From: Eli Zaretskii <eliz@gnu.org>
To: Paul Eggert <eggert@cs.ucla.edu>
Cc: 21707@debbugs.gnu.org
Subject: bug#21707: include-file cleanup for src directory
Date: Tue, 20 Oct 2015 22:15:54 +0300 [thread overview]
Message-ID: <83vba1e4gl.fsf@gnu.org> (raw)
In-Reply-To: <56268CFA.6040708@cs.ucla.edu>
> Cc: 21707@debbugs.gnu.org
> From: Paul Eggert <eggert@cs.ucla.edu>
> Date: Tue, 20 Oct 2015 11:50:34 -0700
>
> Eli Zaretskii wrote:
> > Would it help if we avoid including any of our headers in any
> > other of our headers, so that the headers included by a particular C
> > file are visible by just looking at that single C file?
>
> That would run afoul of a more important design goal, which is that we should be
> able to include headers in any order.
Is that really more important? Even standard C headers sometimes
require order, for example, sys/types.h should be included before
sys/stat.h.
> And anyway, compiling a typical C file ordinarily brings in oodles
> of system headers that the C file doesn't explicitly ask for
That is true, but C headers change much slower than our headers. So
keeping track of the changes in our headers requires more attention
and work.
> so regardless of our style the only practical way to see what headers a
> C file includes is to compile it and see what gcc -E outputs.
The problem with "gcc -E" is that it's system dependent. I thought we
could perhaps come up with a method that would allow anyone easily see
which headers are needed on every supported platform, and thus lower
the probability of unintended breakage.
Of course, it isn't a catastrophe to continue the way we have been
doing this till now.
> Ideally a C source file should include all headers that define symbols the
> source file directly uses, and no headers other than that. This should be true
> for both .h and .c files. We're not there by a long shot, but that should be the
> goal.
I think the real problem is to keep this state once we get there.
One-time efforts to rectify things are relatively easy; making that
stick is harder, especially since (AFAIU) there's no automated way for
producing a report about unnecessary headers.
next prev parent reply other threads:[~2015-10-20 19:15 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-10-19 7:05 bug#21707: include-file cleanup for src directory Paul Eggert
2015-10-19 7:23 ` Eli Zaretskii
2015-10-19 12:54 ` Andy Moreton
2015-10-19 13:23 ` Eli Zaretskii
2015-10-20 5:54 ` Paul Eggert
2015-10-20 8:23 ` Andy Moreton
2015-10-20 15:06 ` Eli Zaretskii
2015-10-20 15:06 ` Eli Zaretskii
2015-10-20 18:50 ` Paul Eggert
2015-10-20 19:15 ` Eli Zaretskii [this message]
2015-10-20 20:52 ` Paul Eggert
2015-10-20 19:28 ` Andy Moreton
2015-10-21 1:38 ` Paul Eggert
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=83vba1e4gl.fsf@gnu.org \
--to=eliz@gnu.org \
--cc=21707@debbugs.gnu.org \
--cc=eggert@cs.ucla.edu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
Code repositories for project(s) associated with this external index
https://git.savannah.gnu.org/cgit/emacs.git
https://git.savannah.gnu.org/cgit/emacs/org-mode.git
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.