Amiga support

[ozkanocakli]

bluscsi's current firmware amiga hdtoolbox recognize bluscsi but when try to format harddrive bluesci hdd activily led turn on solid and wait

[erichelgeson]

Thanks for the report. No "core" developers have Amiga's so will compatibility likely be hit or miss. I know some have reported success but it seems to depend on system/card/etc.

[erichelgeson]

Hi all, There has been quite a few changes to the main branch since this last conversation, which may help with Amiga (or may not). I do hope to have a beta firmware out sometime this week/weekend that hopefully an Amiga user can try out and report back.

[C128Guy]

I've got an Amiga sitting here waiting :)

[Retro-updates]

Thanks. Will for sure test them out and report back the results!

[idrougge]

For what it's worth, the original ArdSCSIno-STM32 works on Amiga, so if BlueSCSI doesn't, that is something introduced by the fork.

[C128Guy]

SCSI2SD works well, I have it in an A2000 and get 2mb/sec transfers. Just none to be had at the moment!

[prickle]

I also tried on an Amiga. In fact, I tried a few systems.

*Amiga 500 with A590 expansion - in HDToolbox, bluescsi does not show up at all
*Sun Sparcstation 10 - using probe-scsi, bluescsi does not show up at all
*486 pc with Adaptec AHA-1542 - in it's BIOS bluescsi is seen, but "Command not supported" errors for format or verify

As these are the only SCSI systems I have, I can't be sure the bluescsi itself is working ok. I guess I will try to turn the serial debugging on and see what it has to say for itself.

[ea4gge]

It may need Eprom update. Here, in this video, my friend Armand install one of the BlueSCSI's i did in a A590:

https://www.youtube.com/watch?v=pbIozWBvdXw

I am also in that video giving an explaination about

[prickle]

So, on the AHA-1542
verify:
Command:2F:0:0:0:0:0:0:8:0:0:ID 1:LUN 0 [*Unknown]
format:
Command:15:10:0:0:C:0:ID 1:LUN 0 [*Unknown]

On the Amiga
during boot:
Command:80:8:0:0:x:1:ID 1:LUN 0 [*Unknown]
16 times where x increments from 0 to F

This time the drive showed up in HDtoolbox. When starting up it sends these commands:
Selection MsgOut2 Command:80:12:0:0:0:24:ID 1:LUN 0 [*Unknown] Sts MsgIn BusFree Selection MsgOut2 Command:80:37:0:8:0:0:ID 1:LUN 1 [*Unknown] Sts MsgIn BusFree Selection MsgOut2 Command:80:37:0:8:0:0:ID 1:LUN 1 [*Unknown] Sts MsgIn BusFree
Trying to ID the drive sends these messages:
Selection MsgOut2 Command:80:1A:0:3:0:FF:ID 1:LUN 0 [*Unknown] Sts MsgIn BusFree Selection MsgOut2 Command:80:12:0:0:0:24:ID 1:LUN 0 [*Unknown] Sts MsgIn BusFree
and responds "Insufficient data returned by inquiry".

[prickle]

Now, after fiddling (removed a previous debug attempt), the amiga does the following on boot:
BusReset! BusFree Selection MsgOut2 Command:12:0:0:0:FE:0:ID 1:LUN 0 [Inquiry] DATAIN PHASE
and hangs there, waiting for something?

[prickle]

Adding some debug prints like this:

inline void writeHandshake(byte d)
{
  GPIOB->regs->BSRR = db_bsrr[d]; // setup DB,DBP (160ns)
  SCSI_DB_OUTPUT() // (180ns)
  // ACK.Fall to DB output delay 100ns(MAX)  (DTC-510B)
  SCSI_OUT(vREQ,inactive) // setup wait (30ns)
  SCSI_OUT(vREQ,inactive) // setup wait (30ns)
  SCSI_OUT(vREQ,inactive) // setup wait (30ns)
  SCSI_OUT(vREQ,active)   // (30ns)
  //while(!SCSI_IN(vACK)) { if(m_isBusReset){ SCSI_DB_INPUT() return; }}
  Serial.println(" -- vREQ Active");
  while(!m_isBusReset && !SCSI_IN(vACK));
  Serial.println(" -- got vACK");
  // ACK.Fall to REQ.Raise delay 500ns(typ.) (DTC-510B)
  GPIOB->regs->BSRR = DBP(0xff);  // DB=0xFF , SCSI_OUT(vREQ,inactive)
  // REQ.Raise to DB hold time 0ns
  SCSI_DB_INPUT() // (150ns)
  while( SCSI_IN(vACK)) { if(m_isBusReset) return; }
  Serial.println(" -- release vACK");
}

gives me this:

BusReset!
Selection
MsgOut2
Command:12:0:0:0:FE:0:ID 1:LUN 0
[Inquiry]
DATAIN PHASE
Writing 36 Bytes.
 -- vREQ Active
 -- got vACK
 -- release vACK
 -- vREQ Active
 -- got vACK
 -- release vACK
 -- vREQ Active
 -- got vACK
 -- release vACK
 -- vREQ Active
 -- got vACK
 -- release vACK
 -- vREQ Active
 -- got vACK
 -- release vACK
 -- vREQ Active
 -- got vACK
 -- release vACK
 -- vREQ Active

Seems like six bytes are sent, then hangs waiting for vACK. My guess is the controller misses the first byte of the response then misinterprets the additional data length as zero, hence it's expecting a total message length of six and refuses to take any more. Maybe?

[prickle]

Line breaks in code blocks not working for me? It's not my day.

[ozkanocakli]

In my setup, a590 recognise on hdtoolbox, I setup partions and it writes data, then I try to format drive, but it stucks on process

[prickle]

You get further than me although our setups should be very similar. What could be different? Some edge case timing issues? Floating lines? Perhaps I am using a poor quality STM clone board.

[ea4gge]

Hi, all.

I have already working the BlueSCSI in an Amiga CDTV. it is working with the CDTV SCSI interface, 2MB of CHIP ram, 8 MB of Fast ram and Kick/WB 3.2. Basically the CDTV SCSI device is an A590/A2090 card. The code was uploaded to the BluePill with USB to serial interface and the Arduino IDE (code without modiffications) and the only issue i has for to start was the SCSI's terminations. I hast to test some configurations but... finally... Controller withouth terminators, JP1 not shorted, JP4 and 5 shorted, JP9 shorted, and blueSCSI powered by USB (i have to test another config but that is at the moment).

The img used for the test is a 100MB img named as: HD10_100.hda, partitioned with the HDToolBox (yes, is directly recogniced).

Edit:

The log file from the SD:

BlueSCSI <-> SD - https://github.com/erichelgeson/BlueSCSI
VERSION: 1.1-SNAPSHOT
DEBUG:0 SCSI_SELECT:0 SDFAT_FILE_TYPE:3
SdFat version: 2.1.0
SdFat Max FileName Length: 32
Initialized SD Card - lets go!
Sd MID:2 OID:TM
Sd Name:SD01G
Sd Date:11/20010
Sd Serial:2711534736
Not an image: LOG.txt
Imagefile: HD10_100.hda / 94371840bytes / 92160KiB / 90MiB
!! More than one image did not specify a SCSI ID. Last file will be used at ID 1. !!
ID:LUN0:LUN1:
0:----:----:
1: 512:----:
2:----:----:
3:----:----:
4:----:----:
5:----:----:
6:----:----:
Finished initialization of SCSI Devices - Entering main loop.

[ea4gge]

Ok... now i have correctly renamed the IMG file to HD00_512.hda. The log:

BlueSCSI <-> SD - https://github.com/erichelgeson/BlueSCSI
VERSION: 1.1-SNAPSHOT
DEBUG:0 SCSI_SELECT:0 SDFAT_FILE_TYPE:3
SdFat version: 2.1.0
SdFat Max FileName Length: 32
Initialized SD Card - lets go!
Sd MID:2 OID:TM
Sd Name:SD01G
Sd Date:11/20010
Sd Serial:2711534736
Not an image: LOG.txt
Imagefile: HD00_512.hda / 94371840bytes / 92160KiB / 90MiB
!! More than one image did not specify a SCSI ID. Last file will be used at ID 1. !!
ID:LUN0:LUN1:
0: 512:----:
1:----:----:
2:----:----:
3:----:----:
4:----:----:
5:----:----:
6:----:----:
Finished initialization of SCSI Devices - Entering main loop.

