- 15 Jun, 2015 21 commits
-
-
Arend van Spriel authored
Making it more clear by freeing the ifp in same place where the vif object is freed. Reviewed-by:
Hante Meuleman <meuleman@broadcom.com> Reviewed-by:
Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by:
Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by:
Arend van Spriel <arend@broadcom.com> Signed-off-by:
Kalle Valo <kvalo@codeaurora.org>
-
Arend van Spriel authored
The bus interface functions txctl and rxctl may be used while the device can not be accessed, eg. upon driver .remove() callback. This patch will immediately return -EIO when this is the case which speeds up the module unload. Reviewed-by:
-
Arend van Spriel authored
The .get_station() cfg80211 callback is used in several scenarios. In managed mode it can obtain information about the access-point and its BSS parameters. In managed mode it can also obtain information about TDLS peers. In AP mode it can obtain information about connected clients. Reviewed-by:
Daniel (Deognyoun) Kim <dekim@broadcom.com> Reviewed-by:
-
Pontus Fuchs authored
Switch from using semi hard coded interface combinations. This makes it easier to announce what the firmware actually supports. This fixes the case where brcmfmac announces p2p but the firmware doesn't support it. Reviewed-by:
Pontus Fuchs <pontusf@broadcom.com> Signed-off-by:
-
Pontus Fuchs authored
Add a feature flag to reflect the firmware's p2p capability. Reviewed-by:
-
Jakub Kicinski authored
We expect EEPROM per-rate power table to be filled with s6 values and warn user if values are invalid. However, there appear to be devices which don't have this section of EEPROM initialized. In such case we should ignore the values and leave the driver power tables set to zero. Note that vendor driver doesn't care about this case but mt76x2 skips 0xff per value. We take mt76x2's approach. Signed-off-by:
Jakub Kicinski <kubakici@wp.pl> Signed-off-by:
-
Stanislaw Gruszka authored
rt2x00.serialmonkey.com will be shutdown. Since traffic on rt2x00 mailing list is very low, we can use only linux-wireless list for any rt2x00 related topics. Thanks for Luis Correia, Ivo van Doorn and Mark Wallis for maintaining rt2x00 servers for years! Signed-off-by:
Stanislaw Gruszka <sgruszka@redhat.com> Signed-off-by:
-
Chunfan Chen authored
If timeshare coexistance between bluetooth and WLAN gets enabled, firmware will give host an event to reduce Rx AMPDU BA window size. The event is handled in this patch. Signed-off-by:
Chunfan Chen <jeffc@marvell.com> Signed-off-by:
Cathy Luo <cluo@marvell.com> Signed-off-by:
Amitkumar Karwar <akarwar@marvell.com> Signed-off-by:
-
Chun-Yeow Yeoh authored
Add the support of channel switching functionality, similar to ath9k support. Tested with TP-Link TL-WN722N and TL-WN821N. Signed-off-by:
Chun-Yeow Yeoh <yeohchunyeow@gmail.com> Signed-off-by:
-
Arend van Spriel authored
Some time ago the function debugfs_create_devm_seqfile() was introduced in debugfs. The caller simply needs to provide a device pointer and read function. The function brcmf_debugfs_add_entry() is now simply a wrapper only doing the work for CONFIG_BRCMDBG. Reviewed-by:
-
Arend van Spriel authored
The watchdog reset as done in brcmf_pcie_buscoreprep() is not sufficient. It needs to modify PCIe core registers as well which is properly done by brcmf_pcie_reset_device() after the chip recognition is done. So the faulty watchdog reset can be removed as it was causing driver reload to fail and hang the system requiring a power-cycle. Instead the call to to the brcmf_pcie_reset_device() function is done twice in the unload. Reviewed-by:
-
Arend van Spriel authored
The information provided by chipinfo is also provided by the revinfo debugfs entry. Removing it from debugfs. Reviewed-by:
-
Hante Meuleman authored
On device to host data using msgbuf the read pointer gets updated once all data is processed. Updating this pointer more frequently allows the firmware to add more data quicker. This will result in slightly higher and more stable throughput on CPU bounded host processors. Reviewed-by:
-
Taehee Yoo authored
The variable "fw_version" is used in the _ResetDigitalProcedure1(). but It is not initialized. so I add init codes for "fw_version" and "fw_subversion". Signed-off-by:
Taehee Yoo <ap420073@gmail.com> Acked-by:
Larry Finger <Larry.Finger@lwfinger.net> Signed-off-by:
-
Taehee Yoo authored
rtl92cu_rx_query_desc set a isampdu twice. but second code is related to isfirst_ampdu. so i change it. Signed-off-by:
-
Taehee Yoo authored
In the rtlwifi/rtl8192cu, I change debug message "RTL8192CE" to "RTL8192CU". Signed-off-by:
-
Taehee Yoo authored
in the _rtl92c_phy_rf6052_config_parafile(), cases RF90_PATH_A and RF90_PATH_B call the same routine. so i remove one of these routine. also the return routine is duplicated. so i remove it. Signed-off-by:
-
Hans Ulli Kroll authored
While working on getting my rtl8821au driver in pretty shape for inclusion, it is dicosvered that the tm_trigger flag is used only for the first device using this driver. This flag handles the thermal power management in the hardware. To change this add a entry in sttruct rtl_dm, so each device can handle is separately. Signed-off-by:
Hans Ulli Kroll <ulli.kroll@googlemail.com> Signed-off-by:
-
Taehee Yoo authored
in the rtlwifi/rtl8192cu, INTF_PCI and INTF_USB is unnecessary. because RTL8192CU chipset is only USB interface. Signed-off-by:
-
Taehee Yoo authored
_InitBeaconParameters() and rtl92cu_init_beacon_parameters() is same routine. I remove both functions. then i add _rtl92cu_init_beacon_parameters() in the hw.c. _rtl92cu_init_beacon_parameters() is same routine with _InitBeaconParameters(). Signed-off-by:
-
Taehee Yoo authored
in the rtlwifi/rtl8192cu, IS_HARDWARE_TYPE_8192CE and IS_HARDWARE_TYPE_8192CU is unnecessary. because rtlwifi/rtl8192cu codes aren't shared. Signed-off-by:
-
- 08 Jun, 2015 19 commits
-
-
Rafał Miłecki authored
This extension of BCMA_DRIVER_PCI has no reason to depend on BCMA_HOST_PCI. User may just want to have PCI device attached to SoC registered without enabling any extra client mode code. This can be useful when having non-bcma PCI device attached or when using other PCI driver. Signed-off-by:
Rafał Miłecki <zajec5@gmail.com> Signed-off-by:
-
Rafał Miłecki authored
Signed-off-by:
-
Rafał Miłecki authored
It seems Broadcom released two devices with conflicting device id. There are for sure 14e4:4321 PCI devices with BCM4321 (N-PHY) chipset, they can be found in routers, e.g. Netgear WNR834Bv2. However, according to Broadcom public sources 0x4321 is also used for 5 GHz BCM4306 (G-PHY). It's unsure if they meant PCI device id, or "virtual" id (from SPROM). To distinguish these devices lets check PHY type (G vs. N). Signed-off-by:
-
Rafał Miłecki authored
As we plan to add support for platform NVRAM we should store direct data pointer without the extra struct firmware layer. This will allow us to support other sources with the only requirement being u8 buffer. Signed-off-by:
-
Jakub Kicinski authored
Most drivers use FIF_OTHER_BSS to set promiscous mode. Let us follow their lead even though it doesn't match exactly the HW filter flags. Signed-off-by:
-
Jakub Kicinski authored
Make sure .disconnect() doesn't cleanup the device if .resume() failed. This may happen when device is removed during suspend. Signed-off-by:
-
Jakub Kicinski authored
Users of older Ralink devices report that received frames sometimes have zero length. Watch out for that. Signed-off-by:
-
Jakub Kicinski authored
Signed-off-by:
-
Rafał Miłecki authored
This allows e.g. user space to use /sys/class/ieee80211/*/macaddress Signed-off-by:
-
Rafał Miłecki authored
Recently Broadcom added support for NVRAMs with entries for multiple PCIe devices. One of the supported formats is based on prefixes defined like: devpath0=pcie/1/4/ and entries like 0:foo=bar 0:baz=qux etc. Unfortunately there are also a bit older devices using different way of defining prefixes, e.g. SmartRG SR400ac (2 x BCM43602) with entries: devpath0=pci/1/1/ devpath1=pci/2/1 Broadcom stated this old format will never be used/supported by brcmfmac but given the simplicity of this patch I'll insist on supporting it. Signed-off-by:
-
Avinash Patil authored
It was observed that AP beacons would not reflect correct regulatory information upon starting AP in A band. This was because of missing AP config band update in set_channel of start_ap. Signed-off-by:
Avinash Patil <patila@marvell.com> Signed-off-by:
-
Zhaoyang Liu authored
This patch fixes semantic warning for debugging data dump feature. Previous code is based on the assumption that skb is not null. New change makes sure that we already have data buffer. Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
Zhaoyang Liu <liuzy@marvell.com> Signed-off-by:
-
Taehee Yoo authored
After physical reconnect, the rtl8192cu chipset shows low transmission rates.It cause is that variable "iqk_initialized" do not de-initialized. So I add this code. Signed-off-by:
-
Xinming Hu authored
Negative adapter->tx_pending is observed while running data traffic, because tx_pending is decreased once more for AMSDU packet. since tx_pending have been decreased for all the source MSDU packets, we don't need to update once more for AMSDU packet. Signed-off-by:
Xinming Hu <huxm@marvell.com> Signed-off-by:
-
Xinming Hu authored
This patch fixes issue with the accessing correct ra_list by downgrading corresponding tid number. Alternatively, ra lists are created in mwifiex_wmm_add_buf_txqueue using downgraded tid number. Signed-off-by:
-
Avinash Patil authored
This was missing and would cause issue in WMM handling. Signed-off-by:
-
Avinash Patil authored
This would enable driver to enter powersave as soon as connected. Signed-off-by:
-
Avinash Patil authored
We often see ADDBA request packets coming to driver because driver has registered for action frame subtype. We dont process BA action frames in host; drop such frames. Signed-off-by:
-
Avinash Patil authored
Earlier only RSN, WPA and channel switch IEs from tail buffer would be downloaded to FW. This patch adds support for downloading more IEs from tail buffer. Signed-off-by:
-
