TOS RAM test routine ?

[Badwolf]

TF536 /50Mhz 64MB = 4m40s

Wow. That is extreme. My 128MB test only takes a couple of minutes IIRC & my RAM is slower than Stephen's.

BW

[exxos]

ZONESIZE corresponds to one character (- or X) printed, I think.

Not sure I follow that...

$20000 = 128 So kinda assume thats the block size the RAM test is using.

I guess if @frank.lukas As a EPROM I could just change at number and see if it makes any difference...

I cannot seem to get Hatari to play nice testing these TOS versions out

[Badwolf]

ZONESIZE corresponds to one character (- or X) printed, I think.

Not sure I follow that...
$20000 = 128 So kinda assume thats the block size the RAM test is using.
I guess if @frank.lukas As a EPROM I could just change at number and see if it makes any difference...

Ah, OK I must have missed that that was hard coded -- the amount of memory tested per character on the screen is calculated, obviously.

I cannot seem to get Hatari to play nice testing these TOS versions out

Hatari will try to apply patches to TOS versions it know about even with --patch-tos false set. You may need to change the version number to something it doesn't recognise.

I actively commented out the patching routines when doing my TOS4 hack.

BW

[exxos]

Hatari will try to apply patches to TOS versions it know about even with --patch-tos false set. You may need to change the version number to something it doesn't recognise.

I actively commented out the patching routines when doing my TOS4 hack.

Yeah turn that off but did not help. I either get four bombs with 206, or with a TT and patched to 306, it gets to desktop but does not ever do any memory test Why cannot something just be simple for once

Someone will just have to try that version patched to see if it speeds up RAM test.

[exxos]

This version has that number sent to $80000.

The hard drive wait was set to 12 seconds so I changed that to 3 seconds.

There was also another fix which I turned on..

/*
* TP_49: Fix memory size display
*
* The memory test at boot time could
* not correctly print values > 64MB
*/

tos206uk.zip

(127.26 KiB) Downloaded 55 times

[Badwolf]

Hatari will try to apply patches to TOS versions it know about even with --patch-tos false set. You may need to change the version number to something it doesn't recognise.

Yeah turn that off but did not help. I either get four bombs with 206, or with a TT and patched to 306, it gets to desktop but does not ever do any memory test Why cannot something just be simple for once

Perhaps you misread me: you can't turn off all patching. Also IIRC memvalid/ramvalid are automatically set so you never get a RAM test.

BW

[Badwolf]

TF536 /50Mhz 64MB = 4m40s

Wow. That is extreme. My 128MB test only takes a couple of minutes IIRC & my RAM is slower than Stephen's.

BW

Just tested the TOS4 TT-RAM test routine (normally never happens, obviously) in DFB1 -- 50s for 128MB@50MHz.

BW

[exxos]

Just tested the TOS4 TT-RAM test routine (normally never happens, obviously) in DFB1 -- 50s for 128MB@50MHz.

It must be running a different branch of the memory tests somewhere then

[stephen_usher]

Here are my results (64MB TT-RAM):

• ST536 TOS2.06 - 4 minutes 40 seconds.
• PAK/3 TOS3.06 - 23 seconds.

Just to note that the ST RAM test is similarly slow. with the PAK/3 TOS3.06 ROM taking only ~8 seconds to test 4MB.

I wonder if the 2.06 test is using 16 bit words during the test and the 3.06 test is doing 32 bit for a start. Maybe it's also turning the cache on before the test so that the whole loop of the program stays in cache.

[exxos]

Here are my results:

• ST536 TOS2.06 - 4 minutes 40 seconds.
• PAK/3 TOS3.06 - 23 seconds.

Just to note that the ST RAM test is similarly slow. with the PAK/3 TOS3.06 ROM taking only ~8 seconds to test 4MB.

I wonder if the 2.06 test is using 16 bit words during the test and the 3.06 test is doing 32 bit for a start. Maybe it's also turning the cache on before the test so that the whole loop of the program stays in cache.

Thanks.. its possible cache could be a issue... There is some experimental stuff in the TOS code relating to that, but I didn't turn it on.

Code: Select all

/*
 * TP_32: CACHE_0X0: Activate caches for all cpus >= 68020
 *
 * This patch is still experimental. Note also that with this patch active,
 * you won't be able to activate/deactivate the blitter from the desktop
 * when you have a cpu >= 68020.
 */
#ifndef TP_32
#define TP_32 0
#endif