LinuxQuestions.org - intel 536ep modem with suse 9.3 pro

- Linux - Newbie (https://www.linuxquestions.org/questions/linux-newbie-8/)

- - intel 536ep modem with suse 9.3 pro (https://www.linuxquestions.org/questions/linux-newbie-8/intel-536ep-modem-with-suse-9-3-pro-323265/)

Hi Emmanuel_uk,

Thanks for your feedback.

I've e-mailed Klaus regarding the access error issue
during the "make install" to see if he can share some
of his idea.

I'll update this thread accordingly if I hear from him.

Meanwhile, if anybody has any idea/suggestion, your
comments will be greatly appreciated.

Regards,

Hi,
Sorry for being late to respond, <I am actually busy with me exams currently>.

Here's what happend with me.
I Applied the patch . then I did the "make" steps (clean, 536, install).
Something was wrong during the install, similar to what happened with pcandpc. I found out later that there was some confusion in that there was something called /dev/536ep0 while it was supposed to be /dev/536ep .
So I did some manual steps described in that readme file that came with the driver, something like:

Code:

insmod -f Intel536.ko

rm /dev/536ep

mknod /dev/536ep c 240 1

ln -s /dev/536ep /dev/modem

That made the /dev/536ep driver and /dev/modem linking to it. I felt that I acheived something at that point, so I started the kppp, used /dev/modem device, when I tried to connect, the kppp crashed (it started alocating huge amounts of memory endlessly). I also remember that there was a "Query" button in kppp configuration, when I pressed that butten it walked through several steps "succesfully apparently", then displayed a dialog box showing 7 empty strings. (AT1,AT2,...etc).

PS: I tried doing all the steps twice, with and with out commenting the line pm_access(power_dev). I got the same results in both cases.

Hi nawwar,

Thanks for your resourceful updates.

I havn't heard from Klaus yet, but when I do, I'll certainly
update this on-going thread.

Regards,

PcandPC
A user on kernel-devel-2.6.11-1.1366_FC4.i686
has commented on linmodem having to comment
out pm_access as well, so "my fix" is necessary.

I am keeping en eye on the list so if anything props up about suse
I will let you know.

You know that group error, I do not if it matters or if this
can be resolve by changing permissions? Did you look
if /dev/536ep was part of the dial group (or whichever group is needed)?

>>So, I went ahead and tried to initiate the dial-up account but the connection failed

Can you post the error messages?
I suggest you use wvdial rather than kppp, if you can, it is easier to diagnose
problems during dial-up. Mave be nothing is wrong with the driver,
but still you could have a problem with the setup of kppp.
Do the ATI command works (in Kppp)?

You may need to trick wvdialconfig if you want to use it by
creating a few symbolics link to ttyS0 (???) because that the
only ports wvdialconfig can scan.

Hi Emmanuel_uk,

Actually, the /dev/536ep was never created in my case.
I was misled by not correctly reading the directory listing.

As far as my comment on "connection failed", I simply meant
that the connection was never successfully made by observing
the icon on the panel.

I've used the KPPP since 8.2 Pro and quite familiar with its GUI
configuration than any other dial-up utility, but I'll consider the
wvdial per your suggestion.

As it stands now, the ATIs do not matter because my modem is
never detected by 9.3 Pro to begin with.

I just hope I can hear something from Klaus.

Thanks.

PcandPc, did you try Nawwar manual method, i.e.
mknod /dev/536ep c 240 1
before or after the make install (try both?)
A shot in the dark, to see if that changes the error message

I tried wvdial, it failed and printed error message "Segmentation fault"

This is a sinppet from the syslog:

Code:

Jun  8 00:01:49 linux kernel:  <6>ACPI: PCI interrupt 0000:02:09.0[A] -> GSI 21 (level, low) -> IRQ 225

Jun  8 00:01:49 linux kernel: divide error: 0000 [#5]

Jun  8 00:01:49 linux kernel: SMP

Jun  8 00:01:49 linux kernel: Modules linked in: ppp_generic slhc Intel536 thinkpad nvidia speedstep_lib freq_table thermal processor fan button ipv6 battery ac nvram joydev st evdev af_packet sg snd_pcm_oss snd_mixer_oss edd sk98lin ohci1394 ieee1394 uhci_hcd ehci_hcd i2c_i801 usbcore i2c_core intel_agp agpgart hw_random parport_pc lp parport snd_intel8x0 snd_ac97_codec snd_pcm snd_timer snd soundcore snd_page_alloc nls_cp437 vfat fat nls_utf8 ntfs sr_mod dm_mod ext3 jbd ide_cd cdrom ide_disk ata_piix libata piix ide_core sd_mod scsi_mod

