hashcat Forum

Pages: 1 2

when I run the catalyst installation .run file, the installation kicks off like it should and then errors out. what I'm getting is

error: X Server version cannot be detected. (default:v2:x86_64:lib::none:3.11.0-12-generic :
Installation will not proceed.

Removing temporary directory: fglrx-install.ojAb46

Its a fresh server installation. I did not attempt to install default ati drivers. Ati driver version I'm attempting to install amd-catalyst-13.12-linux-x86.x86_64.run. I've attempted other versions and newest beta version as well. Same error. I haven't attempted to use dummy plugs or pull any of the cards out. Has anyone gotten 4 7990's to run under Ubuntu or seen this problem before? Any help would be greatly appreciated. I've been pulling my hair out on this issue and could really use some help.

Since you installed the server version, did you go back and manually install Xorg? That would certainly be one reason why the X server version cannot be detected.

Also, per the release notes for AMD Catalyst, the AMD driver is only supported on Ubuntu 12.04 and 13.04. Ubuntu 12.04 is highly recommended since it is an LTS release. You can make it work on 13.10, but it is unsupported and it will be a pain in the ass. And since it's not an LTS release you'll have to rebuild the box in a year when 13.10 reaches EOL.

You know, you just asked me in PM like two days ago which distribution to use, and I pointed you to the Hashcat wiki on setting up a headless Linux server. It doesn't look like you've followed any of that guide. I would highly recommend following the guide step-by-step until you at least become familiar with how this all works.

So, i have had this problem as well.. I tried installing on The-Distribution-Which-Does-Not-Handle-OpenCL-Well (Kali) and had the help of a super mod on here (philsmd [spent 6 hours on thsi total and finally got it working thank so much from him]) and he took me step by step on what to do but after failing so many times it didn't work, so i switched back to the distro i had before, Ubuntu 13.10. I also had the same problem when installing the normal driver, then i actually looked at what was supported and saw it was the BETA driver i needed... So don't install the normal linux driver, download the BETA version and install it (it will install with no errors, or if it does post em here). Anyways, it installed perfectly without no errors (as long as you did the following before installing it:

Code:
sudo apt-get update

sudo apt-get upgrade && sudo apt-get dist-upgrade

then if not already install the linux-headers (linux-headers-generic) i think is its name.
then reboot and do the following: (should download the BETA driver at this point but do not install):

Code:
sudo apt-get install build-essential cdbs fakeroot dh-make debhelper debconf libstdc++6 dkms libqtgui4 wget execstack libelfg0 linux-headers-generic xorg

Then:

Code:
sudo dpkg --get-selections | grep fglrx (if anything comes up, do: apt-get remove --purge fglrx*)

then reboot one more time and do the following:

Code:
cd /usr

sudo ln -svT lib /usr/lib64

Now go to the beta driver and install it:

Code:
chmod +x amd-*

./amd-*

Just install the normal way, (do not create a specfic package and do automatic) and when it is done it will say it is finished, THEN MOST IMPORTANT!!!"

Code:
aticonfig --initial -f --adapters=all    (might be adapter i don't remember)

NOW REBOOT!!
and have FUN!!!

I have it working perfectly with dual Radeon HD 7870's on Ubuntu 13.10 64-bit and ocl works perfect!

... or, you could just follow the wiki.

epixoip,

First I would like to apologize for not heeding your instructions on the PM to follow the wiki step-by-step! Got all 4 7990's working after following your instructions to the letter from the wiki site. Feel like an ass after not listening to you and wasting countless hours trying to get ver 13 to work! Again, thank you so so so much! Now time for HASHCAT!!!!!!!!!!!!!!!!!!!!!!

actually spoke to soon. all went as planned, followed the wiki excactly as instructed. when I issued the reboot after

amdconfig --adapter=all --initial -f

I get a black screen, sort of back lite lcd black. I can ssh into the box no problem. when I issue a clinfo it hangs up

when I issue the amdconfig --adapter=all --adgt, it says I need xserver running in order to perform. I'm not a hardcore Linux guy so I don't know what the issue may be. can anyone point me in the right direction?

btw, the amd installation went smoothly, no errors

(01-20-2014, 12:38 AM)mtmhashcat Wrote: [ -> ]actually spoke to soon. all went as planned, followed the wiki excactly as instructed. when I issued the reboot after

amdconfig --adapter=all --initial -f

I get a black screen, sort of back lite lcd black. I can ssh into the box no problem. when I issue a clinfo it hangs up

when I issue the amdconfig --adapter=all --adgt, it says I need xserver running in order to perform. I'm not a hardcore Linux guy so I don't know what the issue may be. can anyone point me in the right direction?

