On Tue, 24 Dec 2002, Ronald Baljeu wrote:
>
> > I doubt its the driver itself per-se. Do you have any processor load
> > figures? Does the driver take a lot of time? I suspect a potential buggy
> > USB chipset that windows copes with better. What chipset have you got?
>
> It's a D850MV board from Intel (850 chipset, 82801BA I/O controller ICH2).
> The load remains low (0.03). After 7 minutes downloading only
> pppoa shows a CPU time of 0:03 (also low); the rest
> (modem_run, pppd) show 0:00. The syslog shows this:
>
> > Dec 24 15:20:35 pcronald pppoa2[249]: CRC error in an AAL5 frame
> > Dec 24 15:20:52 pcronald last message repeated 9 times
> > Dec 24 15:21:54 pcronald pppoa2[249]: CRC error in an AAL5 frame
> > Dec 24 15:22:52 pcronald last message repeated 14 times
> > Dec 24 15:23:27 pcronald last message repeated 2 times
> > Dec 24 15:26:04 pcronald pppoa2[249]: CRC error in an AAL5 frame
> > Dec 24 15:27:06 pcronald last message repeated 7 times
> > Dec 24 15:27:52 pcronald last message repeated 34 times
> > Dec 24 15:29:19 pcronald last message repeated 318 times
> > Dec 24 15:30:19 pcronald last message repeated 331 times
> > Dec 24 15:31:20 pcronald last message repeated 375 times
> > Dec 24 15:32:19 pcronald last message repeated 386 times
> > Dec 24 15:33:19 pcronald last message repeated 355 times
> > Dec 24 15:34:19 pcronald last message repeated 360 times
> > Dec 24 15:35:19 pcronald last message repeated 374 times
> > Dec 24 15:36:19 pcronald last message repeated 362 times
> > Dec 24 15:37:20 pcronald last message repeated 352 times
> > Dec 24 15:38:19 pcronald last message repeated 386 times
>
I think there is a problem somewhere but I'm not sure where.
I have almost never seen this CRC error. I have my logs back to
Dec 1 and I have one CRC error on 22nd Dec.
Last night I attempted to upgrade my firewall and install
a new USB card as I have:
USB Controller: Intel Corp. 82371SB PIIX3 USB [Natoma/Triton II] (rev 1).
which causes problems.
Unfortunately, I discovered that I have no free PCI slots so instead
I fiddled with other settings - reduced the processor speed from
133 to 100.
I also upgraded my personal "hacked" version of pppoa3 to the CVS one
(and unfortunately didn't keep a copy of my code which was basically
an old pppoa3 + my "endpoint" patch. Finally I fiddled with the
bios settings but AFAIR the only thing I have changed is disabling
the floppy drive controller (the floppy drive was already disabled)
In addition I upgraded from a custom 2.4.14 kernel to the latest RH7.2
2.4.18-19.7.x (I think)
Here is what I got last night:
Jan 3 23:30:55 heisenberg pppd[2202]: secondary DNS address 213.120.62.99
Jan 3 23:36:06 heisenberg pppoa3[2228]: CRC error in an AAL5 frame
Jan 3 23:36:38 heisenberg last message repeated 40 times
Jan 3 23:50:21 heisenberg last message repeated 70 times
Jan 4 00:00:19 heisenberg last message repeated 8 times
Jan 4 00:10:19 heisenberg last message repeated 3 times
Jan 4 00:20:19 heisenberg last message repeated 35 times
Jan 4 00:30:25 heisenberg last message repeated 5 times
Jan 4 00:40:23 heisenberg last message repeated 21 times
Jan 4 00:50:21 heisenberg last message repeated 32 times
Jan 4 01:00:35 heisenberg last message repeated 29 times
Jan 4 01:10:19 heisenberg last message repeated 33 times
Jan 4 01:20:17 heisenberg last message repeated 6 times
Jan 4 10:15:52 heisenberg syslogd 1.4.1: restart.
