RE: [ECOS] Recommended Version of GCC

RE: [ECOS] Recommended Version of GCC

[Jay Foster]

For what it's worth, I too am using gcc 3.4.3 for ARM7TDMI.  I switched from
3.2.1 because 3.2.1 had a bug that generated incorrect ARM code.  I forget
the details now.  Inlining also works with 3.4.3.

Jay

-----Original Message-----
From: Birahim Larou Fall [mailto:BLFall@scmmicro.fr]
Sent: Friday, May 12, 2006 2:05 AM
To: Moussa A. Ba
Cc: ecos-discuss@ecos.sourceware.org
Subject: Re: [ECOS] Recommended Version of GCC


I use this version (gcc 3.4.3) for ARM7TDMI and ARM9TDMI:
$ arm-elf-gcc --version
arm-elf-gcc (GCC) 3.4.3
Copyright (C) 2004 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. 
I have no problem with this version. May be you can find better.

Fall Birahim




"Moussa A. Ba" <mba@embedded-zone.com> 
Sent by: ecos-discuss-owner@ecos.sourceware.org
11/05/2006 18:29

To
ecos-discuss@ecos.sourceware.org
cc

Subject
[ECOS] Recommended Version of GCC






I am about to port ecos to an Oki eval board, ARM7TDMI, which version of 
GCC do you recommend I use.  Should I just use the one distributed from 
the stable ecos release or can I use a more recent version from 
www.gnuarm.com or codesourcery.com

Moussa

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss




-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

[ECOS] Two minor memory allocation issues

[Andy Jackson]

I've just been experimenting with memory allocation (malloc and friends) on 
eCos and have spotted two minor issues:

1. A malloc request for 0x7FFFFFFF bytes succeeds and allocates a small 
amount of memory. Reducing the request to 0x7FFFFFF0 causes it to fail as 
expected. My guess is that this is a function of the allocation padding 
calculation, but I haven't got time at present to investigate further.

2. In the reference manual section 'eCos Memory Pools' on the page for 
'stdlib malloc Pools, the prototype for calloc is given as:

void *calloc(size_t size);

but looking in stdlib.h it appears that it should be:

void *calloc(size_t nmemb, size_t size);

    Andy..



-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss

Re: [ECOS] Two minor memory allocation issues

[Andrew Lunn]

On Tue, May 16, 2006 at 11:15:02AM +0100, Andy Jackson wrote:
> I've just been experimenting with memory allocation (malloc and friends) on 
> eCos and have spotted two minor issues:
> 
> 1. A malloc request for 0x7FFFFFFF bytes succeeds and allocates a small 
> amount of memory. Reducing the request to 0x7FFFFFF0 causes it to fail as 
> expected. My guess is that this is a function of the allocation padding 
> calculation, but I haven't got time at present to investigate further.

It looks like internally dlmalloc is using cyg_int32 to represent
size. There is a macro request2size() which does arithmatic on the
size and 0x7FFFFFFF it is probably wrapping around to negative and so
it allocates the minimal amount of space. 

 
> 2. In the reference manual section 'eCos Memory Pools' on the page for 
> 'stdlib malloc Pools, the prototype for calloc is given as:
> 
> void *calloc(size_t size);

Yes, this is wrong. I will update the documentation.

     Andrew

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss