues' on the application side. Each file descriptor
instance will maintain its own list of queues. Queues cannot be
shared between file descriptors.
For backwards compatibility context id '0' is defined as a default
context specifying no priority and no special flags. This is
intended to be the usual configuration for 99% of applications so
that a garden variety application can function correctly without
creating a queue. Only those applications requiring the specific
benefit of different queues need create one.
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
Buffer object specific resources like pages, domains, sg list
need not be protected with struct_mutex. They can be protected
with a buffer object level lock. This simplifies locking and
makes it easier to avoid potential recursive locking scenarios
for SVM involving mmap_sem and struct_mutex. This also removes
unnecessary serialization when creating buffer objects, and also
between buffer object creation and GPU command submission.
Signed-off-by: Sushmita Susheelendra <ssusheel@codeaurora.org>
[robclark: squash in handling new locking for shrinker]
Signed-off-by: Rob Clark <robdclark@gmail.com>
Now that the msm_gem supports an arbitrary number of vma's, we no longer
need to assign an id (index) to each address space. So rip out the
associated code.
Signed-off-by: Rob Clark <robdclark@gmail.com>
It means we have to do a list traversal where we once had an index into
a table. But the list will normally have one or two entries.
Signed-off-by: Rob Clark <robdclark@gmail.com>
No functional change, that will come later. But this will make it
easier to deal with dynamically created address spaces (ie. per-
process pagetables for gpu).
Signed-off-by: Rob Clark <robdclark@gmail.com>
There are reasons for a memory object to outlive the file descriptor
that created it and so the address space that a buffer object is
attached to must also outlive the file descriptor. Reference count
the address space so that it can remain viable until all the objects
have released their addresses.
Signed-off-by: Jordan Crouse <jcrouse@codeaurora.org>
Signed-off-by: Rob Clark <robdclark@gmail.com>
For a5xx the gpu is 64b so we need to change iova to 64b everywhere. On
the display side, iova is still 32b so it can ignore the upper bits.
(Although all the armv8 devices have an iommu that can map 64b pa to 32b
iova.)
Signed-off-by: Rob Clark <robdclark@gmail.com>
We can have various combinations of 64b and 32b address space, ie. 64b
CPU but 32b display and gpu, or 64b CPU and GPU but 32b display. So
best to decouple the device iova's from mmap offset.
Signed-off-by: Rob Clark <robdclark@gmail.com>
For some optimizations coming on the userspace side, splitting larger
draw or gmem cmds into multiple cmdstream buffers, we need to support
much more than the previous small/arbitrary limit.
Signed-off-by: Rob Clark <robdclark@gmail.com>
For a first step, only purge obj->madv==DONTNEED objects. We could be
more agressive and next try unpinning inactive objects.. but that is
only useful if you have swap.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Track the pid per submit, so we can print the name of the task which
submitted the batch that caused the gpu to hang.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Split up locking and pinning buffers in the submit path. This is needed
because we'll want to insert fencing in between the two steps.
This makes things end up looking more similar to etnaviv submit code
(which was originally modelled on the msm code but has already added
'struct fence' support).
Signed-off-by: Rob Clark <robdclark@gmail.com>
Track the list of in-flight submits. If the gpu hangs, retire up to an
including the offending submit, and then re-submit the remainder. This
way, for concurrently running piglit tests (for example), one failing
test doesn't cause unrelated tests to fail simply because it's submit
was queued up after one that triggered a hang.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Add support to use the VRAM carveout (if specified in dtb) for fbdev
scanout buffer. This allows drm/msm to take over a bootloader splash-
screen, and avoids corruption on screen that results if the kernel uses
memory that is still being scanned out for itself.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Give ourselves a way to wait for certain fence #.. makes it easier to
wait on a set of bo's, which we'll need for atomic.
Signed-off-by: Rob Clark <robdclark@gmail.com>
To ease debugging, add debugfs file which can be cat/tail'd to log
submits, along with fence #. If GPU hangs, you can look at 'gpu'
debugfs file to find last completed fence and current register state,
and compare with logged rd file to narrow down the DRAW_INDX which
triggered the GPU hang.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Add a VRAM carveout that is used for systems which do not have an IOMMU.
The VRAM carveout uses CMA. The arch code must setup a CMA pool for the
device (preferrably in highmem.. a 256m-512m VRAM pool in lowmem is not
cool). The user can configure the VRAM pool size using msm.vram module
param.
Technically, the abstraction of IOMMU behind msm_mmu is not strictly
needed, but it simplifies the GEM code a bit, and will be useful later
when I add support for a2xx devices with GPUMMU, so I decided to keep
this part.
It appears to be possible to configure the GPU to restrict access to
addresses within the VRAM pool, but this is not done yet. So for now
the GPU will refuse to load if there is no sort of mmu. Once address
based limits are supported and tested to confirm that we aren't giving
the GPU access to arbitrary memory, this restriction can be lifted
Signed-off-by: Rob Clark <robdclark@gmail.com>
Re-arrange things a bit so that we can get work requested after a bo
fence passes, like pageflip, done before retiring bo's. Without any
sort of bo cache in userspace, some games can trigger hundred's of
transient bo's, which can cause retire to take a long time (5-10ms).
Obviously we want a bo cache.. but this cleanup will make things a
bit easier for atomic as well and makes things a bit cleaner.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Acked-by: David Brown <davidb@codeaurora.org>
The userspace API already had everything needed to handle read vs write
synchronization. This patch actually bothers to hook it up properly, so
that we don't need to (for example) stall on userspace read access to a
buffer that gpu is also still reading.
Signed-off-by: Rob Clark <robdclark@gmail.com>
Add initial support for a3xx 3d core.
So far, with hardware that I've seen to date, we can have:
+ zero, one, or two z180 2d cores
+ a3xx or a2xx 3d core, which share a common CP (the firmware
for the CP seems to implement some different PM4 packet types
but the basics of cmdstream submission are the same)
Which means that the eventual complete "class" hierarchy, once
support for all past and present hw is in place, becomes:
+ msm_gpu
+ adreno_gpu
+ a3xx_gpu
+ a2xx_gpu
+ z180_gpu
This commit splits out the parts that will eventually be common
between a2xx/a3xx into adreno_gpu, and the parts that are even
common to z180 into msm_gpu.
Note that there is no cmdstream validation required. All memory access
from the GPU is via IOMMU/MMU. So as long as you don't map silly things
to the GPU, there isn't much damage that the GPU can do.
Signed-off-by: Rob Clark <robdclark@gmail.com>
The snapdragon chips have multiple different display controllers,
depending on which chip variant/version. (As far as I can tell, current
devices have either MDP3 or MDP4, and upcoming devices have MDSS.) And
then external to the display controller are HDMI, DSI, etc. blocks which
may be shared across devices which have different display controller
blocks.
To more easily add support for different display controller blocks, the
display controller specific bits are split out into a "kms" module,
which provides the kms plane/crtc/encoder objects.
The external HDMI, DSI, etc. blocks are part encoder, and part connector
currently. But I think I will pull in the drm_bridge patches from
chromeos tree, and split them into a bridge+connector, with the
registers that need to be set in modeset handled by the bridge. This
would remove the 'msm_connector' base class. But some things need to be
double checked to make sure I could get the correct ON/OFF sequencing..
This patch adds support for mdp4 crtc (including hw cursor), dtv encoder
(part of MDP4 block), and hdmi.
Signed-off-by: Rob Clark <robdclark@gmail.com>
