[RFC 1/2] glib: upgrade to 2.32.4

[RFC 1/2] glib: upgrade to 2.32.4

[Martin Jansa]

* it's bugfix release, see http://ftp.gnome.org/pub/gnome/sources/glib/2.32/glib-2.32.4.news
  for list of fixed bugs

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
---
 meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch |   11 +++++++----
 .../{glib-2.0_2.32.3.bb => glib-2.0_2.32.4.bb}     |    5 ++---
 2 files changed, 9 insertions(+), 7 deletions(-)
 rename meta/recipes-core/glib-2.0/{glib-2.0_2.32.3.bb => glib-2.0_2.32.4.bb} (92%)

diff --git a/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch b/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch
index 0149767..d310041 100644
--- a/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch
+++ b/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch
@@ -1,9 +1,12 @@
-From 0040842024725c0e74f84c1719ead6c79ba94aba Mon Sep 17 00:00:00 2001
+From e1252cbf65001487ff3b94bbc141acacf6e528b6 Mon Sep 17 00:00:00 2001
 From: Martin Jansa <Martin.Jansa@gmail.com>
 Date: Wed, 16 May 2012 20:03:34 +0200
 Subject: [PATCH] configure: add option to disable libelf support
 
-Upstream-Status: Pending
+Upstream-Status: Backport 
+Applied in master
+http://git.gnome.org/browse/glib/commit/?id=9e1d4da14d8e69fb60da1a9733ac1d58d0dcd8b2
+can be dropped when upgrading to 2.34.* or newer
 
 Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
 ---
@@ -11,7 +14,7 @@ Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
  1 files changed, 4 insertions(+), 0 deletions(-)
 
 diff --git a/configure.ac b/configure.ac
-index aa65648..1bb49bf 100644
+index ed4c10e..210df90 100644
 --- a/configure.ac
 +++ b/configure.ac
 @@ -1815,6 +1815,9 @@ AC_SUBST(XATTR_LIBS)
@@ -22,7 +25,7 @@ index aa65648..1bb49bf 100644
 +              AC_HELP_STRING([--disable-libelf], [build without libelf support]))
 +if test "x$enable_libelf" != "xno"; then
  
- PKG_CHECK_MODULES([LIBELF], [libelf], [have_libelf=yes], [have_libelf=maybe])
+ PKG_CHECK_MODULES([LIBELF], [libelf >= 0.8.12], [have_libelf=yes], [have_libelf=maybe])
  
 @@ -1835,6 +1838,7 @@ fi
  if test $have_libelf = yes; then
diff --git a/meta/recipes-core/glib-2.0/glib-2.0_2.32.3.bb b/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
similarity index 92%
rename from meta/recipes-core/glib-2.0/glib-2.0_2.32.3.bb
rename to meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
index 08aa897..fca5f8d 100644
--- a/meta/recipes-core/glib-2.0/glib-2.0_2.32.3.bb
+++ b/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
@@ -1,6 +1,5 @@
 require glib.inc
 
-PR = "r2"
 PE = "1"
 
 DEPENDS += "libffi python-argparse-native zlib"
@@ -16,8 +15,8 @@ SRC_URI = "${GNOME_MIRROR}/glib/${SHRT_VER}/glib-${PV}.tar.xz \
            file://nodbus.patch \
            file://nolibelf.patch \
           "
-SRC_URI[md5sum] = "9bd3ac535c24b6c162be13b2c43c5adc"
-SRC_URI[sha256sum] = "b65ceb462807e4a2f91c95e4293ce6bbefca308cb44a1407bcfdd9e40363ff4d"
+SRC_URI[md5sum] = "bf84fefd9c1a5b5a7a38736f4ddd674a"
+SRC_URI[sha256sum] = "a5d742a4fda22fb6975a8c0cfcd2499dd1c809b8afd4ef709bda4d11b167fae2"
 
