Quiet address sanitizer warnings

Also, * Fix some doxygen issues * Fix address sanitizer issues in rsmitst Change-Id: Ie6c6fd9af5c418210b7064e79650fb92cd4a5e2b [ROCm/amdsmi commit: 63064b0000]
2020-11-10 14:16:39 -06:00
1 Parent fa475d307a
@@ -311,7 +311,7 @@ typedef enum {
  RSMI_EVT_NOTIF_LAST = RSMI_EVT_NOTIF_GPU_POST_RESET
 } rsmi_evt_notification_type_t;

-/*
+/**
 * Macro to generate event bitmask from event id
 */
 #define RSMI_EVENT_MASK_FROM_INDEX(i) (1ULL << ((i) - 1))
@@ -786,14 +786,25 @@ typedef rsmi_od_volt_freq_data_t rsmi_od_volt_freq_data;
 * @brief The following structures hold the gpu metrics values for a device.
 */

-struct metrics_table_header {
+/**
+ * @brief Size and version information of metrics data
+ */
+struct metrics_table_header_t {
+  // TODO(amd) Doxygen documents
+  /// \cond Ignore in docs.
  uint16_t      structure_size;
  uint8_t       format_revision;
  uint8_t       content_revision;
+  /// \endcond
 };

+/**
+ * @brief The following structure holds the gpu metrics values for a device.
+ */
 typedef struct {
-  struct metrics_table_header common_header;
+  // TODO(amd) Doxygen documents
+  /// \cond Ignore in docs.
+  struct metrics_table_header_t common_header;

 /* Driver attached timestamp (in ns) */
  uint64_t      system_clock_counter;
@@ -842,10 +853,8 @@ typedef struct {
 /* Link width/speed */
  uint8_t       pcie_link_width;
  uint8_t       pcie_link_speed;  // in 0.1 GT/s
-}rsmi_gpu_metrics_t;
-/// \cond Ignore in docs.
-typedef rsmi_gpu_metrics_t rsmi_gpu_metrics;
-/// \endcond
+  /// \endcond
+} rsmi_gpu_metrics_t;

 /**
 * @brief This structure holds error counts.
@@ -1267,8 +1276,8 @@ rsmi_dev_subsystem_name_get(uint32_t dv_ind, char *name, size_t len);
 *  @param[inout] minor a pointer to a uint32_t into which minor number will
 *  be copied
 *
- *  @retval :: RSMI_STATUS_SUCCESS is returned upon successful call.
- *  @retval :: RSMI_STATUS_INIT_ERROR if failed to get minor number during
+ *  @retval ::RSMI_STATUS_SUCCESS is returned upon successful call.
+ *  @retval ::RSMI_STATUS_INIT_ERROR if failed to get minor number during
 *  initialization.
 *  @retval ::RSMI_STATUS_INVALID_ARGS the provided arguments are not valid
 *
@@ -1910,7 +1919,6 @@ rsmi_status_t rsmi_dev_fan_speed_max_get(uint32_t dv_ind,
 */
 rsmi_status_t rsmi_dev_temp_metric_get(uint32_t dv_ind, uint32_t sensor_type,
                      rsmi_temperature_metric_t metric, int64_t *temperature);
-/** @} */  // end of PhysQuer

 /**
 *  @brief Get the voltage metric value for the specified metric, from the
@@ -2297,6 +2305,9 @@ rsmi_dev_power_profile_presets_get(uint32_t dv_ind, uint32_t sensor_ind,
 *  @brief Set the PowerPlay performance level associated with the device with
 *  provided device index with the provided value.
 *
+ *  @deprecated ::rsmi_dev_perf_level_set_v1() is preferred, with an
+ *  interface that more closely  matches the rest of the rocm_smi API.
+ *
 *  @details Given a device index @p dv_ind and an ::rsmi_dev_perf_level_t @p
 *  perf_level, this function will set the PowerPlay performance level for the
 *  device to the value @p perf_lvl.
@@ -2314,6 +2325,72 @@ rsmi_dev_power_profile_presets_get(uint32_t dv_ind, uint32_t sensor_ind,
 rsmi_status_t
 rsmi_dev_perf_level_set(int32_t dv_ind, rsmi_dev_perf_level_t perf_lvl);

