A1200 super slow to boot with 3.1.4 ROM

[StompinSteve]

Bonjour,

I swapped my KS3.1 ROM with an official 3.1.4 ROM today. Since then, the machine does nothing for 41 seconds before it shows the "insert floppy" animation (when I disconnect the SCSI HDD and have nothing in the Gotek).

My A1200 is a C= A1200 with Blizzard1240 @ 40mhz (maprom enabled), SCSI-kit-IV, 64MB fast, unmodified, offical 3.1.4 kickstart ROM Chips.

I have an IDE terminator on the motherboard (from AmigaStore.eu) that prevents the long waits when there is nothing plugged in there.

When I ran with kick 3.1, when I powered on the machine, it began booting from the SCSI HDD immediately. No delays. It also started booting from floppy/gotek right away. When I kept both mouse buttons pressed, I got to the early boot-screen right away. No delays. The "IDE terminator" did it's job perfectly.

I did a "try before you buy" using the "loadmodule romupdate" method. Sure it had to reboot because of it, but the machine always started booting without delay after power-on or reboots. The ROM was still a 3.1 after all.

Now with the 3.1.4 ROM Chips in there, the machine always just sits there for 41 seconds (yes I used a stopwatch) before doing anything. I sit there, with both mouse buttons pressed for 41 seconds and then the early bootscreen is displayed.
After a reboot with 41 seconds of nothingness, it boots from floppy/gotek or the SCSI harddisk. The system works fine after that.

Without the IDE terminator plugged in, the system waited around 18 seconds before it started doing anything. But I have the terminator still plugged in and boot-ups happened 2 or 3 seconds after power-on or reboots with ROM 3.1.

I have been thinking. I have one of those Gayle adapters plugged in on top of Gayle that fixes the PCMCIA reset bug. So running CardPatch and CardReset is no longer needed.
The functionality of CardReset is already covered by the 3.1.4 update from what I understand, soooooo could the 41 seconds delay be caused by the presence of the Gayle adapter???

I need to fix this. Loading the V45 intuition.library triggers a reboot too ofcourse so that's two "41 seconds picking it's nose" wait-times before anything starts to happen. It now takes around 2 minutes in total to boot the machine to Workbench

(I've been doing some googling but found nothing similar. I found Patrik's thread from January 2019 but that's not the same issue). My machines is "dead/frozen" for 41 seconds before anything can be done (boot from hdd or floppy, display early boot).

[thomas]

I would first remove the terminator and see if you get back to 18 seconds. Or even less. I am not sure if it is the case, but maybe the "nothing connected" detection has been improved and your terminator indicates "something connected" and it waits for a reply.

In the worst case I would replace the terminator by a CF-IDE adapter with a CF card which has an RDB with all the "last" flags set but no partitions. It could even contain an emergency system for the case that the SCSI disk fails.

[van_dammesque]

Put your old ROM back in and make sure it boots normally too.

[StompinSteve]

Originally Posted by van_dammesque

Put your old ROM back in and make sure it boots normally too.

It booted normally just before I swapped the ROM chips.

@ Boing-Ball: the 1230scsi.device is 8.5

I'll try Thomas his hunch and remove the IDE terminator as soon as I get the chance. I have a feeling that might be it as well.
They could have changed the code in ROM 3.1 from:
"I sense no IDE device so i'll wait" (which is mitigated by the IDE teminator, hence no wait-time after power-on)
to
"I sense an IDE device (the terminator) but it does not repond thus i'll wait".

I'll report the result after I removed the terminator.

[StompinSteve]

And here is the result of terminating the IDE terminator: 28 seconds of "doi'n notti'n" instead of 40. 12 Seconds improvent

Next I will attach a proper IDE-CF-Card thingamagic to the IDE port and see what that does.

[fxgogo]

I have never heard of IDE terminators. You learn something new everyday. So I have 3.1.4 on my 1200 and it definitely does take longer, but not that much longer. Just thinking out aloud here, bUt what if the roms hAve not been burrned properly and this is the result?

[StompinSteve]

Originally Posted by fxgogo

I have never heard of IDE terminators.

https://amigastore.eu/en/752-amiga-i...or-44-pin.html
https://amigastore.eu/en/716-amiga-i...r-40-pins.html

AmigaKit produces them too I believe.
I have one plugged in into my TF536 in my A500 as well. It only has a GVP SCSI HDD and it reduces the boot-time-timout (let's call it that) significantly as there is no IDE HDD (CF-Card) present on the TF.
The A500 runs 3.1 at the moment.
When I get my hands on an A500 3.1.4 ROM, I will report if its boot-behaviour changed too.

Originally Posted by fxgogo

So I have 3.1.4 on my 1200 and it definitely does take longer, but not that much longer. Just thinking out aloud here, bUt what if the roms hAve not been burrned properly and this is the result?

Not saying that it is impossible, but my 3.1.4 ROM's are the official ones and the system is rock stable otherwise. I don't have a reason to suspect them.

I think Thomas is right and the behaviour has changed in 3.1.4.

[StompinSteve]

Originally Posted by thomas

In the worst case I would replace the terminator by a CF-IDE adapter with a CF card which has an RDB with all the "last" flags set but no partitions. It could even contain an emergency system for the case that the SCSI disk fails.

Hello Thomas,

I connected a CF-Card to the motherboard and viola, boots right up. No delays.
The mere presence of an IDE device on the motherboard removes the delay entirely. No need to partition it or anything. Just physically being there is enough.

It does imply that the behaviour did indeed change in ROM 3.1.4

Thanks for the tip