[sane-devel] Resend: scanjet 5p, io-error

[Peter Kirchgessner]

Hi,

could be a similar problem. But this time the SCSI interface is a 
Initio-9100uw on FreeBSD 4.6.2.

--Peter

abel deuring schrieb:
> Peter Kirchgessner wrote:
> 
>> Hi,
>>
>> seems to be a SCSI problem. Opening the SCSI-device is ok. Writing 
>> seems also to work. But reading is a problem. You can also have a look 
>> at the system log messages. If there are errors reported from the SCSI 
>> drivers.
> 
> 
> 
>>> [hp] scsi_inquire: sending INQUIRE
>>> [hp] scsi_new: sending TEST_UNIT_READY
>>> [hp] scsi_flush: writing 2 bytes:
>>> [hp]  0x0000  1B 45                                            .E
>>> [hp] scsi_flush: writing 7 bytes:
>>> [hp]  0x0000  1B 2A 73 32 35 37 45                             .*s257E
>>> [hp] scl_inq: read failed (Error during device I/O)
>>> [hp] scl_errcheck: Can't read SCL error stack: Error during device I/O
>>> [hp] sanei_hp_device_new: SCL reset failed
>>> [hp] scsi_close: closing fd 0
>>
> 
> [disclaimer: I'm writing this without having a look into older mails, so 
> can easily be wrong, but:] Isn't this error similiar to the one we have 
> with 2.4 Linux kernels and the Buslogic driver? The latter was caused by 
> a "too paranoid" handling of a data over- or underrun by the Buslogic 
> driver.
> 
> The FREEBSD_CAM_INTERFACE implementation of sanei_scsi_cmd2 in 
> sanei_scsi.c has a very simple error handling; especially there is no 
> attempt to figure out, which error occured, except for "device busy". 
> Perhaps the problem could be fixed by inspecting the CCB in more detail 
> in sanei_scsi_cmd2.
> 
> 
> Abel
> 
> _______________________________________________
> Sane-devel mailing list
> Sane-devel at www.mostang.com
> http://www.mostang.com/mailman/listinfo/sane-devel
> 


-- 
Peter Kirchgessner
http://www.kirchgessner.net
mailto:peter at kirchgessner.net