[CLOSED] Problem with Agere HDA modem on HP550 laptop

Message
Author
dialeddown
Posts: 20
Joined: Mon 26 Apr 2010, 00:39

#16 Post by dialeddown »


User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

Problem returns on lupu5.2.8

#17 Post by peebee »

Hi Richard (rerwin)

You asked in the other thread Drivers for Puppy 5.1-5.2.8 with 2.6.33.2 kernel http://www.murga-linux.com/puppy/viewto ... 248#564248:
peebee,
I see your statements of a year ago about success with wary 060 or 070, or such. Where did all that end up? Does wary 5.12 work for you?

BTW, to keep this thread dedicated to providing drivers without the clutter of debugging conversations, we should take this discussion elsewhere. How about to your original thread on this subject?
http://www.murga-linux.com/puppy/viewto ... 455#322455
Richard
I can confirm that all Wary's up to and including Wary 5.1.4.1 have been working fine - it will be interesting to see what happens when 5.2 / Racy is posted as I think Barry is moving to a later kernel there too (but I think Barry has an incentive as he has a laptop with this modem??)

I'll test your alternative driver tonight (got to go to work now).

Cheers
peebee
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
BarryK
Puppy Master
Posts: 9392
Joined: Mon 09 May 2005, 09:23
Location: Perth, Western Australia
Contact:

#18 Post by BarryK »

peebee,
No, no change in kernel for Wary, it will continue with the 2.6.32.x kernel indefinitely.

Racy is something else, a "Wary on steroids", but because it has recent Xorg and kernel it looses the support for some older hardware that Wary has.
[url]https://bkhome.org/news/[/url]

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#19 Post by peebee »

Hi Richard

I tried the .pet you posted - I'm afraid it crashed just like the previous version.

Perhaps Barry could tell you how he got the Agere code working for Wary??

One difference I notice in Wary is that the modem is called ttyAGS3 in dmesg when its loading but in pupdial its called ttyS_AGS3 ......

This item on Barry's blog http://bkhome.org/blog/?viewDetailed=01890 talks about missing symbols and also some Alsa dependency....

http://bkhome.org/blog/?viewDetailed=01961

Cheers
Peter

dmesg output:

Code: Select all

serial8250_register_ports: BaseAddress 0x4004 Irq 29 
ttyAGS3 at I/O 0x4004 (irq = 29) is a AgereModem
agrserial - ret_val 0, call: lt_modem_ops.init_modem
Loading module Agere Modem Interface driver version 2.1.80.0 (2007-10-01)
==> codecType = 0x32
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<f882c9bb>] AzlDmaAllocation+0x81/0x2f9 [agrmodem]
*pde = 00000000 
Oops: 0002 [#1] SMP 
last sysfs file: /sys/devices/pci0000:00/0000:00:1c.1/0000:10:00.0/ssb0:0/net/wlan0/statistics/tx_bytes
Modules linked in: i915 drm_kms_helper drm i2c_algo_bit i2c_core agrserial arc4 ecb agrmodem b43 mac80211 cfg80211 led_class ohci_hcd snd_hda_codec_analog snd_hda_intel snd_hda_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_seq_dummy snd_seq_oss snd_seq_midi snd_rawmidi snd_seq_midi_event snd_seq snd_timer snd_seq_device snd soundcore snd_page_alloc serio_raw pcspkr fbcon tileblit font bitblit softcursor shpchp pci_hotplug sg e1000e intel_agp agpgart hp_wmi wmi battery fan container video output thermal evdev button processor ac fuse aufs nls_iso8859_1 nls_cp437 usbhid usb_storage squashfs ssb uhci_hcd ehci_hcd usbcore

