Roger Leigh wrote:
Sorry, wrong version of second patch; I've attached the correct one.
This was pretty close. I've cleaned it up in the smarter-targets
branch, including inlining away recursive calls to debian/rules for
implicit targets. See what you think. Note: I have not decided if this
is
On Mon, Jun 13, 2011 at 05:30:53PM -0400, Joey Hess wrote:
Roger Leigh wrote:
Sorry, wrong version of second patch; I've attached the correct one.
This was pretty close. I've cleaned it up in the smarter-targets
branch, including inlining away recursive calls to debian/rules for
implicit
Roger Leigh wrote:
While I've not yet had time to debug it, the attached rules file
You forgot the attachement.
This is due to calling rules_explicit_target before you check if the
sequence is valid (debhelper-fail-me) which leads to infinite
forking since each dh runs rules_explicit_target
Roger Leigh wrote:
Good point. In the attached patch I've retained the scheme for
the build and install targets, where it serves a purpose, and
dropped it for binary.
$sequences{binary} = ['rules:install', @ba, @b];
$sequences{'binary-indep'} = ['rules:install-indep', @b];
Modestas Vainius wrote:
But this all seems like re-implementation of make in perl
Well yes, that's essentially what dh has always been guilty of. :)
Anyway, I guess it is a good thing to be policy complaint. So I suggest
adding
a dh restriction that standard targets have to be explicit.
Hi Joey,
On Sun, Jun 12, 2011 at 11:47:54AM -0400, Joey Hess wrote:
Roger Leigh wrote:
Good point. In the attached patch I've retained the scheme for
the build and install targets, where it serves a purpose, and
dropped it for binary.
$sequences{binary} = ['rules:install', @ba, @b];
On Sun, Jun 12, 2011 at 08:28:01PM +0100, Roger Leigh wrote:
On Sun, Jun 12, 2011 at 11:47:54AM -0400, Joey Hess wrote:
One possible approach here would be to call
rules_explicit_target(build-(arch|indep)) and and then if they exist,
we can skip the entire build sequence up to invoking
Hello,
On sekmadienis 12 Birželis 2011 03:14:08 Joey Hess wrote:
Modestas Vainius wrote:
Personally I don't understand what's the issue here. If debian/rules has
no explicit build,install,binary etc. targets (which will continue to be
= 90% of cases), current dh implementation works just
On Sat, Jun 11, 2011 at 07:30:38PM -0400, Joey Hess wrote:
Roger Leigh wrote:
Notice that the initial 'debian/rules install' performed
by dh binary isn't repeated on the second invocation. But AFAICS
there's no .log file that would make dh skip this, so I'm unsure
why it's skipped. The
On Sat, Jun 11, 2011 at 08:33:19PM -0400, Joey Hess wrote:
Roger Leigh wrote:
+my @bd = qw{
dh_testdir
dh_auto_configure
dh_auto_build
dh_auto_test
-}],
-$sequences{'build-indep'} = [@{$sequences{build}}];
-$sequences{'build-arch'} = [@{$sequences{build}}];
+};
On Sat, Jun 11, 2011 at 08:19:32PM -0400, Joey Hess wrote:
Roger Leigh wrote:
+If running a configure script, it may be necessary to prevent it being
+run twice, once for architecture-independent packages, and again for
+architecture-dependent packages. This may be accomplished by
On Tue, Jun 07, 2011 at 03:52:01PM +0100, Roger Leigh wrote:
On Tue, Jun 07, 2011 at 01:47:41PM +0100, Roger Leigh wrote:
On Tue, Jun 07, 2011 at 08:12:15AM -0400, Joey Hess wrote:
Roger Leigh wrote:
I think I may have proposed something similar to this patch before, but
I can't find
Hello,
On sekmadienis 12 Birželis 2011 00:10:34 Roger Leigh wrote:
On Tue, Jun 07, 2011 at 03:52:01PM +0100, Roger Leigh wrote:
On Tue, Jun 07, 2011 at 01:47:41PM +0100, Roger Leigh wrote:
On Tue, Jun 07, 2011 at 08:12:15AM -0400, Joey Hess wrote:
Roger Leigh wrote:
I think I may
Roger Leigh wrote:
Note that there's one issue with the code that confuses me. This might
be a pre-existing issue with dh that the patch exposes:
% fakeroot dh binary --no-act
debian/rules install
debian/rules override_dh_strip
dh_makeshlibs
dh_shlibdeps
dh_installdeb
Modestas Vainius wrote:
This has a potential to go very deep with make (binary) - perl (dh) - make
(install) - perl (dh) - make (build) - perl (dh). We have already seen
what negative effect this has on e.g. make -jN. IMHO, bugs will be hard to
understand and fix with such level of
Roger Leigh wrote:
+If running a configure script, it may be necessary to prevent it being
+run twice, once for architecture-independent packages, and again for
+architecture-dependent packages. This may be accomplished by
+overriding Ldh_autoconfigure(1):
+
+
Roger Leigh wrote:
+my @bd = qw{
dh_testdir
dh_auto_configure
dh_auto_build
dh_auto_test
-}],
-$sequences{'build-indep'} = [@{$sequences{build}}];
-$sequences{'build-arch'} = [@{$sequences{build}}];
+};
+$sequences{build} = [@bd, 'rules:build-arch',
Roger Leigh wrote:
I think I may have proposed something similar to this patch before, but
I can't find it in the BTS.
It caused significant slowdown (due to running every debhelper command
twice with -a and -i) and quite likely broke packages due to
running make install twice, so was removed.
On Tue, Jun 07, 2011 at 08:12:15AM -0400, Joey Hess wrote:
Roger Leigh wrote:
I think I may have proposed something similar to this patch before, but
I can't find it in the BTS.
It caused significant slowdown (due to running every debhelper command
twice with -a and -i) and quite likely
On Tue, Jun 07, 2011 at 01:47:41PM +0100, Roger Leigh wrote:
On Tue, Jun 07, 2011 at 08:12:15AM -0400, Joey Hess wrote:
Roger Leigh wrote:
I think I may have proposed something similar to this patch before, but
I can't find it in the BTS.
It caused significant slowdown (due to
Package: debhelper
Version: 8.1.6
Severity: normal
Tags: patch
Hi Joey,
I think I may have proposed something similar to this patch before, but
I can't find it in the BTS. This was something I was working on last
year WRT the build-arch/build-indep patches you applied already.
Possibly it was
21 matches
Mail list logo