A few miscellaneous fixes
PATCH 1/3 autotools: fix multi-word arguments for EXTRA_OECONF
In reply to Christopher LarsonThis is needed to better support things like the following (with a multi-word BUILD_CC):
EXTRA_OECONF += '"ac_cv_prog_CC_FOR_BUILD=${BUILD_CC}"'--- meta/classes/autotools.bbclass | 14 ++++++-------- 1 files changed, 6 insertions(+), 8 deletions(-)
meta/classes/autotools.bbclass +6 -8
@@ -70,14 +70,12 @@ CONFIGUREOPT_DEPTRACK = "--disable-dependency-tracking" oe_runconf () { - if [ -x ${S}/configure ] ; then - cfgcmd="${S}/configure \ - ${CONFIGUREOPTS} ${EXTRA_OECONF} $@" - bbnote "Running $cfgcmd..." - $cfgcmd || bbfatal "oe_runconf failed" - else - bbfatal "no configure script found" - fi + if [ -x ${S}/configure ] ; then + bbnote "Running ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} $@" + ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed" + else + bbfatal "no configure script found" + fi } autotools_do_configure() {
Signature · 1 line
1.7.4.1
PATCH 2/3 autoconf: no need to hardcode the full path to m4
In reply to Christopher LarsonThis way autom4te uses m4 as it finds it in the PATH, rather than hardcoding any particular path.
--- meta/recipes-devtools/autoconf/autoconf_2.68.bb | 4 +++- 1 files changed, 3 insertions(+), 1 deletions(-)
meta/recipes-devtools/autoconf/autoconf_2.68.bb +3 -1
@@ -1,6 +1,6 @@ require autoconf.inc -PR = "r2" +PR = "r3" PARALLEL_MAKE = ""
@@ -27,4 +27,6 @@ RDEPENDS_${PN}_virtclass-native = "m4-native gnu-config-native" SRC_URI_append_virtclass-native = " file://fix_path_xtra.patch" +EXTRA_OECONF += "ac_cv_path_M4=m4" + BBCLASSEXTEND = "native"
Signature · 1 line
1.7.4.1
cover A few miscellaneous fixes
The following changes since commit cc626b9e1671670a931ea3e528ea4b0f7b2e923b:
webkit-gtk: Enable dependency tracking since the webkit makefiles have bugs (2011-10-05 14:36:18 +0100)
are available in the git repository at:
http://github.com/kergoth/oe-core misc-fixes
Christopher Larson (3):
autotools: fix multi-word arguments for EXTRA_OECONF
autoconf: no need to hardcode the full path to m4 oe.patch: drop bb.msg.domain reference
meta/classes/autotools.bbclass | 14 ++++++-------- meta/lib/oe/patch.py | 4 ++-- meta/recipes-devtools/autoconf/autoconf_2.68.bb | 4 +++- 3 files changed, 11 insertions(+), 11 deletions(-)
Signature · 1 line
1.7.4.1
PATCH 3/3 oe.patch: drop bb.msg.domain reference
--- meta/lib/oe/patch.py | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-)
meta/lib/oe/patch.py +2 -2
@@ -376,8 +376,8 @@ class UserResolver(Resolver): os.environ['SHELLCMDS'] = "bash --rcfile " + rcfile rc = os.system(bb.data.getVar('TERMCMDRUN', self.patchset.d, 1)) if os.WIFEXITED(rc) and os.WEXITSTATUS(rc) != 0: - bb.msg.fatal(bb.msg.domain.Build, ("Cannot proceed with manual patch resolution - '%s' not found. " \ - + "Check TERMCMDRUN variable.") % bb.data.getVar('TERMCMDRUN', self.patchset.d, 1)) + bb.fatal(("Cannot proceed with manual patch resolution - '%s' not found. " + + "Check TERMCMDRUN variable.") % bb.data.getVar('TERMCMDRUN', self.patchset.d, 1)) # Construct a new PatchSet after the user's changes, compare the # sets, checking patches for modifications, and doing a remote
Signature · 1 line
1.7.4.1
Re: [PATCH 1/3] autotools: fix multi-word arguments for EXTRA_OECONF
In reply to Christopher LarsonOp 7 okt 2011, om 23:02 heeft Christopher Larson het volgende geschreven:
Quoted 32 lines
This is needed to better support things like the following (with a multi-word
BUILD_CC):EXTRA_OECONF += '"ac_cv_prog_CC_FOR_BUILD=${BUILD_CC}"'Signed-off-by Christopher Larson <[email protected]>---
meta/classes/autotools.bbclass | 14 ++++++--------1 files changed, 6 insertions(+), 8 deletions(-)meta/classes/autotools.bbclass +0
-0@@ -70,14 +70,12 @@ CONFIGUREOPT_DEPTRACK = "--disable-dependency-tracking"
oe_runconf() {meta/classes/autotools.bbclass +2
-8- if [ -x ${S}/configure ] ; then - cfgcmd="${S}/configure \ - ${CONFIGUREOPTS} ${EXTRA_OECONF} $@" - bbnote "Running $cfgcmd..." - $cfgcmd || bbfatal "oe_runconf failed" - else - bbfatal "no configure script found" - fi + if [ -x ${S}/configure ] ; then + bbnote "Running ${S}/configure ${CONFIGUREOPTS} ${
EXTRA_OECONF} $@"meta/classes/autotools.bbclass +1
-0+ ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" ||bbfatal "
oe_runconffailed"
meta/classes/autotools.bbclass CONFIGUREOPT_DEPTRACK = "--disable-dependency-tracking"
Quoted 3 lines
meta/classes/autotools.bbclass +3
-0+ else + bbfatal "no configure script found" + f
Was the tab->spaces conversion intentional?
