Re: [hatari-devel] Hatari on constrained devices

[Christian Zietz <czietz@xxxxxxx>]

Christian Zietz schrieb:

> I would argue that a difference of 1-2% (in favor of small mem) is
> perhaps measurable but not noticeable. But there is no big performance
> hit, as one might have feared.

Further profiling shows that almost all memory accesses go through these
memory_get_...() functions, particularly through the 'else' clauses:
https://git.tuxfamily.org/hatari/hatari.git/tree/src/cpu/memory.c#n1982
.... which don't care at all whether the memory bank base address
'ab->baseaddr_direct_r' points to a statically or a dynamically
allocated ST RAM array.

Regards
Christian
--
Christian Zietz  -  CHZ-Soft  -  czietz@xxxxxxx
WWW: https://www.chzsoft.de/
PGP/GnuPG-Key-ID: 0x52CB97F66DA025CA / 0x6DA025CA