+/**
+ *  @brief Set the PowerPlay performance level associated with the device with
+ *  provided device index with the provided value.
+ *
+ *  @details Given a device index @p dv_ind and an ::rsmi_dev_perf_level_t @p
+ *  perf_level, this function will set the PowerPlay performance level for the
+ *  device to the value @p perf_lvl.
+ *
+ *  @param[in] dv_ind a device index
+ *
+ *  @param[in] perf_lvl the value to which the performance level should be set
+ *
+ *  @retval ::RSMI_STATUS_SUCCESS is returned upon successful call.
+ *  @retval ::RSMI_STATUS_NOT_SUPPORTED installed software or hardware does not
+ *  support this function with the given arguments
+ *  @retval ::RSMI_STATUS_PERMISSION function requires root access
+ *
+ */
+rsmi_status_t
+rsmi_dev_perf_level_set_v1(uint32_t dv_ind, rsmi_dev_perf_level_t perf_lvl);
+
+/**
+ *  @brief Set the overdrive percent associated with the device with provided
+ *  device index with the provided value. See details for WARNING.
+ *
+ *  @deprecated This function is deprecated. ::rsmi_dev_overdrive_level_set_v1
+ *  has the same functionaltiy, with an interface that more closely
+ *  matches the rest of the rocm_smi API.
+ *
+ *  @details Given a device index @p dv_ind and an overdrive level @p od,
+ *  this function will set the overdrive level for the device to the value
+ *  @p od. The overdrive level is an integer value between 0 and 20, inclusive,
+ *  which represents the overdrive percentage; e.g., a value of 5 specifies
+ *  an overclocking of 5%.
+ *
+ *  The overdrive level is specific to the gpu system clock.
+ *
+ *  The overdrive level is the percentage above the maximum Performance Level
+ *  to which overclocking will be limited. The overclocking percentage does
+ *  not apply to clock speeds other than the maximum. This percentage is
+ *  limited to 20%.
+ *
+ *   ******WARNING******
+ *  Operating your AMD GPU outside of official AMD specifications or outside of
+ *  factory settings, including but not limited to the conducting of
+ *  overclocking (including use of this overclocking software, even if such
+ *  software has been directly or indirectly provided by AMD or otherwise
+ *  affiliated in any way with AMD), may cause damage to your AMD GPU, system
+ *  components and/or result in system failure, as well as cause other problems.
+ *  DAMAGES CAUSED BY USE OF YOUR AMD GPU OUTSIDE OF OFFICIAL AMD SPECIFICATIONS
+ *  OR OUTSIDE OF FACTORY SETTINGS ARE NOT COVERED UNDER ANY AMD PRODUCT
+ *  WARRANTY AND MAY NOT BE COVERED BY YOUR BOARD OR SYSTEM MANUFACTURER'S
+ *  WARRANTY. Please use this utility with caution.
+ *
+ *  @param[in] dv_ind a device index
+ *
+ *  @param[in] od the value to which the overdrive level should be set
+ *
+ *  @retval ::RSMI_STATUS_SUCCESS call was successful
+ *  @retval ::RSMI_STATUS_NOT_SUPPORTED installed software or hardware does not
+ *  support this function with the given arguments
+ *  @retval ::RSMI_STATUS_PERMISSION function requires root access
+ *
+ */
+rsmi_status_t rsmi_dev_overdrive_level_set(int32_t dv_ind, uint32_t od);
+
 /**
 *  @brief Set the overdrive percent associated with the device with provided
 *  device index with the provided value. See details for WARNING.
@@ -2353,7 +2430,7 @@ rsmi_dev_perf_level_set(int32_t dv_ind, rsmi_dev_perf_level_t perf_lvl);
 *  @retval ::RSMI_STATUS_PERMISSION function requires root access
 *
 */
-rsmi_status_t rsmi_dev_overdrive_level_set(int32_t dv_ind, uint32_t od);
+rsmi_status_t rsmi_dev_overdrive_level_set_v1(uint32_t dv_ind, uint32_t od);

