* [Buildroot] [PATCH 1/2] toolchain/gdb: strip tarball component when extracting
From: Peter Korsgaard @ 2011-10-05 14:47 UTC (permalink / raw)
To: buildroot
In-Reply-To: <1315951050-19272-2-git-send-email-yann.morin.1998@anciens.enib.fr>
>>>>> "Yann" == Yann E MORIN <yann.morin.1998@anciens.enib.fr> writes:
Yann> For some tarballs of gdb (see next patch), the extracted directory is
Yann> *not* named after the version string (eg. gdb-6.6a extract into gdb-6.6/)
Yann> Create the appropriate directory first, then use --strip-{components,path}
Yann> when extracting gdb (the same way it is done for the generic package
Yann> infrastructure).
Yann> At the same time, get rid of the snapshot special case, because:
Yann> 1- it's no longer available in the menu
Yann> 2- it would be handled by the above change
Committed, thanks.
--
Bye, Peter Korsgaard
^ permalink raw reply
* [Buildroot] [git commit] toolchain/gdb: strip tarball component when extracting
From: Peter Korsgaard @ 2011-10-05 14:46 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=c5da6f9553ac82fefbefb655ac0207bc58615ae4
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
For some tarballs of gdb (see next patch), the extracted directory is
*not* named after the version string (eg. gdb-6.6a extract into gdb-6.6/)
Create the appropriate directory first, then use --strip-{components,path}
when extracting gdb (the same way it is done for the generic package
infrastructure).
At the same time, get rid of the snapshot special case, because:
1- it's no longer available in the menu
2- it would be handled by the above change
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
toolchain/gdb/gdb.mk | 9 ++-------
1 files changed, 2 insertions(+), 7 deletions(-)
diff --git a/toolchain/gdb/gdb.mk b/toolchain/gdb/gdb.mk
index 4622616..85b892f 100644
--- a/toolchain/gdb/gdb.mk
+++ b/toolchain/gdb/gdb.mk
@@ -28,13 +28,8 @@ $(DL_DIR)/$(GDB_SOURCE):
gdb-unpacked: $(GDB_DIR)/.unpacked
$(GDB_DIR)/.unpacked: $(DL_DIR)/$(GDB_SOURCE)
- mkdir -p $(TOOLCHAIN_DIR)
- $(GDB_CAT) $(DL_DIR)/$(GDB_SOURCE) | tar -C $(TOOLCHAIN_DIR) $(TAR_OPTIONS) -
-ifeq ($(GDB_VERSION),snapshot)
- GDB_REAL_DIR=$(shell \
- tar jtf $(DL_DIR)/$(GDB_SOURCE) | head -1 | cut -d"/" -f1)
- ln -sf $(TOOLCHAIN_DIR)/$(shell tar jtf $(DL_DIR)/$(GDB_SOURCE) | head -1 | cut -d"/" -f1) $(GDB_DIR)
-endif
+ mkdir -p $(GDB_DIR)
+ $(GDB_CAT) $(DL_DIR)/$(GDB_SOURCE) | tar -C $(GDB_DIR) $(TAR_STRIP_COMPONENTS)=1 $(TAR_OPTIONS) -
ifneq ($(wildcard $(GDB_PATCH_DIR)),)
support/scripts/apply-patches.sh $(GDB_DIR) $(GDB_PATCH_DIR) \*.patch $(GDB_PATCH_EXTRA)
endif
^ permalink raw reply related
* [Buildroot] [PATCH] gdb: bump to 7.3.1
From: Peter Korsgaard @ 2011-10-05 14:28 UTC (permalink / raw)
To: buildroot
In-Reply-To: <4E8C5C31.5040703@visionsystems.de>
>>>>> "Yegor" == Yegor Yefremov <yegor_sub1@visionsystems.de> writes:
Hi,
Yegor> Have you looked at http://www.gnu.org/s/gdb/ 7.3 release notes?
Yegor> " ia64 HP-UX (native) and Blackfin (target) support
Yegor> GDBserver support for PowerPC LynxOS, i686 LynxOS, and Blackfin Linux "
Yegor> Can some one with blackfin test this? If it is working, we could
Yegor> remove the dependence on !blfn.
Yes, I noticed as well. Bfin users, please speak up.
--
Bye, Peter Korsgaard
^ permalink raw reply
* [Buildroot] [git commit] gdb: misc cleanup
From: Peter Korsgaard @ 2011-10-05 14:14 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=37f0d1bb7b5a379161e647b730f5a57b814122bb
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
Only use 6.6 on bfin, remove 6.7.1, mark 6.8 + 7.0 as deprecated and
default to 7.3.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
.../gdb/6.7.1/600-fix-compile-flag-mismatch.patch | 31 --------------------
toolchain/gdb/Config.in | 13 +++-----
2 files changed, 5 insertions(+), 39 deletions(-)
diff --git a/toolchain/gdb/6.7.1/600-fix-compile-flag-mismatch.patch b/toolchain/gdb/6.7.1/600-fix-compile-flag-mismatch.patch
deleted file mode 100644
index 13b72bb..0000000
--- a/toolchain/gdb/6.7.1/600-fix-compile-flag-mismatch.patch
+++ /dev/null
@@ -1,31 +0,0 @@
---- a/gdb/gdbserver/configure
-+++ b/gdb/gdbserver/configure
-@@ -1239,7 +1239,7 @@
- ac_cache_corrupted=: ;;
- ,);;
- *)
-- if test "x$ac_old_val" != "x$ac_new_val"; then
-+ if test "`echo x$ac_old_val`" != "`echo x$ac_new_val`"; then
- { echo "$as_me:$LINENO: error: \`$ac_var' has changed since the previous run:" >&5
- echo "$as_me: error: \`$ac_var' has changed since the previous run:" >&2;}
- { echo "$as_me:$LINENO: former value: $ac_old_val" >&5
---- a/gdb/configure
-+++ b/gdb/configure
-@@ -272,7 +272,7 @@
- PACKAGE_BUGREPORT=
-
- ac_unique_file="main.c"
--ac_subdirs_all="$ac_subdirs_all doc testsuite"
-+ac_subdirs_all="$ac_subdirs_all doc"
- # Factoring default headers for most tests.
