drm/amdgpu: Fix wait for RLCG command completion

if (!(tmp & flag)) condition will always evaluate to true
when the flag is 0x0 (AMDGPU_RLCG_GC_WRITE). Instead check
that address bits are cleared to determine whether
the command is complete.

Signed-off-by: Victor Skvortsov <victor.skvortsov@amd.com>
Tested-by: Bokun Zhang <bokun.zhang@amd.com>
Reviewed by: Shaoyun.liu <Shaoyun.liu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Victor Skvortsov 2022-02-03 21:13:40 +00:00 committed by Alex Deucher
parent fa39f936db
commit aa79d3808e
2 changed files with 3 additions and 1 deletions

View File

@ -902,7 +902,7 @@ static u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v
for (i = 0; i < timeout; i++) {
tmp = readl(scratch_reg1);
if (!(tmp & flag))
if (!(tmp & AMDGPU_RLCG_SCRATCH1_ADDRESS_MASK))
break;
udelay(10);
}

View File

@ -43,6 +43,8 @@
#define AMDGPU_RLCG_WRONG_OPERATION_TYPE 0x2000000
#define AMDGPU_RLCG_REG_NOT_IN_RANGE 0x1000000
#define AMDGPU_RLCG_SCRATCH1_ADDRESS_MASK 0xFFFFF
/* all asic after AI use this offset */
#define mmRCC_IOV_FUNC_IDENTIFIER 0xDE5
/* tonga/fiji use this offset */