From: Andrew \ruskie\ L. <ru...@ma...> - 2005-09-16 17:57:23
|
I have recently bought a stir4210 usb 2.0 dongle and can't get it configu= red. There is a GPLed driver available but it seems it's only for the 2.4 kern= el. And I'm sad to say I'm no coder... I was trying to get it to work with the stir4200 driver from the 2.6.12-rc6 kernel but had no luck. It does detect it, and irdadump is finding the computer from which it's running b= ut it doesn't detect any irda devices I'm throwing at it(atm only a nokia 61= 10). This is the device I have: http://www.equip-info.net/english/index.php?main=3D0&suche=3Dirda&prod=3D= 1508 Also on that site is a link to the drivers. If you require more info please let me know what I can do to help. --=20 Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id =3D 9A5117F8 Key fingerprint =3D 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 |
From: Andrew \ruskie\ L. <ru...@ma...> - 2005-09-16 18:32:16
|
Andrew "ruskie" Levstik wrote : > I have recently bought a stir4210 usb 2.0 dongle and can't get it confi= gured. > There is a GPLed driver available but it seems it's only for the 2.4 ke= rnel. > And I'm sad to say I'm no coder... I was trying to get it to work with > the stir4200 driver from the 2.6.12-rc6 kernel but had no luck. It does > detect it, and irdadump is finding the computer from which it's running= but > it doesn't detect any irda devices I'm throwing at it(atm only a nokia = 6110). >=20 > This is the device I have: > http://www.equip-info.net/english/index.php?main=3D0&suche=3Dirda&prod=3D= 1508 >=20 > Also on that site is a link to the drivers. >=20 > If you require more info please let me know what I can do to help. >=20 Ups sorry... forgot to include the dmesg output along with some errors(?) irda_init() NET: Registered protocol family 23 usbcore: registered new driver stir4200 usb 1-4: USB disconnect, address 9 usb 1-4: new high speed USB device using ehci_hcd and address 10 IRDA-USB found at address 10, Vendor: 66f, Product: 4210 irda_usb_parse_endpoints(), And our endpoints are : in=3D02, out=3D01 (51= 2), int=3D00 Assertion failed! drivers/net/irda/irda-usb.c:irda_usb_parse_endpoints:12= 52=20 self->bulk_out_mtu =3D=3D 64 irda_usb_init_qos(), dongle says speed=3D0x17F, size=3D0x20, window=3D0x8= , bofs=3D0x1, turn=3D0x4 IrDA: Registered device irda0 usbcore: registered new driver irda-usb USB IrDA support registered IrCOMM protocol (Dag Brattli) irlap_change_speed(), setting speed to 9600 ircomm_tty_attach_cable() ircomm_tty_ias_register() ircomm_tty_close() ircomm_tty_shutdown() ircomm_tty_detach_cable() ircomm_close() __ircomm_tty_cleanup() ircomm_tty_shutdown() usbcore: deregistering driver irda-usb irda_usb_receive(), RX status -108,transfer_flags 0x0000=20 irda_usb_submit(), Failed to submit Rx URB -90 irda_usb_disconnect(), USB IrDA Disconnected usbcore: deregistering driver stir4200 NET: Unregistered protocol family 23 --=20 Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id =3D 9A5117F8 Key fingerprint =3D 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 |
From: William A. <wl...@ea...> - 2005-09-17 19:27:15
Attachments:
usbirda
usbirda.usermap
|
You need hotplug and a current version of irdautils. If you are still using devfsd, your driver module will get loaded. If udev, be sure hotplug loads it. Put the attached scripts in the appropriate hotplug locations. Reboot. ----- Original Message ----- From: "Andrew "ruskie" Levstik" <ru...@ma...> To: <ird...@li...> Sent: Friday, September 16, 2005 11:31 AM Subject: Re: [irda-users] stir4210 device Andrew "ruskie" Levstik wrote : > I have recently bought a stir4210 usb 2.0 dongle and can't get it > configured. > There is a GPLed driver available but it seems it's only for the 2.4 > kernel. > And I'm sad to say I'm no coder... I was trying to get it to work with > the stir4200 driver from the 2.6.12-rc6 kernel but had no luck. It does > detect it, and irdadump is finding the computer from which it's running > but > it doesn't detect any irda devices I'm throwing at it(atm only a nokia > 6110). > > This is the device I have: > http://www.equip-info.net/english/index.php?main=0&suche=irda&prod=1508 > > Also on that site is a link to the drivers. > > If you require more info please let me know what I can do to help. > Ups sorry... forgot to include the dmesg output along with some errors(?) irda_init() NET: Registered protocol family 23 usbcore: registered new driver stir4200 usb 1-4: USB disconnect, address 9 usb 1-4: new high speed USB device using ehci_hcd and address 10 IRDA-USB found at address 10, Vendor: 66f, Product: 4210 irda_usb_parse_endpoints(), And our endpoints are : in=02, out=01 (512), int=00 Assertion failed! drivers/net/irda/irda-usb.c:irda_usb_parse_endpoints:1252 self->bulk_out_mtu == 64 irda_usb_init_qos(), dongle says speed=0x17F, size=0x20, window=0x8, bofs=0x1, turn=0x4 IrDA: Registered device irda0 usbcore: registered new driver irda-usb USB IrDA support registered IrCOMM protocol (Dag Brattli) irlap_change_speed(), setting speed to 9600 ircomm_tty_attach_cable() ircomm_tty_ias_register() ircomm_tty_close() ircomm_tty_shutdown() ircomm_tty_detach_cable() ircomm_close() __ircomm_tty_cleanup() ircomm_tty_shutdown() usbcore: deregistering driver irda-usb irda_usb_receive(), RX status -108,transfer_flags 0x0000 irda_usb_submit(), Failed to submit Rx URB -90 irda_usb_disconnect(), USB IrDA Disconnected usbcore: deregistering driver stir4200 NET: Unregistered protocol family 23 -- Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id = 9A5117F8 Key fingerprint = 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 ------------------------------------------------------- SF.Net email is sponsored by: Tame your development challenges with Apache's Geronimo App Server. Download it for free - -and be entered to win a 42" plasma tv or your very own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php _______________________________________________ irda-users mailing list ird...@li... http://lists.sourceforge.net/lists/listinfo/irda-users |
From: Andrew \ruskie\ L. <ru...@ma...> - 2005-09-17 22:29:57
|
> You need hotplug and a current version of irdautils. > If you are still using devfsd, your driver module will get loaded. > If udev, be sure hotplug loads it. > Put the attached scripts in the appropriate hotplug locations. > Reboot. >=20 Thank you... But sadly... it didn't work.... :( I have hotplug... and irda-utils 0.9.16. So I wonder if I'm doing something wrong. The thing is that the GPLed driver I got with is is for the 2.4 kernel on= ly I'm using 2.6.12-rc6 so I can't even build it... and this is a 4210 devic= e. And the kernel already autoloaded(I just love kernel autoloading) the required modules. So not sure what else I could do... btw should it create a /dev device? --=20 Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id =3D 9A5117F8 Key fingerprint =3D 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 |
From: William A. <wl...@ea...> - 2005-09-18 01:47:56
|
it appears that the difference between the 4200 and the 4210 is that the 4210 is a USB 2.0 device capable of 16m transfer speeds. to support the 4200 i included usb support for uhci, ehci and ohci. not all three are necessary in my case but you certainly will need 2.0 support. udev device discovery does create an entry in /dev for devices that are present at boot(which ones and how depend on its configuration files). it is a little slow on the uptake for devices which are hotplugged. i don't run the daemon so it usually takes udev restart to make it reread /sys and find the new entries. additionally these irda options are set: CONFIG_IRDA=m CONFIG_IRNET=m CONFIG_IRCOMM=m CONFIG_IRDA_CACHE_LAST_LSAP=y CONFIG_IRDA_FAST_RR=y CONFIG_IRDA_DEBUG=y CONFIG_IRTTY_SIR=m CONFIG_SIGMATEL_FIR=m my guess is that the 2.6 4200 driver should work for you unless the 4210 returns a specific signature as a consequence of a device probe the driver simply can not recognize. in any event the core usb drivers (in your case usb 2.0) should see the device if it's plugged in and you should be able to see that in /sys if you have it mounted to the kernel's /proc device directory. ----- Original Message ----- From: "Andrew "ruskie" Levstik" <ru...@ma...> To: <ird...@li...> Sent: Saturday, September 17, 2005 3:08 PM Subject: Re: [irda-users] stir4210 device > You need hotplug and a current version of irdautils. > If you are still using devfsd, your driver module will get loaded. > If udev, be sure hotplug loads it. > Put the attached scripts in the appropriate hotplug locations. > Reboot. > Thank you... But sadly... it didn't work.... :( I have hotplug... and irda-utils 0.9.16. So I wonder if I'm doing something wrong. The thing is that the GPLed driver I got with is is for the 2.4 kernel only I'm using 2.6.12-rc6 so I can't even build it... and this is a 4210 device. And the kernel already autoloaded(I just love kernel autoloading) the required modules. So not sure what else I could do... btw should it create a /dev device? -- Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id = 9A5117F8 Key fingerprint = 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 ------------------------------------------------------- SF.Net email is sponsored by: Tame your development challenges with Apache's Geronimo App Server. Download it for free - -and be entered to win a 42" plasma tv or your very own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php _______________________________________________ irda-users mailing list ird...@li... http://lists.sourceforge.net/lists/listinfo/irda-users |
From: William A. <wl...@ea...> - 2005-09-18 03:04:16
|
Andrew: I missed this post earlier. The 4200 driver sees the device and appears to be having considerable negotiation problems with it. if changing the irda parameters doesn't clear it up(unlikely), then upgrading the 4200 driver or bringing the 2.4 4210 driver forward may be your only alternative. the net drivers create a 'netdev' symbolic like 'eth0' which 'irattach' and ifconfig manipulate. unlikely you'll ever see anything in /dev for these. i don't think the changes to the 4200 driver would be too extensive. famous last words. ----- Original Message ----- From: "Andrew "ruskie" Levstik" <ru...@ma...> To: <ird...@li...> Sent: Friday, September 16, 2005 11:31 AM Subject: Re: [irda-users] stir4210 device Andrew "ruskie" Levstik wrote : > I have recently bought a stir4210 usb 2.0 dongle and can't get it > configured. > There is a GPLed driver available but it seems it's only for the 2.4 > kernel. > And I'm sad to say I'm no coder... I was trying to get it to work with > the stir4200 driver from the 2.6.12-rc6 kernel but had no luck. It does > detect it, and irdadump is finding the computer from which it's running > but > it doesn't detect any irda devices I'm throwing at it(atm only a nokia > 6110). > > This is the device I have: > http://www.equip-info.net/english/index.php?main=0&suche=irda&prod=1508 > > Also on that site is a link to the drivers. > > If you require more info please let me know what I can do to help. > Ups sorry... forgot to include the dmesg output along with some errors(?) irda_init() NET: Registered protocol family 23 usbcore: registered new driver stir4200 usb 1-4: USB disconnect, address 9 usb 1-4: new high speed USB device using ehci_hcd and address 10 IRDA-USB found at address 10, Vendor: 66f, Product: 4210 irda_usb_parse_endpoints(), And our endpoints are : in=02, out=01 (512), int=00 Assertion failed! drivers/net/irda/irda-usb.c:irda_usb_parse_endpoints:1252 self->bulk_out_mtu == 64 irda_usb_init_qos(), dongle says speed=0x17F, size=0x20, window=0x8, bofs=0x1, turn=0x4 IrDA: Registered device irda0 usbcore: registered new driver irda-usb USB IrDA support registered IrCOMM protocol (Dag Brattli) irlap_change_speed(), setting speed to 9600 ircomm_tty_attach_cable() ircomm_tty_ias_register() ircomm_tty_close() ircomm_tty_shutdown() ircomm_tty_detach_cable() ircomm_close() __ircomm_tty_cleanup() ircomm_tty_shutdown() usbcore: deregistering driver irda-usb irda_usb_receive(), RX status -108,transfer_flags 0x0000 irda_usb_submit(), Failed to submit Rx URB -90 irda_usb_disconnect(), USB IrDA Disconnected usbcore: deregistering driver stir4200 NET: Unregistered protocol family 23 -- Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id = 9A5117F8 Key fingerprint = 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 ------------------------------------------------------- SF.Net email is sponsored by: Tame your development challenges with Apache's Geronimo App Server. Download it for free - -and be entered to win a 42" plasma tv or your very own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php _______________________________________________ irda-users mailing list ird...@li... http://lists.sourceforge.net/lists/listinfo/irda-users |
From: Thomas I. <il...@gm...> - 2005-09-18 22:21:36
|
AFAIK the STIR 4200 and STIR 421X devices are completely different: the 4200 has some proprietary (but documented) usb interface (eg. register settings, etc) while the 421X series conform to the irda-usb class specification. BUT they need some firmware patches loaded before they can be used. I think the driver provided by sigmatel is a patched version of Linux' usb-irda driver that will support the preloading of the firmware patches. maybe these changes have made there way into the most recent stock linux usb-irda drivers, (i have no idea if they did recently, but they are missing at least in 2.6.12) but anyway, you need these firmware patches. They are _NOT_ GPL, and can be downloaded from http://www.sigmatel.com/documents/stir4210_4220_4116_patch_files.tar.gz as they are not GPL'ed, i suppose that your distribution does not include them, so you need to download these firmware files, install them soewhere, and set up hotplug to load these patches. if sigmatel's patches have not found their way into the stock kernel driver and you try to write support on your own, you need to modify the usb-irda driver, not the stir4200 driver. good luck, tom William Adams schrieb: > Andrew: > > I missed this post earlier. > The 4200 driver sees the device and appears to be having considerable > negotiation problems with it. if changing the irda parameters doesn't > clear it up(unlikely), then upgrading the 4200 driver or bringing > the 2.4 4210 driver forward may be your only alternative. > > the net drivers create a 'netdev' symbolic like 'eth0' which 'irattach' > and ifconfig manipulate. unlikely you'll ever see anything in /dev > for these. > > i don't think the changes to the 4200 driver would be too extensive. > > famous last words. > ----- Original Message ----- From: "Andrew "ruskie" Levstik" > <ru...@ma...> > To: <ird...@li...> > Sent: Friday, September 16, 2005 11:31 AM > Subject: Re: [irda-users] stir4210 device > > > Andrew "ruskie" Levstik wrote : > >> I have recently bought a stir4210 usb 2.0 dongle and can't get it >> configured. >> There is a GPLed driver available but it seems it's only for the 2.4 >> kernel. >> And I'm sad to say I'm no coder... I was trying to get it to work with >> the stir4200 driver from the 2.6.12-rc6 kernel but had no luck. It does >> detect it, and irdadump is finding the computer from which it's >> running but >> it doesn't detect any irda devices I'm throwing at it(atm only a >> nokia 6110). >> >> This is the device I have: >> http://www.equip-info.net/english/index.php?main=0&suche=irda&prod=1508 >> >> Also on that site is a link to the drivers. >> >> If you require more info please let me know what I can do to help. >> > > Ups sorry... forgot to include the dmesg output along with some errors(?) > irda_init() > NET: Registered protocol family 23 > usbcore: registered new driver stir4200 > usb 1-4: USB disconnect, address 9 > usb 1-4: new high speed USB device using ehci_hcd and address 10 > IRDA-USB found at address 10, Vendor: 66f, Product: 4210 > irda_usb_parse_endpoints(), And our endpoints are : in=02, out=01 > (512), int=00 > Assertion failed! > drivers/net/irda/irda-usb.c:irda_usb_parse_endpoints:1252 > self->bulk_out_mtu == 64 > irda_usb_init_qos(), dongle says speed=0x17F, size=0x20, window=0x8, > bofs=0x1, turn=0x4 > IrDA: Registered device irda0 > usbcore: registered new driver irda-usb > USB IrDA support registered > IrCOMM protocol (Dag Brattli) > irlap_change_speed(), setting speed to 9600 > ircomm_tty_attach_cable() > ircomm_tty_ias_register() > ircomm_tty_close() > ircomm_tty_shutdown() > ircomm_tty_detach_cable() > ircomm_close() > __ircomm_tty_cleanup() > ircomm_tty_shutdown() > usbcore: deregistering driver irda-usb > irda_usb_receive(), RX status -108,transfer_flags 0x0000 > irda_usb_submit(), Failed to submit Rx URB -90 > irda_usb_disconnect(), USB IrDA Disconnected > usbcore: deregistering driver stir4200 > NET: Unregistered protocol family 23 > > |
From: William A. <wl...@ea...> - 2005-09-18 22:54:54
|
an examination of the sigmatel 4210 driver suggests this is correct. however it seems the loading of the firmware is about all it does do. stripping the surrounding wrapper from the sigmatel patch files, uppercasing them and using fxload via hotplug to load them is what comes to mind along with a few parameter changes to the 4200 driver. it's not guaranteed to work and the non gpl nature of the patch files is certainly a problem. the idea of a gpl driver that does little besides loading non gpl code is disturbing. ----- Original Message ----- From: "Thomas Ilnseher" <il...@gm...> To: <ird...@li...> Sent: Sunday, September 18, 2005 3:21 PM Subject: Re: [irda-users] stir4210 device > AFAIK the STIR 4200 and STIR 421X devices are completely different: > > the 4200 has some proprietary (but documented) usb interface (eg. register > settings, etc) > while the 421X series conform to the irda-usb class specification. BUT > they need some > firmware patches loaded before they can be used. I think the driver > provided by sigmatel > is a patched version of Linux' usb-irda driver that will support the > preloading of the firmware > patches. maybe these changes have made there way into the most recent > stock linux usb-irda drivers, > (i have no idea if they did recently, but they are missing at least in > 2.6.12) > but anyway, you need these firmware patches. They are _NOT_ GPL, and can > be downloaded from > > http://www.sigmatel.com/documents/stir4210_4220_4116_patch_files.tar.gz > > as they are not GPL'ed, i suppose that your distribution does not include > them, so you need to download these > firmware files, install them soewhere, and set up hotplug to load these > patches. > > if sigmatel's patches have not found their way into the stock kernel > driver and you try to write support on your own, > you need to modify the usb-irda driver, not the stir4200 driver. > > good luck, > > tom > > > William Adams schrieb: > >> Andrew: >> >> I missed this post earlier. >> The 4200 driver sees the device and appears to be having considerable >> negotiation problems with it. if changing the irda parameters doesn't >> clear it up(unlikely), then upgrading the 4200 driver or bringing >> the 2.4 4210 driver forward may be your only alternative. >> >> the net drivers create a 'netdev' symbolic like 'eth0' which 'irattach' >> and ifconfig manipulate. unlikely you'll ever see anything in /dev >> for these. >> >> i don't think the changes to the 4200 driver would be too extensive. >> >> famous last words. >> ----- Original Message ----- From: "Andrew "ruskie" Levstik" >> <ru...@ma...> >> To: <ird...@li...> >> Sent: Friday, September 16, 2005 11:31 AM >> Subject: Re: [irda-users] stir4210 device >> >> >> Andrew "ruskie" Levstik wrote : >> >>> I have recently bought a stir4210 usb 2.0 dongle and can't get it >>> configured. >>> There is a GPLed driver available but it seems it's only for the 2.4 >>> kernel. >>> And I'm sad to say I'm no coder... I was trying to get it to work with >>> the stir4200 driver from the 2.6.12-rc6 kernel but had no luck. It does >>> detect it, and irdadump is finding the computer from which it's running >>> but >>> it doesn't detect any irda devices I'm throwing at it(atm only a nokia >>> 6110). >>> >>> This is the device I have: >>> http://www.equip-info.net/english/index.php?main=0&suche=irda&prod=1508 >>> >>> Also on that site is a link to the drivers. >>> >>> If you require more info please let me know what I can do to help. >>> >> >> Ups sorry... forgot to include the dmesg output along with some errors(?) >> irda_init() >> NET: Registered protocol family 23 >> usbcore: registered new driver stir4200 >> usb 1-4: USB disconnect, address 9 >> usb 1-4: new high speed USB device using ehci_hcd and address 10 >> IRDA-USB found at address 10, Vendor: 66f, Product: 4210 >> irda_usb_parse_endpoints(), And our endpoints are : in=02, out=01 (512), >> int=00 >> Assertion failed! >> drivers/net/irda/irda-usb.c:irda_usb_parse_endpoints:1252 >> self->bulk_out_mtu == 64 >> irda_usb_init_qos(), dongle says speed=0x17F, size=0x20, window=0x8, >> bofs=0x1, turn=0x4 >> IrDA: Registered device irda0 >> usbcore: registered new driver irda-usb >> USB IrDA support registered >> IrCOMM protocol (Dag Brattli) >> irlap_change_speed(), setting speed to 9600 >> ircomm_tty_attach_cable() >> ircomm_tty_ias_register() >> ircomm_tty_close() >> ircomm_tty_shutdown() >> ircomm_tty_detach_cable() >> ircomm_close() >> __ircomm_tty_cleanup() >> ircomm_tty_shutdown() >> usbcore: deregistering driver irda-usb >> irda_usb_receive(), RX status -108,transfer_flags 0x0000 >> irda_usb_submit(), Failed to submit Rx URB -90 >> irda_usb_disconnect(), USB IrDA Disconnected >> usbcore: deregistering driver stir4200 >> NET: Unregistered protocol family 23 >> >> > > > > ------------------------------------------------------- > SF.Net email is sponsored by: > Tame your development challenges with Apache's Geronimo App Server. > Download it for free - -and be entered to win a 42" plasma tv or your very > own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php > _______________________________________________ > irda-users mailing list > ird...@li... > http://lists.sourceforge.net/lists/listinfo/irda-users |
From: Andrew \ruskie\ L. <ru...@ma...> - 2005-09-20 10:46:22
|
William Adams wrote : > an examination of the sigmatel 4210 driver suggests this is correct. > however it seems the loading of the firmware is about all it does do. > stripping the surrounding wrapper from the sigmatel patch files, > uppercasing them and using fxload via hotplug to load them is > what comes to mind along with a few parameter changes to the > 4200 driver. >=20 > it's not guaranteed to work and the non gpl nature of the patch > files is certainly a problem. >=20 > the idea of a gpl driver that does little besides loading non gpl > code is disturbing. >=20 > ----- Original Message ----- > [---=3D| TOFU protection by t-prot: 143 lines snipped |=3D---] >=20 >=20 So is there any way I can get this working. I can't code myself so I can't really do much myself... but I'd like to have this working. Any help would be much appreciated. --=20 Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id =3D 9A5117F8 Key fingerprint =3D 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 |
From: Stephen H. <she...@os...> - 2005-09-19 16:53:55
|
On Mon, 19 Sep 2005 00:21:27 +0200 Thomas Ilnseher <il...@gm...> wrote: > AFAIK the STIR 4200 and STIR 421X devices are completely different: > > the 4200 has some proprietary (but documented) usb interface (eg. > register settings, etc) > while the 421X series conform to the irda-usb class specification. BUT > they need some > firmware patches loaded before they can be used. I think the driver > provided by sigmatel > is a patched version of Linux' usb-irda driver that will support the > preloading of the firmware > patches. maybe these changes have made there way into the most recent > stock linux usb-irda drivers, > (i have no idea if they did recently, but they are missing at least in > 2.6.12) > but anyway, you need these firmware patches. They are _NOT_ GPL, and can > be downloaded from Then they are in violation of GPL, the existing irda-usb driver is GPL, and the copyright holders should contact gpl-violations.org. I am serious, vendors building proprietary drivers out of GPL code is wrong. |
From: Thomas I. <il...@gm...> - 2005-09-20 15:31:59
|
Stephen Hemminger schrieb: >On Mon, 19 Sep 2005 00:21:27 +0200 >Thomas Ilnseher <il...@gm...> wrote: > > > >>AFAIK the STIR 4200 and STIR 421X devices are completely different: >> >>the 4200 has some proprietary (but documented) usb interface (eg. >>register settings, etc) >>while the 421X series conform to the irda-usb class specification. BUT >>they need some >>firmware patches loaded before they can be used. I think the driver >>provided by sigmatel >>is a patched version of Linux' usb-irda driver that will support the >>preloading of the firmware >>patches. maybe these changes have made there way into the most recent >>stock linux usb-irda drivers, >>(i have no idea if they did recently, but they are missing at least in >>2.6.12) >>but anyway, you need these firmware patches. They are _NOT_ GPL, and can >>be downloaded from >> >> > >Then they are in violation of GPL, the existing irda-usb driver >is GPL, and the copyright holders should contact gpl-violations.org. > >I am serious, vendors building proprietary drivers out of GPL code >is wrong. > > Maybe i said it the wrong way, so you all get me wrong ;) this hardware, the STIR421x chips, are based upon some microprozessor core. this microprocessor core executes some code, called "firmware" which is also (?) built in the chip. but this firmware seams to be _B0RKED_ which means the device will not work out of the box. So the driver needs to upload a patch for this _firmware_ running on the device (aka on your irda dongle) so that the device works properly. after this has happened, the device works like any other irda-usb compilant device. so the Firmware patches do not violate GPL. look at this as something like an hardware upgrade. All that is executed by your host cpu (aka the intel / AMD/ via thingie) is GPL, as their modified driver _IS_ GPL. all what needs to be done is to update irda-usb driver in 2.6 to upload this firmware patches. get your hands off the STIR4200 driver, as the STIR4200's are totally different from the STIR421X devices ! using fxload is AFAIK not possible, cause fxload was designed to upload firmware into the Cypress AN2312 chips, which i think use a different method of uploading the firmware then this devices do. to say it short: SigmaTel is NOT wrong. what they did is legal and good. (My opinion) Tommy > >------------------------------------------------------- >SF.Net email is sponsored by: >Tame your development challenges with Apache's Geronimo App Server. >Download it for free - -and be entered to win a 42" plasma tv or your very >own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php >_______________________________________________ >irda-users mailing list >ird...@li... >http://lists.sourceforge.net/lists/listinfo/irda-users > > > > |
From: Stephen H. <she...@os...> - 2005-09-20 16:04:44
|
> Maybe i said it the wrong way, so you all get me wrong ;) Your right, I am wrong. No body but the Debian "all firmware must be open source" fanatics will care. The more general fix for this seems to be either fixing fxload and/or getting the hotplug firmware loader to work for USB devices. |
From: William A. <wl...@gm...> - 2005-09-20 16:24:23
|
----- Original Message ----- From: "Thomas Ilnseher" <il...@gm...> To: <ird...@li...> Sent: Tuesday, September 20, 2005 8:31 AM Subject: Re: [irda-users] stir4210 device > Stephen Hemminger schrieb: > >>On Mon, 19 Sep 2005 00:21:27 +0200 >>Thomas Ilnseher <il...@gm...> wrote: >> >> >>>AFAIK the STIR 4200 and STIR 421X devices are completely different: >>> >>>the 4200 has some proprietary (but documented) usb interface (eg. >>>register settings, etc) >>>while the 421X series conform to the irda-usb class specification. BUT >>>they need some >>>firmware patches loaded before they can be used. I think the driver >>>provided by sigmatel >>>is a patched version of Linux' usb-irda driver that will support the >>>preloading of the firmware >>>patches. maybe these changes have made there way into the most recent >>>stock linux usb-irda drivers, >>>(i have no idea if they did recently, but they are missing at least in >>>2.6.12) >>>but anyway, you need these firmware patches. They are _NOT_ GPL, and can >>>be downloaded from >>> >> >>Then they are in violation of GPL, the existing irda-usb driver >>is GPL, and the copyright holders should contact gpl-violations.org. >> >>I am serious, vendors building proprietary drivers out of GPL code >>is wrong. >> > Maybe i said it the wrong way, so you all get me wrong ;) > > this hardware, the STIR421x chips, are based upon some microprozessor > core. this microprocessor core > executes some code, called "firmware" which is also (?) built in the chip. > but this firmware seams to be > _B0RKED_ which means the device will not work out of the box. So the > driver needs to upload a patch for this > _firmware_ running on the device (aka on your irda dongle) so that the > device works properly. after this has happened, > the device works like any other irda-usb compilant device. > > so the Firmware patches do not violate GPL. look at this as something like > an hardware upgrade. > All that is executed by your host cpu (aka the intel / AMD/ via thingie) > is GPL, as their modified driver _IS_ GPL. > > all what needs to be done is to update irda-usb driver in 2.6 to upload > this firmware patches. > > get your hands off the STIR4200 driver, as the STIR4200's are totally > different from the STIR421X devices ! > if the 4200 and the 4210 were as different as you suggest, it seems unlikely that the 4200 driver could see which it obviously does. also the warning on the Sigmatel website that the 2.4 4210 driver will prevent your 4200 from working suggests that the 4200 will take a firmware upload just as the 4210 does. > using fxload is AFAIK not possible, cause fxload was designed to upload > firmware into the Cypress AN2312 chips, > which i think use a different method of uploading the firmware then this > devices do. > even if the 4210 doesn't support a cypress based load, an io primitive of the same kind that the 2.4 driver uses will work. if Sigmatel wants to load something into the device, why couch it in mystery? > to say it short: SigmaTel is NOT wrong. what they did is legal and good. > (My opinion) > > Tommy > >> >>------------------------------------------------------- >>SF.Net email is sponsored by: >>Tame your development challenges with Apache's Geronimo App Server. >>Download it for free - -and be entered to win a 42" plasma tv or your very >>own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php >>_______________________________________________ >>irda-users mailing list >>ird...@li... >>http://lists.sourceforge.net/lists/listinfo/irda-users >> >> >> > > > > ------------------------------------------------------- > SF.Net email is sponsored by: > Tame your development challenges with Apache's Geronimo App Server. > Download it for free - -and be entered to win a 42" plasma tv or your very > own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php > _______________________________________________ > irda-users mailing list > ird...@li... > http://lists.sourceforge.net/lists/listinfo/irda-users |
From: William A. <wl...@gm...> - 2005-09-20 16:04:38
|
----- Original Message ----- From: "Thomas Ilnseher" <il...@gm...> To: <ird...@li...> Sent: Tuesday, September 20, 2005 8:31 AM Subject: Re: [irda-users] stir4210 device > Stephen Hemminger schrieb: > >>On Mon, 19 Sep 2005 00:21:27 +0200 >>Thomas Ilnseher <il...@gm...> wrote: >> >> >>>AFAIK the STIR 4200 and STIR 421X devices are completely different: >>> >>>the 4200 has some proprietary (but documented) usb interface (eg. >>>register settings, etc) >>>while the 421X series conform to the irda-usb class specification. BUT >>>they need some >>>firmware patches loaded before they can be used. I think the driver >>>provided by sigmatel >>>is a patched version of Linux' usb-irda driver that will support the >>>preloading of the firmware >>>patches. maybe these changes have made there way into the most recent >>>stock linux usb-irda drivers, >>>(i have no idea if they did recently, but they are missing at least in >>>2.6.12) >>>but anyway, you need these firmware patches. They are _NOT_ GPL, and can >>>be downloaded from >>> >> >>Then they are in violation of GPL, the existing irda-usb driver >>is GPL, and the copyright holders should contact gpl-violations.org. >> >>I am serious, vendors building proprietary drivers out of GPL code >>is wrong. >> > Maybe i said it the wrong way, so you all get me wrong ;) > > this hardware, the STIR421x chips, are based upon some microprozessor > core. this microprocessor core > executes some code, called "firmware" which is also (?) built in the chip. > but this firmware seams to be > _B0RKED_ which means the device will not work out of the box. So the > driver needs to upload a patch for this > _firmware_ running on the device (aka on your irda dongle) so that the > device works properly. after this has happened, > the device works like any other irda-usb compilant device. > > so the Firmware patches do not violate GPL. look at this as something like > an hardware upgrade. > All that is executed by your host cpu (aka the intel / AMD/ via thingie) > is GPL, as their modified driver _IS_ GPL. > > all what needs to be done is to update irda-usb driver in 2.6 to upload > this firmware patches. > > get your hands off the STIR4200 driver, as the STIR4200's are totally > different from the STIR421X devices ! > if the 4200 and the 4210 were as different as you suggest, it seems unlikely that the 4200 driver could see which it obviously does. also the warning on the Sigmatel website that the 2.4 4210 driver will prevent your 4200 from working suggests that the 4200 will take a firmware upload just as the 4210 does. > using fxload is AFAIK not possible, cause fxload was designed to upload > firmware into the Cypress AN2312 chips, > which i think use a different method of uploading the firmware then this > devices do. even if the 4210 doesn't support a cypress based load, an io primitive of the same kind that the 2.4 driver uses will work. if Sigmatel wants to load something into the device, why couch it in mystery? > > to say it short: SigmaTel is NOT wrong. what they did is legal and good. > (My opinion) > > Tommy > >> >>------------------------------------------------------- >>SF.Net email is sponsored by: >>Tame your development challenges with Apache's Geronimo App Server. >>Download it for free - -and be entered to win a 42" plasma tv or your very >>own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php >>_______________________________________________ >>irda-users mailing list >>ird...@li... >>http://lists.sourceforge.net/lists/listinfo/irda-users >> >> >> > > > > ------------------------------------------------------- > SF.Net email is sponsored by: > Tame your development challenges with Apache's Geronimo App Server. > Download it for free - -and be entered to win a 42" plasma tv or your very > own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php > _______________________________________________ > irda-users mailing list > ird...@li... > http://lists.sourceforge.net/lists/listinfo/irda-users |
From: William A. <wl...@ea...> - 2005-09-20 11:44:46
|
----- Original Message ----- From: "Andrew "ruskie" Levstik" <ru...@ma...> To: <ird...@li...> Sent: Tuesday, September 20, 2005 3:44 AM Subject: Re: [irda-users] stir4210 device William Adams wrote : > an examination of the sigmatel 4210 driver suggests this is correct. > however it seems the loading of the firmware is about all it does do. > stripping the surrounding wrapper from the sigmatel patch files, > uppercasing them and using fxload via hotplug to load them is > what comes to mind along with a few parameter changes to the > 4200 driver. > > it's not guaranteed to work and the non gpl nature of the patch > files is certainly a problem. > > the idea of a gpl driver that does little besides loading non gpl > code is disturbing. > can the sigmatel patch files legally be used outside the scope of the sigmatel 4210 driver? if so, some alternatives may be available and it may take a little while. |
From: Andrew \ruskie\ L. <ru...@ma...> - 2005-09-20 12:00:30
|
