search

PiSCSI / piscsi

PiSCSI allows a Raspberry Pi to function as emulated SCSI devices (hard disk, CD-ROM, and others) for vintage SCSI-based computers and devices. This is a fork of the RaSCSI project by GIMONS.
https://piscsi.org
BSD 3-Clause "New" or "Revised" License
545 stars 82 forks source link

Macintosh Plus not Seeing PiSCSI #1439

Open Jeremiahdmass opened 8 months ago

Jeremiahdmass commented 8 months ago

Info

Describe the issue

I can't get the Macintosh to recognize the PiScsi at all. I have tried several different disk images, none of which have worked. The Macintosh boots to the question mark no matter what I have attached on the PiScsi. I believe it may have something to do with the Plus's weird SCSI standards.

I think it may have something to do with the Unit Attention, SCSI Selection Delay, and the SCSI selection pulses, because that has caused issues on the SCSI2SD. I have no idea how these settings could be adjusted on the PiScsi though. [https://www.codesrc.com/mediawiki/index.php/Mac_Plus]

rdmark commented 8 months ago

@Jeremiahdmass This sounds like either a SCSI termination or cablage issue. Please go through all the steps in https://github.com/PiSCSI/piscsi/wiki/Troubleshooting and see if it makes a difference.

FWIW there are known quirks and limitations with piscsi and the Mac Plus. See https://github.com/PiSCSI/piscsi/issues/656 for more info.

Jeremiahdmass commented 8 months ago

Thanks for the reply. I put a glob of solder on the termination pads and have the db25 connector plugged directly into the Mac's SCSI port so I don't think either of those could be an issue. However, I forgot to mention that when I look in the logs, there's nothing there besides it saying that I mounted whatever I was trying to boot from.

rdmark commented 8 months ago

Are other SCSI mass storage devices working properly with the same Plus and SCSI cablage?

Jeremiahdmass commented 8 months ago

I have a 20mb seagate scsi drive but it doesn't work. The mac boots to a sad mac saying its not bootable. The HD SC software on mac os sees it but can't format it. Thats the only other scsi thing I have.

Jeremiahdmass commented 8 months ago

I tried some images formatted with the silver lining driver, and still nothing. I would think a driver issue would boot to a sad Mac instead of just waiting for a floppy.

rdmark commented 8 months ago

Your symptoms make me think the signals aren’t getting through on the SCSI bus. Which suggests it’s either bad cablage, heavy oxidation on the DB25 ports, fault on the piscsi board or fault on the Plus logic board.

What’d I would do first is to try other SCSI cables and other Macs to eliminate those factors. And also work some contact cleaner into those DB25s to clear out any oxidation.

Jeremiahdmass commented 8 months ago

Unfortunately I don't have another Mac to try it on. I have worked some contact cleaner into the port but I guess I can try some more. I definitely am going to buy some more cables though.

Jeremiahdmass commented 8 months ago

I have confirmed that my connections are good. The PiScsi is now showing some activity. When the Mac is searching for a SCSI device, the ENB light flashes on once. But the Mac still says it can't find any SCSI devices. I've tried several different hard drive configurations and disk images but to no avail.

rdmark commented 8 months ago

Did you try the Plus specific Silverlining formatted image that I shared in the issue ticket linked above?

If that one doesn't work either, please share piscsi trace logs from when the Plus is powered on.

Jeremiahdmass commented 8 months ago

I have tried a Silverlining formatted disk. However, there is nothing in the logs except the attached device even after the light blinks and the Mac can't find it. Here are the logs. Mar 23 23:25:52 piscsi PISCSI[495]: [2024-03-23 23:25:52.109] [info] Validating: operation=ATTACH, command params='locale=en', 'token=???', device=6:0, type=SCHD, device params='file=HD1-100MB_HFS.hda', vendor='', product='', revision='', block size=0 Mar 23 23:25:52 piscsi PISCSI[495]: [2024-03-23 23:25:52.110] [info] Executing: operation=ATTACH, command params='locale=en', 'token=???', device=6:0, type=SCHD, device params='file=HD1-100MB_HFS.hda', vendor='', product='', revision='', block size=0 Mar 23 23:25:52 piscsi PISCSI[495]: [2024-03-23 23:25:52.111] [info] Attached SCHD 6:0

rdmark commented 8 months ago

"HD1-100MB_HFS.hda" doesn't look like the file name of the Silverlining formatted image. Are the logs from a different attempt? Anyways, there's nothing in this log message that gives any hint. Can you please make sure you enabled trace level logging?

Where did you get the piscsi board from? You might want to double check with them that the board was tested before it was shipped.

Jeremiahdmass commented 8 months ago

That isn't the image from that issue ticket, but I have also tried it and gotten nothing. The "HD1-100MB_HFS.hda" image was from somewhere random but it should also be Silverlining formatted.

