Random fixes from the Mentor Graphics push queue

Random fixes from the Mentor Graphics push queue

[Christopher Larson]

The following changes since commit f991d2d60b74f5ebd990f77aecd3324b1a4533e9:

  libpng: set reasonable SUMMARY (2013-11-27 11:51:08 +0000)

are available in the git repository at:

  https://github.com/kergoth/oe-core.git random-fixes
  https://github.com/kergoth/oe-core/tree/random-fixes

Christopher Larson (7):
  update-rc.d: process symlinks recursively
  quota: apply patch to obey tcp-wrappers config
  python, python-native: fix PARALLEL_MAKEINST failure
  cairo: add/use packageconfig for valgrind support
  perf: remove /usr/local/include from default makefile includes
  qemu: handle CLOEXEC/NONBLOCK if unavailable on host
  pulseaudio: fix RDEPENDS traversal for consolekit

 .../update-rc.d/check-if-symlinks-are-valid.patch  | 26 +++---
 .../recipes-devtools/python/python-native_2.7.3.bb |  1 +
 .../python/parallel-makeinst-create-bindir.patch   | 19 +++++
 meta/recipes-devtools/python/python_2.7.3.bb       |  1 +
 ...Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch | 92 ++++++++++++++++++++++
 meta/recipes-devtools/qemu/qemu.inc                |  1 +
 .../quota/quota/config-tcpwrappers.patch           | 75 ++++++++++++++++++
 meta/recipes-extended/quota/quota_4.01.bb          |  3 +-
 meta/recipes-graphics/cairo/cairo.inc              |  1 +
 meta/recipes-kernel/perf/perf.bb                   | 11 +--
 meta/recipes-multimedia/pulseaudio/pulseaudio.inc  |  6 +-
 11 files changed, 219 insertions(+), 17 deletions(-)
 create mode 100644 meta/recipes-devtools/python/python/parallel-makeinst-create-bindir.patch
 create mode 100644 meta/recipes-devtools/qemu/files/linux-user-Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch
 create mode 100644 meta/recipes-extended/quota/quota/config-tcpwrappers.patch

-- 
1.8.3.4

[PATCH 1/7] update-rc.d: process symlinks recursively

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

When processing startup scripts which use update-alternatives, we need to keep
resolving the symlink recursively until we hit a real file, due to the
alternatives indirection. This fixes the ability to run certain postinsts at
do_rootfs time, which is needed for good read-only-rootfs support.

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 .../update-rc.d/check-if-symlinks-are-valid.patch  | 26 +++++++++++++---------
 1 file changed, 16 insertions(+), 10 deletions(-)

diff --git a/meta/recipes-core/update-rc.d/update-rc.d/check-if-symlinks-are-valid.patch b/meta/recipes-core/update-rc.d/update-rc.d/check-if-symlinks-are-valid.patch
index 6f402dd..075171a 100644
--- a/meta/recipes-core/update-rc.d/update-rc.d/check-if-symlinks-are-valid.patch
+++ b/meta/recipes-core/update-rc.d/update-rc.d/check-if-symlinks-are-valid.patch
@@ -14,25 +14,31 @@ actually exists in rootfs path and then continue.
 Upstream-Status: Pending
 
 Signed-off-by: Laurentiu Palcu <laurentiu.palcu@intel.com> 
+Signed-off-by: Christopher Larson <chris_larson@mentor.com>
 
 Index: git/update-rc.d
 ===================================================================
---- git.orig/update-rc.d	2013-01-16 12:12:58.349814356 +0200
-+++ git/update-rc.d	2013-01-16 13:02:42.490864939 +0200
-@@ -147,13 +147,29 @@
+--- git.orig/update-rc.d
++++ git/update-rc.d
+@@ -147,13 +147,34 @@ fi
  bn=$1
  shift
  
 +sn=$initd/$bn
 +if [ -L "$sn" -a -n "$root" ]; then
-+	readlink=$(which readlink)
++	if which readlink >/dev/null; then
++		while true; do
++			linksn="$(readlink "$sn")"
++			if [ -z "$linksn" ]; then
++				break
++			fi
 +
-+	if [ -n "$readlink" ]; then
-+		sn=$($readlink "$sn")
-+		case "$sn" in
-+			/*) sn=${root}${sn} ;;
-+			*)  sn=$initd/$sn ;;
-+		esac
++			sn="$linksn"
++			case "$sn" in
++				/*) sn="$root$sn" ;;
++				*)  sn="$initd/$sn" ;;
++			esac
++		done
 +	else
 +		echo "update-rc.d: readlink tool not present, cannot check whether \
 +				$sn symlink points to a valid file." >&2
-- 
1.8.3.4

[PATCH 2/7] quota: apply patch to obey tcp-wrappers config

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

Without this, the tcpwrappers argument wasn't obeyed, and as such the build
wasn't as deterministic as we'd prefer.