William Adams wrote : >=20 > ----- Original Message ----- > From: "Andrew "ruskie" Levstik" <ru...@ma...> > To: <ird...@li...> > Sent: Tuesday, September 20, 2005 3:44 AM > Subject: Re: [irda-users] stir4210 device >=20 >=20 > William Adams wrote : >=20 > > an examination of the sigmatel 4210 driver suggests this is correct. > > however it seems the loading of the firmware is about all it does do. > > stripping the surrounding wrapper from the sigmatel patch files, > > uppercasing them and using fxload via hotplug to load them is > > what comes to mind along with a few parameter changes to the > > 4200 driver. > > > > it's not guaranteed to work and the non gpl nature of the patch > > files is certainly a problem. > > > > the idea of a gpl driver that does little besides loading non gpl > > code is disturbing. > > >=20 > can the sigmatel patch files legally be used outside the scope of the > sigmatel > 4210 driver? >=20 > if so, some alternatives may be available and it may take a little whil= e. >=20 >=20 > ------------------------------------------------------- > SF.Net email is sponsored by: > Tame your development challenges with Apache's Geronimo App Server. > Download it for free - -and be entered to win a 42" plasma tv or your v= ery > own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.p= hp > _______________________________________________ > irda-users mailing list > ird...@li... > http://lists.sourceforge.net/lists/listinfo/irda-users >=20 >=20 There is no COPYING or LICENSE file in the tarball of the patches availab= le from their site[http://www.sigmatel.com/products/tech-support.htm]: STIR4210/4220/4116 TA4 and TA5 Linux Patch Files (tar, Version 1.0, 10070= 4, 40 Kb) The file contains Firmware Patches for the STIR4210 TA4 and the STIR4210/= 4220/4116 TA5. This=20 patch is required for normal operation of the Linux Driver. http://www.sigmatel.com/documents/stir4210_4220_4116_patch_files.tar.gz As far as I can tell(from the README in the driver) the patch firmware is applied by the sgtlpatch application in the tools directory of the dri= ver. And they still haven't replied to my inquriy re: 2.6 support... I'll let = you know if I get a response from them in any way. --=20 Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id =3D 9A5117F8 Key fingerprint =3D 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 |
From: William A. <wl...@gm...> - 2005-09-21 04:00:45
|
a look into the history of the stir4200 suggests it is not and never will be irda spec compliant. kludges were installed in the irda code to make it run. the 4210 seems to have rectified this problem and with the exception of needing the patch will run without the aid of anything else. hotplug seems a viable option but testing will be necessary and i don't have a 4210. after the 1st, i'll try and bring the "driver" forward for you unless you have succeeded elsewhere. wouldn't you love to know what happens on Windows? ----- Original Message ----- From: "Andrew "ruskie" Levstik" <ru...@ma...> To: <ird...@li...> Sent: Tuesday, September 20, 2005 3:44 AM Subject: Re: [irda-users] stir4210 device William Adams wrote : > an examination of the sigmatel 4210 driver suggests this is correct. > however it seems the loading of the firmware is about all it does do. > stripping the surrounding wrapper from the sigmatel patch files, > uppercasing them and using fxload via hotplug to load them is > what comes to mind along with a few parameter changes to the > 4200 driver. > > it's not guaranteed to work and the non gpl nature of the patch > files is certainly a problem. > > the idea of a gpl driver that does little besides loading non gpl > code is disturbing. > > ----- Original Message ----- > [---=| TOFU protection by t-prot: 143 lines snipped |=---] > > So is there any way I can get this working. I can't code myself so I can't really do much myself... but I'd like to have this working. Any help would be much appreciated. -- Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id = 9A5117F8 Key fingerprint = 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 ------------------------------------------------------- SF.Net email is sponsored by: Tame your development challenges with Apache's Geronimo App Server. Download it for free - -and be entered to win a 42" plasma tv or your very own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php _______________________________________________ irda-users mailing list ird...@li... http://lists.sourceforge.net/lists/listinfo/irda-users |
From: Andrew \ruskie\ L. <ru...@ma...> - 2005-09-21 05:16:31
|
> a look into the history of the stir4200 suggests it > is not and never will be irda spec compliant. > kludges were installed in the irda code to make it run. > the 4210 seems to have rectified this problem and > with the exception of needing the patch will run > without the aid of anything else. hotplug seems a viable > option but testing will be necessary and i don't have a > 4210. after the 1st, i'll try and bring the "driver" forward > for you unless you have succeeded elsewhere. > wouldn't you love to know what happens on Windows? >=20 I'll prolly try it on my work machine which is still running windows to see how it works. If you need any output or maybe remote access to a machine with the stir4210 device I can offer it. --=20 Andrew "ruskie" Levstik Source Mage GNU/Linux Games grimoire guru Key id =3D 9A5117F8 Key fingerprint =3D 6731 FEF2 99A8 4672 5962 69AB 3DAF DA67 9A51 17F8 |