Skip to contentSkip to conversation

gphoto2_2.4.8.bb: Stop configure from looking in /usr/local/include

PATCH In review Graham Gower 6 participants 17 messages Nov 2, 2010 – Nov 5, 2010

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 gphoto2_2.4.8.bb: Stop configure from looking in /usr/local/include

* Point configure at where it can find jpeg.
* Explicitly disable cdk and aalib.
* Fix license while here.

---
 recipes/gphoto2/gphoto2_2.4.8.bb |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)
recipes/gphoto2/gphoto2_2.4.8.bb +7 -2
index ea54713..836ea85 100644
@@ -1,7 +1,8 @@
 DESCRIPTION = "gphoto2 is a command-line utility to fetch pictures from digital cameras"
 SECTION = "console/utils"
-LICENSE = "GPL"
-DEPENDS = "libtool libgphoto2 popt"
+LICENSE = "GPLv2+"
+DEPENDS = "libtool libgphoto2 popt jpeg"
+PR = "r1"
 
 SRC_URI = "${SOURCEFORGE_MIRROR}/gphoto/gphoto2-${PV}.tar.bz2;name=gphoto2 \
            file://gphoto-popt-fixup.patch"
@@ -11,6 +12,10 @@ SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4
 
 inherit autotools
 
+EXTRA_OECONF="--without-cdk \
+		--without-aalib \
+		--with-jpeg-prefix=${STAGING_EXECPREFIXDIR}"
+
 do_configure_append() {
         find ${S} -name Makefile | xargs sed -i 's:/usr/include:${STAGING_INCDIR}:'
 }
Signature · 1 line
1.7.1
#1 lore raw

2010/11/2 Graham Gower <[email protected]>:

Quoted 16 lines

 * Point configure at where it can find jpeg.
 * Explicitly disable cdk and aalib.
 * Fix license while here.

---
 recipes/gphoto2/gphoto2_2.4.8.bb |    9 +++++++--

 1 files changed, 7 insertions(+), 2 deletions(-)
recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
index ea54713..836ea85 100644
@@ -1,7 +1,8 @@

 DESCRIPTION = "gphoto2 is a command-line utility to fetch pictures from digital cameras" SECTION = "console/utils"

recipes/gphoto2/gphoto2_2.4.8.bb line 4

Quoted 4 lines
recipes/gphoto2/gphoto2_2.4.8.bb +2 -2
-LICENSE = "GPL"
-DEPENDS = "libtool libgphoto2 popt"
+LICENSE = "GPLv2+"
+DEPENDS = "libtool libgphoto2 popt jpeg"

The original recipe would build without the depends on jpeg adding it to depends should be mentioned in the commit message

recipes/gphoto2/gphoto2_2.4.8.bb SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4 · line 18

Quoted 12 lines

+PR = "r1"

 SRC_URI = "${SOURCEFORGE_MIRROR}/gphoto/gphoto2-${PV}.tar.bz2;name=gphoto2 \
           file://gphoto-popt-fixup.patch"

recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
@@ -11,6 +12,10 @@ SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4
 

 inherit autotools

recipes/gphoto2/gphoto2_2.4.8.bb +4 -0
+EXTRA_OECONF="--without-cdk \
+               --without-aalib \
+               --with-jpeg-prefix=${STAGING_EXECPREFIXDIR}"
+

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

from --help:

Optional Features:
--disable-option-checking ignore unrecognized --enable/--with options
--disable-FEATURE do not include FEATURE (same as --enable-FEATURE=no)
--enable-FEATURE[=ARG] include FEATURE [ARG=yes] --disable-dependency-tracking speeds up one-time build
--enable-dependency-tracking do not reject slow dependency extractors
--enable-shared[=PKGS] build shared libraries [default=yes] --enable-static[=PKGS] build static libraries [default=yes] --enable-fast-install[=PKGS]
optimize for fast installation [default=yes]

  --disable-libtool-lock  avoid locking (might break parallel builds)
  --disable-largefile     omit support for large files
  --disable-nls           do not use Native Language Support
  --disable-rpath         do not hardcode runtime library paths

Optional Packages:

  --with-PACKAGE[=ARG]    use PACKAGE [ARG=yes]
  --without-PACKAGE       do not use PACKAGE (same as --with-PACKAGE=no)
  --with-pic              try to use only PIC/non-PIC objects [default=use

both]
--with-gnu-ld assume the C compiler uses GNU ld [default=no]
--with-gnu-ld assume the C compiler uses GNU ld default=no
--with-libiconv-prefix[=DIR] search for libiconv in DIR/include and DIR/lib
--without-libiconv-prefix don't search for libiconv in
includedir and libdir
--with-libintl-prefix[=DIR] search for libintl in DIR/include and DIR/lib
--without-libintl-prefix don't search for libintl in includedir and libdir
--with-libgphoto2=PREFIX
where to find libgphoto2, "no" or "auto"
--without-pthread Don't use pthread
--without-cdk Do not use cdk
--with-cdk-prefix=PREFIX
Location of cdk
--without-jpeg Do not use jpeg
--with-jpeg-prefix=PREFIX
Location of jpeg

  --with-libexif=PREFIX   where to find libexif, "no" or "auto"
  --without-aalib         Do not use aalib
  --without-readline      Do not use readline

All things that are not mentioned in DEPENDS should (IMHO) be explicitly be turned off.
And for the things in DEPENDS I would encourage to still add them. E.g. --with-jpeg. Yes I know these will always be picked up when a DEPENDS = "jpeg" is present, but making it explict does not hurt and it trips an error if jpeg is not there and someone does a bitbake -b of the recipe (instead of silently building a package without jpeg).

Thanks for your contribution & apologies for using your commit for a rant on configure flags.

Frans.

Quoted 12 lines

 do_configure_append() {
        find ${S} -name Makefile | xargs sed -i 's:/usr/include:${STAGING_INCDIR}:' }
--
1.7.1

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#2 lore raw
Quoted 33 lines · Frans Meulenbroeks

2010/11/2 Graham Gower <[email protected]>:

Quoted 16 lines

 * Point configure at where it can find jpeg.
 * Explicitly disable cdk and aalib.
 * Fix license while here.

---
 recipes/gphoto2/gphoto2_2.4.8.bb |    9 +++++++--

 1 files changed, 7 insertions(+), 2 deletions(-)
recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
index ea54713..836ea85 100644
@@ -1,7 +1,8 @@

 DESCRIPTION = "gphoto2 is a command-line utility to fetch pictures from digital cameras" SECTION = "console/utils"

recipes/gphoto2/gphoto2_2.4.8.bb line 4

Quoted 4 lines
recipes/gphoto2/gphoto2_2.4.8.bb +2 -2
-LICENSE = "GPL"
-DEPENDS = "libtool libgphoto2 popt"
+LICENSE = "GPLv2+"
+DEPENDS = "libtool libgphoto2 popt jpeg"

The original recipe would build without the depends on jpeg adding it to depends should be mentioned in the commit message

Quoted 8 lines

+PR = "r1"

 SRC_URI = "${SOURCEFORGE_MIRROR}/gphoto/gphoto2-${PV}.tar.bz2;name=gphoto2 \
           file://gphoto-popt-fixup.patch"

recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
@@ -11,6 +12,10 @@ SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4
 

 inherit autotools

Quoted 10 lines

recipes/gphoto2/gphoto2_2.4.8.bb SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4 · line 18

Quoted 4 lines
recipes/gphoto2/gphoto2_2.4.8.bb +4 -0
+EXTRA_OECONF="--without-cdk \
+               --without-aalib \
+               --with-jpeg-prefix=${STAGING_EXECPREFIXDIR}"
+

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

this is the case with any package that uses configure tools. But assumption here is that when you are building you are not building stuff thats not meant to be in
the given image.

Thx
-Khem

#15 lore raw
In reply to Frans Meulenbroeks
Quoted 93 lines · Frans Meulenbroeks

2010/11/2 Graham Gower <[email protected]>:

Quoted 16 lines

 * Point configure at where it can find jpeg.
 * Explicitly disable cdk and aalib.
 * Fix license while here.