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 .../quota/quota/config-tcpwrappers.patch           | 75 ++++++++++++++++++++++
 meta/recipes-extended/quota/quota_4.01.bb          |  3 +-
 2 files changed, 77 insertions(+), 1 deletion(-)
 create mode 100644 meta/recipes-extended/quota/quota/config-tcpwrappers.patch

diff --git a/meta/recipes-extended/quota/quota/config-tcpwrappers.patch b/meta/recipes-extended/quota/quota/config-tcpwrappers.patch
new file mode 100644
index 0000000..94dc3f5
--- /dev/null
+++ b/meta/recipes-extended/quota/quota/config-tcpwrappers.patch
@@ -0,0 +1,75 @@
+Upstream-status: Pending
+
+--- quota-tools.orig/configure.in
++++ quota-tools/configure.in
+@@ -151,33 +151,46 @@ AC_SUBST(QUOTA_NETLINK_PROG)
+ AC_SUBST(NETLINKLIBS)
+
+ AC_SEARCH_LIBS(gethostbyname, nsl)
+-AC_MSG_CHECKING(for host_access in -lwrap)
+-AC_CACHE_VAL(ac_cv_lib_wrap_main,
+-	saved_LIBS="$LIBS"
+-	LIBS="$LIBS -lwrap"
+-	[AC_LINK_IFELSE([AC_LANG_PROGRAM([[
+-	#include <stdio.h>
+-	#include <sys/types.h>
+-	#include <sys/socket.h>
+-	#include <netinet/in.h>
+-	#include <tcpd.h>
+-	struct request_info request;
+-	int deny_severity, allow_severity;]],[[hosts_access(&request);]])],
+-	dnl We always restore LIBS as we add -lwrap in the next check
+-	[ac_cv_lib_wrap_main=yes; LIBS="$saved_LIBS"; AC_MSG_RESULT(yes)],
+-	[ac_cv_lib_wrap_main=no; LIBS="$saved_LIBS"; AC_MSG_RESULT(no)])
+-	])
+
+-if test ${ac_cv_lib_wrap_main} = yes; then
+-	AC_CHECK_HEADER(tcpd.h,, [
+-		echo 'ERROR: could not find tcpd.h - missing TCP wrappers package'
+-		exit 1
+-	])
+-	LIBS="$LIBS -lwrap"
+-	AC_DEFINE([HOSTS_ACCESS], 1, [Use hosts.allow and hosts.deny for access checking of rpc.rquotad])
+-	COMPILE_OPTS="$COMPILE_OPTS HOSTS_ACCESS"
++AC_ARG_WITH(tcpwrappers,
++	[  --with-tcpwrappers=[yes/no/try]   Use hosts.allow and hosts.deny for access checking of rpc.rquota [default=yes, if available.]],
++	,
++	with_tcpwrappers="try")
++
++if test "x$with_tcpwrappers" != xno; then
++	AC_MSG_CHECKING(for host_access in -lwrap)
++	AC_CACHE_VAL(ac_cv_lib_wrap_main,
++		saved_LIBS="$LIBS"
++		LIBS="$LIBS -lwrap"
++		[AC_LINK_IFELSE([AC_LANG_PROGRAM([[
++		#include <stdio.h>
++		#include <sys/types.h>
++		#include <sys/socket.h>
++		#include <netinet/in.h>
++		#include <tcpd.h>
++		struct request_info request;
++		int deny_severity, allow_severity;]],[[hosts_access(&request);]])],
++		dnl We always restore LIBS as we add -lwrap in the next try
++		[ac_cv_lib_wrap_main=yes; LIBS="$saved_LIBS"; AC_MSG_RESULT(yes)],
++		[ac_cv_lib_wrap_main=no; LIBS="$saved_LIBS"; AC_MSG_RESULT(no)])
++		])
++
++	if test "x$ac_cv_lib_wrap_main" = xyes; then
++		AC_CHECK_HEADER(tcpd.h,, [
++			if test "x$with_tcpwrappers" != xtry; then
++				AC_MSG_ERROR([could not find tcpd.h - missing TCP wrappers package])
++			fi
++		])
++
++		LIBS="$LIBS -lwrap"
++		AC_DEFINE([HOSTS_ACCESS], 1, [Use hosts.allow and hosts.deny for access checking of rpc.rquotad])
++		COMPILE_OPTS="$COMPILE_OPTS HOSTS_ACCESS"
++	elif test "x$with_tcpwrappers" != xtry; then
++		AC_MSG_ERROR([could not find libwrap - missing TCP wrappers package])
++	fi
+ fi
+
++
+ dnl Checks for typedefs, structures, and compiler characteristics.
+ AC_C_CONST
+ AC_C_INLINE
diff --git a/meta/recipes-extended/quota/quota_4.01.bb b/meta/recipes-extended/quota/quota_4.01.bb
index 9835f06..a5f69b3 100644
--- a/meta/recipes-extended/quota/quota_4.01.bb
+++ b/meta/recipes-extended/quota/quota_4.01.bb
@@ -8,7 +8,8 @@ LIC_FILES_CHKSUM = "file://quota.c;beginline=1;endline=33;md5=331c7d77744bfe0ad2
                     file://svc_socket.c;beginline=1;endline=17;md5=24d5a8792da45910786eeac750be8ceb"
 PR = "r1"
 
