Re: [avr-libc-dev] fplib and MUL

[Nicolas Schodet]

* Björn Haase <address@hidden> [051011 23:40]:
> > I would like to inform that i have implemented the multiplication but I
> > have not tested it. As you can quess it was not hard. I have been
> > looking for the simulavr but looks like it is lots of work to make some
> > tests running with it. It seems easier to run the tests on real
> > hardware. For any case, I'll post the modified files from release 1.2.5
> If you are interested: Here you will be finding sample scripts for 
> automatically running numerics testsuites with simulavr. I have been using it 
> for testing my fractional numerics libraries at work, so you might not need 
> to make too many adaptions. You probably will however have to know a little 
> bit about gdb and maybe a little on the expect language.

I have also develloped a small fixed point lib and I use simulavrxx for
simulation.

I had only a few problems to compile simulavrxx on my debian sarge
system using the packaged avr tools, just install the binutils-dev
package for bfd, and modify two lines in avrdevice.cpp (122 and 131):

        size=sec->size;

with:

        size=sec->_cooked_size;

as the version of binutils shipped with debian is older than the one
used by the developpers.

Then I use the -W and -R simulavrxx options to output a quite large
amount of test cases. These options add two registers to the simulated
device to read from stdin or write to stdout. Very convenient!

Nicolas.