- ac_includes_default="\
- #include <stdio.h>
-@@ -3077,7 +3077,7 @@
-
-
-
--subdirs="$subdirs doc testsuite"
-+subdirs="$subdirs doc"
-
-
- # Provide defaults for some variables set by the per-host and per-target
diff --git a/toolchain/gdb/Config.in b/toolchain/gdb/Config.in
index 6bf7fd4..28279a3 100644
--- a/toolchain/gdb/Config.in
+++ b/toolchain/gdb/Config.in
@@ -26,20 +26,16 @@ config BR2_PACKAGE_GDB_HOST
choice
prompt "GDB debugger Version"
- default BR2_GDB_VERSION_6_8 if !BR2_avr32
+ default BR2_GDB_VERSION_6_6 if !BR2_bfin
default BR2_GDB_VERSION_6_7_1_AVR32_2_1_5 if BR2_avr32
+ default BR2_GDB_VERSION_7_3
depends on BR2_PACKAGE_GDB || BR2_PACKAGE_GDB_SERVER || BR2_PACKAGE_GDB_HOST
help
Select the version of gdb you wish to use.
config BR2_GDB_VERSION_6_6
bool "gdb 6.6"
- depends on !BR2_avr32
- depends on BR2_DEPRECATED || BR2_bfin
-
- config BR2_GDB_VERSION_6_7_1
- depends on !(BR2_avr32 || BR2_bfin)
- bool "gdb 6.7.1"
+ depends on BR2_bfin
config BR2_GDB_VERSION_6_7_1_AVR32_2_1_5
depends on BR2_avr32
@@ -48,10 +44,12 @@ choice
config BR2_GDB_VERSION_6_8
bool "gdb 6.8"
depends on !(BR2_avr32 || BR2_bfin)
+ depends on BR2_DEPRECATED
config BR2_GDB_VERSION_7_0_1
bool "gdb 7.0.1"
depends on !BR2_bfin
+ depends on BR2_DEPRECATED
select BR2_PTHREAD_DEBUG if !BR2_PTHREADS_NONE
config BR2_GDB_VERSION_7_1
@@ -69,7 +67,6 @@ endchoice
config BR2_GDB_VERSION
string
default "6.6" if BR2_GDB_VERSION_6_6
- default "6.7.1" if BR2_GDB_VERSION_6_7_1
default "6.7.1-avr32-2.1.5" if BR2_GDB_VERSION_6_7_1_AVR32_2_1_5
default "6.8" if BR2_GDB_VERSION_6_8
default "7.0.1" if BR2_GDB_VERSION_7_0_1
^ permalink raw reply related
* [Buildroot] [PATCH] gdb: bump to 7.3.1
From: Peter Korsgaard @ 2011-10-05 14:07 UTC (permalink / raw)
To: buildroot
In-Reply-To: <1317719147-19210-1-git-send-email-yegorslists@googlemail.com>
>>>>> "yegorslists" == yegorslists <yegorslists@googlemail.com> writes:
yegorslists> From: Yegor Yefremov <yegorslists@googlemail.com>
yegorslists> this version fixes compilation issue on some old build systems like
yegorslists> openSUSE 10.3 saying some host libraries were too old
Thanks, committed. I dumped the minor version so it's 7.3 similar to how
we're doing it for the kernel headers or busybox.
--
Bye, Peter Korsgaard
^ permalink raw reply
* [Buildroot] [PATCH] gdb: bump to 7.3.1
From: Yegor Yefremov @ 2011-10-05 13:31 UTC (permalink / raw)
To: buildroot
In-Reply-To: <878vozr32n.fsf@macbook.be.48ers.dk>
Am 05.10.2011 15:16, schrieb Peter Korsgaard:
>>>>>> "Yegor" == Yegor Yefremov <yegorslists@googlemail.com> writes:
>
> Yegor> Thanks for comments. I'll rework the patch next week. By the way
> Yegor> shouldn't we get rid of 6.8 due to this issue:
> Yegor> http://comments.gmane.org/gmane.comp.gdb.devel/30648. That was the
> Yegor> reason for trying 7.x version.
>
> Yes, probably. It seems we'll need to keep 6.6 around for bfin though.
I've sent a proper patch using git (there is still plenty to learn about).
Have you looked at http://www.gnu.org/s/gdb/ 7.3 release notes?
" ia64 HP-UX (native) and Blackfin (target) support
GDBserver support for PowerPC LynxOS, i686 LynxOS, and Blackfin Linux "
Can some one with blackfin test this? If it is working, we could remove the dependence on !blfn.
Best regards,
Yegor
^ permalink raw reply
* [Buildroot] [git commit] gdb: bump to 7.3.1
From: Peter Korsgaard @ 2011-10-05 13:20 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=3903c3a73dabcdbdf13d3ff09dedef6a8589ab4a
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
this version fixes compilation issue on some old build systems like
openSUSE 10.3 saying some host libraries were too old
[Peter: drop bugfix number from config name, similar to kernel-headers]
Signed-off-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
toolchain/gdb/Config.in | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/toolchain/gdb/Config.in b/toolchain/gdb/Config.in
index 8295f79..6bf7fd4 100644
--- a/toolchain/gdb/Config.in
+++ b/toolchain/gdb/Config.in
@@ -59,6 +59,11 @@ choice
depends on !BR2_bfin
select BR2_PTHREAD_DEBUG if !BR2_PTHREADS_NONE
+ config BR2_GDB_VERSION_7_3
+ bool "gdb 7.3.x"
+ depends on !BR2_bfin
+ select BR2_PTHREAD_DEBUG if !BR2_PTHREADS_NONE
+
endchoice
config BR2_GDB_VERSION
@@ -69,3 +74,4 @@ config BR2_GDB_VERSION
default "6.8" if BR2_GDB_VERSION_6_8
default "7.0.1" if BR2_GDB_VERSION_7_0_1
default "7.1" if BR2_GDB_VERSION_7_1
+ default "7.3.1" if BR2_GDB_VERSION_7_3
^ permalink raw reply related
* [Buildroot] [PATCH] gdb: bump to 7.3.1
From: Peter Korsgaard @ 2011-10-05 13:16 UTC (permalink / raw)
To: buildroot
In-Reply-To: <CAGm1_ktCdcObbXCYQryUOU5h=Sxpe+aE598NuLXU4HCQ9jCxrQ@mail.gmail.com>
>>>>> "Yegor" == Yegor Yefremov <yegorslists@googlemail.com> writes:
Yegor> Thanks for comments. I'll rework the patch next week. By the way
Yegor> shouldn't we get rid of 6.8 due to this issue:
Yegor> http://comments.gmane.org/gmane.comp.gdb.devel/30648. That was the
Yegor> reason for trying 7.x version.
Yes, probably. It seems we'll need to keep 6.6 around for bfin though.
--
Bye, Peter Korsgaard
^ permalink raw reply
* [Buildroot] [PATCH] package/libcgi/Config.in: help text should be indented by <tab> and 2 spaces
From: Peter Korsgaard @ 2011-10-05 12:03 UTC (permalink / raw)
To: buildroot
In-Reply-To: <ADE657CA350FB648AAC2C43247A983F001F39DC56E5B@AUSP01VMBX24.collaborationhost.net>
>>>>> "H" == H Hartley Sweeten <hartleys@visionengravers.com> writes:
H> Fix the help text for package/libcgi/Config.in.
H> Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Committed, thanks.
--
Bye, Peter Korsgaard
^ permalink raw reply
* [Buildroot] [git commit] package/libcgi/Config.in: help text should be indented by <tab> and 2 spaces
From: Peter Korsgaard @ 2011-10-05 12:02 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=3b3d129c377e657fa1c2d64c45e9ccd90a1b8077
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
Fix the help text for package/libcgi/Config.in.
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
package/libcgi/Config.in | 12 ++++++------
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/package/libcgi/Config.in b/package/libcgi/Config.in
index 491ae2e..430039e 100644
--- a/package/libcgi/Config.in
+++ b/package/libcgi/Config.in
@@ -1,11 +1,11 @@
config BR2_PACKAGE_LIBCGI
bool "libcgi"
help
- LibCGI is a library written from scratch to easily make
- CGI applications in C.
+ LibCGI is a library written from scratch to easily make
+ CGI applications in C.
- There are a lot of functions like string manipulation,
- session and cookie support, GET and POST methods manipulation
- etc..., to help you to quickly write powerful CGI programs.
+ There are a lot of functions like string manipulation,
+ session and cookie support, GET and POST methods manipulation
+ etc..., to help you to quickly write powerful CGI programs.
- http://libcgi.sourceforge.net/
+ http://libcgi.sourceforge.net/
^ permalink raw reply related
* [Buildroot] [PATCH] pkg-stats: update grep tests for package type
From: Peter Korsgaard @ 2011-10-05 11:58 UTC (permalink / raw)
To: buildroot
In-Reply-To: <ADE657CA350FB648AAC2C43247A983F001F39DCFE6AF@AUSP01VMBX24.collaborationhost.net>
>>>>> "H" == H Hartley Sweeten <hartleys@visionengravers.com> writes:
H> Update the grep tests used to determine the package type.
H> The package name and directory are now worked out magically due to:
H> package: add helper functions to get package name and directory magically
H> Because of this the extra arguments were removed by patches:
H> package: remove useless arguments from GENTARGETS
H> package: remove useless arguments from AUTOTARGETS
H> package: remove useless arguments from CMAKETARGETS
Committed, thanks.
--
Bye, Peter Korsgaard
^ permalink raw reply
* [Buildroot] [git commit] pkg-stats: update grep tests for package type
From: Peter Korsgaard @ 2011-10-05 11:57 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=6e3e5a99f7c148ef2fcdb6a033c36e9fe6d7ba94
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
Update the grep tests used to determine the package type.