 /**
 * @brief Control the set of allowed frequencies that can be used for the
@@ -92,9 +92,9 @@ class Event {
    explicit Event(rsmi_event_type_t event, uint32_t dev_ind);
    ~Event(void);

-    uint32_t openPerfHandle();
-    uint32_t startCounter(void);
-    uint32_t stopCounter(void);
+    int32_t openPerfHandle();
+    int32_t startCounter(void);
+    int32_t stopCounter(void);
    uint32_t getValue(rsmi_counter_value_t *val);
    uint32_t dev_file_ind(void) const {return dev_file_ind_;}
    uint32_t dev_ind(void) const {return dev_ind_;}
@@ -111,8 +111,8 @@ class Event {
    int32_t fd_;
    perf_event_attr attr_;

-    uint32_t get_event_file_info(void);
-    uint32_t get_event_type(uint32_t *ev_type);
+    int32_t get_event_file_info(void);
+    int32_t get_event_type(uint32_t *ev_type);
 };


@@ -199,6 +199,8 @@ class Device {
    void set_evt_notif_anon_file_ptr(FILE *f) {evt_notif_anon_file_ptr_ = f;}
    FILE *evt_notif_anon_file_ptr(void) const {return evt_notif_anon_file_ptr_;}
    void set_evt_notif_anon_fd(int fd) {evt_notif_anon_fd_ = fd;}
+    void set_evt_notif_anon_fd(uint32_t fd) {
+                                   evt_notif_anon_fd_ = static_cast<int>(fd);}
    int evt_notif_anon_fd(void) const {return evt_notif_anon_fd_;}
    void fillSupportedFuncs(void);
    void DumpSupportedFunctions(void);
@@ -78,7 +78,7 @@ class RocmSMI {
    std::vector<std::shared_ptr<amd::smi::Device>>&
                                  monitor_devices() {return monitor_devices_;}
    uint32_t DiscoverAmdgpuDevices(void);
-    uint32_t DiscoverAMDPowerMonitors(bool force_update = false);
+    int DiscoverAMDPowerMonitors(bool force_update = false);

    // Will execute "func" for every Device object known about, or until func
    // returns non-zero;
@@ -101,16 +101,16 @@ class Monitor {
    const std::string path(void) const {return path_;}
    int readMonitor(MonitorTypes type, uint32_t sensor_ind, std::string *val);
    int writeMonitor(MonitorTypes type, uint32_t sensor_ind, std::string val);
-    uint32_t setTempSensorLabelMap(void);
+    int32_t setTempSensorLabelMap(void);
    uint32_t getTempSensorIndex(rsmi_temperature_type_t type);
    rsmi_temperature_type_t getTempSensorEnum(uint64_t ind);
-    uint32_t setVoltSensorLabelMap(void);
+    int32_t setVoltSensorLabelMap(void);
    uint32_t getVoltSensorIndex(rsmi_voltage_type_t type);
    rsmi_voltage_type_t getVoltSensorEnum(uint64_t ind);
    void fillSupportedFuncs(SupportedFuncMap *supported_funcs);

 private:
-    std::string MakeMonitorPath(MonitorTypes type, int32_t sensor_id);
+    std::string MakeMonitorPath(MonitorTypes type, uint32_t sensor_id);
    std::string path_;
    const RocmSMI_env_vars *env_;
    std::map<rsmi_temperature_type_t, uint32_t> temp_type_index_map_;
@@ -83,7 +83,7 @@ GetDevValueVec(amd::smi::DevInfoTypes type,
 rsmi_status_t
 GetDevBinaryVec(amd::smi::DevInfoTypes type,
                         uint32_t dv_ind, std::vector<unsigned char> *val_vec);
-rsmi_status_t ErrnoToRsmiStatus(uint32_t err);
+rsmi_status_t ErrnoToRsmiStatus(int err);

 struct pthread_wrap {
 public: