Merge pull request #326 from pshep/error_counters Error counters
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266
diff --git a/adl.c b/adl.c
index 9a0c9b2..d5e83a9 100644
--- a/adl.c
+++ b/adl.c
@@ -1070,9 +1070,7 @@ static bool fan_autotune(int gpu, int temp, int fanpercent, int lasttemp, bool *
applog(LOG_WARNING, "Overheat detected on GPU %d, increasing fan to 100%", gpu);
newpercent = iMax;
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_DEV_OVER_HEAT;
- cgpu->dev_over_heat_count++;
+ dev_error(cgpu, REASON_DEV_OVER_HEAT);
} else if (temp > ga->targettemp && fanpercent < top && tdiff >= 0) {
applog(LOG_DEBUG, "Temperature over target, increasing fanspeed");
if (temp > ga->targettemp + opt_hysteresis)
@@ -1176,17 +1174,12 @@ void gpu_autotune(int gpu, enum dev_enable *denable)
applog(LOG_WARNING, "Hit thermal cutoff limit on GPU %d, disabling!", gpu);
*denable = DEV_RECOVER;
newengine = ga->minspeed;
-
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_DEV_THERMAL_CUTOFF;
- cgpu->dev_thermal_cutoff_count++;
+ dev_error(cgpu, REASON_DEV_THERMAL_CUTOFF);
} else if (temp > ga->overtemp && engine > ga->minspeed) {
applog(LOG_WARNING, "Overheat detected, decreasing GPU %d clock speed", gpu);
newengine = ga->minspeed;
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_DEV_OVER_HEAT;
- cgpu->dev_over_heat_count++;
+ dev_error(cgpu, REASON_DEV_OVER_HEAT);
} else if (temp > ga->targettemp + opt_hysteresis && engine > ga->minspeed && fan_optimal) {
applog(LOG_DEBUG, "Temperature %d degrees over target, decreasing clock speed", opt_hysteresis);
newengine = engine - ga->lpOdParameters.sEngineClock.iStep;
diff --git a/cgminer.c b/cgminer.c
index 134212c..83159b2 100644
--- a/cgminer.c
+++ b/cgminer.c
@@ -5487,10 +5487,7 @@ void *miner_thread(void *userdata)
gettimeofday(&getwork_start, NULL);
if (api->thread_init && !api->thread_init(mythr)) {
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_THREAD_FAIL_INIT;
- cgpu->thread_fail_init_count++;
-
+ dev_error(cgpu, REASON_THREAD_FAIL_INIT);
goto out;
}
@@ -5561,11 +5558,7 @@ void *miner_thread(void *userdata)
if (unlikely(hashes == -1)) {
applog(LOG_ERR, "%s %d failure, disabling!", api->name, cgpu->device_id);
cgpu->deven = DEV_DISABLED;
-
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_THREAD_ZERO_HASH;
- cgpu->thread_zero_hash_count++;
-
+ dev_error(cgpu, REASON_THREAD_ZERO_HASH);
mt_disable(mythr, thr_id, api);
}
@@ -6130,9 +6123,7 @@ static void *watchdog_thread(void __maybe_unused *userdata)
applog(LOG_ERR, "%s: Idle for more than 60 seconds, declaring SICK!", dev_str);
gettimeofday(&thr->sick, NULL);
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_DEV_SICK_IDLE_60;
- cgpu->dev_sick_idle_60_count++;
+ dev_error(cgpu, REASON_DEV_SICK_IDLE_60);
#ifdef HAVE_ADL
if (adl_active && cgpu->has_adl && gpu_activity(gpu) > 50) {
applog(LOG_ERR, "GPU still showing activity suggesting a hard hang.");
@@ -6148,9 +6139,7 @@ static void *watchdog_thread(void __maybe_unused *userdata)
applog(LOG_ERR, "%s: Not responded for more than 10 minutes, declaring DEAD!", dev_str);
gettimeofday(&thr->sick, NULL);
- cgpu->device_last_not_well = time(NULL);
- cgpu->device_not_well_reason = REASON_DEV_DEAD_IDLE_600;
- cgpu->dev_dead_idle_600_count++;
+ dev_error(cgpu, REASON_DEV_DEAD_IDLE_600);
} else if (now.tv_sec - thr->sick.tv_sec > 60 &&
(cgpu->status == LIFE_SICK || cgpu->status == LIFE_DEAD)) {
/* Attempt to restart a GPU that's sick or dead once every minute */
diff --git a/driver-bitforce.c b/driver-bitforce.c
index ae9fa01..4cb1721 100644
--- a/driver-bitforce.c
+++ b/driver-bitforce.c
@@ -420,10 +420,7 @@ static bool bitforce_get_temp(struct cgpu_info *bitforce)
if (unlikely(bitforce->cutofftemp > 0 && temp > bitforce->cutofftemp)) {
applog(LOG_WARNING, "BFL%i: Hit thermal cutoff limit, disabling!", bitforce->device_id);
bitforce->deven = DEV_RECOVER;
-
- bitforce->device_last_not_well = time(NULL);
- bitforce->device_not_well_reason = REASON_DEV_THERMAL_CUTOFF;
- bitforce->dev_thermal_cutoff_count++;
+ dev_error(bitforce, REASON_DEV_THERMAL_CUTOFF);
}
}
} else {
@@ -431,9 +428,7 @@ static bool bitforce_get_temp(struct cgpu_info *bitforce)
* our responses are out of sync and flush the buffer to
* hopefully recover */
applog(LOG_WARNING, "BFL%i: Garbled response probably throttling, clearing buffer", bitforce->device_id);
- bitforce->device_last_not_well = time(NULL);
- bitforce->device_not_well_reason = REASON_DEV_THROTTLE;
- bitforce->dev_throttle_count++;
+ dev_error(bitforce, REASON_DEV_THROTTLE);
/* Count throttling episodes as hardware errors */
bitforce->hw_errors++;
bitforce_clear_buffer(bitforce);
@@ -568,9 +563,7 @@ static int64_t bitforce_get_result(struct thr_info *thr, struct work *work)
if (elapsed.tv_sec > BITFORCE_TIMEOUT_S) {
applog(LOG_ERR, "BFL%i: took %dms - longer than %dms", bitforce->device_id,
tv_to_ms(elapsed), BITFORCE_TIMEOUT_MS);
- bitforce->device_last_not_well = time(NULL);
- bitforce->device_not_well_reason = REASON_DEV_OVER_HEAT;
- bitforce->dev_over_heat_count++;
+ dev_error(bitforce, REASON_DEV_OVER_HEAT);
if (!pdevbuf[0]) /* Only return if we got nothing after timeout - there still may be results */
return 0;
@@ -673,9 +666,7 @@ static int64_t bitforce_scanhash(struct thr_info *thr, struct work *work, int64_
if (ret == -1) {
ret = 0;
applog(LOG_ERR, "BFL%i: Comms error", bitforce->device_id);
- bitforce->device_last_not_well = time(NULL);
- bitforce->device_not_well_reason = REASON_DEV_COMMS_ERROR;
- bitforce->dev_comms_error_count++;
+ dev_error(bitforce, REASON_DEV_COMMS_ERROR);
bitforce->hw_errors++;
/* empty read buffer */
bitforce_clear_buffer(bitforce);
diff --git a/driver-icarus.c b/driver-icarus.c
index c3adafa..c979a68 100644
--- a/driver-icarus.c
+++ b/driver-icarus.c
@@ -669,9 +669,7 @@ static int64_t icarus_scanhash(struct thr_info *thr, struct work *work,
if (icarus->device_fd == -1)
if (!icarus_prepare(thr)) {
applog(LOG_ERR, "ICA%i: Comms error", icarus->device_id);
- icarus->device_last_not_well = time(NULL);
- icarus->device_not_well_reason = REASON_DEV_COMMS_ERROR;
- icarus->dev_comms_error_count++;
+ dev_error(icarus, REASON_DEV_COMMS_ERROR);
// fail the device if the reopen attempt fails
return -1;
@@ -691,9 +689,7 @@ static int64_t icarus_scanhash(struct thr_info *thr, struct work *work,
if (ret) {
do_icarus_close(thr);
applog(LOG_ERR, "ICA%i: Comms error", icarus->device_id);
- icarus->device_last_not_well = time(NULL);
- icarus->device_not_well_reason = REASON_DEV_COMMS_ERROR;
- icarus->dev_comms_error_count++;
+ dev_error(icarus, REASON_DEV_COMMS_ERROR);
return 0; /* This should never happen */
}
@@ -713,9 +709,7 @@ static int64_t icarus_scanhash(struct thr_info *thr, struct work *work,
if (ret == ICA_GETS_ERROR) {
do_icarus_close(thr);
applog(LOG_ERR, "ICA%i: Comms error", icarus->device_id);
- icarus->device_last_not_well = time(NULL);
- icarus->device_not_well_reason = REASON_DEV_COMMS_ERROR;
- icarus->dev_comms_error_count++;
+ dev_error(icarus, REASON_DEV_COMMS_ERROR);
return 0;
}
diff --git a/driver-modminer.c b/driver-modminer.c
index 9c4c45b..afdd68b 100644
--- a/driver-modminer.c
+++ b/driver-modminer.c
@@ -560,17 +560,13 @@ static uint64_t modminer_process_results(struct thr_info *thr)
applog(LOG_WARNING, "%s%u.%u: Hit thermal cutoff limit (%f) at %f, disabling device!", modminer->api->name, modminer->device_id, fpgaid, MODMINER_CUTOFF_TEMP, state->temp);
modminer_delta_clock(thr, true, MODMINER_OVERHEAT_CLOCK, true);
+ dev_error(modminer, REASON_DEV_THERMAL_CUTOFF);
modminer->deven = DEV_RECOVER;
- modminer->device_last_not_well = time(NULL);
- modminer->device_not_well_reason = REASON_DEV_THERMAL_CUTOFF;
- modminer->dev_thermal_cutoff_count++;
} else {
applog(LOG_WARNING, "%s%u.%u Overheat limit (%f) reached %f", modminer->api->name, modminer->device_id, fpgaid, MODMINER_OVERHEAT_TEMP, state->temp);
modminer_delta_clock(thr, true, MODMINER_CLOCK_DOWN, true);
- modminer->device_last_not_well = time(NULL);
- modminer->device_not_well_reason = REASON_DEV_OVER_HEAT;
- modminer->dev_over_heat_count++;
+ dev_error(modminer, REASON_DEV_OVER_HEAT);
}
}
}
diff --git a/util.c b/util.c
index d5c2e54..d9d1d48 100644
--- a/util.c
+++ b/util.c
@@ -1423,3 +1423,46 @@ out:
return ret;
}
+
+
+void dev_error(struct cgpu_info *dev, enum dev_reason reason)
+{
+ dev->device_last_not_well = time(NULL);
+ dev->device_not_well_reason = reason;
+
+
+ switch (reason)
+ {
+ case REASON_THREAD_FAIL_INIT:
+ dev->thread_fail_init_count++;
+ break;
+ case REASON_THREAD_ZERO_HASH:
+ dev->thread_zero_hash_count++;
+ break;
+ case REASON_THREAD_FAIL_QUEUE:
+ dev->thread_fail_queue_count++;
+ break;
+ case REASON_DEV_SICK_IDLE_60:
+ dev->dev_sick_idle_60_count++;
+ break;
+ case REASON_DEV_DEAD_IDLE_600:
+ dev->dev_dead_idle_600_count++;
+ break;
+ case REASON_DEV_NOSTART:
+ dev->dev_nostart_count++;
+ break;
+ case REASON_DEV_OVER_HEAT:
+ dev->dev_over_heat_count++;
+ break;
+ case REASON_DEV_THERMAL_CUTOFF:
+ dev->dev_thermal_cutoff_count++;
+ break;
+ case REASON_DEV_COMMS_ERROR:
+ dev->dev_comms_error_count++;
+ break;
+ case REASON_DEV_THROTTLE:
+ dev->dev_throttle_count++;
+ break;
+ }
+
+}
diff --git a/util.h b/util.h
index d5ac54c..61dc668 100644
--- a/util.h
+++ b/util.h
@@ -43,11 +43,14 @@
#endif
struct pool;
+enum dev_reason;
+struct cgpu_info;
bool stratum_send(struct pool *pool, char *s, ssize_t len);
char *recv_line(struct pool *pool);
bool parse_method(struct pool *pool, char *s);
bool extract_sockaddr(struct pool *pool, char *url);
bool auth_stratum(struct pool *pool);
bool initiate_stratum(struct pool *pool);
+void dev_error(struct cgpu_info *dev, enum dev_reason reason);
#endif /* __UTIL_H__ */