The package name and directory are now worked out magically due to:
package: add helper functions to get package name and directory magically
Because of this the extra arguments were removed by patches:
package: remove useless arguments from GENTARGETS
package: remove useless arguments from AUTOTARGETS
package: remove useless arguments from CMAKETARGETS
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
support/scripts/pkg-stats | 12 ++++++------
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/support/scripts/pkg-stats b/support/scripts/pkg-stats
index bb1c27d..8602b0f 100755
--- a/support/scripts/pkg-stats
+++ b/support/scripts/pkg-stats
@@ -98,27 +98,27 @@ for i in $(find package/ -name '*.mk') ; do
is_manual_target=0
is_manual_host=0
- if grep -E "\(call AUTOTARGETS,[^,]*,[^,]*,host\)" $i > /dev/null ; then
+ if grep -E "\(call AUTOTARGETS,host\)" $i > /dev/null ; then
is_auto_host=1
fi
- if grep -E "\(call AUTOTARGETS,[^,]*,[^,]*(,target|)\)" $i > /dev/null ; then
+ if grep -E "\(call AUTOTARGETS\)" $i > /dev/null ; then
is_auto_target=1
fi
- if grep -E "\(call GENTARGETS,[^,]*,[^,]*,host\)" $i > /dev/null ; then
+ if grep -E "\(call GENTARGETS,host\)" $i > /dev/null ; then
is_pkg_host=1
fi
- if grep -E "\(call GENTARGETS,[^,]*,[^,]*(,target|)\)" $i > /dev/null ; then
+ if grep -E "\(call GENTARGETS\)" $i > /dev/null ; then
is_pkg_target=1
fi
- if grep -E "\(call CMAKETARGETS,[^,]*,[^,]*,host\)" $i > /dev/null ; then
+ if grep -E "\(call CMAKETARGETS,host\)" $i > /dev/null ; then
is_cmake_host=1
fi
- if grep -E "\(call CMAKETARGETS,[^,]*,[^,]*(,target|)\)" $i > /dev/null ; then
+ if grep -E "\(call CMAKETARGETS\)" $i > /dev/null ; then
is_cmake_target=1
fi
^ permalink raw reply related
* [Buildroot] [git commit] gst-plugins-bad: fix bzip2 typo
From: Peter Korsgaard @ 2011-10-05 9:05 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=6b19842c678e3edb6413906f5d3c72abf5414be9
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
.../multimedia/gst-plugins-bad/gst-plugins-bad.mk | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/package/multimedia/gst-plugins-bad/gst-plugins-bad.mk b/package/multimedia/gst-plugins-bad/gst-plugins-bad.mk
index bebff40..20aad76 100644
--- a/package/multimedia/gst-plugins-bad/gst-plugins-bad.mk
+++ b/package/multimedia/gst-plugins-bad/gst-plugins-bad.mk
@@ -458,7 +458,7 @@ endif
ifeq ($(BR2_PACKAGE_GST_PLUGINS_BAD_PLUGIN_BZ2),y)
GST_PLUGINS_BAD_CONF_OPT += --enable-bz2
-GST_PLUGINS_BAD_DEPENDENCIES += bz2ip
+GST_PLUGINS_BAD_DEPENDENCIES += bzip2
else
GST_PLUGINS_BAD_CONF_OPT += --disable-bz2
endif
^ permalink raw reply related
* [Buildroot] [git commit] mpg123: tweak audio module selection
From: Peter Korsgaard @ 2011-10-05 7:39 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=fca4aa233659fa15c7518a3cc83bc2331beb87bf
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
Fix build if host has libarts-dev, and add optional sdl+portaudio support.
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
package/multimedia/mpg123/mpg123.mk | 22 +++++++++++++++++++---
1 files changed, 19 insertions(+), 3 deletions(-)
diff --git a/package/multimedia/mpg123/mpg123.mk b/package/multimedia/mpg123/mpg123.mk
index 53b7b39..c01c8ad 100644
--- a/package/multimedia/mpg123/mpg123.mk
+++ b/package/multimedia/mpg123/mpg123.mk
@@ -9,6 +9,7 @@ MPG123_SOURCE = mpg123-$(MPG123_VERSION).tar.bz2
MPG123_SITE = http://$(BR2_SOURCEFORGE_MIRROR).dl.sourceforge.net/sourceforge/mpg123
MPG123_CONF_OPT = --with-optimization=0 --disable-lfs-alias
MPG123_INSTALL_STAGING = YES
+
MPG123_CPU = $(if $(BR2_SOFT_FLOAT),generic_nofpu,generic_fpu)
ifeq ($(BR2_arm),y)
@@ -34,11 +35,26 @@ endif
MPG123_CONF_OPT += --with-cpu=$(MPG123_CPU)
-# Check if ALSA is built, then we should configure after alsa-lib so
-# ./configure can find alsa-lib.
+MPG123_AUDIO = dummy oss
+
+ifeq ($(BR2_PACKAGE_PORTAUDIO),y)
+MPG123_AUDIO += portaudio
+MPG123_CONF_OPT += --with-default-audio=portaudio
+MPG123_DEPENDENCIES += portaudio
+endif
+
+ifeq ($(BR2_PACKAGE_SDL),y)
+MPG123_AUDIO += sdl
+MPG123_CONF_OPT += --with-default-audio=sdl
+MPG123_DEPENDENCIES += sdl
+endif
+
ifeq ($(BR2_PACKAGE_ALSA_LIB),y)
-MPG123_CONF_OPT += --with-audio=alsa
+MPG123_AUDIO += alsa
+MPG123_CONF_OPT += --with-default-audio=alsa
MPG123_DEPENDENCIES += alsa-lib
endif
+MPG123_CONF_OPT += --with-audio=$(shell echo $(MPG123_AUDIO) | tr ' ' ,)
+
$(eval $(call AUTOTARGETS))
^ permalink raw reply related
* [Buildroot] Question: := vs = in package *.mk files
From: Thomas Petazzoni @ 2011-10-05 7:04 UTC (permalink / raw)
To: buildroot
In-Reply-To: <ADE657CA350FB648AAC2C43247A983F001F39DCFE6DB@AUSP01VMBX24.collaborationhost.net>
Le Tue, 4 Oct 2011 19:55:48 -0500,
H Hartley Sweeten <hartleys@visionengravers.com> a ?crit :
> The Buildroot documentation shows using '=' when setting variables
> in the package *.mk file. Some of the existing packages use ':='
> instead. What's the difference? Should the existing packages be
> updated?
The difference is explained in the make documentation:
"""
The first flavor of variable is a "recursively expanded" variable.
