Is it possible that you add dev_loss_tmo to the scsi_dev timeout? Say scsi_dev->timeout = dev_loss_tmo + X seconds?
No. This would imply setting it in slave_configure, and which means commands artificially have a longer timeout at all times, not just when we're in a loss of connectivity. I had considered having eh_timed_out return a time adder, so we timeout once rather than periodically based on the io timeout value. However, the calculations to come up with how much time remains on the dev_loss_tmo value, coupled with adding a return value to this function, didn't seem worth it. -- james s - : 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
