OSDev.org

Hi!

here is the link:
http://www.distantvoices.org/download/b ... mg.tar.bz2

source code is as usual located on my website: www.distantvoices.org

To clarify: it should work (the tests prove the contrary) on real hardware and in VMWare. (three out of four test machines at my availability are running it without problems) It refuses to work in bochs. Need to do some testing with qemu. The gui doesn't start per se. you needto enter 'gui' at the command line in order to start it.

'been successful in doing the vfs thing - now, I'm implementing some gui widgets. Some of them are already implemented:

drawing control aka canvas (i call it doodle control, don't ask me why)
text control
button control

grid layout manager
tab layout manager (this one is cool) *gg*

well, enough of the bragging.

known bugs:
the gui console window (the one where the gui service spits out its messages) disappears. Just move the active window over and it appears again.
if running the shell in gui: the placement of the (not present) cursor is broken. It doesn't work at all. Need to do some work here.
The shell history has some issue witht he scrolling back throu entered commands. as soon as the first entry is reached and you hit the up arrow once again, you canna move away from it - neither up nor down. That's mainly because I've done this thing in the middle of the night instead of going to bed and having a nap.

edit
Have found a bug which caused the whole system to crap out, if a file grew over a certain size. well, it is solved, and of course, other issues might reveal in turn their evil faces and mock on me - but that's the os developers fate, isn't it? The sourcecode and floppy image are as usual on my web.
/edit


thanks for your feedback, criticism and occasional kicks in the butt for inefficient and crappy code. I admit, BlueIllusion is full of crappy code, not the least amongst it my malloc.c. Well, I do my best to redo it as soon as possible.

important: BlueIllusion doesn't work with the Bochs. It is a problem with the floppy driver and I canna find out what the heck it expects me to do. $%?&!!

some screenshots:
an older one:
[img height=230 width=350]http://www.distantvoices.org/bilder/bi0036_GUI_draw.bmp[/img]

from today:
[img height=230 width=350]http://www.distantvoices.org/bilder/bi0 ... ttered.bmp[/img]

Fails to boot on Bochs/Qemu.

Real hardware (Athlon-xp, nforce2 chipset) fails with the message:

fssrv: LESE Modus und kein Block gefunden ... 0 zur[sup]n[/sup]ckgeben.

My German is abominable, but I figured it might have been a disk error so I tried with two other floppies, but received the same error.

Got to the point:

fssrv: init..

then got this

fehlerhafte adresse: 53F88A08 - raus mit dem typen! 100

have no idea what it syas but I assume it is an error of some kind. But considering I can't read the language it could say anything.

Is the GUI supposed to start automatically?

Cheers.

XStream wrote: then got this

fehlerhafte adresse: 53F88A08 - raus mit dem typen! 100

My translation from german:

Error at address 53F88A08 - let's get that guy out of here! 100

Seems an error message that a page fault occured at a more or less random place (to me at least)

@XStream: Thanks! Its an unresolved pagefault and this leads to termination of the faulting process.

The gui isn't supposed to start up per se. you have to enter 'gui' at the command line to start it up. I leave it for the user to choose whether to operate in cli or gui mode.

Unfortunately, the faulting process in our case is the fs service, so the entire system craps out and canna continue.

I'm just a bit disturbed by that: I've tested it in VMWare (don't do any bochs testing) and on four computers - one out of them fails to load it and that's my toshiba laptop. Well, Laptops are weird. This can have very different reasons: one - I don't clean up memory before start up, so there can stand anything in memory. second - I miss initializing of crucial variables, and this can lead to the experienced results. third - some pointers are not initialized to NULL. *mmpf*

Thanks again for testing!

No gui starts in Bochs (2.1.1) and in QEMU (0.6.0).

I attach a ZIP file containing the screenshot of the Bochs and QEMU output and the bochs log, if they can be useful.

Greetings

Hi,

It worked well on my p1 133 mhz
Only a few things:

- I couldn't test the GUI, because of that I had VESA 1.2 (and I think your requirements are VESA 2.0, oh well....)
- Your FD parameters are incorrect, I read that there were 36 sectors per track, 2heads(correct).
With the standard of 80 cylinders this gives me 5760 sectors or 2880 kb or a 2.88 MB.
Maybe this has something todo why the FD is a bit slow....
- calc freezes up by doing 11+2*44 (don't know sure), get an error, don't know the registers anymore, have to take a look at it

Good work, only those bugs (2 and 3) could be fixed....

OT: why are you mixing English and German ? (not that I'm not able to understand it )

@DennisCGc:
why this mixture of languages? *gg* I'm kinda inconsequent concerning language. This stems from me being austrian native and speaking a few languages quite well, if not fluent (well sometimes words like sidewalk refuse to spring into my mind if I utterly need them *rofl*)

thanks for pointing out this bug with the fdc. It gives me some trouble for I don't seem to be able to get it running on the bochs. Maybe it has to do with those false parameters?

stay safe

rolling a test on my Toshiba Laptop (48MB ram). initialisation halts after "boo auf console1" is displayed ... I dunno if i should be expecting a shell right now ... typing "gui[ENTER]" did simply displayed and scrolled.

After Initialization of the File System Service it is supposed to start the shell processes. Yes, a shell can be expected.

but alas, Pype, I suffer the same syndrome if I dare to start it on my own Toshiba Laptop. *grmpf* Notonly it tells me to go away, it craps out with some nested page fault - and this at a position where it works wonderfully on any pc around me. (to say, three of them) and in VMWare. Such bugs are nasty, but I think there is one possibility: memcpyd and length not divided by 4. This to coding til midnight and becoming a moron due to fatigue.

i have not tested it, but it looks impressive!
and if i am not misstaking, you have done it i relative short time? i think that we became members almost at the same time..

relative short time ... . If it were a bit more friendly to other computers than my own test machines, I'd be happier.

Hi,

I tried it out on a few computers. Here's my results (split into 2 posts):

Crashes on Bochs. I tried it with Bochs configured for 2 80686 CPUs, 1 80686 CPU and 1 80586 CPU and it crashes the same each time:

p[CPU0 ]c0001086 >>PANIC<< exception(): 3rd (14) exception with no resolution
i[SYS ]c0001086 Last time is 1094597669
i[CPU0 ]c0001086 protected mode
i[CPU0 ]c0001086 CS.d_b = 32 bit
i[CPU0 ]c0001086 SS.d_b = 32 bit
i[CPU0 ]c0001086 | EAX=00000000 EBX=00000000 ECX=00000004 EDX=0000000c
i[CPU0 ]c0001086 | ESP=c002214c EBP=c002214c ESI=d00b8fa0 EDI=d00b8fa0
i[CPU0 ]c0001086 | IOPL=0 NV UP EI NG NZ AC PO NC
i[CPU0 ]c0001086 | SEG selector base limit G D
i[CPU0 ]c0001086 | SEG sltr(index|ti|rpl) base limit G D
i[CPU0 ]c0001086 | DS:0010( 0002| 0| 0) 00000000 000fffff 1 1
i[CPU0 ]c0001086 | ES:0010( 0002| 0| 0) 00000000 000fffff 1 1
i[CPU0 ]c0001086 | FS:0010( 0002| 0| 0) 00000000 000fffff 1 1
i[CPU0 ]c0001086 | GS:0010( 0002| 0| 0) 00000000 000fffff 1 1
i[CPU0 ]c0001086 | SS:0010( 0002| 0| 0) 00000000 000fffff 1 1
i[CPU0 ]c0001086 | CS:0008( 0001| 0| 0) 00000000 000fffff 1 1
i[CPU0 ]c0001086 | EIP=c0001086 (c0001086)
i[CPU0 ]c0001086 | CR0=0xe0000011 CR1=0x00000000 CR2=0xc0013040
i[CPU0 ]c0001086 | CR3=0x00000000 CR4=0x00000000
i[CPU0 ]c0001086 (instruction unavailable) page not present



COMPUTER G
16 Mb memory
Cyrix 6x86 PR200 CPU
1.44 Mb floppy
Sound card - creative soundblaster 16
4GB hard drive with bad cluster after first 2 Gb (roughly)
S3 Trio64V+ video card
NETGEAR FA311/FA312 PCI Ethernet card
20x IDE CDROM

Booted ok, but triple faulted when I type "help" at the command prompt. I tried it again and type "gui" - this time it displayed "VESA nicht unterstuetzt!" for a while and then repeatedly displayed "pageframe-pop - problem!!! kein freier speicher! 0" over and over and over.. The third time I typed "GUI" (in capitals this time) and it triple faulted again.



COMPUTER D
AMD-K6 3D processor 166Mhz
16Mb Memory
1.44 Mb floppy
CDROM using IDE interface - working (looks crappy)
SiS 6326 video card
NE2000 compatible ethernet card

Crashed on boot - looked like it was continuously repeating error messages (which did change occasionally). It scrolled so fast I couldn't read it properly, but it looked like a register dump.



COMPUTER H
160Mhz Cryix/Pentium
16Mb Memory
1.44 Mb floppy
514Mb Hard disk
CDrom using IDE interface
Cirrus Logic 5430/40 PCI video card
ESS ES688 Audiodrive sound card
NE2000 compatible ethernet card

Booted ok until it got to "fssrv: init...." and stopped (didn't respond to CAPS LOCK either). I waited 23 minutes for it to do something before trying again (did the same 3 times in a row).



COMPUTER N
AMD-K6(tm) 3D processor 300 Mhz
64 Mb memory
32x CD-rom
1335Mb hard disk
S3 video card
ethernet card

Triple faulted during boot. It displayed "FSSRV: LESE Modus und kein Block gefunden ... 0 zurnckgeben..." then loaded more from the boot floppy and triple faulted. I left it running so it must've done the same thing 5 times before I stopped it.



COMPUTER O
Pentium Pro 200 MHz
128 Mb RAM
Adaptec AHA-2940 SCSI controller
Compaq ST32550W SCSI hard-drive
IBM DCAS-34330W SCSI hard-drive
IBM DGHS09U SCSI hard-drive
Sony CDU-76S SCSI CD-ROM
Ethernet on motherboard

Looked like it booted OK, but refused to respond to the keyboard (tried it 3 times).


{continued}

{continued}

COMPUTER P
Dual Pentium III 1 GHz
512 Mb RAM
Adaptec AIC-7899 dual-channel SCSI on motherboard
2 * 9 Gb Quantum ATLAS SCSI hard-drives
IDE CD-ROM (50x generic)
Ethernet on motherboard plus 2 PCI ethernet cards

Boot OK. Typing "help" worked the same as Computer Q. During boot it said "ATI Video Device found. Not yet supported!!", but I typed "gui" anyway. It tried to work, but the screen was all messed up (looked like it was on LSD). I could see a blue vertical line that moved with the mouse, and I could click on the pink/purple squares, which changed to yellow squares (like they had become the active window).

Because this computer seemed to work I started looking for more commands to try. I tried "type" (as *nix "cat" wasn't there), but when I press 'Y' it displays a 'Z' (and "tzpe" wasn't a valid command). Then I tried "ls" and it worked and gave me a list of stuff to try. I tried "calc" and it displayed the usage, so I started pressing the cursor up/down keys to get the command back. This was working, but somehow I ended up line saying "da hinten sind wir vetzt - vorm free ..." and it stopped responding to the keyboard (although CAPS LOCK did change the LED, so it wasn't completely locked up).

I rebooted and tried again, typing "80*50*2+123", but it displayed "80(50(2`123" and gave me an error about mismatched parenthisis. I tried "1/2" which displayed as "1-2" and returned to the prompt without displaying anything. After having several keys messed up I typed "abcdefghijklmnopqrstuvwxyz", which displayed as "abcdefghijklmnopqrstuvwxzy" (correct apart from 'y' <> 'z'). Then I tried "1234567890!@#$%^&*()", which displayed as "1234567890!"$%&/()=", where the '#' didn't display anything. Not sure if your OS is expecting a german keyboard instead of my english/US keyboard.

Anyway, I though I'd try "semtest" after this, which gave me usage instructions. I tried "semtest -c l", "semtest -t 1" and "semtest -d 1" (by pressing '/' to get the '-') which all worked fine. Then I pressed the cursor keys again (habit) and locked it up with the "da hinten sind wir vetzt - vorm free ..." thing again.

After rebooting I tried out "argvtest" with a few different combinations of arguments, which worked well. Then I tried "rechner", which locked it up without displaying anything (it did respond to CAPS LOCK though, so it wasn't completely locked up).

The commands "ps" and "ls" always worked well.



COMPUTER Q (Compaq Proliant 1600)
Dual Pentium II 400 MHz
384 Mb RAM
Dual-channel SCSI on motherboard
5 * 4.5 Gb SCSI hard-drives
IDE CD-ROM
Ethernet on motherboard

Booted OK. Typed "help" and it said "available commands:" and "press any key to continue" then returned to the prompt. Then I typed "gui" and it displayed "VESA nicht unterstuetzt!" and locked up. I pressed the CAPS LOCK key to see how locked up it was and the LED came on, but pressing CAPS LOCK after that didn't do anything, so I guess it wasn't completely locked up until I pressed a key.

I rebooted, and tried out "argvtest" with a few different combinations of arguments. Somehow I got that "da hinten sind wir vetzt - vorm free ..." while I was typing (I don't think I touched the cursor keys) and it locked up. The "rechner" command locked it up same as before.

The commands "ps" and "ls" always worked well.



WORK MACHINE
Pentium IV 1.66 Ghz
256 Mb RAM
40 Gb hard drive
8 Gb hard drive
NVIDEA RIVA TNT2 Model64 Pro video card
NETGEAR FA311/FA312 PCI Ethernet card

Triple faulted during boot. It displayed "FSSRV: LESE Modus und kein Block gefunden ... 0 zurnckgeben..." then loaded more from the boot floppy and triple faulted. Tried it a few times..



Cheers,

Brendan

Man, Brendan, thats a cool bunch of tests! thanks!

Your tests remind me of something: I should build something into the GUI applications to test whether the Gui service is active or not - and have them exit with a nice comment. DOnt want to fire up gui, just because the app is started by error.

Current gui appz: rechner,guitest,gui_starter.bin,launcher

As I seem to have uploaded a debugging snapshot, the help program is mutilated. It should display also some stuff to try. *gg*.

the cat program will come soon, I promise. I just need to implement writing file system access (for memory mapped files too) prior to that. In the meantime, there's a subset of 'more' available.