Since I could get connected I decided to update to a newer kernel:
=============================================================================
Package Arch Version Repository Size
=============================================================================
Installing:
kernel x86_64 2.6.25.4-30.fc9 updates 18 M
kernel-devel x86_64 2.6.25.4-30.fc9 updates 5.2 M
Updating:
kernel-headers x86_64 2.6.25.4-30.fc9 updates 711 k
Transaction Summary
=============================================================================
Install 2 Package(s)
Update 1 Package(s)
Remove 0 Package(s)
Total download size: 24 M
Is this ok [y/N]: y
Downloading Packages:
I have installed it and rebooted and all is well. It is incredible
and I feel very happy and thankful to all who have been involved. I
could not connect and was getting dissapointed. Thanks to grujakg for
pointing out the Makefile part of the package. IT was what I needed
to successfully connect from Fedora. I'll attach ModemData.txt so
that it can be helpful also, I kindly ask Marv to change the driver to
at least martian-full-20080407.tar.gz instead of the recommended
martian-full-20071011.tar.gz. Not a big thing, but more useful to new
users which may encounter this modem. Ubuntu users will continue with
2.6.24-X kernels as it seems like, unless of course the users that
test their systems as technology advances.
This part
\begin{quote}
The modem has a Lucent/Agere/LSI Mars or Apollo DSP (digital signal processing)
chipset.
Support packages for 2.6.n kernels are at:
http://linmodems.technion.ac.il/packages/ltmodem/kernel-2.6/martian/ , with cur
rent update martian-full-20071011.tar.gz
See DOCs/AgereDSP.txt for Details.
\end{quote}
Regards,
Antonio
On Mon, Jun 9, 2008 at 6:15 PM, Antonio Olivares
<olivares14031@xxxxxxxxx> wrote:
> In looking for a response to what they mean/do, the following message
> might explain things a bit
>
> http://lkml.org/lkml/2007/9/30/114
>
> The thread below also illustrates this, I guess it was done to make
> the code more functional
>
> http://www.ussg.iu.edu/hypermail/linux/kernel/0709.3/2581.html
>
> In my humble opinion, the Makefile found in kmodule directory after
> martian-200804047.tar.gz is extracted is the one. However, it needs
> to be tested across several distributions and has to work for most of
> them if not all of them. I know that it works in at least two of the
> many distributions, Slax and Fedora namely.
>
> Regards,
>
> Antonio
>
> On Mon, Jun 9, 2008 at 5:30 PM, Marvin Stodolsky
> <marvin.stodolsky@xxxxxxxxx> wrote:
>> RE:
>> EXTRA_CFLAGS += -DKMARTIAN_STAMP=20061202
>> EXTRA_CPPFLAGS += -DKMARTIAN_STAMP=20061202
>>
>> I just need the syxtax for alternate choosing between these tol
>> depending on SUBLEVEL
>> EXTRA_CFLAGS and EXTRA_CPPFLAGS specifications during the compile,
>> the result on compiling for the different SUBLEVELs of the kernels is obvious.
>>
>> I'll do the martian-full-20080407 update and send it out for testing
>> once the Conditional syntax is resolved.
>>
>> MarvS
>>
>> While not fully understanding what the difference is between
>>
>>
>> On Mon, Jun 9, 2008 at 5:00 PM, Antonio Olivares
>> <olivares14031@xxxxxxxxx> wrote:
>>> It is the other way around Jacques. the EXTRA_CPPFLAGS is for kernels
>>> 2.6.24 and above while the other is for previous kernels pre 2.6.24.
>>> Take a look at grujakg's message in the thread while I will quote
>>>
>>> \begin{QUOTE}
>>>
>>> You would also need to keep in mind the patch for Makefile already
>>> applied to martian-full-20080407.
>>>
>>> diff -u -w -r martian-full-20080407/kmodule/Makefile
>>> martian/kmodule/Makefile
>>> --- martian-full-20080407/kmodule/Makefile 2008-04-08
>>> 02:52:13.000000000 +0200
>>> +++ martian/kmodule/Makefile 2006-12-03 00:39:24.000000000 +0100
>>> @@ -1,14 +1,8 @@
>>> KRELEASE = $(shell uname -r)
>>> -
>>> -ifdef KERNEL_DIR
>>> -KBUILD_DIR = $(KERNEL_DIR)
>>> -else
>>> KBUILD_DIR = /lib/modules/$(KRELEASE)/build
>>> -endif
>>> -
>>> MSRC_DIR = $(shell pwd)
>>>
>>> -EXTRA_CFLAGS += -DKMARTIAN_STAMP=20061202
>>> +EXTRA_CPPFLAGS += -DKMARTIAN_STAMP=20061202
>>> CPPFLAGS += -DKMARTIAN_STAMP=20061202
>>>
>>> # kbuild
>>>
>>> Think that EXTRA_CFLAGS make a difference.
