kernelmode fonctionne avec l'encapsulation IP over atm. Reste a le faire fonctionner correctement avec PPP.
les test effectuer par Pekka Sakara oin donn�s : DonwLoad a 52 K/S Upload a 26 K/S ----- Message retransmis par Pekka Sakara <[EMAIL PROTECTED]> ----- Date: Fri, 20 Sep 2002 09:18:30 -0200 From: Pekka Sakara <[EMAIL PROTECTED]> Reply-To: Pekka Sakara <[EMAIL PROTECTED]> Subject: Re: [Eciadsl-devel] Re: are u still interested in kernelmode driver for eci modem To: Jean-S�abstien Valette <jeanseb.valette@@UNKNOWN,[email protected]. Well, looks like 1.53 is working pretty much correctly now! :) ofcoz for testing I enabled debugging and now there is no endless loops. Download speed is 52k and upload 26k just like it should be!!! btw, I test all the time under X, so this looks real good now. btw2, this mail is send using kernelmode eci-driver :) -Pekka On Thu, 19 Sep 2002 23:00:05 +0200 Jean-S�abstien Valette <[EMAIL PROTECTED]> wrote: > I'm currently under 1.53 without debug. > If you've got this mail it semms to say that it work with rescheduling in callback :). > I've added some info in bulk callback. > Plmease test again a gime the new log if you still have problems > > On Thu, 19 Sep 2002 19:29:18 -0200 > "Pekka Sakara" <[EMAIL PROTECTED]> nous a dis: > > > I tested 1.52 version and results are littlebit better, upload speed is now from 1k to 17k (my upstream max is 26k) but there is still infinite loop in ckeck queue in bulk callback. > > > > I tried ping and ftp-upload, I attach my syslog from friday 13th ;) when I tested this last time. In the end I unplugged modem and system hang again. > > > > btw, If you don't get enough information from eci to your syslog, change all KERN_DEBUG to KERN_INFO in eci.c . This might be distribution depend thing but atleast in by redhat box I need to change this in order to get anything from eci to my syslog. > > > > -Pekka > > > > On Thu, 12 Sep 2002 20:47:00 +0200 > > Jean-S�abstien Valette <[EMAIL PROTECTED]> wrote: > > > > > > > > today version work with rescheduling. > > > was a stupid , very stuipid thing. I was testing in computing but assume that there is at least on cell in queue as queue was scheduled was data comes. > > > So I had a check on queue length in bulk callback to avoid rescheduling for doing nothing. > > > > > > On Thu, 12 Sep 2002 21:17:09 -0200 > > > "Pekka Sakara" <[EMAIL PROTECTED]> nous a dis: > > > > > > > No problem, like you said there might be change to hang my box when trying this :-) > > > > > > > > Any ideas what modifications code might need to avoid endless loop at empty bulk urb? Code has become so complicated and jumping aroud so I have no clue how to follow it anymore ;-) > > > > > > > > btw, have you checked other ATM-adapter drivers how they manage queing and other stuff we need to do while packets travels through driver? And other usb-modems too? Any advice from other drivers? > > > > > > > > -Pekka > > > > > > > > On Wed, 11 Sep 2002 22:52:45 +0200 > > > > Jean-S�abstien Valette <[EMAIL PROTECTED]> wrote: > > > > > > > > > OK it seem that I have a pobleme when checking the queue in the bottom half, this create a endless loop of empty bulk urb. > > > > > I was thinking about that but I never manage to have logs when this happens. > > > > > I'm stupid enought not to think about unpluging modem :( > > > > > > > > > > On Wed, 11 Sep 2002 21:47:32 -0200 > > > > > "Pekka Sakara" <[EMAIL PROTECTED]> nous a dis: > > > > > > > > > > > ok, I tried that modification and things went bad. I just only tried to ping and here is my syslog: > > > > > > > > > > > > Sep 11 21:32:38 brain kernel: eci.c eci_atm_open in [0.100] > > > > > > Sep 11 21:32:38 brain kernel: eci.c params after find : [0.100] > > > > > > Sep 11 21:32:38 brain kernel: eci.c eci_atm_open out > > > > > > Sep 11 21:32:53 brain kernel: eci.c eci_atm_send in > > > > > > Sep 11 21:32:53 brain kernel: eci.c queue AAL5 data for bh_atm > > > > > > Sep 11 21:32:53 brain kernel: eci.c eci_atm_send out > > > > > > Sep 11 21:32:53 brain kernel: eci.c eci_bh_atm IN > > > > > > Sep 11 21:32:53 brain kernel: eci.c Manage new AAL5 data > > > > > > Sep 11 21:32:53 brain kernel: eci.c No availlable cell create new one > > > > > > Sep 11 21:32:53 brain last message repeated 2 times > > > > > > Sep 11 21:32:53 brain kernel: eci.c AAL5 sent > > > > > > Sep 11 21:32:53 brain kernel: eci.c eci_bh_atm OUT > > > > > > Sep 11 21:32:53 brain kernel: eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:53 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:53 brain kernel: eci.c Bufpos = 64 > > > > > > Sep 11 21:32:53 brain kernel: eci.c Bufpos = 128 > > > > > > Sep 11 21:32:53 brain kernel: eci.c Bufpos = 192 > > > > > > Sep 11 21:32:53 brain kernel: ECI_USB [eci.c:1693] : new bulk [192] > > > > > > Sep 11 21:32:53 brain kernel: :00 00 06 40 ec aa aa 03 - ...@.... > > > > > > Sep 11 21:32:53 brain kernel: :00 00 00 08 00 45 00 00 - .....E.. > > > > > > Sep 11 21:32:53 brain kernel: :54 00 00 40 00 40 01 01 - T..@.@.. > > > > > > Sep 11 21:32:53 brain kernel: :1c 0a 68 0b c2 0a 63 19 - ..h...c. > > > > > > Sep 11 21:32:53 brain kernel: :01 08 00 32 0c f5 04 00 - ...2.... > > > > > > Sep 11 21:32:54 brain kernel: :00 a5 d2 7f 3d b6 db 0a - ....