> James Simmons wrote: > > The low_latency is used by the drivers in the case where its not in a > > interrupt context. Well we are trusting the drivers. > > Now if it is true what you said then tty_flip_buffer_push has > > a bug. Looking at several drivers including serial devices > > they set the low_latency flag. > > The generic serial driver (8250) is the one that was > dead locking when that code originally existed. > It was setting low_latency and calling from interrupt context.
serial8250_interrupt -> serial8250_handle_port -> receive_chars -> tty_flip_buffer_push It could still dead lock. The low_latency flag is too weak of a test. I patched tty_flip_buffer_push to fix this. > > > And the initial schedule has no reason to add the extra delay. > > > > So do you support a non delay work queue as well? > > No, the delay work must be used for flush_to_ldisc() > so it makes no sense to define two different work queues > (one delayed and one not) for the same work. Sorry I mean move to a just a non delay work queue only. Which we won't anyways. > I support your patch. > The current stuff works and your patch works. > With your patch, you actually reduce initial > latency for processing receive data. Thanks. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
