Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Wed, 30 Oct 2013 19:15:07 +0100
From: magnum <>
Subject: Re: Compile john in Windows 64 bits

On 2013-10-30 00:43, magnum wrote:
> On 2013-10-29 18:14, magnum wrote:
>> Nailed it. It turned out to be much worse than alignment and in
>> hindsight it is obvious: Microsoft uses their own ABI, different from
>> what anyone else uses.
>> I think I have solved most of it (waiting for Dhiru to test, I'm just
>> dry running). It might be that some invocation of EPILOGUE should be
>> removed from the DES stuff - please review.
> A few tweaks ended up in problem solved at 3f42f04 - presuming a
> self-test hits all code paths. Otherwise I'm still not 100% sure about
> all epilogues.
> Some more details in


We are trying to resolve BSDIcrypt problems on Win64. It seems to get 
stuck somewhere in an endless loop on Win64. DEScrypt, LM and Deepika's 
formats are fine.

I can't see any pop corresponding to this push:

	leaq DES_bs_crypt_25_body(%rip),tmp1
	pushq tmp1

What's with that? And while we're at it,

	cmpl $0,DES_bs_all_keys_changed(%rip)
	jz DES_bs_crypt_body
	pushq %rdi
	call DES_bs_finalize_keys
	popq %rdi

What good does the push/pop of %rdi do here? I can't see it being used 


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.