=... > > > > > > Sep 11 21:32:54 brain kernel: :00 08 09 0a 0b ff ff ff - ........ > > > > > > Sep 11 21:32:54 brain kernel: :ff ff ff ff ff ff ff ff - ........ > > > > > > Sep 11 21:32:54 brain kernel: :00 00 06 40 ec 0c 0d 0e - ...@.... > > > > > > Sep 11 21:32:54 brain kernel: :0f 10 11 12 13 14 15 16 - ........ > > > > > > Sep 11 21:32:54 brain kernel: :17 18 19 1a 1b 1c 1d 1e - ........ > > > > > > Sep 11 21:32:54 brain kernel: :1f 20 21 22 23 24 25 26 - . !"#$%%& > > > > > > Sep 11 21:32:54 brain kernel: :27 28 29 2a 2b 2c 2d 2e - '()*+,-. > > > > > > Sep 11 21:32:54 brain kernel: :2f 30 31 32 33 34 35 36 - /0123456 > > > > > > Sep 11 21:32:54 brain kernel: :37 00 00 00 00 ff ff ff - 7....... > > > > > > Sep 11 21:32:54 brain kernel: :ff ff ff ff ff ff ff ff - ........ > > > > > > Sep 11 21:32:54 brain kernel: :00 00 06 42 e2 00 00 00 - ...B.... > > > > > > Sep 11 21:32:54 brain kernel: :00 00 00 00 00 00 00 00 - ........ > > > > > > Sep 11 21:32:54 brain last message repeated 4 times > > > > > > Sep 11 21:32:54 brain kernel: :5c 7e 71 51 7f ff ff ff - \~qQ.... > > > > > > Sep 11 21:32:54 brain kernel: :ff ff ff ff ff ff ff ff - ........ > > > > > > Sep 11 21:32:54 brain kernel: > > > > > > Sep 11 21:32:54 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:54 brain kernel: eci.c Bulk URB Completed, length 192<6>eci.c eci_b > > > > > > h_bulk IN > > > > > > Sep 11 21:32:55 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:55 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:55 brain kernel: > > > > > > Sep 11 21:32:55 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:55 brain kernel: B [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:55 brain kernel: > > > > > > Sep 11 21:32:55 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:55 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:55 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:55 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:55 brain kernel: > > > > > > Sep 11 21:32:55 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:55 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:55 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:55 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:55 brain kernel: > > > > > > Sep 11 21:32:55 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:55 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:55 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:55 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:55 brain kernel: > > > > > > Sep 11 21:32:55 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:55 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:55 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:55 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:55 brain kernel: > > > > > > Sep 11 21:32:56 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:56 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:56 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:56 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:56 brain kernel: > > > > > > Sep 11 21:32:56 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:56 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > Sep 11 21:32:56 brain kernel: eci.c Buf = c0dfe000 > > > > > > Sep 11 21:32:56 brain kernel: ECI_USB [eci.c:1693] : new bulk [0] > > > > > > Sep 11 21:32:56 brain kernel: > > > > > > Sep 11 21:32:56 brain kernel: eci.c eci_bh_bulk OUT > > > > > > Sep 11 21:32:56 brain kernel: eci.c Bulk URB Completed, length 0<6>eci.c eci_bh_bulk IN > > > > > > > > > > > > And this goes on forever until I unplug usb-cable from modem and then my system hang without any messages to syslog about freeze. > > > > > > > > > > > > Regards, > > > > > > Pekka > > > > > > > > > > > > > > > > > > > > > > > > On Wed, 11 Sep 2002 07:42:44 +0200 (MEST) > > > > > > Jeanseb <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > > Thanks for this test. > > > > > > > In fr�ance we ve got freeze with X windows but we arent sur that tehey are du to > > > > > > > driver. > > > > > > > I know about problems with sending because of the way queuing is done. > > > > > > > Could u try to uncoment call to taskelet_schedule in the bulk callback (and line > > > > > > > about spin lock and instance pointer before and after if commented too) to check > > > > > > > if it work much better. Be careful it may freeze with this. > > > > > > > > > > > > > > > > > > > > > > > > > > > > ------- > > > > > jeanseb > > > > > I know it look like I'm insane > > > > > Take a closer look I'm not to blame. > > > > > If love is blind I guess I'll buy myself a cane > > > > > (slash / axl rose ) > > > > > > > > > > > > > ------- > > > jeanseb > > > I know it look like I'm insane > > > Take a closer look I'm not to blame. > > > If love is blind I guess I'll buy myself a cane > > > (slash / axl rose ) > > > > > ------- > jeanseb > I know it look like I'm insane > Take a closer look I'm not to blame. > If love is blind I guess I'll buy myself a cane > (slash / axl rose ) ----- Fin de message retransmis ----- =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Driver Linux pour modem ECI ADSL USB: http://eciadsl.flashtux.org Archives de cette liste de diffusion: http://liste.eci.free.fr Pour se desinscrire: mailto:[EMAIL PROTECTED]?subject=unsubscribe
