[OE-core] [PATCH] lsof: define linux C library type when using eglibc

[Liu Ting-B28495]

-----Original Message-----
From: Liu Ting-B28495 
Sent: Friday, July 13, 2012 10:47 AM
To: 'Saul Wold'; Patches and discussions about the oe-core layer
Cc: McClintock Matthew-B29882
Subject: RE: [OE-core] [PATCH] lsof: define linux C library type when using eglibc

-----Original Message-----
From: Saul Wold [mailto:sgw at linux.intel.com]
Sent: Friday, July 13, 2012 3:31 AM
To: Patches and discussions about the oe-core layer
Cc: Liu Ting-B28495; McClintock Matthew-B29882
Subject: Re: [OE-core] [PATCH] lsof: define linux C library type when using eglibc

On 07/11/2012 11:17 PM, b28495 at freescale.com wrote:
> From: Ting Liu <b28495 at freescale.com>
>
> lsof tries to compile a temp c source file and execute the binary to 
> determine linux C library type (file Configure, line 2689-2717).
> It is inpracticable for cross-compilation and may have build issue on 
> some distros since it depends on host settings.
>
> Fix below error when building for 64bit target on 64bit host:
> [...]
> | dsock.c:481:44: error: 'TCP_LISTEN' undeclared (first use in this
> | function)
> | dsock.c:482:45: error: 'TCP_CLOSING' undeclared (first use in this
> | function)
> [...]
> | make: *** [dsock.o] Error 1
>
> The actual issue exists in do_configure:
> [...]
> Testing C library type with cc ... done Cannot determine C library 
> type; assuming it is not glibc.
>
> Which is in turn caused by missing 'gnu/stubs-32.h" when compiling the 
> temp c source file on host:
> [...]
> fatal error: gnu/stubs-32.h: No such file or directory compilation terminated.
>
> file gnu/stubs-32.h is provided by 32bit glibc.
>
> Signed-off-by: Ting Liu <b28495 at freescale.com>
> ---
>   meta/recipes-extended/lsof/lsof_4.85.bb |    7 ++++++-
>   1 files changed, 6 insertions(+), 1 deletions(-)
>
> diff --git a/meta/recipes-extended/lsof/lsof_4.85.bb
> b/meta/recipes-extended/lsof/lsof_4.85.bb
> index 5823725..d28d2cb 100644
> --- a/meta/recipes-extended/lsof/lsof_4.85.bb
> +++ b/meta/recipes-extended/lsof/lsof_4.85.bb
> @@ -3,7 +3,7 @@ DESCRIPTION = "Lsof is a Unix-specific diagnostic tool. \
>   Its name stands for LiSt Open Files, and it does just that."
>   SECTION = "devel"
>   LICENSE = "BSD"
> -PR = "r1"
> +PR = "r2"
>
>   SRC_URI = "ftp://lsof.itap.purdue.edu/pub/tools/unix/lsof/lsof_${PV}.tar.bz2"
>
> @@ -31,6 +31,11 @@ export LSOF_INCLUDE = "${STAGING_INCDIR}"
>   do_configure () {
>   	export LSOF_AR="${AR} cr"
>   	export LSOF_RANLIB="${RANLIB}"
> +        if [ "x${EGLIBCVERSION}" != "x" ];then
> +                LINUX_CLIB=${EGLIBCVERSION/\./}
> +                LINUX_CLIB="-DGLIBCV=${LINUX_CLIB}"
Does this really do what you intent it to?

why not set the EGLIBCVERSION directly on the second line, what's the purpose of the /\./?  It renders the version empty in my tests.

Sau!

[Liu Ting-B28495] LINUX_CLIB is set to "-DGLIBCV=213" when lsof can figure out the right Linux c library type. String 213 corresponds to the eglibc version 2.13. the purpose of /\./ is to get rid of '.'.
Here is my test:
-bash-3.2$ EGLIBCVERSION="2.13"
-bash-3.2$ echo ${EGLIBCVERSION/\./}
213
-bash-3.2$

[Liu Ting-B28495] ok, I find why it renders the version empty:
In temp/run.configure:
....
export BUILD_STRIP="strip"
do_configure() {
        if [ "x2.13" != "x" ];then
                LINUX_CLIB=${EGLIBCVERSION/\./}
                LINUX_CLIB="-DGLIBCV=${LINUX_CLIB}"
                export LINUX_CLIB
        fi
        yes | ./Configure linux

}

Variable EGLIBCVERSION is expanded instead of exported.