 SRC_URI_append_virtclass-native = " file://glib-gettextize-dir.patch"
 BBCLASSEXTEND = "native nativesdk"
-- 
1.7.8.6

[RFC 2/2] glib: package bash-completion files in separate package and move ${PN}-codegen to .inc

[Martin Jansa]

* 2.32.4 has completion files in ${datadir}/bash-completion
  and dbus also packages bash-competion files in separate package

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
---
 meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb |    5 +----
 meta/recipes-core/glib-2.0/glib.inc           |    9 +++++++--
 2 files changed, 8 insertions(+), 6 deletions(-)

diff --git a/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb b/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
index fca5f8d..85352ed 100644
--- a/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
+++ b/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
@@ -1,5 +1,6 @@
 require glib.inc
 
+PR = "r1"
 PE = "1"
 
 DEPENDS += "libffi python-argparse-native zlib"
@@ -46,7 +47,3 @@ do_install_append() {
     sed -i -e '1s,#!.*perl,#! ${PERLPATH},' ${D}${bindir}/glib-mkenums
   fi
 }
-
-PACKAGES += "${PN}-codegen"
-FILES_${PN}-codegen = "${libdir}/gdbus-2.0/codegen/*.py"
-FILES_${PN} += "${datadir}/glib-2.0/gettext/mkinstalldirs ${datadir}/glib-2.0/gettext/po/Makefile.in.in"
diff --git a/meta/recipes-core/glib-2.0/glib.inc b/meta/recipes-core/glib-2.0/glib.inc
index e7d547a..9ae02d9 100644
--- a/meta/recipes-core/glib-2.0/glib.inc
+++ b/meta/recipes-core/glib-2.0/glib.inc
@@ -16,7 +16,7 @@ SECTION = "libs"
 DEPENDS = "glib-2.0-native gtk-doc-native zip virtual/libiconv"
 DEPENDS_virtclass-native = "gtk-doc-native pkgconfig-native gettext-native"
 DEPENDS_virtclass-nativesdk = "libtool-nativesdk"
-PACKAGES =+ "${PN}-utils "
+PACKAGES =+ "${PN}-utils ${PN}-bash-completion ${PN}-codegen"
 
 LEAD_SONAME = "libglib-2.0.*"
 FILES_${PN}-utils = "${bindir}/* ${datadir}/glib-2.0/gettext"
@@ -31,12 +31,17 @@ EXTRA_OECONF = "--disable-debug --enable-included-printf=no ${CORECONF}"
 EXTRA_OECONF_virtclass-native = "${CORECONF}"
 EXTRA_OECONF_linuxstdbase = "--enable-included-printf=no ${CORECONF}"
 
-FILES_${PN} = "${libdir}/lib*${SOLIBS} ${sysconfdir}/bash_completion.d ${datadir}/glib-2.0/schemas"
+
+FILES_${PN} = "${libdir}/lib*${SOLIBS} ${datadir}/glib-2.0/schemas \
+               ${datadir}/glib-2.0/gettext/mkinstalldirs ${datadir}/glib-2.0/gettext/po/Makefile.in.in"
 FILES_${PN}-dev += "${libdir}/glib-2.0/include \
                     ${libdir}/gio/modules/lib*${SOLIBSDEV} \
                     ${libdir}/gio/modules/*.la"
 FILES_${PN}-dbg += "${datadir}/glib-2.0/gdb ${datadir}/gdb \
                     ${libdir}/gio/modules/.debug"
+FILES_${PN}-codegen = "${libdir}/gdbus-2.0/codegen/*.py"
+FILES_${PN}-bash-completion = "${sysconfdir}/bash_completion.d \
+                               ${datadir}/bash-completion"
 
 ARM_INSTRUCTION_SET = "arm"
 USE_NLS = "yes"
-- 
1.7.8.6

Re: [RFC 1/2] glib: upgrade to 2.32.4

[Saul Wold]

On 07/16/2012 09:07 AM, Martin Jansa wrote:
> * it's bugfix release, see http://ftp.gnome.org/pub/gnome/sources/glib/2.32/glib-2.32.4.news
>    for list of fixed bugs
>
> Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
> ---
>   meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch |   11 +++++++----
>   .../{glib-2.0_2.32.3.bb => glib-2.0_2.32.4.bb}     |    5 ++---
>   2 files changed, 9 insertions(+), 7 deletions(-)
>   rename meta/recipes-core/glib-2.0/{glib-2.0_2.32.3.bb => glib-2.0_2.32.4.bb} (92%)
>
> diff --git a/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch b/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch
> index 0149767..d310041 100644
> --- a/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch
> +++ b/meta/recipes-core/glib-2.0/glib-2.0/nolibelf.patch
> @@ -1,9 +1,12 @@
> -From 0040842024725c0e74f84c1719ead6c79ba94aba Mon Sep 17 00:00:00 2001
> +From e1252cbf65001487ff3b94bbc141acacf6e528b6 Mon Sep 17 00:00:00 2001
>   From: Martin Jansa <Martin.Jansa@gmail.com>
>   Date: Wed, 16 May 2012 20:03:34 +0200
>   Subject: [PATCH] configure: add option to disable libelf support
>
> -Upstream-Status: Pending
> +Upstream-Status: Backport
> +Applied in master
> +http://git.gnome.org/browse/glib/commit/?id=9e1d4da14d8e69fb60da1a9733ac1d58d0dcd8b2
> +can be dropped when upgrading to 2.34.* or newer
>
>   Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
>   ---
> @@ -11,7 +14,7 @@ Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
>    1 files changed, 4 insertions(+), 0 deletions(-)
>
>   diff --git a/configure.ac b/configure.ac
> -index aa65648..1bb49bf 100644
> +index ed4c10e..210df90 100644
>   --- a/configure.ac
>   +++ b/configure.ac
>   @@ -1815,6 +1815,9 @@ AC_SUBST(XATTR_LIBS)
> @@ -22,7 +25,7 @@ index aa65648..1bb49bf 100644
>   +              AC_HELP_STRING([--disable-libelf], [build without libelf support]))
>   +if test "x$enable_libelf" != "xno"; then
>
> - PKG_CHECK_MODULES([LIBELF], [libelf], [have_libelf=yes], [have_libelf=maybe])
> + PKG_CHECK_MODULES([LIBELF], [libelf >= 0.8.12], [have_libelf=yes], [have_libelf=maybe])
>
>   @@ -1835,6 +1838,7 @@ fi
>    if test $have_libelf = yes; then
> diff --git a/meta/recipes-core/glib-2.0/glib-2.0_2.32.3.bb b/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
> similarity index 92%
> rename from meta/recipes-core/glib-2.0/glib-2.0_2.32.3.bb
> rename to meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
> index 08aa897..fca5f8d 100644
> --- a/meta/recipes-core/glib-2.0/glib-2.0_2.32.3.bb
> +++ b/meta/recipes-core/glib-2.0/glib-2.0_2.32.4.bb
> @@ -1,6 +1,5 @@
>   require glib.inc
>
> -PR = "r2"
>   PE = "1"
>
>   DEPENDS += "libffi python-argparse-native zlib"
> @@ -16,8 +15,8 @@ SRC_URI = "${GNOME_MIRROR}/glib/${SHRT_VER}/glib-${PV}.tar.xz \
>              file://nodbus.patch \
>              file://nolibelf.patch \
>             "
> -SRC_URI[md5sum] = "9bd3ac535c24b6c162be13b2c43c5adc"
> -SRC_URI[sha256sum] = "b65ceb462807e4a2f91c95e4293ce6bbefca308cb44a1407bcfdd9e40363ff4d"
> +SRC_URI[md5sum] = "bf84fefd9c1a5b5a7a38736f4ddd674a"
> +SRC_URI[sha256sum] = "a5d742a4fda22fb6975a8c0cfcd2499dd1c809b8afd4ef709bda4d11b167fae2"
>
>   SRC_URI_append_virtclass-native = " file://glib-gettextize-dir.patch"
>   BBCLASSEXTEND = "native nativesdk"
>

This patch set was merged into OE-Core, but it appears to cause a 
problem with the gdbus-codegen getting auto-tools substitution with the 
wrong PYTHON.

