refactor(amdsmi): remove driver reload API and related functions

projects/amdsmi/CHANGELOG.md

@@ -31,6 +31,11 @@ Full documentation for amd_smi_lib is available at [https://rocm.docs.amd.com/pr
   - The per-process KFD lookup rebuilt the entire KFD node topology (an expensive sysfs walk) on every call just to translate the device BDF into its KFD GPU id.
   - The caller already knows this value, so it is now passed through to `gpuvsmi_get_pid_info()`, eliminating one full topology discovery per process per refresh. Falls back to the original discovery path when the id is unavailable.
 
+### Removed
+
+- **Removed the `amdsmi_gpu_driver_reload()` API and its Python binding** (breaking).  
+  - Reload the amdgpu driver out of band with `sudo modprobe -r amdgpu && sudo modprobe amdgpu` to apply memory partition changes instead.
+
 ### Resolved Issues
 
 - **Fixed `amd-smi set --ptl-status` silently failing to change PTL state**.  

projects/amdsmi/docs/conceptual/partition.md

@@ -460,8 +460,6 @@ Changing partition settings has strict requirements:
   sudo modprobe -r amdgpu && sudo modprobe amdgpu
   ```
 
-  Alternatively, call `amdsmi_gpu_driver_reload()` from the library.
-
 - A driver reload affects **all GPUs in the hive** -- every GPU in the system is reconfigured
   to the new memory partition configuration at once.
 
@@ -475,7 +473,7 @@ set API automatically triggered an immediate driver reload on invocation; the CL
 reloaded after the user explicitly requested the partition change. The API-level reload was
 separated to give applications control over when the disruptive reload occurs. Additionally,
 as of [**ROCm 7.13.0**](https://github.com/ROCm/rocm-systems/blob/develop/projects/amdsmi/CHANGELOG.md#amd_smi_lib-for-rocm-7130), `amd-smi reset -r` is no longer available for driver reloading — use
-`sudo modprobe -r amdgpu && sudo modprobe amdgpu` or `amdsmi_gpu_driver_reload()` instead.
+`sudo modprobe -r amdgpu && sudo modprobe amdgpu` instead.
 ```
 
 ## Workload isolation and assignment
@@ -695,7 +693,8 @@ int main() {
 
     // Step 4: Reload the driver -- required to apply the memory partition change.
     // Stop all GPU workloads first. The reload may reset the accelerator partition.
-    amdsmi_gpu_driver_reload();
+    // The reload needs to occur out of band via calls to `modprobe -r amdgpu` and
+    // then `modprobe amdgpu`
 
     // Step 5: Re-initialize to pull in the updated topology (new device count/handles)
     amdsmi_shut_down();
@@ -731,7 +730,6 @@ APIs, see
   index (obtained from {c:func}`amdsmi_get_gpu_accelerator_partition_profile_config`).
 - {c:func}`amdsmi_get_gpu_memory_partition_config` -- Query the current NPS mode and supported NPS modes.
 - {c:func}`amdsmi_set_gpu_memory_partition_mode` -- Set the NPS memory partition mode.
-- {c:func}`amdsmi_gpu_driver_reload` -- Reload the amdgpu driver to apply memory partition changes.
 
 **Bare metal only:**
 - {c:func}`amdsmi_get_gpu_compute_partition` -- Query the current compute partition setting as a string.
@@ -769,7 +767,8 @@ amdsmi.amdsmi_set_gpu_memory_partition_mode(gpu, amdsmi.AmdSmiMemoryPartitionTyp
 
 # Step 4: Reload the driver -- required to apply the memory partition change.
 # Stop all GPU workloads first. The reload may reset the accelerator partition.
-amdsmi.amdsmi_gpu_driver_reload()
+# The reload needs to occur out of band via calls to `modprobe -r amdgpu` and
+# then `modprobe amdgpu`
 