-SRC_URI = "${SOURCEFORGE_MIRROR}/project/linuxquota/quota-tools/${PV}/quota-${PV}.tar.gz"
+SRC_URI = "${SOURCEFORGE_MIRROR}/project/linuxquota/quota-tools/${PV}/quota-${PV}.tar.gz \
+           file://config-tcpwrappers.patch"
 
 SRC_URI[md5sum] = "5c2c31e321d2e1322ce12d69ae5c66d6"
 SRC_URI[sha256sum] = "a36300bbc126b79b745bf937245092808b4585aa3309ef3335d4ab9d873cd206"
-- 
1.8.3.4

[PATCH 3/7] python, python-native: fix PARALLEL_MAKEINST failure

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

When using make -j with the 'install' target, it's possible for altbininstall
(which normally creates BINDIR) and libainstall (which doesn't, though it
installs python-config there) to race, resulting in a failure due to
attempting to install python-config into a nonexistent BINDIR. Ensure it also
exists in the libainstall target.

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 meta/recipes-devtools/python/python-native_2.7.3.bb   |  1 +
 .../python/parallel-makeinst-create-bindir.patch      | 19 +++++++++++++++++++
 meta/recipes-devtools/python/python_2.7.3.bb          |  1 +
 3 files changed, 21 insertions(+)
 create mode 100644 meta/recipes-devtools/python/python/parallel-makeinst-create-bindir.patch

diff --git a/meta/recipes-devtools/python/python-native_2.7.3.bb b/meta/recipes-devtools/python/python-native_2.7.3.bb
index ef891b2..6b87a16 100644
--- a/meta/recipes-devtools/python/python-native_2.7.3.bb
+++ b/meta/recipes-devtools/python/python-native_2.7.3.bb
@@ -16,6 +16,7 @@ SRC_URI += "\
            file://multilib.patch \
            file://add-md5module-support.patch \
            file://builddir.patch \
+           file://parallel-makeinst-create-bindir.patch \
            "
 S = "${WORKDIR}/Python-${PV}"
 
diff --git a/meta/recipes-devtools/python/python/parallel-makeinst-create-bindir.patch b/meta/recipes-devtools/python/python/parallel-makeinst-create-bindir.patch
new file mode 100644
index 0000000..951cb46
--- /dev/null
+++ b/meta/recipes-devtools/python/python/parallel-makeinst-create-bindir.patch
@@ -0,0 +1,19 @@
+When using make -j with the 'install' target, it's possible for altbininstall
+(which normally creates BINDIR) and libainstall (which doesn't, though it
+installs python-config there) to race, resulting in a failure due to
+attempting to install python-config into a nonexistent BINDIR. Ensure it also
+exists in the libainstall target.
+
+Upstream-Status: Pending
+
+--- Python-2.7.3.orig/Makefile.pre.in
++++ Python-2.7.3/Makefile.pre.in
+@@ -1008,7 +1008,7 @@ LIBPL=		$(LIBP)/config
+ LIBPC=		$(LIBDIR)/pkgconfig
+
+ libainstall:	all python-config
+-	@for i in $(LIBDIR) $(LIBP) $(LIBPL) $(LIBPC); \
++	@for i in $(LIBDIR) $(LIBP) $(LIBPL) $(LIBPC) $(BINDIR); \
+	do \
+		if test ! -d $(DESTDIR)$$i; then \
+			echo "Creating directory $$i"; \
diff --git a/meta/recipes-devtools/python/python_2.7.3.bb b/meta/recipes-devtools/python/python_2.7.3.bb
index 887e02e..ff41122 100644
--- a/meta/recipes-devtools/python/python_2.7.3.bb
+++ b/meta/recipes-devtools/python/python_2.7.3.bb
@@ -32,6 +32,7 @@ SRC_URI += "\
   file://run-ptest \
   file://CVE-2013-4073_py27.patch \
   file://pypirc-secure.patch \
+  file://parallel-makeinst-create-bindir.patch \
 "
 