Pid: 10792, comm: modem-stats Tainted: P           2.6.33.2 #1 3618/HP 550
EIP: 0060:[<f882c9bb>] EFLAGS: 00210202 CPU: 0
EIP is at AzlDmaAllocation+0x81/0x2f9 [agrmodem]
EAX: 00000001 EBX: 00000000 ECX: f4fccd20 EDX: 00000000
ESI: f88d40c4 EDI: f88d4028 EBP: f03bbd34 ESP: f03bbd0c
 DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Process modem-stats (pid: 10792, ti=f03ba000 task=f4ff0c20 task.ti=f03ba000)
Stack:
 00000000 f88de5f5 f88de5d8 f88de5e0 f03bbd28 f03bbda3 000000ff f88d40c4
 f88d40c4 f88d4028 f03bbd54 f882d2ad f88d40c4 00000000 f88d4028 f03bbdc4
 c128a594 f03bbda3 f03bbdc4 f88161e1 f88d40c4 f88c7048 00000008 00000001
Call Trace:
 [<f882d2ad>] ? CAzlIntelInit+0x381/0x3c4 [agrmodem]
 [<c128a594>] ? printk+0xe/0x11
 [<f88161e1>] ? LXHardwareStart+0x11de/0x13fb [agrmodem]
 [<c1055bcd>] ? __alloc_pages_nodemask+0xd6/0x470
 [<f8811998>] ? linux_modem_open+0x45/0x10e [agrmodem]
 [<c102c857>] ? mod_timer+0xea/0xf3
 [<f876bc11>] ? modemPortOpen+0x5/0xe [agrmodem]
 [<f88e7a6f>] ? serial8250_startup+0x74/0x250 [agrserial]
 [<c11a9348>] ? uart_startup+0x68/0xff
 [<c11a9acd>] ? uart_open+0x10b/0x302
 [<c119a53a>] ? tty_ldisc_setup+0x50/0x58
 [<c11969b9>] ? tty_init_dev+0x17c/0x1e9
 [<c1196d17>] ? tty_open+0x2f1/0x43e
 [<c1073ab0>] ? chrdev_open+0x11f/0x135
 [<c1073991>] ? chrdev_open+0x0/0x135
 [<c107041a>] ? __dentry_open+0x10c/0x1f0
 [<c107058e>] ? nameidata_to_filp+0x29/0x39
 [<c107a6d3>] ? do_filp_open+0x44b/0x82e
 [<c106028a>] ? __do_fault+0x301/0x331
 [<c10616be>] ? handle_mm_fault+0x238/0x4d6
 [<c108168d>] ? alloc_fd+0x52/0xb4
 [<c1070211>] ? do_sys_open+0x49/0xe7
 [<c10184ae>] ? do_page_fault+0x25f/0x275
 [<c10702f3>] ? sys_open+0x1e/0x23
 [<c1002694>] ? sysenter_do_call+0x12/0x26
Code: a5 00 00 89 44 24 08 8d 86 31 a5 00 00 89 44 24 04 c7 04 24 00 00 00 00 e8 63 57 01 00 85 c0 0f 84 71 02 00 00 8b 9e 14 a5 00 00 <c7> 03 01 00 00 00 c7 44 24 04 01 00 00 00 89 34 24 e8 f4 fe fe 
EIP: [<f882c9bb>] AzlDmaAllocation+0x81/0x2f9 [agrmodem] SS:ESP 0068:f03bbd0c
CR2: 0000000000000000
---[ end trace 692ed8da9155f2f2 ]---
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#20 Post by rerwin »

peebee,
Thanks for the links to Barry's blog items about the Agere driver. Here is the key to the problem:
The problem arose because I upgraded the ALSA drivers to 1.0.23, but the Agere drivers were compiled against the kernel's source ALSA code.

This time though, for the 2.6.31.14 kernel, I have not upgraded ALSA, avoiding this problem.
Lupu's ALSA is 1.0.24.2 whereas the kernel source must be something earlier, perhaps 1.0.16 or 1.0.20! It appears the ALSA version at run time must be the same as contained in the kernel source used for driver compilation. Maybe we can update the kernel-source's version to contain ALSA 1.0.24.2 without recompiling the kernel, but I doubt it. Since we do not want to change the ALSA version from 1.0.24.2, if we cannot change the kernel source successfully, then we are probably stuck.

