|
From: | Omkar Wagh |
Subject: | Re: [Libjit] libjit about 1000 times slower than native functions |
Date: | Thu, 19 Sep 2013 14:47:42 -0400 |
User-agent: | Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130625 Thunderbird/17.0.7 |
Hey On 09/18/2013 02:20 AM, Patrick van Beem wrote:
Thanks Patrick for pointing that out. Yes, jit_function_to_closure did bring it within comparable levels of our hand-coded interpreters. It's not more efficient but at least we have more flexibility.If possible (if you have a known fixed number of arguments), you can use closures for a fast call to your compiled functions: void * jit_function_to_closure (jit_function_t func)
Regards, Omkar
[Prev in Thread] | Current Thread | [Next in Thread] |