 S = "${WORKDIR}/Python-${PV}"
-- 
1.8.3.4

[PATCH 4/7] cairo: add/use packageconfig for valgrind support

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

It was currently autodetecting.

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 meta/recipes-graphics/cairo/cairo.inc | 1 +
 1 file changed, 1 insertion(+)

diff --git a/meta/recipes-graphics/cairo/cairo.inc b/meta/recipes-graphics/cairo/cairo.inc
index 2149997..5212f8b 100644
--- a/meta/recipes-graphics/cairo/cairo.inc
+++ b/meta/recipes-graphics/cairo/cairo.inc
@@ -16,6 +16,7 @@ PACKAGECONFIG ??= "${@base_contains('DISTRO_FEATURES', 'x11', 'x11', '', d)} \
            ${@base_contains('DISTRO_FEATURES', 'directfb', 'directfb', '', d)}"
 PACKAGECONFIG[x11] = "--with-x=yes,--without-x,${X11DEPENDS}"
 PACKAGECONFIG[directfb] = "--enable-directfb=yes,,directfb"
+PACKAGECONFIG[valgrind] = "--enable-valgrind=yes,--disable-valgrind,valgrind"
 BBCLASSEXTEND = "native"
 
 #check for TARGET_FPU=soft and inform configure of the result so it can disable some floating points 
-- 
1.8.3.4

[PATCH 5/7] perf: remove /usr/local/include from default makefile includes

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 meta/recipes-kernel/perf/perf.bb | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/meta/recipes-kernel/perf/perf.bb b/meta/recipes-kernel/perf/perf.bb
index 903ffa6..4fb8956 100644
--- a/meta/recipes-kernel/perf/perf.bb
+++ b/meta/recipes-kernel/perf/perf.bb
@@ -65,16 +65,17 @@ TUI_DEFINES = "${@perf_feature_enabled('perf-tui', '', 'NO_NEWT=1',d)}"
 # supported kernel.
 LDFLAGS="-ldl -lutil"
 
-EXTRA_OEMAKE = \
-		'-C ${S}/tools/perf \
+EXTRA_OEMAKE = "\
+		-C ${S}/tools/perf \
 		O=${B} \
 		CROSS_COMPILE=${TARGET_PREFIX} \
 		ARCH=${ARCH} \
-		CC="${CC}" \
-		AR="${AR}" \
+		'CC=${CC}' \
+		'AR=${AR}' \
 		perfexecdir=${libexecdir} \
 		NO_GTK2=1 ${TUI_DEFINES} NO_DWARF=1 ${SCRIPTING_DEFINES} \
-		'
+		'INCLUDES=-I. $(CONFIG_INCLUDES)' \
+		"
 
 EXTRA_OEMAKE += "\
 	'prefix=${prefix}' \
-- 
1.8.3.4

[PATCH 6/7] qemu: handle CLOEXEC/NONBLOCK if unavailable on host

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 ...Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch | 92 ++++++++++++++++++++++
 meta/recipes-devtools/qemu/qemu.inc                |  1 +
 2 files changed, 93 insertions(+)
 create mode 100644 meta/recipes-devtools/qemu/files/linux-user-Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch

diff --git a/meta/recipes-devtools/qemu/files/linux-user-Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch b/meta/recipes-devtools/qemu/files/linux-user-Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch
new file mode 100644
index 0000000..eb63896
--- /dev/null
+++ b/meta/recipes-devtools/qemu/files/linux-user-Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch
@@ -0,0 +1,92 @@
+Upstream-Status: Backport
+
+From 53d09b761f032f50c4424e8649396a9041070bae Mon Sep 17 00:00:00 2001
+From: "Edgar E. Iglesias" <edgar.iglesias@gmail.com>
+Date: Mon, 23 Sep 2013 14:11:53 +0200
+Subject: [PATCH] linux-user: Handle SOCK_CLOEXEC/NONBLOCK if unavailable on
+ host
+
+If the host lacks SOCK_CLOEXEC, bail out with -EINVAL.
+If the host lacks SOCK_ONONBLOCK, try to emulate it with fcntl()
+and O_NONBLOCK.
+
+Signed-off-by: Edgar E. Iglesias <edgar.iglesias@gmail.com>
+Signed-off-by: Riku Voipio <riku.voipio@linaro.org>
+---
+ linux-user/syscall.c | 40 +++++++++++++++++++++++++++++++++++++---
+ 1 file changed, 37 insertions(+), 3 deletions(-)
+
+diff --git a/linux-user/syscall.c b/linux-user/syscall.c
+index b3822b3..4a14a43 100644
+--- a/linux-user/syscall.c
++++ b/linux-user/syscall.c
+@@ -1773,7 +1773,7 @@ static void unlock_iovec(struct iovec *vec, abi_ulong target_addr,
+     free(vec);
+ }
+
+-static inline void target_to_host_sock_type(int *type)
++static inline int target_to_host_sock_type(int *type)
+ {
+     int host_type = 0;
+     int target_type = *type;
+@@ -1790,22 +1790,56 @@ static inline void target_to_host_sock_type(int *type)
+         break;
+     }
+     if (target_type & TARGET_SOCK_CLOEXEC) {
++#if defined(SOCK_CLOEXEC)
+         host_type |= SOCK_CLOEXEC;
++#else
++        return -TARGET_EINVAL;
++#endif
+     }
+     if (target_type & TARGET_SOCK_NONBLOCK) {
++#if defined(SOCK_NONBLOCK)
+         host_type |= SOCK_NONBLOCK;
++#elif !defined(O_NONBLOCK)
++        return -TARGET_EINVAL;
++#endif
+     }
+     *type = host_type;
++    return 0;
++}
++
++/* Try to emulate socket type flags after socket creation.  */
++static int sock_flags_fixup(int fd, int target_type)
++{
++#if !defined(SOCK_NONBLOCK) && defined(O_NONBLOCK)
++    if (target_type & TARGET_SOCK_NONBLOCK) {
++        int flags = fcntl(fd, F_GETFL);
++        if (fcntl(fd, F_SETFL, O_NONBLOCK | flags) == -1) {
++            close(fd);
++            return -TARGET_EINVAL;
++        }
++    }
++#endif
++    return fd;
+ }
+
+ /* do_socket() Must return target values and target errnos. */
+ static abi_long do_socket(int domain, int type, int protocol)
+ {
+-    target_to_host_sock_type(&type);
++    int target_type = type;
++    int ret;
++
++    ret = target_to_host_sock_type(&type);
++    if (ret) {
++        return ret;
++    }
+
+     if (domain == PF_NETLINK)
+         return -EAFNOSUPPORT; /* do not NETLINK socket connections possible */
+-    return get_errno(socket(domain, type, protocol));
++    ret = get_errno(socket(domain, type, protocol));
++    if (ret >= 0) {
++        ret = sock_flags_fixup(ret, target_type);
++    }
++    return ret;
+ }
+
+ /* do_bind() Must return target values and target errnos. */
+--
+1.8.2.1
diff --git a/meta/recipes-devtools/qemu/qemu.inc b/meta/recipes-devtools/qemu/qemu.inc
index 747d6b2..0852722 100644
--- a/meta/recipes-devtools/qemu/qemu.inc
+++ b/meta/recipes-devtools/qemu/qemu.inc
@@ -18,6 +18,7 @@ SRC_URI = "\
     file://no-strip.patch \
     file://larger_default_ram_size.patch \
     file://disable-grabs.patch \
+    file://linux-user-Handle-SOCK_CLOEXEC-NONBLOCK-if-unavailab.patch \
     "
 
 SRC_URI_append_class-native = "\
-- 
1.8.3.4

[PATCH 7/7] pulseaudio: fix RDEPENDS traversal for consolekit

[Christopher Larson]

From: Christopher Larson <chris_larson@mentor.com>

Include the console-kit module in PACKSGES explicitly so bitbake can map to
the RDEPENDS we define for it in this recipe, and thereby ensure that when
adding the console-kit module to an image, we also get the necessary
consolekit package produced.

