2.6.32.28: agrsm048pci modem driver

I reported earlier that I had managed to compile all of the analog modem drivers for the 2.6.32.28 kernel, except for the agrsm048pci driver.

Well, I have now managed to compile it. Somehow I overlooked the source package dated '2.1.60_20100108' here:
http://linmodems.technion.ac.il/packages/ltmodem/11c11040/

I commented-out some 'printk' lines that overlap the normal boot messages, and uploaded my patched source here:
http://bkhome.org/sources/kernel-2.6.32.28/3rd-party-UPDATES-10MAR2011/

I have updated the 2.6.32.28 PET, that is used in Woof when building a Puppy (30.0MB):
http://distro.ibiblio.org/quirky/pet_packages-wary5/linux_kernel-2.6.32.28-3-w5.pet

...the driver is big, bumps the PET up from 29.4MB to 30.0MB.


Posted on 10 Mar 2011, 17:28


Comments:

Posted on 10 Mar 2011, 19:27 by Terryphi
Linux kernel 2.6.32.29
Do you intend moving 2.6.32.x to the latest 2.6.32.9 for the next Wary?


Posted on 10 Mar 2011, 20:42 by BarryK
Re 2.6.32.x
The latest is 2.6.32.32.

No, I will wait awhile. Perhaps for Wary 5.2.



Posted on 10 Mar 2011, 24:21 by perthie
Udev Bug
All woof-built puppies contain the udev rules file 55-cups.rules. It contains the invalid pattern match "lp[0-15]*". This is effectively the same as "lp[0-1]*" and only allows lp0 and lp1. A third printer is ignored. The correct syntax should be "lp[0-9]*".



Posted on 11 Mar 2011, 14:19 by BarryK
Re udev rule
perthie,
Thanks, I fixed it.
What is also weird, the next line has lp[0-7]*



Posted on 11 Mar 2011, 21:09 by perthie
Udev Printer Rule
I suspect that the 15 and 7 refer to the maximum number of allowable devices. For that matter, I don't know what the * wildcard is trying to match.


Posted on 12 Mar 2011, 8:03 by scsijon
udev rule
I suspect the first one should have been lp[0-f] as they are in hexidecimal format, somewhere it was translated into binary.


Posted on 12 Mar 2011, 9:47 by broomdodger
udev rule
I suspect the first one should have been lp[0-f] as they are in hexidecimal format, somewhere it was translated into binary.

No that would not be correct to find hex, use [0-9a-f]

[0-f] would find all decimal numbers, some punctuation, all uppercase letters, more punctuation, and lowercase a-f



Posted on 12 Mar 2011, 9:50 by broomdodger
how to quote
oops!
How do I quote someone's comment?
code does not work.


Posted on 12 Mar 2011, 21:18 by perthie
Udev CUPS Rule Ad Nauseum
Udev applies the standard rules for regular expressions. So lp[0-15] is actually matching lp0,lp1 and lp5.