 # Step 5: Re-initialize to pull in the updated topology (new device count/handles)
 amdsmi.amdsmi_shut_down()

projects/amdsmi/example/amd_smi_drm_example.cc

@@ -742,18 +742,9 @@ int main() {
             std::cout << "\tamdsmi_set_gpu_memory_partition(" << gpu_number << ", "
                       << memoryPartitionString(updatePartition) << "): " << err_str << "\n\n";
 
-            // Reload only if the memory partition was set successfully
             if (ret_set == AMDSMI_STATUS_SUCCESS) {
-              std::cout << "\t**Reloading GPU driver to apply memory "
-                        << "partition change, this may take some time... **\n";
-              amdsmi_status_t reload_status = amdsmi_gpu_driver_reload();
-              amdsmi_status_code_to_string(reload_status, &err_str);
-              if (reload_status == AMDSMI_STATUS_SUCCESS) {
-                PRINT_AMDSMI_RET(reload_status)
-                std::cout << "\tamdsmi_gpu_driver_reload(): " << err_str << "\n\n";
-              } else {
-                std::cout << "\tamdsmi_gpu_driver_reload(): " << err_str << "\n\n";
-              }
+              std::cout << "\t** Memory partition staged. Run to apply:\n"
+                        << "\t   sudo modprobe -r amdgpu && sudo modprobe amdgpu\n\n";
             }
 
             // Get the current memory partition
@@ -841,18 +832,9 @@ int main() {
         }
         std::cout << "\tamdsmi_set_gpu_memory_partition(" << gpu_number << ", "
                   << memoryPartitionString(orig_partition) << "): " << err_str << "\n\n";
-        // Reload only if the memory partition was set successfully
         if (ret_set == AMDSMI_STATUS_SUCCESS) {
-          std::cout << "\t**Reloading GPU driver to apply memory "
-                    << "partition change, this may take some time... **\n";
-          amdsmi_status_t reload_status = amdsmi_gpu_driver_reload();
-          amdsmi_status_code_to_string(reload_status, &err_str);
-          if (reload_status == AMDSMI_STATUS_SUCCESS) {
-            PRINT_AMDSMI_RET(reload_status)
-            std::cout << "\tamdsmi_gpu_driver_reload(): " << err_str << "\n\n";
-          } else {
-            std::cout << "\tamdsmi_gpu_driver_reload(): " << err_str << "\n\n";
-          }
+          std::cout << "\t** Memory partition staged. Run to apply:\n"
+                    << "\t   sudo modprobe -r amdgpu && sudo modprobe amdgpu\n\n";
         }
         // Get the current memory partition
         char current_memory_partition[AMDSMI_MAX_STRING_LENGTH];

projects/amdsmi/example/amd_smi_partition_example.cc

@@ -41,7 +41,7 @@
 // invalid and must not be used.
 //
 //   Memory partition (NPS mode):
-//     Requires amdsmi_gpu_driver_reload() to take effect. The driver reload
+//     Requires driver reload to take effect. The driver reload
 //     tears down and rebuilds all kernel device objects, which also destroys
 //     any existing handles. amdsmi_shut_down() + amdsmi_init() is therefore
 //     mandatory before querying or changing anything further.
@@ -363,18 +363,6 @@ static void print_available_partition_modes(const std::vector<amdsmi_processor_h
   return ret == AMDSMI_STATUS_SUCCESS;
 }
 