>>> \end{QUOTE}
>>>
>>> Regards,
>>>
>>> Antonio
>>>
>>> On 6/9/08, Jacques Goldberg <Jacques.Goldberg@xxxxxxx> wrote:
>>>> Marv,
>>>> I will try tomorrow but let me please candidly ask:
>>>> Is EXTRA_CPPFLAGS used for kernels < 24 and is EXTRA_CFLAGS used with 24
>>>> and above?
>>>> This would mean that not the same compiler is used by the Makefile, or
>>>> that I missed something.
>>>> Why couldn't you then just define both?
>>>> Or do you have a typo and mean to define a different value for
>>>> DKMARTIAN_STAMP in the same variable EXTRA_CFLAGS or EXTRA_CPPFLAGS?
>>>>
>>>> Jacques
>>>>
>>>> Marvin Stodolsky wrote:
>>>>> Folks
>>>>>
>>>>> Is any erudite enough in Makefile syntax to help with a needed
>>>>> Conditional.
>>>>>
>>>>> With SUBLEVEL being the third component of the Kernel-Version
>>>>> which can be read out with
>>>>> SUBLEVEL := $(shell uname -r | cut -d. -f3 | cut -d. -f1 | cut -d- -f1)
>>>>> with cogency in the 22 24+ range
>>>>>
>>>>> a Makefile conditional is needed to use
>>>>> EXTRA_CFLAGS += -DKMARTIAN_STAMP=20061202
>>>>> for SUBLEVEL < 24 ,
>>>>> and for greater or equal 24
>>>>> EXTRA_CPPFLAGS += -DKMARTIAN_STAMP=20061202
>>>>>
>>>>> Otherwise we would need separate martian packages for 2.6.24 and later
>>>>> kernels
>>>>> versus 2.6.22 and earlier
>>>>>
>>>>> Sadly I'm not adequately knowledgeable in Makefile syntax.
>>>>>
>>>>> Incidentally, in the kmodule/Makefile , the role of this section:
>>>>>
>>>>> ifdef KERNEL_DIR
>>>>> KBUILD_DIR = $(KERNEL_DIR)
>>>>> else
>>>>> KBUILD_DIR = /lib/modules/$(KRELEASE)/build
>>>>> endif
>>>>>
>>>>> is to enable compiling for other kernels, other than the current boot
>>>>> kernel.
>>>>> The syntax is like:
>>>>> $ make KERNEL_DIR=/usr/src/linux-headers-SomeKernelVersion
>>>>> that is
>>>>> KERNEL_DIR is the path to the linux-headers for the desired kernel
>>>>> version.
>>>>> This saves rebooting when one wants to test compiling for several
>>>>> different linux-headers
>>>>>
>>>>> IIf the command is only:
>>>>> $ make
>>>>> then there is used the default /lib/modules/kernel-version-bootup/build
>>>>>
>>>>> Thus I will retain this section in the update.
>>>>>
>>>>> MarvS
>>>>>
>>>>> On Sat, Jun 7, 2008 at 3:08 PM, Ivica Grujic <grujakg@xxxxxxxxx> wrote:
>>>>>> Antonio I'm thinking that others should know about your good deed, so
>>>>>> I've
>>>>>> posted this on the thread also ;)
>>>>>>
>>>>>> You would also need to keep in mind the patch for Makefile already
>>>>>> applied to martian-full-20080407.
