Wanted to follow up that I've tried the below fix and have been logging how long it runs for, CPU/GPU utilization, and temps.
The latest source doesn't work properly for high-end AMD cards (7XXX and greater). Solution is to change line 459 in oclengine.c from: return quirks; to: return quirks & ~VG_OCL_AMD_BFI_INT;
https://en.bitcoin.it/wiki/VanitygenIt still seems to stall out the same amount of times per day although I've written a script that restarts the process since I'm able to check the current GPU temperature versus expected temperature. In my case the temperature while searching is ~73-76c and when the temperature drops below a certain number I can be sure that oclvanitygen has stalled.
The output of my restart script log shows the below while I'm using another script to log the utilization of CPU/GPU.
# # # # # Fri Jul 21 01:42:45 EDT 2017 # # # # # # # # # #
STARTING oclvanitygen
Starting GPU Temp: 52c
# # # # # Fri Jul 21 03:28:05 EDT 2017 # # # # # # # # # #
RESTARTING oclvanitygen
Current GPU Temp: 68c
Run time: 1 hrs, 45 min, 15 sec
Lowest Temp: 61c - Highest Temp: 76c
# # # # # Fri Jul 21 09:38:07 EDT 2017 # # # # # # # # # #
RESTARTING oclvanitygen
Current GPU Temp: 68c
Run time: 6 hrs, 9 min, 15 sec
Lowest Temp: 61c - Highest Temp: 76c
# # # # # Fri Jul 21 12:45:22 EDT 2017 # # # # # # # # # #
RESTARTING oclvanitygen
Current GPU Temp: 68c
Run time: 3 hrs, 6 min, 35 sec
Lowest Temp: 61c - Highest Temp: 76c
# # # # # Fri Jul 21 15:01:59 EDT 2017 # # # # # # # # # #
RESTARTING oclvanitygen
Current GPU Temp: 68c
Run time: 2 hrs, 16 min, 0 sec
Lowest Temp: 61c - Highest Temp: 75c
Restarting it with an auto-restart script that checks GPU temp every 5 seconds solves the issue of losing search time but I'm wondering if there's some debugging mode or logging associated with vanitygen I might be missing. Thanks!
Nobody. It's impossible, man.
