[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [GNUnet-developers] sizeof HELO Message wrong

From: Blake Matheny
Subject: Re: [GNUnet-developers] sizeof HELO Message wrong
Date: Fri, 25 Jul 2003 14:06:35 -0500
User-agent: Mutt/1.4i

I thought Rick had used gnunet under sparc-64?


Whatchu talkin' 'bout, Willis?
> Am Freitag, 25. Juli 2003 14:23 schrieben Sie:
> > Christian Grothoff wrote:
> > >Hmm. Best reason I can see would be the compiler doing 64-bit alignment.
> > > But if they are using a binary package that works on other systems, I'll
> > > declare myself puzzled. The check just tests if the compiler packed the
> > > struct the way we expect it to, which is without any holes.
> >
> > That "not too many more details" is:
> >
> > $uname -a
> > Linux skye 2.4.19 #5 Fri May 9 20:48:01 EDT 2003 alpha GNU/Linux
> >
> > It is very possible that the compiler did 64-bit alignment. :-)
> >
> > Have anybody used gnunet on a 64-bit processor ever?
> Not to my knowledge. And fixing all structs to be 64-bit aligned would break 
> the protocol (they are 32-bit aligned, which was pretty much as much as we 
> could reasonably do). Now gcc can be tweaked to align things at 32-bit 
> boundaries (definitely with gcc extensions like __attribute__(packed) or 
> something like that), but I wonder if there's not a better way to do this 
> (without making the code too ugly).
> Either way, the alignment is definitely to blame if this is a 64 bit 
> architecture. Can you check if there are any obscure gcc flags that may help? 
> Otherwise we'll have to do something to the code...
> Christian
> _______________________________________________
> GNUnet-developers mailing list
> address@hidden

Blake Matheny           "... one of the main causes of the fall of the
address@hidden      Roman Empire was that, lacking zero, they had   no way to indicate successful termination of  their C programs." --Robert Firth

reply via email to

[Prev in Thread] Current Thread [Next in Thread]