Linmodems Mailing list Archives


Return-Path: <>
Mailing-List: contact; run by ezmlm
Delivered-To: mailing list
Received: (qmail 8316 invoked from network); 16 Apr 2004 21:19:45 -0000
Received: from (
  by with SMTP; 16 Apr 2004 21:19:45 -0000
Received: from ([]
	by with esmtp (Exim 3.35 #7)
	id 1BEakM-00003T-00; Fri, 16 Apr 2004 17:19:39 -0400
Message-ID: <>
Date: Fri, 16 Apr 2004 17:19:37 -0400
From: Marvin Stodolsky <>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.0) Gecko/20020623 Debian/1.0.0-0.woody.1
X-Accept-Language: en
MIME-Version: 1.0
To: Robert Aldridge <>
Subject: Re: New laptop owner...
References: <>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit


Class 0703: 1002:434d   Modem: ATI Technologies Inc: Unknown device 434d (rev 01) (prog-if 00 [Generic])
  SubSystem 103c:006b   Hewlett-Packard Company: Unknown device 006b
	Flags: bus master, 66MHz, slow devsel, latency 64, IRQ 5
	Memory at d0003400 (32-bit, non-prefetchable) [size=256]

 The soft modem Subsystem operates under a controller
     1002:434d  ATI   with Subsystem chipset POSSIBLY from:

The  Modem: ATI Technologies  is a controller for an 1002:434d 
Subsystem, possibly an Intel 537 type , but could be a variety of others.
So let's try the Intel 537 first.  I happen to have those resources on 
this laptop, so I'll send them to you.

Re Modem symbolic link is: /dev/modem -> /dev/ttyS0
This is only for serial port modems, so remove the symbolic link, as Root
# rm /dev/modem

----active COMM services are ------------

eth0      Link encap:Ethernet  HWaddr 00:02:3F:6E:95:98 

This will conflict with DomainNameService during modem usages. So before 
using the modem
# ifconfig
# ifconfig eth0 down
# ifconfig
to check

Install the wvdial.deb package for later modem testing.

scanModem maintainer

Robert Aldridge wrote:

>Hello all.  I'm a new owner of an HP Pavilion zv5120us
>laptop and am running Debian Sid on it.  I need to get
>the modem working from Linux. I ran scanModem which
>wrote out the two files ModemData.txt and
>ModemData.txt.2.  The Linmodems support page says,
>"These files are frightening for a newcomer, so if you
>do not understand them, send ModemData.txt and
>ModemData.txt.2 to our discussion list  and volunteers
>will help you."
>So, that's what I'm doing.  My data files are
>attached.  I would appreciate any help you can
>Robert Aldridge
>Do you Yahoo!?
>Yahoo! Tax Center - File online by April 15th
> The most recent update of scanModem accessed at
> should ALWAYS be used!!  Current  UPDATE=2004 April 2
> URLs to cogent advice are regularly entered and updated, and your problem may be solved therein.
> If further help is needed, send a description of your problem to:
> with this ModemData.txt Attached, using the following in the email Subject Line:
>      scanModem, debian kernel 2.4.25-fe-dm1-ar
> Always send this ENTIRE ModemData.txt. It includes subtle diagnostic ouputs needed to best guide you.
> If your PC is a laptop, please provide Make and Modem information.
> For instructions to UNSUBSCRIBE from,
>   send an email to:
> --------------  System information ------------------------
> debian unstable/experimental
> on System with processor: i686
> supported by kernel:      2.4.25-fe-dm1-ar
> assembled with compiler:  3.3.3
> with current System compiler GCC=3.3.3
>    /usr/bin/gcc -> gcc-3.3
> Modem symbolic link is:  /dev/modem -> /dev/ttyS0
>----active COMM services are ------------
>eth0      Link encap:Ethernet  HWaddr 00:02:3F:6E:95:98  
>  There are Debian packages with modem drivers from SmartLink:
>      sl-modem-daemon - SmartLink software modem daemon
>      sl-modem-source - SmartLink software modem driver - module building source
>  MANY modem subSystems serving under AC'97/MC97 Controllers are also supported.
> An AC'97 modem codec was not detected
>Path to lspci is: /usr/bin/lspci
>--------- lspci scan ----------------
>0000:00:00.0 Host bridge: ATI Technologies Inc: Unknown device 5833 (rev 02)
>0000:00:01.0 PCI bridge: ATI Technologies Inc: Unknown device 5838
>0000:00:13.0 USB Controller: ATI Technologies Inc: Unknown device 4347 (rev 01)
>0000:00:13.1 USB Controller: ATI Technologies Inc: Unknown device 4348 (rev 01)
>0000:00:14.0 SMBus: ATI Technologies Inc ATI SMBus (rev 16)
>0000:00:14.1 IDE interface: ATI Technologies Inc: Unknown device 4349
>0000:00:14.3 ISA bridge: ATI Technologies Inc: Unknown device 434c
>0000:00:14.4 PCI bridge: ATI Technologies Inc: Unknown device 4342
>0000:00:14.5 Multimedia audio controller: ATI Technologies Inc SoundMAX Integrated Digital Audio
>0000:00:14.6 Modem: ATI Technologies Inc: Unknown device 434d (rev 01)
>0000:01:05.0 VGA compatible controller: ATI Technologies Inc: Unknown device 5835
>0000:02:02.0 Network controller: Broadcom Corporation BCM94306 802.11g (rev 03)
>0000:02:03.0 Ethernet controller: Realtek Semiconductor Co., Ltd. RTL-8139/8139C/8139C+ (rev 10)
>0000:02:04.0 CardBus bridge: Texas Instruments PCI1250 PC card Cardbus Controller (rev 01)
>0000:02:04.1 CardBus bridge: Texas Instruments PCI1250 PC card Cardbus Controller (rev 01)
>0000:02:07.0 USB Controller: NEC Corporation USB (rev 43)
>0000:02:07.1 USB Controller: NEC Corporation USB (rev 43)
>0000:02:07.2 USB Controller: NEC Corporation USB 2.0 (rev 04)
>Modem candidates are at PCI_buses:  0000:00:14.6
>Providing detail for device at PCI_bus 0000:00:14.6
>  with vendor-ID:device-ID
>	    ----:----
>Class 0703: 1002:434d   Modem: ATI Technologies Inc: Unknown device 434d (rev 01) (prog-if 00 [Generic])
>  SubSystem 103c:006b   Hewlett-Packard Company: Unknown device 006b
>	Flags: bus master, 66MHz, slow devsel, latency 64, IRQ 5
>	Memory at d0003400 (32-bit, non-prefetchable) [size=256]
> Internal_test:   ./scanModem test 1002:434d 103c:006b debian 2.4.25-fe-dm1-ar 3.3.3 3.3.3
> The soft modem Subsystem operates under a controller
>     1002:434d  ATI   with Subsystem chipset POSSIBLY from:
>	Intel
> First accessing chipset provider and modem assembler information
> for the modem SubSystem under the AC'97 or MC97 Controller.
