[PATCH] Improve performance of vector code on A15

[Silviu Baranga]

Hi Renato,

I only looked at correctness for now when running LNT.

The LNT results aggregator script gives some random error and doesn't produce the
full performance report (probably needs some further configuration).

If there will be any problem with the compile time I'm sure there are ways to
Improve it. Also, since we're disabling the  ExecutionDependencyFix pass it may
actually compile faster.

Cheers,
Silviu

From: Renato Golin [mailto:renato.golin at linaro.org]
Sent: 14 March 2013 13:05
To: Silviu Baranga
Cc: Jakob Stoklund Olesen; Commit Messages and Patches for LLVM
Subject: Re: [PATCH] Improve performance of vector code on A15

On 14 March 2013 12:24, Silviu Baranga <silbar01 at arm.com<mailto:silbar01 at arm.com>> wrote:
Would this be close to a state where it can get committed?

Hi Silviu,

I didn't look very thoroughly, but I think it's specific enough that any further problem can be fixed after commit.

It seems you're taking enough care In the common parts, so that only A15 will go down that path, so it should be pretty safe.

Do you have any improvements on LNT or other benchmarks? What's the slow-down you're seeing on compile time on A15? I'm guessing both answers on non-A15 is zero...

cheers,
--renato

-- IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130314/9ccd8eb1/attachment.html>