Skip to contentSkip to conversation

kernel bbclass: split do_compile into do_compile and do_compile_modules

PATCH In review Koen Kooi 4 participants 14 messages Oct 11, 2010 – Oct 12, 2010

lore
Share & tools
lore

Koen Kooi

author
#0 lore raw

PATCH v3 1/3 kernel bbclass: split do_compile into do_compile and do_compile_modules

This allows recipes to insert a custom task in between building *Image and modules

---
 classes/kernel.bbclass |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)
classes/kernel.bbclass +5 -0
index cf6b8cd..83e3f44 100644
@@ -89,12 +89,17 @@ kernel_do_compile() {
 		oe_runmake dep CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	fi
 	oe_runmake ${KERNEL_IMAGETYPE} CC="${KERNEL_CC}" LD="${KERNEL_LD}"
+}
+
+do_compile_kernelmodules() {
+	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
 	if (grep -q -i -e '^CONFIG_MODULES=y$' .config); then
 		oe_runmake modules  CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	else
 		oenote "no modules to compile"
 	fi
 }
+addtask compile_kernelmodules after do_compile before do_install
 
 kernel_do_install() {
 	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
Signature · 1 line
1.6.6.1
author
#1 lore raw

PATCH v3 2/3 kernel bbclass: make 'kernel-modules' meta-package use RRECOMMENDS to please insane.bbclass

---
 classes/kernel.bbclass |    8 ++++----
 1 files changed, 4 insertions(+), 4 deletions(-)
classes/kernel.bbclass +4 -4
index 83e3f44..7039d27 100644
@@ -489,12 +489,12 @@ python populate_packages_prepend () {
 		for i in l:
 			pkg = module_pattern % legitimize_package_name(re.match(module_regex, os.path.basename(i)).group(1))
 			blacklist.append(pkg)
-	metapkg_rdepends = []
+	metapkg_rrecommends = []
 	packages = bb.data.getVar('PACKAGES', d, 1).split()
 	for pkg in packages[1:]:
-		if not pkg in blacklist and not pkg in metapkg_rdepends and not any(pkg.endswith(post) for post in depchains):
-			metapkg_rdepends.append(pkg)
-	bb.data.setVar('RDEPENDS_' + metapkg, ' '.join(metapkg_rdepends), d)
+		if not pkg in blacklist and not pkg in metapkg_rrecommends and not any(pkg.endswith(post) for post in depchains):
+			metapkg_rrecommends.append(pkg)
+	bb.data.setVar('RRECOMMENDS_' + metapkg, ' '.join(metapkg_rrecommends), d)
 	bb.data.setVar('DESCRIPTION_' + metapkg, 'Kernel modules meta package', d)
 	packages.append(metapkg)
 	bb.data.setVar('PACKAGES', ' '.join(packages), d)
Signature · 1 line
1.6.6.1
Acked-by on v3 2/3
#2 lore raw

Re: [PATCHv3 2/3] kernel bbclass: make 'kernel-modules' meta-package use RRECOMMENDS to please insane.bbclass

2010/10/11 Koen Kooi <[email protected]>:

Quoted 20 lines

---
 classes/kernel.bbclass |    8 ++++----

 1 files changed, 4 insertions(+), 4 deletions(-)
classes/kernel.bbclass +0 -0
index 83e3f44..7039d27 100644
@@ -489,12 +489,12 @@ python populate_packages_prepend () {

               for i in l:
                       pkg = module_pattern % legitimize_package_name(re.match(module_regex, os.path.basename(i)).group(1)) blacklist.append(pkg)
-       metapkg_rdepends = []
+       metapkg_rrecommends = []
       packages = bb.data.getVar('PACKAGES', d, 1).split()
       for pkg in packages[1:]:
-               if not pkg in blacklist and not pkg in metapkg_rdepends and not any(pkg.endswith(post) for post in depchains):
-                       metapkg_rdepends.append(pkg)
-       bb.data.setVar('RDEPENDS_' + metapkg, ' '.join(metapkg_rdepends), d)

Quoted 8 lines

+               if not pkg in blacklist and not pkg in metapkg_rrecommends and not any(pkg.endswith(post) for post in depchains):
+                       metapkg_rrecommends.append(pkg)
+       bb.data.setVar('RRECOMMENDS_' + metapkg, ' '.join(metapkg_rrecommends), d) bb.data.setVar('DESCRIPTION_' + metapkg, 'Kernel modules meta package', d)
       packages.append(metapkg)
       bb.data.setVar('PACKAGES', ' '.join(packages), d)
--
1.6.6.1

author
#3 lore raw

PATCH v3 3/3 multi-kernel: adapt to kernel_do_compile_kernelmodules change in kernel.bbclass

