llvm dev - Feb 2009 - [LLVMdev] Creating an LLVM backend for a very small stack machine

On Monday 23 February 2009 03:23:59 Mark Shannon wrote:
> I've done quite a lot of work on register allocation for stack
machines.
> You might want to look at my papers:
> http://www.dcs.gla.ac.uk/~marks/euroforth.pdf
> http://www.dcs.gla.ac.uk/~marks/thesis.pdf
Hi Mark,

I've read your papers, and in fact they were part of the data that convinced
me that I really could go with a stack machine for the work I'm doing.
I'm
planning on implementing register allocation based partially on your paper.

Actually, I was wondering if any of your lcc implementation work was 
publicly available. 

-- 
Wesley J. Landaker <wjl at icecavern.net> <xmpp:wjl at
icecavern.net>
OpenPGP FP: 4135 2A3B 4726 ACC5 9094  0097 F0A9 8A4C 4CD6 E3D2
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20090223/5428d77f/attachment.sig>

Wesley,

Regarding access to the source code;
I would send you the code, but I might be stepping on a few toes.
The person to speak to is Chris Bailey at the University of York (in the 
UK).

However it is written for the lcc tree-based IR, rather than the 
SSA-based IR of LLVM, so I don't think it will be that much use.

A lot of the analysis it does is to find information that is explicit in 
the SSA form anyway, such as generating a flow graph.

The SSA form already provides a head start for chosing stack allocation 
candidates. For example, PHI-nodes are obvious candidates for edges in 
the flow-graph.

I would be interested to see how good SSA-form is for stack allocation,
as that was my intended direction if I had stayed longer at York.

Cheers,
Mark.

Wesley J. Landaker wrote:
> On Monday 23 February 2009 03:23:59 Mark Shannon wrote:
>> I've done quite a lot of work on register allocation for stack
machines.
>> You might want to look at my papers:
>> http://www.dcs.gla.ac.uk/~marks/euroforth.pdf
>> http://www.dcs.gla.ac.uk/~marks/thesis.pdf
> 
> Hi Mark,
> 
> I've read your papers, and in fact they were part of the data that
convinced
> me that I really could go with a stack machine for the work I'm doing.
I'm
> planning on implementing register allocation based partially on your paper.
> 
> Actually, I was wondering if any of your lcc implementation work was 
> publicly available. 
> 
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev

On Tuesday 24 February 2009 07:49:17 Mark Shannon wrote:
> Regarding access to the source code;
> I would send you the code, but I might be stepping on a few toes.
> The person to speak to is Chris Bailey at the University of York (in the
> UK).
>
> However it is written for the lcc tree-based IR, rather than the
> SSA-based IR of LLVM, so I don't think it will be that much use.
>
> A lot of the analysis it does is to find information that is explicit in
> the SSA form anyway, such as generating a flow graph.
That's okay, I wasn't sure it would be much use either; mostly I was
just
curious and thought maybe I could learn something from it if it was readily 
available.
> The SSA form already provides a head start for chosing stack allocation
> candidates. For example, PHI-nodes are obvious candidates for edges in
> the flow-graph.
I'm a fan of SSA already (used it a bit in a hardware synthesis context), 
but this is my first attempt to use it an LLVM context, or to target a 
stack machine. I guess I'll see how it goes.
> I would be interested to see how good SSA-form is for stack allocation,
> as that was my intended direction if I had stayed longer at York.
I hope to eventually get to this on my project. For my first cut I'm going 
to do a rather dumb allocation just to get things working. In my niche, 
most programs for this machine will be under 256 (!) instructions, and 
speed doesn't matter much. So on one hand, if it fits, optimizations
don't
matter much. But on the other hand, it has to fit! =)

-- 
Wesley J. Landaker <wjl at icecavern.net> <xmpp:wjl at
icecavern.net>
OpenPGP FP: 4135 2A3B 4726 ACC5 9094  0097 F0A9 8A4C 4CD6 E3D2
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: This is a digitally signed message part.
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20090224/238b8dc3/attachment.sig>