Re: [Madwifi-users] Problems with Madwifi and IBM 11 a/b/g Mini PCI Adapter II
Status: Beta
Brought to you by:
otaku
From: POiSOnER <poi...@po...> - 2005-07-28 14:44:21
|
Hi! On Thu, Jul 28, 2005 at 03:38:59PM +0200, Paolo wrote: > On Thu, Jul 28, 2005 at 12:28:24AM +0200, POiSOnER wrote: > > Debian GNU/Linux unstable=20 > > Linux 2.6.12.3 vanilla (compiled with gcc 4) >=20 > dunno is Debian/unstable is compiled w/ gcc-4.x - afaik should be gcc-3.x Not the whole system. But Debian/unstable switched to gcc-4.0.1. Most packets for unstable are now compiled with gcc-4.=20 > it's generally advisable to use the same compiler, same main release numb= er > at least. Others, also on other lists, got different behaviour going from > gcc-4 to gcc-3.[34]. > This problem might be unrelated, though on debugging it's worth reducing > number of varaibles at min ;) I know that the ABI has been changed from gcc-3.x to gcc-4.x. So I compiled the kernel with gcc-4 after upgrading to unstable. The effects when loading ath_pci with params are the same if kernel and madwifi are compiled with gcc-3.3 (standard on stable(sarge) and testing(et= ch)). iwlist ath0 channel is empty and the kernel produces an oops as soon as waproamd tries to access the device. Regards, Holger --=20 "A Microsoft Certified Systems Engineer is to computing expertise what a McDonalds Certified Food Specialist is to fine cuisine."=20 - J. R. Helgeson |