[Fwd: Re: [PATCH 3 of 3] Add IBM s390x architecture and sample]

[Harold Grovesteen]

-------- Original Message --------
Subject: 	Re: [PATCH 3 of 3] Add IBM s390x architecture and sample
Date: 	Wed, 11 Nov 2009 16:36:17 -0600
From: 	Harold Grovesteen <h.grovsteen@tx.rr.com>
To: 	Yann E. MORIN <yann.morin.1998@anciens.enib.fr>
References: 	<b39275d5a408cfaae1ec.1257961230@dev.site> 
<200911112024.36968.yann.morin.1998@anciens.enib.fr>





Yann E. MORIN wrote:

>Harold,
>All,
>
>All-in-all, the same comments as for the previous patch still apply,
>plus the ones below.
>
>On Wednesday 11 November 2009 18:40:30 Harold Grovesteen wrote:
>  
>
>>+# Target options
>>+#
>>+CT_ARCH="s390x"
>>+CT_ARCH_64=y
>>+# CT_ARCH_SUPPORTS_BOTH_MMU is not set
>>+# CT_ARCH_SUPPORTS_BOTH_ENDIAN is not set
>>+# CT_ARCH_SUPPORT_ARCH is not set
>>+# CT_ARCH_SUPPORT_ABI is not set
>>+# CT_ARCH_SUPPORT_CPU is not set
>>+# CT_ARCH_SUPPORT_TUNE is not set
>>+CT_ARCH_SUPPORT_FPU=y
>>+CT_ARCH_DEFAULT_HAS_MMU=y
>>+CT_ARCH_DEFAULT_BE=y
>>+# CT_ARCH_DEFAULT_LE is not set
>>+CT_ARCH_FPU=""
>>+CT_ARCH_FLOAT_HW=y
>>+# CT_ARCH_FLOAT_SW is not set
>>+CT_TARGET_CFLAGS="-O"
>>    
>>
>
>Hmmm. I don't know. -O is overriden by glibc/eglibc to -O2, and
>I'm not sure gcc makes (a proper) use of CT_TARGET_CFLAGS.
>
This might have been a hold over from the crosstool run that used this.  
I can again remove and see how it goes.

>Also applies to your previous patch.
>
>  
>
>>+CT_BINUTILS_EXTRA_CONFIG="--enable-64-bit-bfd"
>>    
>>
>
>This should be done in the binutils build script, when the target is
>64-bit. Prepare a separate patch.
>
It is easy to remove the option.  For what should I create a separate 
patch? a new patch for the sample?

>Regards,
>Yann E. MORIN.
>
>  
>


--
For unsubscribe information see http://sourceware.org/lists.html#faq