RE: [PATCH newlib]An alternative implementation of malloc family's functions with small foot-print

["Bin Cheng" <bin dot cheng at arm dot com>]

> -----Original Message-----
> From: newlib-owner@sourceware.org [mailto:newlib-owner@sourceware.org] On
> Behalf Of Corinna Vinschen
> Sent: Tuesday, May 28, 2013 6:45 PM
> To: newlib@sourceware.org
> Subject: Re: [PATCH newlib]An alternative implementation of malloc family's
> functions with small foot-print
> 
> On May 13 15:41, Bin Cheng wrote:
> > Hi all,
> >
> > As a part of our library work in GNU Tools for ARM Embedded
> > Processors, we implemented an alternative implementation of malloc
> > family's functions with small foot-print.
> >
> > We have discussed before how to integrate the code with newlib and it
> > seems reasonable to have a configuration option to control it.
> >
> > So this patch introduces new implementation of malloc functions in
> > file "nano-mallocr.c" and a new option "--enable-newlib-nano-malloc"
> > to control it. The option is disabled by default so it won't change
> > the default behavior.
> >
> > We have done lots of internal review work and the code has been
> > applied in GNU Tools for ARM Embedded Processor and used for a while,
> > is it OK? Any comments will be highly appreciated.
> >
> > As for the documentation, I will try to describe newlib's
> > configuration options including this one in readme later.
> 
> Two comments:
> 
> > +  --enable-newlib-nano-malloc    use nano version malloc implementation
> 
> Using the word "nano" here *might* imply that the malloc implementation is a
> low-size one, but it isn't really clear, IMHO.  Wouldn't something like "tiny"
> be more clear?  Alternatively, if you like the nano name, I think it would be
> helpful to at least add to the description here, along the lines of
> 
>   --enable-newlib-nano-malloc    use small-footprint nano-malloc
> implementation

Since we have used the word "nano" elsewhere, so the name is kept in case any confusion. The description has been modified as your request.

> 
> > +#ifndef max
> > +#define max(a,b) ((a) >= (b) ? (a) : (b)) #endif
> 
> I'd prefer if you just include <sys/param.h> and use the uppercase MAX() macro
> instead.
Unfortunately I has to keep the macro definition in nano-mallocr.c because <sys/param.h> is overridden by target specific one, which may do not define MAX (for example, on ARM).
The macro name is modified into uppercase as your request.


> Other than that, I think this is ok for inclusion.
> 
The new patch is attached, is it OK?

Thanks.

Attachment: nano-mallocr-20130529.cvs.txt
Description: Text document