[AMD Official Use Only - AMD Internal Distribution Only]
connected_to_gpu is legacy design - I understand UMC IP version is also used so it works. Checking is_app_apu flag is good enough at this stage.
It would be better to move the code change to amdgpu_ras_fs_init.
In addition, it's more accurate to say that bad_pae_thread is not applicable to specific app platform, or something like that. Bad bad page feature is supported on the APU, but just GPU driver is not involved.
Regards,
Hawking
-----Original Message-----
From: amd-gfx <amd-gfx-bounces@xxxxxxxxxxxxxxxxxxxxx> On Behalf Of Tao Zhou
Sent: Monday, September 9, 2024 19:05
To: amd-gfx@xxxxxxxxxxxxxxxxxxxxx
Cc: Zhou1, Tao <Tao.Zhou1@xxxxxxx>
Subject: [PATCH] drm/amdgpu: disable RAS bad page feature for specific APU
The feature is unsupported on specific APU.
Signed-off-by: Tao Zhou <tao.zhou1@xxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
index dbfc41ddc3c7..d46f216a33b1 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ras.c
@@ -3679,6 +3679,12 @@ int amdgpu_ras_init(struct amdgpu_device *adev)
amdgpu_ras_init_reserved_vram_size(adev);
+ /* bad page feature is unspported on specific APU */
+ if ((adev->gmc.xgmi.connected_to_cpu ||
+ adev->gmc.is_app_apu) &&
+ amdgpu_ip_version(adev, UMC_HWIP, 0) == IP_VERSION(12, 0, 0))
+ amdgpu_bad_page_threshold = 0;
+
if (amdgpu_ras_fs_init(adev)) {
r = -EINVAL;
goto release_con;
--
2.34.1
