[<prev] [<thread-prev] [day] [month] [year] [list]
Message-Id: <FB52A08F-411A-41D9-8C52-17B35E9F9EC2@xmtservices.net>
Date: Wed, 4 Jun 2025 21:20:57 -0700
From: Shawn <shawn@...services.net>
To: john-users@...ts.openwall.com
Subject: Re: md5crypt-long works in John but not in hashcat
Thank you!
Shawn
> On Jun 4, 2025, at 20:57, atom <atom@...hcat.net> wrote:
>
> Yes, it works for all attack-modes.
>
>> $ echo verylongpasswordmorethan16 > wordlist.txt
>>
>> $ ./hashcat '$1$4YFHDKR0$v7SN48VopUztrNFFaIMzq1' wordlist.txt
> --potfile-disable
>> ...
>> Minimum password length supported by kernel: 0
>> Maximum password length supported by kernel: 256
>> ...
>> Dictionary cache built:
>> * Filename..: wordlist.txt
>> * Passwords.: 1
>> * Bytes.....: 27
>> * Keyspace..: 1
>> * Runtime...: 0 secs
>> ...
>> $1$4YFHDKR0$v7SN48VopUztrNFFaIMzq1:verylongpasswordmorethan16
>>
>> Session..........: hashcat
>> Status...........: Cracked
>> ...
>
>> On Thu, Jun 5, 2025 at 1:30 AM Shawn <shawn@...services.net> wrote:
>>
>> Atom,
>>
>> Does this apply to wordlist cracks as well?
>>
>>
>> Shawn
>>
>>>> On Jun 4, 2025, at 00:08, atom <atom@...hcat.net> wrote:
>>>
>>> Most important: I'm 100% sure, magnum, you're not too stupid.
>>>
>>> I just wanted to respond because it was said Hashcat can't handle long
>>> passwords. It can. Most people these days just use -O from muscle memory
>>> and forget the implications.
>>>
>>> Make sure you omit -O, and you'll be able to crack passwords longer than
>> 16
>>> characters.
>>>
>>>> $ echo -n verylongpasswordmorethan16 | mkpasswd -s -5
>>>
>>>> $1$4YFHDKR0$v7SN48VopUztrNFFaIMzq1
>>>
>>> Test:
>>>
>>>> $ ./hashcat -m 500 '$1$4YFHDKR0$v7SN48VopUztrNFFaIMzq1' -a 3
>>> ?l?l?l?l?longpasswordmorethan16 --adv
>>>
>>>> hashcat (v6.2.6-941-gad38c9253) starting
>>>> ...
>>>> Minimum password length supported by kernel: 0
>>>> Maximum password length supported by kernel: 256
>>>
>>> Above you can see the supported password length (depending on the given
>>> command line parameters).
>>>
>>>> ...
>>>
>>> $1$4YFHDKR0$v7SN48VopUztrNFFaIMzq1:verylongpasswordmorethan16
>>>
>>>> ...
>>>
>>>
>>>
>>>> On Tue, Jun 3, 2025 at 12:30 PM magnum <magnumripper@...hmail.com>
>> wrote:
>>>>
>>>>> On 2025-06-02 07:05, Shawn Tayler wrote:
>>>>> Just curious what you are doing different in JTR on your
>>>>> format md5crypt-long. JTR cracks the hashes straight away but
>>>>> I don't have cuda support with it. I have cuda on hashcat but
>>>>> it doesn't crack over there.
>>>>>
>>>>> What are you guys doing differently?
>>>>>
>>>>> Is there a document to walk me through getting CUDA support on
>>>>> JTR?
>>>>>
>>>>> Thanks for all the great help in the past.
>>>>
>>>> Does any JtR format work with CUDA? If so, you're actually running
>>>> OpenCL then (which comes with CUDA). Our actual CUDA support was
>>>> experimental and only had a few formats. We dropped it years ago in
>>>> favor of OpenCL as the latter is portable.
>>>>
>>>> Other than that, the md5crypt-long format is CPU-only. You're supposed
>>>> to run md5crypt-opencl and then (or in parallel) md5crypt-long with
>>>> --min-len=16. We should definitely support passwords longer that 15 with
>>>> OpenCL but no-one has contributed that and I'm too stupid to understand
>>>> the optimizations so my tries have failed.
>>>>
>>>> magnum
>>>>
>>>>
>>>
>>> --
>>> atom
>>
>>
>
> --
> atom
Powered by blists - more mailing lists
Confused about mailing lists and their use?
Read about mailing lists on Wikipedia
and check out these
guidelines on proper formatting of your messages.
