Hi!
On Sat, 2015-01-24 at 11:42:32 +0100, Hector Oron wrote:
> 2015-01-22 2:38 GMT+01:00 Hector Oron <hector.o...@gmail.com>:
> > Right that was planned for jessie+1.
> > It should happen on upcoming upload to experimental 7.8.2-1.
>
> Just wanted to note that the above has already happened.
Thanks, although it seems the important part (dropping the
Build-Depends) did not happen. :) I'm attaching an untested patch
doing that and removing the now unused code.
> >> There's also already a gdb-multiarch package which should be able to
> >> replace the gdb64 package in some circumstances. But for when that's
> >> still not enough, it would be possible to make the gdb package be
> >> Multi-Arch:same, and make it ship triplet-qualified filenames for
> >> files varying per arch, as in /usr/bin/i586-linux-gnu-gdb, which would
> >> allow to co-install them with ones from foreign architectures, like
> >> /usr/bin/x86_64-linux-gnu-gdb.
> >>
> >> The maintainer scripts could then use alternatives to create the gdb
> >> and other slave filenames, and assign a higher priority to the native
> >> architecture (the one defined by «dpkg --print-architecture»).
> >>
> >> In addition this would give a 64-bit package on 32-bit arches that's
> >> fully functional, as currently the gdb64 packages lack for example
> >> python support, among other things.
> >>
> >> What do you think?
> >
> > That should be good!
> > Could you provide a patch?
>
> Any news about a patch for adding triplet-qualified filenames? Do you
> have enough bandwidth/resources to work on it?
I can try to allocate some time to this, as I think it would be nice
to remove as much multilib stuff as we can to help simplify our
builds. I've started to have a look to see if this might be actually
trivially fixed in a sit down.
Thanks,
Guillem
From 80ecfb7b24ae6b55aa8f51a57ef268c78a4cd237 Mon Sep 17 00:00:00 2001
From: Guillem Jover <guil...@debian.org>
Date: Sat, 24 Jan 2015 18:30:26 +0100
Subject: [PATCH] Remove now unused multilib Build-Depends and code
---
debian/control | 7 +++----
debian/control.in | 7 +++----
debian/rules | 25 +------------------------
3 files changed, 7 insertions(+), 32 deletions(-)
diff --git a/debian/control b/debian/control
index c012d9c..28568f2 100644
--- a/debian/control
+++ b/debian/control
@@ -18,7 +18,6 @@ Build-Depends:
dejagnu,
flex | flex-old,
procps,
- g++-multilib [i386 powerpc s390 sparc],
# Do we really care that much about running the Java tests?
# gcj-jdk | gcj,
gobjc,
@@ -27,9 +26,9 @@ Build-Depends:
texinfo (>= 4.7-2.2),
texlive-base,
# Libdev deps
- libexpat1-dev, lib64expat1-dev [i386 powerpc s390 sparc],
- libncurses5-dev, lib64ncurses5-dev [i386 powerpc s390 sparc],
- libreadline-dev, lib64readline6-dev [i386 powerpc s390 sparc],
+ libexpat1-dev,
+ libncurses5-dev,
+ libreadline-dev,
zlib1g-dev,
liblzma-dev,
libbabeltrace-dev [amd64 armel armhf i386 kfreebsd-amd64 kfreebsd-i386 mips mipsel powerpc s390x],
diff --git a/debian/control.in b/debian/control.in
index 8a1e07c..7ad0033 100644
--- a/debian/control.in
+++ b/debian/control.in
@@ -18,7 +18,6 @@ Build-Depends:
dejagnu,
flex | flex-old,
procps,
- g++-multilib [i386 powerpc s390 sparc],
# Do we really care that much about running the Java tests?
# gcj-jdk | gcj,
gobjc,
@@ -27,9 +26,9 @@ Build-Depends:
texinfo (>= 4.7-2.2),
texlive-base,
# Libdev deps
- libexpat1-dev, lib64expat1-dev [i386 powerpc s390 sparc],
- libncurses5-dev, lib64ncurses5-dev [i386 powerpc s390 sparc],
- libreadline-dev, lib64readline6-dev [i386 powerpc s390 sparc],
+ libexpat1-dev,
+ libncurses5-dev,
+ libreadline-dev,
zlib1g-dev,
liblzma-dev,
libbabeltrace-dev [amd64 armel armhf i386 kfreebsd-amd64 kfreebsd-i386 mips mipsel powerpc s390x],
diff --git a/debian/rules b/debian/rules
index d1a5e62..8732563 100755
--- a/debian/rules
+++ b/debian/rules
@@ -66,25 +66,6 @@ ifdef GDB_TARGET
TP = $(DEB_TARGET_ALIAS)-
TS = -$(DEB_TARGET_ALIAS)
CROSS_FORCE = FORCE
-else
-ifneq (,$(filter $(DEB_HOST_ARCH),i386 powerpc sparc s390))
- build64 = yes
- CC64 = gcc -m64
- CFLAGS64 = $(CFLAGS)
- BUILDDIR64 = $(ALL_BUILDDIR)/objdir64
- ifeq ($(DEB_HOST_ARCH),i386)
- HOST64 = x86_64-linux-gnu
- endif
- ifeq ($(DEB_HOST_ARCH),powerpc)
- HOST64 = powerpc64-linux-gnu
- endif
- ifeq ($(DEB_HOST_ARCH),sparc)
- HOST64 = sparc64-linux-gnu
- endif
- ifeq ($(DEB_HOST_ARCH),s390)
- HOST64 = s390x-linux-gnu
- endif
-endif
endif
distribution := $(shell lsb_release -is)
@@ -140,10 +121,6 @@ EXTRA_FLAGS := --disable-gdbtk --disable-shared \
DEB_CONFIGURE_EXTRA_FLAGS := --host=$(DEB_HOST_GNU_TYPE) $(EXTRA_FLAGS) \
--enable-tui --with-zlib --with-lzma --with-python=python3
-# 64-bit flags
-DEB_CONFIGURE_FLAGS_64 := --host=$(HOST64) $(EXTRA_FLAGS) \
- --enable-tui --without-babeltrace
-
# multiarch targets; this is taken from the binutils-multiarch package but
# doesn't seem like a terribly nice list; see
# <20110117211551.ga7...@bee.dooz.org> for discussion -- locally updated
@@ -242,7 +219,7 @@ endif
touch $@
clean::
- rm -f check-stamp build64-stamp build-multiarch-stamp
+ rm -f check-stamp build-multiarch-stamp
rm -rf $(ALL_BUILDDIR)
if test -f gdb/version.in.backup; then \
--
2.2.1.209.g41e5f3a
