- 15 Aug, 2013 40 commits
-
-
John W. Linville authored
The driver fails to check the results of DMA mapping in twp places, which results in the following warning: [ 28.078515] ------------[ cut here ]------------ [ 28.078529] WARNING: at lib/dma-debug.c:937 check_unmap+0x47e/0x930() [ 28.078533] bcma-pci-bridge 0000:0e:00.0: DMA-API: device driver failed to check map error[device address=0x00000000b5d60d6c] [size=1876 bytes] [mapped as single] [ 28.078536] Modules linked in: bnep bluetooth vboxpci(O) vboxnetadp(O) vboxnetflt(O) vboxdrv(O) ipv6 b43 brcmsmac rtl8192cu rtl8192c_common rtlwifi mac802 11 brcmutil cfg80211 snd_hda_codec_conexant rng_core snd_hda_intel kvm_amd snd_hda_codec ssb kvm mmc_core snd_pcm snd_seq snd_timer snd_seq_device snd k8temp cordic joydev serio_raw hwmon sr_mod sg pcmcia pcmcia_core soundcore cdrom i2c_nforce2 i2c_core forcedeth bcma snd_page_alloc autofs4 ext4 jbd2 mbcache crc1 6 scsi_dh_alua scsi_dh_hp_sw scsi_dh_rdac scsi_dh_emc scsi_dh ata_generic pata_amd [ 28.078602] CPU: 1 PID: 2570 Comm: NetworkManager Tainted: G O 3.10.0-rc7-wl+ #42 [ 28.078605] Hardware name: Hewlett-Packard HP Pavilion dv2700 Notebook PC/30D6, BIOS F.27 11/27/2008 [ 28.078607] 0000000000000009 ffff8800bbb03ad8 ffffffff8144f898 ffff8800bbb03b18 [ 28.078612] ffffffff8103e1eb 0000000000000002 ffff8800b719f480 ffff8800b7b9c010 [ 28.078617] ffffffff824204c0 ffffffff81754d57 0000000000000754 ffff8800bbb03b78 [ 28.078622] Call Trace: [ 28.078624] <IRQ> [<ffffffff8144f898>] dump_stack+0x19/0x1b [ 28.078634] [<ffffffff8103e1eb>] warn_slowpath_common+0x6b/0xa0 [ 28.078638] [<ffffffff8103e2c1>] warn_slowpath_fmt+0x41/0x50 [ 28.078650] [<ffffffff8122d7ae>] check_unmap+0x47e/0x930 [ 28.078655] [<ffffffff8122de4c>] debug_dma_unmap_page+0x5c/0x70 [ 28.078679] [<ffffffffa04a808c>] dma64_getnextrxp+0x10c/0x190 [brcmsmac] [ 28.078691] [<ffffffffa04a9042>] dma_rx+0x62/0x240 [brcmsmac] [ 28.078707] [<ffffffffa0479101>] brcms_c_dpc+0x211/0x9d0 [brcmsmac] [ 28.078717] [<ffffffffa046d927>] ? brcms_dpc+0x27/0xf0 [brcmsmac] [ 28.078731] [<ffffffffa046d947>] brcms_dpc+0x47/0xf0 [brcmsmac] [ 28.078736] [<ffffffff81047dcc>] tasklet_action+0x6c/0xf0 --snip-- [ 28.078974] [<ffffffff813891bd>] SyS_sendmsg+0xd/0x20 [ 28.078979] [<ffffffff81455c24>] tracesys+0xdd/0xe2 [ 28.078982] ---[ end trace 6164d1a08148e9c8 ]--- [ 28.078984] Mapped at: [ 28.078985] [<ffffffff8122c8fd>] debug_dma_map_page+0x9d/0x150 [ 28.078989] [<ffffffffa04a9322>] dma_rxfill+0x102/0x3d0 [brcmsmac] [ 28.079001] [<ffffffffa047a13d>] brcms_c_init+0x87d/0x1100 [brcmsmac] [ 28.079010] [<ffffffffa046d851>] brcms_init+0x21/0x30 [brcmsmac] [ 28.079018] [<ffffffffa04786e0>] brcms_c_up+0x150/0x430 [brcmsmac] As the patch adds a new failure mechanism to dma_rxfill(). When I changed the comment at the start of the routine to add that information, I also polished the wording. Signed-off-by:
Larry Finger <Larry.Finger@lwfinger.net> Cc: Stable <stable@vger.kernel.org> Cc: Brett Rudley <brudley@broadcom.com> Cc: Franky (Zhenhui) Lin <frankyl@broadcom.com> Cc: Hante Meuleman <meuleman@broadcom.com> Cc: brcm80211-dev-list@broadcom.com Acked-by:
Arend van Spriel <arend@broadcom.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Mark Schulte authored
Adding type casts to suppress sparse warnings: * warning: cast to restricted __le32/__le16 Signed-off-by:
Mark Schulte <schultetwin@gmail.com> Acked-by:
-
Sujith Manoharan authored
Since the rx_fifo queue is accessed only using the various lockless SKB queue routines, there is no need to initialize the lock and __skb_queue_head_init() can be used. Signed-off-by:
Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by:
-
Sujith Manoharan authored
The documentation for antenna diversity says: "The decision of diversity is done at 802.11 preamble. So, for 11G/11B, for every MAC packet hardware will do a decision. But in 11N with aggregation, the decision is made only at the preamble and all other MPDUs will use the same LNA as the first MPDU." Make use of rs_firstaggr to avoid needlessly scanning for LNA changes. Signed-off-by:
-
Sujith Manoharan authored
Signed-off-by:
-
Sujith Manoharan authored
