@Framiga

Repulse is not compatible. IIRC Xeron tried it and it's listed as not compatible since the list appeared.

_________________
The only spanish amiga news web page/club: CUAZ

@Spirantho

uh?? Highway works quite well... I tested it.

Do you have the "Romulus" add-on? if you use a Romulus add on simply disable it presing right or left mouse button (it depends on your config). Maybe you should update Luciferin "BootModule".

_________________
The only spanish amiga news web page/club: CUAZ

So here some questions (and answers, after some tweaking) regarding my personal experience while trying to install the OS4 on a classic Amiga.

The Amiga is an A4000D with X-surf, Algor and CV3D. I`ve also a 2,1GB Seagate HD & LG CDR on the mobo IDE, and 4GB Fujitsu HD on the CSPPC SCSI.

The CD should do everything, in case you have any problems the floppy mounts and loads the kickstart. An easy installation I`d say...

The first problem was at the very begining, when I tried to format/partition and finally save the contents to the SCSI disk : I`ve reserved a partition especially for the OS4 there, the first 100MB were OS3.9, the last 1,5 as WORK and the in between frees space as OS4 partitions. The contents wouldn`t just be saved there (the SCSI RDB)! One temporary/quick overcome to this problem was to transfer my IDE Programs: partition to the SCSI reserved space, and create the OS partitions to the IDE disk. It worked fine!

Next issue, was the network : after running the wizard, the installed card wouldn`t respond to the test button! It appeared ok and functional at the boot menu, and after running the ShowConfig The fix : the device was missing! I`ve copied the 68K x-surf.device to the devs:networks/
After setting the DNS servers the gateway etc. everything run smoothly... If only I could use the USB wheel mouse!

Next step, the USB stack : pretty straightforward, thanks to petunia (kudos to the OS4 team!), I`ve installed the latest Poseidon (the best USB stack), and used the FreeWheel utility (68K binary). Don`t forget to mount a FAT32 FS system, in order to read those PC formatted sticks!

Now what`s left?
- That kickstart ROM modules were to be loaded on a Kickflash eprom in order to save boot time and warm resets... How could it be done this with the Algor Flash memory? Any ideas (what files are used, which ones are to be copied etc.)

- The SCSI partitions : still won`t write to these (not even save an .info file , like snapshot, or even format a partition!). The strangest thing is that I can do anything to that disk from the 3.x side!
As a hint, I`ve tried months ago to install NetBSD to that disk , but I got the screen flooded with the message:
Quote:

sd0(cbiiisc0:0:0:0): Check Condition on CDB: 0x2a 00 00 4d 00 hex 00 00 hex 00 {where hex, various numbers on each repeated line, e.g. d8/30/10/78} SENSE KEY: Aborted Command ASC/ASCQ: AS 0x47 ASCQ 0xa1

I thought the disk was faulty, and I`ve posted to the NetBSD ML. The said that it might had been the cabling, or the terminators. That wouldn`t explain the faultless use through OS 3.x . Another suggetsion was about" an old Fujitsu disk that reports 0x47/0xa1 for a SCSI phase error in synchronous transfer mode." As a matter of fact it is an old Fujitsu disk, and I tried through the CSPPC flash rom to set the SCSI in asynchronous mode, but no luck...

Any ideas are welcome

_________________
Ordell Robbie: Is she dead, yes or no?
Louis: Pretty much.

@Cass

Quote:

Cass wrote: - That kickstart ROM modules were to be loaded on a Kickflash eprom in order to save boot time and warm resets... How could it be done this with the Algor Flash memory? Any ideas (what files are used, which ones are to be copied etc.)

Documentation/C/BootLoader doc explains which files are loaded etc. However,
BootLoader will always load the OS4 kickstart files from disk, and they would probably
not fit into the flash memory anyway.

