[speedtouch] Re: Getting close Was Not your normal Speedtouch Woes!
So. One option is to do a fresh install of RH Enterprise 3 then install the Speedbundle download. Is that correct? I note there are some problems with Speedbundle and RH9, is this something I should worry about? Also the included documentation for speedbundle (ie README) looks a bit skimpy to me, is there a dummies version anywhere? > -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Duncan Sands > Sent: Monday, 17 November 2003 9:46 PM > To: [EMAIL PROTECTED]; Bill Bennett > Subject: [speedtouch] Re: Getting close Was Not your normal Speedtouch > Woes! > > > There are two drivers for the speedtouch modem: the "kernel mode driver" > which uses a kernel module and a special pppd, and Benoit's "user mode > driver" > which does not need the speedtch kernel module and uses an ordinary pppd. > The kernel mode driver's module is available in the latest 2.4 kernels. > This > makes your life easier if you are using the kernel mode driver (see > http://linux-usb.sf.net/SpeedTouch), but it doesn't help you if you are > using > the user mode driver as you are. In fact it makes things worse because > the > two drivers can fight over who gets to drive the modem. So not only do > you > still need to do all the usual steps to get the user mode driver working, > you may also need to delete (or unload) the speedtch module (it lives > in a directory under /lib/modules and is called speedtch.o). > > All the best, > > Duncan. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Hello Edouard, Just a quick reply, more tomorrow. > I just dropped the test. See if it makes CRC problems appear again. Just dropping the test for pos > 1536 shouldn't change anything afaict. Remember I inserted the else after that check. Also I always saw both "Idiotic frame length"s and CRC errors at the same time (before I inserted the else). That would suggest that dropping the first test shouldn't make a difference. > The lbuf usage is still needed, it's the reading buffer. unused_cell is > just a pointer that keeps tracks of pppoa3 progress into the read buffer if > there are still ATM cells left in. Then that should be changed back. The current patch doesn't seem to use lbuf at all. > I have just a doubt about the case where: > - num_bytes_read < 0 > > It seems to me, that this case should be checked and invalidate the > unused_cell data. Isn't unused_cells set to NULL by aal5_frame_from_atm_cell() after the whole frame has been processed? I did not check that. Again, I only submitted this patch because it addresses the CRC issue, I am not stating that it is entirely correct yet. This probably needs a little discussion (please refer to Ok Overbeek as well) and shouldn't be pushed to beta3 in 24 hours. > Can you check if a test like this one hurst the CRC solving: > if (num_bytes_read<0) unused_cells = NULL; > after the pos = 0; statement. I'll look at this tomorrow. Plus see above. By the way, num_bytes_read is the absolute number of bytes read so it will never be less than 0. And again, I assume unused_cells is set to NULL by aal5_frame_from_atm_cell(). Bye, Leonard. -- How clean is a war when you shoot around nukelar waste? Stop the use of depleted uranium ammo! End all weapons of mass destruction. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Leonard den Ottolander ([EMAIL PROTECTED]) wrote: > Is there a way to check the frame length against the negotiated length, or > do you just drop this check? No, pppoa daemons are just sort of bridging pipe. It's just a dumb piece of software: read data on one side, copy it to the other side. I just dropped the test. See if it makes CRC problems appear again. > > - some cleanup concerning the effective payload used > > The way aal5_frame_from_atm_cells() is called probably needs checking. I > am not sure if the use of unused_cells instead of lbuf is correct. Might > just work for small frame sizes. Also it seems that currently lbuf is not > used at all so if the use of unused_cells instead of lbuf is correct you > might just drop all references to lbuf. The lbuf usage is still needed, it's the reading buffer. unused_cell is just a pointer that keeps tracks of pppoa3 progress into the read buffer if there are still ATM cells left in. I have just a doubt about the case where: - num_bytes_read < 0 It seems to me, that this case should be checked and invalidate the unused_cell data. > Once again, this patch was not written by me, I just did some cleanup > and separation of functionality. I did not verify all the details of > its inner working, but can verify that it fixes the CRC error problem > I saw when using the KQD6P2.eni firmware with the unpatched > driver. But it definitely needs another check ;) . Can you check if a test like this one hurst the CRC solving: if (num_bytes_read<0) unused_cells = NULL; after the pos = 0; statement. -- Edouard Gomez Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Hello Edouard, > Ok concerning the second patch, i'm going to apply it with slight > changes: > - the 'idiotic' message isn't good as this driver can be used on >networks with mtu > 1500. Not all the networks are Ethernet, and it's up >to your ISP to negotiate the MTU value through pppd or whatever. Is there a way to check the frame length against the negotiated length, or do you just drop this check? > - some cleanup concerning the effective payload used The way aal5_frame_from_atm_cells() is called probably needs checking. I am not sure if the use of unused_cells instead of lbuf is correct. Might just work for small frame sizes. Also it seems that currently lbuf is not used at all so if the use of unused_cells instead of lbuf is correct you might just drop all references to lbuf. Once again, this patch was not written by me, I just did some cleanup and separation of functionality. I did not verify all the details of its inner working, but can verify that it fixes the CRC error problem I saw when using the KQD6P2.eni firmware with the unpatched driver. But it definitely needs another check ;) . > -> use of constants defined in atm.h Sure. > and move n1 to the inner if(pti==1) >subcase. n1 has been renamed to nb_cells as well. And sure, sure again :-) . Bye, Leonard. -- How clean is a war when you shoot around nukelar waste? Stop the use of depleted uranium ammo! End all weapons of mass destruction. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Gilles Espinasse ([EMAIL PROTECTED]) wrote: > And the patch to modem_run send a few days ago by Howard Wilkinson? Sure. I just turned the -T option into -t to have lower cases option only (and -t was not used). And added some comments (2 iirc) and turned the last return statement surrounded by if, with a ()?:; operator (nicer ;-) Do you see anything else important i miss and that should be part of beta3 ? I give a 24h delay so people can complain, then i'll release beta3 if no major problem is found. 2003-11-17 23:14:56 GMT patch-13 Summary: Added timeout in modem_run. Revision: speedtouch--trunk--1.2--patch-13 Added timeout alarm to give up connecting in modem_run. modified files: ChangeLog doc-linux/man/modem_run.1 src/modem_run.c -- Edouard Gomez Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
And the patch to modem_run send a few days ago by Howard Wilkinson? Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Leonard den Ottolander ([EMAIL PROTECTED]) wrote: > Please consider the patches I supplied via sourceforge before jumping > to beta3. TIA. Ok concerning the second patch, i'm going to apply it with slight changes: - the 'idiotic' message isn't good as this driver can be used on networks with mtu > 1500. Not all the networks are Ethernet, and it's up to your ISP to negotiate the MTU value through pppd or whatever. - some cleanup concerning the effective payload used -> use of constants defined in atm.h and move n1 to the inner if(pti==1) subcase. n1 has been renamed to nb_cells as well. -- Edouard Gomez Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: Getting close Was Not your normal Speedtouch Woes!
On Monday 17 November 2003 22:28, Edouard Gomez wrote: > Bill Bennett ([EMAIL PROTECTED]) wrote: > > Is there a good reason NOT to use the kernel driver? > > - If you use a BSD > - If you use an old Linux kernel (<2.4.20 and 2.2.x) < 2.4.16 > then the userspace driver is only choice you have. You can (safely) use > the kernel driver with linux >2.4.20 and 2.6. Duncan. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: Dodgy Link
On Monday 17 November 2003 22:24, Bill Bennett wrote: > The link works, but the page is empty. Internet explorer has a bug which causes the page contents to be shifted down. Just scroll down to the bottom of the page. Duncan. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: Getting close Was Not your normal Speedtouch Woes!
Bill Bennett ([EMAIL PROTECTED]) wrote: > Is there a good reason NOT to use the kernel driver? - If you use a BSD - If you use an old Linux kernel (<2.4.20 and 2.2.x) then the userspace driver is only choice you have. You can (safely) use the kernel driver with linux >2.4.20 and 2.6. -- Edouard Gomez Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Dodgy Link
The link works, but the page is empty. > -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Duncan Sands > Sent: Monday, 17 November 2003 9:46 PM > To: [EMAIL PROTECTED]; Bill Bennett > Subject: [speedtouch] Re: Getting close Was Not your normal Speedtouch > Woes! > This > makes your life easier if you are using the kernel mode driver (see > http://linux-usb.sf.net/SpeedTouch), but it doesn't help you if you are Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: Getting close Was Not your normal Speedtouch Woes!
Is there a good reason NOT to use the kernel driver? > -Original Message- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Duncan Sands > Sent: Monday, 17 November 2003 9:46 PM > To: [EMAIL PROTECTED]; Bill Bennett > Subject: [speedtouch] Re: Getting close Was Not your normal Speedtouch > Woes! > > > On Monday 17 November 2003 01:17, Bill Bennett wrote: > > My kernel is 2.4.21-4.EL. > > > > I read (deep inside the Redhat support forum) that this kernel has > support > > for the Speedtouch modem. So rather than fuss about patching, I went > back > > to first steps and did a clean install, bypassing all the kernel parts > of > > Benoit Papillault's Speedtouch HOWTO, got all the way to: > > There are two drivers for the speedtouch modem: the "kernel mode driver" > which uses a kernel module and a special pppd, and Benoit's "user mode > driver" > which does not need the speedtch kernel module and uses an ordinary pppd. > The kernel mode driver's module is available in the latest 2.4 kernels. > This > makes your life easier if you are using the kernel mode driver (see > http://linux-usb.sf.net/SpeedTouch), but it doesn't help you if you are > using > the user mode driver as you are. In fact it makes things worse because > the > two drivers can fight over who gets to drive the modem. So not only do > you > still need to do all the usual steps to get the user mode driver working, > you may also need to delete (or unload) the speedtch module (it lives > in a directory under /lib/modules and is called speedtch.o). > > All the best, > > Duncan. > > > # pppd call adsl > > > > This seemed fine. Then tried: > > > > # ifconfig ppp0 > > > > This gave me: > > > > ppp0: error fetching interface information: device not found. > > > > This is hardly surprising as my dev directory has ippp0 through to > ippp63 > > and ppp but no ppp0. I ran MAKEDEV ppp and simply got ipppx0 through to > > ipppx63. What's going on here? > > > > I looked at messages and found that the ADSL line is working fine, but > > after ppp0 is started by root I get > > > > ...can't locate module tty-ldisc-14 > > > > Working backwards I found my /etc/modules.conf contains the line: > > > > Alias tty-ldisc-14 ppp_synctty > > > > So why isn't this being picked up? > > > > > -Original Message- > > > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] > On > > > Behalf Of Gilles Espinasse > > > Sent: Saturday, 15 November 2003 9:11 PM > > > To: [EMAIL PROTECTED] > > > Subject: [speedtouch] Re: Not your normal Speedtouch Woes! > > > > > > > > > > > > - Original Message - > > > From: "Bill Bennett" <[EMAIL PROTECTED]> > > > To: <[EMAIL PROTECTED]> > > > Sent: Saturday, November 15, 2003 6:51 AM > > > Subject: [speedtouch] Not your normal Speedtouch Woes! > > > > > > > Well maybe this is a common problem. Like most other people, I did > > > > everything according to the script, but without success. However, > the > > > > process bombed out in a way that I've not seen described elsewhere. > > > > > > > > Modprobe couldn't find the n_hdlc module. > > > > > > > > I'm a bit uncertain about this as I'm running Redhat Enterprise 3 > > > > Linux, which is about as recent as distros get and SOME Speedtouch > > > > > > documentation > > > > > > > says this is already in recent kernels. > > > > > > What kernel version are you running? (uname -r) > > > You don't need to patch on recent kernel but you may need to set > > > CONFIG_N_HDLC=m in you kernel .config and recompile your kernel > > > > > > > > > > > > Liste de diffusion modem ALCATEL SpeedTouch USB > > > Pour se désinscrire : mailto:speedtouch- > > > [EMAIL PROTECTED] > > > > Liste de diffusion modem ALCATEL SpeedTouch USB > > Pour se désinscrire : > > mailto:[EMAIL PROTECTED] > > > Liste de diffusion modem ALCATEL SpeedTouch USB > Pour se désinscrire : mailto:speedtouch- > [EMAIL PROTECTED] > > Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Leonard den Ottolander ([EMAIL PROTECTED]) wrote: > Please consider the patches I supplied via sourceforge before jumping > to beta3. TIA. Thanks to point me to them, as it's very rare to have patches on the sf site :-) The cleanup patch is not a problem, accepted as is. The second one will require a bit of review. I just need to check two or three things in the code i've just cross read. -- Edouard Gomez Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: [CVS commits] Various small things
Hello Edouard, > It's been a while since the last commits. This is probably going to be > released as beta3. Then i'll do rc releases for a final 1.2. Please consider the patches I supplied via sourceforge before jumping to beta3. TIA. Bye, Leonard. -- How clean is a war when you shoot around nukelar waste? Stop the use of depleted uranium ammo! End all weapons of mass destruction. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: Getting close Was Not your normal Speedtouch Woes!
On Monday 17 November 2003 01:17, Bill Bennett wrote: > My kernel is 2.4.21-4.EL. > > I read (deep inside the Redhat support forum) that this kernel has support > for the Speedtouch modem. So rather than fuss about patching, I went back > to first steps and did a clean install, bypassing all the kernel parts of > Benoit Papillault's Speedtouch HOWTO, got all the way to: There are two drivers for the speedtouch modem: the "kernel mode driver" which uses a kernel module and a special pppd, and Benoit's "user mode driver" which does not need the speedtch kernel module and uses an ordinary pppd. The kernel mode driver's module is available in the latest 2.4 kernels. This makes your life easier if you are using the kernel mode driver (see http://linux-usb.sf.net/SpeedTouch), but it doesn't help you if you are using the user mode driver as you are. In fact it makes things worse because the two drivers can fight over who gets to drive the modem. So not only do you still need to do all the usual steps to get the user mode driver working, you may also need to delete (or unload) the speedtch module (it lives in a directory under /lib/modules and is called speedtch.o). All the best, Duncan. > # pppd call adsl > > This seemed fine. Then tried: > > # ifconfig ppp0 > > This gave me: > > ppp0: error fetching interface information: device not found. > > This is hardly surprising as my dev directory has ippp0 through to ippp63 > and ppp but no ppp0. I ran MAKEDEV ppp and simply got ipppx0 through to > ipppx63. What's going on here? > > I looked at messages and found that the ADSL line is working fine, but > after ppp0 is started by root I get > > ...can't locate module tty-ldisc-14 > > Working backwards I found my /etc/modules.conf contains the line: > > Alias tty-ldisc-14 ppp_synctty > > So why isn't this being picked up? > > > -Original Message- > > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > > Behalf Of Gilles Espinasse > > Sent: Saturday, 15 November 2003 9:11 PM > > To: [EMAIL PROTECTED] > > Subject: [speedtouch] Re: Not your normal Speedtouch Woes! > > > > > > > > - Original Message - > > From: "Bill Bennett" <[EMAIL PROTECTED]> > > To: <[EMAIL PROTECTED]> > > Sent: Saturday, November 15, 2003 6:51 AM > > Subject: [speedtouch] Not your normal Speedtouch Woes! > > > > > Well maybe this is a common problem. Like most other people, I did > > > everything according to the script, but without success. However, the > > > process bombed out in a way that I've not seen described elsewhere. > > > > > > Modprobe couldn't find the n_hdlc module. > > > > > > I'm a bit uncertain about this as I'm running Redhat Enterprise 3 > > > Linux, which is about as recent as distros get and SOME Speedtouch > > > > documentation > > > > > says this is already in recent kernels. > > > > What kernel version are you running? (uname -r) > > You don't need to patch on recent kernel but you may need to set > > CONFIG_N_HDLC=m in you kernel .config and recompile your kernel > > > > > > > > Liste de diffusion modem ALCATEL SpeedTouch USB > > Pour se désinscrire : mailto:speedtouch- > > [EMAIL PROTECTED] > > Liste de diffusion modem ALCATEL SpeedTouch USB > Pour se désinscrire : > mailto:[EMAIL PROTECTED] Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] [CVS commits] Various small things
It's been a while since the last commits. This is probably going to be released as beta3. Then i'll do rc releases for a final 1.2. Please test, first reports will decided wheter or not i release that as beta3 :-) 2003-11-17 13:32:57 GMT patch-11 Summary: On the road to beta3 Revision: speedtouch--trunk--1.2--patch-11 * Added SIGHUP and SIGPIPE trapping for GNU/Linux. It seemed SIGTERM wasn't enough. * Removed linux kernel includes. That solves the problem of driver building on distros using 2.6 headers. * Bump up the version numbers. new files: src/.arch-ids/pusb-linux.h.id src/pusb-linux.h modified files: ChangeLog VERSION configure configure.in src/Makefile.in src/pppoa2.c src/pppoa3.c src/pusb-linux.c -- Edouard Gomez Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
[speedtouch] Re: New Firmware information
Hello Gilles, Stef, > Beware that on the new V3.0 of the windows driver alcaudsl.sys does not > contain anymore the firmware. It's KQD6P2.eni (761 389 octets) like in the > files available for download but biger. KQD6P2.eni does not work well with > the user mode driver (I don't know for the kernel driver). Please apply the patches that I submitted via sourceforge and earlier through this mailing list as well. I have been using them for months and they do fix these CRC errors, due to some changes in the byte counting. An older version of this patch can be found at http://home.hetnet.nl/~ottolander/speedtouch/speedtouch.html . I will soon make the new (trimmed) patch and a spec file available for the latest version. Bye, Leonard. -- How clean is a war when you shoot around nukelar waste? Stop the use of depleted uranium ammo! End all weapons of mass destruction. Liste de diffusion modem ALCATEL SpeedTouch USB Pour se désinscrire : mailto:[EMAIL PROTECTED]
