Novatel Merlin V620 CDMA PC Card: How To?

georgecmu · 07-21-2005, 05:45 PM

Quote:

Originally posted by gpetme
Does anyone have the ability to port this patch to 2.6?

Thanks,

Greg

You should be able to just make the simple edit in /drivers/usb/serial/usb-serial.c

Look for line
buffer_size = endpoint->wMaxPacketSize;

and replace it with the four lines I posted above.
So far it worked for me.

sipples · 07-22-2005, 10:30 AM

I've opened a discussion on the Linux USB serial driver developer list. We'll see what the collective wisdom suggests for an "official" patch. Ultimately this will probably mean that the Linux USB serial driver will get a fix to handle this situation.

Question: anyone know what the Kyocera's vendor and product IDs are? Are they also 1410 and 1110, like the Novatel?

gpetme · 07-23-2005, 11:58 AM

Quote:

Originally posted by georgecmu
Look for line
buffer_size = endpoint->wMaxPacketSize;
[/B]

I got it working with 2.6. I couldn't find the exact place you referenced, but I hacked the change in. I'll wait for a real fix after the kernel developers have a look at this problem. For now I'm working.

Thanks everyone.

Does anyone know how to speed up the connection? Someone mentioned that they were getting ~ 1.5 Mbps in Windows but only 200-400 Kbps in Linux. I'm seeing the same problem. Anyone have any suggestions?

Thanks,

Greg

Bini2 · 07-26-2005, 08:36 PM

Quote:

Originally posted by gpetme
I got it working with 2.6. I couldn't find the exact place you referenced, but I hacked the change in. I'll wait for a real fix after the kernel developers have a look at this problem. For now I'm working.

Thanks everyone.

Does anyone know how to speed up the connection? Someone mentioned that they were getting ~ 1.5 Mbps in Windows but only 200-400 Kbps in Linux. I'm seeing the same problem. Anyone have any suggestions?

Thanks,

Greg

Any luck with the speed issue Greg?

I seem to be having the same issue.

I used tcpdump and there appears to be no issues with the communication. I am stumped on this one.

Anyone have some trouble shooting suggestions?

gpetme · 07-27-2005, 08:05 PM

No luck yet with getting the V620 to speed up. Are other people (besides Bini2 and myself) having performance problems?

Greg

sheltontechDev · 07-28-2005, 06:07 PM

I'm having the same problem, but my partner who made the exact same changes is not.

I have a HP Pavillion zv6005 (AMD) running FC x86_64

He's using a similiar HP with a 32 bit Intel chip.

Is there any thing all of the systems with speed problems have in common?

Bini2 · 07-30-2005, 03:04 PM

Quote:

Originally posted by sheltontechDev
I'm having the same problem, but my partner who made the exact same changes is not.

I have a HP Pavillion zv6005 (AMD) running FC x86_64

He's using a similiar HP with a 32 bit Intel chip.

Is there any thing all of the systems with speed problems have in common?

What kind of speeds are you getting, shelton?

sheltontechDev · 07-31-2005, 08:09 PM

I'm getting 200-400, Usualy around 250 kbs.

The intel setup is reporting min 600 to max 800 with a 675 avg.

We got these numnbers from toast.net

Louis-Charles D · 08-08-2005, 10:20 PM

Quote:

Originally posted by jpriddis
Using the above information, I have mine connecting, and using Ethereal, I can see an IP address for PPP0. (Running under Fedora Core 2, with the Ricoh PCI-to-PCMCIA adapter, and Kernel version 2.6.58 on a Shuttle X mini-desktop) However, it does not seem to be pulling the DNS information. When in a Winders environment with the Verizon software, it does fine. Is there an additional line that needs to be in the verizon file under /etc/ppp/peers ? At first, it was trying to pull DNS from the onboard NIC settings as ETH0. Once I removed those, I tried hard coding the DNS server info into the Lan config information, but it still isn't resolving DNS. Any suggestions on what I may be missing? Thanks in advance for everyone's posts.

Hi ,

I'm trying to get the ricoh adapter to work with verizon. Everything seems ok but when I try to connect, I get the following:

pppd call 1xevdo
Failed to open /dev/ttyUSB0: No such device

... the /dev/ttyUSB0 device exist and I have been following the instruction from phil karn at http://www.ka9q.net/5220.html .

The internet access has been tested on a notebook without the PCI2PCMCIA adapter and everything worked fine. Any idea?

thank you,
Louis-Charles Dumais

junxiondeveloper · 08-09-2005, 11:48 AM

Quote:

Originally posted by sipples
Here's a slightly modified version of the patch suggested by Junxion. Call this Version 0.2, I suppose. :-) I've made the following changes:

On edit: Why is this code in an #ifdef, by the way? It would seem to be harmless enough for setting buffer_size generally since the driver is going to pick the higher of the device-reported max packet size or the user-specified size. Is there any harm to having a "too large" packet size?

On 2nd edit: Should pktsize default to 1? Are there any devices that would report 0 -- and which should use 0? Or is there some real-world minimum that this code should set?