>>>>>>
>>>>>> diff -u -w -r martian-full-20080407/kmodule/Makefile
>>>>>> martian/kmodule/Makefile
>>>>>> --- martian-full-20080407/kmodule/Makefile 2008-04-08
>>>>>> 02:52:13.000000000 +0200
>>>>>> +++ martian/kmodule/Makefile 2006-12-03 00:39:24.000000000 +0100
>>>>>> @@ -1,14 +1,8 @@
>>>>>> KRELEASE = $(shell uname -r)
>>>>>> -
>>>>>> -ifdef KERNEL_DIR
>>>>>> -KBUILD_DIR = $(KERNEL_DIR)
>>>>>> -else
>>>>>> KBUILD_DIR = /lib/modules/$(KRELEASE)/build
>>>>>> -endif
>>>>>> -
>>>>>> MSRC_DIR = $(shell pwd)
>>>>>>
>>>>>> -EXTRA_CFLAGS += -DKMARTIAN_STAMP=20061202
>>>>>> +EXTRA_CPPFLAGS += -DKMARTIAN_STAMP=20061202
>>>>>> CPPFLAGS += -DKMARTIAN_STAMP=20061202
>>>>>>
>>>>>> # kbuild
>>>>>>
>>>>>> Think that EXTRA_CFLAGS make a difference.
>>>>>>
>>>>>> This Makefile is all I had to keep, not to build but to successfully run
>>>>>> martian.
>>>>>>
>>>>>> Thanks again for great work.
>>>>>>
>>>>>> Regards
>>>>>>
>>>>>> grujakg
>>>>>>
>>>>>> Antonio Olivares wrote:
>>>>>>> If compiling fails, there is a file that asks for # linux/ioctl32.h>
>>>>>>>
>>>>>>> comment it out and the compilation will succeed.
>>>>>>> /* # linux/ioctl32.h> */
>>>>>>>
>>>>>>> I forgot to mention this in the previous emails.
>>>>>>>
>>>>>>> Regards,
>>>>>>>
>>>>>>> Antonio
>>>>>>>
>>>>>>> --- On Sat, 6/7/08, Ivica Grujic <grujakg@xxxxxxxxx> wrote:
>>>>>>>
>>>>>>>> From: Ivica Grujic <grujakg@xxxxxxxxx>
>>>>>>>> Subject: Re: Fw: [Fwd: [Fwd: Re: Another Patch will be needed for
>>>>>>>> kernel
>>>>>>>> version 2.6.25]]
>>>>>>>> To: olivares14031@xxxxxxxxx
>>>>>>>> Date: Saturday, June 7, 2008, 8:27 AM
>>>>>>>> Yes, I got it.
>>>>>>>>
>>>>>>>> Thanks a lot.
>>>>>>>>
>>>>>>>> Regards,
>>>>>>>>
>>>>>>>> grujakg
>>>>>>>>
>>>>>>>> Antonio Olivares wrote:
>>>>>>>>> Done,
>>>>>>>>>
>>>>>>>>> please let me know if you receive it without problems.
>>>>>>>> :)
>>>>>>>>> Regards,
>>>>>>>>>
>>>>>>>>> Antonio
>>>>>>>>>
>>>>
>>>
>>
>
Only plain text email is forwarded by the Discuss@xxxxxxxxxxxxx List Server,
as HTML can contain viruses. Use as the email Subject Line:
YourName, YourCountry Fedora release 9 (Sulphur)
Kernel kernel 2.6.25.4-30.fc9.x86_64
With this Subject Line cogent experts will be alerted, and useful case names left in the Archive.
YourCountry will enable Country specific guidance. Linux experts in YourCountry
can be found through: http://www.linux.org/groups/index.html.
They will know your Country's modem code, which may be essential for dialup service.
Responses from Discuss@xxxxxxxxxxxxx are sometimes blocked by an Internet Provider mail filters.
So in a day, also check the Archived responses at http://www.linmodems.org
-------------------------- System information ----------------------------
CPU=x86_64, Fedora release 9 (Sulphur)
Kernel
Linux version 2.6.25.4-30.fc9.x86_64 (mockbuild@) (gcc version 4.3.0 20080428 (Red Hat 4.3.0-8) (GCC) ) #1 SMP Wed May 21 17:34:18 EDT 2008
scanModem update of: 2008_06_05
Some modem drivers can only be used in 32 bit modem on x86_64 systems,
while some others are competent on x86_64 Systems. Cases are:
1) http://linmodems.technion.ac.il/bigarch/archive-seventh/msg03119.html
for the snd-hda-intel audio+modem driver. Also applicable to AC97 modem controllers.
In both cases, 32 bit libraries must be installed to support the slmodemd helper having a precompiled 32 bit component.