(Every ten minutes one of my machines goes off to collect news and mail
and I also ran it manually at 23:36 before going to bed)
The machine locked up solid last night and there is nothing in any of
the logs to give me any clues as to what went wrong.
I have now reverted to my custom 2.4.14 kernel:
Jan 4 10:22:43 heisenberg rc: Starting adsl.sh: succeeded
Jan 4 10:25:09 heisenberg pppoa3[786]: CRC error in an AAL5 frame
Jan 4 10:25:40 heisenberg last message repeated 43 times
Jan 4 10:26:00 heisenberg last message repeated 31 times
Jan 4 10:26:01 heisenberg sshd(pam_unix)[1009]: session opened for user root by
(uid=0)
Jan 4 10:26:02 heisenberg pppoa3[786]: CRC error in an AAL5 frame
Jan 4 10:26:33 heisenberg last message repeated 52 times
Jan 4 10:28:00 heisenberg last message repeated 103 times
Jan 4 10:28:10 heisenberg sshd(pam_unix)[1009]: session closed for user root
Jan 4 10:36:29 heisenberg sshd(pam_unix)[1146]: session opened for user root by
(uid=0)
Jan 4 10:40:12 heisenberg pppoa3[786]: CRC error in an AAL5 frame
So it doesn't appear to be a kernel problem.
I will revert to my old hacked pppoa3 tonight to eliminate that and then,
if necessary, dismantle the machine again and return the processor
speed to its previous setting.
N.B. I _do_ have a known buggy chipset which I have managed to work around
at the cost of reduced bandwidth previously by using delays in pppoa3 but
with the latest pppoa3 from CVS these delays don't seem to work any more.
Maybe pppoa3 is now running much faster and that is causing more problems.
IIRC we are now using EP1 rather than EP3 (I need to check the code
to be sure)
I: If#= 1 Alt= 1 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=usbdevfs
E: Ad=06(O) Atr=02(Bulk) MxPS= 64 Ivl= 0ms
E: Ad=07(O) Atr=02(Bulk) MxPS= 64 Ivl= 0ms
E: Ad=87(I) Atr=02(Bulk) MxPS= 64 Ivl= 0ms
I: If#= 1 Alt= 2 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=usbdevfs
E: Ad=06(O) Atr=02(Bulk) MxPS= 32 Ivl= 0ms
E: Ad=07(O) Atr=02(Bulk) MxPS= 32 Ivl= 0ms
E: Ad=87(I) Atr=02(Bulk) MxPS= 64 Ivl= 0ms
I: If#= 1 Alt= 3 #EPs= 3 Cls=ff(vend.) Sub=00 Prot=00 Driver=usbdevfs
E: Ad=06(O) Atr=02(Bulk) MxPS= 16 Ivl= 0ms
E: Ad=07(O) Atr=02(Bulk) MxPS= 16 Ivl= 0ms
E: Ad=87(I) Atr=02(Bulk) MxPS= 64 Ivl= 0ms
>From my /proc/bus/usb/devices it appears that EP1 supports bigger
transfers than EP3 so this may well be the speedup cause.
Alternatively it could be the modem buffering that has been enabled.
I haven't looked at this. I think Ed Gomez found out how to do this
from Thompson? It might be that buggy(er) chipsets need this
disabled to keep the speed down a bit.
Regards,
Tim.
p.s. I'm supposed to have gone out half an hour ago so I won't get
chance to play with any of these possibilities until tomorrow.
--
God said, "div D = rho, div B = 0, curl E = - @B/@t, curl H = J + @D/@t,"
and there was light.
http://tjw.hn.org/ http://www.locofungus.btinternet.co.uk/
Liste de diffusion modem ALCATEL SpeedTouch USB
Pour se d�sinscrire : mailto:[EMAIL PROTECTED]?subject=unsubscribe
