3.1.4 SetPatch is very slow

[patrik]

Originally Posted by Scrambler77

Hallo everyone. I'm having that issue too: KS 3.1.4/Workbench 3.1.4/SetPatch 45.15. 3.2 sec delay in startup-sequence.
Clean install. My A1200 has a Cobra 030 installed.

Upgrade to 3.1.4.1 and add the argument NODRIVELEDPATCH to the SetPatch command in your S:Startup-Sequence.

[AmigaHope]

Originally Posted by patrik

Example test runs:

9.Ram Disk:> SYS:UHC/C/time echo test
test
0.056383s
9.Ram Disk:> SYS:UHC/C/time wait 1
1.060052s
9.Ram Disk:> SYS:UHC/C/time wait 3
3.090985s
9.Ram Disk:> SYS:UHC/C/time wait 10
10.164959s

Not related to SetPatch, but is something wrong with your system clock? "wait 10" taking 10.16 seconds to complete seems incredibly wrong. Then again maybe the stock wait command isis just really badly written. The reason I wonder about your system clock is because the error seems to scale with the length of time you're running wait. It's almost as if the system clock and system bus crystal are mismatched.

[patrik]

Originally Posted by AmigaHope

Not related to SetPatch, but is something wrong with your system clock? "wait 10" taking 10.16 seconds to complete seems incredibly wrong. Then again maybe the stock wait command isis just really badly written. The reason I wonder about your system clock is because the error seems to scale with the length of time you're running wait. It's almost as if the system clock and system bus crystal are mismatched.

What is measured is the execution time of the Wait command. Apart from the actual waiting, this time will also include loading the Wait executable into ram plus any initialization/cleanup.

Wait can only guarantee that execution takes at least the time you specify.

Also, as this is a multitasking system, the execution time of a command will vary somewhat.

So you should always expect a greater execution time of Wait than the time you specified.

With that said, it is interesting that the measured time has a greater difference between the wait time specified and the time measured, the longer wait time you specify.

[Mark sealey]

Ive found this problem too...replaced 314 Setpatch with 3.9 last night....no lag in booting.....020 Furia

[gulliver]

Originally Posted by Mark sealey

Ive found this problem too...replaced 314 Setpatch with 3.9 last night....no lag in booting.....020 Furia

Dont do that!

The 3.9 setpatch does not apply some fixes and has a few bugs of its own compared to the 3.1.4.1 version.

They are not meant to be interchangeable.

[Mark sealey]

Originally Posted by gulliver

Dont do that!

The 3.9 setpatch does not apply some fixes and has a few bugs of its own compared to the 3.1.4.1 version.

They are not meant to be interchangeable.

Thanks for the reply....ant idea what the bug fixes were...seems to work so far.

[gulliver]

Originally Posted by Mark sealey

Thanks for the reply....ant idea what the bug fixes were...seems to work so far.

This is a small list that I can remember now, there is much more than
this, but I am not at home to take a look at the changelog
(also remember that the 3.9 setpatch is buggy).

So, the 3.1.4.1 SetPatch:

-Fixes audio.device bad channel allocation (doomsnd.library)
-Fixes shell FAILAT bug
-Patches gfx/Draw() bug
-Patches ExAll() bug
-Fixes for graphics/EraseRect().
-Fix for cardres/CardForceChange()
-Fixes some return parameters
-Tests and fixes UMult64/SMult64
-Fixes a bug in Alert 0x0100000f
-It enables proper use of the scsi.device LED
-It properly handles the 680?0.library
-It properly manages the FPU
-It can properly manage the system mathlibs
-It allows a 68060 to run and not crash

Also NSDPatch was removed as it was rather useless and a lot of dead
code went out too.

And have in mind that this is what I remember right now, there is for
sure more to it.

So again, don't use SetPatch from 3.9 in 3.1.4.1. :-)

[patrik]

Originally Posted by Mark sealey

Ive found this problem too...replaced 314 Setpatch with 3.9 last night....no lag in booting.....020 Furia

Have you tried the 3.1.4.1 SetPatch in combination with adding the NODRIVELEDPATCH argument to the SetPatch line in S:Startup-Sequence?

This should remove the 3s lag.

[DanyPPC]

Is there a solution to let OS 3.1.4 use Fast Ram on Blizzard1230 when mounting RAD disk ?

Sorry if this problem has just found an answer. In that case could you post the link for that ?
Many thanks,

DanyPPC

[PeterK]

You could try FastRAD:

http://aminet.net/package/util/boot/FastRAD11

[DanyPPC]

Many thanks, I'll try it.

[Metropolis]

Noob question: How does one specifically add NODRIVELEDPATCH to the Setpatch command in the 3.1.4.1 s:startup-sequence? I have the following...

SetPatch QUIET NODRIVELEDPATCH 

...in my s:startup-sequence, but the lag is still there on 3.1.4.1 boot and a direct call of SetPatch doesn't list any mention of NODRIVELEDPATCH.

[Reject72]

Originally Posted by Metropolis

Noob question: How does one specifically add NODRIVELEDPATCH to the Setpatch command in the 3.1.4.1 s:startup-sequence? I have the following...

SetPatch QUIET NODRIVELEDPATCH 

...in my s:startup-sequence, but the lag is still there on 3.1.4.1 boot and a direct call of SetPatch doesn't list any mention of NODRIVELEDPATCH.

Try

setpatch NODRIVELEDPATCH QUIET

this should work.

[PeterK]

Originally Posted by Metropolis

Noob question: How does one specifically add NODRIVELEDPATCH to the Setpatch command in the 3.1.4.1 s:startup-sequence? I have the following...

SetPatch QUIET NODRIVELEDPATCH 

...in my s:startup-sequence, but the lag is still there on 3.1.4.1 boot and a direct call of SetPatch doesn't list any mention of NODRIVELEDPATCH.

Then you still have SetPatch 45.15 from OS 3.1.4, but you need SetPatch 45.25 from OS 3.1.4.1 to use that option.

Just type "Setpatch ?" or use the Version command. Try "Which SetPatch" to see from it is loaded.

[Metropolis]

I have Setpatch 45.25 installed - 3.1.4.1 update was successful and I'm seeing the correct number when queried. I switched my startup-sequence to now read:

setpatch NODRIVELEDPATCH QUIET

but it seems as though the delay is still there. "Setpatch ?" returns the following:

QUIET/S,NOCACHE/S,REVERSE/S,NOAGA/S,NODRIVELEDPATCH/S:

Perhaps I'm doing something wrong?

[Cav]

Is it "correct" and "usual" that SetPatch 45.25 consumes about 2.5 MB of fastram on an 3.1.4.1 installation? Running on an A4000 with A3660 and MuLibs 680x0 46.6.1.

I'm also experiencing a ~2-3 second delay when starting, seem to be related to card.resource that fails (According to SnoopDos). I figure this is not present on an A4000 but only on PCMCIA equipped Amigas?

[Thomas Richter]

Originally Posted by Cav

Is it "correct" and "usual" that SetPatch 45.25 consumes about 2.5 MB of fastram on an 3.1.4.1 installation? Running on an A4000 with A3660 and MuLibs 680x0 46.6.1.

Well, it is not SetPatch requiring that memory, but the MMU tables requiring memory. How much memory that is depends on how much RAM you have in your system, and ditto how long it takes to create the tables.


There may also be other reasons for delays, the drive scan is one of them.

[Cav]

Originally Posted by Thomas Richter

Well, it is not SetPatch requiring that memory, but the MMU tables requiring memory. How much memory that is depends on how much RAM you have in your system, and ditto how long it takes to create the tables.


There may also be other reasons for delays, the drive scan is one of them.

You are of course right =)
I've disabled the 060-libraries and without them SetPatch consumes about 80k ..

I've only got 16 MB of fastram. So I figure there's something bogus going on in my system? What would the recomended version of 060 libraries to use with A4000, A3660, 3.1.4.1 ?

I'm using 46.6.1 right now.
http://aminet.net/package/util/sys/Mu680x0Libs

Br,
Daniel

[Thomas Richter]

Originally Posted by Cav

I've only got 16 MB of fastram. So I figure there's something bogus going on in my system?

It requires a lot of memory to "cut holes" into the MMU table to make room for some system resources that are not "properly mapped". Thus, what is in your system in terms of third party hardware extensions, and what is the output of MuScan?

[Cav]

@ Thomas Richter:

I have a ZZ9000, except from that nothing but the A3660.

MuScan 46.1 (02.07.2016) © THOR

68060 MMU detected.
MMU page size is 0x1000 bytes.

Memory map:
0x00000000 - 0x001FFFFF CacheInhibit Imprecise NonSerial
0x00200000 - 0x00BBFFFF Blank
0x00BC0000 - 0x00BFFFFF CacheInhibit I/O space
0x00C00000 - 0x00D7FFFF Blank
0x00D80000 - 0x00DFFFFF CacheInhibit I/O space
0x00E00000 - 0x00EFFFFF Blank
0x00F00000 - 0x00F7FFFF CacheInhibit
0x00F80000 - 0x00FFFFFF ROM CopyBack Remapped to 0x07F7C000
0x01000000 - 0x06FFFFFF Blank
0x07000000 - 0x07F7BFFF CopyBack
0x07F7C000 - 0x07FFBFFF ROM CopyBack
0x07FFC000 - 0x07FFFFFF CopyBack
0x08000000 - 0x3FFFFFFF Blank
0x40000000 - 0x4FFFFFFF CacheInhibit I/O space
0x50000000 - 0xFFFFFFFF Blank


PROCESSOR: CPU 68060/68882fpu/68060mmu
CUSTOM CHIPS: AA PAL Alice (id=$0023), AA Lisa (id=$00F8)
VERS: Kickstart version 46.143, Exec version 46.45, Disk version 45.194
RAM: Node type $A, Attributes $505 (FAST), at $7000000-$7FFFFFF (16.0 meg)
Node type $A, Attributes $703 (CHIP), at $4000-$1FFFFF (~2.0 meg)
BOARDS:
Board (unidentified): Prod=28014/4($6D6E/$4)
(@$40000000, size 256meg, subsize same)