In reply to Koen Kooi
---
 recipes/linux/multi-kernel.inc |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)
recipes/linux/multi-kernel.inc +1 -0
index 70cfd87..0523a33 100644
@@ -41,6 +41,7 @@ do_compileconfigs () {
         # Build and Install this alternative kernel
         do_configure
         kernel_do_compile
+        do_compile_kernelmodules
         kernel_do_install
 
         # Drop the resulting images in the deploy dir
Signature · 1 line
1.6.6.1
Acked-by on v3 3/3
#4 lore raw

Re: [PATCHv3 3/3] multi-kernel: adapt to kernel_do_compile_kernelmodules change in kernel.bbclass

2010/10/11 Koen Kooi <[email protected]>:

Quoted 15 lines

---
 recipes/linux/multi-kernel.inc |    1 +

 1 files changed, 1 insertions(+), 0 deletions(-)
recipes/linux/multi-kernel.inc +0 -0
index 70cfd87..0523a33 100644
@@ -41,6 +41,7 @@ do_compileconfigs () {
         # Build and Install this alternative kernel

        do_configure
        kernel_do_compile
+        do_compile_kernelmodules
        kernel_do_install

Quoted 2 lines

        # Drop the resulting images in the deploy dir --

This is where I found it would probably better be called kernel_do_compile_modules or so.

Still:

#5 lore raw
In reply to Koen Kooi

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

ping

Quoted 19 lines · Koen Kooi

This allows recipes to insert a custom task in between building *Image and modules

---
 classes/kernel.bbclass |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)
classes/kernel.bbclass +4 -0
index cf6b8cd..83e3f44 100644
@@ -89,12 +89,17 @@ kernel_do_compile() {
 		oe_runmake dep CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	fi
 	oe_runmake ${KERNEL_IMAGETYPE} CC="${KERNEL_CC}" LD="${KERNEL_LD}"
+}
+
+do_compile_kernelmodules() {
+	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

classes/kernel.bbclass kernel_do_compile() { · line 105

Quoted 10 lines
classes/kernel.bbclass +1 -0
 	if (grep -q -i -e '^CONFIG_MODULES=y$' .config); then
 		oe_runmake modules  CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	else
 		oenote "no modules to compile"
 	fi
 }
+addtask compile_kernelmodules after do_compile before do_install
 
 kernel_do_install() {
 	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

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

iD8DBQFMtBfOMkyGM64RGpERAonPAKCPzmCFS67arfwLKKL8Eb7QENuKOgCgmx++ 7TUh52+7x1OV5xjF7Pp/oFk=
=yf9t
-----END PGP SIGNATURE-----

#6 lore raw

2010/10/12 Koen Kooi <[email protected]>:

Quoted 4 lines

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

ping

Haven't had time to test this yet (and won't have tonight as I have to supervise a lab). It is on the to-do list.

[...]

Quoted 4 lines

classes/kernel.bbclass kernel_do_compile() { · line 95

Quoted 4 lines
classes/kernel.bbclass +4 -0
+}
+
+do_compile_kernelmodules() {
+     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

Is it needed/required? I'm a little bit concerned by this. (actually this was the only worry I had after reading the patch, don't want to loose additional CFLAGS settings here (but must say I did not dive into it, so not sure if things are arranged differently; that is also why I wanted to do a test build).

Frans

#7 lore raw

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

Quoted 8 lines · Frans Meulenbroeks

2010/10/12 Koen Kooi <[email protected]>:

Quoted 4 lines

classes/kernel.bbclass kernel_do_compile() { · line 95

Quoted 4 lines
classes/kernel.bbclass +4 -0
+}
+
+do_compile_kernelmodules() {
+     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

Is it needed/required?