Signed-off-by: Christopher Larson <kergoth@gmail.com>
---
 meta/recipes-multimedia/pulseaudio/pulseaudio.inc | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/meta/recipes-multimedia/pulseaudio/pulseaudio.inc b/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
index 7e87ef8..2bee9eb 100644
--- a/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
+++ b/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
@@ -69,8 +69,12 @@ USERADD_PARAM_pulseaudio-server = "--system --home /var/run/pulse \
                               --no-create-home --shell /bin/false \
                               --groups audio,pulse --gid pulse pulse"
 
+# The console-kit module is included here explicitly so bitbake can map to the
+# RDEPENDS we define for it in this recipe, and thereby ensure that when
+# adding the console-kit module to an image, we also get the necessary
+# consolekit package produced.
 PACKAGES =+ "libpulsecore libpulsecommon libpulse libpulse-simple libpulse-mainloop-glib \
-             pulseaudio-server pulseaudio-misc"
+             pulseaudio-server pulseaudio-misc pulseaudio-module-console-kit"
 
 #upgrade path:
 RREPLACES_pulseaudio-server = "libpulse-bin libpulse-conf"
-- 
1.8.3.4

Re: Random fixes from the Mentor Graphics push queue

[Chris Larson]

[-- Attachment #1: Type: text/plain, Size: 890 bytes --]

On Wed, Nov 27, 2013 at 11:07 AM, Christopher Larson <kergoth@gmail.com>wrote:

> The following changes since commit
> f991d2d60b74f5ebd990f77aecd3324b1a4533e9:
>
>   libpng: set reasonable SUMMARY (2013-11-27 11:51:08 +0000)
>
> are available in the git repository at:
>
>   https://github.com/kergoth/oe-core.git random-fixes
>   https://github.com/kergoth/oe-core/tree/random-fixes
>
> Christopher Larson (7):
>   update-rc.d: process symlinks recursively
>   perf: remove /usr/local/include from default makefile includes
>   pulseaudio: fix RDEPENDS traversal for consolekit
>

Were there issues with these, or was it just that further testing was
necessary? I noticed this series was partially applied.
-- 
Christopher Larson
clarson at kergoth dot com
Founder - BitBake, OpenEmbedded, OpenZaurus
Maintainer - Tslib
Senior Software Engineer, Mentor Graphics

[-- Attachment #2: Type: text/html, Size: 1416 bytes --]

Re: Random fixes from the Mentor Graphics push queue

[Richard Purdie]

On Tue, 2013-12-03 at 11:40 -0700, Chris Larson wrote:
> On Wed, Nov 27, 2013 at 11:07 AM, Christopher Larson
> <kergoth@gmail.com> wrote:
>         The following changes since commit
>         f991d2d60b74f5ebd990f77aecd3324b1a4533e9:
>         
>           libpng: set reasonable SUMMARY (2013-11-27 11:51:08 +0000)
>         
>         are available in the git repository at:
>         
>           https://github.com/kergoth/oe-core.git random-fixes
>           https://github.com/kergoth/oe-core/tree/random-fixes
>         
>         Christopher Larson (7):
>           update-rc.d: process symlinks recursively
>           perf: remove /usr/local/include from default makefile
>         includes
>           pulseaudio: fix RDEPENDS traversal for consolekit
> 
> Were there issues with these, or was it just that further testing was
> necessary? I noticed this series was partially applied.

On the most part just being cautious for M1. 

The perf one does worry me as we keep going around in circles on it
where we change one thing, it breaks something else, we fix that,
something else breaks, we fix that and we end up back where we started.
The commit history and/or the mailing list should have details about
what happened there. CC'ing Bruce since I know he has opinions on the
perf fix. Bruce, did we get a fix for that upstream in the end?

Cheers,

Richard

Re: Random fixes from the Mentor Graphics push queue

[Bruce Ashfield]

On 12/3/2013, 5:40 PM, Richard Purdie wrote:
> On Tue, 2013-12-03 at 11:40 -0700, Chris Larson wrote:
>> On Wed, Nov 27, 2013 at 11:07 AM, Christopher Larson
>> <kergoth@gmail.com> wrote:
>>          The following changes since commit
>>          f991d2d60b74f5ebd990f77aecd3324b1a4533e9:
>>
>>            libpng: set reasonable SUMMARY (2013-11-27 11:51:08 +0000)
>>
>>          are available in the git repository at:
>>
>>            https://github.com/kergoth/oe-core.git random-fixes
>>            https://github.com/kergoth/oe-core/tree/random-fixes
>>
>>          Christopher Larson (7):
>>            update-rc.d: process symlinks recursively
>>            perf: remove /usr/local/include from default makefile
>>          includes
>>            pulseaudio: fix RDEPENDS traversal for consolekit
>>
>> Were there issues with these, or was it just that further testing was
>> necessary? I noticed this series was partially applied.
>
> On the most part just being cautious for M1.
>
> The perf one does worry me as we keep going around in circles on it
> where we change one thing, it breaks something else, we fix that,
> something else breaks, we fix that and we end up back where we started.
> The commit history and/or the mailing list should have details about
> what happened there. CC'ing Bruce since I know he has opinions on the
> perf fix. Bruce, did we get a fix for that upstream in the end?

