Age | Commit message (Collapse) | Author | Files | Lines |
|
Spreading the load across multiple SDMA engines can increase memory
transfer performance.
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
|
|
Depending on usage patterns, the current LRU policy may create a
non-injective mapping between userspace ring ids and kernel rings.
This behaviour is undesired as apps that attempt to fill all HW blocks
would be unable to reach some of them.
This change forces the LRU policy to create bijective mappings only.
v2: compress ring_blacklist
v3: simplify amdgpu_ring_is_blacklisted() logic
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
|
|
Use an LRU policy to map usermode rings to HW compute queues.
Most compute clients use one queue, and usually the first queue
available. This results in poor pipe/queue work distribution when
multiple compute apps are running. In most cases pipe 0 queue 0 is
the only queue that gets used.
In order to better distribute work across multiple HW queues, we adopt
a policy to map the usermode ring ids to the LRU HW queue.
This fixes a large majority of multi-app compute workloads sharing the
same HW queue, even though 7 other queues are available.
v2: use ring->funcs->type instead of ring->hw_ip
v3: remove amdgpu_queue_mapper_funcs
v4: change ring_lru_list_lock to spinlock, grab only once in lru_get()
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Add amdgpu_queue_mgr, a mechanism that allows disjointing usermode's
ring ids from the kernel's ring ids.
The queue manager maintains a per-file descriptor map of user ring ids
to amdgpu_ring pointers. Once a map is created it is permanent (this is
required to maintain FIFO execution guarantees for a context's ring).
Different queue map policies can be configured for each HW IP.
Currently all HW IPs use the identity mapper, i.e. kernel ring id is
equal to the user ring id.
The purpose of this mechanism is to distribute the load across multiple
queues more effectively for HW IPs that support multiple rings.
Userspace clients are unable to check whether a specific resource is in
use by a different client. Therefore, it is up to the kernel driver to
make the optimal choice.
v2: remove amdgpu_queue_mapper_funcs
v3: made amdgpu_queue_mgr per context instead of per-fd
v4: add context_put on error paths
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Tonga based asics may experience hangs when an HQD's EOP parameters
are modified.
Workaround this HW issue by avoiding writes to these registers for
tonga asics.
Based on the following ROCm commit:
2a0fb8 - drm/amdgpu: Synchronize KFD HQD load protocol with CP scheduler
From the ROCm git repository:
https://github.com/RadeonOpenCompute/ROCK-Kernel-Driver.git
CC: Jay Cornwall <Jay.Cornwall@amd.com>
Suggested-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
The MQD structure matches the reg layout. Take advantage of this to
simplify HQD programming.
Note that the ACTIVE field still needs to be programmed last.
Suggested-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Instead of taking the first pipe and giving the rest to kfd, take the
first 2 queues of each pipe.
Effectively, amdgpu and amdkfd own the same number of queues. But
because the queues are spread over multiple pipes the hardware will be
able to better handle concurrent compute workloads.
amdgpu goes from 1 pipe to 4 pipes, i.e. from 1 compute threads to 4
amdkfd goes from 3 pipe to 4 pipes, i.e. from 3 compute threads to 4
v2: fix policy comment
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Instead of picking an arbitrary queue for KIQ, search for one according
to policy. The queue must be unused.
Also report the KIQ as an unavailable resource to KFD.
In testing I ran into KCQ initialization issues when using pipes 2/3 of
MEC2 for the KIQ. Therefore the policy disallows grabbing one of these.
v2: fix (ring.me + 1) to (ring.me -1) in amdgpu_amdkfd_device_init
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
The assumption that we are only using the first pipe no longer holds.
Instead, calculate the queue_mask from the queue_bitmap.
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Pipes provide better concurrency than queues, therefore we want to make
sure that apps use queues from different pipes whenever possible.
Optimize for the trivial case where an app will consume rings in order,
therefore we don't want adjacent rings to belong to the same pipe.
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
This information is already available in adev.
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Update the KGD to KFD interface to allow sharing pipes with queue
granularity instead of pipe granularity.
