diff options
| author | Felix Kuehling <Felix.Kuehling@amd.com> | 2017-09-21 22:26:41 +0200 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2017-09-28 22:03:30 +0200 |
| commit | c98171ccf6580407d07a3b5dc8188ce9e1f4f7ca (patch) | |
| tree | 191e13e83451ea87c6e4fdc758425b046a0c4bf8 /drivers/gpu/drm/amd/amdgpu/iceland_ih.c | |
| parent | drm/amd/chash: Fix typo (diff) | |
| download | linux-c98171ccf6580407d07a3b5dc8188ce9e1f4f7ca.tar.xz linux-c98171ccf6580407d07a3b5dc8188ce9e1f4f7ca.zip | |
drm/amdgpu: Handle GPUVM fault storms
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>
Diffstat (limited to 'drivers/gpu/drm/amd/amdgpu/iceland_ih.c')
| -rw-r--r-- | drivers/gpu/drm/amd/amdgpu/iceland_ih.c | 19 |
1 files changed, 17 insertions, 2 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c index 65ed6d3a8f05..bd592cb39f37 100644 --- a/drivers/gpu/drm/amd/amdgpu/iceland_ih.c +++ b/drivers/gpu/drm/amd/amdgpu/iceland_ih.c @@ -216,8 +216,23 @@ static u32 iceland_ih_get_wptr(struct amdgpu_device *adev) */ static bool iceland_ih_prescreen_iv(struct amdgpu_device *adev) { - /* Process all interrupts */ - return true; + u32 ring_index = adev->irq.ih.rptr >> 2; + u16 pasid; + + switch (le32_to_cpu(adev->irq.ih.ring[ring_index]) & 0xff) { + case 146: + case 147: + pasid = le32_to_cpu(adev->irq.ih.ring[ring_index + 2]) >> 16; + if (!pasid || amdgpu_vm_pasid_fault_credit(adev, pasid)) + return true; + break; + default: + /* Not a VM fault */ + return true; + } + + adev->irq.ih.rptr += 16; + return false; } /** |