February 2009 - Oberon - mailman3.inf.ethz.ch

NO & A2 bootSerialTrace for MailList archive.
by Chris Glur 28 May '09

28 May '09

A previous mail described how the bootSerialTrace for NO workED. For my ver: 2001-beta; only the 'trace to own screen' worked - for me. ----------- Date: Tue, 5 Feb 2002 10:50:13 +0100 (MET) From: fischer(a)inf.ethz.ch Subject: Problems with Trace setting stored permanently (Native Alpha 11-Oct-2001) .. Ran these tests with 2 machines attached via null-modem cable. <etc. everything confirmed OK.> ============= For A2 from a CD the bootSerialTraces with 3 different IDEs are given below, as being possibley usefull for other collaborators. I'll try to debug them later, since the documentation seems quite substantial. Thanks, == Chris Glur. ============ Mon Feb 23 15:12:21 SAST 2009 ========= Machine: A2 Revision 1560 (27.08.2008)ult, you have to set them Machine: Detecting memory... 32MB Machine: 1 cores per physical package, 1 threads per core. Machine: Enabling MMU... Ok Heaps: Initializing heap... Ok Machine: Single-processor Kernel: Initialized and started. {P cpuid= 0, pid= -1 PCI: 1 bus(ses) found, PCI version: 02.10} ATADisks: Force PIO mode for ATAPI devices ATADisks: Scanning PCI bus for IDE & SATA class devices ... ATADisks: Found PCI device on bus 0, device 7, function 1 ATADisks: Adding controller 01F0, 03F0, E800, IRQ: 14: IDE0..1, Bus-master enabled ATADisks: Adding controller 0170, 0370, E808, IRQ: 15: IDE2..3, Bus-master enabled ATADisks: Detected devices: IDE0: 38166MB, LBA, ATA/ATAPI-5, MW DMA 2, ST340016A, ver 5.0 IDE3: ATAPI cd-rom device, removable, no DMA, CD-956E/AKV DiskFS: Index not found on RAM0 DiskFS: Scanning RAM0... marking 0 files DiskFS: 31992K of 32000K available on RAM0 DiskVolumes: IDE0#0 invalid partition <------ ?? NO handles it but not DOS !! DiskVolumes: IDE3#0 .[1] TRAP 8 PL 3 ASSERT failed A2 Revision 1560 (27.08.2008) <-- CD @ hdd ?? CS:=0000001B DS:=00000020 ES:=00000020 SS:=0000002B CR0=80040031 FPU=00000000 PC=0067FDD0 ESI=0067FB01 EDI=8411FF05 SP=8411FF2C CR2=8411DFFC PID=00000000 EAX=00180000 EBX=00000117 ECX=00005400 EDX=8411FF34 CR3=01FFE000 LCK=00000000 BP=8411FF5C FS:=00000000 GS:=00000000 ERR=00000000 CR4=00000000 TMR=000026ED FLAGS: P#>_+I +3 {1..2, 9, 12..13, 21} Process: 7 run 0 30067CF90:Commands.Runner ATADisks.Controller.ProtPacketPIO pc=6056 [000017A8H] {0, 28} DisplayLinear.Init pc=568 [00000238H] w=1024 h=768 f=2 mem=2097152 ( 200000H) res=0 ts="" padr=1179927 ( 120117H) vadr=0 State DisplayLinear: @Self=0067F910H d=00000000H DisplayLinear.$$ pc=8 [00000008H] @Self=0067F910H d=00000000H Modules.ThisModule pc=1938 [00000792H] name="DisplayLinear" res=0 msg=""... @PtrReturnType=00000000H m=0067F910H p=0067F910H fileName="DisplayLinear.Obx" body=DisplayLinear.$$ pc=0 [00000000H] new=TRUE i=1 Commands.Runner.@Body pc=671 [0000029FH] @Self=0067CF90H --------------------------------- {P cpuid= 0, pid= -1 Exception during command execution} <---------?? DriverDatabase: Enabled driver lookup service. {P cpuid= 0, pid= 13 Usb: USB driver loaded.} <---- old-box has no USB-BIOS ?? Usb: Driver UsbHub (USB Hub Driver) has been added. Default font installedOpenType 0.45 BBPort /3.12.2004 eos, pl Online 00:11 ============ end of trace ================== ?? Remove the USB info & Try a DOS-able IDE == Machine: A2 Revision 1560 (27.08.2008)ult, you have to set them Machine: Detecting memory... 32MB Machine: 1 cores per physical package, 1 threads per core. Machine: Enabling MMU... Ok Heaps: Initializing heap... Ok Machine: Single-processor Kernel: Initialized and started. {P cpuid= 0, pid= -1 PCI: 1 bus(ses) found, PCI version: 02.10} ATADisks: Force PIO mode for ATAPI devices Machine: A2 Revision 1560 (27.08.2008) Machine: Detecting memory... 32MB Machine: 1 cores per physical package, 1 threads per core. Machine: Enabling MMU... Ok Heaps: Initializing heap... Ok Machine: Single-processor Kernel: Initialized and started. {P cpuid= 0, pid= -1 PCI: 1 bus(ses) found, PCI version: 02.10} ATADisks: Force PIO mode for ATAPI devices ATADisks: Scanning PCI bus for IDE & SATA class devices ... ATADisks: Found PCI device on bus 0, device 7, function 1 ATADisks: Adding controller 01F0, 03F0, E800, IRQ: 14: IDE0..1, Bus-master enabled ATADisks: Adding controller 0170, 0370, E808, IRQ: 15: IDE2..3, Bus-master enabled ATADisks: Detected devices: IDE0: 8063MB, LBA, ATA/ATAPI-4, MW DMA 2, WDC WD84AA, ver 4.0 IDE3: ATAPI cd-rom device, removable, no DMA, CD-956E/AKV DiskFS: Index not found on RAM0 DiskFS: Scanning RAM0... marking 0 files DiskFS: 31992K of 32000K available on RAM0 DiskVolumes: IDE0#0 invalid partition <--------------- ?? DiskVolumes: IDE3#0 .[1] TRAP 8 PL 3 ASSERT failed A2 Revision 1560 (27.08.2008) CS:=0000001B DS:=00000020 ES:=00000020 SS:=0000002B CR0=80040031 FPU=00000000 PC=00680350 ESI=00680081 EDI=8411FF05 SP=8411FF2C CR2=8411DFFC PID=00000000 EAX=00180000 EBX=00000117 ECX=00005400 EDX=8411FF34 CR3=01FFE000 LCK=00000000 BP=8411FF5C FS:=00000000 GS:=00000000 ERR=00000000 CR4=00000000 TMR=00001EB1 FLAGS: P#>_+I +3 {1..2, 9, 12..13, 21} Process: 7 run 0 30067D510:Commands.Runner ATADisks.Controller.ProtPacketPIO pc=6056 [000017A8H] {0, 28} DisplayLinear.Init pc=568 [00000238H] w=1024 h=768 f=2 mem=2097152 ( 200000H) res=0 ts="" padr=1179927 ( 120117H) vadr=0 State DisplayLinear: @Self=0067FE90H d=00000000H DisplayLinear.$$ pc=8 [00000008H] @Self=0067FE90H d=00000000H Modules.ThisModule pc=1938 [00000792H] name="DisplayLinear" res=0 msg=""... @PtrReturnType=00000000H m=0067FE90H p=0067FE90H fileName="DisplayLinear.Obx" body=DisplayLinear.$$ pc=0 [00000000H] new=TRUE i=1 Commands.Runner.@Body pc=671 [0000029FH] @Self=0067D510H --------------------------------- {P cpuid= 0, pid= -1 Exception during command execution} DriverDatabase: Enabled driver lookup service. Default font installedOpenType 0.45 BBPort /3.12.2004 eos, pl ================= Try a small old DOSable-IDE == tested to boot DOS .. ATADisks: Adding controller 01F0, 03F0, E800, IRQ: 14: IDE0..1, Bus-master enabled, LBA not supported, DMA nod ATADisks: InitDevice, res: 0 ATADisks: Adding controller 0170, 0370, E808, IRQ: 15: IDE2..3, Bus-master enabled ATADisks: Detected devices: IDE0: 40MB, 980*5*17, no DMA, ST351A/X IDE3: ATAPI cd-rom device, removable, no DMA, CD-956E/AKV DiskFS: Index not found on RAM0 DiskFS: Scanning RAM0... marking 0 files DiskFS: 31992K of 32000K available on RAM0 IDE0 reset done DiskVolumes: IDE0#0 error 1 <---- ?? DiskVolumes: IDE3#0 .[1] TRAP 8 PL 3 ASSERT failed A2 Revision 1560 (27.08.2008) <---?? CD-ROM CS:=0000001B DS:=00000020 ES:=00000020 SS:=0000002B CR0=80040031 FPU=00000000 PC=0067F530 ESI=0067F261 EDI=8411FF05 SP=8411FF2C CR2=8411DFFC PID=00000000 EAX=00180000 EBX=00000117 ECX=00005400 EDX=8411FF34 CR3=01FFE000 LCK=00000000 BP=8411FF5C FS:=00000000 GS:=00000000 ERR=00000000 CR4=00000000 TMR=0000BE51 FLAGS: P#>_+I +3 {1..2, 9, 12..13, 21} Process: 7 run 0 30067CA70:Commands.Runner ATADisks.Controller.ProtPacketPIO pc=6056 [000017A8H] {0, 28} .. ======================

