[PATCH 0/4] Cleanups: grub, autotools, meta-environment

[PATCH 0/4] Cleanups: grub, autotools, meta-environment

[Darren Hart]

During grub feature development I ran across various bugs, typos, thinkos,
ommisions, etc. that I've collected in this series.

The following changes since commit 2864ff6a4b3c3f9b3bbb6d2597243cc5d3715939:

  getVar/setVar cleanups (2011-11-26 22:42:00 +0000)

are available in the git repository at:
  git://git.yoctoproject.org/user-contrib/dvhart/oe-core cleanup
  http://git.yoctoproject.org/cgit.cgi/user-contrib/dvhart/oe-core/log/?h=cleanup

Darren Hart (4):
  meta-environment: Fix a typo in do_populate_sysroot[noexec]
  autotools.bbclass: Report the missing configure path
  grub: Use COMPATIBLE_MACHINE
  grub: Drop "apply=yes" from patch

 meta/classes/autotools.bbclass             |    9 +++++----
 meta/recipes-bsp/grub/grub_1.99.bb         |   11 +++--------
 meta/recipes-core/meta/meta-environment.bb |    2 +-
 3 files changed, 9 insertions(+), 13 deletions(-)

-- 
1.7.6.4

[PATCH 1/4] meta-environment: Fix a typo in do_populate_sysroot[noexec]

[Darren Hart]

From: Darren Hart <dvhart@linux.intel.com>

I noticed with while grepping for usage of noexec. Replace
do_populage_sysroot with do_populate_sysroot.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
---
 meta/recipes-core/meta/meta-environment.bb |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/meta/recipes-core/meta/meta-environment.bb b/meta/recipes-core/meta/meta-environment.bb
index 39ba96c..8f9391a 100644
--- a/meta/recipes-core/meta/meta-environment.bb
+++ b/meta/recipes-core/meta/meta-environment.bb
@@ -48,4 +48,4 @@ do_unpack[noexec] = "1"
 do_patch[noexec] = "1"
 do_configure[noexec] = "1"
 do_compile[noexec] = "1"
-do_populage_sysroot[noexec] = "1"
+do_populate_sysroot[noexec] = "1"
-- 
1.7.6.4

[PATCH 2/4] autotools.bbclass: Report the missing configure path

[Darren Hart]

From: Darren Hart <dvhart@linux.intel.com>

If the configure script isn't found, report the explicit path tried.
This can help debug subtle errors where the ${S} sourcedir may not
be exactly what is expected.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
---
 meta/classes/autotools.bbclass |    9 +++++----
 1 files changed, 5 insertions(+), 4 deletions(-)

diff --git a/meta/classes/autotools.bbclass b/meta/classes/autotools.bbclass
index 37e7d4b..7536bac 100644
--- a/meta/classes/autotools.bbclass
+++ b/meta/classes/autotools.bbclass
@@ -70,11 +70,12 @@ CONFIGUREOPT_DEPTRACK = "--disable-dependency-tracking"
 
 
 oe_runconf () {
-	if [ -x ${S}/configure ] ; then
-		bbnote "Running ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
-		${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed"
+	cfgscript="${S}/configure"
+	if [ -x "$cfgscript" ] ; then
+		bbnote "Running $cfgscript ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
+		$cfgscript ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed"
 	else
-		bbfatal "no configure script found"
+		bbfatal "no configure script found at $cfgscript"
 	fi
 }
 
-- 
1.7.6.4

[PATCH 3/4] grub: Use COMPATIBLE_MACHINE

[Darren Hart]

From: Darren Hart <dvhart@linux.intel.com>

Drop the anonymous python block in favor of the much simpler
COMPATIBLE_MACHINE mechanism.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tom Zanussi <tom.zanussi@intel.com>
---
 meta/recipes-bsp/grub/grub_1.99.bb |    9 ++-------
 1 files changed, 2 insertions(+), 7 deletions(-)

diff --git a/meta/recipes-bsp/grub/grub_1.99.bb b/meta/recipes-bsp/grub/grub_1.99.bb
index b6aa827..66ad669 100644
--- a/meta/recipes-bsp/grub/grub_1.99.bb
+++ b/meta/recipes-bsp/grub/grub_1.99.bb
@@ -22,6 +22,8 @@ SRC_URI = "ftp://ftp.gnu.org/gnu/grub/grub-${PV}.tar.gz \
 SRC_URI[md5sum] = "ca9f2a2d571b57fc5c53212d1d22e2b5"
 SRC_URI[sha256sum] = "b91f420f2c51f6155e088e34ff99bea09cc1fb89585cf7c0179644e57abd28ff"
 
