- Apr 13, 2011
-
-
roel authored
Check whether index is within bounds before testing the element Signed-off-by:
Roel Kluin <roel.kluin@gmail.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
roel authored
Check whether index is within bounds before testing the element Both spurChans arrays in modalHeader5G and modalHeader2G have 5 elements, AR_EEPROM_MODAL_SPURS is defined 5. So unless a break occurs, in the last iteration (i=5) we tried to access spurChansPtr[5] before testing whether i was within bounds. Fix this. Signed-off-by:
-
Felix Fietkau authored
While the rx/tx tasklet is pending, new unnecessary interrupts may arrive. Decrease the load by temporarily disabling the interrupts until the tasklet has completed. Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Signed-off-by:
-
Felix Fietkau authored
The pointers to the debugfs entries do not need to be saved, because they will be recursively removed when the wiphy is unregistered. Signed-off-by:
-
Felix Fietkau authored
Reusing the configured retry counts from the skb cb is more efficient than reloading the data from uncached memory. Replace ts_longretry (unused) with ts_final_retry which contains the retry count for the final rate only Signed-off-by:
-
Felix Fietkau authored
Use ACCESS_ONCE to reduce the number of redundant loads on uncached memory Signed-off-by:
-
Felix Fietkau authored
Use ACCESS_ONCE to reduce the number of variable reloads on uncached memory Signed-off-by:
-
Felix Fietkau authored
It is no longer necessary for preparing mac80211 tx status Signed-off-by:
-
Felix Fietkau authored
Use local variables to reduce the number of load/store operations on uncached memory. Signed-off-by:
-
Felix Fietkau authored
Set the slot time based on the mac80211 short slot vs long slot setting instead of just forcing long slot for all CCK-enabled channels. This slightly improves 802.11g mode performance in in my tests. Signed-off-by:
-
Felix Fietkau authored
ath5k uses 8 usec as a sifs time, extracted from the initvals, whereas the standard requires a sifs time of 10. The difference originates from the fact that the SIFS register has an offset of 2 usec. Fix the SIFS time definition to use the standard value of 10 usec and subtract 2 usecs when writing the SIFS register. Signed-off-by:
-
Felix Fietkau authored
Subtract the difference in preamble duration (in usec) from the value returned by ieee80211_generic_frame_duration. Signed-off-by:
-
Felix Fietkau authored
During normal operation, minstrel was showing suspicious EWMA probabilities exceeding 100%. It looks like the tx status reporting in ath5k was not properly clearing the rate index for rates which were never attempted. This is caused by uninitialized stale data in the on-stack tx status information, which is reused when more frames are received. To fix this, rely on ts->ts_final_idx to select the last attempted rate, instead of checking whether ts->ts_rate is set. Additionally, the conversion from the driver rate index back to the mac80211 rate index can be dropped, as the mac80211 tx status will still have the original rate index which was used to set up the descriptor. Additionally, one more inaccuracy was fixed - the final rate attempt count only needs to be increased by one if the transmission attempt was successful. Signed-off-by:
-
John W. Linville authored
The ath5k driver does kmalloc allocations for pcal info in a loop. But, if one fails it was simply returning -ENOMEM without freeing already allocated memory. This patch corrects that oversight. Reported-by:
Eugene A. Shatokhin <dame_eugene@mail.ru> Signed-off-by:
Bob Copeland <me@bobcopeland.com>
-
Yogesh Ashok Powar authored
Fix checkpatch errors and warnings comprising of indent errors, spaces and __packed warnings. Also fix 'make C = 2' warnings. Signed-off-by:
Yogesh Ashok Powar <yogeshp@marvell.com> Signed-off-by:
-
Felix Fietkau authored
Older versions have not been sold and the driver does not explicitly check for them anyway, so we can simply ignore the macRev here. Reduces ath9k_hw size on mips by more than 2 KB. Signed-off-by:
-
Mohammed Shafi Shajakhan authored
Add station connected time in debugfs. This will be helpful to get a measure of stability of the connection and for debugging stress issues Cc: Senthilkumar Balasubramanian <Senthilkumar.Balasubramanian@Atheros.com> Signed-off-by:
Mohammed Shafi Shajakhan <mshajakhan@atheros.com> Signed-off-by:
-
Rajkumar Manoharan authored
The commit "ath9k: configure beacons based on hw opmode" introduced a regression which leads to kernel panic. Failed to stop ani timer during the driver unload while any of the beaconing vif is running. Signed-off-by:
Rajkumar Manoharan <rmanoharan@atheros.com> Signed-off-by:
-
Senthil Balasubramanian authored
Update Tx gain 23 for all tx gain table. Signed-off-by:
Senthil Balasubramanian <senthilkumar@atheros.com> Signed-off-by:
-
Senthil Balasubramanian authored
Export RSSI information from all the control and extension channel chains to debugfs. Also add rx antenna information to debugfs. This will be useful for debugging purpose. Signed-off-by:
-
Javier Cardona authored
Also, advertise support for mesh authentication. Signed-off-by:
Javier Cardona <javier@cozybit.com> Signed-off-by:
-
Javier Cardona authored
Notify userspace when a beacon/presp is received from a suitable mesh peer candidate for whom no sta information exists. Userspace can then decide to create a sta info for the candidate. If userspace is not ready to authenticate the peer right away, it can create the sta info with the authenticated flag unset and set it later. Signed-off-by:
-
Javier Cardona authored
Modify the NEW_STATION command to accept PLINK_ACTIONS, in case userspace wants to create stations and initiate a peer link right away (for authenticated stations) or create a blocked station (for debugging). Signed-off-by:
-
Javier Cardona authored
Signed-off-by:
-
Javier Cardona authored
Signed-off-by:
-
Javier Cardona authored
Signed-off-by:
-
Javier Cardona authored
Signed-off-by:
-
Javier Cardona authored
During mesh setup, use NL80211_MESH_SETUP_USERSPACE_AUTH flag to create a secure mesh and route management frames to userspace. Also, NL80211_CMD_GET_WIPHY now returns a flag NL80211_SUPPORT_MESH_AUTH if the wiphy's mesh implementation supports routing of mesh auth frames to userspace. This is useful for forward compatibility between old kernels and new userspace tools. Signed-off-by:
Thomas Pedersen <thomas@cozybit.com> Signed-off-by:
-
Javier Cardona authored
To NL80211_MESH_SETUP_IE. This reflects our ability to insert any ie into a mesh beacon, not simply path selection ies. Signed-off-by:
-
Paul Bolle authored
iwl4965_rate_control_register() prints a message at KERN_ERR level. It looks like it's just a debugging message, so pr_err() seems to be overdone. But none of the similar functions in drivers/net/wireless print a message, so let's just drop it. Signed-off-by:
Paul Bolle <pebolle@tiscali.nl> Signed-off-by:
-
Felix Fietkau authored
rx should only be enabled after enough rx buffers have been given to the hardware, however ath_rx_buf_link was calling ath9k_hw_rxena after every single added buffer. Fix this by calling ath9k_hw_rxena directly from the rx tasklet after completion instead. Signed-off-by:
-
Felix Fietkau authored
I can't find any valid reason for not setting the ATH9K_RX_FILTER_PSPOLL flag on older hardware and neither the documentation nor the reference code mention any reason for excluding older hardware here. Signed-off-by:
-
Yogesh Ashok Powar authored
Instead of having separate key information definitions for each type of key, a common key information bitmap is used. Signed-off-by:
Bing Zhao <bzhao@marvell.com> Signed-off-by:
-
Bing Zhao authored
Before calling host_to_card() to send the cmd to firmware, we use skb_push() to add 4 bytes SDIO interface header at the start of the data buffer. Since cmd_skb data structure will be re-used at a later time, we need to restore its headroom by removing the 4 bytes header. Signed-off-by:
Marc Yang <yangyang@marvell.com> Signed-off-by:
-
Rajkumar Manoharan authored
The problem is that when the attenuation is increased, the rate will start to drop from MCS7 -> MCS6, and finally will see MCS1 -> CCK_11Mbps. When the rate is changed b/w CCK and OFDM, it will use register desired_scale to calculate how much tx gain need to change. The output power with the same tx gain for CCK and OFDM modulated signals are different. This difference is constant for AR9280 but not AR9285/AR9271. It has different PA architecture a constant. So it should be calibrated against this PA characteristic. The driver has to read the calibrated values from EEPROM and set the tx power registers accordingly. Signed-off-by:
-
Vivek Natarajan authored
This function returns true if there is atleast one frame in any one of the tx queues. Signed-off-by:
Vivek Natarajan <vnatarajan@atheros.com> Signed-off-by:
-
Vivek Natarajan authored
In a highly noisy environment, the tx rate of the driver drops and the application slows down since it has not yet received ACKs for the frames already queued in the hardware. Since this ACK may take more than 100ms, stopping the dev queues for entering PS at this stage breaks applications, WMM test cases in my testing. If there are frames already pending in the tx queue, postponing the PS logic helps to avoid redundant queue stops. When power save is enabled by default and in a noisy environment, this API certainly helps in improving the average throughput. Signed-off-by:
-
- Apr 08, 2011
-
-
Johannes Berg authored
If we maintain API properly, then there isn't really a reason to warn about this since we'll just be adding things that are safe to ignore, so downgrade the warning to debug info level. Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
Wey-Yi Guy <wey-yi.w.guy@intel.com>
-
Johannes Berg authored
There's no need to keep both normal and BT statistics versions around all the time in memory when we only use a subset of both. So keep only the subsets that we need in memory, depending on the debug config). Also, in doing so, we can remove all the calls to iwl_bt_statistics() in the driver as we'll just access the copied statistics now. Finally, also remove this call from the one place where it might still be needed and automatically detect what kind of statistics the device is sending based on their size. This way, we don't need to keep track of which devices do what any more, which is good since this is subject to change based on the ucode version (as some ucode even for non-BT devices will in fact use BT statistics). Warn upon encountering a statistics command from the ucode that isn't known, so we will find such issues earlier in the future. Signed-off-by:
Don Fry <donald.h.fry@intel.com> Signed-off-by:
-
Wey-Yi Guy authored
_agn driver should perform both sensitivity and chain noise calib. Signed-off-by:
-