Re: [PATCH 3/3] oe.patch: drop bb.msg.domain reference
In reply to Christopher LarsonHi,
there was already a patch for this [1] (Yes I know, I should have added more descriptions), according to the patch from Richard [2] the message domain can be constructed dynamical now.
Since your patch seems a little bit different (dropping the message domain completely) I can't tell if my resolution was inadequate.
For this patch I think it wouldn't make a big difference for the user, feel free to apply the one that best suits in resolving the problem.
[1] http://lists.linuxtogo.org/pipermail/openembedded-core/2011-September/009474.html
[2] http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=a6c48298b17e6a5844b3638b422fe226e3b67b89
2011/10/7 Christopher Larson <[email protected]>:
Quoted 29 lines
Signed-off-by Christopher Larson <[email protected]>---
meta/lib/oe/patch.py | 4 ++--1 files changed, 2 insertions(+), 2 deletions(-)meta/lib/oe/patch.py +0
-0@@ -376,8 +376,8 @@ class UserResolver(Resolver):os.environ['SHELLCMDS'] = "bash --rcfile " + rcfile
rc =os.system(bb.data.getVar('TERMCMDRUN', self.patchset.d, 1))
ifos.WIFEXITED(rc)andos.WEXITSTATUS(rc)!= 0:
- bb.msg.fatal(bb.msg.domain.Build, ("Cannot proceed with manual patch resolution - '%s' not found. " \
- + "Check TERMCMDRUN variable.") %bb.data.getVar('TERMCMDRUN', self.patchset.d, 1))
+ bb.fatal(("Cannot proceed with manual patch resolution - '%s' not found. "
+ + "Check TERMCMDRUN variable.") %bb.data.getVar('TERMCMDRUN', self.patchset.d, 1))# Construct a new PatchSet after the user's changes, compare the # sets, checking patches for modifications, and doing a remote --
1.7.4.1_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core
Signature · 2 lines
Regards Samuel
Re: [PATCH 1/3] autotools: fix multi-word arguments for EXTRA_OECONF
In reply to Koen KooiQuoted 19 lines · Koen Kooi
Quoted 10 lines
oe_runconf() {
- if [ -x ${S}/configure ] ; then
- cfgcmd="${S}/configure \
- ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
- bbnote "Running $cfgcmd..."
- $cfgcmd || bbfatal "oe_runconffailed"
- else
- bbfatal "no configure script found"
- fiQuoted 8 lines
+ if [ -x ${S}/configure ] ; then
+ bbnote "Running ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
+ ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconffailed"
+ else
+ bbfatal "no configure script found"
+ f
Quoted 1 line
Was the
tab->spacesconversion intentional?
Erm, nope, well spotted, thanks.
Signature · 5 lines
Christopher Larson clarson at kergoth dot com Founder - BitBake, OpenEmbedded, OpenZaurus Maintainer - Tslib Senior Software Engineer, Mentor Graphics
Re: [PATCH 3/3] oe.patch: drop bb.msg.domain reference
In reply to Samuel StirtzelQuoted 7 lines · Samuel Stirtzel
Hi,
there was already a patch for this [1] (Yes I know, I should have added more descriptions), according to the patch from Richard [2] the message domain can be constructed dynamical now.
Since your patch seems a little bit different (dropping the message domain completely) I can't tell if my resolution was inadequate.
Quoted 6 lines
For this patch I think it wouldn't make a big difference for the user, feel free to apply the one that best suits in resolving the problem.
[1] http://lists.linuxtogo.org/pipermail/openembedded-core/2011-September/009474.html
[2] http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=a6c48298b17e6a5844b3638b422fe226e3b67b89
Indeed, your version is superior.
For what it's worth, consider both of those patches of yours acked by me. Getting useful error messages rather than python syntax errors is rather important.
Signature · 5 lines
Christopher Larson clarson at kergoth dot com Founder - BitBake, OpenEmbedded, OpenZaurus Maintainer - Tslib Senior Software Engineer, Mentor Graphics