summary |
shortlog | log |
commit |
commitdiff |
tree
first ⋅ prev ⋅ next
proski [Fri, 2 Nov 2007 02:05:22 +0000 (02:05 +0000)]
Fix support for big-endian systems
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2818
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Thu, 1 Nov 2007 22:25:15 +0000 (22:25 +0000)]
Fix support for 64-bit systems
Fix all uses of long integers for 32-bit values.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2817
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Thu, 1 Nov 2007 21:15:28 +0000 (21:15 +0000)]
Big cleanup of tools/ath_info.c
Reformat with madwifi-indent, it was needed badly. Break some long
lines. Avoid complex structure initializers that GNU indent cannot
format. Avoid using macro for the names table. Remove spaces in the
names, add then with formatted output. Remove spaces before punctuation
signs.
Fix all sparse warnings. Default to base 16 for all numbers specified
on the command line.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2816
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Wed, 31 Oct 2007 05:39:41 +0000 (05:39 +0000)]
Fix sysctl support in Linux 2.6.24
Linux 2.6.24 introduces a sysctl checker that disallows using arbitrary
ctl_name values for sysctl entries. The entries with non-standard
values should set ctl_name to CTL_UNNUMBERED.
On the other hand, using consistent non-zero ctl_name values is required
for older kernels (tested on Linux 2.4.33.3 from Slackware 11).
Thus the solution is to use CTL_UNNUMBERED on Linux 2.6.24+ and the
original values in the older kernels. Move CTL_AUTO and DEV_ATH to
include/compat.h and define them correspondingly.
When copying sysctl entries, check procname for being non-zero, since
ctl_name is zero (CTL_UNNUMBERED) on Linux 2.6.24 and newer.
For the same reason, don't use ctl_name to distinguish entries in
ath_sysctl_template, use extra2 instead.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2814
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mtaylor [Mon, 29 Oct 2007 22:34:55 +0000 (22:34 +0000)]
When a channel switch announcement (CSA) happens either through an action frame or a beacon, the sampling algorithm can get back responses with a rate of zero which result in rate index lookups that fail with -1 as a return value.
We just need to ignore these, as the sample rate control algorithm will recover quickly.
Previously, after a channel switch due to radar or command line channel change, we would see lots of error messages spewed to the logs about invalid rate indexes from the sample alg. This just squelches those messages.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2799
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mtaylor [Mon, 29 Oct 2007 22:33:03 +0000 (22:33 +0000)]
Our ABI for hostapd currently uses an old structure that only has one byte for the key index. This means that the 'magic value' of -1 (255) for NONE is passed rather than the larger type we actually use internally. To make comparisons work properly, I'm putting the magic value back at (uint8_t)-1 again so that a key index of 'none' with a MAC address that is passed into madwifi is interpreted correctly and the keys for that station are properly removed, rather than errors returned.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2798
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Mon, 29 Oct 2007 04:19:27 +0000 (04:19 +0000)]
Fix oops due to sysctl table check failure in Linux 2.6.24
Linux 2.6.24 has a sysctl table checker that rejects the tables
presented by the driver. Make the code survive this condition. Free
devname (or dev_name) before freeing the table. Don't do it at the
unregister time, it's too late.
The sysctl tables still need to be fixed.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2795
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mtaylor [Sun, 28 Oct 2007 20:40:04 +0000 (20:40 +0000)]
Fix for a false replay detection when seq was initialized to zero and first packet is also seq zero.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2794
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mtaylor [Sun, 28 Oct 2007 17:55:54 +0000 (17:55 +0000)]
Add indication of radar, half and quarter rates to wlanconfig list chan
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2793
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mentor [Sat, 27 Oct 2007 17:58:32 +0000 (17:58 +0000)]
Move acquisition of node reference outside macro to ensure node is referenced only once.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2792
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Wed, 17 Oct 2007 16:59:45 +0000 (16:59 +0000)]
Linux 2.6.24 moves proc_net inside init_net
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2756
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Wed, 17 Oct 2007 03:39:50 +0000 (03:39 +0000)]
Fix compilation for the forthcoming Linux 2.6.24
SET_MODULE_OWNER() is not defined in 2.6.24, but it's already a no-op in
2.6.23. Make sure that it will be a no-op in 2.6.23 and newer kernels.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2754
0192ed92-7a03-0410-a25b-
9323aeb14dbd
br1 [Tue, 16 Oct 2007 12:32:26 +0000 (12:32 +0000)]
add dump eeprom option (-d) to ath_info to dump the entire contents
of the eeprom to the screen and to a file
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2750
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mentor [Sun, 14 Oct 2007 17:48:45 +0000 (17:48 +0000)]
Try not to dereference null pointers; the loop above might not have run
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2745
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mentor [Sat, 13 Oct 2007 21:19:49 +0000 (21:19 +0000)]
Refactor beacon buffer generating loop for readability
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2742
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mentor [Thu, 11 Oct 2007 15:40:49 +0000 (15:40 +0000)]
If a rates information element is received that is larger than we will accept, simply take the first maximum size elements and continue, rather than BUG'ing out.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2736
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mentor [Mon, 8 Oct 2007 12:02:22 +0000 (12:02 +0000)]
Significantly lower the PCI_LATENCY_TIMER value that we were setting. This is essentially hte total time we are allowed to hold onto the PCI bus, before we see if anyone else wants a go. Higher means better throuhgput at the expense of the latency of other devices, and vice versa.
This value is roughly around the value that other devices are setting. The previous value appeared to be stupidly high.
This may solve some interactivity problems and machine performance issues.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2732
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mentor [Mon, 8 Oct 2007 11:38:21 +0000 (11:38 +0000)]
Function declaration whitespace
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2731
0192ed92-7a03-0410-a25b-
9323aeb14dbd
proski [Fri, 5 Oct 2007 17:01:13 +0000 (17:01 +0000)]
Fix lock context imbalance in ap_end() found by sparse
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2726
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mtaylor [Fri, 5 Oct 2007 05:52:42 +0000 (05:52 +0000)]
Added support for AP mode scan results to be updated and returned to
wireless tools and other clients. This fixes:
bgscan
iwlist <iface> scan
wlanconfig <iface> list scan
wlanconfig <iface> list peers
etc...
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2724
0192ed92-7a03-0410-a25b-
9323aeb14dbd
mrenzmann [Thu, 4 Oct 2007 13:07:51 +0000 (13:07 +0000)]
Restructuring the repository layout in response to ath5k.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@2721
0192ed92-7a03-0410-a25b-
9323aeb14dbd