Intermittent connections are not "debounced". It is a little known
fact that if you don't include special circuitry to debounce
pushbuttons, the software will see a sequence of quick ons and offs,
sometimes thousands of them, over the period of time of tens to
hundreds of milliseconds. So an intermittent could cause what looks
like a thousand plug-ins and unplug-ins over the course of a very
short period of time.
Just one conjecture: some queue somewhere fills up, and there is an
undetected memory overrun clobbering something just after it in
memory, and the suspend/resume just happens to run through code that
re-initializes the broken variables?
Could even be hardware: that kind of event sequence screws up the
state of the USB-handler chip? And the suspend/resume cuts power to
the chip then powers it up, and that resets the screwed-up state?
Bad hardware can cause software to be crazy, and the craziness that
can result is essentially unlimited.
A word of experience from somebody who did hardware back when a 1024
bit memory chip (2102) and a 1MHz processor (6502) were Hot Stuff...
On Dec 7, 2008, at 3:05 PM, Neil Sikka wrote:
How would the hairline fracture explain the remedy of suspending and
resuming the laptop?
On Sun, Dec 7, 2008 at 2:06 PM, Dmitry Dagunts <[EMAIL PROTECTED]>
wrote:
In agreement with Charles, If it's just one USB connector that stops
working you may be able to trace the paths nearby, but since you are
loosing all of your USB connectors, I think the assumption of a
hairline crack is more probable. Unfortunately, I would imagine
this crack to be somewhere close to the southbridge on the
motherboard and not anywhere near the USB connectors themselves. I
too doubt this can be easily traced/fixed.
On Sun, Dec 7, 2008 at 1:52 PM, Charles Cranston <[EMAIL PROTECTED]> wrote:
My best thought is you have a hairline crack on the motherboard, and
when you flex the machine just-so, it bends the motherboard such
that the crack opens. But first I would take a look at where the
USB connector is connected to the motherboard, because often the
mechanical stress of plugging and unplugging, or even displacing the
USB plug up and down while connected, is enough to cause that solder
joint to fail. With single or double sided boards one can bridge
such a crack with a small piece of wire, but most complex circuit
boards are multi-layer these days, and are essentially unrepairable
by the layman. However, if it's just the USB connector, you might
be able to carefully solder it back.
On Dec 7, 2008, at 1:24 PM, Ed Kohlwey wrote:
Hi Everyone,
As some of the group regulars know I've been having problems with my
laptop "flipping out" lately. When pressure is applied to the case
in a
certain way, anything connected on USB stops working, and I get a slew
of errors in dmesg (attached below). The computer will also refuse to
boot the first 20 or so times that I hit the power button after such
an
event. To get it to turn back on I have to shake it, disconnect the
power supply, etc. a few times. The really weird thing is that I can
negate the effects of the "flip out" by suspending and resuming the
laptop before a complete power down (at least this seems to work
consistently, I'm not completely sure if this is actually connected or
not).
My first intuition was that it might be a short, so I opened up my
case
and covered most of the bare metal with either plastic or electrical
tape. For a while, the computer would also turn itself off, and the
electrical tape fix seems to have at least remedied that behavior.
Does anyone have any ideas on how to troubleshoot this sort of
issue? It
seems almost certain to me that its electrical in nature, but short of
what I've done I don't really have any ideas.
-Ed
====================================================================
[ 2437.744104] usb 1-1: USB disconnect, address 2
[ 2438.052062] usb 1-1: new full speed USB device using uhci_hcd and
address 3
[ 2438.172170] usb 1-1: device descriptor read/64, error -71
[ 2438.396053] usb 1-1: device descriptor read/64, error -71
[ 2438.612070] usb 1-1: new full speed USB device using uhci_hcd and
address 4
[ 2438.732056] usb 1-1: device descriptor read/64, error -71
[ 2438.956131] usb 1-1: device descriptor read/64, error -71
[ 2439.172058] usb 1-1: new full speed USB device using uhci_hcd and
address 5
[ 2439.580046] usb 1-1: device not accepting address 5, error -71
[ 2439.692043] usb 1-1: new full speed USB device using uhci_hcd and
address 6
[ 2440.104051] usb 1-1: device not accepting address 6, error -71
[ 2440.105410] hub 1-0:1.0: unable to enumerate USB device on port 1
[ 2440.106752] hub 3-0:1.0: port 1 disabled by hub (EMI?),
re-enabling...
[ 2440.106763] usb 3-1: USB disconnect, address 5
[ 2440.107160] btusb_intr_complete: hci0 urb ffff880074099600 failed
to
resubmit (19)
[ 2440.111402] btusb_send_frame: hci0 urb ffff880060b29cc0 submission
failed
[ 2440.600069] usb 3-1: new full speed USB device using uhci_hcd and
address 6
[ 2440.720071] usb 3-1: device descriptor read/64, error -71
[ 2440.944056] usb 3-1: device descriptor read/64, error -71
[ 2441.160051] usb 3-1: new full speed USB device using uhci_hcd and
address 7
[ 2441.280063] usb 3-1: device descriptor read/64, error -71
[ 2441.504071] usb 3-1: device descriptor read/64, error -71
[ 2441.720060] usb 3-1: new full speed USB device using uhci_hcd and
address 8
[ 2442.128053] usb 3-1: device not accepting address 8, error -71
[ 2442.240052] usb 3-1: new full speed USB device using uhci_hcd and
address 9
[ 2442.648041] usb 3-1: device not accepting address 9, error -71
[ 2442.649393] hub 3-0:1.0: unable to enumerate USB device on port 1
--
Neil Sikka
