UPGRADE YOUR BROWSER

We have detected your current browser version is not the latest one. Xilinx.com uses the latest web technologies to bring you the best online experience possible. Please upgrade to a Xilinx.com supported browser:Chrome, Firefox, Internet Explorer 11, Safari. Thank you!

cancel
Showing results for 
Search instead for 
Did you mean: 
Highlighted
Visitor sandro17
Visitor
7,767 Views
Registered: ‎02-06-2016

SDK 2015.4 toolchain issue on sotfware-based floating-point support

Hi all,

 

I'm developing some bare-metal applications for the Cortex-A9 processor (Zynq platform) with my own BSP support. In my BSP code I do not enable the floating point hardware support (VFP and Neon). I compile the code with the “-mfloat-abi=soft” so that compiler ignore all FPU and NEON instructions, and use only the core register set and emulate all floating-point operations using library calls.  

The problem is that in the SDK 2015.4 toolchain this is not happening. You can see below the dissasembly code of the __udivdi3 function produced by the compiler for software-based floating point operations: in the first case the code was compiled with the SDK 14.7 (ISE 14.7) and in the second case with SDK 2015.4 (Vivado 2015.4). As you can see, in the second case some instructions related with the floating point hardware support still exists which crashes the application (going to the undef excpetion). In both cases all the code as well as compilation options were exactly the same.

There are any extra option that I need to add in the new compiler or is just an issue/bug of the new toolchain?!

 

Sem Título.0.jpg

Sem Título.jpg

 

Best,

Sandro

0 Kudos