-// Returns true if the driver was reloaded successfully.
-[[nodiscard]] static bool reload_driver() {
-  print_separator("Reload driver");
-  // Mandatory to apply memory partition change. All GPU activity must be
-  // stopped first. The reload may reset the accelerator partition to default.
-  // Root privileges (sudo) are required for driver reload.
-  std::cout << "  Reloading driver, this may take some time...\n";
-  auto ret = amdsmi_gpu_driver_reload();
-  std::cout << "  amdsmi_gpu_driver_reload: " << status_str(ret) << '\n';
-  return ret == AMDSMI_STATUS_SUCCESS;
-}
-
 // For each GPU: build the NPS→profiles map, print every entry, then demonstrate
 // iterating over only the profiles compatible with the currently active NPS mode.
 static void print_profiles_by_nps(const std::vector<amdsmi_processor_handle>& gpus) {
@@ -514,9 +502,8 @@ int main() {
     } else {
       mem_changed = set_memory_partition(gpus.front(), target_memory_partition);
       if (mem_changed) {
-        if (!reload_driver())
-          std::cout << "[warn] Driver reload failed; memory partition change may not have "
-                       "taken effect.\n";
+        std::cout << "  Memory partition staged. Run to apply:\n"
+                  << "    sudo modprobe -r amdgpu && sudo modprobe amdgpu\n";
       } else {
         std::cout << "\n[info] Memory partition unchanged; skipping driver reload.\n";
       }

projects/amdsmi/example/amd_smi_partition_example.py

@@ -29,7 +29,7 @@
 # invalid and must not be used.
 #
 #   Memory partition (NPS mode):
-#     Requires amdsmi_gpu_driver_reload() to take effect. The driver reload
+#     Requires driver reload to take effect. The driver reload
 #     tears down and rebuilds all kernel device objects, which also destroys
 #     any existing handles. amdsmi_shut_down() + amdsmi_init() is therefore
 #     mandatory before querying or changing anything further.
@@ -244,22 +244,6 @@ def set_memory_partition(gpu0: Any, target: amdsmi.AmdSmiMemoryPartitionType) ->
         return False
 
 
-# Returns True if the driver was reloaded successfully.
-def reload_driver() -> bool:
-    print_separator("Reload driver")
-    # Mandatory to apply memory partition change. All GPU activity must be
-    # stopped first. The reload may reset the accelerator partition to default.
-    # Root privileges (sudo) are required for driver reload.
-    print("  Reloading driver, this may take some time...")
-    try:
-        amdsmi.amdsmi_gpu_driver_reload()
-        print("  amdsmi_gpu_driver_reload: success")
-        return True
-    except amdsmi.AmdSmiException as e:
-        print(f"  amdsmi_gpu_driver_reload: {e}")
-        return False
-
-
 # For each GPU: build the NPS->profiles map, print every entry, then demonstrate
 # iterating over only the profiles compatible with the currently active NPS mode.
 def print_profiles_by_nps(gpus: List) -> None:
@@ -388,11 +372,10 @@ def main() -> None:
         else:
             mem_changed = set_memory_partition(gpus[0], target_memory_partition)
             if mem_changed:
-                if not reload_driver():
-                    print(
-                        "[warn] Driver reload failed; memory partition change "
-                        "may not have taken effect."
-                    )
+                print(
+                    "  Memory partition staged. Run to apply:\n"
+                    "    sudo modprobe -r amdgpu && sudo modprobe amdgpu"
+                )
             else:
                 print("\n[info] Memory partition unchanged; skipping driver reload.")
 

projects/amdsmi/include/amd_smi/amdsmi.h

@@ -7112,7 +7112,7 @@ amdsmi_status_t amdsmi_get_gpu_memory_partition(amdsmi_processor_handle processo
  *  Device must be idle and have no workloads when performing set partition operations.
  *
  *  On @platform{gpu_bm_linux} AMDGPU driver restart is REQUIRED to complete updating to
- *  the new memory partition setting. Refer to `amdsmi_gpu_driver_reload()` for more details.
+ *  the new memory partition setting.
  *
  *  @param[in] processor_handle Device which to query
  *
@@ -7163,10 +7163,10 @@ amdsmi_status_t amdsmi_get_gpu_memory_partition_config(amdsmi_processor_handle p
  *  Device must be idle and have no workloads when performing set partition operations.
  *
  *  @details On @platform{gpu_bm_linux} AMDGPU driver restart is REQUIRED to complete updating
- *  to the new memory partition setting. Refer to `amdsmi_gpu_driver_reload()` for more details.
+ *  to the new memory partition setting.
  *
  *  On @platform{gpu_bm_linux} AMDGPU driver restart is REQUIRED to complete updating to
- *  the new memory partition setting. Refer to `amdsmi_gpu_driver_reload()` for more details.
+ *  the new memory partition setting.
  *
  *  @param[in] processor_handle A processor handle
  *
@@ -7859,62 +7859,6 @@ amdsmi_status_t amdsmi_get_gpu_process_list_by_pid(amdsmi_processor_handle* proc
 
