[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
- To: alleg-developers@xxxxxxxxxx
- Subject: Re: [AD] malloc
- From: Chris <chris.kcat@xxxxxxxxxx>
- Date: Sun, 20 Nov 2005 21:05:14 -0800
- Domainkey-signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:from:to:subject:date:user-agent:references:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:message-id; b=D/3ygnEx2p6B9M0no/+JDcWuqILG955dUX+Cf4nhnxSkxsYt9jhBB/oc1DdvvNCyl5X/3jikcQKdTggGbS/pux6YR+dmDBeQvdk68Xk1Y2sms+yAcBEpp4pCdnUNSJPP8+uM8r26gTVWlhcvO+IgBfovKBUUBB5mZUSIWbKdszg=
On Sunday 20 November 2005 08:51 pm, aj wrote:
> then a bool IsAlignedPtr(); may be needed, before any SSE code.
>
> SSE instructions can still be used inside a
> if ( IsAlignedPtr() )
> SSE code
> else
> i386 code
And all that branching could probably negate any speed benefits SSE would
give. And how do you propose to write SSE code in C? AFAIK, assembly is being
dropped, since it's restrictive, compiler-specific, and in most cases not as
good as compiler-generated optimized code.
> for each line.
> SSE code can be used, even without using the -mSSE
> AFAIK -mSSE tells the compiler to generate its own SSE were
> appropriate, thats not going to prevent specific use of SSE.
If its anything like -mmmx, it'll cause the compiler to generate SSE code on
its own, and enable a C interface to using SSE.