| From | Sent On | Attachments |
|---|---|---|
| O. Hartmann | Jun 20, 1999 5:34 am | |
| Richard Cownie | Jun 20, 1999 11:40 am | |
| Tim Vanderhoek | Jun 21, 1999 8:38 am | |
| Kedar Rajadnya | Jun 21, 1999 8:49 am | |
| Kenneth D. Merry | Jun 21, 1999 9:01 am | |
| Kedar Rajadnya | Jun 21, 1999 9:03 am | |
| Richard Cownie | Jun 21, 1999 9:37 am | |
| Steve Chapin | Jun 21, 1999 1:08 pm | |
| M. L. Dodson | Jun 21, 1999 1:26 pm | |
| Mika Nystrom | Jun 21, 1999 2:05 pm | |
| Gary Palmer | Jun 21, 1999 3:42 pm | |
| Gary Palmer | Jun 21, 1999 3:47 pm | |
| Terry Lambert | Jun 21, 1999 5:43 pm | |
| Richard Cownie | Jun 22, 1999 7:48 am | |
| Aaron Grosky | Jun 22, 1999 9:03 am | |
| Terry Lambert | Jun 22, 1999 11:59 am | |
| Steve Chapin | Jun 23, 1999 10:26 am |
| Subject: | Re: SMP, 4GB RAM, 4x CPU | |
|---|---|---|
| From: | Richard Cownie (ti...@ma.ikos.com) | |
| Date: | Jun 22, 1999 7:48:22 am | |
| List: | org.freebsd.freebsd-smp | |
On Mon, 21 Jun 1999, Terry Lambert wrote:
DEC (used to; haven't checked lately) have instruction scheduling code up for download on gatekeeper.dec.com. From memory, running it as an assembler preprocessor was pretty trivial. Of course, DEC may have advanced the technology without posting new code, so YMMV...
To expose enough parallelism to get a good instruction schedule, you probably need to do a lot of stuff earlier in the compilation, e.g. loop unrolling, traversing the expression tree in a different order for code-generation/register-allocation. In particular, once the register allocation is cast in stone you don't have very much freedom to re-order instructions. So I doubt that the assembler pre-processor makes a big difference - if you want good performance from the Alpha (and why else would you want an Alpha ?) the DEC compiler is probably the only game in town.
Unless anyone's tried the assembler scheduler and has figures to suggest otherwise ?
I notice that egcs (soon to be gcc-2.95) has a bunch of new instruction-scheduling stuff, maybe this will do better than the old gcc.
Richard Cownie (ti...@ma.ikos.com)
To Unsubscribe: send mail to majo...@FreeBSD.org with "unsubscribe freebsd-smp" in the body of the message





