[AMD Official Use Only - AMD Internal Distribution Only] Reviewed-by: Asad Kamal <[email protected]>
Thanks & Regards Asad -----Original Message----- From: Lazar, Lijo <[email protected]> Sent: Friday, June 13, 2025 4:37 PM To: [email protected] Cc: Zhang, Hawking <[email protected]>; Deucher, Alexander <[email protected]>; Kamal, Asad <[email protected]> Subject: [PATCH] drm/amdgpu: Extend bus status check to more cases In case of unexpected errors, check if device is alive on the bus. Signed-off-by: Lijo Lazar <[email protected]> --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 15 +++++++++++++++ drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 7 +------ drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c | 5 ++++- drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c | 10 ++++++++++ 4 files changed, 30 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index ea9206ab3ed1..928d28d100f6 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1780,4 +1780,19 @@ extern const struct attribute_group amdgpu_flash_attr_group; void amdgpu_set_init_level(struct amdgpu_device *adev, enum amdgpu_init_lvl_id lvl); + +static inline int amdgpu_device_bus_status_check(struct amdgpu_device +*adev) { + u32 status; + int r; + + r = pci_read_config_dword(adev->pdev, PCI_COMMAND, &status); + if (r || PCI_POSSIBLE_ERROR(status)) { + dev_err(adev->dev, "device lost from bus!"); + return -ENODEV; + } + + return 0; +} + #endif diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 2795a34563f4..a968a9cab27e 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -6066,14 +6066,9 @@ static int amdgpu_device_health_check(struct list_head *device_list_handle) { struct amdgpu_device *tmp_adev; int ret = 0; - u32 status; list_for_each_entry(tmp_adev, device_list_handle, reset_list) { - pci_read_config_dword(tmp_adev->pdev, PCI_COMMAND, &status); - if (PCI_POSSIBLE_ERROR(status)) { - dev_err(tmp_adev->dev, "device lost from bus!"); - ret = -ENODEV; - } + ret |= amdgpu_device_bus_status_check(tmp_adev); } return ret; diff --git a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c index 1c083304ae77..41acc1ce0b3e 100644 --- a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c +++ b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c @@ -353,11 +353,14 @@ static int aqua_vanjaram_query_partition_mode(struct amdgpu_xcp_mgr *xcp_mgr) if (adev->nbio.funcs->get_compute_partition_mode) { mode = adev->nbio.funcs->get_compute_partition_mode(adev); - if (mode != derv_mode) + if (mode != derv_mode) { dev_warn( adev->dev, "Mismatch in compute partition mode - reported : %d derived : %d", mode, derv_mode); + if (derv_mode == AMDGPU_UNKNOWN_COMPUTE_PARTITION_MODE) + amdgpu_device_bus_status_check(adev); + } } return mode; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c index 7eaf58fd7f9a..59f9abd0f7b8 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c @@ -86,6 +86,7 @@ static void smu_cmn_read_arg(struct smu_context *smu, #define SMU_RESP_BUSY_OTHER 0xFC #define SMU_RESP_DEBUG_END 0xFB +#define SMU_RESP_UNEXP (~0U) /** * __smu_cmn_poll_stat -- poll for a status from the SMU * @smu: a pointer to SMU context @@ -171,6 +172,15 @@ static void __smu_cmn_reg_print_error(struct smu_context *smu, dev_err_ratelimited(adev->dev, "SMU: I'm debugging!"); break; + case SMU_RESP_UNEXP: + if (amdgpu_device_bus_status_check(smu->adev)) { + /* print error immediately if device is off the bus */ + dev_err(adev->dev, + "SMU: response:0x%08X for index:%d param:0x%08X message:%s?", + reg_c2pmsg_90, msg_index, param, message); + break; + } + fallthrough; default: dev_err_ratelimited(adev->dev, "SMU: response:0x%08X for index:%d param:0x%08X message:%s?", -- 2.25.1
