9068 and faster code execution and use dual-core

Hi eveyone,

I'm working on a project where a C library is called a vi Labview on computer target. I have two platforms Crio namely 9082 and 9068 who have two operating systems and hardware. I compiled my code with Labwindows and successfully run in the Crio 9082 in 70 Ms when I check with the distributed systems management tool, are both the active processor cores taking on 50% load.

For the second platform Crio, namely 9068, I compiled the code even with Eclipse and run it on 9068. It works and produces the same result, but in Ms. 1400 when I check the CPU usage, a single processor takes almost 100% of load and the other is about 7%. It seems that the CPU allocation is different from 9082. I tried looping timed to manually adjust the cpu but it does not work (namely it allows only 1 cpu usage). My basic problem is to run my code faster in 9068, and of course, the problem may be different by an inefficient use of dual core in 9068. I tried different code optimizations (i.e., - O2,-O3, - Os) in Eclipse gear and they make almost no difference. Any suggestion is appreciated. Thanks in advance.

Gear

Advertisement

OK, so it looks like you don't have that a single thread of execution, then, is that correct?

There are compilation flags that can significantly improve the performance of point floating on the ARM. If you are not already doing so, I recommend the following (in http://www.ni.com/tutorial/14625/en/#toc5)

(Objectives focusing on the ARM) Perform the following steps to improve the performance of the floating-point operations:

  1. Select settings under of generation C/C++ in the left pane of the Properties dialog box.
  2. In the tool settings tab, select various under the cross GCC compiler .
  3. In the text box of other flags , add a space after the existing text and type - mfpu = vfpv3 - mfloat-abi = softfp

 

Tags: NI Software

Similar Questions

Maybe you are looking for


HashFlare