Re: Compiled 3.8.0 from source on Ubuntu 12.10 slower than 3.6.3 from PPA

[Alexander Barth]

I can't really share my code, and the one-liner is close enough to being

representative to serve as a suitable test - at the very least, that
one-liner shows the same "regression"...

For what it is worth, I can confirm that octave 3.8.0 without jit is about 50 % slower than octave 3.6.4 for the one-liner:

octave 3.6.4:
octave:1>> a = 1; b = 1; tic; for i=1:1000; for j=1:1000; a = a + b; end; end; toc
Elapsed time is 1.07851 seconds

octave 3.8.0:

octave:1> a = 1; b = 1; tic; for i=1:1000; for j=1:1000; a = a + b; end; end; toc
Elapsed time is 1.71183 seconds.
octave:2> jit_enable (1)
octave:3> a = 1; b = 1; tic; for i=1:1000; for j=1:1000; a = a + b; end; end; toc
Elapsed time is 0.0122502 seconds.

Both versions of octave are compiled from source using the same libraries. I have also a similar CPU: Intel(R) Core(TM) i3-3120M CPU @ 2.50GHz.

For a vectorized "real world" code however (e.g. test_divand from divand package), I see only a very small difference: 4 or 5% slowdown of 3.8.0 (with or without jit) compared to 3.6.4.

Cheers,
Alex