Re: Call for stge(4) testers

On Sun, Jul 09, 2006 at 08:07:55PM -0700, John-Mark Gurney wrote:
 > Pyun YongHyeon wrote this message on Mon, Jul 10, 2006 at 11:27 +0900:
 > > On Sat, Jul 08, 2006 at 09:24:23PM -0700, Matthew Jacob wrote:
 > >  > Jason THorpe did one for NetBSD.
 > >  > 
 > >  > I did one for Solaris. Nice chip. Too bad it wasn't more popular.
 > > 
 > > I'm satisfied with the performance and easy interface for the chip.
 > > But the lack of Tx interrupt moderation and extra accesses to a
 > > status register to check which kind of Tx errors were occurred are
 > > serious fault. Since the the error condition uses the same status
 > > bit in interrupt status register it's hard to distingush failures
 > > from sucess without extra register accesses.
 > 
 > bah, tx interrupts are so last century... there isn't any reason to
 > have them... just schedule a timeout or wakeup a thread when the
 > number of tx descriptors are low.. and if you ran out, you could
 > even do the minimal work to free them up right in place...
 > 

I'm afraid it's not that simple. If we just reclaim Tx descriptors
based only on number of available Tx descriptors it would hold a lot
of mbuf chains in memory for a long time. My approach to solve this
is hybrid(a watermark which control when to raise a Tx interrupt
plus clean up handler which is called every hz)

-- 
Regards,
Pyun YongHyeon