Richard Drummonds new UAE 0.8.25

[kadi]

I still have a nasty bug.

When i start IBrowse it works perfect for the first time. The new improved bsdsocket emulation seems to work perfect.

But i cannot start IBrowse again after i closed it.

Under KDE the Mousepointer gets freezed and i have to log out and in.

Under Gnome just the emulator task crashes and gets discarded.

theres an output in the shell:

[color=0000CC]

Found 0 joysticks
Found x11pc raw keyboard mapping
Using cooked keymap
Opening cfgfile: /home/kadi/.uaerc...okay.
Testing the RDTSC instruction ... done.
Calibrating delay loop.. ok - 1393,97 BogoMIPS
Testing system sleep function.
Average duration of a 1ms sleep: 20 ms
Enabling busy-waiting for sub-10ms sleeps
Found 0 joysticks
Found x11pc raw keyboard mapping
Using cooked keymap
Follow sprite mode set
No disk in drive 0.
Resetting frame rate hack
Building CPU table for configuration: 68040
1866 CPU functions
Building CPU function table (4 0 0).
Max CPUID level=1 Processor is AuthenticAMD
have_cmov=1, have_rat_stall=1
JIT: Setting signal handler
JIT: Building Compiler function table
Supposedly 42381 compileable opcodes!
Resolution: 800 x 600 x 32
Resetting frame rate hack
reset at 0
Don't care mouse mode set
PAL mode, 50Hz
Byte put to custom register DA8000 PC=00F8010A
Byte put to custom register DA8000 PC=00F80152
Fastmem (32bit): mapped @$10000000: 32 MB Zorro III fast memory
   Card 1 (ZorroIII) done.
UAEGFX-card: mapped @$12000000
   Card 2 (ZorroIII) done.
Filesystem: mapped memory @$e90000.
   Card 3 (ZorroII) done.
Byte put to custom register DFF110 PC=00F8BD96
Byte put to custom register DFF111 PC=00F8BD96
Byte put to custom register DFF112 PC=00F8BD96
Byte put to custom register DFF113 PC=00F8BD96
Byte put to custom register DFF114 PC=00F8BD96
Byte put to custom register DFF115 PC=00F8BD96
Byte put to custom register DFF116 PC=00F8BD96
Byte put to custom register DFF117 PC=00F8BD96
FS: mounted virtual unit st0
FS: mounted virtual unit st1
FS: mounted virtual unit st4
FS: mounted virtual unit st5
PTEST @$1002452c
SetDAC()
Resolution: 1024 x 768 x 32
JIT: can't handle access!
JIT: instruction byte  0 is 8b
JIT: instruction byte  1 is 57
JIT: instruction byte  2 is 04
JIT: instruction byte  3 is 89
JIT: instruction byte  4 is 55
JIT: instruction byte  5 is dc
JIT: instruction byte  6 is 83
JIT: instruction byte  7 is e2
JIT: instruction byte  8 is f8
JIT: instruction byte  9 is 83
Segmentation fault
[/color]

The same AmigaOS-Installation works perfect under winUAE, so i guess its a UAE thing rather tan a AmigaOS thing.

[kadi]

>But i cannot start IBrowse again after i closed it.

Well i finally found that bug.

I foud out that IBrowse crashes becaus of the file /cache/database/GlobalCache. When i delete it IBrowse is able to start and doesnt crash.

I beleve it's somthing to do with linuxfilesystem and missing protection bits on linux directorys.

i havnt tryed to create a hardfile to store the cachestuff on yet, i hope that'll finally fix my problems.

[kadi]

Yes i'll post that on the new mailinglist after i tryed a hardfile.

[kadi]

Hallo
So i already posted it in the ML now, but so far i didn't got the message back. I don't know if internet is just too slow or there is a problem with my mail outgoing server.

but someone els was so kind and already posted this tread to the list. So i better post my ML mail here too.


> I found a problem with a linux filesystem directory used > as a HD in UAE.

> Apparently IBrowse is unable to handle its cache correctly
> and crashes, exits and also freezes the X cursor if the
> cache is located on such a linux folder.

> But it works without any flaws if the IBrowse cache is
> located on a hardfile.

> I'm using ext3 on my Linux setup. I also tryed it with
> WinUAE (ntfs) but i could not reproduce the problem there.