Rockbox mail archiveSubject: Re: Proposal to update GCC for ARM
Re: Proposal to update GCC for ARM
From: Thomas Martitz <thomas.martitz_at_student.htw-berlin.de>
Date: Tue, 09 Mar 2010 00:48:00 +0100
Am 09.03.2010 00:43, schrieb Karl Kurbjun:
> I think that updating to the latest version of GCC would be nice, but
> I am curious what the codec performance is on a "non-ideal" situation.
> I noticed that the tests you performed are against a PP target which,
> as I understand, still uses long calls.
> I would imagine that the long call targets would represent the
> best-case performance by migrating.
> Have you done any tests on the Gigabeat or Sansa AMS to see how the
> codec performance stacks up on the EABI builds verus the 'old' GCC
> setup we use?
> If the results on the Gigabeat or Sansa AMS are similar to the Sansa
> v1 results then I would be for updating the toolchain.
The AMS Sansas are in a bad situation w.r.t. to test_codec. Their low
memory doesn't permit running test_codec on most test files since test
codec wants the full file in ram (you can basically only test upto
128kbit/s). But if someone tested gigabeatfx, it should give quite
similar results since they have almost the same processor.
I don't think the extra function call overhead caused by long calls are
Received on 2010-03-09