After reflection: I think I have been using the wrong kernel source package. I will now try with playdayz's sfs file posted at the start of the lupu thread.
EDIT: Rebuilt the HDA-HSF package contents using playdayz's kernel source. However, the md5sums of the old and new agr* driver modules are the same. So that was not the solution.

EDIT: I found the HowTo for building the Agere HDA driver here;
http://linmodems.technion.ac.il/package ... 0-HDA.html
The part we are not complying with is:
For the 11c11040 case you must determine your ALSA drivers version. You MUST be using the version of ALSA that came with your current kernel. If you have since upgraded to a newer version of the ALSA modules, you will need to revert to the old version (or if this is not possible, upgrade your kernel.) To find your ALSA module version, issue in a terminal the command:
cat /proc/asound/version
In lupu5428 this returns:
Advanced Linux Sound Architecture Driver Version 1.0.21.

Our current ALSA version shown by "alsactl --version" is:
alsactl version 1.0.24.2

Therefore, to get the driver to work we must return lupu528 to running with ALSA 1.0.21 -- quite unlikely. We're dead.

But we could continue the experiment by trying the driver on puppy 4.3.1, whose kernel came with ALSA 1.0.20 and also runs with it. The test would entail using my "4.3.1.1." package along with the associated agrsm package. I think peebee may have tried that without success. But that setup holds the best promise for getting it to work outside of wary.
Richard

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#21 Post by peebee »

Hi Richard

Having Wary 5.1.4.1 available as a frugal install on my disk for when I'm away from home and need to use a modem is fine for me.

It would obviously be nice to have it also work on the current "mainstream" pup but looks like lupu is a no-go.

Perhaps the thing to do is jump a generation and get it working on slacko - if that has the compatible alsa constraint..... I'll investigate later.

So I wouldn't see the need to go back to pup431.....

Regards
Peter
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

tempestuous
Posts: 5464
Joined: Fri 10 Jun 2005, 05:12
Location: Australia

#22 Post by tempestuous »

rerwin wrote:Our current ALSA version shown by "alsactl --version" is:
alsactl version 1.0.24.2
The version of the "alsactl" utility does not necessarily represent the version of ALSA libraries.
The correct way of determining ALSA version is to run this command -

Code: Select all

cat /proc/asound/version
rerwin wrote:Therefore, to get the driver to work we must return lupu528 to running with ALSA 1.0.21
ALSA was upgraded at some point in the 5.2.x series. I suggest you test your modem driver in Puppy 5.1.1, which should still have ALSA 1.0.21.
Then if it works, it's quite possible to downgrade ALSA to the standard version in the 2.6.33.2 kernel source (ALSA 1.0.21). In fact, I still use Lucid 5.1 as my kernel development box, so I have the older ALSA libraries. Let me know if you want me to package these.

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#23 Post by peebee »

For slacko

cat proc/asound/version gives 1.0.23

whereas

alsactl --version gives 1.0.24.2

so I guess that means that slacko will also not be able to support the agere hda modems.....

Three Headed Dog luci001 on the same kernel gives the same results.

whereas squeeze 5.x.6 exprimo gives 1.0.24 and 1.0.23 so presumably there is the opportunity there to upgrade alsa to 1.0.24 and then the agere modem might work....

I'm going to try the new versions of wary and racy tomorrow when I've downloaded and will report the result.

