From b7dfbd2e601f3fee545bc158feceba4f340fe7cf Mon Sep 17 00:00:00 2001 From: Tushar Patel Date: Thu, 17 Mar 2022 15:31:22 -0400 Subject: drm/amdkfd: Fix Incorrect VMIDs passed to HWS Compute-only GPUs have more than 8 VMIDs allocated to KFD. Fix this by passing correct number of VMIDs to HWS v2: squash in warning fix (Alex) Signed-off-by: Tushar Patel Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index bb1c025d9001..b03663f42cc9 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -680,7 +680,7 @@ MODULE_PARM_DESC(sched_policy, * Maximum number of processes that HWS can schedule concurrently. The maximum is the * number of VMIDs assigned to the HWS, which is also the default. */ -int hws_max_conc_proc = 8; +int hws_max_conc_proc = -1; module_param(hws_max_conc_proc, int, 0444); MODULE_PARM_DESC(hws_max_conc_proc, "Max # processes HWS can execute concurrently when sched_policy=0 (0 = no concurrency, #VMIDs for KFD = Maximum(default))"); -- cgit v1.2.3 From 887f75cfd0da44c19dda93b2ff9e70ca8792cdc1 Mon Sep 17 00:00:00 2001 From: Kai-Heng Feng Date: Thu, 7 Apr 2022 20:12:28 +0800 Subject: drm/amdgpu: Ensure HDA function is suspended before ASIC reset DP/HDMI audio on AMD PRO VII stops working after S3: [ 149.450391] amdgpu 0000:63:00.0: amdgpu: MODE1 reset [ 149.450395] amdgpu 0000:63:00.0: amdgpu: GPU mode1 reset [ 149.450494] amdgpu 0000:63:00.0: amdgpu: GPU psp mode1 reset [ 149.983693] snd_hda_intel 0000:63:00.1: refused to change power state from D0 to D3hot [ 150.003439] amdgpu 0000:63:00.0: refused to change power state from D0 to D3hot ... [ 155.432975] snd_hda_intel 0000:63:00.1: CORB reset timeout#2, CORBRP = 65535 The offending commit is daf8de0874ab5b ("drm/amdgpu: always reset the asic in suspend (v2)"). Commit 34452ac3038a7 ("drm/amdgpu: don't use BACO for reset in S3 ") doesn't help, so the issue is something different. Assuming that to make HDA resume to D0 fully realized, it needs to be successfully put to D3 first. And this guesswork proves working, by moving amdgpu_asic_reset() to noirq callback, so it's called after HDA function is in D3. Fixes: daf8de0874ab5b ("drm/amdgpu: always reset the asic in suspend (v2)") Signed-off-by: Kai-Heng Feng Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) (limited to 'drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c') diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c index b03663f42cc9..29e9419a914b 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c @@ -2323,18 +2323,23 @@ static int amdgpu_pmops_suspend(struct device *dev) { struct drm_device *drm_dev = dev_get_drvdata(dev); struct amdgpu_device *adev = drm_to_adev(drm_dev); - int r; if (amdgpu_acpi_is_s0ix_active(adev)) adev->in_s0ix = true; else adev->in_s3 = true; - r = amdgpu_device_suspend(drm_dev, true); - if (r) - return r; + return amdgpu_device_suspend(drm_dev, true); +} + +static int amdgpu_pmops_suspend_noirq(struct device *dev) +{ + struct drm_device *drm_dev = dev_get_drvdata(dev); + struct amdgpu_device *adev = drm_to_adev(drm_dev); + if (!adev->in_s0ix) - r = amdgpu_asic_reset(adev); - return r; + return amdgpu_asic_reset(adev); + + return 0; } static int amdgpu_pmops_resume(struct device *dev) @@ -2575,6 +2580,7 @@ static const struct dev_pm_ops amdgpu_pm_ops = { .prepare = amdgpu_pmops_prepare, .complete = amdgpu_pmops_complete, .suspend = amdgpu_pmops_suspend, + .suspend_noirq = amdgpu_pmops_suspend_noirq, .resume = amdgpu_pmops_resume, .freeze = amdgpu_pmops_freeze, .thaw = amdgpu_pmops_thaw, -- cgit v1.2.3