Please note, this is a STATIC archive of website hashcat.net from 08 Oct 2020, cach3.com does not collect or store any user information, there is no "phishing" involved.

hashcat Forum

Full Version: oclhashcat 1.01 benchmarks fail AMD Radeon 7970
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Hi,
I'm quite new to using oclhashcat so forgive me if I missed something obvious.

I just upgraded from an nVidia Quadro 2000D to a AMD Radeon HD 7970 and I am getting a problem when running the benchmark -b option.

It fails with an error in the same place every time I run it:
PS C:\Users\Administrator\Downloads\oclhashcat\oclHashcat-1.01> .\oclHashcat64.exe -b
oclHashcat v1.01 starting in benchmark-mode...

Device #1: Tahiti, 2048MB, 925Mhz, 32MCU
Hashtype: MD4
Workload: 1024 loops, 256 accel
Speed.GPU.#1.: 16053.9 MH/s
Hashtype: MD5
Workload: 1024 loops, 256 accel
Speed.GPU.#1.: 8014.7 MH/s
Hashtype: SHA1
Workload: 512 loops, 256 accel
Speed.GPU.#1.: 2516.3 MH/s
Hashtype: SHA256
Workload: 256 loops, 256 accel
Speed.GPU.#1.: 1009.6 MH/s
Hashtype: SHA512
Workload: 128 loops, 256 accel
Speed.GPU.#1.: 75153.6 kH/s
Hashtype: SHA-3(Keccak)
Workload: 256 loops, 256 accel
Speed.GPU.#1.: 141.6 MH/s
Hashtype: RipeMD160
Workload: 256 loops, 256 accel
Speed.GPU.#1.: 1629.6 MH/s
Hashtype: Whirlpool
Workload: 256 loops, 256 accel
Speed.GPU.#1.: 31003.7 kH/s
Hashtype: GOST R 34.11-94
Workload: 256 loops, 256 accel
Speed.GPU.#1.: 99653.0 kH/s
Hashtype: SHA-1(Base64), nsldap, Netscape LDAP SHA
Workload: 512 loops, 256 accel
Speed.GPU.#1.: 2516.9 MH/s
Hashtype: SSHA-1(Base64), nsldaps, Netscape LDAP SSHA
Workload: 512 loops, 256 accel
Speed.GPU.#1.: 2502.2 MH/s
ERROR: clEnqueueNDRangeKernel() -4

I have tried a fresh install in case there was a problem with the kernels from the previous card but it has made no difference. I know someone else using the same card and the same config as me but I think they are using older oclhashcat-plus.

Other than this problem the card and oclhashcat seems to function perfectly with example scripts and the hashes I've tried.

Any suggestions much appreciated?

Andy
(02-06-2014, 04:50 PM)andy_rat Wrote: [ -> ]Any suggestions much appreciated?

Andy

using newest amd drivers?
(02-06-2014, 04:53 PM)curlyboi Wrote: [ -> ]
(02-06-2014, 04:50 PM)andy_rat Wrote: [ -> ]Any suggestions much appreciated?

Andy

using newest amd drivers?

Yes. I have 13-12_win7_win8_64_dd_ccc_whql.exe
installed. I should have said. It's running on windows 2012 foundation server (I had a license doing nothing). It is running inside VMware ESXi using PCI-E pass-through. Although I don't think that makes any difference. As I said I know someone using the same setup Post here https://isc.sans.edu/forums/diary/Buildi...loud/16505
>> ERROR: clEnqueueNDRangeKernel() -4
Is there anything I could try to fix this?
I thought I'd try running all the tests individually and exclude the one that seems to be causing me a problem. I put a little bit of scruffy powershell together to do it:
Code:
$allHashTypes = get-content -Path C:\files\oclhashcathashtypes.txt  | foreach {"$(($_ -split'=')[0])"} | foreach {$_.trim()}
$exclusions = Get-Content -Path c:\files\exclusions.txt | foreach {$_.trim()}