> === Checking 103c through chipset providers and modem assemblers ====
> Vendor=8086 is Intel, Inc. producing HaM and 536ep host controller free (HCF) modems, 537 soft modem
> and AC'97 and MC97 controllers managing a varierty of non-Intel soft modem subSystems.
> These subSystems will in have PCI_IDs assigned by the modem assembler, not Intel.
> Download available drivers from:
> Also check at:
> for beta releases and perhaps Already compiled drivers for some Linux distributions
> ---------------------
> The installer code candidate is:
>	Intel-537AA-   for MDC controllers
>	Intel-537EA-   for CNR controllers
> But the Intel 537 readout needs confirmation !!
> Sadly there are a few mis-designations in the pciids dataset, which ./scanModem relies on
> Too many modems are being labeled Intel537 types.
> This includes the distinct Agere, Conexant and SmartLink chipsets types. Conseqeuntly
> YOU must manually obtain further information under Microsoft Windows through modem ATI tests/queries
> as instructed in ModemData.txt.2  Please include this information in any report to
> === Checking 1002 through chipset providers and modem assemblers ====
>  ======= PCI_ID checking completed ====== 
> Update=2004 April 2
>The following information blocks just query some ppp support items.
>   grep ppp /etc/modules.conf
>alias char-major-108	ppp_generic
>alias /dev/ppp		ppp_generic
>alias tty-ldisc-3	ppp_async
>alias tty-ldisc-14	ppp_synctty
>alias ppp-compress-21	bsd_comp
>alias ppp-compress-24	ppp_deflate
>alias ppp-compress-26	ppp_deflate
> Be sure to read the Ethernet section of ModemData.txt.2 
>DEVPPP=crw-rw---- 1 root dip 108, 0 Dec 11 09:19 /dev/ppp
>  The current modem symbolic link is: /dev/modem -> /dev/ttyS0
>  The ports /dev/ttyS0 or 1,2,3 are for standard Controller chip modems
> No devfsd.conf file found, indicated absense of the devfsd daemon package
> for device file system (devfs) symbolic link support.
> ---- dmesg queries -------
>PCI: if you experience problems, try using option 'pci=noacpi' or even 'acpi=off'
>apm: BIOS not found.
> The ModemDriverCompiling.txt  is a MUST READ,
> if you are not experienced in configuring kernel-source/
> or get "unresolved symbols" upon driver insertion.
>  Most recent WinModem fixes are in:
> Do NOT send this  ModemData.txt.2  with common information to
> For instructions to UNSUBSCRIBE from,
>   send an email to:
> ==============
> This ./scanModem service is sadly  necessary because of the pre-compiled format
> (binary, Closed Source)  of some modem supporting code. Intellectual property
> is protected within the binary component.  But this prevents compatibility
> assessments/debugging against Linux Open Source code. Skipping many details,
> the consequence is that some Linux distributions do not deliver Closed Source code,
> or tools adequate to fully identify the modem's chipset.
> This script tries to close  this information gap.
>  Several sections are included in this file:
>   The Summary of modem Support under Linux
>   description of the soft modem identity tests: SIL ids
>   some output from scanModem specific to your System
>   an explanation of the GCC 2.95 compiler warning
>   ethernet Interference with DNS (domain name service)
>   Please Linux Newbies read carefully, the companion
>   ModemDriversCompiling.txt and its
>     Followup Actions section  with dialout examples.
> USB modems will not be detected through the
>   acm.o driver with port /dev/ttyACM0
>   UNTIL proper USB support is installed.
>   See,
>   Linux_kernel_source/Documentation/usb/acm.txt, and
> (page 15)
> Among the USB modems with proprietary interfaces,
> provides Very Good support for
>   SmartUSB56 (ST7554) based modems with vendor/product usb ids: 0483:7554:
> is an installation report.
>   To write out Smartlink information, use the proxy entry
>     ./scanModem test 1131:3400
> PCMCIA card modems can serve if your laptops modem is not supported under Linux.
> The 3COM,Inc 3CXM556 and AgereSystems FM560LK chip modems use the Open Source serial-cs.o driver.
>It is the CHIPSET, and Not the modem BrandName which is informative for Linux support.
> For modems not recognized by ./scanModem, chipset information may be obtained under Microsoft Windows through:
> 1) Start > Settings > Control Panel > Classical View (for WinXP) > Modem  , or alternatively,
>   Start > Settings > Control Panel > Classical View (for WinXP) > System > Devices > Modem
>   Do the modem diagnostics test, if available, looking for manufacturer chipset,
>   It is within information output in lines:  ATI - specifications
>      Try to identify the modem setup file, with name perhaps MODEM.INF
> 2) Open a COMM console, and send ATI commands to the modem (ATI, ATI1, ATI2, etc)
>   which may elicit chipset and driver information. Here is an example
>       ATI3 - Agere SoftModem Version 2.1.22
>       ATI5 - 2.1.22, AMR Intel MB, AC97 ID:SIL REV:0x27
>   successfully identifying an Agere SoftModem chipset, both by name and through
>   the:softmodem SIL ID:              AC97 ID:SIL REV:0x27
> SUPPORT SUMMARY - as of 2004 April 2
> -------------------------------------
> Controller chipsets are used in the most expensive modems and are supported.
>  They utilize the same driver, serial.o , and most commonly the ports /dev/ttyS0 thru 3.
>  Such chipsets are produced by 3Com, AgereSystems (the Venus chipset),
>  Topic Semiconductor Corp., and others. But the same companies may
>  produce unsupported WinModems. Do not depend on the Brand Name.
>  Look for an explicit statement of Linux support
> AVOID the following modem chipsets:
>   ESS -       no formal support since 2.2.2 kernels, though there are kluges:
>   3Com/US Robotics winmodems - never supported under Linux
> SmartLink - newer chipsets are supported:
>   Moreover, the slmodem-2.9.n  drivers will support Many but not all modem Ssubsystems
>   that serve under AC'97/MC97 controllers.
> Conexant -  all modems supported with drivers at:
>   Testcode for slow 14,400 mH support is free, BUT
>   there is a one time charge for the full speed support.
> Intel Inc. -
>   For early releases check at:
>    The HaM modem is supported, but code is no longer being updated;
>      there will likely be failures under emerging 2.6 kernels.
>    The 536ep and five Intel537 modem chipset variants are actively supported.
> Lucent /AgereSystems
>   The Venus chipset modem uses the serial.o driver and is thus supported.
>   Modems with digital signal processing (DSP) chipsets are supported:  http:/
>   For AC'97/MC97 soft modems. Initiate software access through code sponsor IBM:
>   Coding of support for has Not yet begun PCI soft modems with PCI_IDs 11c1:048?
>    and probably will first be available within 2004.
> Motorola support is at:
>   through 2.4.22 kernels , but there have NO success reports as yet.
> PCTel modem business has been sold to Conexant
>   Driver compiler resources are at:
>  However, the pre-compiled binary component is yet to be assembled
>   with the gcc 3.nn used in newer Linux releases,
>   so forced (-f) insertions may be necessary:
>     insmod -f pctel
>   Some/all? of the HSP soft modems are supported by the SmartLink slmdm-2.7.10 or slmodem-2.9.n drivers.
> Broadcom modems in Dell laptops and PCs - can be made to function under 2.4 kernels, but code hasn't been updated recently.
> --------- end chipset section ----------------------
>  SUPPORT under the emerging 2.6.nn kernels.
> The more expensive Controller chipset modems are functioning.
> SmartLink slmodem-2.9.n drivers are functioning.
> Conexant has 2.6.n support
> The Lucent/Agere digital signal processing (DSP) chipsets do have
>   informal service into 2.6.n kernels and an AgereSystems release is soon expected.
> Intel, Agere Systems and Linuxant are firmly expected to continue support into 2.6.nn,
>   though the Intel HaM software will not be updated to 2.6.n service
> ------------------------------------
>  Preparing winmodem drivers will generally require compiling.
>  If you are not familar with this process DEFINITELY read carefully
>  the section below "COMPILING DRIVERS, for Linux Newbies"
> Identifiers for Soft Modems
>Earlier generations of modems had most of their functionality encoded in the modem hardware.
>They could be quickly identified by the PCI IDs, and supporting software Easily recommended.
>There is an increasingly prevalant "soft modem" type, for which much more