[ea4gge]

yes, same here. Termination jumpers closed, J9 closed, powered by USB and the terminations of the controler extracted (in my case) with this config is working

[ozkanocakli]

could you please send picture of your current configuration

[ea4gge]

yes, of course:

Controller and BluSCSI. Sorry, i forget to menction this point: the board is not the same, so the jumper's name will be different. In my case i did the board of:

https://github.com/xunker/bluescsi_pb

Basically the difference is the dual connectors for SCSI so it must to work equal. The terminators are 330 Ohms and 220 Ohms. The terminators of the SCSI CDTV's card are the same, 330 and 220. In all my test when terminators are not correct the isue is exactly the same as you have descrived:

"I removed termination jumpers, and power via usb port, A590 hangs on booting"

So... Can be in your card the terminators values the issue? i has before, in the SD2SCSI previouslly installed, the terminators of the design: 100 ohms, and the SCSI card's terminator installed also and it was working fine but not in this case. The only way for to have it working was the described : SCSI card without terminators and BlueSCSI with them activated.

[ozkanocakli]

both two boards are different from mine, my terminators values are 220 ohm and 390 ohm, so i have to find 330 ohm and test it again, thanx for pictures and feedback

[ea4gge]

Thanks to you. I'll be waiting your update, i hope this comments help you

[Retro-updates]

Anyone more luck in getting the BlueSCSI working with an Amiga SCSI controller? I tried the latest build v1.1-20211215 with a BlueSCSI 1.1-a rev board on a GVP SCSI controller. HDToolbox does find the virtual disks but cannot configure them.

[ea4gge]

In GVP i have not idea. in CDTV, A590, A4000, A2090/2091, A1200 with blizzard and A3000 is working without issues already.

BR. ea4gge/// Alberto

[andabata]

I got it semi working on a G-Force 030 Combo, when i created an image with UAE and used that, it booted. If i tried writing to it, it hangs.

[ea4gge]

Strage... write protection, may be?

[andabata]

That would be to easy :), I used it on a tekram controller (pc) aswell. and there it was fully functional. On an adaptec controller it failed completly.

[Retro-updates]

Did you use any specific virtual image file? Are you using standard bluescsi software?

[ea4gge]

V1.1 of the firm
The IMG must to have 512 bytes per sector

Only that is needed. At the moment i have 6 IMG mounted as the same time as 6 diferent SCSI units in the same BlueSCSI and with different partitions. Yes, it can mount diferent IMG at the same time

[ea4gge]

Ah!! another think for the GVP:

The GVP for the A500 have not Power Supply in the SCSI connector so... You must to modify it or give Power to the BluePill in the USB connector. I don't know in the case of the GVP for A2000. In that case is important to config the jumpers on the blueSCSI for isolate the PS pin on the SCSI connector. In the GVP this pin is GND

[Retro-updates]

Not yet, but i can try of course if you think it could make a difference,

[ea4gge]

may be, because the terminators. Only for test

[Retro-updates]

was away for few days, but will try to test again in the next few days

[Retro-updates]

Tested with BlueSCSI (ID0) as only drive on the GVP SCSI controller.
Booted from AmigaOS3.2 install disk, HDTooldBox scanned for SCSI devices on the GVP scsi device and found the virtual disk on the BlueSCSI.
I was able to read and save the disk configuration.
HDTooldbox showed the Quantum fireball and i was able to partition the drive.

BUT: when i try to save changes to drive and exit HDToolbox, the system seems to freeze (Mouse moves and UI Buttons seem to respond (highlight when clicked), but i cannot Save the changes to drive and cannot Exit HDToolbox.

When simply rebooting the Amiga (2000) and try to install AmigaOS3.2 there is no harddrive found (because i could not save the changes to drive after partitioning).

[Retro-updates]

Today ive'tried the BlueSCSI beta firmware (BlueSCSI-1.1-BETA-20220330.bin) on my A500 with DataFlyer 500 SCSI controller and THAT WORKED!
Booted from the DataFlyer (3.7e) diskprep diskette and then the virtual drive on the BlueSCSI MicroSD was found.
I was able to partition it and able to install AmigaOS3.2 on it without any issues.