2 1

Re. N-O/Aos boot tracing facility ? ++
by Chris Glur 22 Feb '09

22 Feb '09

> Was the boot tracing capabilty continued > on all N-O versions and also Aos ? Of course now I realise that if the 2 args: speed & portNo still exist, then the function using these args does too. For my N-O box: = N-O / V24 sends OK to my LEO [linux ETH Obn] box, at the recommended parameters for the Oberon boot tracing facility - using linux /minicom. = <left shift> instead of the documented <ScrollLock> interrupts booting, to enabling the boot-args. = I use the change-boot-args whenever I change my boot partition, so that's confirmed ok. If I can get some debug-info, perhaps I can decode the correct InitString to CD-boot Aos ? Thanks, == Chris Glur.

2 1

N-O/Aos boot tracing facility ?
by Chris Glur 22 Feb '09

22 Feb '09

I'm investigating Aos installation - with 'wrong InitString error'. The obvious next step is the 'boot tracing facility' via <shift> during boot. I must have done this, for N-O, because I can remember the trace output ser-TXed to the other box, and I've got got mail-list correspondence related. I even [? imagine?] I confimed the BPStrace=-1 for output to local console. My current N-O version is 2001-Beta, and I'm wondering if I did/saw the boot tracing on an earlier version. Was the boot tracing capabilty continued on all N-O versions and also Aos ? Thanks, == Chris Glur.

1 0

Re. NO on a Compaq DeskPro EN
by Chris Glur 16 Feb '09

16 Feb '09

> "Did you try NO before you changed to ROMPaq ?" > > Yes. I had to put a hdd into the machine. > At the initial attempt to boot Oberon0, the > bios or ROM gave a message to boot from the > ROMPaq diskette and update the ROM. After > that, Oberon0 works normally. The only > problem visible is the failure of keyboard > and mouse after booting from the hdd. > > If there are no ideas here I'll try asking > in a hardware forum. Yes, Newsgroups have degenerated by being crowded-out by blogs. While you're waiting for an answer, you could linux: lynx -dump 'http://www.google.com/search?q=ROMPaq+keyboard' > ROMPaq etc........ or the NO equivalent. == Chris Glur.

1 0

Re: [Oberon] Re. NO on a Compaq DeskPro EN
by peasthope＠shaw.ca 16 Feb '09

16 Feb '09

Chris, "Did you try NO before you changed to ROMPaq ?" Yes. I had to put a hdd into the machine. At the initial attempt to boot Oberon0, the bios or ROM gave a message to boot from the ROMPaq diskette and update the ROM. After that, Oberon0 works normally. The only problem visible is the failure of keyboard and mouse after booting from the hdd. If there are no ideas here I'll try asking in a hardware forum. Thanks, ... Peter E. -- http://members.shaw.ca/peasthope/ http://carnot.yi.org/ = http://carnot.pathology.ubc.ca/

1 0

Re. NO on a Compaq DeskPro EN
by Chris Glur 16 Feb '09

16 Feb '09

Peter wrote: > Recently I installed a drive in a Compaq > DeskPro EN, updated the ROM with the ROMPaq > diskette, partitioned the drive with fdisk, > installed DOS on the first partition and > installed NO onto the second partition > using the Oberon0 diskette. Nothing in > the installations was remarkable. > > When the system starts NO from the hard > drive, the keyboard and mouse are not > operational. The display is normal. > The keyboard works properly in DOS. > Haven't tried A2 yet. > > Any ideas to fix this? AFAIK 'PC" was/is a standard [initiated by IBM & then 'taken over' by M$]. And the initial BIOS [in ROM] had the basic device drivers, eg. keyboard and VDU. Then later mouse drivers -- in software rather than firmware [ROM] -- probably used the serial driver of the BIOS. I've got a Compaq old notebook which has some non-standard features, so I suspect that your ROMPaq also deviates from the PC-BIOS. OTOH I doubt that your box would then run standard DOS. Did you try NO before you changed to ROMPaq ? == Chris Glur.

