I see that there is a uhci race condition on the TODO list for
USB. Is anyone working on this? I think my problem may be related:

===========================================================================
Jan 27 21:36:58 wheel pppd[4068]: pppd 2.3.9 started by gary, uid 100
Jan 27 21:36:59 wheel chat[4070]: abort on (BUSY)
Jan 27 21:36:59 wheel chat[4070]: abort on (NO CARRIER)
Jan 27 21:36:59 wheel chat[4070]: abort on (Invalid)
Jan 27 21:36:59 wheel chat[4070]: abort on (NO DIAL)
Jan 27 21:36:59 wheel chat[4070]: abort on (Login Incorrect)
Jan 27 21:36:59 wheel chat[4070]: abort on (Bad Password)
Jan 27 21:36:59 wheel chat[4070]: abort on (Improper)
Jan 27 21:36:59 wheel chat[4070]: send (atd4907700^M)
Jan 27 21:36:59 wheel /netbsd: uhci_device_bulk_transfer: not done, ii=0xf0617d20
	[ This doesn't happen the first time I dial in
	  after a reboot, but does every time afterwards. ]

Jan 27 21:36:59 wheel chat[4070]: expect (ogin:)
Jan 27 21:36:59 wheel chat[4070]: ^M
Jan 27 21:36:59 wheel chat[4070]: OK^M
Jan 27 21:37:02 wheel xntpd[314]: sendto(153.39.63.1): No route to host
Jan 27 21:37:02 wheel xntpd[314]: sendto(192.233.85.1): No route to host
Jan 27 21:37:13 wheel xntpd[314]: sendto(4.0.46.19): No route to host
Jan 27 21:37:29 wheel chat[4070]: atd4907700^M^M
Jan 27 21:37:29 wheel chat[4070]: CONNECT 52000 V42bis^M
Jan 27 21:37:31 wheel chat[4070]: Welcome to Shore.Net. Access restricted to customers and affiliates only.^M
Jan 27 21:37:31 wheel chat[4070]: ^M
Jan 27 21:37:31 wheel chat[4070]: Port of call: gniqncy-s02-0^M
Jan 27 21:37:31 wheel chat[4070]: ^M
Jan 27 21:37:31 wheel chat[4070]: ^M
Jan 27 21:37:31 wheel chat[4070]: northshore login:
Jan 27 21:37:31 wheel chat[4070]:  -- got it 
	[ ... ]
===========================================================================

And it connects, and all appears fine, except sometimes when I
tax the link a bit, I'll get this:

===========================================================================
Jan 27 21:38:02 wheel /netbsd: uhci0: host controller process error
Jan 27 21:38:02 wheel /netbsd: uhci0: host controller halted
Jan 27 21:38:02 wheel last message repeated 5 times
Jan 27 21:38:02 wheel /netbsd: ucomstart: err=IOERROR
Jan 27 21:38:02 wheel /netbsd: uhci0: host controller halted
Jan 27 21:38:33 wheel last message repeated 145 times
Jan 27 21:40:25 wheel last message repeated 293 times
===========================================================================

which naturally wedges the modem. A reboot clears it right up, but it
gets tiresome.
   Here are the relevant bits from dmesg:

===========================================================================
NetBSD 1.4Q (WHEEL) #1: Wed Jan 26 18:47:38 EST 2000
    root@wheel:/usr/tmp/src/sys/arch/i386/compile/WHEEL
cpu0: family 6 model 1 step 2
cpu0: AMD K7 (Athlon) (686-class)
total memory = 127 MB
avail memory = 99164 KB
using 2822 buffers containing 26080 KB of memory
BIOS32 rev. 0 found at 0xfdae0

uhci0 at pci0 dev 7 function 2: VIA Technologies VT83C572 USB Controller (rev. 0x06)
uhci0: interrupting at irq 3
usb0 at uhci0: USB revision 1.0
uhub0 at usb0
uhub0: VIA Technologie UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub0: 2 ports with 2 removable, self powered
uhci1 at pci0 dev 7 function 3: VIA Technologies VT83C572 USB Controller (rev. 0x06)
uhci1: interrupting at irq 3
usb1 at uhci1: USB revision 1.0
uhub1 at usb1
uhub1: VIA Technologie UHCI root hub, class 9/0, rev 1.00/1.00, addr 1
uhub1: 2 ports with 2 removable, self powered

umodem0 at uhub0 port 1 configuration 2 interface 0
umodem0: Diamond Multimedia Systems, Inc. SupraExpress 56e USB V.90, rev 1.00/1.00, addr 2, iclass 2/2
umodem0: data interface 1, has CM over data, has break
ucom0 at umodem0
===========================================================================

  If anyone has pointers as to where to start looking for the bug
or has a patch to test, just let me know and I'll take a look.

					Gary D. Duzan


p.s. I'd do a send-pr, but since something similar is already on
a TODO list that seems a bit redundant.