[AMD Official Use Only - General]
Are these page faults reported after ("b93ed51c32ca drm/amdgpu: fix AGP init order ") ? Jesse also found page faults in Kfdtest after this commit, and can be fixed by below patch:
[PATCH] drm/amdgpu: exclude domain start when calucales offset for AGP aperture BOs
-----Original Message-----
From: amd-gfx <amd-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of Alex Deucher
Sent: Friday, November 10, 2023 4:51 AM
To: amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: Deucher, Alexander <Alexander.Deucher@xxxxxxx>; Wang, Yang(Kevin) <KevinYang.Wang@xxxxxxx>
Subject: [PATCH 3/5] drm/amdgpu/gmc11: disable AGP aperture
We've had misc reports of random IOMMU page faults when this is used. It's just a rarely used optimization anyway, so let's just disable it. It can still be toggled via the module parameter for testing.
v2: leave it configurable via module parameter
Reviewed-by: Yang Wang <kevinyang.wang@xxxxxxx> (v1)
Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
index e1078b53e942..23d7b548d13f 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v11_0.c
@@ -642,7 +642,7 @@ static void gmc_v11_0_vram_gtt_location(struct amdgpu_device *adev,
amdgpu_gmc_gart_location(adev, mc, AMDGPU_GART_PLACEMENT_HIGH);
if (!amdgpu_sriov_vf(adev) &&
(amdgpu_ip_version(adev, GC_HWIP, 0) < IP_VERSION(11, 5, 0)) &&
- (amdgpu_agp != 0))
+ (amdgpu_agp == 1))
amdgpu_gmc_agp_location(adev, mc);
/* base offset of vram pages */
--
2.41.0
