Re: [PATCH] atapi request sense work

[Pat LaVarre <p.lavarre@ieee.org>]

Jeff G:

> Open-ioctl-close of /dev/scd$n now mostly works.
> 
> (Read/ write still has issues I will report separately.)

Ah, never mind there for now.  Digging a little, I see plug 'n play
(e.g. the determination of DVD/ CD disc rewritable or not) may go weird
while yet we neglect to auto sense.  Moving on ...

To my newbie surprise, I have more good news.  Teaching my -bk8 kernel
toclear x04 ATA_DMA_INTR, means my kernel no longer needs my lab hack of
the bk8.good.always.patch.

Failure is ok now, though not yet autosensed.

Sorry to say, failure is ok only if I do NOT apply your autosense
patch.  Still here applying that provokes an indefinite series of
/var/log/messages that begin, for example

kernel: ata_sg_clean: unmapping 1 sg elements
kernel:   Vendor: Iomega    Model: RRD               Rev: 74.B
kernel:   Type:   CD-ROM                             ANSI SCSI revision: 00
kernel: ata_scsi_dump_cdb: CDB (2:0,0,0) 00 00 00 00 00 00 00 00 00
kernel: ata_scsi_translate: ENTER
kernel: ata_dev_select: ENTER, ata2: device 0, wait 1
kernel: ata_tf_load_pio: feat 0x0 nsect 0x0 lba 0x0 0x0 0x0
kernel: ata_tf_load_pio: device 0xA0
kernel: ata_exec_command_pio: ata2: cmd 0xA0
kernel: ata_scsi_translate: EXIT
kernel: atapi_packet_task: busy wait
kernel: atapi_packet_task: send cdb
kernel: ata_host_intr: BUS_NODATA (drv_stat 0x51 host_stat 0x24)
kernel: ata_dma_complete: ENTER
kernel: ata_dma_complete: host 2, host_stat==0x24, drv_stat==0x51
kernel: ata_scsi_error: ENTER
kernel: ata_eng_timeout: ENTER
kernel: ata_tf_load_pio: feat 0x0 nsect 0x0 lba 0x0 0x60 0x0
kernel: ata_tf_load_pio: device 0xA0
kernel: ata_exec: ata2: cmd 0xA0
kernel: ata_exec_command_pio: ata2: cmd 0xA0
kernel: atapi_error: send cdb
kernel: ata_eng_timeout: EXIT
kernel: ata_scsi_error: EXIT
kernel: ata_scsi_dump_cdb: CDB (2:0,0,0) 00 00 00 00 00 00 00 00 00
kernel: ata_scsi_translate: ENTER
kernel: ata_scsi_dump_cdb: CDB (2:0,0,0) 00 00 00 00 00 00 00 00 00
kernel: ata_scsi_translate: ENTER
kernel: ata_scsi_dump_cdb: CDB (2:0,0,0) 00 00 00 00 00 00 00 00 00
kernel: ata_scsi_translate: ENTER
...

I think a significant variable may be whether I have other /dev/scd*
devices present and various modules more or less autoloaded.  I don't
imagine those variables change whether auto sense works or not, but I do
think they change whether Linux itself catches a unit attention or
whether I have to ask to capture a unit attention.

Real life intrudes, bye for now, I'll post again when next I make
progress and/or feel stuck.

Pat LaVarre