$exe="C:\Users\Administrator\Downloads\oclhashcat\oclHashcat-1.01\oclhashcat64.exe"
$arg1 = "-m"
$arg3 = "-b"


foreach ($hashType in $(Compare-Object $allHashTypes $exclusions -PassThru))
{
$allArgs = @($arg1,$hashType,$arg3)
$result = iex "&$exe $allArgs"
$test =""
$speed=""
foreach ($line in $result)
    {
    if ($line.contains("Hashtype"))
        {$test = $line.replace("Hashtype: ","").trim()}
    if ($line.contains("Speed"))
        {
        $speed = $line.replace("Speed.GPU.#1.:","").trim()
        write-host "$hashType, $test, $speed"
        }
    }
}
If anyone wants to use this you just need to copy the list of hash types from the oclhashcat help screen and create a file called exclusions.txt containing the numbers of the hash types you want exclude (1 per line). If I add a single invalid number to the exclusions file I get an error but it will still do all tests).

And guess what? Doing them individually I can run them all without getting the error.
So it only happens if I run oclhashcat with a straight -b option running all tests sequentially.

0, MD5, 8003.5 MH/s
10, md5($pass.$salt), 8054.2 MH/s
20, md5($salt.$pass), 4411.2 MH/s
30, md5(unicode($pass).$salt), 8037.5 MH/s
40, md5($salt.unicode($pass)), 4326.9 MH/s
50, HMAC-MD5 (key = $pass), 1178.7 MH/s
60, HMAC-MD5 (key = $salt), 2236.1 MH/s
100, SHA1, 2510.5 MH/s
110, sha1($pass.$salt), 2499.1 MH/s
120, sha1($salt.$pass), 1606.4 MH/s
130, sha1(unicode($pass).$salt), 2492.9 MH/s
140, sha1($salt.unicode($pass)), 1526.5 MH/s
150, HMAC-SHA1 (key = $pass), 543.7 MH/s
160, HMAC-SHA1 (key = $salt), 1074.5 MH/s
190, sha1(LinkedIn), 2470.2 MH/s
300, MySQL, 1181.3 MH/s
400, phpass, MD5(Wordpress), MD5(phpBB3), 2057.7 kH/s
500, md5crypt, MD5(Unix), FreeBSD MD5, Cisco-IOS MD5, 3489.1 kH/s
900, MD4, 15930.3 MH/s
1000, NTLM, 15480.5 MH/s
1100, DCC, mscash, 4189.9 MH/s
1400, SHA256, 1016.9 MH/s
1410, sha256($pass.$salt), 999.7 MH/s
1420, sha256($salt.$pass), 846.4 MH/s
1430, sha256(unicode($pass).$salt), 990.5 MH/s
1440, sha256($salt.$pass), 829.5 MH/s
1450, HMAC-SHA256 (key = $pass), 237.6 MH/s
1460, HMAC-SHA256 (key = $salt), 497.0 MH/s
1500, descrypt, DES(Unix), Traditional DES, 84696.2 kH/s
1600, md5apr1, MD5(APR), Apache MD5, 3503.3 kH/s
1700, SHA512, 74205.8 kH/s
1710, sha512($pass.$salt), 73085.7 kH/s
1720, sha512($salt.$pass), 70351.3 kH/s
1730, sha512(unicode($pass).$salt), 72978.0 kH/s
1740, sha512($salt.unicode($pass)), 69858.1 kH/s
1750, HMAC-SHA512 (key = $pass), 17864.9 kH/s
