Re: [hatari-devel] Preparing release 2.3 / Slow IDE boot for Falcon

[ Thread Index | Date Index | More Archives ]


On 10/13/20 9:20 PM, Thomas Huth wrote:
Am Tue, 13 Oct 2020 09:59:40 +0200
schrieb Nicolas Pomarède <npomarede@xxxxxxxxxxxx>:
Le 13/10/2020 à 09:49, Eero Tamminen a écrit :
yes, maybe just disabling the IDE IO registers might be enough to
speed up
things in case --fast-boot is enabled.

I think it's likely the best solution for the next release. It's
the previous behavior with --fast-boot enabled, so this should not
be a regression for anybody, and for people who need more accurate
behavior, they
still can disable fast-boot.

Thomas, do you have time to do it this week?

If he doesn't have time, I think I can have a look myself

The patch itself is rather simple, so I quickly did the modifications
and pushed it. It's the testing that takes time here ... and I only did
some quick testing so far. So Eero, if you have some spare minutes, it
would be great if you could do some additional testing with this patch.

I had my machine running through following
TOS-boot tester configuration combinations
whole day:
  --disks floppy,gemdos,acsi,ide,scsi
  --graphics mono,rgb,vga,vdi1,vdi2,vdi4
  --machines st,megast,ste,megaste,tt,falcon
  --memsizes 2,4,8,14
  --ttrams 0,32
  --bool --compatible,--mmu

With all TOS, and EmuTOS version.  For TOS,
I tested both US & non-US variants, for
EmuTOS I tested only UK one.

This was done with the fast boot options enabled:

All EmuTOS variants passed all tests without

Some TOS combinations failed, which I'll check
closer to see whether I just need to increase
tester timeouts slightly (except for one fail,
all of them were with TOS v3.06 + MMU, none with
TOS v3.00 + MMU).

After that I'll do some runs without fast boot
options (with much reduced set of HW combinations
to keep runtime within 1 day).

	- Eero

Mail converted by MHonArc 2.6.19+