Mailing Lists: Apple Mailing Lists
Image of Mac OS face in stamp
 
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Should I "manualy unfold" LoadUnaligned" function?

On Nov 1, 2005, at 3:36 PM, Sanjay Patel wrote:

Last time I checked, neither version of gcc had the logic to remove the "lvsl" instruction from the loop since your "target" is not loop-invariant (just loop-invariant in the last 4 bits most likely).

Take a look at this page:
http://developer.apple.com/hardware/ve/code_optimization.html

In particular, see the section titled "Efficient Safe Unaligned Vector Loads."

Most likely, you'll want to use a loop like the one shown that starts with:
fixAlignment = vec_add( vec_lvsl( 15, ptr ), vec_splat_u8(1) );

Thank you a lot, i definitly overlooked this particular page.
Could you please advice some info/references about how to store unaligned data from loops?
--
Sincerely,
Rustam Muginov

_______________________________________________
Do not post admin requests to the list. They will be ignored.
PerfOptimization-dev mailing list      (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/perfoptimization-dev/email@hidden

This email sent to email@hidden
References: 
 >Should I "manualy unfold" LoadUnaligned" function? (From: Rustam Muginov <email@hidden>)
 >Re: Should I "manualy unfold" LoadUnaligned" function? (From: Sanjay Patel <email@hidden>)



Visit the Apple Store online or at retail locations.
1-800-MY-APPLE

Contact Apple | Terms of Use | Privacy Policy

Copyright © 2007 Apple Inc. All rights reserved.