Variables of this sort are defined by lines using `=' (*note Setting
Variables: Setting.) or by the `define' directive (*note Defining
Variables Verbatim: Defining.). The value you specify is installed
verbatim; if it contains references to other variables, these
references are expanded whenever this variable is substituted (in the
course of expanding some other string).
[...]
"Simply expanded variables" are defined by lines using `:=' (*note
Setting Variables: Setting.). The value of a simply expanded variable
is scanned once and for all, expanding any references to other
variables and functions, when the variable is defined. The actual
value of the simply expanded variable is the result of expanding the
text that you write. It does not contain any references to other
variables; it contains their values _as of the time this variable was
defined_.
"""
See http://www.gnu.org/software/make/manual/make.html#Flavors
In package definition, Peter has said that he preferred the usage of
'=' except when it would create a noticeable performance problem. So
for all package definitions such as _SITE, _VERSION, _SOURCE,
_AUTORECONF, _INSTALL_STAGING, etc. the '=' sign is the one to choose.
Of course, a lot of packages do not comply with this rule, but this
will be fixed progressively over time as we update packages.
Regards,
Thomas
--
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
^ permalink raw reply
* [Buildroot] How to create package that build with scons
From: Thomas Petazzoni @ 2011-10-05 6:45 UTC (permalink / raw)
To: buildroot
In-Reply-To: <201110041820.16324.arnout@mind.be>
Le Tue, 4 Oct 2011 18:20:16 +0200,
Arnout Vandecappelle <arnout@mind.be> a ?crit :
> 'make host-scons' works for me...
>
> There is currently no package that depends on scons, but I guess
> that it was used by some private package of Thomas (or one of his
> customers).
Not for a private package. I submitted as part of a patchset that
included packages for the disko multimedia framework and its demo
programs. However, Peter had some comments of my disko/disko-demos
packages, and he did merge only the scons package. I haven't (yet)
found the time to adjust the disko/disko-demos packages and to submit
them. This explains why there is a host-scons package that isn't used
by any package in the tree.
But at least, with this host-scons package, I was able to build fine
other packages relying on the Scons build system.
Regards,
Thomas
--
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
^ permalink raw reply
* [Buildroot] Bug in pkg-stats?
From: Thomas Petazzoni @ 2011-10-05 6:43 UTC (permalink / raw)
To: buildroot
In-Reply-To: <ADE657CA350FB648AAC2C43247A983F001F39DCFE6AE@AUSP01VMBX24.collaborationhost.net>
Le Tue, 4 Oct 2011 18:05:04 -0500,
H Hartley Sweeten <hartleys@visionengravers.com> a ?crit :
> I think the problem with pkg-stats is due to your patches:
>
> package: remove useless arguments from CMAKETARGETS
> package: remove useless arguments from AUTOTARGETS
> package: remove useless arguments from GENTARGETS
> package: add helper functions to get package name and directory
> magically
Yes.
> The grep tests are still looking for the package name and directory to
> determine the package type.
>
> I'll post a patch shortly.
I already have a patch that fixes pkg-stats, which I intend to post
shortly. But if you are quicker than me, no problem :)
> One other thing with pkg-stats.
>
> Can you think of any way to pull the _VERSION of the packages out of
> the *.mk files and show that information in the output html file?
Yes, shouldn't be too difficult. Why would this be useful ?
Thomas
--
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
^ permalink raw reply
* [Buildroot] [PATCH 2/9] toolchain-external: allow specifying extra external libraries
From: Thomas De Schampheleire @ 2011-10-05 6:20 UTC (permalink / raw)
To: buildroot
In-Reply-To: <201110042336.44134.yann.morin.1998@anciens.enib.fr>
Hello Yann,
On Tue, Oct 4, 2011 at 11:36 PM, Yann E. MORIN
<yann.morin.1998@anciens.enib.fr> wrote:
> Thomas?, All,
>
> On Monday 03 October 2011 09:21:59 Thomas De Schampheleire wrote:
>> On Sun, Oct 2, 2011 at 11:45 PM, Yann E. MORIN
>> <yann.morin.1998@anciens.enib.fr> wrote:
>> > On Sunday 02 October 2011 21:20:10 Thomas Petazzoni wrote:
>> >> From: Thomas De Schampheleire <patrickdepinguin+buildroot@gmail.com>
>> >>
>> >> Custom toolchains may provide extra libraries that need to be copied to the
>> >> target. This patch adds a configuration option for this.
>> >
>> > What kind of libraries are we talking about here? Things such as libz.so,
>> > or like libnss_*.so ?
> [--SNIP--]
>> > So, I'm not really sure we should allow that without further explanations.
>> > (and there was no more in Thomas' original post on the suject...)
>
>> uClibc supports the 'backtrace' functionality. I'm not sure whether it
>> already is in 0.9.32,
>
> Yes, it is.
>
>> but it certainly is in the repository. When you
>> enable the functionality in the uClibc configuration, a library
>> libubacktrace.so is generated. Application code that uses the
>> backtrace() or backtrace_symbols() functions, need to link with this
>> library and thus it needs to be copied from the toolchain.
>
> If an application is linked against libubactrace, then it should have a
> NEEDED entry for it, no? Or is it dlopen-ed on-demand?
I can confirm that there is an appropriate NEEDED entry in the application.
>
> That's what I find odd with how buildroot populates the rootfs. It has
> a somewhat-hard-coded list of libraries to copy. That list is lamost
> unconditionally constructed.
>
> The proper way to handle this would be to scan all the executables aferward,
> looking at the DT_NEEDED flag in the ELF header, and copying the missing
> libraries.
>
> Of course, this is tricky, because of the way buildroot handles the sysroot.