 /** @} End tagProcessInfo */
 
-/*****************************************************************************/
-/** @defgroup tagDriverControl Driver control mechanisms
- *  These functions provide control over the driver. Users should use with
- *  caution as they may cause the driver to become unstable.
- *  @{
- */
-/**
- *  @brief Restart the device driver (kmod module) for all AMD GPUs on the
- *  system.
- *
- *  @ingroup tagDriverControl
- *
- *  @platform{gpu_bm_linux} @platform{guest_1vf} @platform{guest_mvf}
- *
- *  @details This function will reload the AMD GPU driver as described in
- *  the Linux kernel documentation -
- *  https://docs.kernel.org/admin-guide/sysctl/kernel.html#modprobe
- *  with no extra parameters as specified in
- *  https://docs.kernel.org/gpu/amdgpu/module-parameters.html.
- *
- *  Use this function with caution, as it will unload and reload the AMD GPU
- *  driver: `modprobe -r amdgpu && modprobe amdgpu`.
- *
- *  Any process or workload using the AMD GPU driver is REQUIRED to be
- *  stopped before calling this function. Otherwise, function will return
- *  ::AMDSMI_STATUS_AMDGPU_RESTART_ERR could not successfully restart
- *  the amdgpu driver.
- *
- *  User is REQUIRED to have root/admin privileges to call this function.
- *  Otherwise, this function will return ::AMDSMI_STATUS_NO_PERM.
- *
- *  This API will take time to complete, as we are checking the driver's
- *  loading status to confirm it reloaded properly. If
- *  ::AMDSMI_STATUS_AMDGPU_RESTART_ERR is returned, it means the driver
- *  did not reload properly and the user should check dmesg logs.
- *
- *  This function has been created in order to conveniently reload the
- *  AMD GPU driver once `amdsmi_set_gpu_memory_partition()` or
- *  `amdsmi_set_gpu_memory_partition_mode()` successfully has been changed
- *  on Baremetal systems. Now users can control the reload once all GPU
- *  processes/workloads have been stopped on the AMD GPU driver.
- *  A (AMD GPU) driver reload is REQUIRED to complete changing
- *  to the new memory partition configuration
- *  (`amdsmi_set_gpu_memory_partition()`/`amdsmi_set_gpu_memory_partition_mode()`)
- *  operation MUST be successful. This function WILL EFFECT all GPUs in the
- *  hive to be reconfigured with the specified memory partition configuration.
- *
- *  @return ::amdsmi_status_t | ::AMDSMI_STATUS_SUCCESS on success
- *  @return                   | ::AMDSMI_STATUS_NO_PERM function requires admin/sudo privileges
- *  @return                   | ::AMDSMI_STATUS_AMDGPU_RESTART_ERR could not successfully restart
- *                                the amdgpu driver.
- */
-amdsmi_status_t amdsmi_gpu_driver_reload(void);
-
-/** @} End tagDriverControl */
-
 /*****************************************************************************/
 /** @defgroup tagPTL Peak Tops Limiter
  *  @{

projects/amdsmi/py-interface/__init__.py

@@ -188,7 +188,6 @@
 from .amdsmi_interface import amdsmi_set_gpu_perf_level
 from .amdsmi_interface import amdsmi_get_gpu_power_profile_presets
 from .amdsmi_interface import amdsmi_reset_gpu
-from .amdsmi_interface import amdsmi_gpu_driver_reload
 from .amdsmi_interface import amdsmi_set_gpu_perf_determinism_mode
 from .amdsmi_interface import amdsmi_set_gpu_fan_speed
 from .amdsmi_interface import amdsmi_reset_gpu_fan

projects/amdsmi/py-interface/amdsmi_interface.py

@@ -4774,10 +4774,6 @@ def amdsmi_reset_gpu(processor_handle: processor_handle_t):
     _check_res(amdsmi_wrapper.amdsmi_reset_gpu(processor_handle))
 
