On Tue, Apr 9, 2013 at 2:13 AM, wrote:
I'm inclined to reject this. I suspect that optimizing b0,b1,b2,b3 is
just tinkering around the edges: for real speed-up you need to eliminate
the bounds checking and do the equivalent of C++'s UNALIGNED_LOAD32.
Doing that will require re-writing in assembler, or at least using
package unsafe, and would remove any mention of b0,b1,b2,b3.
I'm inclined to reject this. I suspect that optimizing b0,b1,b2,b3 is
just tinkering around the edges: for real speed-up you need to eliminate
the bounds checking and do the equivalent of C++'s UNALIGNED_LOAD32.
Doing that will require re-writing in assembler, or at least using
package unsafe, and would remove any mention of b0,b1,b2,b3.
-j
--
---
You received this message because you are subscribed to the Google Groups "golang-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to golang-dev+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
