[PATCH] Add --disable-decimal-float option to gcc-cross-intermediate/gcc-cross_4.4.2.bb

[PATCH] Add --disable-decimal-float option to gcc-cross-intermediate/gcc-cross_4.4.2.bb

[Guo Hongruan]

1. Add --disable-decimal-float option to gcc-cross-intermediate_4.4.2.bb when LIBC == uclibc and TARGET_ARCH == i[3|4|5|6]86, without it, the compilation will failed due to lack of fenv.h.
---
 recipes/gcc/gcc-cross-intermediate_4.4.2.bb |    7 +++++++
 recipes/gcc/gcc-cross_4.4.2.bb              |    7 ++++++-
 2 files changed, 13 insertions(+), 1 deletions(-)

diff --git a/recipes/gcc/gcc-cross-intermediate_4.4.2.bb b/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
index 3ca59aa..4bb1004 100644
--- a/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
+++ b/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
@@ -1,9 +1,16 @@
 require gcc-cross_${PV}.bb
 require gcc-cross-intermediate.inc
 
+PR = "${INC_PR}.2"
+
 DEPENDS += "gmp-native mpfr-native"
 
 EXTRA_OECONF += " --disable-libmudflap \
 		  --disable-libgomp \
 		  --disable-libssp \
 		"
+
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i386', '--disable-decimal-float', '',d)}"
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i486', '--disable-decimal-float', '',d)}"
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i586', '--disable-decimal-float', '',d)}"
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i686', '--disable-decimal-float', '',d)}"
diff --git a/recipes/gcc/gcc-cross_4.4.2.bb b/recipes/gcc/gcc-cross_4.4.2.bb
index 11eb29d..3e5b15a 100644
--- a/recipes/gcc/gcc-cross_4.4.2.bb
+++ b/recipes/gcc/gcc-cross_4.4.2.bb
@@ -2,10 +2,15 @@ require gcc-${PV}.inc
 require gcc-cross4.inc
 require gcc-configure-cross.inc
 require gcc-package-cross.inc
-PR = "${INC_PR}.1"
+PR = "${INC_PR}.2"
 
 SRC_URI_append_fail-fast = " file://zecke-no-host-includes.patch;patch=1 "
 
 EXTRA_OECONF += " --disable-libunwind-exceptions --with-mpfr=${STAGING_DIR_NATIVE}${prefix_native}"
 
 ARCH_FLAGS_FOR_TARGET += "-isystem${STAGING_DIR_TARGET}${target_includedir}"
+
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i386', '--disable-decimal-float', '',d)}"
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i486', '--disable-decimal-float', '',d)}"
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i586', '--disable-decimal-float', '',d)}"
+EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i686', '--disable-decimal-float', '',d)}"
-- 
1.5.4.3

Re: [PATCH] Add --disable-decimal-float option to gcc-cross-intermediate/gcc-cross_4.4.2.bb

[Khem Raj]

On (02/01/10 15:33), Guo Hongruan wrote:
> 1. Add --disable-decimal-float option to gcc-cross-intermediate_4.4.2.bb when LIBC == uclibc and TARGET_ARCH == i[3|4|5|6]86, without it, the compilation will failed due to lack of fenv.h.
> ---
>  recipes/gcc/gcc-cross-intermediate_4.4.2.bb |    7 +++++++
>  recipes/gcc/gcc-cross_4.4.2.bb              |    7 ++++++-
>  2 files changed, 13 insertions(+), 1 deletions(-)
> 
> diff --git a/recipes/gcc/gcc-cross-intermediate_4.4.2.bb b/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
> index 3ca59aa..4bb1004 100644
> --- a/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
> +++ b/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
> @@ -1,9 +1,16 @@
>  require gcc-cross_${PV}.bb
>  require gcc-cross-intermediate.inc
>  
> +PR = "${INC_PR}.2"
> +
>  DEPENDS += "gmp-native mpfr-native"
>  
>  EXTRA_OECONF += " --disable-libmudflap \
>  		  --disable-libgomp \
>  		  --disable-libssp \
>  		"
> +
> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i386', '--disable-decimal-float', '',d)}"
> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i486', '--disable-decimal-float', '',d)}"
> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i586', '--disable-decimal-float', '',d)}"
> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH', 'i686', '--disable-decimal-float', '',d)}"

It should be disabled for all uclibc based arch/distro combo's
and the change should go to gcc-<version>.inc files as in my original
patch.

Re: [PATCH] Add --disable-decimal-float option to gcc-cross-intermediate/gcc-cross_4.4.2.bb

[Guo Hongruan]

在 Tue, 05 Jan 2010 07:14:53 +0800，Khem Raj <raj.khem@gmail.com> 写道:

> On (02/01/10 15:33), Guo Hongruan wrote:
>> 1. Add --disable-decimal-float option to  
>> gcc-cross-intermediate_4.4.2.bb when LIBC == uclibc and TARGET_ARCH ==  
>> i[3|4|5|6]86, without it, the compilation will failed due to lack of  
>> fenv.h.
>> ---
>>  recipes/gcc/gcc-cross-intermediate_4.4.2.bb |    7 +++++++
>>  recipes/gcc/gcc-cross_4.4.2.bb              |    7 ++++++-
>>  2 files changed, 13 insertions(+), 1 deletions(-)
>>
>> diff --git a/recipes/gcc/gcc-cross-intermediate_4.4.2.bb  
>> b/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
>> index 3ca59aa..4bb1004 100644
>> --- a/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
>> +++ b/recipes/gcc/gcc-cross-intermediate_4.4.2.bb
>> @@ -1,9 +1,16 @@
>>  require gcc-cross_${PV}.bb
>>  require gcc-cross-intermediate.inc
>>
>> +PR = "${INC_PR}.2"
>> +
>>  DEPENDS += "gmp-native mpfr-native"
>>
>>  EXTRA_OECONF += " --disable-libmudflap \
>>  		  --disable-libgomp \
>>  		  --disable-libssp \
>>  		"
>> +
>> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH',  
>> 'i386', '--disable-decimal-float', '',d)}"
>> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH',  
>> 'i486', '--disable-decimal-float', '',d)}"
>> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH',  
>> 'i586', '--disable-decimal-float', '',d)}"
>> +EXTRA_OECONF_append_linux-uclibc += " ${@base_contains('TARGET_ARCH',  
>> 'i686', '--disable-decimal-float', '',d)}"
>
> It should be disabled for all uclibc based arch/distro combo's
> and the change should go to gcc-<version>.inc files as in my original
> patch.

Yes, I agree with you.

My proposal is to keep the effect of my patch as small as possible. For I  
have not validated other platforms, so I can not disable it this way. I  
will try it as soon as possible, if it did has the same problem on other  
platform, I will fix the patch.

By the way, on arm, mips platform, there is no such problem. I have not  
figured our why it does not occur on the platforms.

>
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel


-- 
Guo Hongruan, Embedded Linux Consultant
Skype: camelguo
Twitter: camelguo
http://www.gulessoft.com