Re: Release 1.1 (beta) of my AES implementation
From: Mok-Kong Shen (mok-kong.shen_at_t-online.de)
Date: 06/25/03
- Next message: Marin: "Re: Release 1.1 (beta) of my AES implementation"
- Previous message: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- In reply to: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- Next in thread: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- Reply: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Date: Wed, 25 Jun 2003 17:55:24 +0200
Tom St Denis wrote:
>
> So what did you have in mind?
>
> #include "aes.c"
> #include "sha1.c"
> #include "...."
>
> ???
Exactly. (Unfortunately.)
> You're not going to get any users [re: bug finders] that way.
I am not sure of that. In my past experience lots of
people program that way.
> > I can think that a candidate application of my code
> > would be doing something like:
> >
> > read a block from input file to aesin;
> > aesprocess();
> > write aesout to output file;
> >
> > There is nothing to fork, isn't it?
>
> fork does not a thread make. fork makes processes [hint: there is a
> difference].
>
> If an application is threaded the global variable share the same
> location in memory. If two threads wanted to use AES simultaneously
> they could not with your code.
>
> Want a very realistic application of this?
>
> Sure. Sony Computer Entertainment of America [SCEA], who uses
> LibTomCrypt for their PC and PS2 development intends to use it to
> encrypt packets over UDP/IP going in both directions. They have
> contacted me about several ideas for handling UDP traffic. I think
> their idea is to spawn two threads for each direction and apply the
> cipher at the outgoing/ingoing edges. Anyways, the fact that LTC is
> thread safe [in that respect] was appreciated by the developers.
Sorry that I don't yet understand. If two independent
programs each include my code as header file, one
piece of my code (only the functions actually needed)
would be in each of their memory spaces which are
independent of each other. There couldn't be any
concurrent access to any variable, for there isn't any
common (shared) variable in the first place. Or am I
missing something?
Thanks.
M. K. Shen
- Next message: Marin: "Re: Release 1.1 (beta) of my AES implementation"
- Previous message: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- In reply to: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- Next in thread: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- Reply: Tom St Denis: "Re: Release 1.1 (beta) of my AES implementation"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ] [ attachment ]
Relevant Pages
|