GTX 670 Issues with oclHashcat-plus
#1
Hi,

I am having troubles running v0.81 oclHashcat-plus on a GTX 670 w/ v295.59 Nvidia drivers under Arch Linux x64. I have no issues running oclHashcat-lite.

oclHashcat-plus will run successfully for a few seconds and then stop. It remains possible to request status updates, pause and resume, the quit command will not work. Status updates after it has stopped demonstrate that progress has ceased as this figure remains static but the time and speed alter accordingly.

Code:
./cudaHashcat-plus64.bin -m 1400 -r rules/best64.rule hashes.txt wordlist.txt

cudaHashcat-plus v0.08 by atom starting...

Hashes: 419564
Unique digests: 419564
Bitmaps: 21 bits, 1048576 entries, 0x000fffff mask, 4194304 bytes
Rules: 78
GPU-Loops: 32
GPU-Accel: 8
Password lengths range: 1 - 15
Platform: NVidia compatible platform found
Watchdog: Temperature limit set to 90c
Device #1: GeForce GTX 670, 4095MB, 705Mhz, 7MCU
Device #1: Allocating 8MB host-memory
Device #1: Kernel ./kernels/4318/m1400_a0.sm_30.64.cubin

Scanned dictionary wordlist.txt: 139921496 bytes, 14344390 words, 1118862420 keyspace, starting attack...

[s]tatus [p]ause [r]esume [q]uit => s
Status.......: Running
Rules.Type...: File (rules/best64.rule)
Input.Mode...: File (wordlist.txt)
Hash.Target..: File (hashes.txt)
Hash.Type....: SHA256
Time.Running.: 4 secs
Time.Left....: 18 secs
Time.Util....: 4828.9ms/173.7ms Real/CPU, 3.7% idle
Speed........: 48397.6k c/s Real,   160.1M c/s GPU
Recovered....: 70/419564 Digests, 0/1 Salts
Progress.....: 235447836/1118862420 (21.04%)
Rejected.....: 1742364/235447836 (0.74%)
HW.Monitor.#1:  0% GPU, 46c Temp
[s]tatus [p]ause [r]esume [q]uit => s
Status.......: Running
Rules.Type...: File (rules/best64.rule)
Input.Mode...: File (wordlist.txt)
Hash.Target..: File (hashes.txt)
Hash.Type....: SHA256
Time.Running.: 8 secs
Time.Left....: 31 secs
Time.Util....: 8353.0ms/173.7ms Real/CPU, 2.1% idle
Speed........: 27978.6k c/s Real,   160.1M c/s GPU
Recovered....: 70/419564 Digests, 0/1 Salts
Progress.....: 235447836/1118862420 (21.04%)
Rejected.....: 1742364/235447836 (0.74%)
HW.Monitor.#1:  0% GPU, 45c Temp

Any help is appreciated.
#2
if you have parallel aticonfig --odgt --odgc to check GPU load and temp, can you see that after keying 's', GPU suddenly goes down to NULL%

if you don't key 's' ochHascat+ caries on with the work until work load ends?

You would be the third reporting this status screen behaviour... only from a different angle.

Seems that window users do not have this problem
#3
sounds like a fade-out-wait-because-of-to-small-dictionary. try to reproduce with larger dictionary. see if it happens only on the end
#4
(07-14-2012, 12:52 AM)ntk Wrote: if you don't key 's' ochHascat+ caries on with the work until work load ends?

It seems that it still never ends if I never press the status button. If I am quick it is possible for me to get a few status updates before the program stops working properly.

