drm/sched: implement dynamic job-flow control

Currently, job flow control is implemented simply by limiting the number
of jobs in flight. Therefore, a scheduler is initialized with a credit
limit th

drm/sched: implement dynamic job-flow control

Currently, job flow control is implemented simply by limiting the number
of jobs in flight. Therefore, a scheduler is initialized with a credit
limit that corresponds to the number of jobs which can be sent to the
hardware.

This implies that for each job, drivers need to account for the maximum
job size possible in order to not overflow the ring buffer.

However, there are drivers, such as Nouveau, where the job size has a
rather large range. For such drivers it can easily happen that job
submissions not even filling the ring by 1% can block subsequent
submissions, which, in the worst case, can lead to the ring run dry.

In order to overcome this issue, allow for tracking the actual job size
instead of the number of jobs. Therefore, add a field to track a job's
credit count, which represents the number of credits a job contributes
to the scheduler's credit limit.

Signed-off-by: Danilo Krummrich <[email protected]>
Reviewed-by: Luben Tuikov <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

treewide: Add missing includes masked by cgroup -> bpf dependency

cgroup.h (therefore swap.h, therefore half of the universe)
includes bpf.h which in turn includes module.h and slab.h.
Since we're a

treewide: Add missing includes masked by cgroup -> bpf dependency

cgroup.h (therefore swap.h, therefore half of the universe)
includes bpf.h which in turn includes module.h and slab.h.
Since we're about to get rid of that dependency we need
to clean things up.

v2: drop the cpu.h include from cacheinfo.h, it's not necessary
and it makes riscv sensitive to ordering of include files.

Signed-off-by: Jakub Kicinski <[email protected]>
Signed-off-by: Alexei Starovoitov <[email protected]>
Reviewed-by: Christoph Hellwig <[email protected]>
Acked-by: Krzysztof Wilczyński <[email protected]>
Acked-by: Peter Chen <[email protected]>
Acked-by: SeongJae Park <[email protected]>
Acked-by: Jani Nikula <[email protected]>
Acked-by: Greg Kroah-Hartman <[email protected]>
Link: https://lore.kernel.org/all/[email protected]/ # v1
Link: https://lore.kernel.org/all/[email protected]/ # cacheinfo discussion
Link: https://lore.kernel.org/bpf/[email protected]

show more ...

drm/lima: fix warning when CONFIG_DEBUG_SG=y & CONFIG_DMA_API_DEBUG=y

Otherwise get following warning:

DMA-API: lima 1c40000.gpu: mapping sg segment longer than device claims to support [len=414924

drm/lima: fix warning when CONFIG_DEBUG_SG=y & CONFIG_DMA_API_DEBUG=y

Otherwise get following warning:

DMA-API: lima 1c40000.gpu: mapping sg segment longer than device claims to support [len=4149248] [max=65536]

See: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5496

Reviewed-by: Vasily Khoruzhick <[email protected]>
Reported-by: Roman Stratiienko <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: add pm resume/suspend ops

Add driver pm system and runtime hardware resume/suspend ops.
Note this won't enable runtime pm of the device yet.

v2:
Do clock and power gating when suspend/res

drm/lima: add pm resume/suspend ops

Add driver pm system and runtime hardware resume/suspend ops.
Note this won't enable runtime pm of the device yet.

v2:
Do clock and power gating when suspend/resume.

Tested-by: Bhushan Shah <[email protected]>
Reviewed-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: separate clk/regulator enable/disable function

For being used by both device init/fini and suspend/resume.

Tested-by: Bhushan Shah <[email protected]>
Reviewed-by: Vasily Khoruzhick <anarsoul

drm/lima: separate clk/regulator enable/disable function

For being used by both device init/fini and suspend/resume.

Tested-by: Bhushan Shah <[email protected]>
Reviewed-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: add resume/suspend callback for each ip

For called when PM do resume/suspend.

Tested-by: Bhushan Shah <[email protected]>
Reviewed-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qi

drm/lima: add resume/suspend callback for each ip

For called when PM do resume/suspend.

Tested-by: Bhushan Shah <[email protected]>
Reviewed-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: Clean up redundant pdev pointer

There's no point explicitly tracking the platform device when it can be
trivially derived from the regular device pointer in the couple of
places it's ever

drm/lima: Clean up redundant pdev pointer

There's no point explicitly tracking the platform device when it can be
trivially derived from the regular device pointer in the couple of
places it's ever used.

