• Open Menu Close Menu
  • Apple
  • Shopping Bag
  • Apple
  • Mac
  • iPad
  • iPhone
  • Watch
  • TV
  • Music
  • Support
  • Search apple.com
  • Shopping Bag

Lists

Open Menu Close Menu
  • Terms and Conditions
  • Lists hosted on this site
  • Email the Postmaster
  • Tips for posting to public mailing lists
Re: Stupid Intel question
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Stupid Intel question


  • Subject: Re: Stupid Intel question
  • From: Urs Heckmann <email@hidden>
  • Date: Sat, 18 Mar 2006 13:27:36 +0100

Hiya Frederik,

Frederik Slijkerman wrote:

I started the thread on Xcode-users about strict aliasing. You should know that there is a problem in AUElement::RestoreState that means that you can't use strict aliasing (and therefore also -ftree- vectorize) for now, unless you're willing to modify the CoreAudio SDK. I'll file a bug on that.

Yeah, I've browsed the list archive and I'll probably post some stuff, too, as it's important for me that pointer casting a Float32 to an UInt32 works perfectly. This is certainly a bug in gcc and there's no predictable pattern as to when it happens or not. As I said, it works in all other compilers I have around.


I am not scared to patch the CASDK, but I'm scared to patch all my mixed float/fixed point code ;-)

Regarding the optimization, I've noticed too in the disassembler that GCC 4 seems to produce rather sub-optimal code for Intel, even with -O3 and all the other optimization stuff on. Luckily the processor is quite fast so it doesn't matter that much at the moment, but it would be nice if this is improved in the future, or if it turns out to be due to a particular compiler switch.

Well, I think it does matter for us softsynth makers. The ES2 seems to run faster per clock on Intel cpus while our stuff will probably run slightly slower on Intel/OS X compared to G5s (or the same code compiled for Windows). Word has already spread about these performance tests, so customers will expect our stuff to catch up with this trend.


The MacBook 1.83 GHz feels dramatically faster than my single proc G5 1.8GHz, not only due to the second cpu core. The only software that feels slower is my own stuff, supposedly due to weird stuff going on in XCode/GCC.

Hmmm... does anybody know of a way to convert functions in a Win32 .dll into inline assembly for GCC?

8-))

Later,

;)  Urs


urs heckmann email@hidden www.u-he.com

_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden


  • Follow-Ups:
    • Re: Stupid Intel question
      • From: Steve Checkoway <email@hidden>
References: 
 >Stupid Intel question (From: Urs Heckmann <email@hidden>)
 >Re: Stupid Intel question (From: Steve Checkoway <email@hidden>)
 >Re: Stupid Intel question (From: Urs Heckmann <email@hidden>)
 >Re: Stupid Intel question (From: "Frederik Slijkerman" <email@hidden>)

  • Prev by Date: Re: Stupid Intel question
  • Next by Date: Re: Stupid Intel question
  • Previous by thread: Re: Stupid Intel question
  • Next by thread: Re: Stupid Intel question
  • Index(es):
    • Date
    • Thread