It seems that BlueSCSI firmware is starting to be compatible with more Amiga SCSI controllers!

[marcelv-3]

for the A2000 with GVP, the only way I am able to create partitions and format the drive is to set debug = 2 in bluescsi, upload the image and start the A2000, using faaastPrep to partition/format

Debug = 2 logs to the sd card, very slow, so the problem muat be a timing issue,

After partitioning/formatting you can set debug = 0 and use that image to run the Amiga

can also upload an gvp prepped hda image somewhere

[Retro-updates]

Did you use the standard BlueSCSI-1.1-20211215.bin firmware ?

[marcelv-3]

used the 1.1-SNAPSHOT-20220522
will try later with the latest snapshot

[ozkanocakli]

I tried your image file, a590 it boots, unfortunately i can write any file to hdd, it hangs and corrupt image file

my config
A500
kick3.2
A590 scsi controller

blue retro
firmware BlueSCSI-1.1-20220404
usb powered bluepill

[ea4gge]

A590 must to be updated in the ROM side

A590_2091_PDMA07_Roms.zip

Atached lastest ROMS

[ea4gge]

Did you use the standard BlueSCSI-1.1-20211215.bin firmware ?

I have used the "before compilation"(20210813 cpp file ported to .ino file) 1.1 one (attached). In my case uploaded by arduino IDE RS232 option

BlueSCSI_1-1.zip

If have not changes

[marcelv-3]

Tested also with latest snapshot build 20220617, same problem still exists as in the previous snapshot and the latest build, unable to format/write unless I set debug to 2 (write to file), makes things very slow but everything works, format, read, write,..
In last snapshot the code to log to file is removed so had to put this back.

The problem is in the Data out phase. / readdataloop.
Maybe a timing problem.?

Seems like a more general problem for some non macintosh scsi controllers.

@erichelgeson How to best debug this? Can provide a log with complete boot/read/write (debug=2)

testeds with config:
A2000
kick 2
GVP scsi controller

[ea4gge]

Tested also with latest snapshot build 20220617, same problem still exists as in the previous snapshot and the latest build, unable to format/write unless I set debug to 2 (write to file), makes things very slow but everything works, format, read, write,.. In last snapshot the code to log to file is removed so had to put this back.

The problem is in the Data out phase. / readdataloop. Maybe a timing problem.?

Seems like a more general problem for some non macintosh scsi controllers.

@erichelgeson How to best debug this? Can provide a log with complete boot/read/write (debug=2)

testeds with config: A2000 kick 2 GVP scsi controller

May be but as we have tested... in A4000 (pending to confirm the controller), A590 (ROMs updated), A2091 (ROMs updated), A1200 (Blizzard IV) and CDTV are working fine so... i guess is an issue with the GVP

[marcelv-3]

Tried your "before compilation"(20210813 cpp file ported to .ino file) 1.1
Same results. Hangs on write or gives write error on the first block it is supposed to write.
Am glad it is working on other Amiga/SCSI combinations but want to make this working. Is working fine with a real HD and with debug on, so has to be a software issue.
The ROM (gvpscsi.ddevice) is the latest version 3.15. There is a unofficial 4.15 version so might try that, but as said earlier, the problem occurs on other systems too (samplers / sun, .. see other discussions)

[marcelv-3]

Write to hd is working using the code from the following pull request:
#142

@ozkanocakli Could you try this version to see if this also fixes your write to hd problem.