---
 recipes/gphoto2/gphoto2_2.4.8.bb |    9 +++++++--

 1 files changed, 7 insertions(+), 2 deletions(-)
recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
index ea54713..836ea85 100644
@@ -1,7 +1,8 @@

 DESCRIPTION = "gphoto2 is a command-line utility to fetch pictures from digital cameras" SECTION = "console/utils"

recipes/gphoto2/gphoto2_2.4.8.bb line 4

Quoted 4 lines
recipes/gphoto2/gphoto2_2.4.8.bb +2 -2
-LICENSE = "GPL"
-DEPENDS = "libtool libgphoto2 popt"
+LICENSE = "GPLv2+"
+DEPENDS = "libtool libgphoto2 popt jpeg"

The original recipe would build without the depends on jpeg adding it to depends should be mentioned in the commit message

recipes/gphoto2/gphoto2_2.4.8.bb SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4 · line 18

Quoted 12 lines

+PR = "r1"

 SRC_URI = "${SOURCEFORGE_MIRROR}/gphoto/gphoto2-${PV}.tar.bz2;name=gphoto2 \
           file://gphoto-popt-fixup.patch"

recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
@@ -11,6 +12,10 @@ SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4
 

 inherit autotools

recipes/gphoto2/gphoto2_2.4.8.bb +4 -0
+EXTRA_OECONF="--without-cdk \
+               --without-aalib \
+               --with-jpeg-prefix=${STAGING_EXECPREFIXDIR}"
+

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

from --help:

Optional Features:
 --disable-option-checking  ignore unrecognized --enable/--with options --disable-FEATURE       do not include FEATURE (same as --enable-FEATURE=no) --enable-FEATURE[=ARG]  include FEATURE [ARG=yes] --disable-dependency-tracking  speeds up one-time build --enable-dependency-tracking   do not reject slow dependency extractors --enable-shared[=PKGS]  build shared libraries [default=yes] --enable-static[=PKGS]  build static libraries [default=yes] --enable-fast-install[=PKGS]
                         optimize for fast installation [default=yes] --disable-libtool-lock  avoid locking (might break parallel builds) --disable-largefile     omit support for large files --disable-nls           do not use Native Language Support --disable-rpath         do not hardcode runtime library paths

Optional Packages:
 --with-PACKAGE[=ARG]    use PACKAGE [ARG=yes]
 --without-PACKAGE       do not use PACKAGE (same as --with-PACKAGE=no) --with-pic              try to use only PIC/non-PIC objects [default=use both]
 --with-gnu-ld           assume the C compiler uses GNU ld [default=no] --with-gnu-ld           assume the C compiler uses GNU ld default=no --with-libiconv-prefix[=DIR]  search for libiconv in DIR/include and DIR/lib --without-libiconv-prefix     don't search for libiconv in includedir and libdir
 --with-libintl-prefix[=DIR]  search for libintl in DIR/include and DIR/lib --without-libintl-prefix     don't search for libintl in includedir and libdir --with-libgphoto2=PREFIX
                         where to find libgphoto2, "no" or "auto" --without-pthread         Don't use pthread
 --without-cdk           Do not use cdk
 --with-cdk-prefix=PREFIX
                         Location of cdk
 --without-jpeg          Do not use jpeg
 --with-jpeg-prefix=PREFIX
                         Location of jpeg
 --with-libexif=PREFIX   where to find libexif, "no" or "auto" --without-aalib         Do not use aalib
 --without-readline      Do not use readline

All things that are not mentioned in DEPENDS should (IMHO) be explicitly be turned off.
And for the things in DEPENDS I would encourage to still add them. E.g. --with-jpeg. Yes I know these will always be picked up when a DEPENDS = "jpeg" is present, but making it explict does not hurt and it trips an error if jpeg is not there and someone does a bitbake -b of the recipe (instead of silently building a package without jpeg).

Quoted 5 lines

Thanks for your contribution & apologies for using your commit for a rant on configure flags.