(07-14-2012, 08:32 AM)atom Wrote: sounds like a fade-out-wait-because-of-to-small-dictionary. try to reproduce with larger dictionary. see if it happens only on the end

Running with no rules and a dictionary of 37m words, it still freezes:

Code:
Status.......: Running
Input.Mode...: File (bigger_dictionary.txt)
Hash.Target..: File (hashes.txt)
Hash.Type....: SHA256
Time.Running.: 8 secs
Time.Left....: 4 secs
Time.Util....: 8610.5ms/1067.2ms Real/CPU, 14.1% idle
Speed........:  1739.9k c/s Real, 72404.0k c/s GPU
Recovered....: 3/419564 Digests, 0/1 Salts
Progress.....: 30054106/37964470 (79.16%)
Rejected.....: 15072986/30054106 (50.15%)
HW.Monitor.#1:  0% GPU, 40c Temp
[s]tatus [p]ause [r]esume [q]uit => s
#5
looks good for me. remember that you can not efficiently run fast hashes like sha256 without an mutation amplifier on gpu see here: https://hashcat.net/wiki/oclhashcat_plus#background
#6
(07-14-2012, 11:18 AM)atom Wrote: looks good for me. remember that you can not efficiently run fast hashes like sha256 without an mutation amplifier on gpu see here: https://hashcat.net/wiki/oclhashcat_plus#background

It may not be efficient but surely it should not freeze indefinitely? It should still finish, but for me it does not.
#7
yeah, but freezing is mostly caused by hardware problem. plus requires much more memory on gpu than lite, so maybe there is a problem
#8
(07-15-2012, 09:53 AM)atom Wrote: yeah, but freezing is mostly caused by hardware problem. plus requires much more memory on gpu than lite, so maybe there is a problem

Well I can successfully run the same attack on a GTX 260 with less than 1GB of onboard memory, but the GTX 670 has 4GB. I will see if I can check how much memory is being utilised by Hashcat later.

If you honestly think that maybe there is something wrong with my graphics card, do you have any suggestions on how I could test it? I haven't had any issues with any other CUDA applications or intensive GPU benchmarks.
#9
(07-15-2012, 09:53 AM)atom Wrote: yeah, but freezing is mostly caused by hardware problem. plus requires much more memory on gpu than lite, so maybe there is a problem

The graphics card is using 132MB of onboard memory. I guess it might just be a dodgy card. Time to RMA it I guess.

Thanks for all the help by the way Smile
#10
i have the same card. I am running this in Windows but for some reason, cannot get more than 15k in speed for WPA/WPA2. any way i can boost this up a bit?

C:\Users\lcobb\Desktop\HashCat>cudaHashcat-plus64.exe -m 2500 -a 3 --gpu-devices
1 ZombieNet01.hccap -1 ?u?l?d ?1?1?1?1?1?1?1?1?1?1?1?1?1?1?1
cudaHashcat-plus v0.08 by atom starting...

Hashes: 1
Unique salts: 1
Unique digests: 1
Bitmaps: 8 bits, 256 entries, 0x000000ff mask, 1024 bytes
GPU-Loops: 64
GPU-Accel: 1
Password lengths range: 8 - 15
Platform: NVidia compatible platform found
Watchdog: Temperature limit set to 90c
Device #1: GeForce GTX 670, 2048MB, 705Mhz, 7MCU
Device #2: skipped by user
Device #1: Allocating 1MB host-memory
Device #1: Kernel ./kernels/4318/m2500.sm_30.64.cubin

Status.......: Running
Input.Mode...: Mask (?1?1?1?1?1?1?1?1?1?1?1?1?1?1?1)
Hash.Target..: ZombieNet (XX:XX:XX:XX:XX:XX <-> XX:XX:XX:XX:XX:XX)
Hash.Type....: WPA/WPA2
Time.Running.: 15 secs
Time.Left....: > 10 Years
Time.Util....: 15079.8ms/4.0ms Real/CPU, 0.0% idle
Speed........: 15567 c/s Real, 15552 c/s GPU
Recovered....: 0/1 Digests, 0/1 Salts
Progress.....: 234752/4177248169415651 (0.00%)
Rejected.....: 0/234752 (0.00%)
HW.Monitor.#1: 96% GPU, 55c Temp

Any help to get me boosted up to faster speeds on this? I mean i figured with over 1000 CUDA cores, i would be going about 50k at least