It is, and it's in kernel.bbclass already:

fe938bbe (Michael Lauer 2004-12-14) kernel_do_compile() {
0448ec0d (Richard Purdie 2006-11-20) unset CFLAGS CPPFLAGS CXXFLAGS
LDFLAGS MACHINE
fe938bbe (Michael Lauer 2004-12-14) oe_runmake
include/linux/version.h CC="${KERNEL_CC}" LD="${KERNEL_LD}"

Quoted 3 lines

I'm a little bit concerned by this.
(actually this was the only worry I had after reading the patch, don't want to loose additional CFLAGS settings here

You actually want to do that, since the kernel is bare-metal, not userspace. -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFMtDm/MkyGM64RGpERAuFwAKCBzAXJwVvcQ6oh2DSISA/7Hhh8zACgp9M9 muqzjUd8tGg/RCZoEKmsU/4=
=1SOi
-----END PGP SIGNATURE-----

#8 lore raw

2010/10/12 Koen Kooi <[email protected]>:

Quoted 36 lines

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

Quoted 8 lines · Frans Meulenbroeks

2010/10/12 Koen Kooi <[email protected]>:

Quoted 4 lines

classes/kernel.bbclass kernel_do_compile() { · line 95

Quoted 4 lines
classes/kernel.bbclass +4 -0
+}
+
+do_compile_kernelmodules() {
+     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

Is it needed/required?

It is, and it's in kernel.bbclass already:

fe938bbe (Michael Lauer 2004-12-14) kernel_do_compile() { 0448ec0d (Richard Purdie 2006-11-20)     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
fe938bbe (Michael Lauer 2004-12-14)     oe_runmake include/linux/version.h CC="${KERNEL_CC}" LD="${KERNEL_LD}"

Quoted 3 lines

I'm a little bit concerned by this.
(actually this was the only worry I had after reading the patch, don't want to loose additional CFLAGS settings here

You actually want to do that, since the kernel is bare-metal, not userspace. -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFMtDm/MkyGM64RGpERAuFwAKCBzAXJwVvcQ6oh2DSISA/7Hhh8zACgp9M9 muqzjUd8tGg/RCZoEKmsU/4=
=1SOi
-----END PGP SIGNATURE-----

Quoted 5 lines

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

Thanks for the answer.

Did a quick peek on CFLAGS in the recipes, saw there is a CFLAGS_KERNEL. That one is only used on one place:
linux-wrt_2.4.30.bb:export CFLAGS_KERNEL="-fno-delayed-branch "

Haven't tested this, but my understanding on how things work is good enough and the change simple enough for to give an ack

#9 lore raw

2010/10/12 Frans Meulenbroeks <[email protected]>:

Quoted 45 lines

2010/10/12 Koen Kooi <[email protected]>:

Quoted 36 lines

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

Quoted 8 lines · Frans Meulenbroeks

2010/10/12 Koen Kooi <[email protected]>:

Quoted 4 lines

classes/kernel.bbclass kernel_do_compile() { · line 95

Quoted 4 lines
classes/kernel.bbclass +4 -0
+}
+
+do_compile_kernelmodules() {
+     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

Is it needed/required?

It is, and it's in kernel.bbclass already:

fe938bbe (Michael Lauer 2004-12-14) kernel_do_compile() { 0448ec0d (Richard Purdie 2006-11-20)     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
fe938bbe (Michael Lauer 2004-12-14)     oe_runmake include/linux/version.h CC="${KERNEL_CC}" LD="${KERNEL_LD}"

Quoted 3 lines

I'm a little bit concerned by this.
(actually this was the only worry I had after reading the patch, don't want to loose additional CFLAGS settings here

You actually want to do that, since the kernel is bare-metal, not userspace. -----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (Darwin)

iD8DBQFMtDm/MkyGM64RGpERAuFwAKCBzAXJwVvcQ6oh2DSISA/7Hhh8zACgp9M9 muqzjUd8tGg/RCZoEKmsU/4=
=1SOi
-----END PGP SIGNATURE-----

Quoted 5 lines

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

Thanks for the answer.

Quoted 8 lines

Did a quick peek on CFLAGS in the recipes, saw there is a CFLAGS_KERNEL. That one is only used on one place:
linux-wrt_2.4.30.bb:export CFLAGS_KERNEL="-fno-delayed-branch "

Haven't tested this, but my understanding on how things work is good enough and the change simple enough for to give an ack

Actually on 2nd thought one tiny remark left:
after seeing patch 2:
what about renaming:
do_compile_kernelmodules
to something with a kernel prefix
e.g. kernel_do_compile_modules.
That makes the naming scheme somewhat more consistent.

#10 lore raw

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

Quoted 40 lines · Frans Meulenbroeks

2010/10/12 Frans Meulenbroeks <[email protected]>:

Quoted 1 line

2010/10/12 Koen Kooi <[email protected]>:

Quoted 8 lines · Frans Meulenbroeks
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

2010/10/12 Koen Kooi <[email protected]>:

Quoted 4 lines

classes/kernel.bbclass kernel_do_compile() { · line 95

Quoted 4 lines
classes/kernel.bbclass +4 -0
+}
+
+do_compile_kernelmodules() {
+     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

Is it needed/required?