Frans.

I agree with you here. Often a recipe breaks when a particular subsystem is enabled, which is usually not enabled during regular image builds. And it can be difficult to determine if something should be added to DEPENDS or disabled at configure time.

What I found with gphoto2 (which probably wasn't clear) was the configure script was testing for jpeg and cdk with -I/usr/local/include. So I guessed on what was useful and what wasn't and added jpeg to depends to make it clear that jpeg support was intended (even though it will likely be pulled in by other dependencies).

I assume you just want exif and readline disabled too? (and possibly an essay in the commit message :))

-Graham

#16 lore raw

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

Quoted 113 lines · Graham Gower
Quoted 93 lines · Frans Meulenbroeks

2010/11/2 Graham Gower <[email protected]>:

Quoted 14 lines

* Point configure at where it can find jpeg.
* Explicitly disable cdk and aalib.
* Fix license while here.

---
 recipes/gphoto2/gphoto2_2.4.8.bb |    9 +++++++--
 1 files changed, 7 insertions(+), 2 deletions(-)
recipes/gphoto2/gphoto2_2.4.8.bb +0 -0
index ea54713..836ea85 100644

recipes/gphoto2/gphoto2_2.4.8.bb line 4

Quoted 6 lines
recipes/gphoto2/gphoto2_2.4.8.bb +2 -2
@@ -1,7 +1,8 @@
 DESCRIPTION = "gphoto2 is a command-line utility to fetch pictures from digital cameras"
 SECTION = "console/utils"
-LICENSE = "GPL"
-DEPENDS = "libtool libgphoto2 popt"
+LICENSE = "GPLv2+"
+DEPENDS = "libtool libgphoto2 popt jpeg"

The original recipe would build without the depends on jpeg adding it to depends should be mentioned in the commit message

recipes/gphoto2/gphoto2_2.4.8.bb SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4 · line 18

Quoted 12 lines
recipes/gphoto2/gphoto2_2.4.8.bb +5 -0
+PR = "r1"
 
 SRC_URI = "${SOURCEFORGE_MIRROR}/gphoto/gphoto2-${PV}.tar.bz2;name=gphoto2 \
           file://gphoto-popt-fixup.patch"
@@ -11,6 +12,10 @@ SRC_URI[gphoto2.sha256sum] = "a2e2f70fd6051234cbc3d76dc82df6991e8a07ead56aa8d5c4
 
 inherit autotools
 
+EXTRA_OECONF="--without-cdk \
+               --without-aalib \
+               --with-jpeg-prefix=${STAGING_EXECPREFIXDIR}"
+

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

from --help:

Optional Features:
--disable-option-checking ignore unrecognized --enable/--with options
--disable-FEATURE do not include FEATURE (same as --enable-FEATURE=no)
--enable-FEATURE[=ARG] include FEATURE [ARG=yes] --disable-dependency-tracking speeds up one-time build
--enable-dependency-tracking do not reject slow dependency extractors
--enable-shared[=PKGS] build shared libraries [default=yes] --enable-static[=PKGS] build static libraries [default=yes] --enable-fast-install[=PKGS]
optimize for fast installation [default=yes]

 --disable-libtool-lock  avoid locking (might break parallel builds)
 --disable-largefile     omit support for large files
 --disable-nls           do not use Native Language Support
 --disable-rpath         do not hardcode runtime library paths

Optional Packages:

 --with-PACKAGE[=ARG]    use PACKAGE [ARG=yes]
 --without-PACKAGE       do not use PACKAGE (same as --with-PACKAGE=no)
 --with-pic              try to use only PIC/non-PIC objects [default=use

both]
--with-gnu-ld assume the C compiler uses GNU ld [default=no]
--with-gnu-ld assume the C compiler uses GNU ld default=no
--with-libiconv-prefix[=DIR] search for libiconv in DIR/include and DIR/lib
--without-libiconv-prefix don't search for libiconv in
includedir and libdir
--with-libintl-prefix[=DIR] search for libintl in DIR/include and DIR/lib
--without-libintl-prefix don't search for libintl in includedir and libdir
--with-libgphoto2=PREFIX
where to find libgphoto2, "no" or "auto"
--without-pthread Don't use pthread
--without-cdk Do not use cdk
--with-cdk-prefix=PREFIX
Location of cdk
--without-jpeg Do not use jpeg
--with-jpeg-prefix=PREFIX
Location of jpeg

