[webkit-dev] Using OfflineAsm

[Vienneau, Christopher]

Thanks for the replies Mark, Michael and Filip.

To follow up:
Michael asks –“What processor(s) and OS(es) are you targeting?”
We are developing on game consoles, our focus is on the ps4 and xboxone.  The cpu’s are X86_64, and their OS are ‘similar’ to unix/windows.

Mark wrote – “the C LOOP LLINT emulated a CPU’s behavior using emitted C instructions.  As a result, it incurs a performance penalty that is under 10% (vs the ASM LLINT) if I remember correctly”
A potential 10% savings is something we’re interested in, we really do want it to perform as well as it can within the limitations are working with (no JIT).  We gave it a quick try since it appeared easy to do but encountered some build problems.  We have decided to try again once we re-integrate the recent changes from webkit into our codebase.  This will probably be quite some time from now, so I’ll pick it up again then.

Thanks again for the help

Chris

From: Mark Lam [mailto:mark.lam at apple.com]
Sent: Friday, September 05, 2014 4:14 PM
To: Vienneau, Christopher
Cc: webkit-dev at lists.webkit.org
Subject: Re: [webkit-dev] Using OfflineAsm


On Sep 5, 2014, at 4:05 PM, Mark Lam <mark.lam at apple.com<mailto:mark.lam at apple.com>> wrote:
4)      Should I actually expect a performance increase compared to the “C LOOP”?

Yes.  The ASM LLINT should be faster than the C LOOP LLINT.

To clarify, the C LOOP LLINT emulated a CPU’s behavior using emitted C instructions.  As a result, it incurs a performance penalty that is under 10% (vs the ASM LLINT) if I remember correctly.  So, the perf gain you’ll get comes from recuperating the losses due to the C loop being a CPU emulator.  You should not expect to see any of the more substantive gains that come from the JITs.

Regards,
Mark

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.webkit.org/pipermail/webkit-dev/attachments/20140913/aa4a716b/attachment.html>