1 0

linux oberon Display update interruped ?
by Chris Glur 16 Feb '09

16 Feb '09

For [LEO] linux ETH Oberon when I do an inet-fetch: email or http, and while I wait for it to complete, I 'switch' to another desktop/task; the S3/V4 oberons screen is blank when I switch back to view it, until the Oberon task is completed. I'm guessing that the screen-write/update is suspended while the inet-fetch is busy. Under S3 the inet-task apparently eventually times out, but V4 seems not to. So under V4 I've lost control and need to re-start V4, and lose my 'setup'. S3's 'threaded-mode' version seems to do the same. Is the Oberon OS's inner 'time-slice looping through all tasks' adjusted for eg. inet-fetches, like disabling an interrupt ? Or is it that the display is only updated if the 'display buffer is changed' and nothing is written to the buffer, while it's waiting to complete the inet-fetch ? A mouse-operation would write to the display-buffer, But I know that mouse operations are disabled during inet-fetch. How is the mouse disabled ? Is the 'inner-loop' diabled until the inet-job returns OR a time-out occurs ? What would be a possible work-around to force mouse and keyboard acces ? AFAIK N-O could use the keybrd interrupt: Ctl/<Break>, but LEO can't. == Chris Glur. PS. OK after several minutes both S3 & V4 'time out' and the I/O is again usable.

2 1

NO on a Compaq DeskPro EN
by peasthope＠shaw.ca 15 Feb '09

15 Feb '09

Recently I installed a drive in a Compaq DeskPro EN, updated the ROM with the ROMPaq diskette, partitioned the drive with fdisk, installed DOS on the first partition and installed NO onto the second partition using the Oberon0 diskette. Nothing in the installations was remarkable. When the system starts NO from the hard drive, the keyboard and mouse are not operational. The display is normal. The keyboard works properly in DOS. Haven't tried A2 yet. Any ideas to fix this? Thanks, ... Peter E. -- http://members.shaw.ca/peasthope/ http://carnot.yi.org/ = http://carnot.pathology.ubc.ca/

1 0

A2 System Size (Stauber Sven Philipp)
by Paul Reed 13 Feb '09

13 Feb '09

Hooray! ;-) > From: "Stauber Sven Philipp" <sven.stauber(a)inf.ethz.ch> > Subject: [Oberon] A2 System Size > The absolutely minimal A2 system that currently can be built (the linker > expects BootConsole) includes the following modules: > > Trace, Machine, Heaps, Modules, Objects, Kernel, KernelLog, Streams, > Reflection, Traps, Commands, Files, Loader and BootConsole > > > > System size: 299008 bytes (statically linked boot image) > > > > Whatever else modules you include is application-specific. > > > > Best, > > Sven Stauber

1 0

A2 System Size
by Stauber Sven Philipp 11 Feb '09

11 Feb '09

Dear all, On http://www.bluebottle.ethz.ch/download.html, a new mini A2 CD image is available. Size: 27 MB (uncompressed), 6.7 MB (zip). Please note that this is just a mini - not minimal - system. Size information (object file size / source code size): Kernel: 129 KB / 344 KB System: 819 KB / 1.29 MB Drivers: 897 KB / 1.95MB ApplicationMini: 2.21 MB / 3.91 MB GuiApplicationMini: 1.98 MB / 2.49 MB Build: 130 KB / 224 KB Total (approx.): 6.1 MB / 10.2 MB Resource/data files: 10.7 MB The absolutely minimal A2 system that currently can be built (the linker expects BootConsole) includes the following modules: Trace, Machine, Heaps, Modules, Objects, Kernel, KernelLog, Streams, Reflection, Traps, Commands, Files, Loader and BootConsole System size: 299008 bytes (statically linked boot image) Whatever else modules you include is application-specific. Best, Sven Stauber

1 0