Jun  8 00:01:49 linux kernel: CPU:    0

Jun  8 00:01:49 linux kernel: EIP:    0060:[<e1547fc8>]    Tainted: PF    U VLI

Jun  8 00:01:49 linux kernel: EFLAGS: 00210246  (2.6.11.4-20a-smp)

Jun  8 00:01:49 linux kernel: EIP is at SetPLL+0x1fa/0x2a0 [Intel536]

Jun  8 00:01:49 linux kernel: eax: 00000000  ebx: 00000000  ecx: 00000000  edx: 00000000

Jun  8 00:01:49 linux kernel: esi: c7fd0d80  edi: 00000000  ebp: dc49de58  esp: dc49ddf0

Jun  8 00:01:49 linux kernel: ds: 007b  es: 007b  ss: 0068

Jun  8 00:01:49 linux kernel: Process wvdial (pid: 8355, threadinfo=dc49c000 task=dfb30020)

Jun  8 00:01:49 linux kernel: Stack: 0100ef60 c0404928 c03fdb00 00000000 c012a525 dc49de30 dc49de30 005143f8

Jun  8 00:01:49 linux kernel:        c01360c0 dc49c000 c012a577 005143f8 c030ef2c 00000000 00000001 00000000

Jun  8 00:01:49 linux kernel:        01100100 7f000f00 0000f0b7 0000f0b6 4b87ad6e 00200287 c0136080 e1612f20

Jun  8 00:01:49 linux kernel: Call Trace:

Jun  8 00:01:49 linux kernel:  [<c012a525>] del_timer_sync+0x85/0xc0

Jun  8 00:01:49 linux kernel:  [<c01360c0>] autoremove_wake_function+0x0/0x50

Jun  8 00:01:49 linux kernel:  [<c012a577>] del_singleshot_timer_sync+0x17/0x30

Jun  8 00:01:49 linux kernel:  [<c030ef2c>] schedule_timeout+0x6c/0xc0

Jun  8 00:01:49 linux kernel:  [<c0136080>] finish_wait+0x30/0x70

Jun  8 00:01:49 linux kernel:  [<e1547d1f>] IM_PLL_SetDSPFrequency+0x6d/0x76 [Intel536]

Jun  8 00:01:49 linux kernel:  [<e154a3e8>] IM_Power_Down+0x22/0x8e [Intel536]

Jun  8 00:01:49 linux kernel:  [<e1549eef>] IM_DSP_OFF+0xb/0x34 [Intel536]

Jun  8 00:01:49 linux kernel:  [<e1544d52>] clm_terminate+0x52/0x70 [Intel536]

Jun  8 00:01:49 linux kernel:  [<e15441c5>] close+0x5/0x30 [Intel536]

Jun  8 00:01:49 linux kernel:  [<e1546926>] softserial_close+0x26/0x50 [Intel536]

Jun  8 00:01:49 linux kernel:  [<c0240c12>] release_dev+0x402/0x7d0

Jun  8 00:01:49 linux kernel:  [<c023ee7a>] tty_ldisc_try+0x3a/0x50

Jun  8 00:01:49 linux kernel:  [<c011d4c8>] __wake_up+0x38/0x50

Jun  8 00:01:49 linux kernel:  [<c023effb>] tty_ldisc_deref+0x8b/0xb0

Jun  8 00:01:49 linux kernel:  [<c0241ff7>] tty_ioctl+0x237/0x6b0

Jun  8 00:01:49 linux kernel:  [<c01769e0>] __pollwait+0x0/0xd0

Jun  8 00:01:49 linux kernel:  [<c0175f61>] do_ioctl+0x71/0xa0

Jun  8 00:01:49 linux kernel:  [<c024146f>] tty_release+0xf/0x20

Jun  8 00:01:49 linux kernel:  [<c01643e6>] __fput+0x116/0x160

Jun  8 00:01:49 linux kernel:  [<c0162a6f>] filp_close+0x4f/0x90

Jun  8 00:01:49 linux kernel:  [<c0104079>] sysenter_past_esp+0x52/0x79

Jun  8 00:01:49 linux kernel: Code: e0 89 04 24 0f b6 07 e9 af 00 00 00 3b 1d 48 3c 61 e1 76 06 8b 1d 48 3c 61 e1 89 d8 d1 e8 8b 0d 48 3c 61 e1 01 c8 ba 00 00 00 00 <f7> f3 88 45 d9 fe 4d d9 0f b6 5d d9 8d 73 01 89 c8 ba 00 00 00

Hi nawwar and Emmanuel_uk,

Thanks for your consistent interest and feedback.