 --with-libexif=PREFIX   where to find libexif, "no" or "auto"
 --without-aalib         Do not use aalib
 --without-readline      Do not use readline

All things that are not mentioned in DEPENDS should (IMHO) be explicitly be turned off.
And for the things in DEPENDS I would encourage to still add them. E.g. --with-jpeg. Yes I know these will always be picked up when a DEPENDS = "jpeg" is present, but making it explict does not hurt and it trips an error if jpeg is not there and someone does a bitbake -b of the recipe (instead of silently building a package without jpeg).

Quoted 5 lines

Thanks for your contribution & apologies for using your commit for a rant on configure flags.

Frans.

I agree with you here. Often a recipe breaks when a particular subsystem is enabled, which is usually not enabled during regular image builds. And it can be difficult to determine if something should be added to DEPENDS or disabled at configure time.

What I found with gphoto2 (which probably wasn't clear) was the configure script was testing for jpeg and cdk with -I/usr/local/include. So I guessed on what was useful and what wasn't and added jpeg to depends to make it clear that jpeg support was intended (even though it will likely be pulled in by other dependencies).

Quoted 2 lines

I assume you just want exif and readline disabled too? (and possibly an essay in the commit message :))

For gphoto2 to be usefull it need exif, not sure what readline brings us.

regards,

Koen
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFM0RX/MkyGM64RGpERApdsAJ0UwmJ03Z/Q7LyJoSKssYWZQRg1wQCeL94P FIJjFYYigRNPTPrV1XXemD8=
=GDK5
-----END PGP SIGNATURE-----

#3 lore raw
In reply to Khem Raj

2010/11/2 Khem Raj <[email protected]>:

Quoted 9 lines

On Mon, Nov 1, 2010 at 11:38 PM, Frans Meulenbroeks

Quoted 5 lines

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

Quoted 4 lines

this is the case with any package that uses configure tools. But assumption here is that when you are building you are not building stuff thats not meant to be in
the given image.

True. However we should enforce that. If e.g. gphoto has no depends on exif it autotools will still pick up exif if it happens to be there because some other package build it.
Funtionally this is not necessarily bad (as your image has exif anyway), but from a point of reproducability it is undesirable. E.g. if you build the recipe the first time, you might not get exif and if you rebuild you do get exif. (because the build system does not know the dependency is there so it might not have build exif yet).

Actually the results might even differ for between builds of the same image (especially if you build machines with different speeds etc and/or use different flags for parallellism.
Therefore I feel it is highly desirable to have all optionals that you have in configure either explicitly disabled or explicitly added to depends.

Wrt exif & readline: I agree with koen.

Frans.

#4 lore raw
Quoted 24 lines · Frans Meulenbroeks

2010/11/2 Khem Raj <[email protected]>:

Quoted 9 lines

On Mon, Nov 1, 2010 at 11:38 PM, Frans Meulenbroeks

Quoted 5 lines

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

Quoted 4 lines

this is the case with any package that uses configure tools. But assumption here is that when you are building you are not building stuff thats not meant to be in
the given image.

True. However we should enforce that. If e.g. gphoto has no depends on exif it autotools will still pick up exif if it happens to be there because some other package build it.
Funtionally this is not necessarily bad (as your image has exif anyway), but from a point of reproducability it is undesirable. E.g. if you build the recipe the first time, you might not get exif and if you rebuild you do get exif. (because the build system does not know the dependency is there so it might not have build exif yet).

