[M3devel] "C-generating backend progress report, that nobody asked for" :)

[Daniel Alejandro Benavides D.]

Hi all:
Jay, are you considering Clang as target?
Thanks in advance

--- El lun, 10/9/12, Jay K <jay.krell at cornell.edu> escribió:

De: Jay K <jay.krell at cornell.edu>
Asunto: Re: [M3devel] "C-generating backend progress report, that nobody asked for" :)
Para: "Tony" <hosking at cs.purdue.edu>
CC: "m3devel" <m3devel at elegosoft.com>
Fecha: lunes, 10 de septiembre, 2012 00:47



Please, no.

Stock gdb and stock Visual C++ and stock cdb/ntsd/windbg ought to work well with Modula-3.
I expect they will pretty good pretty soon, and we should try to make them even better maybe...
by making globals visible in a reasonable fashion.


m3gdb isn't supported on Darwin and HP-UX (no stabs), at least.
If the gcc backend becomes obsolete and deleted...m3gdb would be the last of the
forked always-going-stale never-merging GPL stuff... i.e. another ripe target...



 - Jay

Subject: Re: [M3devel] "C-generating backend progress report, that nobody asked for" :)
From: hosking at cs.purdue.edu
Date: Sun, 9 Sep 2012 20:54:12 -0400
CC: dragisha at m3w.org; m3devel at elegosoft.com
To: jay.krell at cornell.edu

Stock debuggers are designed for C.  They generally will not be able to grok M3.  Better to use M3 run-time type information as per m3gdb.  See 10.1145/143103.143112.


On Sep 9, 2012, at 4:07 PM, Jay <jay.krell at cornell.edu> wrote:
The current method stinks for stock debugging. As I understand, globals are combined into structs & the fields all have generated names, & global records are flattened therein.

It works. But stock debuggers see garbage.

 - Jay (briefly/pocket-sized-computer-aka-phone)
On Sep 9, 2012, at 12:22 PM, Antony Hosking <hosking at cs.purdue.edu> wrote:

I mis-spoke.  I think LLVM can cope with much the same as we currently have.  Front-end will continue to compute alignments and layouts.  We can assert the front-end's datalayout by passing it to LLVM explicitly, which will complain if it does not match the actual target.  This is a rather nice feature of LLVM in that it allows us to retain control while having LLVM optimize accordingly.

On Sep 9, 2012, at 11:05 AM, Antony Hosking <hosking at cs.purdue.edu> wrote:
Yes, agreed, these need to be properly typed too.First step will be to lift the M3CG interface.Problem: The compiler needs to control layout so that the run-time system knows where to find things.  This means that we need to assert alignments and layouts produced by the backend are the same as those in the front-end.  LLVM has nice ways to do this.  How will we do it in the C backend?  Does C have sufficient control over alignment?
On Sep 9, 2012, at 10:53 AM, Jay <jay.krell at cornell.edu> wrote:
Btw, can this include "segment"/globals? Can they each be separate named variables? At least some of them?

 - Jay (briefly/pocket-sized-computer-aka-phone)
On Sep 9, 2012, at 7:48 AM, Antony Hosking <hosking at cs.purdue.edu> wrote:

I'm looking at it...
First step is to lift slightly the level of M3CG to use properly typed memory access, instead of untyped address + offset.

On Sep 9, 2012, at 2:32 AM, Dragiša Durić <dragisha at m3w.org> wrote:
I hope somebody will take on LLVM :).




 		 	   		  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://m3lists.elegosoft.com/pipermail/m3devel/attachments/20120910/597e1885/attachment-0002.html>