OclHashcat 1.31 Session restore issue

[Gold_digger]

Hi,

I have updated oclhashcat from 1.30 to 1.31 and I have discovered that session restore does not work at all in both combinator and brute force modes.

I run hashat with this command:

oclHashcat64.exe --hash-type 2611 -a 3 --remove --outfile X:\bruteforce.out --outfile-format 2 K:/hashes.txt ?l?l?l?l?l?d?d --session new --gpu-temp-retain 70 -n 30

When it reached 50% my system crashed. So I did oclHashcat64.exe --session new --restore and it started from the beggining. I have also tried quitting session after it reached 5 % and restoring it. Unfortunately it also started from the beggining.

What is weird session.restore file is modified every minute so it seems that progress is saved...

Same thing when I start combinator attack.

My platform is Win7x64 AMD 7970 Ghz

any help would be apreciated.

[minga]

This is a known bug. I think its even officially confirmed. You can wait for a DEV to respond about that.

Go back to 1.30 where the bug is still partially present, but much much much harder to recreate.

[Gold_digger]

Thanks !

[philsmd]

I can identify several problems here:
1. problem reports etc go to https://hashcat.net/trac/
2. I double-checked if there are/were (closed) any trac tickets and also asked atom (dev) if he known of any reported --restore problem with oclHashcat 1.31 and we came to the conclusion that there are no tickets and were no (further) error reports (besides this one)
3. I also tried to reproduce it (with the commands given above), but was not able to identify any problem with --restore. If I hit 's' immediately after I run oclHashcat --session new --restore I see that the Progress line is way larger than 0%.
4. I have no idea how large those files are, for instance hashes.txt etc
5. did you try to modify the command line i.e. for testing purpose don't use the --remove etc

I would recommend, if you are still experiencing this problem, that you should go to https://hashcat.net/trac/, search for maybe related available tickets and if no such ticket exist, create one with full details...(including number of hashes, if it does depend on some of the switches you use, e.g. --remove etc)
Thx

UPDATE: I did some further tests and I am still convince it is working fine. The only thing that you might need to consider is that oclHashcat needs to pass a so-called checkpoint, only after reaching the new checkpoint the --restore "value" will increase.

It depends on your setup (gpu etc), hash type, number of hashes and especially also the number of salts when the next checkpoint will be reached. This can sometimes happen within some seconds/minutes, but in other cases (large amount of salted hashes, slow hashes etc) only after some hours.

There is currently an open trac ticket https://hashcat.net/trac/ticket/445 which may help some of the users to discover when such a checkpoint was reached, furthermore afaik --status-automat already displays this information