Cheers
Peter
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#24 Post by rerwin »

tempestuous wrote:The version of the "alsactl" utility does not necessarily represent the version of ALSA libraries.
The correct way of determining ALSA version is to run this command -

Code: Select all

cat /proc/asound/version
. . .
ALSA was upgraded at some point in the 5.2.x series. I suggest you test your modem driver in Puppy 5.1.1, which should still have ALSA 1.0.21.
Then if it works, it's quite possible to downgrade ALSA to the standard version in the 2.6.33.2 kernel source (ALSA 1.0.21). In fact, I still use Lucid 5.1 as my kernel development box, so I have the older ALSA libraries. Let me know if you want me to package these.
It appears I jumped to two wrong conclusions and got the facts all scrambled. Thank you for straightening me out. I am back to being confused about how to identify that ALSA has been "upgraded" -- exactly what does that mean? Lupu 5.2.8 shows

Code: Select all

sh-4.1# cat /proc/asound/version
Advanced Linux Sound Architecture Driver Version 1.0.21.
This is what I am working from in an Agere drivers HowTo:
For the 11c11040 case you must determine your ALSA drivers version. You MUST be using the version of ALSA that came with your current kernel. If you have since upgraded to a newer version of the ALSA modules, you will need to revert to the old version (or if this is not possible, upgrade your kernel.) To find your ALSA module version, issue in a terminal the command:
cat /proc/asound/version
How do we identify which ALSA version "came with the kernel"? And in what way would we "revert to the old version?" In the kernel source or in the operational ALSA utilities version, or what else?

And thank you very much for your offer of help in making some progress. I will probably need it to understand more about this.

For testing, it is actually peebee whom I will depend on for that with 5.1.1. I think I will put together a package for that test, that adds the updated agrsm firmware tarball, if it is not up to date in 5.1.1.
Richard

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#25 Post by peebee »

Latest tests:

Wary 5.1.4.1 says 1.0.21 and 1.0.21 - modem works

Wary 5.1.99 says 1.0.21 and 1.0.21 - modem works

Racy says 1.0.24 and 1.0.21 - modem not detected

so Racy is in same state as exprimo - alsa could be updated to 1.0.24 and then the modem might work??

cheers
peter
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#26 Post by rerwin »

tempestuous, peebee,
I have what appears to be good news! I have remastered 5.1.1 with the latest modem-modprobe package (to make it look more like 5.2.8 internally) and the three main "dialup modem" packages (other than the Intel 537s) and seem to run with a healthy HDA modem. The catch is that I do not have such a modem; I only forced loading of the agrmodem module for the test. But the indication is good.

When I run the test on 5.2.8, I see certain error indications. On Wary I see different messages that look like success. (I have an HSF modem installed, which might be fooling puppy into thinking the HDA modem is there.) When I run the test on my new upgraded 5.1.1 installation, I see the same messages as in Wary. This tells me that the modem might actually work in this setup.

However, the two ALSA versions are not the same -- /proc/asound: 1.0.21 and alsactl: 1.0.22. So we need peebee to try this with his real HDA modem, to be sure. If that works, it confirms that the alsactl version is unimportant for our purposes.

modem-modprobe_fix_pack: http://www.murga-linux.com/puppy/viewto ... 074#431074
modem drivers: http://www.murga-linux.com/puppy/viewto ... 112#564112

Remember: To do the remaster, use the option offered by the modem-modprobe package. Do not respond to its "remaster" dialog until the driver (and any other) packages are installed.
Richard

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

5.1.1 experiment results

#27 Post by peebee »

Hi Richard

Tried 5.1.1 - afraid it crashed just like 5.2.8

What I did:
downloaded 5.1.1 iso
downloaded the 2 pets you specified (only the 11c11040 modem pet)
installed 5.1.1 manual frugal by opening the iso and copying the 3 files to a test directory
booted 5.1.1
closed down and created a savefile
rebooted
installed 2 pets
rebooted
checked alsa versions - result = 1.0.21 and 1.0.22 so 5.1.1 alsa version is higher than kernel??
issued the modem-stats -c "ATI3" /dev/ttyAGS3 command in a terminal
said "Killed"

I didn't do a "remaster"......as I never boot off cd

Cheers
Peter

output from dmesg was:

Code: Select all

serial8250_register_ports: BaseAddress 0x4004 Irq 29 xxxx
ttyAGS3 at I/O 0x4004 (irq = 29) is a AgereModem
agrserial - ret_val 0, call: lt_modem_ops.init_modem
Loading module Agere Modem Interface driver version 2.1.80.0 (2007-10-01)
==> codecType = 0x32
BUG: unable to handle kernel NULL pointer dereference at (null)
IP: [<f8f6c9bb>] AzlDmaAllocation+0x81/0x2f9 [agrmodem]
*pde = 00000000 
Oops: 0002 [#1] SMP 
last sysfs file: /sys/devices/pci0000:00/0000:00:1c.1/0000:10:00.0/ssb0:0/net/wlan0/dormant
Modules linked in: agrserial agrmodem i915 drm_kms_helper drm i2c_algo_bit i2c_core snd_hda_codec_analog arc4 ecb snd_hda_intel snd_hda_codec snd_pcm_oss snd_mixer_oss snd_pcm snd_page_alloc snd_seq_dummy snd_seq_oss snd_seq_midi snd_seq_midi_event snd_rawmidi b43 mac80211 led_class cfg80211 snd_seq snd_seq_device snd_timer snd ohci_hcd soundcore serio_raw pcspkr fbcon bitblit softcursor tileblit font shpchp pci_hotplug e1000e intel_agp agpgart hp_wmi wmi fan battery container thermal video output evdev button processor ac fuse aufs nls_iso8859_1 nls_cp437 usbhid usb_storage squashfs ssb uhci_hcd ehci_hcd usbcore

Pid: 18102, comm: modem-stats Tainted: P           2.6.33.2 #1 3618/HP 550
EIP: 0060:[<f8f6c9bb>] EFLAGS: 00210202 CPU: 0
EIP is at AzlDmaAllocation+0x81/0x2f9 [agrmodem]
EAX: 00000001 EBX: 00000000 ECX: ed128920 EDX: 00000000
ESI: f81a00c4 EDI: f81a0028 EBP: ec875d34 ESP: ec875d0c
 DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Process modem-stats (pid: 18102, ti=ec874000 task=f68dc7d0 task.ti=ec874000)
Stack:
 00000000 f81aa5f5 f81aa5d8 f81aa5e0 ec875d28 ec875da3 000000ff f81a00c4
 f81a00c4 f81a0028 ec875d54 f8f6d2ad f81a00c4 00000000 f81a0028 ec875dc4
 c128a594 ec875da3 ec875dc4 f8f561e1 f81a00c4 f8155048 00000008 00000001
Call Trace:
 [<f8f6d2ad>] ? CAzlIntelInit+0x381/0x3c4 [agrmodem]
 [<c128a594>] ? printk+0xe/0x11
 [<f8f561e1>] ? LXHardwareStart+0x11de/0x13fb [agrmodem]
 [<c1055bcd>] ? __alloc_pages_nodemask+0xd6/0x470
 [<f8f51998>] ? linux_modem_open+0x45/0x10e [agrmodem]
 [<c102c857>] ? mod_timer+0xea/0xf3
 [<f8eabc0f>] ? modemPortOpen+0x5/0xe [agrmodem]
 [<f818ba6f>] ? serial8250_startup+0x74/0x250 [agrserial]
 [<c11a9348>] ? uart_startup+0x68/0xff
 [<c11a9acd>] ? uart_open+0x10b/0x302
 [<c119a53a>] ? tty_ldisc_setup+0x50/0x58
 [<c11969b9>] ? tty_init_dev+0x17c/0x1e9
 [<c1196d17>] ? tty_open+0x2f1/0x43e
 [<c1073ab0>] ? chrdev_open+0x11f/0x135
 [<c1073991>] ? chrdev_open+0x0/0x135
 [<c107041a>] ? __dentry_open+0x10c/0x1f0
 [<c107058e>] ? nameidata_to_filp+0x29/0x39
 [<c107a6d3>] ? do_filp_open+0x44b/0x82e
 [<c106028a>] ? __do_fault+0x301/0x331
 [<c10616be>] ? handle_mm_fault+0x238/0x4d6
 [<c108168d>] ? alloc_fd+0x52/0xb4
 [<c1070211>] ? do_sys_open+0x49/0xe7
 [<c10184ae>] ? do_page_fault+0x25f/0x275
 [<c10702f3>] ? sys_open+0x1e/0x23
 [<c128c625>] ? syscall_call+0x7/0xb
Code: a5 00 00 89 44 24 08 8d 86 31 a5 00 00 89 44 24 04 c7 04 24 00 00 00 00 e8 63 57 01 00 85 c0 0f 84 71 02 00 00 8b 9e 14 a5 00 00 <c7> 03 01 00 00 00 c7 44 24 04 01 00 00 00 89 34 24 e8 f4 fe fe 
EIP: [<f8f6c9bb>] AzlDmaAllocation+0x81/0x2f9 [agrmodem] SS:ESP 0068:ec875d0c
CR2: 0000000000000000
---[ end trace 713d15658ad5604f ]---
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#28 Post by rerwin »