1760, HMAC-SHA512 (key = $salt), 34335.5 kH/s
1800, sha512crypt, SHA512(Unix), 12532 H/s
2100, DCC2, mscash2, 102.2 kH/s
2400, Cisco-PIX MD5, 5257.9 MH/s
2500, WPA/WPA2, 131.1 kH/s
2600, Double MD5, 2124.6 MH/s
3000, LM, 1288.0 MH/s
3100, Oracle 7-10g, 358.0 MH/s
3200, bcrypt, Blowfish(OpenBSD), 3533 H/s
5000, SHA-3(Keccak), 141.7 MH/s
5100, Half MD5, 4523.4 MH/s
5200, Password Safe SHA-256, 469.9 kH/s
5300, IKE-PSK MD5, 502.6 MH/s
5400, IKE-PSK SHA1, 273.9 MH/s
5500, NetNTLMv1-VANILLA / NetNTLMv1+ESS, 8024.5 MH/s
5600, NetNTLMv2, 495.3 MH/s
5700, Cisco-IOS SHA256, 1009.9 MH/s
5800, Samsung Android Password/PIN, 1565.7 kH/s
6000, RipeMD160, 1600.1 MH/s
6100, Whirlpool, 30621.5 kH/s
6211, TrueCrypt 5.0+ PBKDF2-HMAC-RipeMD160 + AES, 377.8 kH/s
6221, TrueCrypt 5.0+ PBKDF2-HMAC-SHA512 + AES, 36389 H/s
6231, TrueCrypt 5.0+ PBKDF2-HMAC-Whirlpool + AES, 1319 H/s
6241, TrueCrypt 5.0+ PBKDF2-HMAC-RipeMD160 boot-mode + AES, 745.8 kH/s
6300, AIX {smd5}, 3485.5 kH/s
6400, AIX {ssha256}, 6280.7 kH/s
6500, AIX {ssha512}, 514.6 kH/s
6600, 1Password, 1061.4 kH/s
6700, AIX {ssha1}, 13731.1 kH/s
6800, Lastpass, 953.1 kH/s
6900, GOST R 34.11-94, 98868.7 kH/s
7100, OSX v10.8, 525 H/s
7200, GRUB 2, 1837 H/s
7400, sha256crypt, SHA256(Unix), 78385 H/s
7500, Kerberos 5 AS-REQ Pre-Auth etype 23, 51728.2 kH/s
11, Joomla, 8062.5 MH/s
21, osCommerce, xt:Commerce, 4403.1 MH/s
101, SHA-1(Base64), nsldap, Netscape LDAP SHA, 2516.2 MH/s
111, SSHA-1(Base64), nsldaps, Netscape LDAP SSHA, 2493.0 MH/s
112, Oracle 11g, 2497.8 MH/s
121, SMF > v1.1, 1577.9 MH/s
122, OSX v10.4, v10.5, v10.6, 1603.6 MH/s
131, MSSQL(2000), 2489.9 MH/s
132, MSSQL(2005), 2496.9 MH/s
141, EPiServer 6.x < v4, 1529.6 MH/s
1441, EPiServer 6.x > v4, 828.7 MH/s
1711, SSHA-512(Base64), LDAP {SSHA512}, 73024.9 kH/s
1722, OSX v10.7, 70420.2 kH/s
1731, MSSQL(2012), 72304.1 kH/s
2611, vBulletin < v3.8.5, 2123.2 MH/s
2711, vBulletin > v3.8.5, 1523.3 MH/s
2811, IPB2+, MyBB1.2+, 1523.3 MH/s

Any thoughts?
Trying out oclhashcat and also noticed it fails during the built-in benchmark right after DES on R9 270X, WHQL 13.12 drivers and SDK 2.9.

Hashtype: SSHA-1(Base64), nsldaps, Netscape LDAP SSHA
Workload: 512 loops, 256 accel

Speed.GPU.#1.: 1772.0 MH/s

Hashtype: descrypt, DES(Unix), Traditional DES
Workload: 64 loops, 32 accel

Speed.GPU.#1.: 60941.9 kH/s
ERROR: clEnqueueNDRangeKernel() -4
I wonder if it's a problem with our driver version then? Have you tried running the individual benchmark for the one where you get the error?