>of the functionality is in the software, hence the soft modem designation.
>The overall functional system is comprised of a:
>   AC'97 or MC97 controller which has the primary PCI ID;
>   an inexpensive "soft modem" chip which has the SubSystem PCI  ID,
>   and the CPU (central processing unit of the computer)
>     which does essentially all the signal processing.
>Skipping the remainder of this section is OK, if your modem is Not of this AC'97 or MC97.
>For most soft modems, Silicon Labs. produces the core soft modem chip.
>It is assigned a SIL id as its functionality is refined further by final chipset
>manufacturers including AgereSystems, Conexant, Intel, PCTel, SmartLink and MinorBrands.
>It is this SIL id which is needed for direction to optimal Linux software.
>Unfortunately the primary PCI ID only identifies the Modem controller
>and the soft modem Subsystem ID does often NOT identify the soft modem chip vendor.
>Rather manufactures of the Subsystem often put on their own Vendor ID
>Consequently there are some identification processes that cannot be fully automated as yet.
>YOU must assist by one of the following:
> 1) As described above access information under Microsoft Windows, such as
>   ATI3 - Agere SoftModem Version 2.1.22
>   ATI5 - 2.1.22, AMR Intel MB, AC97 ID:SIL REV:0x27
>  Common 0xNN identifiers are:
> 0x26  or SIL26   Motorola
> 0x22  or SIL22   SmartLink
> 0x24  or SIL24  Broadcom - supported by SmartLink drivers
> 	  BCM64  Broadcom
> 0x21  or SIL21   PCTel  or  SmartLink drivers
> 0x27  or SIL27   AgereSystems
>          cxt29   Conexant -
>          SIL25  Intel 537AA
>          INT65  Intel 537EA   "
> 2) Install the latest Smartlink slmodem-2.9.n driver, which outputs softmodem
> identifiers during bootup diagnostics.  The following commands will display selectively (grep)
> relevant lines from the dmesg diagnostic buffer after bootup:
> # dmesg | grep slamr
> The slmodem drivers will often also functionally serve many
> the modem Subsystem chips from Agere, Intel, Pctel and others
> as the core COMM functionalities encoded in these chips are very similar.
> The remainder of this section relates a method of SIL identification,
> which was mostly of utility to the earliest soft modem releases and rarely
> provides SIL identifiers for the newer softmodems.
>During bootup, kernel diagnostics on the System are stored for later display by:
>   dmesg
>This information may include a SIL_id of modems under AC'97/MC97 Controllers,
>depending upon the type of briding of the modem card to the motherboard.
>The scanModem script processes dmesg output to capture AC'97 modem information,
>parses it into a SIL_id if possible, and then displays of modem chipset information.
>Guidance to sources of modem supporting software may thus be obtained.
>The transfer of the AC97 information to the dmesg buffer requires that modules supporting
>both the digital audio card and the ac97_codec be loaded during bootup:
>   modprobe ac97_codec
>   modprobe audio_drivers  (such as i810_audio)
>This can be checked after bootup with:
>   lsmod
>This script can also be used by entering a test block with nomenclature:
>        ./scanModem SILtest
>with SILtest a text file in This Folder
>containing a section of a dmesg output or /var/log/messages like:
>  i810: Intel ICH 82801AA found at IO 0xdc00 and 0xd800, IRQ 11
>  i810_audio: Audio Controller supports 2 channels.
>  ac97_codec: AC97 Audio codec, id: 0x4144:0x5340 (Analog Devices AD1881)
>  i810_audio: AC'97 codec 0 Unable to map surround DAC's (or DAC's not present), total channels = 2
>  ac97_codec: AC97 Modem codec, id: 0x5349:0x4c22 (Silicon Laboratory Si3036)
>which does include a line beginning with:   ac97_codec: AC97 Modem codec, id:
>---------- End SIL_id section ---------------
> ==================== AC'97/MC97 summary ==========================
> Any of the Controllers in the left most column could support
> a variety of subSystem soft modems.
> Hence the current ambiguity in possible support by drivers under Linux
>VendorID:DeviceID                Controller type
>     PCI_ID
>   ---------- --------------------------------------
>    8086:2416 82801AA ICH >  		+ a p c .