[ozkanocakli]

could you please upload firmware, i could not see it on request 142

[marcelv-3]

will upload tonight

[marcelv-3]

firmware and HD image to test, not an official release, only to check if this solves the write problem

fwpull142.zip
iHD00_512-E.zip

[ozkanocakli]

on stm32 boards still gives read/write errors , on ch32 working, could you please apply this fix to latest firmware that supports cd images

[nd-mk]

Hi,

can someone help me?
I bought some bluescsi cards that were supposed to work with the Amiga, but both on Oktagon 2008 and GVP controllers don't work ... there are four identical cards and I've been trying all day and I can't get them to work
The problem is that they don't allow booting my A2000s, the system hangs, not even allowing booting from the floppy disk
If I then reconnect the previous normal SCSI HDs everything works perfectly
What should I do ?
The LOG file is generated, but then everything freezes, as if these bluescsi drives are incompatible with the SCSI controllers
If someone is kind enough to help me, otherwise I'll send them back to the sender or throw them out the window...

this is an example of LOG that is generated in the root of the SD, as well as many other HD files that I have tried and none have solved the problem

BlueSCSI <-> SD - https://github.com/erichelgeson/BlueSCSI
VERSION: 1.1-SNAPSHOT-20220627
DEBUG:0 SDFAT_FILE_TYPE:3
SdFat version: 2.1.2
Sd Format: FAT32FAT16 - Consider formatting the SD Card with exFAT for improved performance.
SPI speed: 50Mhz
SdFat Max FileName Length: 32
Initialized SD Card - lets go!
Sd MID:9F OID:TI
SD Name:SD32G
Sd Date: 1/2015
SD Serial:3943403648

• HD00_512.hda / 524288000bytes / 512000KiB / 500MiB
  ID:MON0:
  0:512:
  1:----:
  2:----:
  3:----:
  4:----:
  5:----:
  6:----:
  Finished initialization of SCSI Devices - Entering main loop.

OR here an other exemple:

BlueSCSI <-> SD - https://github.com/erichelgeson/BlueSCSI
VERSION: 1.1-SNAPSHOT-20220627
DEBUG:0 SDFAT_FILE_TYPE:3
SdFat version: 2.1.2
Sd Format: FAT32FAT16 - Consider formatting the SD Card with exFAT for improved performance.
SPI speed: 50Mhz
SdFat Max FileName Length: 32
Initialized SD Card - lets go!
Sd MID:9F OID:TI
Sd Name:SD32G
Sd Date:1/2015
Sd Serial:3943403648

• HD00_512-E.hda / 42991616bytes / 41984KiB / 41MiB
  ID:LUN0:
  0: 512:
  1:----:
  2:----:
  3:----:
  4:----:
  5:----:
  6:----:
  Finished initialization of SCSI Devices - Entering main loop.

Oktagon 2008 and GVP controllers are terminated by default, and all tried bluescsi are also terminated, what should I do?

thanks to who helps me

[marcelv-3]

You are using software VERSION: 1.1-SNAPSHOT-20220627
Could you update to the latest version? v1.1-20220917
The latest version works much better in an Amiga with the GVP controller.

You can also try my hdd image a few posts back (iHD00_512-E.zip)

[C128Guy]

I was able to get it to work with the HC+8 card but it was not easy. First I had to initialize the disk with the Amiga, but couldn't get the o/s to install. Then I put the card into my laptop and install the o/s using WinUAE. Only way I could get it to work.

I got absolutely no where with the A2091 card.

[nd-mk]

hi

yes, It's 1.1-SNAPSHOT-20220627
Could you please upload the firmware v1.1-20220917?

could you also please point me to a simple programmer for a few € that I can order from aliexpress to reprogram it?

one last question, did you have to remove the termination resistors in your GVP controller to get your bluescsi card to work properly? or did you have to remove the termination jumpers from the bluescsi card?

Thanks for your help

[marcelv-3]

https://github.com/erichelgeson/BlueSCSI/releases/tag/v1.1-20220917
download links at bottom of page

termination on GVP and BlueScsi are on

For the programmer I use a ST-Link V2, not from ali but from a local store.
Where are you located?

[ozkanocakli]

selamlar samba yı enable etmem rağmen, yazılım cihazı bulamıyor, fsg-3 ün default ip adresi ve password u nedir iyi günler

…

On Fri, 25 Nov 2022, 00:42 marcelv-3, ***@***.***> wrote: https://github.com/erichelgeson/BlueSCSI/releases/tag/v1.1-20220917 download links at bottom of page termination on GVP and BlueScsi are on — Reply to this email directly, view it on GitHub <#34 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ANK32Q7VARQYQCMZALWGDZTWJ7OLJANCNFSM45OR77SQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>