btw, the amd installation went smoothly, no errors

exactly,

Code:
fglrxinfo

and

Code:
glx_gears

works, but kernel doesn't recognize the driver.
Your findings are exactly similar to mine that I posted few days back in here https://hashcat.net/forum/thread-3031-pos...l#pid17846

no worries dude.

the symptoms you have described means there is a gpu that is unresponsive, and the driver is stuck in iowait trying to communicate with that gpu. there are a couple reasons for this.

the first reason is that sometimes the driver doesn't install quite properly, but doesn't provide any error message to let you know. just part of the AMD quality control we've all come to know and love.

so first uninstall the driver:

Code:
stop lightdm

rm -f /etc/X11/xorg.conf*

amdconfig --uninstall=force

reboot

then go back through just the driver installation part of the wiki:

https://hashcat.net/wiki/doku.php?id=linu...d_catalyst

make sure you reboot after re-installing the driver.

if the problem persists, it may be that you have a bad gpu. so i would then recommend taking the system down to one card and testing, then add another card and test, then add another and test, so on and so forth.

but the method in the wiki is proven, we've personally built out scores of boxes with those exact steps without issue, and i know many other hashcat users have followed this guide without incident as well. and we know that this works with quad 7990s as well because we've built around 20 quad 7990s boxes so far, and they all work flawlessly.

Hello,

I've been using hashcat for a little over 2 years now, but I have gotten serious with it in the past 4 months. It's my favorite tool. I use it on two boxes for CTF competitions which I have one tomorrow!

For the past month, I've been wrestling with abnormalities running hashcat with different combinations of OSes (Windows 7 & Ubuntu 14.04, 14.10), AMD drivers (about 5 different versions),

While doing an incremental brute force (i.e. ?a) from 10 to 20 in length (yes I know it could take years, my objective is to stress test it, not to crack it) the abnormalities range from:
a) computer freezes
b) about 10 min to a couple of hours, GPUs working went from 4 to 3 or 1

So I finally decided to follow the wiki guide below, step-by-step given the fact that 'mtmhashcat' and I have the same AMD 7990 cards (I have two, he has four).

(01-20-2014, 09:33 AM)epixoip Wrote: [ -> ]https://hashcat.net/wiki/doku.php?id=linu...d_catalyst

After following all the steps, the job starts off by using all four GPUs (i.e. both cards), but 5 to 10 minutes into the job, 3 of my GPUs flat-line (i.e. 0% usage) and only one is working, when I hit 'q' to quit, the terminal freezes. If i start another terminal, and restart from the beginning (i.e. not restore, but a fresh start, note: I do not delete any generated/cache files from the previous job) this second terminal hangs after printing "generating tables" (or something similar, I'm writing from memory).

I am using oclHashcat for AMD v1.31. The command I use is the following:

Code:
oclHashcat64.bin -m 0 -a 3 -d 1,2,3,4 -p : --session=crypto8md5 --gpu-temp-disable --gpu-accel=800 --gpu-loops=1024 --username -o cracked.txt -i --increment-min=10 --increment-max=20 CRYPTO-8.md5hash.txt ?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a?a

It's worth mentioning that I don't think temperature is an issue given that I keep my GPUs nicely cooled.

Another angle to my problems could be BIOS settings? I don't know much about BIOS settings other than changing the boot priority. And I do remember I defaulted them a while back. I have a funny suspicion that I changed something inadvertently in some BIOS setting. Does anyone have any clues how to tweak BIOS settings to get a favorable result to the problem I'm having?

I'm really hoping somebody sees this message and has some insight on a solution to my problem.

In advanced, thank you very much! Hope to hear from you soon.

(11-14-2014, 03:16 PM)trashcan Wrote: [ -> ]the job starts off by using all four GPUs but 5 to 10 minutes into the job, 3 of my GPUs flat-line (i.e. 0% usage) and only one is working, when I hit 'q' to quit, the terminal freezes.

this is a bug in 1.31. hopefully atom fixes it soon because it is affecting a lot of people.

Pages: 1 2

mtmhashcat

epixoip

3kings

epixoip

mtmhashcat

mtmhashcat

blackMORE

epixoip

trashcan

epixoip