Quote:

- The SCSI partitions : still won`t write to these (not even save an .info file , like snapshot, or even format a partition!). The strangest thing is that I can do anything to that disk from the 3.x side! As a hint, I`ve tried months ago to install NetBSD to that disk , but I got the screen flooded with the message: Quote: sd0(cbiiisc0:0:0:0): Check Condition on CDB: 0x2a 00 00 4d 00 hex 00 00 hex 00 {where hex, various numbers on each repeated line, e.g. d8/30/10/78} SENSE KEY: Aborted Command ASC/ASCQ: AS 0x47 ASCQ 0xa1 I thought the disk was faulty, and I`ve posted to the NetBSD ML. The said that it might had been the cabling, or the terminators. That wouldn`t explain the faultless use through OS 3.x . Another suggetsion was about" an old Fujitsu disk that reports 0x47/0xa1 for a SCSI phase error in synchronous transfer mode." As a matter of fact it is an old Fujitsu disk, and I tried through the CSPPC flash rom to set the SCSI in asynchronous mode, but no luck...

The OS4 cybppc.device contains code from NetBSD so its no surprise when it
shows similar behaviour. The device does not read the CSPPC FlashROM SCSI
settings so you cant force switching it to async mode.

You can verify that your SCSI chain has an active Terminator at each end of the
cable, and that the drive is jumpered to deliver TermPower to the SCSI cable which
is needed by the terminators. The CSPPC has no termination and does not deliver
TermPower.

@Crumb

Quote:

Crumb wrote: @Spirantho uh?? Highway works quite well... I tested it. Do you have the "Romulus" add-on? if you use a Romulus add on simply disable it presing right or left mouse button (it depends on your config). Maybe you should update Luciferin "BootModule".

Nope, don't have any add-ons. Just the Highway which works perfectly under OS 3.9.

It also - as you say - works fine under OS 4 - at least until you do a lot of hard disk access.
No Highway in the machine? Works perfectly.
Highway in the machine but no drivers loaded? Works perfectly.
Highway in the machine with highwayusb.kmod (or whatever it's called)? Crashes solid whenever there's heavy disk access. Particularly formatting a large disk is easiest to reproduce.

Tested it quite a bit now. Unfortuantely OS 4 doesn't support Poseidon (it needs to patch the input.device) so I can't use that. Even tried that mini-USB stack for 68000 Amigas (ANAIIS) but that didn't work.

Hello.
I have a working X-surf (the one with ide and clockport ports) in my amiga4000
In os3.9 works fine but in the 4.0 side freezes the system every time that i transfer some bits.
It works fine the first packets but in a few seconds completelly freezes the system.
Its a 100% clean install.
I use the x-surf.device V1.2 filesize 13448 (040 version) as sugested in the compatibility list of acube.
Any idea?
I've tryed to blacklist it on the prefs/compatibility but I only get 2 more seconds of transfers.
Of course its with the roadshow stack.

Thanks and sorry for my bad english.

I got my classic OS4 copy last weekend....a few things I noticed (4000T):

1) MediaToolbox does not seem to scan for ID15...my CYPPC SCSI HD was not recognized and so had to change the ID to something else....my controller is set to ID7.

Basically, as it is scanning it counts from 1-14 of 15 but never shows 15/15.
(seems like a while/for loop bug)....

2) I have a AriadneII ethernet card -- was thinking that all devices listed in the
new connection wizard had device drivers installed. It shows a bunch of possible cards...
but my ariadneII was not found .... because those drivers in the list was not in devs:networks...anyways was able to pull out my Villiage Tronic floppy disk to get the driver...and then it worked like a charm....

3) Would like to see SCSI driver support for the A4000T .. I have a bunch of scsi stuff
(scanner, tape drive -- external and dvd drive -- internal) that are the old 50-pin fast/slow scsi-2/1...and don't want to slow down the CYPCC SCSI bus...
IDE has always been a problem on the A4000T --- seems to take forever to cold boot when I had a IDE CD-ROM connected..or worse got gurus from some IDE devices...many threads on the net about this...

http://groups.google.com/group/comp.sys.amiga.hardware/browse_thread/thread/d39eb888b4b14e50/cd7baab5199b57e4?lnk=st&q=IDE+A4000T+guru#cd7baab5199b57e4

@my_pc_is_amiga

Can't comment on your Number 1., but as for the Ariadne.device driver I believe it could not be supplied with OS4.0 because it is owned by a 3rd party. Copying the driver from your original floppy supplied with the Ariadne card is the way to go.

