Software Archive
Read-only legacy content
Announcements
FPGA community forums and blogs on community.intel.com are migrating to the new Altera Community and are read-only. For urgent support needs during this transition, please visit the FPGA Design Resources page or contact an Altera Authorized Distributor.
17060 Discussions

Is it possible to vectorize atomic operation on MIC?

Yuan_Y_
Beginner
‎07-30-2013 07:08 PM
619 Views

Hi all,

I use gcc built-in function such as __sync_fetch_and_add in my program for atomic operations. Is it posibble to vectorize these atomic operations?

Thanks.

0 Kudos

Link Copied

1 Reply
jimdempseyatthecove
Honored Contributor III
‎07-31-2013 12:50 PM
619 Views

__sync_fetch_and_add uses:

LOCK; XADD regValue, [location]

LOCK prefix is not supported for PADDB/PADDW/PADD
Additionally, the destination must be an XMM register.

You could use a CMPXCHG16B (or CMPXCHG8B) as a DCAS. This will get you partway to where you want to go.

Jim Dempsey

0 Kudos
Copy link
Reply

Community support is provided Monday to Friday. Other contact methods are available here.

Intel does not verify all solutions, including but not limited to any file transfers that may appear in this community. Accordingly, Intel disclaims all express and implied warranties, including without limitation, the implied warranties of merchantability, fitness for a particular purpose, and non-infringement, as well as any warranty arising from course of performance, course of dealing, or usage in trade.

For more complete information about compiler optimizations, see our Optimization Notice.