From: Rene Rivera (grafik666_at_[hidden])
Date: 2002-06-12 12:08:04
On 2002-06-11 at 04:59 PM, gohmandj_at_[hidden] (Dan Gohman) wrote:
>On Tue, Jun 11, 2002 at 04:43:35PM -0500, Rene Rivera wrote:
>> OK, I get it now. In that case I propose this instead...
>> feature address-model : default 16 32 64 ;
>> This is a more generic term and can later expand to specify other
>> models on other CPUs.
>This sounds good to me.
Cool. I've just updated "features.jam" to include the "address-model"
feature. And also split the "architecture" into "architecture" and
"instruction-set". And added entries for MIPS to the architecture and
instruction-set, using the names that GCC uses/gives.
I also updated the "gcc-tools.jam" to use and generate the flags for MIPS as
>By the way, is there a standard way for a toolset to say that it doesn't
>support a particular option? For example, MIPSpro doesn't support a 16-bit
Not really... But take a look at what I did for gcc... it's a kludge, but it
-- grafik - Don't Assume Anything
-- rrivera_at_[hidden] - grafik_at_[hidden]
-- 102708583_at_icq - Grafik666_at_AIM - Grafik_at_[hidden]
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk