The best AMIGA assembler/c-comp. ? - Page 2

khephren · 22 September 2003, 12:48

I agree with the AsmOne crowd, the best without doubt. Don't remember any major bugs in it, was just not as annoying as DevPac or Argasm.

Anyone remember using one of the thousands of versions of Seka?

Only reason I used it was because that was what Tec/CRB line source was written in it.

I used DICE C on the Amiga (sure it was called that, memory is fading quicker now).

zerohero · 14 February 2004, 10:42

Could someone please re-upload Slayers SAS/C 6.58 package to the Zone, it would be appreciated. I have been considering doing the four level upgrade tap dance but decided to ask if there is a kind person here who could spare me the trouble. I would probably endup breaking the keyboard on my forehead.

Slayer · 14 February 2004, 14:17

my pleasure

please check end of READ.ME for other neccessary
assigns...

good luck!

zerohero · 14 February 2004, 14:55

Hey, thanks....

You spared my keyboard.

AmiGer · 16 February 2004, 10:57

Meanwhile I have made some experiences with (dis-)assemblers/debuggers concerning the CARE project.

For me the best disassembler is IRA. D68k is also great.
And the best and most compatible assembler is PhxAss.
With these great tools it's possible to rebuild a 1:1 executable.
I have the sources of all of these programs (these are not free). So it's possible to tweak them to my own needs.
D68k and PhxAss are programmed in assembler, IRA in C.
All of them are very fast.

The best low-level debugger is ... the WINUAE built-in debugger because it can be used in every situation.

With the last WinUae version the debugger became better and more useable. On a real (older) Amiga I would prefer
the Action Replay as low level debugger.
I don't use symbolic debuggers so I don't know what I would prefer.

zerohero · 16 February 2004, 22:30

I checked out your Care project page and i think it is commendable that you take the old amiga classics and improve them.

Interesting tools too, ira seems to be what i have been looking for

I have tried Resource 6.06 but having problems using non standard gfx modes i.e. those provided by the picasso card so i have to run it in 640*512. Its cleaner with a cmd line tool

compared to a messy gui.

What do you think of a asmpro 1.16i it was mentioned in the ira documentation that phxass was better suited for larger programs? and what linker do you use, phxlnk, dlink??

Do you repair broken disks?? i have a messed up music intro disk which refuses to load. I thought i might salve the disk and disassemble the loader?? do you have any advice?

Anyways, i thought the name "phxass" was cool, what happens if you say it real fast... hmmm

Do you have any new disassembling projects or do you plan on improving elite?

AmiGer · 17 February 2004, 08:06

I think I remember that I had some compatibility problems with AsmPro.
The big advantage of using IRA and PhxAss is, that the authors (btw. both german) worked together until some years ago.
They tweaked their programs until they reached a 1:1 reassembling success.

Normaly I use the PhxAss build-in Linker or PhxLnk.

Disassemble/analyze/debug(winuae) boot block then the loader (if there is a seperate).
If it's a disk motor problem this could help.

To give you an idea: 50000 source lines compiled in less than 45 seconds that's fast... (on emulated A1200)

I concentrate on real life the most time now but I have a new project. I make good progress but it will take much time...
I don't want to reveal to much...

zerohero · 18 February 2004, 12:01

I started to try out winuaes debugger on Tosec, stolen data #8 which has a "software error/Task Held" after selecting the diskmag in the menu. It seems that the program goes into an eternal loop which consists of the following lines:

BSET.B #$0007,(A6,$0124) == $00c0039a #Test a bit and set
#Enable & wait for interrupts, user interrupt vector??
STOP.L #$2000
#Is this Branch, always terminate?
BT .B #$ffffffe6 == 00fc0f7c (TRUE)

Any idea how to proceed from here, perhaps Nop out BT?? and see what happens?

Toni Wilen · 18 February 2004, 12:25

That is Kickstart's idle-loop which is run when there is nothing else to do..

Most likely reason for task held-message is illegal instruction, address error etc.. exception in loader program

AmiGer · 18 February 2004, 14:58

You are debugging the ROM... (look at the PC address e.g. $FC0E00...)

You have debugged too far... Which instructions/library calls/direct hardware accesses etc. of the >user< program caused the jump to this routine... !?

19 February 2004, 00:11

Quote:

Originally posted by AmiGer
For me the best disassembler is IRA.

I have to agree with this. I tried Resource - the so-called creme de la creme of Dissasembling, even v6. No luck whatsoever. I'm sure if you know what you are doing this is the better dissasembler (manuals anyone!?) but IRA worked the first time.....well, actually the second when I downloaded the latter version.

Make sure you get from Aminet the 1.05 version named "ira105_src.lha" as it does contain the exe too, and not the older 1.02 version.

AmiGer · 19 February 2004, 07:58

Quote:

Originally posted by Jim
I have to agree with this. I tried Resource - the so-called creme de la creme of Dissasembling, even v6. No luck whatsoever.