'[2024-03-26 22:04:06.452] [info] Validating: operation=ATTACH, command params='locale=en', 'token=???', device=6:0, type=SCHD, device params='file=HD1-100MB_HFS.hda', vendor='', product='', revision='', block size=0 [2024-03-26 22:04:06.453] [info] Executing: operation=ATTACH, command params='locale=en', 'token=???', device=6:0, type=SCHD, device params='file=HD1-100MB_HFS.hda', vendor='', product='', revision='', block size=0 [2024-03-26 22:04:06.453] [info] Attached SCHD 6:0 [2024-03-26 22:04:06.488] [trace] Received CHECK_AUTHENTICATION command [2024-03-26 22:04:06.492] [trace] Received SERVER_INFO command [2024-03-26 22:04:06.507] [trace] Received DEVICES_INFO command [2024-03-26 22:04:06.511] [trace] Received DEFAULT_IMAGE_FILES_INFO command [2024-03-26 22:04:06.518] [trace] Received DEVICE_TYPES_INFO command [2024-03-26 22:04:06.528] [trace] Received DEFAULT_IMAGE_FILES_INFO command [2024-03-26 22:04:06.535] [trace] Received SERVER_INFO command [2024-03-26 22:04:06.560] [trace] Received NETWORK_INTERFACES_INFO command [2024-03-26 22:04:06.597] [trace] Received DEVICE_TYPES_INFO command [2024-03-26 22:04:06.606] [trace] Received DEVICE_TYPES_INFO command [2024-03-26 22:04:06.614] [trace] Received DEVICE_TYPES_INFO command [2024-03-26 22:04:06.621] [trace] Received DEVICE_TYPES_INFO command [2024-03-26 22:04:06.630] [trace] Received SERVER_INFO command [2024-03-26 22:04:06.985] [trace] Received CHECK_AUTHENTICATION command [2024-03-26 22:04:44.799] [trace] (ID 6) - ++++ Starting processing for initiator ID 7 [2024-03-26 22:04:44.799] [trace] (ID 6) - Selection phase [2024-03-26 22:04:44.799] [trace] (ID 6) - Selection completed [2024-03-26 22:04:44.799] [trace] (ID 6) - Command phase [2024-03-26 22:04:45.275] [trace] (ID 6) - Received unknown command: $0 [2024-03-26 22:04:45.275] [debug] (ID 6) - Error status: Sense Key $05, ASC $20 [2024-03-26 22:04:45.275] [trace] (ID 6) - Status phase, status is $02 [2024-03-26 22:04:45.275] [trace] (ID 6) - Sending data, offset: 0, length: 1 [2024-03-26 22:04:46.216] [trace] (ID 6) - Bus Free phase [2024-03-26 22:04:46.463] [trace] (ID 6) - ++++ Starting processing for initiator ID 7 [2024-03-26 22:04:46.463] [trace] (ID 6) - Selection phase [2024-03-26 22:04:46.463] [trace] (ID 6) - Selection completed [2024-03-26 22:04:46.463] [trace] (ID 6) - Command phase [2024-03-26 22:04:47.799] [trace] (ID 6) - Received unknown command: $0 [2024-03-26 22:04:47.799] [debug] (ID 6) - Error status: Sense Key $05, ASC $20 [2024-03-26 22:04:47.799] [trace] (ID 6) - Status phase, status is $02 [2024-03-26 22:04:47.799] [trace] (ID 6) - Sending data, offset: 0, length: 1 [2024-03-26 22:04:48.625] [trace] (ID 6) - ++++ Starting processing for initiator ID 7 [2024-03-26 22:04:48.625] [trace] (ID 6) - Selection phase [2024-03-26 22:04:48.625] [trace] (ID 6) - Selection completed [2024-03-26 22:04:48.625] [trace] (ID 6) - Command phase Segmentation fault '

I bought a kit from eBay with just the SMD parts soldered and did the rest myself, so it wasn't tested. I have checked it multiple times and my solder work is good.

rdmark commented 8 months ago

Look at that, it’s a segmentation fault at the end of the logs. Would it be possible to get a stack trace from that segfault?

Jeremiahdmass commented 8 months ago

Sorry, I don't really know what that means. I'm not good at this kind of stuff.

rdmark commented 8 months ago

That’s alright. We can walk you though the steps.

First of all, you need to check out the latest development code, and compile it with DEBUG symbols.

Please follow the steps at https://github.com/PiSCSI/piscsi/wiki/Setup-Instructions#user-content-Clone__build_the_PiSCSI_software but when you get to the make step do:

git pull && git checkout develop

Then

DEBUG=1 make

depending on how powerful your RPi is, this process may take an hour or longer

Jeremiahdmass commented 7 months ago

Ok, I have done that. Now what do I need to do? I tried to attach that image again, and the logs are the exact same.

rdmark commented 7 months ago

@Jeremiahdmass My apologies, I realized you need to do a git pull after git checkout develop to get the latest code.

