linux.git/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_arcturus.c, branch v5.8

kernel: move use_mm/unuse_mm to kthread.c

2020-06-11T02:14:18+00:00

Patch series "improve use_mm / unuse_mm", v2.

This series improves the use_mm / unuse_mm interface by better documenting
the assumptions, and my taking the set_fs manipulations spread over the
callers into the core API.

This patch (of 3):

Use the proper API instead.

Link: http://lkml.kernel.org/r/20200404094101.672954-1-hch@lst.de

These helpers are only for use with kernel threads, and I will tie them
more into the kthread infrastructure going forward.  Also move the
prototypes to kthread.h - mmu_context.h was a little weird to start with
as it otherwise contains very low-level MM bits.

Signed-off-by: Christoph Hellwig 
Signed-off-by: Andrew Morton 
Tested-by: Jens Axboe 
Reviewed-by: Jens Axboe 
Acked-by: Felix Kuehling 
Cc: Alex Deucher 
Cc: Al Viro 
Cc: Felipe Balbi 
Cc: Jason Wang 
Cc: "Michael S. Tsirkin" 
Cc: Zhenyu Wang 
Cc: Zhi Wang 
Cc: Greg Kroah-Hartman 
Link: http://lkml.kernel.org/r/20200404094101.672954-1-hch@lst.de
Link: http://lkml.kernel.org/r/20200416053158.586887-1-hch@lst.de
Link: http://lkml.kernel.org/r/20200404094101.672954-5-hch@lst.de
Signed-off-by: Linus Torvalds

AMD KFD: Use fallthrough;

2020-03-13T15:52:35+00:00

Convert the various uses of fallthrough comments to fallthrough;

Done via script
Link: https://lore.kernel.org/lkml/b56602fcf79f849e733e7b521bb0e17895d390fa.1582230379.git.joe@perches.com/

Acked-by: Felix Kuehling 
Signed-off-by: Joe Perches 
Signed-off-by: Alex Deucher

drm/amdkfd: Make get_tile_config() generic

2020-02-28T21:59:20+00:00

Given we can query all the asic specific information from amdgpu_gfx_config,
we can make get_tile_config() generic.

Signed-off-by: Yong Zhao 
Reviewed-by: Alex Deucher 
Signed-off-by: Alex Deucher

drm/amdkfd: use kiq to load the mqd of hiq queue for gfx v9 (v6)

2020-01-16T18:34:50+00:00

There is an issue that CP will check the HIQ queue to be configured and mapped
with KIQ ring, otherwise, it will be unable to read back the secure buffer while
the gfxoff is enabled even with trusted IP blocks.

v1 -> v2:
- Fix to remove surplus set_resources packets.
- Fill the whole configuration in MQD.
- Change the author as Aaron because he addressed the key point of this issue.
- Add kiq ring lock.

v2 -> v3:
- Free the lock while in error return case.
- Remove the programming only needed by the queue is unmapped.

v3 -> v4:
- Remove doorbell programming because it's used for restarting queue.
- Remove CP scheduler programming because map_queue packet will handle this.

v4 -> v5:
- Remove cp_hqd_active because mec ucode will enable it while use map_queues.
- Revise goto out_unlock.
- Correct the right doorbell offset for HIQ that kfd driver assigned in the
  packet.

v5 -> v6:
- Merge Arcturus fix into this patch because it will get oops in Arcturus
  platform.

Reported-by: Lisa Saturday 
Signed-off-by: Aaron Liu 
Signed-off-by: Huang Rui 
Reviewed-and-Tested-by: Aaron Liu 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher

drm/amdgpu: flush TLB functions removal from kfd2kgd interface

2020-01-16T18:34:42+00:00

[Why]
kfd2kgd interface will be deprecated. This removal only covers TLB
invalidation for now. They have been replaced in amdgpu_amdkfd API.

[How]
TLB invalidate functions removed from the different amdkfd_gfx_v*
versions.

Signed-off-by: Alex Sierra 
Acked-by: Christian König 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher

drm/amdkfd: Improve function get_sdma_rlc_reg_offset() (v2)

2020-01-14T15:17:05+00:00

The SOC15_REG_OFFSET() macro needs to dereference adev->reg_offset[IP]
pointer, which is sometimes NULL when there are fewer than 8 sdma engines.
Avoid that by not initializing the array regardless.

v2: squash in warning fixes

Signed-off-by: Yong Zhao 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher

drm/amdkfd: Use Arcturus specific set_vm_context_page_table_base()

2019-12-18T21:09:06+00:00

Since Arcturus has it own function pointer, we can move Arcturus
specific logic to there rather than leaving it entangled with
other GFX9 chips.

Signed-off-by: Yong Zhao 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher

drm/amdkfd: Delete unnecessary pr_fmt switch

2019-10-30T15:06:51+00:00

Given amdkfd.ko has been merged into amdgpu.ko, this switch is no
longer useful.

Signed-off-by: Yong Zhao 
Reviewed-by: Alex Deucher 
Signed-off-by: Alex Deucher

drm/amdkfd: Use array to probe kfd2kgd_calls

2019-10-03T14:11:05+00:00

This is the same idea as the kfd device info probe and move all the
probe control together for easy maintenance.

Signed-off-by: Yong Zhao 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher

drm/amdkfd: Eliminate get_atc_vmid_pasid_mapping_valid

2019-10-03T14:11:04+00:00

get_atc_vmid_pasid_mapping_valid() is very similar to
get_atc_vmid_pasid_mapping_pasid(), so they can be merged into a new
function get_atc_vmid_pasid_mapping_info() to reduce register access
times. More importantly, getting the PASID and the valid bit atomically
with a single read fixes some potential race conditions where the
mapping changes between the two reads.

Signed-off-by: Yong Zhao 
Reviewed-by: Felix Kuehling 
Signed-off-by: Alex Deucher