It is, and it's in kernel.bbclass already:

fe938bbe (Michael Lauer 2004-12-14) kernel_do_compile() {
0448ec0d (Richard Purdie 2006-11-20) unset CFLAGS CPPFLAGS CXXFLAGS
LDFLAGS MACHINE
fe938bbe (Michael Lauer 2004-12-14) oe_runmake
include/linux/version.h CC="${KERNEL_CC}" LD="${KERNEL_LD}"

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

I'm a little bit concerned by this.
(actually this was the only worry I had after reading the patch, don't want to loose additional CFLAGS settings here

You actually want to do that, since the kernel is bare-metal, not userspace.

Quoted 5 lines
Quoted 1 line

Thanks for the answer.

Quoted 8 lines

Did a quick peek on CFLAGS in the recipes, saw there is a CFLAGS_KERNEL. That one is only used on one place:
linux-wrt_2.4.30.bb:export CFLAGS_KERNEL="-fno-delayed-branch "

Haven't tested this, but my understanding on how things work is good enough and the change simple enough for to give an ack

Quoted 7 lines

Actually on 2nd thought one tiny remark left:
after seeing patch 2:
what about renaming:
do_compile_kernelmodules
to something with a kernel prefix
e.g. kernel_do_compile_modules.
That makes the naming scheme somewhat more consistent.

Namespacing the class is only used for methods overridden from base.bbclass. Again, look at the rest of kernel.bbclass methods like uboot-mkimage, menuconfig and friends don't use that namespace.

And it doesn't actually work since _modules is a local override.

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

iD8DBQFMtFxYMkyGM64RGpERArFyAKCr5HipFKmy8xsYIoF/U0NIXFfDlwCfferR yhTvF+MOCupbbMMstKVvkmk=
=6zMs
-----END PGP SIGNATURE-----

#11 lore raw

2010/10/12 Koen Kooi <[email protected]>:

Quoted 52 lines

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

Quoted 40 lines · Frans Meulenbroeks

2010/10/12 Frans Meulenbroeks <[email protected]>:

Quoted 1 line

2010/10/12 Koen Kooi <[email protected]>:

Quoted 8 lines · Frans Meulenbroeks
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

2010/10/12 Koen Kooi <[email protected]>:

Quoted 4 lines

Review

Quoted 4 lines
(quoted diff) +4 -0
+}
+
+do_compile_kernelmodules() {
+     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

Is it needed/required?

It is, and it's in kernel.bbclass already:

fe938bbe (Michael Lauer 2004-12-14) kernel_do_compile() { 0448ec0d (Richard Purdie 2006-11-20)     unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
fe938bbe (Michael Lauer 2004-12-14)     oe_runmake include/linux/version.h CC="${KERNEL_CC}" LD="${KERNEL_LD}"

Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines

I'm a little bit concerned by this.
(actually this was the only worry I had after reading the patch, don't want to loose additional CFLAGS settings here

You actually want to do that, since the kernel is bare-metal, not userspace.

Quoted 5 lines
Quoted 1 line

Thanks for the answer.

Quoted 8 lines

Did a quick peek on CFLAGS in the recipes, saw there is a CFLAGS_KERNEL. That one is only used on one place:
linux-wrt_2.4.30.bb:export CFLAGS_KERNEL="-fno-delayed-branch "

Haven't tested this, but my understanding on how things work is good enough and the change simple enough for to give an ack

Quoted 7 lines

Actually on 2nd thought one tiny remark left:
after seeing patch 2:
what about renaming:
do_compile_kernelmodules
to something with a kernel prefix
e.g. kernel_do_compile_modules.
That makes the naming scheme somewhat more consistent.

Quoted 3 lines

Namespacing the class is only used for methods overridden from base.bbclass. Again, look at the rest of kernel.bbclass methods like uboot-mkimage, menuconfig and friends don't use that namespace.

Well I don't care too much about it, this just bubbled up because this:

Quoted 3 lines

kernel_do_compile
+ do_compile_kernelmodules
kernel_do_install

looks a little bit better/more consistent; i did not evn consider overriding or so, just peeked at the aesthetics.
as I wrote before it is a nit (and absolutely no deal-breaker afaic)

Frans

Quoted 18 lines

And it doesn't actually work since _modules is a local override.

