Then Intel Wireless Pro 4965 and 3945 do not capture any other pakets except beacon frames when in monitor mode when using the gentoo-sources kernel driver. The bug has been reported the the Intel wireless driver team and a tested patch exists. But it looks like it is not yet put into the offical driver not the kernel 2.6.25. It would be nice if you gentoo guys could add it at least for 2.6.25 Reproducible: Always Steps to Reproduce: 1. Compile the kernel driver for Intel Wireless Pro 4965 in sys-kernel/gentoo-sources 2. Set the wireless interface into monitor mode using 'iwconfig wlan0 mode monitor' 3. Use wireshark or any abritary tool to read from the interface. Actual Results: Only beacons of other APs are captured, no traffic. Expected Results: All the other pakets should be also captured. http://www.intellinuxwireless.org/bugzilla/show_bug.cgi?id=1537
Well, our policy is usually to not add patches unless they make the mainline development tree. But the changes in iwlwifi from 2.6.25 to dev tree are so numerous and I don't see how this could even be backported to 2.6.25 once it actually makes it to 2.6.26 dev. So, I'm going to punt (American football reference) to Daniel. :)
I must have been seriously drunk when I wrote the bug-report o.o Sorry about the language. Anyways: Somehow the official patch from the intel bug-report does not work on my 2.6.25 sources =/ I guess that means I have to patch it by hand (ugh x.x) or wait for the 2.6.26 kernel. But I am not too happy always using the bleeding edge kernel just to get my hardware running. Michael
Created attachment 158289 [details, diff] Warning, needs cleanup The original Intel patch does not work with the version of iwlwifi in kernel 2.6.25 The linenumbers were off, one context was not right and a variable was renamed. I manually patched my sources for the iwl4965 driver, but not the 3965 driver (but manually patching it should be just as easy). I created a patch of my current 2.6.25-r4 sources against the 2.6.25-r5 sources (the Intel drivers did not change). This is probally still the wrong way to create a patch, but I hope someone might clean up my file :-) The patch works flawlessly for me, the dirver captures everything on the current channel.
It looks like the patch hasn't been sent/accepted upstream. Can you confirm that the bug still exists on the latest 2.6.26-rc kernel?
I can confirm that this problem still exists in the new kernel 2.6.26. The patch does not work for kernel 2.6.26. I will try to make a new one later.
I think this might be the fix: http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff_plain;h=25b3f57c1f4572d9b2d5671d60e99f26d2a7a26b Which is part of the latest git-sources (latest right now is git-sources-2.6.27_rc3-r2). Can someone please test this kernel version.
Please reopen after you've tested and report the findings.
Works flawlessly with recent kernel versions. (>=2.6.27 I think)