[PATCH AUTOSEL 5.6 14/38] drm/amdgpu: fix wrong vram lost counter increment V2

From: Sasha Levin
Date: Fri Apr 24 2020 - 08:32:45 EST


From: Evan Quan <evan.quan@xxxxxxx>

[ Upstream commit 028cfb2444b94d4f394a6fa4ca46182481236e91 ]

Vram lost counter is wrongly increased by two during baco reset.

V2: assumed vram lost for mode1 reset on all ASICs

Signed-off-by: Evan Quan <evan.quan@xxxxxxx>
Acked-by: Alex Deucher <alexander.deucher@xxxxxxx>
Signed-off-by: Alex Deucher <alexander.deucher@xxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 20 ++++++++++++++++++--
drivers/gpu/drm/amd/amdgpu/cik.c | 2 --
drivers/gpu/drm/amd/amdgpu/nv.c | 4 ----
drivers/gpu/drm/amd/amdgpu/soc15.c | 4 ----
drivers/gpu/drm/amd/amdgpu/vi.c | 2 --
5 files changed, 18 insertions(+), 14 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
index c8bf9cb3cebf2..f184cdca938de 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
@@ -1953,8 +1953,24 @@ static void amdgpu_device_fill_reset_magic(struct amdgpu_device *adev)
*/
static bool amdgpu_device_check_vram_lost(struct amdgpu_device *adev)
{
- return !!memcmp(adev->gart.ptr, adev->reset_magic,
- AMDGPU_RESET_MAGIC_NUM);
+ if (memcmp(adev->gart.ptr, adev->reset_magic,
+ AMDGPU_RESET_MAGIC_NUM))
+ return true;
+
+ if (!adev->in_gpu_reset)
+ return false;
+
+ /*
+ * For all ASICs with baco/mode1 reset, the VRAM is
+ * always assumed to be lost.
+ */
+ switch (amdgpu_asic_reset_method(adev)) {
+ case AMD_RESET_METHOD_BACO:
+ case AMD_RESET_METHOD_MODE1:
+ return true;
+ default:
+ return false;
+ }
}

/**
diff --git a/drivers/gpu/drm/amd/amdgpu/cik.c b/drivers/gpu/drm/amd/amdgpu/cik.c
index 006f21ef7ddf0..62635e58e45ee 100644
--- a/drivers/gpu/drm/amd/amdgpu/cik.c
+++ b/drivers/gpu/drm/amd/amdgpu/cik.c
@@ -1358,8 +1358,6 @@ static int cik_asic_reset(struct amdgpu_device *adev)
int r;

if (cik_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
r = amdgpu_dpm_baco_reset(adev);
} else {
r = cik_asic_pci_config_reset(adev);
diff --git a/drivers/gpu/drm/amd/amdgpu/nv.c b/drivers/gpu/drm/amd/amdgpu/nv.c
index 2d1bebdf1603d..cc3a79029376b 100644
--- a/drivers/gpu/drm/amd/amdgpu/nv.c
+++ b/drivers/gpu/drm/amd/amdgpu/nv.c
@@ -351,8 +351,6 @@ static int nv_asic_reset(struct amdgpu_device *adev)
struct smu_context *smu = &adev->smu;

if (nv_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
ret = smu_baco_enter(smu);
if (ret)
return ret;
@@ -360,8 +358,6 @@ static int nv_asic_reset(struct amdgpu_device *adev)
if (ret)
return ret;
} else {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
ret = nv_asic_mode1_reset(adev);
}

diff --git a/drivers/gpu/drm/amd/amdgpu/soc15.c b/drivers/gpu/drm/amd/amdgpu/soc15.c
index d8945c31b622c..132a67a041a24 100644
--- a/drivers/gpu/drm/amd/amdgpu/soc15.c
+++ b/drivers/gpu/drm/amd/amdgpu/soc15.c
@@ -569,14 +569,10 @@ static int soc15_asic_reset(struct amdgpu_device *adev)

switch (soc15_asic_reset_method(adev)) {
case AMD_RESET_METHOD_BACO:
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
return soc15_asic_baco_reset(adev);
case AMD_RESET_METHOD_MODE2:
return amdgpu_dpm_mode2_reset(adev);
default:
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
return soc15_asic_mode1_reset(adev);
}
}
diff --git a/drivers/gpu/drm/amd/amdgpu/vi.c b/drivers/gpu/drm/amd/amdgpu/vi.c
index 78b35901643bc..3ce10e05d0d6b 100644
--- a/drivers/gpu/drm/amd/amdgpu/vi.c
+++ b/drivers/gpu/drm/amd/amdgpu/vi.c
@@ -765,8 +765,6 @@ static int vi_asic_reset(struct amdgpu_device *adev)
int r;

if (vi_asic_reset_method(adev) == AMD_RESET_METHOD_BACO) {
- if (!adev->in_suspend)
- amdgpu_inc_vram_lost(adev);
r = amdgpu_dpm_baco_reset(adev);
} else {
r = vi_asic_pci_config_reset(adev);
--
2.20.1