Matthew Wilcox wrote:
I still say it's an improvement over scanning the list of scsi hosts
looking for any that have work pending whenever any interrupt comes in.
No, it's not. It's just silly:
You have replaced an in-driver loop with a slower, higher-overhead,
highly unusual loop that no one but advansys does.
Your interrupt handler should do something like
board_info = dev_id;
for (i = 0; i < board_info->n_ports; i++)
advansys_port_intr(board_info, &board_info->port[i]);
or if your hardware can tell you which ports have work pending,
board_info = dev_id;
port_status = readl(hw reg);
for (i = 0; i < board_info->n_ports; i++)
if (port_status & (1 << i))
advansys_port_intr(board_info,
&board_info->port[i]);
Having the kernel interrupt subsystem implement a loop for you is
completely utterly unusual, slow, silly, and likely to get broken when
people futz with the interrupt system core. It probably has edge cases
too, like screwing with the screaming interrupt detection code.
The kernel interrupt subsystem is not meant to be abused in that way.
Jeff
-
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