peebee,
Apparently, my test is not valid as an indication that the driver is working correctly. Instead of entering the modemstats command, could you simply boot up and then find the messages in /var/log/messages that pertain to the driver? I search on "agr" and find only a couple of lines in both wary and lupu 5.1.1, but only that the module was "registered" in lupu 5.2.8. In lupu 5.2.8, the ttyAGR3 node does not get generated, whereas both wary and lupu 5.1.1 generate it automatically.

Remember, my test uses forced loading of agrmodem, set in the boot manager as an "added module." The bottom line is that there is a difference in behavior between 5.2.8 and 5.1.1, with 5.1.1 matching the behavior of wary 5.1.3. That is all I can go on.

As a last gasp, could you boot up with lupu 5.1.1 + the 2 added packages and save off the /var/log/messages file for attachment to a posting. Also do the same on whatever version of Wary you have. Then we can compare the agr-related messages for clues.

Maybe we still have an ALSA mismatch of some sort, but the details of that are not obvious, unless the difference in the reported versions is significant. What bothers me is that lupu 5.2.8 reports that is has ALSA 1.0.21 even though both playdayz and tempestuous state that ALSA was "upgraded", is some sense. Maybe the upgrade omitted setting the correct version in /proc/asound/version. Here is where we need tempestuous' assistance.
Richard

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#29 Post by peebee »

rerwin wrote:As a last gasp, could you boot up with lupu 5.1.1 + the 2 added packages and save off the /var/log/messages file for attachment to a posting. Also do the same on whatever version of Wary you have. Then we can compare the agr-related messages for clues.
Richard
Hi Richard

Attached please find messages files with added .gz to make them postable for:

wary5.1.4.1
lupu5.1.1 with added pets
lupu5.2.8 with added pets

all copied just after the modem-stats command

Hope they give you some clues.

Cheers
Peter
Attachments
messages-lupu528.gz
(58.32 KiB) Downloaded 751 times
messages-lupu511.gz
(58.24 KiB) Downloaded 781 times
messages-wary5141.gz
(58.69 KiB) Downloaded 800 times
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#30 Post by rerwin »

peebee, tempestuous,
Thanks for the message logs. I have compared them line by line for the modem:

Code: Select all

W514 Sep 22 07:09:05 puppypc15574 user.info kernel: usbcore: registered new interface driver agr12dec2006
L511 Sep 21 09:19:35 puppypc user.info kernel: usbcore: registered new interface driver agr12dec2006
L528 Sep 22 00:32:29 puppypc user.info kernel: usbcore: registered new interface driver agr12dec2006
 . . .
W514 Sep 22 07:09:05 puppypc15574 user.warn kernel: Disabling lock debugging due to kernel taint
L511 Sep 21 09:19:37 puppypc user.warn kernel: Disabling lock debugging due to kernel taint
L528 Sep 22 00:32:31 puppypc user.warn kernel: Disabling lock debugging due to kernel taint