My favourite topic. I'm working on the upstream fix, which was that we
shouldn't be using CFLAGS and LDFLAGS to pass build values into perf,
but instead use EXTRA_CLFAGS (which exists) and EXTRA_LDFLAGS (Which
doesn't exist yet, and that's the patch I have queued).

Looking at Chris' posted patch, I can't actually tell what it was 
fixing, since INCLUDES isn't directly used by the perf Makefiles and on
a really quick check, I can't see what CONFIG_INCLUDES is set to either.

I'm sure I'm just being dense (long day) and I'm missing something, so
I can't map the short log's intent to what the patch is actually doing.

The patch won't undo anything that we are working to fix/change
upstream, but since something is whizzing over my head, I can't be sure
about side effects.

Cheers,

Bruce

>
> Cheers,
>
> Richard
>
>
>
>

Re: Random fixes from the Mentor Graphics push queue

[Chris Larson]

[-- Attachment #1: Type: text/plain, Size: 2772 bytes --]

On Tue, Dec 3, 2013 at 7:38 PM, Bruce Ashfield <bruce.ashfield@windriver.com
> wrote:

> On 12/3/2013, 5:40 PM, Richard Purdie wrote:
>
>> On Tue, 2013-12-03 at 11:40 -0700, Chris Larson wrote:
>>
>>> On Wed, Nov 27, 2013 at 11:07 AM, Christopher Larson
>>> <kergoth@gmail.com> wrote:
>>>          The following changes since commit
>>>          f991d2d60b74f5ebd990f77aecd3324b1a4533e9:
>>>
>>>            libpng: set reasonable SUMMARY (2013-11-27 11:51:08 +0000)
>>>
>>>          are available in the git repository at:
>>>
>>>            https://github.com/kergoth/oe-core.git random-fixes
>>>            https://github.com/kergoth/oe-core/tree/random-fixes
>>>
>>>          Christopher Larson (7):
>>>            update-rc.d: process symlinks recursively
>>>            perf: remove /usr/local/include from default makefile
>>>          includes
>>>            pulseaudio: fix RDEPENDS traversal for consolekit
>>>
>>> Were there issues with these, or was it just that further testing was
>>> necessary? I noticed this series was partially applied.
>>>
>>
>> On the most part just being cautious for M1.
>>
>> The perf one does worry me as we keep going around in circles on it
>> where we change one thing, it breaks something else, we fix that,
>> something else breaks, we fix that and we end up back where we started.
>> The commit history and/or the mailing list should have details about
>> what happened there. CC'ing Bruce since I know he has opinions on the
>> perf fix. Bruce, did we get a fix for that upstream in the end?
>>
>
> My favourite topic. I'm working on the upstream fix, which was that we
> shouldn't be using CFLAGS and LDFLAGS to pass build values into perf,
> but instead use EXTRA_CLFAGS (which exists) and EXTRA_LDFLAGS (Which
> doesn't exist yet, and that's the patch I have queued).
>
> Looking at Chris' posted patch, I can't actually tell what it was fixing,
> since INCLUDES isn't directly used by the perf Makefiles and on
> a really quick check, I can't see what CONFIG_INCLUDES is set to either.
>
> I'm sure I'm just being dense (long day) and I'm missing something, so
> I can't map the short log's intent to what the patch is actually doing.
>
> The patch won't undo anything that we are working to fix/change
> upstream, but since something is whizzing over my head, I can't be sure
> about side effects.