Quoted 7 lines

Actually the results might even differ for between builds of the same image (especially if you build machines with different speeds etc and/or use different flags for parallellism.
Therefore I feel it is highly desirable to have all optionals that you have in configure either explicitly disabled or explicitly added to depends.

this is everlasting problem with autoconf and autotools. There is a set of options
which are enabled based on what autoconf guesses on the sysroots. Unfortunately it will be very tedious if you want to explicitly enable and disable every option as well.
and there are quite a lot of packages using autotools. However I think that the sysroot is minimal
when we build from scratch so most of options should be disabled which are decided by autoconf
so we may not get unneeded features which is better than having them enabled at build time
and needed support being missing at run time.

Quoted 9 lines

Wrt exif & readline: I agree with koen.

Frans.

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#5 lore raw

I think the most elegant solution comes from something like USE flags.

See http://gentoo-portage.com/media-gfx/gphoto2 (2_2.4.9 fwiw)

...
IUSE="aalib exif ncurses nls readline"
...
exif? ( media-libs/libexif )
readline? ( sys-libs/readline )
...

Drawback is each recipe needs to be edited.

My 2 Cents

Andrea

#6 lore raw
In reply to Khem Raj
Quoted 34 lines · Khem Raj
Quoted 24 lines · Frans Meulenbroeks

2010/11/2 Khem Raj <[email protected]>:

Quoted 9 lines

On Mon, Nov 1, 2010 at 11:38 PM, Frans Meulenbroeks

Quoted 5 lines

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

Quoted 4 lines

this is the case with any package that uses configure tools. But assumption here is that when you are building you are not building stuff thats not meant to be in
the given image.

True. However we should enforce that. If e.g. gphoto has no depends on exif it autotools will still pick up exif if it happens to be there because some other package build it.
Funtionally this is not necessarily bad (as your image has exif anyway), but from a point of reproducability it is undesirable. E.g. if you build the recipe the first time, you might not get exif and if you rebuild you do get exif. (because the build system does not know the dependency is there so it might not have build exif yet).

Quoted 7 lines

Actually the results might even differ for between builds of the same image (especially if you build machines with different speeds etc and/or use different flags for parallellism.
Therefore I feel it is highly desirable to have all optionals that you have in configure either explicitly disabled or explicitly added to depends.

Quoted 12 lines

this is everlasting problem with autoconf and autotools. There is a set of options
which are enabled based on what autoconf guesses on the sysroots. Unfortunately it will be very tedious if you want to explicitly enable and disable every option as well.
and there are quite a lot of packages using autotools. However I think that the sysroot is minimal
when we build from scratch so most of options should be disabled which are decided by autoconf
so we may not get unneeded features which is better than having them enabled at build time
and needed support being missing at run time.

well but then you rebuild that package ie after just PR bump and feature gets enabled, users gets used to it and with next rebuild from scratch you will be surprised why they are complaining about missing feature (happened to me with python + tk).

So to keep them disabled even for next rebuild after PR bump we should put --disable-feature to EXTRA_OECONF to be sure.

But that would make it conflicting for maintainers who are used to have it "usually enabled" and "usually disabled".

Which can be solved by DISTRO_FEATURES but some shortcuts like gentoo has would be usefull

DEPENDS += tk? ( >=dev-lang/tk-8.0 )
instead of
DEPENDS += ${@base_contains('DISTRO_FEATURES', 'tk', 'tk', '', d)}

$(use_with threads)
instead of
${@base_contains('DISTRO_FEATURES', 'threads', '--with-threads', --without-threads)}

$(use_enable ipv6)
instead of
${@base_contains('DISTRO_FEATURES', 'ipv6', '--enable-ipv6', --disable-ipv6)}

or even
DEPEND="dev-libs/ecore[evas]"
instead of
hoping that ecore was built with evas support

Regards,

Signature · 1 line
Martin 'JaMa' Jansa     jabber: [email protected]
#7 lore raw
Quoted 48 lines · Martin Jansa
Quoted 34 lines · Khem Raj
Quoted 24 lines · Frans Meulenbroeks

2010/11/2 Khem Raj <[email protected]>:

Quoted 9 lines

On Mon, Nov 1, 2010 at 11:38 PM, Frans Meulenbroeks

Quoted 5 lines

I have some concerns here.
For configure it seems best to explicitly set and reset the flags in it, otherwise you may just pick up whatever happened to be build before.