W514 Sep 22 07:09:05 puppypc15574 user.warn kernel: serial8250_register_ports: BaseAddress 0x4004 Irq 16 
L511 Sep 21 09:19:37 puppypc user.warn kernel: serial8250_register_ports: BaseAddress 0x4004 Irq 29 
L528 Sep 22 00:32:31 puppypc user.warn kernel: serial8250_register_ports: BaseAddress 0x4004 Irq 29 

W514 Sep 22 07:09:05 puppypc15574 user.info kernel: ttyAGS3 at I/O 0x4004 (irq = 16) is a AgereModem
L511 Sep 21 09:19:37 puppypc user.info kernel: ttyAGS3 at I/O 0x4004 (irq = 29) is a AgereModem
L528 Sep 22 00:32:31 puppypc user.info kernel: ttyAGS3 at I/O 0x4004 (irq = 29) is a AgereModem

W514 Sep 22 07:09:05 puppypc15574 user.err kernel: agrserial - ret_val 0, call: lt_modem_ops.init_modem
L511 Sep 21 09:19:37 puppypc user.err kernel: agrserial - ret_val 0, call: lt_modem_ops.init_modem
L528 Sep 22 00:32:31 puppypc user.err kernel: agrserial - ret_val 0, call: lt_modem_ops.init_modem

W514 Sep 22 07:09:05 puppypc15574 user.info kernel: Loading module Agere Modem Interface driver version 2.1.80.0 (2007-10-01)
L511 Sep 21 09:19:37 puppypc user.info kernel: Loading module Agere Modem Interface driver version 2.1.80.0 (2007-10-01)
L528 Sep 22 00:32:31 puppypc user.info kernel: Loading module Agere Modem Interface driver version 2.1.80.0 (2007-10-01)

---
W514 Sep 22 07:09:05 puppypc15574 user.info kernel: [drm] Initialized drm 1.1.0 20060810
L511 Sep 22 08:20:01 puppypc user.warn kernel: ==> codecType = 0x32
L528 Sep 22 00:34:00 puppypc user.warn kernel: ==> codecType = 0x32
---

W514 Sep 22 07:09:05 puppypc15574 user.info kernel: pci 0000:00:02.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
L511 Sep 22 08:20:01 puppypc user.alert kernel: BUG: unable to handle kernel NULL pointer dereference at (null)
L528 Sep 22 00:34:00 puppypc user.alert kernel: BUG: unable to handle kernel NULL pointer dereference at (null)

W514 Sep 22 07:09:05 puppypc15574 user.debug kernel: pci 0000:00:02.0: setting latency timer to 64
L511 Sep 22 08:20:01 puppypc user.alert kernel: IP: [<f88339bb>] AzlDmaAllocation+0x81/0x2f9 [agrmodem]
L528 Sep 22 00:34:00 puppypc user.alert kernel: IP: [<f8f6c9bb>] AzlDmaAllocation+0x81/0x2f9 [agrmodem]

W514 Sep 22 07:09:05 puppypc15574 user.debug kernel: pci 0000:00:02.0: irq 29 for MSI/MSI-X
L511 Sep 22 08:20:01 puppypc user.warn kernel: *pde = 00000000 
L528 Sep 22 00:34:00 puppypc user.warn kernel: *pde = 00000000 

L511 Sep 22 08:20:01 puppypc user.emerg kernel: Oops: 0002 [#1] SMP 
L528 Sep 22 00:34:00 puppypc user.emerg kernel: Oops: 0002 [#1] SMP 

L511 Sep 22 08:20:01 puppypc user.emerg kernel: last sysfs file: /sys/devices/pci0000:00/0000:00:1c.1/0000:10:00.0/ssb0:0/net/wlan0/dormant
L528 Sep 22 00:34:00 puppypc user.emerg kernel: last sysfs file: /sys/devices/pci0000:00/0000:00:1c.1/0000:10:00.0/ssb0:0/net/wlan0/dormant
The two lupus are essentially the same, derailed by a null pointer reference. So lupu 5.1.1 is not the solution. What next, tempestuous?