This allows for more interesting pipe/queue splits.
v2: fix overflow check for res.queue_mask
v3: fix shift overflow when setting res.queue_mask
v4: fix comment in is_pipeline_enabled()
v5: clamp res.queue_mask to the first MEC only
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
The current implementation is hardcoded to enable ME1/PIPE0 interrupts
only.
This patch allows amdgpu to enable interrupts for any pipe of ME1.
v2: added gfx9 support
v3: use soc15_grbm_select for gfx9
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Previously the queue/pipe split with kfd operated with pipe
granularity. This patch allows amdgpu to take ownership of an arbitrary
set of queues.
It also consolidates the last few magic numbers in the compute
initialization process into mec_init.
v2: support for gfx9
v3: renamed AMDGPU_MAX_QUEUES to AMDGPU_MAX_COMPUTE_QUEUES
v4: fix off-by-one in num_mec checks in *_compute_queue_acquire
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Make amdgpu the owner of all per-pipe state of the HQDs.
This change will allow us to split the queues between kfd and amdgpu
with a queue granularity instead of pipe granularity.
This patch fixes kfd allocating an HDP_EOP region for its 3 pipes which
goes unused.
v2: support for gfx9
v3: fix gfx7 HPD intitialization
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Take ownership of pipe initialization away from KFD.
Note that hpd_eop_gpu_addr was already large enough to accomodate all
pipes.
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Rename straggler instances of r(adeon)dev to a(mdgpu)dev
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
The return value from copy_form_user is 0 for the success case.
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Use the same gfx_*_mqd_commit function for kfd and amdgpu codepaths.
This removes the last duplicates of this programming sequence.
v2: fix cp_hqd_pq_wptr value
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Acked-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
The gfxv7 contains a slightly different version of cik_mqd called
bonaire_mqd. This can introduce subtle bugs if fixes are not applied in
both places.
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Handle HQD deactivation timeouts instead of ignoring them.
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
The MQD programming sequence currently exists in 3 different places.
Refactor it to absorb all the duplicates.
The success path remains mostly identical except for a slightly
different order in the non-kiq case. This shouldn't matter if the HQD
is disabled.
The error handling paths have been updated to deal with the new code
structure.
v2: the non-kiq path for gfxv8 was dropped in the rebase
v3: split MEC_HPD_SIZE rename, dropped doorbell changes
Reviewed-by: Edward O'Callaghan <funfunctor@folklore1984.net>
Acked-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
Rename MEC_HPD_SIZE to GFXN_MEC_HPD_SIZE to clarify it is specific to a
gfx generation.
Acked-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
|
|
There is no good mechanism to handle the corresponding error.
When signal interrupt happens, unpin is not called.
As a result, inside AMDGPU, the statistic of pin size will be wrong.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Either in cgs functions or for callers of cgs functions:
1. The signal interrupt can affect the expected behaviour
2. There is no good mechanism to handle the corresponding error
3. There is no chance of deadlock in these single BO waiting
4. There is no clear benefit for interruptible waiting
5. Future caller of these functions might have same issue.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
[ 141.420491] BUG: unable to handle kernel NULL pointer dereference at 0000000000000030
