[PATCH 2/2] update check condition of query for ras page retire

[Zhang, Hawking]

[AMD Official Use Only - General]

static ssize_t smu_v13_0_6_get_ecc_info(struct smu_context *smu,
                        void *table)
 {
-       /* Support ecc info by default */
-       return 0;
+       /* we use debug mode flag instead of this interface */
+       return -EOPNOTSUPP;
 }

Shall we just drop the callback implementation? smu_get_ecc_info will return -EOPNOTSUPP if the callback is not supported.

Regards,
Hawking

-----Original Message-----
From: amd-gfx <amd-gfx-bounces at lists.freedesktop.org> On Behalf Of Tao Zhou
Sent: Wednesday, January 17, 2024 17:15
To: amd-gfx at lists.freedesktop.org
Cc: Zhou1, Tao <Tao.Zhou1 at amd.com>
Subject: [PATCH 2/2] update check condition of query for ras page retire

Support page retirement handling in debug mode.

Signed-off-by: Tao Zhou <tao.zhou1 at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c              | 9 +++++++--
 drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c | 4 ++--
 2 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
index 41139bac7643..6df32f0afd89 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_umc.c
@@ -90,12 +90,16 @@ static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev,  {
        struct ras_err_data *err_data = (struct ras_err_data *)ras_error_status;
        struct amdgpu_ras *con = amdgpu_ras_get_context(adev);
+       unsigned int error_query_mode;
        int ret = 0;

+       amdgpu_ras_get_error_query_mode(adev, &error_query_mode);
+
        mutex_lock(&con->page_retirement_lock);

        ret = amdgpu_dpm_get_ecc_info(adev, (void *)&(con->umc_ecc));
-       if (ret == -EOPNOTSUPP) {
+       if (ret == -EOPNOTSUPP &&
+           error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY) {
                if (adev->umc.ras && adev->umc.ras->ras_block.hw_ops &&
                    adev->umc.ras->ras_block.hw_ops->query_ras_error_count)
                    adev->umc.ras->ras_block.hw_ops->query_ras_error_count(adev, ras_error_status); @@ -119,7 +123,8 @@ static void amdgpu_umc_handle_bad_pages(struct amdgpu_device *adev,
                         */
                        adev->umc.ras->ras_block.hw_ops->query_ras_error_address(adev, ras_error_status);
                }
-       } else if (!ret) {
+       } else if (error_query_mode == AMDGPU_RAS_FIRMWARE_ERROR_QUERY ||
+           (!ret && error_query_mode == AMDGPU_RAS_DIRECT_ERROR_QUERY)) {
                if (adev->umc.ras &&
                    adev->umc.ras->ecc_info_query_ras_error_count)
                    adev->umc.ras->ecc_info_query_ras_error_count(adev, ras_error_status); diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
index c560f4af214d..d86c9e7fc64b 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_6_ppt.c
@@ -2909,8 +2909,8 @@ static int smu_v13_0_6_select_xgmi_plpd_policy(struct smu_context *smu,  static ssize_t smu_v13_0_6_get_ecc_info(struct smu_context *smu,
                        void *table)
 {
-       /* Support ecc info by default */
-       return 0;
+       /* we use debug mode flag instead of this interface */
+       return -EOPNOTSUPP;
 }

 static const struct pptable_funcs smu_v13_0_6_ppt_funcs = {
--
2.35.1