+COMPATIBLE_HOST = '(x86_64.*|i.86.*)-(linux|freebsd.*)'
+
 inherit autotools
 inherit gettext
 
@@ -31,13 +33,6 @@ do_configure() {
     oe_runconf
 }
 
-python __anonymous () {
-    import re
-    host = d.getVar('HOST_SYS', 1)
-    if not re.match('x86.64.*-linux', host) and not re.match('i.86.*-linux', host):
-        raise bb.parse.SkipPackage("incompatible with host %s" % host)
-}
-
 do_install_append () {
     install -m 0755 ${WORKDIR}/40_custom ${D}${sysconfdir}/grub.d/40_custom
 }
-- 
1.7.6.4

[PATCH 4/4] grub: Drop "apply=yes" from patch

[Darren Hart]

From: Darren Hart <dvhart@linux.intel.com>

The "apply=yes" doesn't appear to be necessary, drop it in favor of
a simpler SRC_URI specification.

Signed-off-by: Darren Hart <dvhart@linux.intel.com>
CC: Tom Zanussi <tom.zanussi@intel.com>
---
 meta/recipes-bsp/grub/grub_1.99.bb |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/meta/recipes-bsp/grub/grub_1.99.bb b/meta/recipes-bsp/grub/grub_1.99.bb
index 66ad669..49c26f6 100644
--- a/meta/recipes-bsp/grub/grub_1.99.bb
+++ b/meta/recipes-bsp/grub/grub_1.99.bb
@@ -16,7 +16,7 @@ RDEPENDS_${PN} = "diffutils freetype"
 PR = "r2"
 
 SRC_URI = "ftp://ftp.gnu.org/gnu/grub/grub-${PV}.tar.gz \
-          file://grub-install.in.patch;apply=yes \
+          file://grub-install.in.patch \
           file://40_custom"
 
 SRC_URI[md5sum] = "ca9f2a2d571b57fc5c53212d1d22e2b5"
-- 
1.7.6.4

Re: [PATCH 3/4] grub: Use COMPATIBLE_MACHINE

[Koen Kooi]

[-- Attachment #1: Type: text/plain, Size: 457 bytes --]


Op 30 nov. 2011, om 22:49 heeft Darren Hart het volgende geschreven:

> From: Darren Hart <dvhart@linux.intel.com>
> 
> Drop the anonymous python block in favor of the much simpler
> COMPATIBLE_MACHINE mechanism.

minor nit: when I read the subject my first thought was "please use COMPATIBLE_HOST instead", but the actual patch is doing the right thing. Could you please amend the commit summary to has HOST instead of MACHINE?

regards,

Koen

[-- Attachment #2: Message signed with OpenPGP using GPGMail --]
[-- Type: application/pgp-signature, Size: 169 bytes --]

Re: [PATCH 3/4] grub: Use COMPATIBLE_MACHINE

[Darren Hart]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 11/30/2011 01:57 PM, Koen Kooi wrote:
> 
> Op 30 nov. 2011, om 22:49 heeft Darren Hart het volgende
> geschreven:
> 
>> From: Darren Hart <dvhart@linux.intel.com>
>> 
>> Drop the anonymous python block in favor of the much simpler 
>> COMPATIBLE_MACHINE mechanism.
> 
> minor nit: when I read the subject my first thought was "please
> use COMPATIBLE_HOST instead", but the actual patch is doing the
> right thing. Could you please amend the commit summary to has HOST
> instead of MACHINE?

Thanks for the catch Koen!