It seems to be getting the native sysroot python instead of the more 
correct "#! /usr/bin/env python", this is causing a dependency failure.

If you could look at this please.

Thanks
	Sau!

Re: [RFC 1/2] glib: upgrade to 2.32.4

[Colin Walters]

On Tue, 2012-07-17 at 17:34 -0700, Saul Wold wrote:

> It seems to be getting the native sysroot python instead of the more 
> correct "#! /usr/bin/env python", this is causing a dependency failure.
> 
> If you could look at this please.

I ran into this a while ago with flex/bison:

https://lists.yoctoproject.org/pipermail/poky/2012-January/007403.html

Apparently the "correct" fix is to set an environment variable like e.g.
PYTHON="/usr/bin/env python" ./configure --prefix=foo ?

That seems fairly gross though.  Some more generic solution would be
nice...I don't have any great ideas offhand though.  I wonder if there's
any actual use case for hardcoding the paths found at build time into
the generated binaries.

See also:
https://bugzilla.gnome.org/show_bug.cgi?id=650763
https://bugzilla.gnome.org/show_bug.cgi?id=670718

Re: [RFC 1/2] glib: upgrade to 2.32.4

[Richard Purdie]

On Tue, 2012-07-17 at 20:47 -0400, Colin Walters wrote:
> On Tue, 2012-07-17 at 17:34 -0700, Saul Wold wrote:
> 
> > It seems to be getting the native sysroot python instead of the more 
> > correct "#! /usr/bin/env python", this is causing a dependency failure.
> > 
> > If you could look at this please.
> 
> I ran into this a while ago with flex/bison:
> 
> https://lists.yoctoproject.org/pipermail/poky/2012-January/007403.html
> 
> Apparently the "correct" fix is to set an environment variable like e.g.
> PYTHON="/usr/bin/env python" ./configure --prefix=foo ?
> 
> That seems fairly gross though.  Some more generic solution would be
> nice...I don't have any great ideas offhand though.  I wonder if there's
> any actual use case for hardcoding the paths found at build time into
> the generated binaries.

Generally we kind of cheat and encode "/usr/bin/env xxx" into
native/nativesdk scripts and ${bindir}/env xxx into target scripts.

This relies on /usr/bin/env being available but so far its been the best
option we've found as yet.

Cheers,

Richard

Re: [RFC 1/2] glib: upgrade to 2.32.4

[Colin Walters]

On Wed, 2012-07-18 at 19:13 +0100, Richard Purdie wrote:

> Generally we kind of cheat and encode "/usr/bin/env xxx" into
> native/nativesdk scripts and ${bindir}/env xxx into target scripts.

How is that accomplished?  Can you link me an example?

Re: [RFC 1/2] glib: upgrade to 2.32.4

[Richard Purdie]

On Wed, 2012-07-18 at 14:18 -0400, Colin Walters wrote:
> On Wed, 2012-07-18 at 19:13 +0100, Richard Purdie wrote:
> 
> > Generally we kind of cheat and encode "/usr/bin/env xxx" into
> > native/nativesdk scripts and ${bindir}/env xxx into target scripts.
> 
> How is that accomplished?  Can you link me an example?

Basically using sed. I didn't claim it was pretty, it does get out of
various immediate problems. You can see the scope of the problem in:

http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=ac7a0311825e20c544d17bfd8be63546ad36e665

We're pretty sensitive to picking up that kind of issue.

Cheers,

Richard