[Buildroot] [git commit] acpid: unavailable for external uClibc toolchains

[Buildroot] [git commit] acpid: unavailable for external uClibc toolchains

[Peter Korsgaard]

commit: http://git.buildroot.net/buildroot/commit/?id=ec6aaa4706d6826aca4e48ea65bbebbd893f76d6
branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master

External non-patched uClibc toolchains lack the MSG_CMSG_CLOEXEC
definition, so does internal uClibc <0.9.33, so exclude it for those
scenarios. Fixes:
http://autobuild.buildroot.net/results/6d4/6d471942788fa05f324649ab15e6ca382e46df9c/

Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
---
 package/acpid/Config.in |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/package/acpid/Config.in b/package/acpid/Config.in
index 5aec7f1..205a6ae 100644
--- a/package/acpid/Config.in
+++ b/package/acpid/Config.in
@@ -1,6 +1,9 @@
 config BR2_PACKAGE_ACPID
 	bool "acpid"
 	depends on BR2_x86_64 || BR2_i386
+	# Usually lacks MSG_CMSG_CLOEXEC
+	depends on !BR2_TOOLCHAIN_EXTERNAL_UCLIBC && \
+		!BR2_UCLIBC_VERSION_0_9_32
 	help
 	  Advanced Configuration and Power Interface event daemon.
 

[Buildroot] [git commit] acpid: unavailable for external uClibc toolchains

[Thomas Petazzoni]

Dear Peter Korsgaard,

On Thu, 24 Apr 2014 22:35:33 +0200, Peter Korsgaard wrote:
> commit: http://git.buildroot.net/buildroot/commit/?id=ec6aaa4706d6826aca4e48ea65bbebbd893f76d6
> branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
> 
> External non-patched uClibc toolchains lack the MSG_CMSG_CLOEXEC
> definition, so does internal uClibc <0.9.33, so exclude it for those
> scenarios. Fixes:
> http://autobuild.buildroot.net/results/6d4/6d471942788fa05f324649ab15e6ca382e46df9c/

Huh? This build failure is due to a difference between a patched
version of uClibc (available in the internal toolchain backend) and an
external uClibc toolchain. Haven't we said that we shouldn't really on
our uClibc internal toolchain having feature patches, and instead patch
the packages?

I'm sorry, but I disagree with the solution chosen here.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

[Buildroot] [git commit] acpid: unavailable for external uClibc toolchains

[Peter Korsgaard]

>>>>> "Thomas" == Thomas Petazzoni <thomas.petazzoni@free-electrons.com> writes:

Hi,

 > Dear Peter Korsgaard,
 > On Thu, 24 Apr 2014 22:35:33 +0200, Peter Korsgaard wrote:
 >> commit: http://git.buildroot.net/buildroot/commit/?id=ec6aaa4706d6826aca4e48ea65bbebbd893f76d6
 >> branch: http://git.buildroot.net/buildroot/commit/?id=refs/heads/master
 >> 
 >> External non-patched uClibc toolchains lack the MSG_CMSG_CLOEXEC
 >> definition, so does internal uClibc <0.9.33, so exclude it for those
 >> scenarios. Fixes:
 >> http://autobuild.buildroot.net/results/6d4/6d471942788fa05f324649ab15e6ca382e46df9c/

 > Huh? This build failure is due to a difference between a patched
 > version of uClibc (available in the internal toolchain backend) and an
 > external uClibc toolchain. Haven't we said that we shouldn't really on
 > our uClibc internal toolchain having feature patches, and instead patch
 > the packages?

 > I'm sorry, but I disagree with the solution chosen here.

So what do you suggest instead?

depends on BR2_TOOLCHAIN_USES_GLIBC || BR2_UCLIBC_VERSION_SNAPSHOT

?

-- 
Bye, Peter Korsgaard