The next step would be to run piscsi in the gdb debugger, e.g. sudo gdb --args piscsi -ID0 /path/to/harddrive.hda

Then you attempt to boot your system. Once the piscsi process crashes with the segfault, type "bt" in the debugger to get the backtrace. Then share the backtrace here.

Jeremiahdmass commented 7 months ago

I have fixed that and gotten the latest code, but when I run piscsi with the debugger it doesn't work. I don't think its attaching the disk image, because the activity indicator isn't doing anything. When I run the service normally, the light will blink, but only if an image is mounted.

rdmark commented 7 months ago

Did you execute the “run” command in the debugger?

Jeremiahdmass commented 7 months ago

I didn't know that needed to be done🤦‍♂️.

Thread 1 "piscsi" received signal SIGSEGV, Segmentation fault. 0x00000000004dd130 in GPIOBUS::CommandHandShake(std::vector<unsigned char, std::allocator >&) () (gdb) bt

0 0x00000000004dd130 in GPIOBUS::CommandHandShake(std::vector<unsigned char, std::allocator >&) ()

1 0x00000000004a16bc in ScsiController::Command() ()

2 0x00000000004a15a8 in ScsiController::Selection() ()

3 0x00000000004a7db0 in PhaseHandler::ProcessPhase() const ()

4 0x00000000004a11f0 in ScsiController::Process(int) ()

5 0x00000000004aaec8 in AbstractController::ProcessOnController(int) ()

6 0x00000000004a916c in ControllerManager::ProcessOnController(int) const ()

7 0x00000000004836c8 in Piscsi::Process() ()

8 0x0000000000483030 in Piscsi::run(std::span<char*, 18446744073709551615ul>) ()

9 0x00000000004e2d58 in main ()

(gdb)

Jeremiahdmass commented 7 months ago

When I boot the piscsi after the Mac and then try to use the hdsc app the service doesn't crash but the logs are different. I don't know if that means anything but here they are.

[2024-04-14 00:06:10.923] [trace] (ID 6) - ++++ Starting processing for initiator ID 7 [2024-04-14 00:06:10.923] [trace] (ID 6) - Selection phase [2024-04-14 00:06:10.924] [trace] (ID 6) - Selection completed [2024-04-14 00:06:10.924] [trace] (ID 6) - Command phase [2024-04-14 00:06:14.923] [trace] (ID 6) - Received unknown command: $0 [2024-04-14 00:06:14.923] [debug] (ID 6) - Error status: Sense Key $05, ASC $20 [2024-04-14 00:06:14.924] [trace] (ID 6) - Status phase, status is $02 [2024-04-14 00:06:14.924] [trace] (ID 6) - Sending data, offset: 0, length: 1 [2024-04-14 00:06:22.923] [trace] (ID 6) - Bus Free phase

rdmark commented 6 months ago

Very nice stacktrace, thank you! The piscsi process seems to be doing absolutely normal things there, and fails at a very early stage when accessing the RPi's GPIO bus. Not to try and lay the blame on your setup, but I wonder if your RPi could be faulty, or if you could have poor connectivity on one or more of the GPIO pins? Have you used all the GPIO pins with other RPi hats, by any chance? Do you have more RPis lying around that you could try?

Jeremiahdmass commented 6 months ago

I tried a different RPi, and I got the same results. I used the SD card from the other Pi I was using and used jumpers to connect the piscsi board to the GPIO. I triple-checked all the jumper connections before testing and ensured they were all good.

rdmark commented 6 months ago

Thanks for the thorough tests.

I released a new stable version of the software earlier this week. May I ask you to upgrade and see if this changes anything? If it looks the same, we may have to start trying older software versions and see if there was an earlier breakage.

Jeremiahdmass commented 6 months ago

I updated it and I'm still getting the exact same results. What do I need to do to install the older versions?

rdmark commented 6 months ago

Do you have spare micro SD cards? We have pre-made images for all release versions that you can flash and have a working setup with an old version with little effort.

Jeremiahdmass commented 6 months ago

I installed the February 2022 version (v22.02.02-lite) and still, nothing.

I got the trace logs and it looks like a very similar error to what I was getting on the latest version.

"[2024-05-06 22:08:53.629] [trace] virtual void SCSIDEV::Selection() Selection Phase ID=7 (with device) [2024-05-06 22:08:53.629] [trace] virtual void SASIDEV::Command() Command Phase [2024-05-06 22:08:54.351] [trace] virtual void SCSIDEV::Selection() Selection Phase ID=7 (with device) [2024-05-06 22:08:54.351] [trace] virtual void SASIDEV::Command() Command Phase [2024-05-06 22:08:55.629] [trace] virtual void SCSIDEV::Error(ERROR_CODES::sense_key, ERROR_CODES::asc) Error (to status phase) [2024-05-06 22:08:55.629] [trace] void SASIDEV::Status() Status phase [2024-05-06 22:08:55.629] [trace] void SASIDEV::Status() Status Phase $02 [2024-05-06 22:08:55.629] [trace] virtual void SCSIDEV::Send() sending handhake with offset 0, length 1 "