- 07 Jan, 2013 37 commits
-
-
Alexey Khoroshilov authored
If pci_set_dma_mask() or pci_set_consistent_dma_mask() fails in p54p_probe(), it breaks off initialization, deallocates all resources, but returns zero. Similar issue is if check for returned value of pci_resource_len() fails. The patch implements proper error code propagation. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by:
Alexey Khoroshilov <khoroshilov@ispras.ru> Acked-by:
Christian Lamparter <chunkeey@googlemail.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Hante Meuleman authored
On interface up dongle gets inialized. Move UP command to common routine and update these common init routines using ifp. Reviewed-by:
Arend Van Spriel <arend@broadcom.com> Reviewed-by:
Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by:
Hante Meuleman <meuleman@broadcom.com> Signed-off-by:
-
Hante Meuleman authored
Reviewed-by:
-
Hante Meuleman authored
Netdev xmit routine brcfm_netdev_start_xmit was defined incorrectly and used wrong return codes. Always eat the packet and return ok. Remove drvr_up check since it is not relevant. Reviewed-by:
-
Hante Meuleman authored
CDC errors are retained. However, it is never used so it can be removed. Reviewed-by:
-
Arend van Spriel authored
Stumbled into a curly braces used for if statement with only one conditional statement. Removing them. Reviewed-by:
-
Arend van Spriel authored
The header of the ethernet packet is processed conditionally, but the check is wrong as it checks skb length is at least ETH_ALEN. It should check it is at least sizeof struct ethhdr instead. Reviewed-by:
-
Arend van Spriel authored
The driver had some global definitions in dhd.h to map event identifiers to event names. With redesign of firmware event processing this has all moved to fweh module so these definitions can be removed. Reviewed-by:
-
Hante Meuleman authored
CDC debug is partly done with TRACE and partly with CTL, however CDC hardly ever needs debugging. Use dedicated level CDC and replace TRACE and CTL in dhd_cdc with that. Reviewed-by:
-
Arend van Spriel authored
The use of the function brcmf_proto_hdrpull() is moved to the common part of the driver and consequently it can be removed from the bus interface. Reviewed-by:
-
Arend van Spriel authored
The bus interface provided a wrapper function to pass a single packet to the common driver part filling a skb queue with one packet. For clarity the caller now sets up the skb queue and call the rx bus interface function. Reviewed-by:
-
Arend van Spriel authored
Several source files (but not all) define the pr_fmt() macro in exactly the same way. Instead this commit defines it in a header file so driver logging is consistent. Cc: Joe Perches <joe@perches.com> Reviewed-by:
-
Arend van Spriel authored
Upon unloading the brcmfmac module the USB firmware should be reset as the device remains powered. The reset assures a known device state when a new brcmfmac driver load is being done. Reviewed-by:
-
Hante Meuleman authored
Set bitrate_mask is not desired anymore. The firmware will determine the correct rates to be used. Reviewed-by:
-
Hante Meuleman authored
When USB device gets removed rx complete comes with 0-length packets. Do not refill those packets. In some rare cases it can cause infinite loop. Reviewed-by:
-
Piotr Haber authored
Dead code after AMPDU restructure. Reviewed-by:
Piotr Haber <phaber@broadcom.com> Signed-off-by:
-
Arend van Spriel authored
The interface address of the wireless device is determined by the permanent address stored in the device. This patch allows it to be overridden from user-space. Reported-by:
Maximilian Engelhardt <maxi@daemonizer.de> Reviewed-by:
-
Arend van Spriel authored
All callbacks that access driver functions should do that under perimeter lock protection. The add_interface() callback was lacking this lock. Reviewed-by:
-
Piotr Haber authored
On hardware reintialization reference count of already existing timers would be increased again. This leads to problems on module unloading. Reviewed-by:
-
Gabor Juhos authored
Both the rtt2x00usb_eeprom_read and the ioremap functions are allowed to fail, however their return values are not checked in the read_eeprom functions in the rt2800{pci,usb} drivers. The patch adds the missing checks, and converts all read_eeprom functions to return an int value, so the error values can be propagated up to the 'rt2800_validate_eeprom' function. Signed-off-by:Gabor Juhos <juhosg@openwrt.org> Signed-off-by:
-
