On Tue, 2006-10-17 at 10:45 -0400, Jes Sorensen wrote: > I have a bug report about the qla1280 driver dropping the per device > negotiated settings past a bus reset. I could in principle try and > keep a list of configured devices in the driver, but it seems to me > that it would be reasonable that the SCSI layer should call > ->slave_configure() again for each configured device after a bus > reset? Really, no. A lot of drivers allocate resources in slave_configure() on the assumption that it is just called once before the device begins to be used in earnest, so this routine isn't idempotent. > Anyone else seeing this in other drivers and have any oppinion on it? What most SPI drivers do is to call spi_schedule_dv() from the reset detection routines. This schedules domain validation to run after the reset processing has completed. Unfortunately, you actually have to attach to the transport class for this to work ... which the qla1280 doesn't. James - To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