Quoted 4 lines

this is the case with any package that uses configure tools. But assumption here is that when you are building you are not building stuff thats not meant to be in
the given image.

True. However we should enforce that. If e.g. gphoto has no depends on exif it autotools will still pick up exif if it happens to be there because some other package build it.
Funtionally this is not necessarily bad (as your image has exif anyway), but from a point of reproducability it is undesirable. E.g. if you build the recipe the first time, you might not get exif and if you rebuild you do get exif. (because the build system does not know the dependency is there so it might not have build exif yet).

Quoted 7 lines

Actually the results might even differ for between builds of the same image (especially if you build machines with different speeds etc and/or use different flags for parallellism.
Therefore I feel it is highly desirable to have all optionals that you have in configure either explicitly disabled or explicitly added to depends.

Quoted 12 lines

this is everlasting problem with autoconf and autotools. There is a set of options
which are enabled based on what autoconf guesses on the sysroots. Unfortunately it will be very tedious if you want to explicitly enable and disable every option as well.
and there are quite a lot of packages using autotools. However I think that the sysroot is minimal
when we build from scratch so most of options should be disabled which are decided by autoconf
so we may not get unneeded features which is better than having them enabled at build time
and needed support being missing at run time.

Quoted 5 lines

well but then you rebuild that package ie after just PR bump and feature gets enabled, users gets used to it and with next rebuild from scratch you will be surprised why they are complaining about missing feature (happened to me with python + tk).

yes there are many other packages including eglibc/glibc which will not configured same on rebuild.

Quoted 17 lines

So to keep them disabled even for next rebuild after PR bump we should put --disable-feature to EXTRA_OECONF to be sure.

But that would make it conflicting for maintainers who are used to have it "usually enabled" and "usually disabled".

Which can be solved by DISTRO_FEATURES but some shortcuts like gentoo has would be usefull

DEPENDS += tk? ( >=dev-lang/tk-8.0 )
instead of
DEPENDS += ${@base_contains('DISTRO_FEATURES', 'tk', 'tk', '', d)}

$(use_with threads)
instead of
${@base_contains('DISTRO_FEATURES', 'threads', '--with-threads', --without-threads)}

Quoted 9 lines

$(use_enable ipv6)
instead of
${@base_contains('DISTRO_FEATURES', 'ipv6', '--enable-ipv6', --disable-ipv6)}

or even
DEPEND="dev-libs/ecore[evas]"
instead of
hoping that ecore was built with evas support

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

Quoted 10 lines

Regards,

--
Martin 'JaMa' Jansa     jabber: [email protected]

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#8 lore raw

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

Frans

#9 lore raw
Quoted 8 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

Quoted 9 lines

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

numbers means not so much when it comes down to this. I don't know if any distribution that even does that
so I take comfort in everyone being wrong.

Quoted 7 lines

Frans

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#10 lore raw

2010/11/3 Khem Raj <[email protected]>:

Quoted 20 lines
Quoted 8 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

Quoted 9 lines

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

Quoted 4 lines

numbers means not so much when it comes down to this. I don't know if any distribution that even does that
so I take comfort in everyone being wrong.

I agree that numbers don't mean that much. Just wanted to give an indication of the possible effort needed.
Also no idea how debian does this.
One way to get a minimal package is to have everything in packaged staging then bake the recipe.
that'll only drag in stuff from depends.

Frans

#11 lore raw
Quoted 21 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 20 lines
Quoted 17 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

Quoted 8 lines
Quoted 4 lines

numbers means not so much when it comes down to this. I don't know if any distribution that even does that
so I take comfort in everyone being wrong.

I agree that numbers don't mean that much. Just wanted to give an indication of the possible effort needed.
Also no idea how debian does this.

Does it do it in first place ?

Quoted 3 lines

One way to get a minimal package is to have everything in packaged staging then bake the recipe.
that'll only drag in stuff from depends.