>    8086:2426 82801AB ICH0 > 		+ a .
>    8086:2446 82801BA ICH2 > 		+ a p c .
>    8086:2486 82801CA/CAM AC'97 ICH3 >  + a p c i .
>    8086:24c6 82801DB ICH4 > 		+ a   c i b .
>    8086:24d6 82801EB ICH5 > 		+       i .
>    8086:7196 82440 - 443MX >  		+ a  .
>    1543:3052 1543:3052 SI3052 >                i .
>    1039:7013  SIS 630 >               	+ a p   i .
>    1039:7018  SIS 960 >               	+       i .
>    10de:01c1  Nvidia Corp >           	+       i .
>    1106:3068  VIA >			+ a p c i .
>    1022:7446  AMD >
>    10b9:5453  ALI 5453 >                   p c ? .
>    10b9:5457  ALI 5457 >		+   p c i .
>    e159:0001  TigerJet >			i .
>    1002:434d  ATI >				i .
>  --------------------------------------------------------
>  p
>  c   Conexant -
>  +   SmartLink - http://www.smlink.com10b9:5450
>  i   Intel -
>  b   Broadcom
> ======================================================================
> Soft modems under AC'97 or MC97 Controllers
> =============================
> If a modem_codec SIL_id was not output earlier, please reboot Linux to freshen the
> bootup dmesg buffer and then rerun scanModem, to acquire an expected SIL_id.
> However
> The roughly classified with AC'97/MC97 modems may be housed within an
> AMR connector, the MDC connector, IBM's CDC connector, or a mini PCI slot.
> These so-called "soft modems" lack a DSP (digital singal processing) chipset.  Rather the DSP processes are implemented in the much larger base of software code.
> This greatly reduces the cost of the modem hardware and specialization to the
> diverse telephone line standards around the world.
> However the primary PCI_ID 1002:434d identifies Only the modem controller,
> and often FAILS to identify the chipset of the modem SubSystem!!
> Queries under MS windows may be informative, by the following routes:
>   Open a COMM console and send ATI commands to the modem,
>     which may provide identifying information: ATI, ATI1, ATI2 etc.
>     For a good report see:
>   Get whatever detail information is available through:
>     Start > Settings > Control Panel > Modem
>        There may be a TestModem option available.
>     Start > Settings > Control Panel > System > Devices > Modem
>   Try to identify the modem setup file, with name perhaps MODEM.INF
>     Information therein including driver names may help to identify the chipset.
> If the Primary and Subsystem Vendor information was not adeqaute,
> it may be useful to search at
>Analyzing information for PCMCIA device at PCI Bus 0000:02:04.0
>GREPping for an inserted PCMCIA modem with filter:        ommunication
>Analyzing information for PCMCIA device at PCI Bus 0000:02:04.1
>GREPping for an inserted PCMCIA modem with filter:        ommunication
> If a PCMCIA modem is currently inserted and the sockets activated by
>    /etc/init.d/pcmcia start
> then the PCMCIA bridge is NOT transparent.
> If the modem is known to have a Lucent digital signal processing chipset,
> then PCMCIA.tar.gz variant assembled by Joern Wustenfeld is necessary,
> rather than the standard ltmodem-8.26a9.tar.gz at
>  -----------------------------------------------------
>  The System has Ethernet capability. If not expert, 
>  shut down ethernet before initiated modem usage with:
>  # ifconfig eth0 down
>  Attempted of effective networking links are displayed by command:
>  #  /sbin/ifconfig
>  A block with "lo" is an internal loopback test and harmless.
>  However, ethernet "eth0" can be problematic for PPP connections,
>  because of competition for DNS (domain name service).
>  The default is to use the DNS specified for etherenet and
>  without expert configuration, this will block browser naviagation through PPP.
>  ========== ifconfig test =============
>  eth0      Link encap:Ethernet  HWaddr 00:02:3F:6E:95:98  
>lo        Link encap:Local Loopback  
>  If is wisest to disable bootup establishment of ethernet in your Control Center.
>  Depending on your Linux distribution,
>      one of the following Root commands way alternatively be effective:
>  # ifdown eth0
>  # ifconfig eth0 down
>  # /etc/init.d/network stop
>  # /etc/init.d/networking stop

Webmaster: Russell Nelson
Last modified: Wed Jul 30 11:02:43 EDT 2003