Signed-off-by: Robin Murphy <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/8d9073cc91c10fc70910587fd1794e0e8f32b467.1587509150.git.robin.murphy@arm.com

show more ...

drm/lima: Clean up IRQ warnings

Use the optional form of platform_get_irq() for blocks that legitimately
may not be present, to avoid getting an annoying barrage of spurious
warnings for non-existen

drm/lima: Clean up IRQ warnings

Use the optional form of platform_get_irq() for blocks that legitimately
may not be present, to avoid getting an annoying barrage of spurious
warnings for non-existent PPs on configurations like Mali-450 MP2.

Signed-off-by: Robin Murphy <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/de475904091400ef6c123285f221094654d96d35.1587509150.git.robin.murphy@arm.com

show more ...

drm/lima: Add optional devfreq and cooling device support

Most platforms with a Mali-400 or Mali-450 GPU also have support for
changing the GPU clock frequency. Add devfreq support so the GPU clock

drm/lima: Add optional devfreq and cooling device support

Most platforms with a Mali-400 or Mali-450 GPU also have support for
changing the GPU clock frequency. Add devfreq support so the GPU clock
rate is updated based on the actual GPU usage when the
"operating-points-v2" property is present in the board.dts.

The actual devfreq code is taken from panfrost_devfreq.c and modified so
it matches what the lima hardware needs:
- a call to dev_pm_opp_set_clkname() during initialization because there
are two clocks on Mali-4x0 IPs. "core" is the one that actually clocks
the GPU so we need to control it using devfreq.
- locking when reading or writing the devfreq statistics because (unlike
than panfrost) we have multiple PP and GP IRQs which may finish jobs
concurrently.

Signed-off-by: Martin Blumenstingl <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: save task info dump when task fail

Save all information to start a task which can be exported to user
for debug usage. Dump file data format is specified in lima_dump.h

v2:
Add include he

drm/lima: save task info dump when task fail

Save all information to start a task which can be exported to user
for debug usage. Dump file data format is specified in lima_dump.h

v2:
Add include header to address build robot complain.

Tested-by: Andreas Baierl <[email protected]>
Reviewed-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: add __GFP_NOWARN flag to all dma_alloc_wc

This prevent CMA printing dumy "PFNs busy" info which is
caused by alloc fail re-try case.

Reviewed-by: Vasily Khoruzhick <[email protected]>
Si

drm/lima: add __GFP_NOWARN flag to all dma_alloc_wc

This prevent CMA printing dumy "PFNs busy" info which is
caused by alloc fail re-try case.

Reviewed-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: Add support for multiple reset lines

Some SoCs like HiKey have 2 reset lines, so update
to use the devm_reset_control_array_* variant of the
API so that multiple resets can be specified in

drm/lima: Add support for multiple reset lines

Some SoCs like HiKey have 2 reset lines, so update
to use the devm_reset_control_array_* variant of the
API so that multiple resets can be specified in DT.

Cc: Qiang Yu <[email protected]>
Cc: David Airlie <[email protected]>
Cc: Daniel Vetter <[email protected]>
Cc: [email protected]
Cc: [email protected]
Signed-off-by: Peter Griffin <[email protected]>
Signed-off-by: John Stultz <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: Reduce the amount of logs on deferred probe of clocks and reset controller

There is no point to print deferred probe messages as errors. Adjust
the printks for error paths of obtaining cl

drm/lima: Reduce the amount of logs on deferred probe of clocks and reset controller

There is no point to print deferred probe messages as errors. Adjust
the printks for error paths of obtaining clocks and reset controller.
This removes the error message of lima_clk_init() call in favor or
specific failure messages inside.

Signed-off-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: Reduce number of PTR_ERR() calls

Store the PTR_ERR() result in local variable in clock init error path.
This makes the code consistent with similar section in regulator init
code.

Signed-

drm/lima: Reduce number of PTR_ERR() calls

Store the PTR_ERR() result in local variable in clock init error path.
This makes the code consistent with similar section in regulator init
code.

Signed-off-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: Reduce the amount of logs on deferred probe

There is no point to print deferred probe (and its failures to get
resources) as an error. For example getting a regulator causes three
unneede

drm/lima: Reduce the amount of logs on deferred probe

There is no point to print deferred probe (and its failures to get
resources) as an error. For example getting a regulator causes three
unneeded error messages:

lima 13000000.gpu: failed to get regulator: -517
lima 13000000.gpu: regulator init fail -517
lima 13000000.gpu: Fatal error during GPU init

Also do not print clock rates before the initialization finishes
because they will be duplicated after deferral. Each probe step already
prints error so remove the final error message "Fatal error during GPU
init".

In case of multiple probe tries this would pollute the dmesg.

Fixes: a1d2a6339961 ("drm/lima: driver for ARM Mali4xx GPUs")
Signed-off-by: Krzysztof Kozlowski <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]

show more ...

drm/lima: driver for ARM Mali4xx GPUs

- Mali 4xx GPUs have two kinds of processors GP and PP. GP is for
OpenGL vertex shader processing and PP is for fragment shader
processing. Each processor h

drm/lima: driver for ARM Mali4xx GPUs

- Mali 4xx GPUs have two kinds of processors GP and PP. GP is for
OpenGL vertex shader processing and PP is for fragment shader
processing. Each processor has its own MMU so prcessors work in
virtual address space.
- There's only one GP but multiple PP (max 4 for mali 400 and 8
for mali 450) in the same mali 4xx GPU. All PPs are grouped
togather to handle a single fragment shader task divided by
FB output tiled pixels. Mali 400 user space driver is
responsible for assign target tiled pixels to each PP, but mali
450 has a HW module called DLBU to dynamically balance each
PP's load.
- User space driver allocate buffer object and map into GPU
virtual address space, upload command stream and draw data with
CPU mmap of the buffer object, then submit task to GP/PP with
a register frame indicating where is the command stream and misc
settings.
- There's no command stream validation/relocation due to each user
process has its own GPU virtual address space. GP/PP's MMU switch
virtual address space before running two tasks from different
user process. Error or evil user space code just get MMU fault
or GP/PP error IRQ, then the HW/SW will be recovered.
- Use GEM+shmem for MM. Currently just alloc and pin memory when
gem object creation. GPU vm map of the buffer is also done in
the alloc stage in kernel space. We may delay the memory
allocation and real GPU vm map to command submission stage in the
furture as improvement.
- Use drm_sched for GPU task schedule. Each OpenGL context should
have a lima context object in the kernel to distinguish tasks
from different user. drm_sched gets task from each lima context
in a fair way.

mesa driver can be found here before upstreamed:
https://gitlab.freedesktop.org/lima/mesa

v8:
- add comments for in_sync
- fix ctx free miss mutex unlock

v7:
- remove lima_fence_ops with default value
- move fence slab create to device probe
- check pad ioctl args to be zero
- add comments for user/kernel interface

v6:
- fix comments by checkpatch.pl

v5:
- export gp/pp version to userspace
- rebase on drm-misc-next

v4:
- use get param interface to get info
- separate context create/free ioctl
- remove unused max sched task param
- update copyright time
- use xarray instead of idr
- stop using drmP.h

v3:
- fix comments from kbuild robot
- restrict supported arch to tested ones

v2:
- fix syscall argument check
- fix job finish fence leak since kernel 5.0
- use drm syncobj to replace native fence
- move buffer object GPU va map into kernel
- reserve syscall argument space for future info
- remove kernel gem modifier
- switch TTM back to GEM+shmem MM
- use time based io poll
- use whole register name
- adopt gem reservation obj integration
- use drm_timeout_abs_to_jiffies

Cc: Eric Anholt <[email protected]>
Cc: Rob Herring <[email protected]>
Cc: Christian König <[email protected]>
Cc: Daniel Vetter <[email protected]>
Cc: Alex Deucher <[email protected]>
Cc: Sam Ravnborg <[email protected]>
Cc: Rob Clark <[email protected]>
Cc: Dave Airlie <[email protected]>
Signed-off-by: Andreas Baierl <[email protected]>
Signed-off-by: Erico Nunes <[email protected]>
Signed-off-by: Heiko Stuebner <[email protected]>
Signed-off-by: Marek Vasut <[email protected]>
Signed-off-by: Neil Armstrong <[email protected]>
Signed-off-by: Simon Shields <[email protected]>
Signed-off-by: Vasily Khoruzhick <[email protected]>
Signed-off-by: Qiang Yu <[email protected]>
Reviewed-by: Eric Anholt <[email protected]>
Reviewed-by: Rob Herring <[email protected]>
Signed-off-by: Eric Anholt <[email protected]>
Link: https://patchwork.freedesktop.org/patch/291200/

show more ...