 
-def amdsmi_gpu_driver_reload():
-    _check_res(amdsmi_wrapper.amdsmi_gpu_driver_reload())
-
-
 def amdsmi_set_gpu_fan_speed(processor_handle: processor_handle_t, sensor_idx: int, fan_speed: int):
     if not isinstance(processor_handle, amdsmi_wrapper.amdsmi_processor_handle):
         raise AmdSmiParameterException(processor_handle, amdsmi_wrapper.amdsmi_processor_handle)

projects/amdsmi/py-interface/amdsmi_wrapper.py

@@ -4252,12 +4252,6 @@ class struct_amdsmi_cper_hdr_t(Structure):
     amdsmi_get_gpu_process_list_by_pid.argtypes = [ctypes.POINTER(ctypes.POINTER(None)), uint32_t, ctypes.POINTER(struct_amdsmi_proc_info_by_pid_t), ctypes.POINTER(ctypes.c_uint32)]
 except AttributeError:
     pass
-try:
-    amdsmi_gpu_driver_reload = _libraries['libamd_smi.so'].amdsmi_gpu_driver_reload
-    amdsmi_gpu_driver_reload.restype = amdsmi_status_t
-    amdsmi_gpu_driver_reload.argtypes = []
-except AttributeError:
-    pass
 try:
     amdsmi_get_gpu_ptl_state = _libraries['libamd_smi.so'].amdsmi_get_gpu_ptl_state
     amdsmi_get_gpu_ptl_state.restype = amdsmi_status_t
@@ -5309,15 +5303,14 @@ class struct_amdsmi_ttm_info_t(Structure):
     'amdsmi_gpu_block_t', 'amdsmi_gpu_cache_info_t',
     'amdsmi_gpu_control_counter',
     'amdsmi_gpu_counter_group_supported', 'amdsmi_gpu_create_counter',
-    'amdsmi_gpu_destroy_counter', 'amdsmi_gpu_driver_reload',
-    'amdsmi_gpu_metrics_t', 'amdsmi_gpu_ras_policy_info_t',
-    'amdsmi_gpu_ras_policy_v4_0_t', 'amdsmi_gpu_read_counter',
-    'amdsmi_gpu_validate_ras_eeprom', 'amdsmi_gpu_xcp_metrics_t',
-    'amdsmi_gpu_xgmi_error_status', 'amdsmi_hsmp_driver_version_t',
-    'amdsmi_hsmp_freqlimit_src_names', 'amdsmi_hsmp_metrics_table_t',
-    'amdsmi_init', 'amdsmi_init_flags_t',
-    'amdsmi_init_gpu_event_notification', 'amdsmi_io_bw_encoding_t',
-    'amdsmi_is_P2P_accessible',
+    'amdsmi_gpu_destroy_counter', 'amdsmi_gpu_metrics_t',
+    'amdsmi_gpu_ras_policy_info_t', 'amdsmi_gpu_ras_policy_v4_0_t',
+    'amdsmi_gpu_read_counter', 'amdsmi_gpu_validate_ras_eeprom',
+    'amdsmi_gpu_xcp_metrics_t', 'amdsmi_gpu_xgmi_error_status',
+    'amdsmi_hsmp_driver_version_t', 'amdsmi_hsmp_freqlimit_src_names',
+    'amdsmi_hsmp_metrics_table_t', 'amdsmi_init',
+    'amdsmi_init_flags_t', 'amdsmi_init_gpu_event_notification',
+    'amdsmi_io_bw_encoding_t', 'amdsmi_is_P2P_accessible',
     'amdsmi_is_gpu_power_management_enabled', 'amdsmi_kfd_info_t',
     'amdsmi_link_id_bw_type_t', 'amdsmi_link_metrics_t',
     'amdsmi_link_status_t', 'amdsmi_link_type_t',

