Mark Fletcher wrote:
I haven't made much headway with this, other than there does seem to be a change between the sarge gcc and the etch / sid gcc in this area.--- tom arnall <kloro@cox.net> wrote:Looks like whether or not there's a working Linux driver for your wireless card depends on thehardwarerevision of your particular card. For your cardthereappear to be 3 revisions and only 1 (C1) isreportedto have a working native Linux driver. Check outthislink to help you figure out which revision youhave:http://support.dlink.com/products/DWLG650.asp If you are lucky and have the C1 revision, use the MadWiFi Linux driver. If not, try the Prism orMadwifidrivers and if you can't get them to work, fallbackto the ndiswrapper approach -- but if at allpossibleuse a native Linux driver. Once you've figured out if you have the C1revision ornot, post back here if you need help with gettingthedriver included in the kernel and other setup, orifyou need help with ndiswrapper. HTH Markindeed, i could use some help setting up the madwifidriver. specifically, in the error output from the compiler:OK let me look into this and get back to you.Mark
Since I'm running sarge I can't immediately try it myself.One idea though -- given the level of support and Debian-friendliness I've found from the last few days' googling for Madwifi, I'd be stunned if the guys who put this driver together would deliver something that wouldn't work with etch or sid (the usual disclaimers about testing and unstable notwithstanding). Especially given that the available packages are in /unstable and as such are targetting sid...
There's a sarge-friendly version out there I noticed from my googling -- where did you get the version of the driver you are trying to compile? Is there any chance you've picked up a sarge version of the driver and are trying to compile it under etch or sid?
Just a thought... Mark
make[2]: Entering directory `/usr/src/kernel-headers-2.6.8-2-386'CC [M] /home/kloro/linux/madwifi-ng/ath_hal/ah_osdep.oIn file included from include/asm/thread_info.h:16, from include/linux/thread_info.h:21, from include/linux/spinlock.h:12, from include/linux/capability.h:45, from include/linux/sched.h:7, from include/linux/module.h:10, from /home/kloro/linux/madwifi-ng/ath_hal/ah_osdep.c:46: include/asm/processor.h:93: error: array type has incomplete element type make[3]: *** [/home/kloro/linux/madwifi-ng/ath_hal/ah_osdep.o] Error 1 make[2]: *** [_module_/home/kloro/linux/madwifi-ng/ath_hal] Error 2 make[2]: Leaving directory `/usr/src/kernel-headers-2.6.8-2-386' make[1]: *** [all] Error 2 make[1]: Leaving directory `/home/kloro/linux/madwifi-ng/ath_hal' make: *** [modules] Error 1 the line 'include/asm/processor.h:93: error: arraytype has incomplete element type' refers to the following bit of code in'processor.h': extern struct tss_struct init_tss[NR_CPUS]; i did further research which seems to indicate thatversion 4 of the gcc compiler doesn't like declarations of the form:extern struct foobar array[]; but wants to see instead the form: extern struct foobar *array; but i don't know enough about gcc to determine ifindeed this kind of problem is the problem with the declaration in 'processor.h'. and if it is, how does the principle apply to the declaration in 'processor.h'? i.e., wtf do i need to change in the header file? further, i cannot find on the internet anything like a comprehensive list of gcc error messages.thanks in advance for your ideas, tom arnall north spit, ca -- To UNSUBSCRIBE, email todebian-user-REQUEST@lists.debian.org with a subject of "unsubscribe". Trouble? Contactlistmaster@lists.debian.org