I'm so impressed! By the way, I'm in the middle of communication with Klaus.
Basically, he suggested the same methods as in Emmanuel_uk's patch link.
This took care of my previous script errors, but still my modem connection is
not made at all, and I conveyed this information to Klaus, and currently, I'm
awaiting for his feedback.

So, Emmanul_uk, you've made the patch working under Mandriva 2005 LE?
Lucky for you!

Besides this modem connection, I'm having other issues here and there on
SuSE 9.3 Pro. And, perhaps, I'll switch to other distro eventually if this modem
connection is not resolved. You know, I can't just try this and that indefinitely.

Again, thanks you all folks.

Regards,

Hi you two,

All I know about segmentation fault is that it _may_ be
a sign of the driver being unsuitable (i.e. code wrong or unsuitable).
Those bl..y patch are a pain when they come to line wrapping.
Can you try to patch by hand and see what the make steps give then?
(one of my mail explain how you can do the patch by hand)

I am sorry this 536ep is causing you so much trouble. As Klaus
had no problem, I assumed it would be the same for you two...
It seems the problem is out of my league.

Unless you can find what is different between Klaus system and yours,
I suggest
- you download scanmodem
- make uninstall (remove the /dev/536* if still there

then "record" all the steps (I mean the command line you used + 1 or 2
output lines if informative, you know what I mean)
- which intel file you are using
- state if you had error or not during the patch
(maybe best you take the patch from the ubuntu website and
disable the pm_access by hand)
- the exact errors with a few context lines during compilation
- the errors using wvdial
- make ref to http://linmodems.technion.ac.il/arc...h/msg02026.html
- make ref to linmodem email dated 04/06/05
"RE: Problem with Intel 536EP and Fedora Core 4 (RC3): Working! (Sort of)"
stating that it is the third confirmation that pm_access needs disabling
with 2.6.11 kernel for the 536ep

then run scanmodem

email all that + modemdata.txt to the discussion list of
http://linmodems.technion.ac.il

These guys there are fantastic, they surely will have further suggestions.
On top it is two of you with the same chip, same distro,
so it should make their life easier

I am looking forward to see what they suggest

If this fail (I do not think so), then this will go direct to
the intel developer

Hi Emmanuel_uk,

It'd be nice if Intel promptly updates their Linux drivers, but
then being Linux issues, I assume that they're under no obligation
at all other than their "voluntary" commitment, which is still appreciative.

By the way, I'm in the middle of communication with Klaus as I stated
in one of my previous posts. Hopefully, things iron out smoothly. His
suggestion is no different from what you, Emmanuel_uk, suggested in
the patch 1 link.

After all, he appears to be the one who made this modem work under
SuSE 9.3 Pro, and I'm diligently following up with him.

I'll certainly update this thread whether the result is successful or otherwise.

As a last resort, I'll consider working with the Linmodem.org guys. But then,
if I decide to switch over to the other distros, this is it for me.

So, the whole deal here is the kernel 2.6.10 and later with Intel 536EP modem
driver on SuSE 9.3 Pro, huh? Well, I don't have this issue at all under FC3 with
the same latest Intel driver. I suppose this is because FC3's kernel version is
2.6.9-something.

Hopefully, things work out for good with Intel drivers for kernels 2.6.10 and later
sooner or later.

Regards,

Hi,

I've been working with Klaus on this modem issue
since he appears to have made the modem working
on SuSE 9.3.

He was very helpful, and I made my modem working
the first time ever after reinstalling SuSE and carefully
following his suggestions.

His suggestions were no different from the earlier ones
by Emmanuel_uk on applying the patch.

However, my working with the modem didn't last long.
After having made work the first time, restarting the
modem conncetion failed. Since then, the modem hasn't
worked at all -- back to square one.

Now, I've decided to replace this SuSE with another
distro and sadly given up on this issue.

Hopefully, there'll be newer, officialy driver by Intel soon.

Thanks all for your efforts and helpful suggestions.

Regards,

Thanks pcandps for your follow-up,

However, I've decided to change my modem, I'm thinking of an external one.

Thanks again everyone.

-Nawwar

Hi all,

If anyone's following this issue, I have a semi-bright news!

I've come back to this issue after some time lapse.

The 536EP modem is detected successfully under SuSE 9.3 Pro
when the DSL modem connection is on simultaneously.

As soon as the DSL modem is off, the modem wouldn't be able
to be detected nor would it work.

This is sad ...

Hi all,

Good news at last!

Intel appears to have released an updated
driver (4.71) that works nicely. I just tried it.

But, it appears the modem driver is not loaded
on each reboot ...

Please anyone look into this and advise us how
to fix the driver load on each reboot.

Thanks.

This is good news indeed.
re starting up, what about editing
/etc/modprobe.preload
# /etc/modprobe.preload: kernel modules to load at boot time.