Date: Thu, 16 Jun 2011 09:04:11 +0200 From: "O. Hartmann" <ohartman@zedat.fu-berlin.de> To: Doug Rabson <dfr@rabson.org> Cc: "freebsd-performance@freebsd.org" <freebsd-performance@freebsd.org>, FreeBSD Current <freebsd-current@freebsd.org> Subject: Re: FreeBSD 9.0-CUR/amd64 CLANG: howto use gcc __builtin_ia32? Message-ID: <4DF9AAEB.5090801@zedat.fu-berlin.de> In-Reply-To: <BANLkTikf-KqC4yYe8ab=vwee4z9U%2BT65iw@mail.gmail.com> References: <4DF4EC9F.9020604@zedat.fu-berlin.de> <BANLkTikf-KqC4yYe8ab=vwee4z9U%2BT65iw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On 06/15/11 20:36, Doug Rabson wrote: > You could try using the standard<xmmintrin.h> header - that has inline > functions which should cover all the SSE instructions. > > On 12 June 2011 17:43, Hartmann, O.<ohartman@zedat.fu-berlin.de> wrote: > >> I use some numerical code utilizing the SIMD units of modern X86 >> architectures. Code compiles well using gcc/gcc46, >> but clang does not know about the __builtin_ia32_xxxxx() statements. How to >> treat those in clang and how to make >> C code compiling with clang utilizing those __builtin_ia32 statements? >> >> Thanks, >> Oliver Thanks for this hint. I tried, and its a negative. I'm quite new to the usage of builtins and I oriented myself using gcc style __builtins. Especially __builtin_ia32_mulpd(()), __builtin_ia32_addpd(()). The aim is to add, mul, div and sub two 64 Bit doubles residing in a 128 Bit xmm register; by doing this, a struct contains two unions, each 128 bit, to simulate a 4d vector, so I' capable of adding, multiplying etc. vectors using the SIMD units. It seems that there is a more 'sophisticated' way than simply apllying the __builtin assembly equivalences. Oliver
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4DF9AAEB.5090801>