New stuff for 5.1.
amdgpu:
- DC bandwidth formula updates
- Support for DCC on scanout surfaces
- Support for multiple IH rings on soc15 asics
- Fix xgmi locking
- Add sysfs interface to get pcie usage stats
- Simplify DC i2c/aux code
- Initial support for BACO on vega10/20
- New runtime SMU feature debug interface
- Expand existing sysfs power interfaces to new clock domains
- Handle kexec properly
- Simplify IH programming
- Rework doorbell handling across asics
- Drop old CI DPM implementation
- DC page flipping fixes
- Misc SR-IOV fixes
amdkfd:
- Simplify the interfaces between amdkfd and amdgpu
ttm:
- Add a callback to notify the driver when the lru changes
sched:
- Refactor mirror list handling
- Rework hw fence processing
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190125231517.26268-1-alexander.deucher@amd.com
amdgpu only uses shared-fences internally, but dmabuf importers rely on
implicit write hazard tracking via the reservation_object.fence_excl.
For example, the importer use the write hazard for timing a page flip to
only occur after the exporter has finished flushing its write into the
surface. As such, on exporting a dmabuf, we must either flush all
outstanding fences (for we do not know which are writes and should have
been exclusive) or alternatively create a new exclusive fence that is
the composite of all the existing shared fences, and so will only be
signaled when all earlier fences are signaled (ensuring that we can not
be signaled before the completion of any earlier write).
v2: reservation_object is already locked by amdgpu_bo_reserve()
v3: Replace looping with get_fences_rcu and special case the promotion
of a single shared fence directly to an exclusive fence, bypassing the
fence array.
v4: Drop the fence array ref after assigning to reservation_object
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107341
Testcase: igt/amd_prime/amd-to-i915
References: 8e94a46c17 ("drm/amdgpu: Attach exclusive fence to prime exported bo's. (v5)")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: "Christian König" <christian.koenig@amd.com>
Reviewed-by: "Christian König" <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
sriov need to restrict max_pfn below AMDGPU_GMC_HOLE.
access the hole results in a range fault interrupt IIRC.
Signed-off-by: Wentao Lou <Wentao.Lou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
After removing unnecessary VM size calculations,
vm_manager.max_pfn would reach 0x10,0000,0000
max_pfn << AMDGPU_GPU_PAGE_SHIFT exceeding AMDGPU_GMC_HOLE_START
would cause GPU reset.
Signed-off-by: wentalou <Wentao.Lou@amd.com>
Reviewed-by: Monk Liu <monk.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add missing power_average to visible check for power
attributes for APUs. Was missed before.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This is only used by drm_irq_install(), which is an optional helper.
For legacy pci devices this is required (due to interrupt sharing without
msi/msi-x), and just making this the default exactly matches the behaviour
of all existing drivers using the drm_irq_install() helpers. In case that
ever becomes wrong drivers can roll their own irq handling, as many
drivers already do (for other reasons like needing a threaded interrupt
handler, or having an entire pile of different interrupt sources).
v2: Rebase
v3: Improve commit message (Emil)
Cc: Sam Ravnborg <sam@ravnborg.org>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190129104248.26607-3-daniel.vetter@ffwll.ch
Decauple sched threads stop and start and ring mirror
list handling from the policy of what to do about the
guilty jobs.
When stoppping the sched thread and detaching sched fences
from non signaled HW fenes wait for all signaled HW fences
to complete before rerunning the jobs.
v2: Fix resubmission of guilty job into HW after refactoring.
v4:
Full restart for all the jobs, not only from guilty ring.
Extract karma increase into standalone function.
v5:
Rework waiting for signaled jobs without relying on the job
struct itself as those might already be freed for non 'guilty'
job's schedulers.
Expose karma increase to drivers.
v6:
Use list_for_each_entry_safe_continue and drm_sched_process_job
in case fence already signaled.
Call drm_sched_increase_karma only once for amdgpu and add documentation.
v7:
Wait only for the latest job's fence.
Suggested-by: Christian Koenig <Christian.Koenig@amd.com>
Signed-off-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The power smu7 powerplay code is much more robust and has
been the default for a while now. Remove the old code.
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add missing power_average to visible check for power
attributes for APUs. Was missed before.
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Replace the last bool type parameter with a general flags parameter,
to make the last parameter be able to contain more information.
v2: drop setting need_ctx_switch = false
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Signed-off-by: Jack Xiao <Jack.Xiao@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
sriov would meet guest driver load failure,
if calling amdgpu_asic_reset in amdgpu_device_init.
sriov should skip asic_reset in device_init.