[ 141.420532] IP: [<ffffffff81579ee1>] fence_remove_callback+0x11/0x60
[ 141.420563] PGD 20a030067
[ 141.420575] PUD 2088ca067
[ 141.420587] PMD 0
[ 141.420599] Oops: 0000 [#1] SMP
[ 141.420612] Modules linked in: amdgpu(OE) ttm(OE) drm_kms_helper(E) drm(E) i2c_algo_bit(E) fb_sys_fops(E) syscopyarea(E) sysfillrect(E) sysimgblt(E) rpcsec_gss_krb5(E) nfsv4(E) nfs(E) fscache(E) eeepc_wmi(E) asus_wmi(E) sparse_keymap(E) snd_hda_codec_realtek(E) video(E) snd_hda_codec_generic(E) snd_hda_codec_hdmi(E) snd_hda_intel(E) joydev(E) snd_hda_codec(E) snd_seq_midi(E) snd_seq_midi_event(E) snd_hda_core(E) snd_hwdep(E) snd_rawmidi(E) snd_pcm(E) kvm(E) irqbypass(E) crct10dif_pclmul(E) snd_seq(E) crc32_pclmul(E) ghash_clmulni_intel(E) snd_seq_device(E) snd_timer(E) aesni_intel(E) aes_x86_64(E) lrw(E) gf128mul(E) glue_helper(E) ablk_helper(E) cryptd(E) snd(E) soundcore(E) serio_raw(E) shpchp(E) i2c_piix4(E) i2c_designware_platform(E) 8250_dw(E) i2c_designware_core(E) mac_hid(E) binfmt_misc(E)
[ 141.420948] nfsd(E) auth_rpcgss(E) nfs_acl(E) lockd(E) grace(E) sunrpc(E) parport_pc(E) ppdev(E) lp(E) parport(E) autofs4(E) hid_generic(E) usbhid(E) hid(E) psmouse(E) r8169(E) ahci(E) mii(E) libahci(E) wmi(E)
[ 141.421042] CPU: 14 PID: 223 Comm: kworker/14:2 Tainted: G OE 4.9.0-custom #4
[ 141.421074] Hardware name: System manufacturer System Product Name/PRIME B350-PLUS, BIOS 0606 04/06/2017
[ 141.421146] Workqueue: events amd_sched_job_timedout [amdgpu]
[ 141.421169] task: ffff88020b03ba80 task.stack: ffffc900016f4000
[ 141.421193] RIP: 0010:[<ffffffff81579ee1>] [<ffffffff81579ee1>] fence_remove_callback+0x11/0x60
[ 141.421229] RSP: 0018:ffffc900016f7d30 EFLAGS: 00010202
[ 141.421250] RAX: ffff8801c049fc00 RBX: ffff8801d4d8dc00 RCX: 0000000000000000
[ 141.421278] RDX: 0000000000000001 RSI: ffff8801c049fcc0 RDI: 0000000000000000
[ 141.421307] RBP: ffffc900016f7d48 R08: 0000000000000000 R09: 0000000000000000
[ 141.421334] R10: 00000020ed512a30 R11: 0000000000000001 R12: 0000000000000000
[ 141.421362] R13: ffff880209ba4ba0 R14: ffff880209ba4c58 R15: ffff8801c055cc60
[ 141.421390] FS: 0000000000000000(0000) GS:ffff88021ef80000(0000) knlGS:0000000000000000
[ 141.421421] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 141.421443] CR2: 0000000000000030 CR3: 000000020b554000 CR4: 00000000003406e0
[ 141.421471] Stack:
[ 141.421480] ffff8801d4d8dc00 ffff880209ba4c48 ffff880209ba4ba0 ffffc900016f7d78
[ 141.421513] ffffffffa0697920 ffff880209ba0000 0000000000000000 ffff880209ba2770
[ 141.421549] ffff880209ba4b08 ffffc900016f7df0 ffffffffa05ce2ae ffffffffa0509eb7
[ 141.421583] Call Trace:
[ 141.421628] [<ffffffffa0697920>] amd_sched_hw_job_reset+0x50/0xb0 [amdgpu]
[ 141.421676] [<ffffffffa05ce2ae>] amdgpu_gpu_reset+0x8e/0x690 [amdgpu]
[ 141.421712] [<ffffffffa0509eb7>] ? drm_printk+0x97/0xa0 [drm]
[ 141.421770] [<ffffffffa0698156>] amdgpu_job_timedout+0x46/0x50 [amdgpu]
[ 141.421829] [<ffffffffa0696a07>] amd_sched_job_timedout+0x17/0x20 [amdgpu]
[ 141.421859] [<ffffffff81095493>] process_one_work+0x153/0x3f0
[ 141.421884] [<ffffffff81095c5b>] worker_thread+0x12b/0x4b0
[ 141.421907] [<ffffffff81095b30>] ? rescuer_thread+0x350/0x350
[ 141.421931] [<ffffffff8109b423>] kthread+0xd3/0xf0
[ 141.421951] [<ffffffff8109b350>] ? kthread_park+0x60/0x60
[ 141.421975] [<ffffffff817e1ee5>] ret_from_fork+0x25/0x30
[ 141.421996] Code: ac 81 e8 a3 1f b0 ff 48 c7 c0 ea ff ff ff e9 48 ff ff ff 0f 1f 80 00 00 00 00 0f 1f 44 00 00 55 48 89 e5 41 55 41 54 49 89 fc 53 <48> 8b 7f 30 48 89 f3 e8 73 7c 26 00 48 8b 13 48 39 d3 41 0f 95
[ 141.422156] RIP [<ffffffff81579ee1>] fence_remove_callback+0x11/0x60
[ 141.422183] RSP <ffffc900016f7d30>
[ 141.422197] CR2: 0000000000000030
[ 141.433483] ---[ end trace bc0949bf7ddd6d4b ]---
if the job is reset twice, then the parent could be NULL.