I also have an A4000T and have tried many different interface settings (while Beta testing). The A4000T onboard scsi does not work and is a know problem. The onboard IDE performs well and does not affect boot up as long as you have a HDD with a CD drive. The SCSI i/f on the CS PPC also works well with the A4000T combination. The main problem is getting the correct termination with older narrow scsi devices on the CS SCSI bus. But once setup again works well.

My set up includes both CS SCSI HDD & DVD-RW and ide HDD & CD-ROM.

As for the X-surf device the standard OS3.9 deriver is not compatible with OS4.0 (or MorphOS PowerUp). I Understand that an OS4.0 native driver is being worked on and should be available soon? In the meantime there is a patched x-surf device available as a temporary fix. Again this is owned by a 3rd party.

Good luck anyway.

TrevorDick

Last edited by TrevorDick on 26-Dec-2007 at 10:26 AM.

_________________
No, I don't need no reason, I'm just breezin'

@doraemon

Quote:

I use the x-surf.device V1.2 filesize 13448 (040 version) as sugested in the compatibility list of acube. Any idea?

See my posting #188 above.

TrevorDick

_________________
No, I don't need no reason, I'm just breezin'

@TrevorDick

i have XSurf up and running with the XSurf 68K driver patch i use for MOS.

The only problem is how getting it (legally or better, respecting the not-public state of it).

Contact me privately (my address is in my profile) and we will see how to manage this.

Edit- shameless plug ... does anybody is in touch with Aliedesign?

Edit2- the mails to Aliendesign are bounced

Last edited by Framiga on 26-Dec-2007 at 11:13 AM.
Last edited by Framiga on 26-Dec-2007 at 10:37 AM.

_________________

@TrevorDick

Thanks for the advice...I only have a CDROM connected to the IDE port...

For the Internet "new connection" ... I understand but would be nice to have something that tells you to install the 3rd party drivers for those users (like me) that didn't know there wasn't any native ones.... or at least show in the *.db list which dirvers are installed and which would need to be installed.....(i.e. like with a check mark or something but each item)....

Well, Now im trying to use my Algor USB Pro in OS4.
It make nothing. The usb devices don't get power.
The only config that i can found is sys/prefs/usb but seems very small.
I think that I need to configure the algorusb.device in some place, no?
Its supported by the built-in stack or I need to install the poseidon-68k ?

A lot of thanks!

@TrevorDick

Now my A4000 is up and running with OS4 and xsurf :)
Thanks!

@Spirantho

Quote:

Unfortuantely OS 4 doesn't support Poseidon (it needs to patch the input.device)

IIRC Poseidon works on OS4 as long as you don't load Sirion. Poseidon can work without patching input.device but you probably won't be able to use usb keyboards/mice

_________________
The only spanish amiga news web page/club: CUAZ

@TetiSoft

Quote:

The OS4 cybppc.device contains code from NetBSD so its no surprise when it shows similar behaviour. The device does not read the CSPPC FlashROM SCSI settings so you cant force switching it to async mode.

That makes sense!

Quote:

You can verify that your SCSI chain has an active Terminator at each end of the cable, and that the drive is jumpered to deliver TermPower to the SCSI cable which is needed by the terminators. The CSPPC has no termination and does not deliver TermPower.

The last time I checked, the cabling was Terminator---CSPPC---Fujitsu
When shipped from the factory, the drive is terminated. I left it as default setting (nothing changed regarding the jumpers). If this is working faultless on 3.9 (read/write), does it make any sense to have such errors on 4.0?

Follows a report of the device ( q-device):
Quote:

Driver: cybppc.device Unit: 0 Vendor: FUJITSU Product: M2954Q-512 Revision: 0153 Type: DIRECT ACCESS (eg Disk) Removable Media: No Vendor Type Modifier: 0 ISO Version: 0 ECMA Version: 0 ANSI Version: 2 (SCSI Version) AENC Support: No TrmIOP Support: No Relative Addressing: No 32bit Support: No 16bit Support Yes 8bit Support Yes Synchronous Support: Yes Linked Support: Yes Command Queuing Support: Yes Supported Reset Type: Hard ------------------------------------------------------ Mode Sense Cmdsize: 10 bytes Mode Data Length: 144 bytes Inserted Medium Type: Default Write Protected: No Descriptor Length: 8 bytes Density Code: Reserved or vendor specific Number of Blocks: 8498506 Block Length: 512 bytes =z Information Page: Read/Write Error Recovery Page Code: $1 Save to NVRAM Supported: Yes Page Length: 10 bytes Automatic Write Reallocation: No Automatic Read Reallocation: No Transfer Block (TB): Transfer erroneous block to initiator Read Continuous (RC): Tolerate error recovery delays Enable Early Recovery (EER): Always use fastest error recovery method Post Recoverable Error (PER): Don't report recovered errors Disable Transfer on Error (DTE): Don't terminate data phase on error Disable Correction (DCR): Use error correction codes Retries on Read Error: 55 Correction Span: 41 bits Head Offset Count: 0 Data Strobe Offset Count: 0 Retries on Write Error: 0 Recovery Time Limit: 0 ms Έ=z΄ Information Page: Disconnection and Reconnection Parameters Page Code: $2 Save to NVRAM Supported: Yes Page Length: 14 bytes Buffer Full Ratio: 128 Buffer Empty Ratio: 128 Bus Inactivity Limit: 50 *100 us Disconnect Time Limit: 0 *100 us Connect Time Limit: 0 *100 us Maximum Burst Size: 0 *512 bytes Data Transfer Disconnect Control: Not Used Information Page: Formatting Parameters Page Code: $3 Save to NVRAM supported: Yes Page Length: 22 bytes Tracks Per Zone: 9 Alternate Sectors Per Zone: 11 Alternate Tracks Per Zone: 0 Alternate Tracks Per Logical Unit: 9 Sectors Per Track: 186 Bytes Per Physical Sector: 512 Interleave: 1 Track Skew Factor: 36 sectors Cylinder Skew Factor: 36 sectors Supports Soft Sector Formatting: No Supports Hard Sector Formatting: Yes Supports Removable Media: No Surface Bit: Off ΄=z° Information Page: Rigid Disk Geometry Parameters Page Code: $4 Save to NVRAM supported: Yes Page Length: 22 bytes Cylinders: 5714 Heads: 9 Start Write PreComp: 0 Start Reduced Write: 0 Step Rate: Device Default Landing Cylinder: 0 Rotational Offset: 0 Rotation Rate: 7200 rpm =z” Information Page: Verify Error Recovery Page Code: $7 Save to NVRAM supported: Yes Page Length: 10 bytes Enable Early Recovery (EER): Always use fastest error recovery method Post Recoverable Error (PER): Don't report recovered errors Disable Transfer on Error (DTE): Don't terminate data phase on error Disable Correction (DCR): Use error correction codes Verify Retries: 55 Verify Correction Span: 41 bits Verify Recovery Time Limit: 0 ms ΄=z° Information Page: Cache Page Code: $8 Save to NVRAM supported: Yes Page Length: 18 bytes Read Cache: Enabled Multiplication Factor: Min/Max pre-fetch in terms of logical blocks Write Cache: Disabled or Unavailable Demand Read Retention Priority: 0 Write Retention Priority: 0 Disable Pre-Fetch Transfer Length: 65535 blocks Min Pre-Fetch: 0 bytes/scalar Max Pre-Fetch: 249 bytes/scalar Max Pre-Fetch Ceiling: 65535 blocks p=z” Information Page: Control Mode Page Code: $A Save to NVRAM supported: Yes Page Length: 6 bytes Report Log Exceptions: Yes Queuing Algorithm: Restricted reordering Queue Error Management: Resume after contingent allegiance Queuing: Enabled Extended Contingent Allegiance: Disabled Ready AEN Permission: No AEN on init sequence completion Unit Attention AEN Permission: Will not issue AEN instead of UAC Error AEN Permission: Will not issue AEN on DEC detection AEN Holdoff Period: 0 ms Θ=zΔ Information Page: Extended Error Recovery Warning: Page format may be device specific. Info may be incorrect Reference: FUJITSU MAB 3091SC, MAE 3182LC, MAG 3091L Page Code: $21 Save to NVRAM supported: Yes Page Length: 2 bytes Retry Count on Seek Error: 15 ?=z” Information Page: Vendor Specific Page Code: $22 Page Length: 2 bytes ?=z” Information Page: Vendor Specific Page Code: $38 Page Length: 4 bytes

@doraemon
Poseidon works fine with Algor

Last edited by Cass on 29-Dec-2007 at 05:01 PM.