2) For USB modems using the slusb.ko driver. 32 bit libraries must be installed to support the slmodemd helper having a precompiled 32 bit component
3) The hsfmodem and hcflinmodem drivers for Conexant chipsest modes are x86_64 competent.
There are no blacklisted modem drivers in /etc/modprobe* files
Attached USB devices are:
ID 1d6b:0002 Linux Foundation 2.0 root hub
ID 1d6b:0001 Linux Foundation 1.1 root hub
ID 1d6b:0001 Linux Foundation 1.1 root hub
ID 1d6b:0001 Linux Foundation 1.1 root hub
ID 1d6b:0001 Linux Foundation 1.1 root hub
USB modems not recognized
For candidate card in slot 04:05.0, firmware information and bootup diagnostics are:
PCI slot PCI ID SubsystemID Name
---------- --------- --------- --------------
04:05.0 11c1:044e 11c1:044c Communication controller: Agere Systems LT WinModem
Modem interrupt assignment and sharing:
18: 67228 IO-APIC-fasteoi 164x
--- Bootup diagnostics for card in PCI slot 04:05.0 ----
ACPI: PCI Interrupt 0000:04:05.0[A] -> GSI 18 (level, low) -> IRQ 18
For candidate card in slot 80:01.0, firmware information and bootup diagnostics are:
PCI slot PCI ID SubsystemID Name
---------- --------- --------- --------------
80:01.0 1106:3288 1462:7253 Audio device: VIA Technologies, Inc. VIA High Definition Audio Controller
Modem interrupt assignment and sharing:
17: 1096 IO-APIC-fasteoi HDA Intel
--- Bootup diagnostics for card in PCI slot 80:01.0 ----
ACPI: PCI Interrupt 0000:80:01.0[A] -> GSI 17 (level, low) -> IRQ 17
PCI: Setting latency timer of device 0000:80:01.0 to 64
===== Advanced Linux Sound Architecture (ALSA) diagnostics =====
The ALSA packages provide audio support and also drivers for some modems.
ALSA diagnostics are written during bootup to /proc/asound/ folders.
PCI slot 80:01.0 has a High Definition Audio Card
---ALSA bootup diagnostics ---
The ALSA verion is 1.0.16
The modem cards detected by "aplay -l" are: None
The /proc/asound/pcm file reports:
-----------------------
00-02: ALC883 Analog : ALC883 Analog : capture 1
00-01: ALC883 Digital : ALC883 Digital : playback 1
00-00: ALC883 Analog : ALC883 Analog : playback 1 : capture 1
about /proc/asound/cards:
------------------------
0 [VT82xx ]: HDA-Intel - HDA VIA VT82xx
HDA VIA VT82xx at 0xbfffc000 irq 17
=== Finished firmware and bootup diagnostics, next deducing cogent software. ===
Predictive diagnostics for card in bus 04:05.0:
Modem chipset detected on
CLASS="Class 0780: 11c1:044e"
NAME="Communication controller: Agere Systems LT WinModem "
SUBSYS=11c1:044c
PCIDEV=11c1:044e
IRQ=18
IDENT=Agere.DSP
For candidate modem in: 04:05.0
Class 0780: 11c1:044e Communication controller: Agere Systems LT WinModem
Primary device ID: 11c1:044e
Support type needed or chipset: Agere.DSP
----------------end Softmodem section --------------
The modem has a Lucent/Agere/LSI Mars or Apollo DSP (digital signal processing) chipset.
Support packages for 2.6.n kernels are at:
http://linmodems.technion.ac.il/packages/ltmodem/kernel-2.6/martian/ , with current update martian-full-20071011.tar.gz
See DOCs/AgereDSP.txt for Details.
Predictive diagnostics for card in bus 80:01.0:
Modem chipset not detected on
CLASS="Class 0403: 1106:3288"
NAME="Audio device: VIA Technologies, Inc. VIA High Definition Audio Controller "
SUBSYS=1462:7253
PCIDEV=1106:3288
IRQ=17
HDA=1106:3288
SOFT=1106:3288.HDA
High Definition Audio (HDA) cards MAY host a modem chip in their Subsystem,
and many are supported by the ALSA audio+modem driver snd-hda-intel
A modem was not detected on HDA card 1106:3288.
