Page 2 of 3

Re: GOLD2.5 Experimental Build

Posted: Sat 17 Jun 2017 23:00
by TrashyMG
SpacemonkeyUK wrote:
Fri 16 Jun 2017 22:15
TrashyMG wrote:
Wed 14 Jun 2017 19:02
I think so, if you mean the IDEFIX, which is just Gold 2 with the IDE stuff.
No it enable vampiretool Fastide to test faster IDE transfers
AFIK the "vampiretool fastide" setting on a V600 doesn't do anything. it only works on the V500. the passed around IDEFIX core is for the V600 only And the approach for enabling and setting up the appropriate values was a bit more manual.

Re: GOLD2.5 Experimental Build

Posted: Sat 17 Jun 2017 23:28
by SpacemonkeyUK
sorry but the Gold 3 b1 i have double the speed of the IDE in Sysinfo when i run vampiretool fastide

Re: GOLD2.5 Experimental Build

Posted: Sun 18 Jun 2017 00:19
by TrashyMG
Alright, I was talking about the idefix.jic core that was passed around for the V600, I wasn't aware of a Gold3 beta being passed Around.

I guess you're in the inner circle.

Re: GOLD2.5 Experimental Build

Posted: Sun 18 Jun 2017 00:31
by SpacemonkeyUK
no just asked on IRC to test IDE

Re: GOLD2.5 Experimental Build

Posted: Sun 18 Jun 2017 13:15
by SpacemonkeyUK
After updating to Gold 2.5 i now get some screen corpution on the boot screen towards the end of booting. This was not present on Gold 2 or 3b1

Image

Re: GOLD2.5 Experimental Build

Posted: Sun 18 Jun 2017 19:18
by Uigiflip
I get the same

Re: GOLD2.5 Experimental Build

Posted: Mon 19 Jun 2017 13:59
by TrashyMG
Yeah I get that too, although just a single strip down on the bottom.

Re: GOLD2.5 Experimental Build

Posted: Mon 19 Jun 2017 14:00
by TrashyMG
Yeah I get that too, although just a single strip down on the bottom. It doesn't seem to cause any issues on the WB or in any games from what I can tell.

Re: GOLD2.5 Experimental Build

Posted: Wed 21 Jun 2017 21:27
by flype
hmm, interesting guys, thanks for the report.

i can understand why : something new is allocating memory at end of fastram (using AllocMem MEMF_REVERSE flag) on boot process that was not before. No big issue, it is a 'legal' symptom. The BootPic allocates memory at end of fastram to make it displayable, but it is released as fast as possible. So any program that allocates memory reverse, after the BootPic is executed, will override the picture. Will look at this deeper when i can.

One question : does it happens when no CF is connected or without startup-sequence ?

Re: GOLD2.5 Experimental Build

Posted: Wed 21 Jun 2017 21:33
by bax
This symptom is a little bit annoying and can be fixed. But it's also harmless. The boot picture does free the allocated memory after showing the picture. Thus, when another program in the startup-sequence allocates FastMem from the top of the list, it will get some that falls into the boot picture framebuffer range. This causes the curruptions. Rest assured, the video out is read-only.