[m-rev.] for prelim review: "Recursive make considered harmful" support

[Peter Ross]

On Wed, Nov 09, 2005 at 10:28:46AM +1100, Ralph Becket wrote:
> Peter Ross, Tuesday,  8 November 2005:
> > 
> > Estimated hours taken: 4
> > Branches: main, release
> > 
> > Implement support for writing Mmakefile in the style of
> > "Recursive make considered harmful".
> > 
> > This consists of two parts, getting mmake to ignore the .d files in the
> > current directory.  The second part is adding support to the compiler to
> > write out just the .d file.
> 
> Can you include some explanation for why these changes to .d file
> handling are necessary?
> 
How about:
	We need to ignore the .d files in the current directory because
	the correct .d file needs to in the same directory as the .m
	file, and if we don't ignore these files we get rules being
	defined twice once from the .d file in the current directory and
	once in the directory where the .m is located.

This change would not be needed if one can show me how to write a
generic form of the make rule using "%.d: %.m"

eg I want to have the following.

main.d: ant/main.m
	mmc --generate-dependencies ant/main.m

For me the second part is self explanatory, sometimes we want the
ability to just write the .d file of the current module, and do no other
processing on the module.

> > scripts/mmake.in:
> > 	If the environment variable MMAKE_USE_DS is set to no then don't
> > 	include the ds in the temporary Mmakefile.
> 
> By `ds' I presume you mean the .d files?  This isn't clear.
> 
Yes, do you have a better name for the environment variable?
--------------------------------------------------------------------------
mercury-reviews mailing list
post:  mercury-reviews at cs.mu.oz.au
administrative address: owner-mercury-reviews at cs.mu.oz.au
unsubscribe: Address: mercury-reviews-request at cs.mu.oz.au Message: unsubscribe
subscribe:   Address: mercury-reviews-request at cs.mu.oz.au Message: subscribe
--------------------------------------------------------------------------