08-10-2017, 07:26 PM
Has anyone come across the error "clGetEventProfilingInfo(): CL_UNKNOWN_ERROR" on the brand new EVGA GeForce GTX 1080 Ti SC2 HYBRID GAMING, 11G-P4-6598-KR, 11GB GDDR5X, HYBRID & LED board?
I'm running it headless (although lowering the boards priority to 3 or 2 from 4 doesn't help), have installed multiple NVIDIA driver versions on the board, downgraded Hashcat versions to 3.0 from 3.6.0 - without success. Windows 7 64 bit
I am using it to run 6213 Truecrypt hash. It won't even get to the computing part and crashes with this error before you get to interactive mode.
My setup works without errors with my EVA gtx970. When the new board is enabled in pairing, or by itself - I get the same error. But the board will run benchmarks just fine and renders correctly. I'm not overclocking it at all.
This is run using the batch file - and no, it also fails in direct mode, very simple stuff - two predefined character sets, and 4 characters total on substitution. This should run in just a few minutes. No dictionary either.
I have two batch files and run all sequences of the hash I need. I have converted the two into the final form at the end.
HASHDO.BAT
call m:\H360\HashdoFile2.bat 6213 filler_enc.bin ?a FILLER_NOFILE_GAN ?1andalf?2?2?2 8_12_3 M:\Work\ Gg
HASHDO2.BAT
m:\h360\hashcat64.exe -a 3 -m %1 %7%2 -1 %8 -2 %3 %5 --session=%4_%1_%6 -p : --force -o "%7Password_%4_%1_%6.txt" --outfile-format=4 --remove -w 4 --gpu-temp-abort=86
CONVERTED LINE:
m:\h360\hashcat64.exe -a 3 -m 6213 m:\work\filler_enc.bin -1 Gg -2 ?a ?1andalf?2?2?2 --session=FILLER_NOFILE_GAN_6213_8_12_3 -p : --force -o "m:\work\Password_FILLER_NOFILE_GAN_6213_8_12_3.txt" --outfile-format=4 --remove -w 4 --gpu-temp-abort=86
I have run out of ideas. All system updates are installed. No virus activity that I can tell - the board sits inactive like it should when not queried. I could not include the log file or the binary, in .rar or native modes - not sure why.
=====================================================================
UPDATE: 08/30/17
I have returned the 1080ti AGAIN to Newegg (this is the last time they will allow it)
- This new 1080ti benchmarks properly with EVGA Scanner, also works under memtestg80 (to 4094 memory)
- I have made sure there are no directives in the environment under win7 to OPENCL
- I have disabled all software services that are not needed (vmware, True Image, Chute, Etc.)
- I have downloaded a new binary 3.6 setup from scratch.
- Both my gtx 970 and gtx 108ti are on 385.12
- I have made sure the 1080ti hardware slot is auto configured, and have swapped slots between the 970
--> the gtx 970 works, the 1080ti still does not work.
It does EXACTLY the same as the last card. I'm am guessing now it is either software, or this motherboard - x64
Remember that I did have a 1080ti (same model) run for a week originally before it actually started to lock up my machine - that one was returned for the #2 replacement, this is #3.
I'm running it headless (although lowering the boards priority to 3 or 2 from 4 doesn't help), have installed multiple NVIDIA driver versions on the board, downgraded Hashcat versions to 3.0 from 3.6.0 - without success. Windows 7 64 bit
I am using it to run 6213 Truecrypt hash. It won't even get to the computing part and crashes with this error before you get to interactive mode.
My setup works without errors with my EVA gtx970. When the new board is enabled in pairing, or by itself - I get the same error. But the board will run benchmarks just fine and renders correctly. I'm not overclocking it at all.
This is run using the batch file - and no, it also fails in direct mode, very simple stuff - two predefined character sets, and 4 characters total on substitution. This should run in just a few minutes. No dictionary either.
I have two batch files and run all sequences of the hash I need. I have converted the two into the final form at the end.
HASHDO.BAT
call m:\H360\HashdoFile2.bat 6213 filler_enc.bin ?a FILLER_NOFILE_GAN ?1andalf?2?2?2 8_12_3 M:\Work\ Gg
HASHDO2.BAT
m:\h360\hashcat64.exe -a 3 -m %1 %7%2 -1 %8 -2 %3 %5 --session=%4_%1_%6 -p : --force -o "%7Password_%4_%1_%6.txt" --outfile-format=4 --remove -w 4 --gpu-temp-abort=86
CONVERTED LINE:
m:\h360\hashcat64.exe -a 3 -m 6213 m:\work\filler_enc.bin -1 Gg -2 ?a ?1andalf?2?2?2 --session=FILLER_NOFILE_GAN_6213_8_12_3 -p : --force -o "m:\work\Password_FILLER_NOFILE_GAN_6213_8_12_3.txt" --outfile-format=4 --remove -w 4 --gpu-temp-abort=86
I have run out of ideas. All system updates are installed. No virus activity that I can tell - the board sits inactive like it should when not queried. I could not include the log file or the binary, in .rar or native modes - not sure why.
=====================================================================
UPDATE: 08/30/17
I have returned the 1080ti AGAIN to Newegg (this is the last time they will allow it)
- This new 1080ti benchmarks properly with EVGA Scanner, also works under memtestg80 (to 4094 memory)
- I have made sure there are no directives in the environment under win7 to OPENCL
- I have disabled all software services that are not needed (vmware, True Image, Chute, Etc.)
- I have downloaded a new binary 3.6 setup from scratch.
- Both my gtx 970 and gtx 108ti are on 385.12
- I have made sure the 1080ti hardware slot is auto configured, and have swapped slots between the 970
--> the gtx 970 works, the 1080ti still does not work.
It does EXACTLY the same as the last card. I'm am guessing now it is either software, or this motherboard - x64
Remember that I did have a 1080ti (same model) run for a week originally before it actually started to lock up my machine - that one was returned for the #2 replacement, this is #3.
