Now that I've fixed the misassembly of diveby3.asm I can do timings
comparing tcc and gcc on Windows 32:

Multiply 2000 digit numbers

tcc: 13 ms
gcc: 12 ms

Multiply 200000 digit numbers

tcc: 9532 ms
gcc: 8902 ms

Compute 2000000!

tcc: 3373 ms
gcc: 2793 ms

So it's certainly within about 20% of the gcc timings for most things.
Some of that could be due to the bloating caused by tcc in the
assembly code, which probably pushes off a lot of the c/l timings.

For some reason when I try to multiply 1000000 digit numbers the tcc
version crashes, though the gcc version does not. I've no idea why
that would be. Probably another tcc bug. I'm not going to try tracking
it down at this point. The next step will be to get the test suite
working so we can see how good or bad things really are.

Bill.

--

You received this message because you are subscribed to the Google Groups 
"mpir-devel" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/mpir-devel?hl=en.


Reply via email to