> Every time the compiler is called, the sysroot is overloaded to point to the
> staging area. This totaly highjacks the purpose of the sysroot.
>
> In that case, there is no need to pre-populate the rootfs with the libraries
> from the sysroot, it should be done only at the end.
Note that this step would need to be done at each 'make' execution, to
cater for external applications that were copied to the rootfs
manually (which is precisely my case).
>
> For example., that's what the 'populate' script in crosstool-NG is for:
> create a copy of the rootfs, and scan the sysroot for missing libraries.
>
>> The appropriate header is located next to the other toolchain headers,
>> which are already copied by default:
>> usr/powerpc-linux/sysroot/usr/include/execinfo.h
>
> Yes, because it is a system library, so it is copied as part of the initial
> staging filling.
>
>> Depending on the uClibc configuration, less or more libraries may need
>> to be copied. I can imagine that a similar thing is needed for other C
>> libraries. Hence this generic option I added.
>>
>> Given this background information, how do you feel about this patch?
>
> Sadly, this has to go in for now. I'll be creating a new thread with an
> alternate proposal, but it will require a bit more work, so I'll Ack
> this for now.
Ok. Thanks for your ack.
> Currently, this hack applies only to custom external toolchain, because
> none of the "automatic" external toolchain currently bundles uClibc-0.9.32.
I think this would be stated more correctly as:
"... because none of the automatic external toolchains currently add
libraries that are not present in the explicit LIB_EXTERNAL_LIB /
USR_LIB_EXTERNAL_LIB variables."
Best regards,
Thomas
^ permalink raw reply
* [Buildroot] [PATCH] package: add the duma memory debugging library
From: Baruch Siach @ 2011-10-05 6:20 UTC (permalink / raw)
To: buildroot
Cc: Yann E. MORIN <yann.morin.1998@anciens.enib.fr>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
Hi Yann,
Might this collide somehow with the optional duma support of crosstool-ng?
baruch
package/Config.in | 1 +
package/duma/Config.in | 15 +++++++++++
package/duma/duma-2_5_15-cross-compile.patch | 36 ++++++++++++++++++++++++++
package/duma/duma-2_5_15-separate-cpp.patch | 23 ++++++++++++++++
package/duma/duma.mk | 36 ++++++++++++++++++++++++++
5 files changed, 111 insertions(+), 0 deletions(-)
create mode 100644 package/duma/Config.in
create mode 100644 package/duma/duma-2_5_15-cross-compile.patch
create mode 100644 package/duma/duma-2_5_15-separate-cpp.patch
create mode 100644 package/duma/duma.mk
diff --git a/package/Config.in b/package/Config.in
index fd920c3..d565216 100644
--- a/package/Config.in
+++ b/package/Config.in
@@ -20,6 +20,7 @@ menu "Debugging, profiling and benchmark"
source "package/bonnie/Config.in"
source "package/dhrystone/Config.in"
source "package/dmalloc/Config.in"
+source "package/duma/Config.in"
source "package/kexec/Config.in"
source "package/lmbench/Config.in"
source "package/lsof/Config.in"
diff --git a/package/duma/Config.in b/package/duma/Config.in
new file mode 100644
index 0000000..7f7a765
--- /dev/null
+++ b/package/duma/Config.in
@@ -0,0 +1,15 @@
+config BR2_PACKAGE_DUMA
+ bool "duma"
+ help
+ D.U.M.A. - Detect Unintended Memory Access. A fork of the Electric
+ Fence library. Detects buffer overflow and underflow, and also memory
+ leaks.
+
+ http://duma.sourceforge.net
+
+if BR2_PACKAGE_DUMA
+
+config BR2_PACKAGE_DUMA_NO_LEAKDETECTION
+ bool "disable memory leak detection"
+
+endif # BR2_PACKAGE_DUMA
diff --git a/package/duma/duma-2_5_15-cross-compile.patch b/package/duma/duma-2_5_15-cross-compile.patch
new file mode 100644
index 0000000..d50ac20
--- /dev/null
+++ b/package/duma/duma-2_5_15-cross-compile.patch
@@ -0,0 +1,36 @@
+Allow cross compilation. Taken from crosstool-ng.
+
+Signed-off-by: Baruch Siach <baruch@tkos.co.il>
+---
+diff -durN duma_2_5_15.orig/GNUmakefile duma_2_5_15/GNUmakefile
+--- duma_2_5_15.orig/GNUmakefile 2008-08-03 21:22:38.000000000 +0200
++++ duma_2_5_15/GNUmakefile 2009-06-19 15:32:23.000000000 +0200
+@@ -93,10 +93,6 @@
+ # also define 'WIN32'
+
+ # some defaults:
+-CC=gcc
+-CXX=g++
+-AR=ar
+-RANLIB=ranlib
+ INSTALL=install
+ RM=rm
+ RMFORCE=rm -f
+@@ -471,7 +467,7 @@
+
+ createconf$(EXEPOSTFIX): createconf.o
+ - $(RMFORCE) createconf$(EXEPOSTFIX)
+- $(CC) $(CFLAGS) $(DUMA_OPTIONS) createconf.o -o createconf$(EXEPOSTFIX)
++ $(HOSTCC) $(HOSTCFLAGS) $(DUMA_OPTIONS) createconf.o -o createconf$(EXEPOSTFIX)
+
+ tstheap$(EXEPOSTFIX): libduma.a tstheap.o
+ - $(RMFORCE) tstheap$(EXEPOSTFIX)
+@@ -532,7 +528,7 @@
+ # define rules how to build objects for createconf
+ #
+ createconf.o:
+- $(CC) $(CFLAGS) $(DUMA_OPTIONS) -c createconf.c -o $@
++ $(HOSTCC) $(HOSTCFLAGS) $(DUMA_OPTIONS) -c createconf.c -o $@
+
+
+ #
diff --git a/package/duma/duma-2_5_15-separate-cpp.patch b/package/duma/duma-2_5_15-separate-cpp.patch
new file mode 100644
index 0000000..ad10882
--- /dev/null
+++ b/package/duma/duma-2_5_15-separate-cpp.patch
@@ -0,0 +1,23 @@
+Allow build without C++ support. Taken from crosstool-ng.