*shrug*, it’s possible it was to fix warnings only with older kernels, as
we use our own BSPs in some cases. I’m not opposed to you dropping that
patch, it was only a warning fix for us anyway :)
-- 
Christopher Larson
clarson at kergoth dot com
Founder - BitBake, OpenEmbedded, OpenZaurus
Maintainer - Tslib
Senior Software Engineer, Mentor Graphics

[-- Attachment #2: Type: text/html, Size: 3764 bytes --]

Re: [PATCH 7/7] pulseaudio: fix RDEPENDS traversal for consolekit

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 2160 bytes --]

On Wed, Nov 27, 2013 at 11:07:07AM -0700, Christopher Larson wrote:
> From: Christopher Larson <chris_larson@mentor.com>
> 
> Include the console-kit module in PACKSGES explicitly so bitbake can map to
> the RDEPENDS we define for it in this recipe, and thereby ensure that when
> adding the console-kit module to an image, we also get the necessary
> consolekit package produced.
> 
> Signed-off-by: Christopher Larson <kergoth@gmail.com>
> ---
>  meta/recipes-multimedia/pulseaudio/pulseaudio.inc | 6 +++++-
>  1 file changed, 5 insertions(+), 1 deletion(-)
> 
> diff --git a/meta/recipes-multimedia/pulseaudio/pulseaudio.inc b/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> index 7e87ef8..2bee9eb 100644
> --- a/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> +++ b/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> @@ -69,8 +69,12 @@ USERADD_PARAM_pulseaudio-server = "--system --home /var/run/pulse \
>                                --no-create-home --shell /bin/false \
>                                --groups audio,pulse --gid pulse pulse"
>  
> +# The console-kit module is included here explicitly so bitbake can map to the
> +# RDEPENDS we define for it in this recipe, and thereby ensure that when
> +# adding the console-kit module to an image, we also get the necessary
> +# consolekit package produced.
>  PACKAGES =+ "libpulsecore libpulsecommon libpulse libpulse-simple libpulse-mainloop-glib \
> -             pulseaudio-server pulseaudio-misc"
> +             pulseaudio-server pulseaudio-misc pulseaudio-module-console-kit"

Can we add PACKAGECONFIG for this and add it to PACKAGES only
conditionally?

It causes hard dependency on consolekit -> libx11 so pulseaudio now
cannot be built for distros without x11 DISTRO_FEATURE.

>  #upgrade path:
>  RREPLACES_pulseaudio-server = "libpulse-bin libpulse-conf"
> -- 
> 1.8.3.4
> 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 205 bytes --]

Re: [PATCH 7/7] pulseaudio: fix RDEPENDS traversal for consolekit

[Chris Larson]

[-- Attachment #1: Type: text/plain, Size: 2169 bytes --]

On Wed, Dec 11, 2013 at 2:22 PM, Martin Jansa <martin.jansa@gmail.com>wrote:

> On Wed, Nov 27, 2013 at 11:07:07AM -0700, Christopher Larson wrote:
> > From: Christopher Larson <chris_larson@mentor.com>
> >
> > Include the console-kit module in PACKSGES explicitly so bitbake can map
> to
> > the RDEPENDS we define for it in this recipe, and thereby ensure that
> when
> > adding the console-kit module to an image, we also get the necessary
> > consolekit package produced.
> >
> > Signed-off-by: Christopher Larson <kergoth@gmail.com>
> > ---
> >  meta/recipes-multimedia/pulseaudio/pulseaudio.inc | 6 +++++-
> >  1 file changed, 5 insertions(+), 1 deletion(-)
> >
> > diff --git a/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> b/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> > index 7e87ef8..2bee9eb 100644
> > --- a/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> > +++ b/meta/recipes-multimedia/pulseaudio/pulseaudio.inc
> > @@ -69,8 +69,12 @@ USERADD_PARAM_pulseaudio-server = "--system --home
> /var/run/pulse \
> >                                --no-create-home --shell /bin/false \
> >                                --groups audio,pulse --gid pulse pulse"
> >
> > +# The console-kit module is included here explicitly so bitbake can map
> to the
> > +# RDEPENDS we define for it in this recipe, and thereby ensure that when
> > +# adding the console-kit module to an image, we also get the necessary
> > +# consolekit package produced.
> >  PACKAGES =+ "libpulsecore libpulsecommon libpulse libpulse-simple
> libpulse-mainloop-glib \
> > -             pulseaudio-server pulseaudio-misc"
> > +             pulseaudio-server pulseaudio-misc
> pulseaudio-module-console-kit"
>
> Can we add PACKAGECONFIG for this and add it to PACKAGES only
> conditionally?
>
> It causes hard dependency on consolekit -> libx11 so pulseaudio now
> cannot be built for distros without x11 DISTRO_FEATURE.


Sorry about this, I’ll send a patch shortly.
-- 
Christopher Larson
clarson at kergoth dot com
Founder - BitBake, OpenEmbedded, OpenZaurus
Maintainer - Tslib
Senior Software Engineer, Mentor Graphics

[-- Attachment #2: Type: text/html, Size: 2830 bytes --]