I've rebased and pushed this series to the same location:

http://git.yoctoproject.org/cgit.cgi/user-contrib/dvhart/oe-core/commit/?h=cleanup&id=46d4be3925dec1c9b6fd228b0fd603e8f2416ed5


- -- 
Darren "Dufus" Hart
Intel Open Source Technology Center
Yocto Project - Linux Kernel
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iQEcBAEBAgAGBQJO1qhZAAoJEKbMaAwKp364fxwH/0/8mhWaZkm3xu8+PRZxpCle
8lSiLZ8unO4qfUtXdXG96tlAHWlhAQfl1y3Zo4EoKcXoiSnV6rQX67nFrRnUAdge
+E0uLYYJlLQUa9Yrbt8NbuR7TktiP28AK9/An7ljk/hNbyDEQQDTDwac02IhJmaC
rBc1L3P/c6LuiN8hqCdZu4OXfahMndMkK3lzXaU0TVddfNr5o+jYiPmDRL9Dn0/W
b75ogSHpKAzHhFEsOf563DPpdfagiz2e0B4t6SMBIxRDcgHQHmNIo6lvjfDfzFYc
e7B3rA3nmWnHctkrDjs4GTO5wecfO1Xj31bpsPX6C7jW4f0BKrmdpk+PV/TGeUA=
=Ingz
-----END PGP SIGNATURE-----

Re: [PATCH 0/4] Cleanups: grub, autotools, meta-environment

[Richard Purdie]

On Wed, 2011-11-30 at 13:49 -0800, Darren Hart wrote:
> During grub feature development I ran across various bugs, typos, thinkos,
> ommisions, etc. that I've collected in this series.
> 
> The following changes since commit 2864ff6a4b3c3f9b3bbb6d2597243cc5d3715939:
> 
>   getVar/setVar cleanups (2011-11-26 22:42:00 +0000)
> 
> are available in the git repository at:
>   git://git.yoctoproject.org/user-contrib/dvhart/oe-core cleanup
>   http://git.yoctoproject.org/cgit.cgi/user-contrib/dvhart/oe-core/log/?h=cleanup
> 
> Darren Hart (4):
>   meta-environment: Fix a typo in do_populate_sysroot[noexec]
>   autotools.bbclass: Report the missing configure path
>   grub: Use COMPATIBLE_MACHINE
>   grub: Drop "apply=yes" from patch

Merged to master thanks (using v2 of the COMPATIBLE_MACHINE/HOST one).

I'm moving a little more rapidly with patches as there are a *lot*
appearing at the moment and I want to get some of the simpler ones
merged sooner than later. 

Cheers,

Richard

Re: [PATCH 2/4] autotools.bbclass: Report the missing configure path

[Khem Raj]

On Wed, Nov 30, 2011 at 1:49 PM, Darren Hart <darren@dvhart.com> wrote:
> From: Darren Hart <dvhart@linux.intel.com>
>
> If the configure script isn't found, report the explicit path tried.
> This can help debug subtle errors where the ${S} sourcedir may not
> be exactly what is expected.
>
> Signed-off-by: Darren Hart <dvhart@linux.intel.com>
> ---
>  meta/classes/autotools.bbclass |    9 +++++----
>  1 files changed, 5 insertions(+), 4 deletions(-)
>
> diff --git a/meta/classes/autotools.bbclass b/meta/classes/autotools.bbclass
> index 37e7d4b..7536bac 100644
> --- a/meta/classes/autotools.bbclass
> +++ b/meta/classes/autotools.bbclass
> @@ -70,11 +70,12 @@ CONFIGUREOPT_DEPTRACK = "--disable-dependency-tracking"
>
>
>  oe_runconf () {
> -       if [ -x ${S}/configure ] ; then
> -               bbnote "Running ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
> -               ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed"
> +       cfgscript="${S}/configure"
> +       if [ -x "$cfgscript" ] ; then
> +               bbnote "Running $cfgscript ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
> +               $cfgscript ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed"
>        else
> -               bbfatal "no configure script found"
> +               bbfatal "no configure script found at $cfgscript"

why not just do one line change something like

bbfatal "configure script ${S}/configure not found"

>        fi
>  }
>
> --
> 1.7.6.4
>
>
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Re: [PATCH 2/4] autotools.bbclass: Report the missing configure path

[Darren Hart]

On 11/30/2011 02:31 PM, Khem Raj wrote:
> On Wed, Nov 30, 2011 at 1:49 PM, Darren Hart <darren@dvhart.com> wrote:
>> From: Darren Hart <dvhart@linux.intel.com>
>>
>> If the configure script isn't found, report the explicit path tried.
>> This can help debug subtle errors where the ${S} sourcedir may not
>> be exactly what is expected.
>>
>> Signed-off-by: Darren Hart <dvhart@linux.intel.com>
>> ---
>>  meta/classes/autotools.bbclass |    9 +++++----
>>  1 files changed, 5 insertions(+), 4 deletions(-)
>>
>> diff --git a/meta/classes/autotools.bbclass b/meta/classes/autotools.bbclass
>> index 37e7d4b..7536bac 100644
>> --- a/meta/classes/autotools.bbclass
>> +++ b/meta/classes/autotools.bbclass
>> @@ -70,11 +70,12 @@ CONFIGUREOPT_DEPTRACK = "--disable-dependency-tracking"
>>
>>
>>  oe_runconf () {
>> -       if [ -x ${S}/configure ] ; then
>> -               bbnote "Running ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
>> -               ${S}/configure ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed"
>> +       cfgscript="${S}/configure"
>> +       if [ -x "$cfgscript" ] ; then
>> +               bbnote "Running $cfgscript ${CONFIGUREOPTS} ${EXTRA_OECONF} $@"
>> +               $cfgscript ${CONFIGUREOPTS} ${EXTRA_OECONF} "$@" || bbfatal "oe_runconf failed"
>>        else
>> -               bbfatal "no configure script found"
>> +               bbfatal "no configure script found at $cfgscript"
> 
> why not just do one line change something like
> 
> bbfatal "configure script ${S}/configure not found"

I prefer to use a variable when I need a value more than once. It avoids
future bugs where one is changed and the others were missed. As this
string is used 4 times, it seemed like the right call.

--
Darren

> 
>>        fi
>>  }
>>
>> --
>> 1.7.6.4
>>
>>
>> _______________________________________________
>> Openembedded-core mailing list
>> Openembedded-core@lists.openembedded.org
>> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

-- 
Darren Hart
Intel Open Source Technology Center
Yocto Project - Linux Kernel

Re: [PATCH 2/4] autotools.bbclass: Report the missing configure path

[Khem Raj]

>>> -               bbfatal "no configure script found"
>>> +               bbfatal "no configure script found at $cfgscript"
>>
>> why not just do one line change something like
>>
>> bbfatal "configure script ${S}/configure not found"
>
> I prefer to use a variable when I need a value more than once. It avoids
> future bugs where one is changed and the others were missed. As this
> string is used 4 times, it seemed like the right call.

ok with variable still the cfgscript itself is the script and the
message "no configure script found at $cfgscript"
seems like $cfgscript is a location. you may want to make it
"$cfgscript configure script not found" or something

Re: [PATCH 2/4] autotools.bbclass: Report the missing configure path

[Darren Hart]

On 11/30/2011 02:48 PM, Khem Raj wrote:
>>>> -               bbfatal "no configure script found"
>>>> +               bbfatal "no configure script found at $cfgscript"
>>>
>>> why not just do one line change something like
>>>
>>> bbfatal "configure script ${S}/configure not found"
>>
>> I prefer to use a variable when I need a value more than once. It avoids
>> future bugs where one is changed and the others were missed. As this
>> string is used 4 times, it seemed like the right call.
> 
> ok with variable still the cfgscript itself is the script and the
> message "no configure script found at $cfgscript"
> seems like $cfgscript is a location. you may want to make it
> "$cfgscript configure script not found" or something

I can see that (although I think can be argued this is correct as well -
a path with a filename is still a location). I was copying the language
used elsewhere... but from where is eluding my search at the moment. RP
has pulled these already I believe, so if the language here truly
bothers you, we can address it in a follow-on. Personally I don't think
it's really worth the trouble.

-- 
Darren Hart
Intel Open Source Technology Center
Yocto Project - Linux Kernel

Re: [PATCH 2/4] autotools.bbclass: Report the missing configure path

[Khem Raj]

> I can see that (although I think can be argued this is correct as well -
> a path with a filename is still a location). I was copying the language
> used elsewhere... but from where is eluding my search at the moment. RP
> has pulled these already I believe, so if the language here truly
> bothers you, we can address it in a follow-on. Personally I don't think
> it's really worth the trouble.

yeah its fine