When many wavefronts cause VM faults at the same time, it can
overwhelm the interrupt handler and cause IH ring overflows before
the driver can notify or kill the faulting application.
As a workaround I'm introducing limited per-VM fault credit. After
that number of VM faults have occurred, further VM faults are
filtered out at the prescreen stage of processing.
This depends on the PASID in the interrupt packet, so it currently
only works for KFD contexts.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch provides a guarantee that the first n queues allocated by
an application will be on different pipes. Where n is the number of
pipes available from the hardware.
This helps avoid ring aliasing which can result in work executing in
time-sliced mode instead of truly parallel mode.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Useful for testing the effects of multipipe compute without recompiling.
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
A performance regression for OpenCL tests on Polaris11 had this feature
disabled for all asics.
Instead, disable it selectively on the affected asics.
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Andres Rodriguez <andresx7@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The error handling for virtual functions assumed a single
vf per VM and didn't properly account for bare metal. Make
the error arrays per device and add locking.
Reviewed-by: Gavin Wan <gavin.wan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
It was not clear. The rest of the driver is MIT/X11.
Reviewed-by: Christian König <christian.koenig@amd.com>
Acked-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v2: drop hdp invalidate/flush.
v3: honor pgoff during prime mmap. Add a barrier after cpu access.
v4: drop begin/end_cpu_access() for now, revisit later.
Signed-off-by: Samuel Li <Samuel.Li@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Just set the CPU access required flag when we pin it.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
this fix memory leak due to request_firmware after driver
unloaded
v2:
release gmc firmware for gmc6/7/8 as well
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
fix missing finish uvd enc_ring.
v2:
since the adev pointer check in already in ring_fini
so drop the check outsider
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
this way after KIQ MQD released in drv unloading, CPC
can still let KIQ access this MQD thus RLCV SAVE_VF
will not fail
v2:
always use VRAM domain for KIQ MQD no matter BM or SRIOV
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
v2:
move kcq_disable out of SRIOV, make it genearal
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
only with this way we can debug the VMC page fault issue
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use it to replace the hard coded value in amdgpu_vm_bo_update_mapping().
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When max_bytes is not 8 bytes aligned and bo size is larger than
max_bytes, the last 8 bytes in a ttm node may be left unchanged.
For example, on pre SDMA 4.0, max_bytes = 0x1fffff, and the bo size
is 0x200000, the problem will happen.
In order to fix the problem, we separately store the max nums of
PTEs/PDEs a single operation can set in amdgpu_vm_pte_funcs
structure, rather than inferring it from bytes limit of SDMA
constant fill, i.e. fill_max_bytes.
Together with the fix, we replace the hard code value "10" in
amdgpu_vm_bo_update_mapping() with the corresponding values from
structure amdgpu_vm_pte_funcs.
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Yong Zhao <yong.zhao@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Use 2MB fragment size by default for older hardware generations as well.
Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: John Bridgman <john.bridgman@amd.com>
Reviewed-by: Roger He <Hongbo.He@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
SRIOV doesn't implement PMC capability of PCIe, so it can't update
power state by reading PMC register.
Currently, amdgpu driver doesn't disable pci device when removing
driver, the enable_cnt of pci device will not be decrease to 0.
When reloading driver, pci_enable_device will do nothing as
enable_cnt is not zero. And power state will not be updated as PMC
is not support.
So current_state of pci device is not D0 state and pci_enable_msi
return fail.
Add pci_disable_device when remmoving driver to fix the issue.
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>
Signed-off-by: Eric Huang <JinHuiEric.Huang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
We discovered that on some devices even with iommu enabled
you can access all of system memory through the iommu translation.
Therefore, we revert the read method to the translation only service
and drop the write method completely.
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Christan König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
GPU reset will require all hw doing hw_init thus
ucode_init_bo will be invoked again, which lead to
memory leak
skip the fw_buf allocation during sriov gpu reset to avoid
memory leak.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
otherwise a gpu hang will make application couldn't be killed
under timedout=0 mode
v2:
Fix memoryleak job/job->s_fence issue
unlock mn
remove the ERROR msg after waiting being interrupted
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
RLC need CSB registers initiated under SRIOV during world switch
otherwise the clear state buffer behav will not be recovered to
current VF scheme after switch back
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
increase timeout to 12 seconds,because there may have multiple
FLR waiting for done, the waiting time of events may be long,
increase to 12s to reduce timeout failure.
Signed-off-by: Horace Chen <horace.chen@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
bo_free on csa is too late to put in amdgpu_fini because that
time ttm is already finished,
Move it earlier to avoid the page fault.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Signed-off-by: Horace Chen <horace.chen@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
FRAME_CONTROL(begin) is needed for vega10 due to ucode logic change,
it can fix some CTS random fail under gfx preemption enabled mode.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
At least for SRIOV we found reload PSP fw during
gpu reset cause PSP hang.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
currently in_reset is only used in sriov gpu reset, and it
will be used for other non-gfx hw component later, like
PSP, so move it from gfx to adev and rename to in_sriov_reset
make more sense.
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Monk Liu <Monk.Liu@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
V2
Signed-off-by: Ken Wang <Ken.Wang@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(v2): Add domain to iova debugfs
(v3): Add true read/write methods to access system memory of pages
mapped to the device
(v4): Move get_domain call out of loop and return on error
(v5): Just use kmap/kunmap
Signed-off-by: Tom St Denis <tom.stdenis@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
(v2): add domains and avoid strcmp
fix checkpatch.pl WARNING:
Prefer 'unsigned int' to bare use of 'unsigned'
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Requested by SRIOV, the clearance of the bit moved into firmware
Signed-off-by: Leo Liu <leo.liu@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Port it from sdma4 for wptr polling usage.
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>
When hypervisor triggering FLR for one of VFs, need to enable sdma
wptr polling to avoid missing wptr update if enabling doorbell.
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>
IH tracks pending retry faults in a hash table for fast lookup in
interrupt context. Each VM has a short FIFO of pending VM faults for
processing in a bottom half.
The IH prescreening stage adds retry faults and filters out repeated
retry interrupts to minimize the impact of interrupt storms.
It's the VM's responsibility remove pending faults once they are
handled. For now this is only done when the VM is destroyed.
v2:
- Made the hash table smaller and the FIFO longer. I never want the
FIFO to fill up, because that would make prescreen take longer.
128 pending page faults should be enough to keep migrations busy.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Christian König <christian.koenig@amd.com> (v1)
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
To filter out high-frequency interrupts that can be safely ignored.
v2: squash in trivial typo fix for si (Alex)
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Allows assigning a PASID to a VM for identifying VMs involved in page
faults. The global PASID manager is also exported in the KFD
interface so that AMDGPU and KFD can share the PASID space.
PASIDs of different sizes can be requested. On APUs, the PASID size
is deterined by the capabilities of the IOMMU. So KFD must be able
to allocate PASIDs in a smaller range.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Make sure vm->root.bo is not left reserved if amdgpu_bo_kmap fails.
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
currently, for CI asics,
use dpm by default, amdgpu.dpm=-1.
when set amdgpu.dpm=1, enable powplay.
when set amdgpu.dpm=0, disable both dpm and powerplay.
when powerplay is stable on CI asics, ci_dpm will
be removed.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
put amd_pm_funcs table in struct powerplay for all
asics.
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>