Hi,

>>>>> On Tue, 16 Jan 2001 23:59:21 -0500, Allen Briggs <briggs@wasabisystems.com> said:

>> These errors seem to happen to me most often when moving a large amounts of
>> data to my drive over the network.  I don't know if this helps at all but it
>> may be something to look into.

Allen> That jives with what other people are saying: it's happening primarily
Allen> (only?) on writes to the disk.

There is no error in my environment with -current at 2000.01.13 24:00+9:00
and LC630.

But I don't have tried the case which is over network. My impression
is "Good jobs, Allen!" on now.

I think that esp driver performance is nearly equal to wd(c)
driver(PIO mode) one.

wd0:<ST34321A>
              -------Sequential Output-------- ---Sequential Input-- --Random--
              -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
Machine    MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU  /sec %CPU
          100   629 95.8  1157 91.2   717 91.0   685 97.2  1333 92.9  58.4 34.5

sd0:<QUANTUM, FIREBALL ST3.2S, 0F0C>
              -------Sequential Output-------- ---Sequential Input-- --Random--
              -Per Char- --Block--- -Rewrite-- -Per Char- --Block--- --Seeks---
Machine    MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU  /sec %CPU
          100   653 95.7  1065 71.4   622 53.0   785 96.9  1536 69.9  63.3 22.0

Allen> Can someone who can easily reproduce this try the attached patch?  If
Allen> it doesn't work, can you also try it without sync.?  (set sc->sc_minsync
Allen> to 0 in the espattach() function)

If I have a spair time, I am going to try it.

Takeshi Shibagaki
ie9t-sbgk@asahi-net.or.jp