| Re: [hatari-devel] SCSI emulation |
[ Thread Index |
Date Index
| More lists.tuxfamily.org/hatari-devel Archives
]
- To: hatari-devel@xxxxxxxxxxxxxxxxxxx
- Subject: Re: [hatari-devel] SCSI emulation
- From: Uwe Seimet <Uwe.Seimet@xxxxxxxxx>
- Date: Sat, 1 Sep 2018 14:27:42 +0200
- Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; t=1535804862; s=strato-dkim-0002; d=seimet.de; h=In-Reply-To:References:Message-ID:Subject:To:From:Date: X-RZG-CLASS-ID:X-RZG-AUTH:From:Subject:Sender; bh=tNM4bwZwhduoL4R9FUoygfJZdJw0czBosidYGFevKmI=; b=tuV+Yk1BHpLv0jk3dXXGfLGY9PI0CGa4CbcH4e7meivljsST0drptO6HqC6jdxpf1C ECNzzW3eo4XqYdC1cLsW0c09dIQ2bWaK6f6pXx7+8g/UznLM3kSyE4d741fPkgF3MQZt lI5KiY2oNU38domGsl6fVrHh7DsfDGf5U430w/yxSkrJwavOrvFEaLquz3ho1+jPLAdC zwCw9F1c7x0Rh/dA8NY1E8Jg05IZ7Gu5spbZfs4aI3qQ3HkQoUN+aRSQALMzqdwV9/vl kMtF9DUgY/zBt+fAti1BDqOxQa8tsXtimVuUNNeIDDxNC1CPo0+42jmEtvsxTgNQuU9I 9TTA==
Hi Thomas,
> > When running SDRVTEST (available for download on
> > https://www.hddriver.net/en/downloads.html) it reported some problems
> > with the emulation, please see the attached log.
>
> Ok, thanks a lot for the hint, I think I've found and fixed the bug with
> the unsupported commands now. SDRVTEST now reports "OK" for a SCSI drive
> attached with ID 0.
> There still seem to be some problems with IDs > 0, though. E.g. with
> https://www.hddriver.net/downloads/hddrdemo101_big_endian.img the driver is
> initially loaded, but then fails to recognize the partition table. Do
> you know what could be wrong here?
Yes, things look better now. On my machine (Hatari) the partition table
of the hddrdemo101_big_endian.img is detected and I can access the root
directory/folder.
Note that my copy of HDDRIVER is still configured for ID 1. Did you configure
your copy to check SCSI ID 1? See "Devices and Partitions". Your problem
with IDs > 0 might just be a configuration issue.
What does not work for me is launching an application (HDDRUTIL.APP from
this image) or loading one of the text files. When trying to launch HDDRUTIL
the desktop reappears and HDDRUTIL is not started, and when trying to load
a text file (README or LIESMICH) from that image, Thing reports "Diese
Anwendung kann das angegebene Object nicht finden".
What I noticed is that START/STOP UNIT is sent from HDDRIVER, which it
would usually not do, unless triggered by the status returned by a device.
When trying to display LIESMICH with SCSI this is what Hatari displays:
HDC: READ SECTOR (t=1, lun=0, opc=40, cnt=0x1, ctrl=0x0) with LBA 0x6 -> OK (0)
HDC: SHIP (t=1, lun=0, opc=27, cnt=0x1, ctrl=0x0).
HDC: SHIP (t=1, lun=0, opc=27, cnt=0x1, ctrl=0x0).
HDC: SHIP (t=1, lun=0, opc=27, cnt=0x1, ctrl=0x0).
HDC: SHIP (t=1, lun=0, opc=27, cnt=0x1, ctrl=0x0).
When I use the same image with ACSI I get this output:
HDC: READ SECTOR (t=1, lun=0, opc=40, cnt=0x1, ctrl=0x0) with LBA 0x6 -> OK (0)
HDC: READ SECTOR (t=1, lun=0, opc=40, cnt=0x4, ctrl=0x0) with LBA 0x28a -> OK (0)
HDC: READ SECTOR (t=1, lun=0, opc=40, cnt=0x1, ctrl=0x0) with LBA 0x28e -> OK (0
In this case it works. I assume there is still something wrong with the
SCSI or DMA status returned by Hatari. In general, the log output should
be identical for ACSI and SCSI as long as the same drive image is used.
By the way, can you change the logging so that it's possible to
distinguish between SCSI and ACSI log messages? Don't know about others,
but personally I'd prefer a hex value also for the 'opc' field.
Best regards
Uwe