Signed-off-by: Wentao Lou <Wentao.Lou@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The entries are ignored for now, but it at least stops crashing the
hardware when somebody tries to push something to the other IH rings.
v2: limit ring size, add TODO comment
v3: only program rings if they are actually allocated
v4: limit the ring init to Vega10
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Otherwise we run into a non-retry fault on access.
It seems to be a hardware bug that the executable bit has
higher priority than the valid bit.
v2: handle clears as well
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
User can use "pp_dpm_dcefclk" to retrieve and adjust dcefclock power
levels.
V2: expose this interface for Vega10 and later ASICs only
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
User can use "pp_dpm_fclk" to retrieve and adjust fclock power
levels.
V2: expose this interface for Vega20 and later ASICs only
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
User can use "pp_dpm_socclk" to retrieve and adjust SOC clock power
levels.
V2: expose this interface for Vega10 and later ASICs only
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
User can use "ppfeatures" sysfs interface to retrieve and set enabled
powerplay features.
V2: expose this feature for Vega10 and later dGPUs
V3: squash in removal of unused variable (Alex)
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
In some cases, psp response status is not 0 even there is no
problem while the command is submitted. Some version of PSP FW
doesn't write 0 to that field.
So here we would like to only print a warning instead of an error
during psp initialization to avoid breaking hw_init and it doesn't
return -EINVAL.
Signed-off-by: Aaron Liu <aaron.liu@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Xiangliang Yu<Xiangliang.Yu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Feifei Xu <Feifei.Xu@amd.com>
Reviewed-by: Paul Menzel <pmenzel+amd-gfx@molgen.mpg.de>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
HW doorbell writing routing policy: writing to doorbell
not in SDMA/IH/MM/ACV doorbell range will be routed to CP.
So CP doorbell routing depends on doorbell range setting
of above blocks. Setting doorbell range of above blocks
earlier (soc15_common_hw_init) to make sure CP doorbell
writing be routed to CP block.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Different ASIC has different SDMA queue number so
different SDMA doorbell range. Introduce an extra
parameter to sdma_doorbell_range function and set
sdma doorbell range correctly.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Different ASIC has different sdma doorbell range. Add
a per device sdma_doorbell_range field and initialize
it.
Signed-off-by: Oak Zeng <Oak.Zeng@amd.com>
Reviewed-by: Philip Yang <Philip.Yang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It was at the same time too strict (for linear tiling modes, where no
height alignment is required) and too lenient (for 2D tiling modes,
where height may need to be aligned to values > 8).
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SOC15 chips require a reset if the driver was previously loaded
because the PSP can only be loaded once between each reset.
v2: rebase, handle multiple asic funcs
Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
VI chips require a reset if the driver was previously loaded
because the SMU can only be loaded once between each reset.
v2: rebase
Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
CIK chips require a reset if the driver was previously loaded
because the SMU can only be loaded once between each reset.
v2: rebase
Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Used to determine if we need to reset the asic on init due
to the driver having been previously loaded or not shutdown
cleanly. E.g., kexec or VM passthrough.
v2: rebase
Reviewed-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Expose sclk (gfx clock) and mclk (memory clock) via
hwmon compatible interface. hwmon does not actually
formally specify a frequency type attribute, but these
are compatible with the format of the other attributes
exposed via hwmon. Units are hertz.
freq1_input - GPU gfx/compute clock in hertz
freq2_input - GPU memory clock in hertz (dGPU only)
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Add a sysfs file that reports the number of bytes transmitted and
received in the last second. This can be used to approximate the PCIe
bandwidth usage over the last second.
v2: Clarify use of mps as estimation of bandwidth
v3: Don't make the file on APUs
v4: Early exit for APUs in the read function, change output to
display "packets-received packets-sent mps"
v5: fix missing header for si (Alex)
Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We need these offsets for PCIE perf counters, so include them as well as
the the previously-used defines from the nbio_*.c files
v2: Return NBIF definitions back to previous files
Signed-off-by: Kent Russell <kent.russell@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v2: Move locks around in other functions so that this
function can stand on its own. Also only hold the hive
specific lock for add/remove device instead of the driver
global lock so you can't add/remove devices in parallel from
one hive.
v3: add reset_lock
Acked-by: Shaoyun.liu < Shaoyun.liu@amd.com>
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