If another modem card is present, then most likely 1106:3288 does not host a modem.
If another modem card has not been detected, then possibilities are:
1) A Conexant modem chip is present on 1106:3288, as Conexant chips
are frequently not detectable by ALSA diagnostics
2) The modem may be of the older non-PCI Controller Chipset (hardware) type.
Try detection with Root permission:
wvdialconf /etc/wvdial.conf
For candidate modem in: 80:01.0
Class 0403: 1106:3288 Audio device: VIA Technologies, Inc. VIA High Definition Audio Controller
Primary device ID: 1106:3288
Subsystem PCI_id 1462:7253
Softmodem codec or chipset from diagnostics:
from Archives:
Support type needed or chipset:
Support can likely be achieved through two mutually exclusive alternatives:
1) The hsfmodem software for Conexant chipset modems: Read DOCs/Conexant.txt
The following ALSA alternative CANNOT work with Conexant modems.
2) An ALSA modem driver plus slmodemd. Read DOCs/Smartlink.txt for details, and
to test get the package SLMODEMD.gcc4.3.tar.gz from:
http://linmodems.technion.ac.il/packages/smartlink/
----------------end Softmodem section --------------
Writing DOCs/Smartlink.txt
============ end Smartlink section =====================
Completed candidate modem analyses.
The base of the UDEV device file system is: /dev/.udev
Versions adequately match for the compiler installed: 4.3.0
and the compiler used in kernel assembly: 4.3.0
Minimal compiling resources appear complete:
make utility - /usr/bin/make
Compiler version 4.3
linuc_headers base folder /lib/modules/2.6.25.4-30.fc9.x86_64/build
However some compilations and executable functions may need additional files,
in the FileNames.h (so called kernel "h"eaders) collection installed in /usr/include/ .
For martian_modem, additional required packages are libc6-dev (and for Debian/Ubuntu,
linux-libc-dev). The also required headers of package libc6 are commonly installed by default.
Compiling hsfmodem drivers does require linux-libc-dev and libc6-dev packages, for kernels 2.6.24 and later versions.
In not included on your install CD, search for them at http://packages.ubuntu.com
or comparable Repository for other Linux distros.
When compiling ALSA drivers, the utility "patch" will also be needed.
If a driver compilation fails, with message including some lack of some FileName.h (stdio.h for example), then
Some additional kernel-header files need installation to /usr/include. The minimal additional packages are libc6-dev
and any of its dependents, under Ubuntu linux-libc-dev
If an alternate ethernet connection is available,
$ apt-get update
$ apt-get -s install linux-kernel-devel
will install needed packages.
For Debian/Ubuntu related distributions, run the following command to display the needed package list:
Otherwise packages have to be found through http://packages.ubuntu.com
Once downloaded and transferred into a Linux partition,
they can be installed alltogether with:
$ sudo dpkg -i *.deb
Checking pppd properties:
-r-xr-xr-x 1 root root 349064 2008-05-13 04:23 /usr/sbin/pppd
In case of an "error 17" "serial loopback" problem, see:
http://linmodems.technion.ac.il/linmodems/archive-sixth/msg02637.html
To enable dialout without Root permission do:
$ su - root (not for Ubuntu)
chmod a+x /usr/sbin/pppd
or under Ubuntu related Linuxes
chmod a+x /usr/sbin/pppd
Checking settings of: /etc/ppp/options
asyncmap 0
crtscts
lock
modem
proxyarp
lcp-echo-interval 30
lcp-echo-failure 4
In case of a message like:
Warning: Could not modify /etc/ppp/pap-secrets: Permission denied
see http://linmodems.technion.ac.il/bigarch/archive-sixth/msg04656.html
Read Modem/DOCs/YourSystem.txt concerning other COMM channels: ppp0
Which can interfere with Browser naviagation.
Don't worry about the following, it is for experts should trouble shooting be necessary.
==========================================================
Checking for modem support lines:
--------------------------------------
/device/modem symbolic link:
slmodemd created symbolic link /dev/ttySL0:
Within /etc/udev/ files:
/etc/udev/rules.d/50-udev-default.rules:KERNEL=="mwave", NAME="modems/mwave", GROUP="uucp"
Within /etc/modprobe.conf files:
Within any ancient /etc/devfs files:
Within ancient kernel 2.4.n /etc/module.conf files:
--------- end modem support lines --------
