lp:madwifi
- Get this branch:
- bzr branch lp:madwifi
Branch merges
Related bugs
Related blueprints
Branch information
Import details
Last successful import was .
Whiteboard
branch no longer exists
Recent revisions
- 1030. By mentor
-
Don't add a WDS node if it is not necessary when initialising the state machine.
Signed-off-by: Jiri Fojtasek <email address hidden>
#1578Edited for trunk and error message added by me. The formatting, also by me, makes me cry.
- 1029. By mtaylor
-
Backport of changes from madwifi-dfs that should
really have been put into trunk instead.Adds per-packet noise and TSF calculations.
Benoit and I worked on this as part of getting the
necessary radar detection algorithms to work.The TSF adjustment is context aware and can at least
keep TSF values from ever going backwards over successive
packets when the 15-bit hardware counter for RX packets
has rolled over several times between interrupts.- M
- 1028. By mtaylor
-
This patch fixes a bug in the usage of wireless extensions to pass structures
via driver-specific extensions (iwpriv).The structures were being identified to wext as an array of bytes, which was ok
as long as the byte count fit within the bit field (see IW_PRIV_SIZE_MASK). If
the size exceeded the bit field, we could get a buffer overflow.WEXT looks at the size of the element and the number of elements and allocates a
buffer large enough to hold the data. It handles copying the data to/from the
user's provided buffer and provides an "extra" pointer to the handler.The poblem would occur when the request lists a buffer length greater than this
maximum number of bytes (again, see IW_PRIV_SIZE_MASK) but the actual buffer
provided by wext was smaller.This patch provides macros for constructing size and type encodings for structs
to be exposed in this manner to wireless extensions.It prefers to list structures as byte arrays whenever possible, but degenerates
to 4-byte integer arrays when necessary in order to correctly encode the buffer
space needed. - 1024. By mickflemm
-
* Add write capability to ath_info
* Does not need driver to be loaded anymore
* Added man page for ath_info
Thanx to Joerg Albert for his work ! - 1023. By mentor
-
Decapsulate packets using a DEC's OUI in SNAP packets. As per discussion on madwifi-devel, subject: "possible error when de-capsulating packets?"
- 1022. By derek
-
Improve minstrel algorithm so throughput is improved for
those times when the link is ideal.
Improve startup performance, so initially tries the lower rates,
rather than repeatedly pounding on the very high (and probably not working) rates.
Algorithm will quickly determine ideal rates to use.
Fix spelling errors, document revised algorithm.