+
+Signed-off-by: Baruch Siach <baruch@tkos.co.il>
+---
+diff -durN duma_2_5_15.orig/GNUmakefile duma_2_5_15/GNUmakefile
+--- duma_2_5_15.orig/GNUmakefile 2009-06-19 16:41:53.000000000 +0200
++++ duma_2_5_15/GNUmakefile 2009-06-19 16:43:53.000000000 +0200
+@@ -289,9 +289,13 @@
+ dumatest.c tstheap.c thread-test.c testmt.c dumatestpp.cpp testoperators.cpp \
+ createconf.c
+
+-OBJECTS=dumapp.o duma.o sem_inc.o print.o
++OBJECTS = duma.o sem_inc.o print.o
++SO_OBJECTS = duma_so.o sem_inc_so.o print_so.o
+
+-SO_OBJECTS=dumapp_so.o duma_so.o sem_inc_so.o print_so.o
++ifeq ($(DUMA_CPP),1)
++OBJECTS += dumapp.o
++SO_OBJECTS += dumapp_so.o
++endif
+
+ # Make all the top-level targets the makefile knows about.
+ all: libduma.a tstheap$(EXEPOSTFIX) dumatest$(EXEPOSTFIX) thread-test$(EXEPOSTFIX) testmt$(EXEPOSTFIX) dumatestpp$(EXEPOSTFIX) testoperators$(EXEPOSTFIX) $(DUMA_DYN_DEPS)
diff --git a/package/duma/duma.mk b/package/duma/duma.mk
new file mode 100644
index 0000000..3b8675b
--- /dev/null
+++ b/package/duma/duma.mk
@@ -0,0 +1,36 @@
+#############################################################
+#
+# duma
+#
+#############################################################
+DUMA_VERSION = 2_5_15
+DUMA_SOURCE = duma_$(DUMA_VERSION).tar.gz
+DUMA_SITE = http://$(BR2_SOURCEFORGE_MIRROR).dl.sourceforge.net/sourceforge/duma
+DUMA_INSTALL_STAGING = YES
+
+DUMA_OPTIONS = $(if $(BR2_PACKAGE_DUMA_NO_LEAKDETECTION), \
+ -DDUMA_LIB_NO_LEAKDETECTION)
+
+ifeq ($(BR2_INSTALL_LIBSTDCPP),y)
+DUMA_CPP = DUMA_CPP=1
+else
+DUMA_CPP =
+endif
+
+# The dependency of some source files in duma_config.h, which is generated at
+# build time, is not specified in the Makefile. Force non-parallel build.
+define DUMA_BUILD_CMDS
+ $(MAKE1) CC=$(TARGET_CC) CXX=$(TARGET_CXX) AR=$(TARGET_AR) \
+ RANLIB=$(TARGET_RANLIB) DUMA_OPTIONS=$(strip $(DUMA_OPTIONS)) \
+ $(DUMA_CPP) -C $(@D)
+endef
+
+define DUMA_INSTALL_STAGING_CMDS
+ $(MAKE) prefix=$(STAGING_DIR) install -C $(@D)
+endef
+
+define DUMA_INSTALL_TARGET_CMDS
+ $(MAKE) prefix=$(TARGET_DIR) install -C $(@D)
+endef
+
+$(eval $(call GENTARGETS))
--
1.7.6.3
^ permalink raw reply related
* [Buildroot] modprobe doesn't work with MIPS codesourcery
From: Satoshi Matsushita @ 2011-10-05 1:16 UTC (permalink / raw)
To: buildroot
Hello,
I'd like to use buildroot-2011.05 with Linux2.6.28
on the MIPS CPU system.
As for the toolchain, I planed to use "CodeSourcery
MIPS 2011.03". So I changed options as follows from
default settings:
Target Architecture (mipsel)
Target Architecture Variant (mips 32r2)
Toolchain type (External toolchain)
Toolchain (CodeSourcery MIPS 2011.03)
Filesystem images (jffs2 root filesystem)
As the result, the build process itself has been
finished successfully. But when the Linux boot with
this image, the following messages was displayed and
boot process was failed.
------------------------------------------------------
VFS: Mounted root (jffs2 filesystem) readonly.
Freeing prom memory: 956k freed
Freeing unused kernel memory: 136k freed
request_module: runaway loop modprobe binfmt-464c
request_module: runaway loop modprobe binfmt-464c
request_module: runaway loop modprobe binfmt-464c
request_module: runaway loop modprobe binfmt-464c
request_module: runaway loop modprobe binfmt-464c
------------------------------------------------------
If I could have any suggestions to solve this problem,
it would very helpful for me.
BTW, when I used crosstool-ng as toolchain, Linux booted
without problem.
Best Regards,
Satoshi Matsushita
^ permalink raw reply
* [Buildroot] Question: := vs = in package *.mk files
From: H Hartley Sweeten @ 2011-10-05 0:55 UTC (permalink / raw)
To: buildroot
Hello all,
The Buildroot documentation shows using '=' when setting variables
in the package *.mk file. Some of the existing packages use ':='
instead. What's the difference? Should the existing packages be
updated?