I think the best way to handle this would be to look for the product/vendor IDs for the Novatel and Kyocera cards and then set the packet size correctly specifically for those devices, PLUS this user-specified parameter for generic devices. But I've got to dig into the code for that, and I'm not near the usbserial.c source at the moment. Anyone know what the Kyocera product/vendor IDs are?

The modifications are in #ifdef in case there are any other specific usb serial drivers that are not implementing this function and defaulting to the generically provided function.

You could create custom drivers for each PC card modem and then hard set what wMaxPacketSize should be. It's much easier for us to use a single driver for all cards of this class. We just wanted more flexibility.

junxiondeveloper · 08-09-2005, 11:56 AM

Quote:

Originally posted by georgecmu
IMHO this code would make more sense:

if(serial->vendor == 0x1410 && serial->product == 0x1110) {
buffer_size = 2048;
} else {
buffer_size = endpoint->wMaxPacketSize;
}

This way there's no need for an extra module parameter and an extra usbserial module, since the buffer_size is only hardcoded for the specific device Novatel V620.

The patch is intended to work with as many usb serial devices as possible. The above code change works but only for the Novatel V620. There are a variety of EVDO capable cards in the market today.

Sierra/Audiovox PC5220 (0xf3d/0x112)
Novatel V620/S620 (0x1410/0x1110)
Sierra AC 580 (0x1199/0x112)
Kyocera KPC650 (0xc88/0x17da)

H.T.

junxiondeveloper · 08-09-2005, 12:07 PM

All,

Thanks to Ren Roderick and an un-named friend for sending me the patch for 2.6.x kernel series for this issue. It is consistent with the 2.4.x patch I previously posted.

http://www.junxion.com/opensource/li...usbserial.html

H.T.

sipples · 08-13-2005, 09:59 AM

Some good news: the latest SuSE Linux 9.3 kernel (2.6.11.4-21.8), which was issued to fix some security-related issues, appears to work fine as-is with the Novatel V620. I left a download going overnight with this new kernel version (unpatched), and there were no lockups or other oddities, so it all seems stable.

I did run a speed test that suggested a somewhat slower speed than what should be available (versus the result I got when I hacked the previous kernel's usbserial driver). That's similar to what others are reporting here. However, I'm not 100% sure this speed test reflects reality. (Still investigating.) And at least the modem card works without kernel driver hacking now.

Bear in mind that there's still one mandatory configuration change: commenting (#) out the "lcp" options in the /etc/ppp/options file. There are four lcp-related lines. If you don't do this the connection will terminate after a minute or two. Also, just use "stupid mode" in your favorite PPP dialer. (I use KInternet.) That works fine.

If you want the modem automatically recognized when you insert it into a slot (or when you boot up with it installed), then add this line to the end of the /etc/modprobe.conf.local file:

options usbserial vendor=0x1410 product=0x1110

Works like a charm.

By the way, I believe that you can update the PRL (roaming list) by dialing *22899, even under Linux. Just use a terminal program (like Minicom) and type ATDT*22899 as a command string to the modem. Wait a bit until the modem downloads the PRL, updates, then resets. This way you don't need a Windows machine to update the roaming list as your carrier adds 1xEV-DO towers to their network. I'm not sure yet how to check the PRL version -- possibly an ATI command of some sort -- but perhaps it's possible.

sipples · 08-20-2005, 10:42 AM

Spoke too soon: the latest SuSE kernel still needs the patch to usb-serial.c. But then it works fine, and adding the "options" line (as described above) makes everything automatic.

tempire · 09-17-2005, 05:05 PM

It seems the topic has moved more towards getting higher speeds, but I am stuck on getting the v620 to work. I have tried every combination previously suggested, but I continue to get the following message:

usb.c: USB device 2 (vend/prod 0x1410/0x1110) is not claimed by any active driver.

I am currently using the stompbox flash image, based on pebble linux, which I believe is based on debian.

The full usb detection messages are below:

Starting hotplug subsystem: usbusb-ohci.c: USB OHCI at membase 0xc4898000, IRQ 11
usb-ohci.c: usb-02:00.0, PCI device 1033:0035
usb.c: new USB bus registered, assigned bus number 1
hub.c: USB hub found
hub.c: 3 ports detected
usb-ohci.c: USB OHCI at membase 0xc489a000, IRQ 11
usb-ohci.c: usb-02:00.1, PCI device 1033:0035
usb.c: new USB bus registered, assigned bus number 2
hub.c: USB hub found
hub.c: 2 ports detected
hub.c: new USB device 02:00.0-1, assigned address 2
usb.c: USB device 2 (vend/prod 0x1410/0x1110) is not claimed by any active driver.
usb-uhci.c: $Revision: 1.275 $ time 23:23:37 Mar 21 2005
usb-uhci.c: High bandwidth mode enabled
usb-uhci.c: v1.275:USB Universal Host Controller Interface driver
uhci.c: USB Universal Host Controller Interface driver v1.1

Not quite sure where to go @ this point, ideas are appreciated.

-g