BTW, I compared the md5sums of the 11c11040 module between compilations with both the generic puppy version of the 2.6.33.2 source package and the one playdayz used for lucid pup -- they were the same. So, I conclude that the choice of source code package was not a factor. Is there a way to create a dot-pet package that restores the 1.0.21 versions of the ALSA components that got "upgraded"? Or is it messier than that?
Richard

tempestuous
Posts: 5464
Joined: Fri 10 Jun 2005, 05:12
Location: Australia

#31 Post by tempestuous »

rerwin, let's take the technical discussion of ALSA-compatibility offline. I have sent you a PM.

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#32 Post by rerwin »

peebee,
I have something for you to try. Tempestuous sent me a dot-pet of the ALSA 1.0.21 library and utilities that I have tested on my PC. The same test as before, forcing loading of agrmodem, produced a better-looking set of messages -- and no distress calls!
  • Sep 23 17:38:30 puppypc user.info kernel: usbcore: registered new interface driver agr12dec2006
    Sep 23 17:38:30 puppypc user.warn kernel: serial8250_register_ports: BaseAddress 0x0400 Irq 3
    Sep 23 17:38:30 puppypc user.info kernel: ttyAGS3 at I/O 0x400 (irq = 3) is a AgereModem
    Sep 23 17:38:30 puppypc user.err kernel: agrserial - ret_val 0, call: lt_modem_ops.init_modem
    Sep 23 17:38:30 puppypc user.info kernel: Loading module Agere Modem Interface driver version 2.1.80.0 (2007-10-01)
This suggests that the package may solve the problem. However, it restores ALSA 1.0.21 for everything, so new features provided by 1.0.24 will be lost.

I have modified the package slightly, to add scripts to deal with the fact that lupu puts 2 files in /sbin, while the package has them in /usr/sbin. If you uninstall the package, the mods will be undone. The package can be installed on top of everything else, so you need not start afresh. Here's hoping.
Richard

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#33 Post by peebee »

rerwin wrote:peebee,
I have something for you to try. Here's hoping.
Richard
So so sorry Richard - but I still get a crash even with the alsa 1.0.21 pet

Messages attached for lupu528
Attachments
messages230911.gz
(58.52 KiB) Downloaded 1027 times
dmesg230911.gz
(30.19 KiB) Downloaded 1035 times
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

#34 Post by rerwin »

peebee,
I am surprised by the failure of the ALSA 1.0.21 package to resolve the crash problem. The idea was to have all parts -- library, utilities and driver -- at the same version. If the agrsm driver was coded for 1.0.20, 1.0.21 must be too different for it to handle.

Although I doubt it will help, you might also try installing the new compiled version of the ALSA 1.0.24 driver into a normal lupu 528, to match the library and utilities there. But since it crashes with 1.0.21, I don't expect any better from matching ALSA 1.0.24 components, either.

I think we will have to wait for the agrsm driver maintainer to upgrade it to at least ALSA 1.0.24.
Richard

User avatar
peebee
Posts: 4370
Joined: Sun 21 Sep 2008, 12:31
Location: Worcestershire, UK
Contact:

#35 Post by peebee »

Hi Richard

I guess the advice we should give anybody with the AGERE HDA modem is to use Wary in the first instance.

If they really want lupu or slacko then it is possible to switch their kernels to use the wary kernel as described in:
http://www.murga-linux.com/puppy/viewtopic.php?t=60180

I've put the kernel from wary5.1.103 into both lupu528 and slacko4.99.7 today and both seem to work fine AND have a working modem - see the attached screen dump.

I'll put this thread back to CLOSED.

Cheers
peter
Attachments
screen.png
(147 KiB) Downloaded 741 times
ImageLxPup = Puppy + LXDE
Main version used daily: LxPupSc; Assembler of UPups, ScPup & ScPup64, LxPup, LxPupSc & LxPupSc64

Post Reply