Hi,
I have three rt2573 wireless devices:
. TP-Link WN321G
. D-Link DWA-110
. Built-in inside Intel ClassmatePC
The D-Link device loads only the rt73usb module and
works really fine. The other two loads rt73usb AND
rt2500usb. The correct module for all is rt73usb.
Loading rt2500usb AND rt73usb don't have any functional
problem. All wireless devices works fine. But is very
annoying the additional modules loaded.
Both modules are loaded because both handle 148f:2573
devices. Of course my universe is very small, only two
devices, but... devices 148f:2573 that works with
rt2500usb exists? Those devices can work with rt73usb?
If the those 148f:2573 USB devices doesn't exists or
if they exists and works fine with rt73usb, can we
remove 148f:2573 from rt2500usb?
Thanks!
Piter Punk
Hi all,
for completeness sake, here's the output of my devices as well.
Luis Correia
rt2x00 project admin
On Sun, Dec 28, 2008 at 22:32, Ivo van Doorn <[email protected]> wrote:
> On Sunday 28 December 2008, Pavel Roskin wrote:
>> On Sun, 2008-12-28 at 10:34 +0100, Ivo van Doorn wrote:
>> > On Sunday 28 December 2008, Piter PUNK wrote:
>>
>> > > If the those 148f:2573 USB devices doesn't exists or
>> > > if they exists and works fine with rt73usb, can we
>> > > remove 148f:2573 from rt2500usb?
>> >
>> > No we can't, you can blame manufacturers for shipping
>> > USB sticks with different chipsets but the exact same USB ID.
>>
>> Maybe there are some hints on the USB level that would make it possible
>> to distinguish between the devices without loading the driver?
>> Something like the number of endpoints? Then we could ask USB
>> developers to provide a way to specify them in the USB devece table.
>
> Number of endpoints varies, your rt73usb device only has 2 endpoints, while
> mine appears to have 5.
> The only thing the driver is doing now is reading the EEPROM and double checking
> the first 2 bytes to see which device has been loaded.
>
>> I have a device that works with rt73usb, but not rt2500usb. I'm
>> attaching the output of "lsusb -v" for the device. If we look at
>> similar information for other devices, we could find out how to
>> distinguish them.
>
> Please note that some manufacturers (most notorious example Linksys) managed
> to get 5 different chipsets under a single USB ID. chipsets came from Broadcom,
> Ralink and Prism.
>
> For what it is worth, I am attaching the lsusb -v output from my hardware
> (both rt73usb as well as rt2500usb).
>
> Ivo
>
On Sunday 28 December 2008, Piter PUNK wrote:
>
> Hi,
>
> I have three rt2573 wireless devices:
>
> . TP-Link WN321G
> . D-Link DWA-110
> . Built-in inside Intel ClassmatePC
>
> The D-Link device loads only the rt73usb module and
> works really fine. The other two loads rt73usb AND
> rt2500usb. The correct module for all is rt73usb.
>
> Loading rt2500usb AND rt73usb don't have any functional
> problem. All wireless devices works fine. But is very
> annoying the additional modules loaded.
>
> Both modules are loaded because both handle 148f:2573
> devices. Of course my universe is very small, only two
> devices, but... devices 148f:2573 that works with
> rt2500usb exists? Those devices can work with rt73usb?
rt2500usb and rt73usb drivers are for different chipsets,
so when rt73usb works for a particular chip the rt2500usb
driver will not work for that same chip.
> If the those 148f:2573 USB devices doesn't exists or
> if they exists and works fine with rt73usb, can we
> remove 148f:2573 from rt2500usb?
No we can't, you can blame manufacturers for shipping
USB sticks with different chipsets but the exact same USB ID.
Ivo
On Sunday 28 December 2008, Pavel Roskin wrote:
> On Sun, 2008-12-28 at 10:34 +0100, Ivo van Doorn wrote:
> > On Sunday 28 December 2008, Piter PUNK wrote:
>
> > > If the those 148f:2573 USB devices doesn't exists or
> > > if they exists and works fine with rt73usb, can we
> > > remove 148f:2573 from rt2500usb?
> >
> > No we can't, you can blame manufacturers for shipping
> > USB sticks with different chipsets but the exact same USB ID.
>
> Maybe there are some hints on the USB level that would make it possible
> to distinguish between the devices without loading the driver?
> Something like the number of endpoints? Then we could ask USB
> developers to provide a way to specify them in the USB devece table.
Number of endpoints varies, your rt73usb device only has 2 endpoints, while
mine appears to have 5.
The only thing the driver is doing now is reading the EEPROM and double checking
the first 2 bytes to see which device has been loaded.
> I have a device that works with rt73usb, but not rt2500usb. I'm
> attaching the output of "lsusb -v" for the device. If we look at
> similar information for other devices, we could find out how to
> distinguish them.
Please note that some manufacturers (most notorious example Linksys) managed
to get 5 different chipsets under a single USB ID. chipsets came from Broadcom,
Ralink and Prism.
For what it is worth, I am attaching the lsusb -v output from my hardware
(both rt73usb as well as rt2500usb).
Ivo
On Sun, 2008-12-28 at 10:34 +0100, Ivo van Doorn wrote:
> On Sunday 28 December 2008, Piter PUNK wrote:
> > If the those 148f:2573 USB devices doesn't exists or
> > if they exists and works fine with rt73usb, can we
> > remove 148f:2573 from rt2500usb?
>
> No we can't, you can blame manufacturers for shipping
> USB sticks with different chipsets but the exact same USB ID.
Maybe there are some hints on the USB level that would make it possible
to distinguish between the devices without loading the driver?
Something like the number of endpoints? Then we could ask USB
developers to provide a way to specify them in the USB devece table.
I have a device that works with rt73usb, but not rt2500usb. I'm
attaching the output of "lsusb -v" for the device. If we look at
similar information for other devices, we could find out how to
distinguish them.
--
Regards,
Pavel Roskin
On Sun, Dec 28, 2008 at 05:14:54PM -0500, Pavel Roskin wrote:
> On Sun, 2008-12-28 at 10:34 +0100, Ivo van Doorn wrote:
> > On Sunday 28 December 2008, Piter PUNK wrote:
>
> > > If the those 148f:2573 USB devices doesn't exists or
> > > if they exists and works fine with rt73usb, can we
> > > remove 148f:2573 from rt2500usb?
> >
> > No we can't, you can blame manufacturers for shipping
> > USB sticks with different chipsets but the exact same USB ID.
>
> Maybe there are some hints on the USB level that would make it possible
> to distinguish between the devices without loading the driver?
> Something like the number of endpoints? Then we could ask USB
> developers to provide a way to specify them in the USB devece table.
Seems like a trouble to solve a non-problem...
> > On Sunday 28 December 2008, Piter PUNK wrote:
> > > Loading rt2500usb AND rt73usb don't have any functional
> > > problem. All wireless devices works fine. But is very
> > > annoying the additional modules loaded.
I'm not sure I see how it is annoying, but I'm sorry it annoys you.
Nevertheless, I don't think it is worth much effort to fix it.
John
--
John W. Linville Linux should be at the core
[email protected] of your literate lifestyle.