_________________
Ordell Robbie: Is she dead, yes or no?
Louis: Pretty much.

@Cass

stupid question (perhaps), what filesystem are you using?

And about setting the Syncron flag in the RDB, use RDBFlags from Aminet

5.AOS4:C> RDBFlags DEVICE cybppc.device 0
device "cybppc.device", unit 0
sector size is 512 bytes
found RDB at sector 2
drive FUJITSU MAB3091SP 0106

last disk: no
last lun: yes
last target: no
reselection: yes
synchronous: yes

_________________

@Framiga

FFS
Here my settings:
Quote:

5.OS4:> "RAM Disk:RDBFlags" device "scsi.device", unit 0 sector size is 512 bytes found RDB at sector 0 drive ST32122A 0.42 last disk: yes last lun: yes last target: yes reselection: yes synchronous: no 5.OS4:> "RAM Disk:RDBFlags" DEVICE cybppc.device 0 device "cybppc.device", unit 0 sector size is 512 bytes found RDB at sector 0 drive FUJITSU M2954Q-512 0153 last disk: yes last lun: yes last target: yes reselection: yes synchronous: no

_________________
Ordell Robbie: Is she dead, yes or no?
Louis: Pretty much.

@Cass

Quote:

Cass wrote: @TetiSoft Quote: You can verify that your SCSI chain has an active Terminator at each end of the cable, and that the drive is jumpered to deliver TermPower to the SCSI cable which is needed by the terminators. The CSPPC has no termination and does not deliver TermPower. The last time I checked, the cabling was Terminator---CSPPC---Fujitsu When shipped from the factory, the drive is terminated. I left it as default setting (nothing changed regarding the jumpers). If this is working faultless on 3.9 (read/write), does it make any sense to have such errors on 4.0?

The OS4 device is "more picky" than the FlashROM device, ergo "works under OS3"
does not mean "is setup correct". I dont know the default settings of any SCSI drive,
you should verify that it delivers TermPower to the SCSI cable, AFAIK no software
can tell you if the voltage is present or not. And while you are at it, please verify that
your drive does terminate the bus ACTIVE. If in doubt, use a separate ACTIVE terminator.

I am trying to install os4 here onto my amiga4000.
I did manage to get it installed onto an ide harddisk and LG CDRW.
This boots ok when I do a reset after the first reset off the bootprocess..otherwise it just hangs there.
I also managed to get a scsi harddisk recognised in mediatoolbox.
This only worked after I removed all old partitions(SFS) with an old OS3.9 setup.
But after I changed the partions into the OS4 template computer refuses to boot up.
Again after reset it hangs and no reset helps.
Only thing what helps is removal off the scsi cable.
I tried different scsi harddisks all same problems.
Also tried no IDE drives but then I have no cdrom to boot from.
IS there a problem for OS4 to boot if there are two harddisks(or partitions) which are both setup with this OS4 template?
Scsi harddisk is active terminated and there is an active terminator on the other side off the cable so that should be fine? Or maybe just to be sure look for an active terminator..it always worked fine before.
Staying onto ide for now to give it a go but would prefer the scsibus offcourse.
Maybe on next meeting in belgium someone can help me out on this.

@Cass

effetively all is more picky under AmigaOS4 cybppc.device.

I never got that message error (even if i did a lot of tweaking with the UWSCSI cable lately ... result, it was half borked)

Hence i would check/replace the flat cable first, then (if you can) use 2 active terminators at both ends.

About the synchron flag, you can set it with the RDBFlags tool:

RDBFlags DEVICE cybppc.device 0 Sync

It will write the Sync flag into your RDB (better to fix the error issue before)

Good luck and Happy New Year!

Looking again at your rdbflag output, i see that the RDB is seen at sector 0 ... have you used SCSIConfig in the past?

_________________