Sujith Manoharan authored
Signed-off-by:
Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by:
-
Geert Uytterhoeven authored
drivers/net/wireless/brcm80211/brcmsmac/main.c: In function ‘brcms_b_recv’: drivers/net/wireless/brcm80211/brcmsmac/main.c:7636: warning: ‘morepending’ may be used uninitialized in this function Signed-off-by:
Geert Uytterhoeven <geert@linux-m68k.org> Acked-by:
-
Stanislaw Gruszka authored
Power save support was removed from iwlegacy due to possible firmware crashes problems it cause. I use to plan first inspect code to find reason of problems, fix them and then allow to enable PS. But realistically - code inspection will not happen, so let's do it, and wait for eventual bug reports. Signed-off-by:
Stanislaw Gruszka <sgruszka@redhat.com> Signed-off-by:
-
Stanislaw Gruszka authored
Dump implementation of flush, which just wait until all TX queues become empty. Signed-off-by:
-
Gabor Juhos authored
Some 3-column initval arrays have wrong comments. The column of these arrays is indexed by the 'freqIndex' variable in 'ar5008_hw_process_ini' which only depends on the actual band. The 'initvals' tool from 'qca-swiss-army-knife' prints the correct comment lines for these arrays, since commit 'atheros-initvals: fix comments for non-fastclock 3-column tables' however the comments were not refreshed in ath9k. The patch contains no functional changes. Signed-off-by:
-
Christian Lamparter authored
This patch moves parts of carl9170_update_beacon into separate subroutines, so the parts become more manageable. Signed-off-by:
-
Christian Lamparter authored
The commit: "mac80211: introduce IEEE80211_NUM_TIDS and use it" introduced a generic NUM_TID definitions for all everyone. Signed-off-by:
-
Christian Lamparter authored
While the driver supports HW offload in a single P2P client configuration, it doesn't support HW offload in the concurrent P2P GO+CLIENT configuration. Signed-off-by:
-
Christian Lamparter authored
Previously, op_start would set disable_offload always to false, even if it was set to true by the fw parser. Signed-off-by:
-
Christian Lamparter authored
Otherwise carl9170 triggers a warning in cfg80211, from net/wireless/core.c /* Combinations with just one interface aren't real */ if (WARN_ON(c->max_interfaces < 2)) Note: The number of supported interfaces is set by the carl9170 firmware. The default number of supported interfaces for all current firmwares is 2. Therefore this warning can only be observed with custom firmwares. Signed-off-by:
-
Christian Lamparter authored
Janusz Dziedzic reported that after a change in wpa_supplicant ["nl80211: Automatically use concurrent P2P if possible"], carl9170 was no longer able to host a P2P network. This patch tackles the problem by allowing GO interfaces to be registered, long after the P2P_CLIENT interface is brought up. Reported-by:
Janusz Dziedzic <janusz.dziedzic@gmail.com> Signed-off-by:
-
Christian Lamparter authored
Import new headers from my firmware branch: <https://github.com/chunkeey/carl9170fw> visit our wiki at: <http://wireless.kernel.org/en/users/Drivers/carl9170.fw> Signed-off-by:
-
Christian Lamparter authored
Based on a quick test [ath9k and carl9170], TDLS seemed to be working fine. And while we are at it, let's move the wiphy feature flag set from carl9170_alloc into a single place in carl9170_fw. Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Acked-by:
Luis R. Rodriguez <mcgrof@qca.qualcomm.com> Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
-
Stanislaw Gruszka authored
rt2800_init_rfcsr() is too big, split RF initialization into per chip functions. Code will change, but we should setup the same values onto RF registers and in the same order as before. Signed-off-by:
Stanislaw Gruszka <stf_xl@wp.pl> Signed-off-by:
-
-
- 03 Jan, 2013 3 commits
-
-
Emmanuel Grumbach authored
This allows to let sparse check that the NIC access is always released. Signed-off-by:
Emmanuel Grumbach <emmanuel.grumbach@intel.com> Signed-off-by:
Johannes Berg <johannes.berg@intel.com>
-
Eytan Lifshitz authored
Fix typo in the macro name and the wrong value. Signed-off-by:
Eytan Lifshitz <eytan.lifshitz@intel.com> Signed-off-by:
-
Johannes Berg authored
Synchronizing the IRQ is pointless when we will then enable the RF-Kill interrupt again, but is needed before we free it and the data needed to handle IRQs; move it to the free function. Simiarly, cancelling the replenish work struct can move to the function that frees the RX data structures. Reviewed-by:
-
