[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [MiNT] Libcurl-ssl / RPM package
- To: mint <mint@lists.fishpool.fi>
- Subject: Re: [MiNT] Libcurl-ssl / RPM package
- From: Vincent Rivière <vincent.riviere@freesbee.fr>
- Date: Mon, 14 Jun 2010 17:13:50 +0200
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=gamma; h=domainkey-signature:received:received:sender:message-id:date:from :user-agent:mime-version:to:subject:references:in-reply-to :content-type:content-transfer-encoding; bh=q0BP1OxHPafGLMM8aXFRD+hVkgHztUPJ91N35G6Q6PE=; b=h0nEbftsUFTswGsvd97CDJ5vlvAz01pGttz+8HInKjxumOk2ewFI6DjjJWFpglljkn 5+p46Leuqf1s4NRcvqtDFp6UOc4UKeCfDbyMKBiKnIsh0O5tm4zdShpEw+o95IFMZAPQ kNQbHdi6Jg2uFLV1gl65FhC5nGbElwFx3IwOE=
- Domainkey-signature: a=rsa-sha1; c=nofws; d=gmail.com; s=gamma; h=sender:message-id:date:from:user-agent:mime-version:to:subject :references:in-reply-to:content-type:content-transfer-encoding; b=SoOeGIKYIq+ShwWEVe4f9R3cH6t4Uy9DslEr7Q0iirwKrwSN6MKOprlIxcZWTmjgPU 0kt9ejmdzK1Wgamv38FXallec3swzeMg5rIb8XQ+PytnWE38qxNG47RMYtTNq2VzNGW8 E9Mzcz2ucx51ixDB0oNEDerE+u5IkOQDquFCs=
- In-reply-to: <845495def3bc16e5a0fcba375627015b-EhVcX1lFRQVaRwYcDTpQCEFddQZLVF5dQUBFBDBTXF5bVkYOWkF0A1w6XF1YR0AAWF9bSA==-webmailer2@server08.webmailer.hosteurope.de>
- List-help: <mailto:ecartis@lists.fishpool.fi?Subject=help>
- List-id: <mint.lists.fishpool.fi>
- List-owner: <mailto:tjhukkan@fishpool.fi>
- List-post: <mailto:mint@lists.fishpool.fi>
- List-subscribe: <mailto:mint-request@lists.fishpool.fi?Subject=subscribe>
- List-unsubscribe: <mailto:mint-request@lists.fishpool.fi?Subject=unsubscribe>
- References: <ed8542bdc154dd7693c8d3210c31d89b-EhVcX1lFRQVaRwYcDTpQCEFddQZLVF5dQUBFBDBTXF5bVkYOWkF0A1w6XF1YR0AAWlhfQw==-webmailer2@server08.webmailer.hosteurope.de> <4C163DF4.9060102@freesbee.fr> <845495def3bc16e5a0fcba375627015b-EhVcX1lFRQVaRwYcDTpQCEFddQZLVF5dQUBFBDBTXF5bVkYOWkF0A1w6XF1YR0AAWF9bSA==-webmailer2@server08.webmailer.hosteurope.de>
- Sender: mint-bounce@lists.fishpool.fi
- User-agent: Thunderbird 2.0.0.24 (Windows/20100228)
m0n0 a écrit :
That mean's when compiling for -m68020, the FPU is completly ignored, too.
Right?
On GCC 2.x, -m68020 means 68020 without FPU.
On GCC 4.x, -m68020 means 68020 with a mandatory FPU.
This is a difference.
And I highly recommend to compile any FPU software with -m68020-60 since
this is the option we have chosen for standard multilibs.
I know that an FPU can speed up floating point operations dramatically,
and floating point math is involved alot in encryption / decryption - as
far as I know.
The problem with floating point is the fact it is imprecise by design:
there are rounding errors. Most of times, it doesn't matter because they
are extremely small.
But encryption/decryption software needs to be 100% accurate. I suspect
these algorithms to use integer math only.
Well, I don't know.
Anyone here has numbers regarding the speedup when using the FPU? I would
at least expect values like 100% speed inprovement ;) But maybe I'm
totally wrong?
For standard software: no gain.
For math/3D software: huge gain.
--
Vincent Rivière