It depends also on the sequence of staging. So there are multiple factors and no two stagings will look same

Quoted 8 lines

Frans

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#12 lore raw

2010/11/3 Khem Raj <[email protected]>:

Quoted 28 lines
Quoted 26 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 24 lines
Quoted 8 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

Quoted 9 lines

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

numbers means not so much when it comes down to this. I don't know if any distribution that even does that
so I take comfort in everyone being wrong.

Quoted 5 lines
Quoted 3 lines

I agree that numbers don't mean that much. Just wanted to give an indication of the possible effort needed.
Also no idea how debian does this.

Does it do it in first place ?

I peeked briefly at the sources.
Debian seems to resolve the problem by going for the max dependencies so autotools can pick up everything.
E.g. for gphoto2, lenny has a depends on both libexif and libreadline (see http://packages.debian.org/lenny/gphoto2)
So basically they go for the max config.

We could do the same. Then again for deeply embedded systems it might be desirable to allow more finegrained tuning.
(did I hear someone say USE flags ? )

Enjoy! Frans

#13 lore raw
Quoted 42 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 28 lines
Quoted 26 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 24 lines
Quoted 8 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

Quoted 9 lines

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

numbers means not so much when it comes down to this. I don't know if any distribution that even does that
so I take comfort in everyone being wrong.

Quoted 5 lines
Quoted 3 lines

I agree that numbers don't mean that much. Just wanted to give an indication of the possible effort needed.
Also no idea how debian does this.

Does it do it in first place ?

I peeked briefly at the sources.
Debian seems to resolve the problem by going for the max dependencies so autotools can pick up everything.
E.g. for gphoto2, lenny has a depends on both libexif and libreadline (see http://packages.debian.org/lenny/gphoto2)
So basically they go for the max config.

Quoted 4 lines

We could do the same. Then again for deeply embedded systems it might be desirable to allow more finegrained tuning.
(did I hear someone say USE flags ? )

which is good for things that dont care about size. Not so good for embedded systems I think.
and there are some packages I know which have catch-22 situation so adding max depends is
also not a wholesome solution.

Quoted 7 lines

Enjoy! Frans

_______________________________________________
Openembedded-devel mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel

#14 lore raw

2010/11/5 Khem Raj <[email protected]>:

Quoted 49 lines
Quoted 42 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 28 lines
Quoted 26 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 24 lines
Quoted 8 lines · Frans Meulenbroeks

2010/11/3 Khem Raj <[email protected]>:

Quoted 6 lines

yes a more detailed solution is better. you have to go through the configure of every
package and understand the behavior then act upon. As I said before its a tedious task

Quoted 9 lines

I'm well aware of this. That is also why in the infrastructure thread I suggested using automated testing to find misbehaving packages.

Btw there are 355 .inc files that contain the word autotools and 1899 .bb files. Of course there are lots of dups in there.
And maybe for the base recipes we can convince the yocto people that this is a serious QA issue (and they have some people employed to work on this if I understood properly).

numbers means not so much when it comes down to this. I don't know if any distribution that even does that
so I take comfort in everyone being wrong.

Quoted 5 lines
Quoted 3 lines

I agree that numbers don't mean that much. Just wanted to give an indication of the possible effort needed.
Also no idea how debian does this.

Does it do it in first place ?

I peeked briefly at the sources.
Debian seems to resolve the problem by going for the max dependencies so autotools can pick up everything.
E.g. for gphoto2, lenny has a depends on both libexif and libreadline (see http://packages.debian.org/lenny/gphoto2)
So basically they go for the max config.

Quoted 4 lines

We could do the same. Then again for deeply embedded systems it might be desirable to allow more finegrained tuning.
(did I hear someone say USE flags ? )

Quoted 6 lines

which is good for things that dont care about size. Not so good for embedded systems I think.
and there are some packages I know which have catch-22 situation so adding max depends is
also not a wholesome solution.

Catch22 situations seem mostly to appear because one packag exports both a lib and a (sample) program.
One example that I saw recently was libcdio and libcddb.

Probably the only other way of catch 22 situations are triple or quadruple staged builds.(build A, build B rebuild A maybe rebuild B)

Frans