bug-make
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: GNU Make: A Program for Directing Recompilation, GNU make Version 3.


From: Dave Korn
Subject: RE: GNU Make: A Program for Directing Recompilation, GNU make Version 3.81, April 2006 Documentation
Date: Sun, 6 May 2007 18:17:12 +0100

On 06 May 2007 14:50, Bob wrote:

> Some documentation issues for GNU Make: A Program for Directing
> Recompilation, GNU make Version 3.81, April 2006.
> 
> 1) On page 34, Section 4.8 Special Built-in Target Names, there's
> repeated text. The text under ".SUFFIXES" is also the second paragraph
> under ".SECONDEXPANSION".

  So it is!

> 2) On page 87, Section 8.8 The eval Function, there's a problem with the
> code example if "make clean" is issued, because the $(ALL_OBJS) variable
> in the rm command under clean, won't be initialized (and I'm curious how
> you'd do it).

  ALL_OBJS is /always/ initialised.  The foreach always runs, the ALL_OBJS+=
statements generated by the call PROGRAM_template are always eval'd, where do
you see a problem?

> 3) On page 106, Section 10.3 Variables Used by Implicit Rules, there's
> repeated text. The CO entry is repeated twice, once after CC and another
> time after CXX.

  Yep, it sure is.
 
> 4) A general recommendation, when lists are given, such as on page 106,
> Section 10.3, and the order of the entries is arbitrary, sort them in
> alphabetic sequence.

  Well, they aren't entirely arbitrary, they're roughly grouped into similar
functionality:

(binutils AR, AS)
(compiler CC, CXX, CPP, FC)
(rcs CO, GET)
(parsing LEX, YACC, LINT)
(couple more less-used complers M2C, PC)
(docs build tools MAKEINFO, TEX, TEXI2DVI, WEAVE, CWEAVE, TANGLE, CTANGLE)
(misc RM)

..with CO doubled and slightly misplaced.  In that particular case, straight
alpha order would scatter related things all over the place.  I'd suggest that
for this one, we move the CO entry so it's just before GET (kill the dup), and
perhaps bring M2C and PC back up to join the other language compilers.

  Attached patch does just that, and fixes 1) as well.


    cheers,
      DaveK
-- 
Can't think of a witty .sigline today....

Attachment: docs-patch.diff
Description: Binary data


reply via email to

[Prev in Thread] Current Thread [Next in Thread]