@Xenic

Should work, nonetheless. I tried with lha (2.12) on 68k. Maybe there is some glitch in lha, I have no time yet to run 2.15 using Enforcer.

In the worst case broadblue´s line would be recursive archiving, but as it works well on 68k, lha is apparently recursive trap proof.

Last edited by whose on 04-Jan-2011 at 11:14 PM.
Last edited by whose on 04-Jan-2011 at 11:12 PM.
Last edited by whose on 04-Jan-2011 at 11:11 PM.

@ChrisH

the file linked in that thread crashed the new xad archiver.
the file has been updated in the meantime - i don't know if it still crashes it.

thanks and regards
michael

_________________
Michael Merkel
(Michael.Merkel@gmx.net Home)
Member of Amiga-Freunde Pfalz, OS4 Welt

@broadblues

Quote:

Using your patch I get a ISI if open an test archive in RAM: created with the new lha

I have this same problem.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

@MichaelMerkel

Tried it with the one at the top of that thread and it works here. You don't happen to know which iBatch beta archive had the problem do you?

I dismissed this as a non-issue since only Andy had reported it and failed to reproduce it later.

I'd be interested to know if similar problems are happening under MorphOS (with the new LhA client internally).

Chris

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y
But tompeli have the same crash as well as i can read.

@Tompely
Maybe you can reproduce such crash, and send crash-log to cris_y ?

_________________
Join us to improve dopus5!
zerohero's mirror of os4/os3 crosscompiler suites

@kas1e
It crashes on my AmigaONE XE/G4

@Chris_Y
I've sent you the crashlog via PM.

@raumfahre

Crashlog seems to be the same as Andy's:

lha_external.c:534

Can't see anything particulary wrong which would be causing it, probably just stack space.

If you download again, you will get a version compiled with optimisations on - see if that makes any difference.

Otherwise check Prefs/DOS and up the process stack size (mine's 64K). If that doesn't help it may not be stack, in which case I'm at a loss as to what to suggest.

It's not my code, so I haven't extensively studied what it is doing, and I haven't been able to reproduce it here yet which makes debugging virtually impossible.

Chris

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y

If I pack anything with -H1 or -H2 (or without -H with this new version of Lha) then UnArc crashes. Stack pointer is inside boundaries and red zone is not touched. I tried with normal stack size and 1000000 stack size. I tried with max settings in DOS prefs. No difference. Then if I pack the same content with -H0 then UnArc works fine. Crash is ISI in null address.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

@Chris_Y

Quote:

Chris_Y wrote: @MichaelMerkel Tried it with the one at the top of that thread and it works here. You don't happen to know which iBatch beta archive had the problem do you?

i think it was the "a-beta" archive. currently only the "b-beta" is available... tried b-beta and had no problems.
unfortunately i don't have the a-beta any more

regards...

_________________
Michael Merkel
(Michael.Merkel@gmx.net Home)
Member of Amiga-Freunde Pfalz, OS4 Welt

@Tomppeli

I just tried again and got another crash, so it looks to be slightly random ...

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad

@broadblues

and again very similar crash log, my isi is not on zero address though.

Crash log for task "sys:utilities/UnArc"
Generated by GrimReaper 53.2
Crash occured in module  at address 0x94744BD4
Type of crash: ISI (Instruction Storage Interrupt) exception

Register dump:
GPR (General Purpose Registers):
   0: 00000000 5343E520 00000000 5883E610 5343E468 FFFFFFFF 021C801C 00000000 
   8: A4FAA960 545FBC58 021BD806 3E259DBA 28044042 54470000 54470000 54470000 
  16: 54470000 53C93BF8 54470000 54470000 53BDB498 5343EBD0 00000000 5446E740 
  24: 5446E540 5343E8B8 584DEA5C 536C3708 58836610 6FD1C070 00000000 5343E520 


FPR (Floating Point Registers, NaN = Not a Number):
   0:              nan                0    -5.48615e+303                0 
   4:                0            575.8             30.8     7.38464e-315 
   8:     3.39522e-313     2.78132e-309     2.64619e-260                0 
  12:     7.29112e-304     3.39522e-313                0                0 
  16:                0                0                0                0 
  20:                0                0                0                0 
  24:              255      2.14748e+09      1.52588e-05       4.5036e+15 
  28:            65536       4.5036e+15                0                0 

FPSCR (Floating Point Status and Control Register): 0x82004000


SPRs (Special Purpose Registers):
           Machine State (msr) : 0x0002F030
                Condition (cr) : 0x28044082
      Instruction Pointer (ip) : 0x94744BD4
       Xtended Exception (xer) : 0x00000008
                   Count (ctr) : 0x94744BD7
                     Link (lr) : 0x6FD1B570
            DSI Status (dsisr) : 0x00000000
            Data Address (dar) : 0x5883E6A0



680x0 emulated registers:
DATA: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 
ADDR: 021BD806 536C3708 00000000 00000000 00000000 00000000 00000000 5343E5C0 
FPU0:                0                0                0                0 
FPU4:                0                0                0                0 



Symbol info:
Instruction pointer 0x94744BD4 belongs to module "" (HUNK/Kickstart)

Stack trace:
    
    LhA:LhA_GetInfo()+0xB0C (section 1 @ 0x2b5c)
    module LIBS:xadmaster.library at 0x6FD2C5A8 (section 5 @ 0x9588)
    module LIBS:xadmaster.library at 0x6FD23128 (section 5 @ 0x108)
    module sys:utilities/UnArc at 0x6EF55744 (section 5 @ 0x3724)
    module sys:utilities/UnArc at 0x6EF57D34 (section 5 @ 0x5D14)
    UnArc:_start()+0xC08 (section 1 @ 0xe44)
    native kernel module dos.library.kmod+0x0001c210
    native kernel module kernel+0x00036290
    native kernel module kernel+0x00036310
    

PPC disassembly:
 94744bcc: XXXXXXXX   illegal address
 94744bd0: XXXXXXXX   illegal address
 94744bd4: XXXXXXXX   illegal address
 94744bd8: XXXXXXXX   illegal address
 94744bdc: XXXXXXXX   illegal address

System information:

CPU 
 Model: AMCC PPC440EP V1.3 
 CPU speed: 733 MHz 
 FSB speed: 146 MHz 
 Extensions:  

Machine 
 Machine name: Sam440EP 
 Memory: 1048576 KB 
 Extensions: bus.pci 

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad

@broadblues

I can reproduce it every time. Last line in stack trace is actually an empty line. Usually previous line is Lha_GetInfo but last time it was Lha.

Last edited by Tomppeli on 05-Jan-2011 at 07:53 PM.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

@Tomppeli

Weird....

Either there is still a bug in the XAD LhA client, or one of the changes I've made or the compiler I've used has introduced a problem.

Until I can reproduce it, I can't do much more, sorry

Source code is in the new archive though.

Chris

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y

Can't unpack that archive because UnArc crashes.....

okay probably I can use lha to do it, but that means Chris you have a test archive that should cause the crash.

What system are you running?

Same question to Tompeli

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad

@Chris_Y

Hmm you using clib2 does the rest of xad?, why that particular choice of library? Might be irrelavent ofcourse

I'll see if I can get the thing to build here, looks like I need to check out the rest of xad first....[quote]
broadblues wrote:

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad

@CHris_Y

Okay I'm confused, where abouts in the source tree do you place the sousrce in you archive, in order to get it to build with your lha_compile script?

_________________
BroadBlues On Blues BroadBlues On Amiga Walker Broad

@broadblues

portable/clients

I used clib2 because it's easier

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz

@Chris_Y

The process stack was set at 64K.
I've downloaded xad_lha.lha again and replaced in libs/xad and UnArc no longer crashes!

Last edited by raumfahre on 06-Jan-2011 at 10:19 AM.

@Chris_Y

Quote:

If you download again, you will get a version compiled with optimisations on - see if that makes any difference.

I downloaded it again now and UnArc works without crashes now. It looks like different optimisations made difference.

_________________
Rock lobster bit me. My Workbench has always preferences. X1000 + AmigaOS4.1 FE
"Anyone can build a fast CPU. The trick is to build a fast system." -Seymour Cray

Yay! Simple as that, then!

_________________
"Miracles we do at once, the impossible takes a little longer" - AJS on Hyperion
Avatar is Tabitha by Eric W Schwartz