Thanks,
Hartley
^ permalink raw reply
* [Buildroot] [PATCH] pkg-stats: update grep tests for package type
From: H Hartley Sweeten @ 2011-10-04 23:10 UTC (permalink / raw)
To: buildroot
Update the grep tests used to determine the package type.
The package name and directory are now worked out magically due to:
package: add helper functions to get package name and directory magically
Because of this the extra arguments were removed by patches:
package: remove useless arguments from GENTARGETS
package: remove useless arguments from AUTOTARGETS
package: remove useless arguments from CMAKETARGETS
Signed-off-by: H Hartley Sweeten <hsweeten@visionengravers.com>
---
diff --git a/support/scripts/pkg-stats b/support/scripts/pkg-stats
index bb1c27d..8602b0f 100755
--- a/support/scripts/pkg-stats
+++ b/support/scripts/pkg-stats
@@ -98,27 +98,27 @@ for i in $(find package/ -name '*.mk') ; do
is_manual_target=0
is_manual_host=0
- if grep -E "\(call AUTOTARGETS,[^,]*,[^,]*,host\)" $i > /dev/null ; then
+ if grep -E "\(call AUTOTARGETS,host\)" $i > /dev/null ; then
is_auto_host=1
fi
- if grep -E "\(call AUTOTARGETS,[^,]*,[^,]*(,target|)\)" $i > /dev/null ; then
+ if grep -E "\(call AUTOTARGETS\)" $i > /dev/null ; then
is_auto_target=1
fi
- if grep -E "\(call GENTARGETS,[^,]*,[^,]*,host\)" $i > /dev/null ; then
+ if grep -E "\(call GENTARGETS,host\)" $i > /dev/null ; then
is_pkg_host=1
fi
- if grep -E "\(call GENTARGETS,[^,]*,[^,]*(,target|)\)" $i > /dev/null ; then
+ if grep -E "\(call GENTARGETS\)" $i > /dev/null ; then
is_pkg_target=1
fi
- if grep -E "\(call CMAKETARGETS,[^,]*,[^,]*,host\)" $i > /dev/null ; then
+ if grep -E "\(call CMAKETARGETS,host\)" $i > /dev/null ; then
is_cmake_host=1
fi
- if grep -E "\(call CMAKETARGETS,[^,]*,[^,]*(,target|)\)" $i > /dev/null ; then
+ if grep -E "\(call CMAKETARGETS\)" $i > /dev/null ; then
is_cmake_target=1
fi
^ permalink raw reply related
* [Buildroot] Bug in pkg-stats?
From: H Hartley Sweeten @ 2011-10-04 23:05 UTC (permalink / raw)
To: buildroot
In-Reply-To: <ADE657CA350FB648AAC2C43247A983F001F39DC56F52@AUSP01VMBX24.collaborationhost.net>
On Tuesday, October 04, 2011 3:36 PM, H Hartley Sweeten wrote:
>
> Hello all,
>
> The output of pkg-stats shows almost every package with the following status:
>
> AUTOTARGETS GENTARGETS CMAKETARGETS manual
> target host target host target host target Actions
> NO NO NO NO NO NO NO YES * convert to generic target
>
> I don't think this output is correct but I can't figure out what's wrong
> with the script.
Thomas,
I think the problem with pkg-stats is due to your patches:
package: remove useless arguments from CMAKETARGETS
package: remove useless arguments from AUTOTARGETS
package: remove useless arguments from GENTARGETS
package: add helper functions to get package name and directory magically
The grep tests are still looking for the package name and directory to
determine the package type.
I'll post a patch shortly.
One other thing with pkg-stats.
Can you think of any way to pull the _VERSION of the packages out of the *.mk
files and show that information in the output html file?
Regards,
Hartley
^ permalink raw reply
* [Buildroot] Bug in pkg-stats?
From: H Hartley Sweeten @ 2011-10-04 22:35 UTC (permalink / raw)
To: buildroot
Hello all,
The output of pkg-stats shows almost every package with the following status:
AUTOTARGETS GENTARGETS CMAKETARGETS manual
target host target host target host target Actions
NO NO NO NO NO NO NO YES * convert to generic target
I don't think this output is correct but I can't figure out what's wrong
with the script.
Regards,
Hartley
^ permalink raw reply
* [Buildroot] [git commit] imagemagick: add optional fftw support
From: Peter Korsgaard @ 2011-10-04 22:15 UTC (permalink / raw)
To: buildroot
commit: http://git.buildroot.net/buildroot/commit/?id=d1c54ade86e8cae6f0aa1bf1709941c70cea1af7
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
Signed-off-by: Peter Korsgaard <jacmet@sunsite.dk>
---
package/imagemagick/imagemagick.mk | 9 +++++++++
1 files changed, 9 insertions(+), 0 deletions(-)
diff --git a/package/imagemagick/imagemagick.mk b/package/imagemagick/imagemagick.mk
index 0bb76f3..b2a65e1 100644
--- a/package/imagemagick/imagemagick.mk
+++ b/package/imagemagick/imagemagick.mk
@@ -80,6 +80,15 @@ else
IMAGEMAGICK_CONF_OPT += --without-tiff
endif
+ifeq ($(BR2_PACKAGE_FFTW),y)
+# configure script misdetects these leading to build errors
+IMAGEMAGICK_CONF_ENV += ac_cv_func_creal=yes ac_cv_func_cimag=yes
+IMAGEMAGICK_CONF_OPT += --with-fftw
+IMAGEMAGICK_DEPENDENCIES += fftw
+else
+IMAGEMAGICK_CONF_OPT += --without-fftw
+endif
+
define IMAGEMAGICK_REMOVE_CONFIG_SCRIPTS
$(RM) -f $(addprefix $(TARGET_DIR)/usr/bin/, \
$(addsuffix -config, \
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
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).