projects/amdsmi/rocm_smi/include/rocm_smi/rocm_smi.h

@@ -5929,56 +5929,6 @@ rsmi_status_t rsmi_dev_metrics_log_get(uint32_t dv_ind);
 
 /** @} */  // end of DevMetricsHeaderInfoGet
 
-/*****************************************************************************/
-/** @defgroup DriverControl Driver control mechanisms
- *  These functions provide control over the driver. Users should use with
- *  caution as they may cause the driver to become unstable.
- *  @{
- */
-/**
- *  @brief Restart the device driver (kmod module) for all AMD GPUs on the
- *  system.
- *
- *  @details This function will reload the AMD GPU driver as described in
- *  the Linux kernel documentation -
- *  https://docs.kernel.org/admin-guide/sysctl/kernel.html#modprobe
- *  with no extra parameters as specified in
- *  https://docs.kernel.org/gpu/amdgpu/module-parameters.html.
- *
- *  Use this function with caution, as it will unload and reload the AMD GPU
- *  driver: `modprobe -r amdgpu && modprobe amdgpu`.
- *
- *  Any process or workload using the AMD GPU driver is REQUIRED to be
- *  stopped before calling this function. Otherwise, function will return
- *  ::RSMI_STATUS_AMDGPU_RESTART_ERR could not successfully restart
- *  the amdgpu driver.
- *
- *  User is REQUIRED to have root/admin privileges to call this function.
- *  Otherwise, this function will return ::RSMI_STATUS_PERMISSION.
- *
- *  This API will take time to complete, as we are checking the driver's
- *  loading status to confirm it reloaded properly. If
- *  ::RSMI_STATUS_AMDGPU_RESTART_ERR is returned, it means the driver
- *  did not reload properly and the user should check dmesg logs.
- *
- *  This function has been created in order to conveniently reload the
- *  AMD GPU driver once `rsmi_dev_memory_partition_set()`
- *  successfully has been changed on Baremetal systems.
- *  Now users can control the reload once all GPU processes/workloads
- *  have been stopped on the AMD GPU driver. A (AMD GPU) driver reload
- *  is REQUIRED to complete changing to the new memory partition
- *  configuration (`rsmi_dev_memory_partition_set()`) operation MUST
- *  be successful. This function WILL EFFECT all GPUs in the hive to
- *  be reconfigured with the specified memory partition configuration.
- *
- *  @retval ::RSMI_STATUS_SUCCESS call was successful
- *  @retval ::RSMI_STATUS_PERMISSION function requires root access
- *  @retval ::RSMI_STATUS_AMDGPU_RESTART_ERR could not successfully restart
- *            the amdgpu driver.
- */
-rsmi_status_t rsmi_dev_amdgpu_driver_reload(void);
-/** @} */  // end of DriverControl
-
 #ifdef __cplusplus
 }
 #endif  // __cplusplus

projects/amdsmi/rocm_smi/include/rocm_smi/rocm_smi_device.h

@@ -253,8 +253,6 @@ class Device {
   void fillSupportedFuncs(void);
   void DumpSupportedFunctions(void);
   bool DeviceAPISupported(std::string name, uint64_t variant, uint64_t sub_variant);
-  rsmi_status_t restartAMDGpuDriver(void);
-  rsmi_status_t isRestartInProgress(bool* isRestartInProgress, bool* isAMDGPUModuleLive);
   rsmi_status_t storeDevicePartitions(uint32_t dv_ind);
   template <typename T>
   std::string readBootPartitionState(uint32_t dv_ind);