linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Andrew Morton <andrewm@uow.edu.au>
To: Keith Owens <kaos@ocs.com.au>
Cc: lkml <linux-kernel@vger.kernel.org>
Subject: 2.4.2-pre1 mkdep and symlinked $TOPDIR
Date: Sun, 04 Mar 2001 16:24:57 +1100	[thread overview]
Message-ID: <3AA1D1A9.3A35557A@uow.edu.au> (raw)

Keith,

I do builds in /usr/src/linux, which is a symlink
to /usr/src/linux-akpm.  The recent `mkdep' changes
have broken this practice most horridly.  When searching
.hdepend, `make' doesn't recognise that nested headers
have changed. This is because .hdepend has things like

/usr/src/linux/include/asm/byteorder.h: \
   /usr/src/linux-akpm/include/asm/types.h \

So when it looks for dependencies for
	/usr/src/linux-akpm/include/asm/byteorder.h

it doesn't find anything.  In earlier kernels, the
whole toolchain honoured the symlink form of the name.

The net effect of all this is that changes to included
headers are *not* causing recompiles, and use of a
symlinked $TOPDIR is not feasible.

The workaround at this time is to do a `cd' to the
real $TOPDIR directory before starting development, 
rather than a `cd' to the symlink.

If it is your intention that everything in the
build system use the "real" pathname then I suggest
we need to make it happen consistently.  Or can we
restore the old behaviour?



--- linux-2.4.3-pre1/scripts/mkdep.c	Sat Mar  3 20:52:24 2001
+++ linux-akpm/scripts/mkdep.c	Sun Mar  4 16:16:00 2001
@@ -218,19 +218,12 @@
 void add_path(const char * name)
 {
 	struct path_struct *path;
-	char resolved_path[PATH_MAX+1];
 	const char *name2;
 
-	if (strcmp(name, ".")) {
-		name2 = realpath(name, resolved_path);
-		if (!name2) {
-			fprintf(stderr, "realpath(%s) failed, %m\n", name);
-			exit(1);
-		}
-	}
-	else {
+	if (strcmp(name, "."))
+		name2 = name;
+	else
 		name2 = "";
-	}
 
 	path_array = realloc(path_array, (++paths)*sizeof(*path_array));
 	if (!path_array) {

             reply	other threads:[~2001-03-04  5:25 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-03-04  5:24 Andrew Morton [this message]
2001-03-05  0:21 ` 2.4.2-pre1 mkdep and symlinked $TOPDIR Keith Owens

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=3AA1D1A9.3A35557A@uow.edu.au \
    --to=andrewm@uow.edu.au \
    --cc=kaos@ocs.com.au \
    --cc=linux-kernel@vger.kernel.org \
    /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.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).