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

iD8DBQFMtFxYMkyGM64RGpERArFyAKCr5HipFKmy8xsYIoF/U0NIXFfDlwCfferR yhTvF+MOCupbbMMstKVvkmk=
=6zMs
-----END PGP SIGNATURE-----

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

#12 lore raw
In reply to Koen Kooi

Am Dienstag, den 12.10.2010, 10:09 +0200 schrieb Koen Kooi:

Quoted 30 lines

ping

Quoted 27 lines · Koen Kooi

This allows recipes to insert a custom task in between building *Image and modules

---
 classes/kernel.bbclass |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)
classes/kernel.bbclass +5 -0
index cf6b8cd..83e3f44 100644
@@ -89,12 +89,17 @@ kernel_do_compile() {
 		oe_runmake dep CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	fi
 	oe_runmake ${KERNEL_IMAGETYPE} CC="${KERNEL_CC}" LD="${KERNEL_LD}"
+}
+
+do_compile_kernelmodules() {
+	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
 	if (grep -q -i -e '^CONFIG_MODULES=y$' .config); then
 		oe_runmake modules  CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	else
 		oenote "no modules to compile"
 	fi
 }
+addtask compile_kernelmodules after do_compile before do_install
 
Quoted 2 lines

classes/kernel.bbclass kernel_do_compile() { · line 105

Quoted 2 lines
classes/kernel.bbclass +0 -0
 kernel_do_install() {
 	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

I am using MACHINE = "beagleboard".

I did

	$ bitbake -c clean virtual/kernel
	$ bitbake virtual/kernel

with your patches applied for minimal-{libc,eglibc} and I did not see any regressions.

The build for angstrom-2010.x is still running. angstrom-2008.1 hangs right now, though. Could you test that again, please. I will take a look in a few more hours, if the build succeeded.

	NOTE: Running task 892 of 911 (ID: 13, /home/paul/oe/openembedded/recipes/linux)
	NOTE: package linux-omap-psp-2.6.32-r90+gitra6bad4464f985fdd3bed72e1b82dcbfc004d…
	NOTE: package linux-omap-psp-2.6.32-r90+gitra6bad4464f985fdd3bed72e1b82dcbfc004d…

Thanks,

Paul

Tested-by
#13 lore raw

Am Dienstag, den 12.10.2010, 16:26 +0200 schrieb Paul Menzel:

Quoted 44 lines

Am Dienstag, den 12.10.2010, 10:09 +0200 schrieb Koen Kooi:

Quoted 30 lines

ping

Quoted 27 lines · Koen Kooi

This allows recipes to insert a custom task in between building *Image and modules

---
 classes/kernel.bbclass |    5 +++++
 1 files changed, 5 insertions(+), 0 deletions(-)
classes/kernel.bbclass +5 -0
index cf6b8cd..83e3f44 100644
@@ -89,12 +89,17 @@ kernel_do_compile() {
 		oe_runmake dep CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	fi
 	oe_runmake ${KERNEL_IMAGETYPE} CC="${KERNEL_CC}" LD="${KERNEL_LD}"
+}
+
+do_compile_kernelmodules() {
+	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE
 	if (grep -q -i -e '^CONFIG_MODULES=y$' .config); then
 		oe_runmake modules  CC="${KERNEL_CC}" LD="${KERNEL_LD}"
 	else
 		oenote "no modules to compile"
 	fi
 }
+addtask compile_kernelmodules after do_compile before do_install
 
Quoted 2 lines

classes/kernel.bbclass kernel_do_compile() { · line 105

Quoted 2 lines
classes/kernel.bbclass +0 -0
 kernel_do_install() {
 	unset CFLAGS CPPFLAGS CXXFLAGS LDFLAGS MACHINE

I am using MACHINE = "beagleboard".

I did

	$ bitbake -c clean virtual/kernel
	$ bitbake virtual/kernel

with your patches applied for minimal-{libc,eglibc} and I did not see any regressions.

Quoted 7 lines

The build for angstrom-2010.x is still running. angstrom-2008.1 hangs right now, though. Could you test that again, please. I will take a look in a few more hours, if the build succeeded.

	NOTE: Running task 892 of 911 (ID: 13, /home/paul/oe/openembedded/recipes/linux)
	NOTE: package linux-omap-psp-2.6.32-r90+gitra6bad4464f985fdd3bed72e1b82dcbfc004d…
	NOTE: package linux-omap-psp-2.6.32-r90+gitra6bad4464f985fdd3bed72e1b82dcbfc004d…

It finished just fine.

But you already committed this in 615876 [1] and onwards.

Thanks,

Paul

[1] http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=615876fe218dc3feb4a3df9e6546a7b1a6376800