I don't know "Resource". Is it only for disassembling or for resource extraction, too ? Where can I get it... ? Is it free... ?

Quote:

Make sure you get from Aminet the 1.05 version named "ira105_src.lha" as it does contain the exe too, and not the older 1.02 version..

Sorry, the newest version is 1.07beta. No sources or manual available (lost by author). I don't know the differences but this version is from 1997... Please PM who's interested...

zerohero · 19 February 2004, 08:50

>That is Kickstart's idle-loop which is run when there is nothing else to do..

Yep, that explains the eternal loop, i'll continue debugging tonight

I'll set my breakpoints a bit earlier.

I'll upload Resource to the zone if you want it..?

AmiGer · 19 February 2004, 08:58

Quote:

Originally posted by zerohero
i'll continue debugging tonight

Good luck.

Quote:

I'll upload Resource to the zone if you want it..?

I found V5 on my own but if you have a newer one (what is the newest ?) an upload would be nice...

zerohero · 19 February 2004, 11:12

I think 6.06 is the newest version and i have applied a hex patch found on aminet, can't find it now, but i will upload what i have
when i get home from work

zerohero · 19 February 2004, 12:16

Resource 6.06 + patch + manual is in the zone..

19 February 2004, 14:47

Quote:

Originally posted by zerohero
Resource 6.06 + patch + manual is in the zone..

Thanks

Will see how far I get with this one.....

AmiGer · 19 February 2004, 14:55

Thanks from me, too.

For me it looks like V5 but with big installation process and many addons...
Unfortunately it seems that it can't handle overlay hunks completely, but it's good anyway...

zerohero · 19 February 2004, 23:31

Ok... i have had another debugging session and i started this time setting breakpoints in the menu selection loop and it seems
to be an illegal instruction.

JSR.L (A6,$fdd8) == $00dfedd8 # prefetch fdd8 (ILLEGAL) 4eae (JSR), is this a failed memory alloc? AlloAbs(bytesize,location)(d0,a1)

The offset $fdd8 seems too large? is it two complement btw?

See attached logfile for more details.

AmiGer · 20 February 2004, 08:08

prefetch fdd8 (ILLEGAL) 4eae (JSR)
0005e494 4eae fdd8 JSR.L (A6,$fdd8) == $00c0004e
next PC: 0005e498

I don't know anything about these prefetch messages (ask Toni) but this JSR is
a standard exec _LVOOpenLibrary and that call was ok in your debug session (look at the return address in D0)...
It opens the dos.library !
The $FDD8 is not a large value. It's just a negative word: _LVOOpenLibrary EQU -552

The second JSR is an dos.library LoadSeg:
_LVOLoadSeg EQU -150

Then the program jumps into this new loaded routine.
Your fault is that you are debugging with the command "z" at the last JSR (A0). Use "t" and continue debugging...

Which disk image you're using exactly ?

16 February 2004, 10:57	#25
AmiGer Registered User Join Date: Sep 2002 Location: Germany Posts: 349	Meanwhile I have made some experiences with (dis-)assemblers/debuggers concerning the CARE project. For me the best disassembler is IRA. D68k is also great. And the best and most compatible assembler is PhxAss. With these great tools it's possible to rebuild a 1:1 executable. I have the sources of all of these programs (these are not free). So it's possible to tweak them to my own needs. D68k and PhxAss are programmed in assembler, IRA in C. All of them are very fast. The best low-level debugger is ... the WINUAE built-in debugger because it can be used in every situation. With the last WinUae version the debugger became better and more useable. On a real (older) Amiga I would prefer the Action Replay as low level debugger. I don't use symbolic debuggers so I don't know what I would prefer. Last edited by AmiGer; 25 November 2004 at 08:21.

16 February 2004, 22:30	#26
zerohero Registered User Join Date: Aug 2003 Location: schtaan/sweden Posts: 57	I checked out your Care project page and i think it is commendable that you take the old amiga classics and improve them. Interesting tools too, ira seems to be what i have been looking for I have tried Resource 6.06 but having problems using non standard gfx modes i.e. those provided by the picasso card so i have to run it in 640512. Its cleaner with a cmd line tool compared to a messy gui. What do you think of a asmpro 1.16i it was mentioned in the ira documentation that phxass was better suited for larger programs? and what linker do you use, phxlnk, dlink?? Do you repair broken disks?? i have a messed up music intro disk which refuses to load. I thought i might salve the disk and disassemble the loader?? do you have any advice? Anyways, i thought the name "phxass" was cool, what happens if you say it real fast... hmmm Do you have any new disassembling projects or do you plan on improving elite? Last edited by zerohero; 17 February 2004 at 07:46.*

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Assembler for Amiga OS 3.9	NubeCheCorre	support.Other	2	22 February 2012 01:39
Amiga Macro Assembler	Canyon Matt	request.Apps	0	06 January 2009 18:40
Total Amiga Assembler	pmc	Coders. General	7	24 May 2007 09:40
Comp Pro Amiga Classix Bundle	amigakit.com	MarketPlace	1	15 November 2006 02:27
comp.sys.amiga.games	kolorabi	Amiga scene	19	29 September 2003 19:25

22 September 2003, 12:48	#21
khephren Junior Member Join Date: Jan 2002 Location: Exeter, UK Age: 52 Posts: 128	I agree with the AsmOne crowd, the best without doubt. Don't remember any major bugs in it, was just not as annoying as DevPac or Argasm. Anyone remember using one of the thousands of versions of Seka? Only reason I used it was because that was what Tec/CRB line source was written in it. I used DICE C on the Amiga (sure it was called that, memory is fading quicker now).

14 February 2004, 10:42	#22
zerohero Registered User Join Date: Aug 2003 Location: schtaan/sweden Posts: 57	Could someone please re-upload Slayers SAS/C 6.58 package to the Zone, it would be appreciated. I have been considering doing the four level upgrade tap dance but decided to ask if there is a kind person here who could spare me the trouble. I would probably endup breaking the keyboard on my forehead.

14 February 2004, 14:17	#23
Slayer Amiga Member Join Date: Aug 2003 Location: New Zealand Age: 57 Posts: 695	my pleasure please check end of READ.ME for other neccessary assigns... good luck!

14 February 2004, 14:55	#24
zerohero Registered User Join Date: Aug 2003 Location: schtaan/sweden Posts: 57	Hey, thanks.... You spared my keyboard.

17 February 2004, 08:06	#27
AmiGer Registered User Join Date: Sep 2002 Location: Germany Posts: 349	I think I remember that I had some compatibility problems with AsmPro. The big advantage of using IRA and PhxAss is, that the authors (btw. both german) worked together until some years ago. They tweaked their programs until they reached a 1:1 reassembling success. Normaly I use the PhxAss build-in Linker or PhxLnk. Disassemble/analyze/debug(winuae) boot block then the loader (if there is a seperate). If it's a disk motor problem this could help. To give you an idea: 50000 source lines compiled in less than 45 seconds that's fast... (on emulated A1200) I concentrate on real life the most time now but I have a new project. I make good progress but it will take much time... I don't want to reveal to much...

18 February 2004, 12:25	#29
Toni Wilen WinUAE developer Join Date: Aug 2001 Location: Hämeenlinna/Finland Age: 49 Posts: 26,588	That is Kickstart's idle-loop which is run when there is nothing else to do.. Most likely reason for task held-message is illegal instruction, address error etc.. exception in loader program

18 February 2004, 14:58	#30
AmiGer Registered User Join Date: Sep 2002 Location: Germany Posts: 349	You are debugging the ROM... (look at the PC address e.g. $FC0E00...) You have debugged too far... Which instructions/library calls/direct hardware accesses etc. of the >user< program caused the jump to this routine... !?

19 February 2004, 08:50	#33
zerohero Registered User Join Date: Aug 2003 Location: schtaan/sweden Posts: 57	>That is Kickstart's idle-loop which is run when there is nothing else to do.. Yep, that explains the eternal loop, i'll continue debugging tonight I'll set my breakpoints a bit earlier. I'll upload Resource to the zone if you want it..?

19 February 2004, 11:12	#35
zerohero Registered User Join Date: Aug 2003 Location: schtaan/sweden Posts: 57	I think 6.06 is the newest version and i have applied a hex patch found on aminet, can't find it now, but i will upload what i have when i get home from work

19 February 2004, 12:16	#36
zerohero Registered User Join Date: Aug 2003 Location: schtaan/sweden Posts: 57	Resource 6.06 + patch + manual is in the zone..

19 February 2004, 14:55	#38
AmiGer Registered User Join Date: Sep 2002 Location: Germany Posts: 349	Thanks from me, too. For me it looks like V5 but with big installation process and many addons... Unfortunately it seems that it can't handle overlay hunks completely, but it's good anyway...

20 February 2004, 08:08	#40
AmiGer Registered User Join Date: Sep 2002 Location: Germany Posts: 349	prefetch fdd8 (ILLEGAL) 4eae (JSR) 0005e494 4eae fdd8 JSR.L (A6,$fdd8) == $00c0004e next PC: 0005e498 I don't know anything about these prefetch messages (ask Toni) but this JSR is a standard exec _LVOOpenLibrary and that call was ok in your debug session (look at the return address in D0)... It opens the dos.library ! The $FDD8 is not a large value. It's just a negative word: _LVOOpenLibrary EQU -552 The second JSR is an dos.library LoadSeg: _LVOLoadSeg EQU -150 Then the program jumps into this new loaded routine. Your fault is that you are debugging with the command "z" at the last JSR (A0). Use "t" and continue debugging... Which disk image you're using exactly ?

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)