Signed-off-by: Chunming Zhou <David1.Zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Reviewed-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Roger.He <Hongbo.He@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
1. The signal interrupt can affect the expected behaviour.
2. There is no good mechanism to handle the corresponding error.
When signal interrupt happens, unpin is not called.
As a result, inside AMDGPU, the statistic of pin size will be wrong.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
1. The signal interrupt can affect the expected behaviour.
2. There is no good mechanism to handle the corresponding error.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
1. The signal interrupt can affect the expected behaviour.
2. There is no good mechanism to handle the corresponding error.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
1. The signal interrupt can affect the expected behaviour.
2. There is no mechanism to handle the corresponding error.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
If amdgpu_bo_reserve function is interrupted by signal,
amdgpu_bo_kunmap function is not called.
Signed-off-by: Alex Xie <AlexBin.Xie@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Avoid big roundoff errors in scanline/hactive durations for
high pixel clocks, especially for >= 500 Mhz, and thereby
program more accurate display fifo watermarks.
This is a port of the corresponding amdgpu patch.
Implemented for DCE 4,6,8.
Tested on Evergreen/DCE-4 with Radeon HD-5770.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
At dot clocks > approx. 250 Mhz, some of these calcs will overflow and
cause miscalculation of latency watermarks, and for some overflows also
divide-by-zero driver crash. Make calcs more overflow resistant.
This is a direct port of the corresponding patch from amdgpu-kms,
copy-paste for cik from dce-8 and si from dce-6, with a slightly
simpler variant for evergreen dce-4/5.
Only tested on DCE-4 evergreen with a Radeon HD-5770.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
This apparently got lost when implementing the new DCE-6 support
and would cause failures in pageflip scheduling and timestamping.
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Function radeon_fence_emit() returns -ENOMEM if there is no enough
memory. And in this case, function radeon_ring_unlock_undo() rather than
function radeon_ring_unlock_commit() should be called. However, in
function radeon_test_create_and_emit_fence(), the return value of
radeon_fence_emit() is ignored. This patch adds the check.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Function radeon_ring_lock() returns an errno on failure, and its return
value should be validated. However, in functions r420_cp_errata_init()
and r420_cp_errata_fini(), its return value is not checked. This patch
adds the checks.
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Pan Bian <bianpan2016@163.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Enable UVD block for SRIOV.
Signed-off-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Add UVD hw init.
Signed-off-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Add UVD doorbell for SRIOV.
Signed-off-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Add UVD initialization for SRIOV.
Signed-off-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Used virt_alloc_mm_table function to allocate MM table memory.
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Add two functions to allocate & free MM table memory.
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Move mm table construction functions into mmsch header file so that
UVD can reuse it.
Signed-off-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Fixed PSP loading issue for sriov.
Signed-off-by: Daniel Wang <Daniel.Wang2@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Now GPU hypervisor will load SDMA and RLCG ucode, so skip it
in guest.
Signed-off-by: Daniel Wang <Daniel.Wang2@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
Change place of virt_init_setting function so that can cover the
cg and pg flags configuration.
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|
|
GPU hypervisor cover all settings of CG and PG, so guest doesn't
need to do anything. Bypass it.
Signed-off-by: Frank Min <Frank.Min@amd.com>
Signed-off-by: Xiangliang Yu <Xiangliang.Yu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
|