RE: RAID controller safety

[Salyzyn, Mark]

Kenny Simpson [mailto:theonetruekenny@xxxxxxxxx] sez:
> Won't the i2o_block driver use i2o_block_device_flush to 
> flush the devices' cache (by issuing a
> I2O_CMD_BLOCK_CFLUSH), or this this function used in some 
> very different context?

We support I2O_BSA_CACHE_FLUSH, which is the i2o spec definition of this
identifier. It is merely internally re-issued as a SCSI
SYNCHRONIZE_CACHE command issued to the block device TID.

> Oddly enough, I see I2O_CMD_BLOCK_CFLISH #define'd to 0x37 in 
> both the i2o driver
> (include/linux/i2o.h), AND in the dpt driver 
> (drivers/scsi/dpt/dpti_i2o.h).  However, I do not see
> the dpt driver using this value anywhere.

The dpt_i2o driver is a *SCSI* driver, and the card accepts SCSI
commands to all the devices (including block). The dpt_i2o driver uses
the SCSI synchronize as the path for this action, that is why you see no
utilization of I2O_BSA_CACHE_FLUSH.

A DPT private message is used to issue these SCSI commands to the
controller.

Sincerely -- Mark Salyzyn
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/