The combined RSSI can be invalid which is indicated by the value -128. Use RX_FLAG_NO_SIGNAL_VAL for such cases. Signed-off-by:
-
Sujith Manoharan authored
Signed-off-by:
-
Sujith Manoharan authored
In case a descriptor has the "done" bit clear and the next descriptor has it set, we drop both of them. If the packet that is received after these two packets is dropped for some reason, "discard_next" will not cleared. Fix this. Signed-off-by:
-
Sujith Manoharan authored
The MIC/PHYERR/CRC error bits are valid only for the last desc. for chained packets. Check this early in the preprocess() routine and bail out. Signed-off-by:
-
Sujith Manoharan authored
Signed-off-by:
-
Sujith Manoharan authored
Make sure that chained descriptors are handled correctly before the packet is parsed to determine if it is a beacon. Signed-off-by:
-
Sujith Manoharan authored
Handle chained descriptors and increment the RX counter only for valid packets. Since this is used only by MCI, use CONFIG_ATH9K_BTCOEX_SUPPORT. Signed-off-by:
-
Sujith Manoharan authored
The various error bits that ath_debug_stat_rx() checks are valid only for the last descriptor for a chained packet, handle this correctly. Signed-off-by:
-
Sujith Manoharan authored
Parse the PHY error details only for the last fragment in case descriptors are chained. Signed-off-by:
-
Sujith Manoharan authored
Signed-off-by:
-
Sujith Manoharan authored
There is no need to calculate the mactime for chained descriptor packets, so make sure that this is done only for the last fragment of valid packets. Signed-off-by:
-
Sujith Manoharan authored
The keymiss events are valid only in the last descriptor of a packet. Fix this by making sure that we return early in case of chained descriptors. Signed-off-by:
-
Sujith Manoharan authored
Frames with invalid or zero length can be discarded early, there is no need to check the crypto bits. Signed-off-by:
-
Sujith Manoharan authored
Since the DFS code appears to process the phy errors ATH9K_PHYERR_RADAR and ATH9K_PHYERR_FALSE_RADAR_EXT, check for the correct phyerr status in the main RX tasklet routine. Signed-off-by:
-
Sujith Manoharan authored
Signed-off-by:
-
Oleksij Rempel authored
If usb auto suspend is enabled or system run in to suspend/resume cycle, ath9k-htc adapter will stop to response. It is reproducible on xhci HCs. Host part of problem: XHCI do timing calculation based on Transfer Type and bInterval, immediately after device was detected. Ath9k-htc try to overwrite this parameters on module probe and some changes in FW, since we do not initiate usb reset from the driver this changes are not took to account. So, before any kind of suspend or reset, host controller will operate with old parameters. Only after suspend/resume and if interface id stay unchanged, new parameters will by applied. Host will send bulk data with no intervals (?), which will cause overflow on FIFO of EP4. Firmware part of problem: By default, ath9k-htc adapters configured with EP3 and EP4 as interrupt endpoints. Current firmware will try to overwrite ConfigDescriptor to make EP3 and EP4 bulk. FIFO for this endpoints stay not reconfigured, so under the hood it is still Int EP. This patch is revert of 4a0e8ecc commit which trying to reduce CPU usage on some systems. Since it will produce more bug as fixes, we will need to find other way to fix it. here is comment from kernel source which has some more explanation: * Some buggy high speed devices have bulk endpoints using * maxpacket sizes other than 512. High speed HCDs may not * be able to handle that particular bug, so let's warn... in our case EP3 and EP4 have maxpacket sizes = 64!!! Signed-off-by:
Oleksij Rempel <linux@rempel-privat.de> Signed-off-by:
-
Vladimir Kondratiev authored
Fix for TCP iperf from Windows to Linux stall after about 1sec Hardware reports false errors in some situations: Microsoft IP stack, in violation of RFC 1624, set TCP checksum that should be 0x0 as 0xffff. hardware report Rx csum error. If HW csum absolutely trusted, this frame can be never received, as re-transmitted one will have same csum problem. In addition, it mess up block ack reorder buffer, as if packet dropped, it is not score boarded there. Signed-off-by:
Vladimir Kondratiev <qca_vkondrat@qca.qualcomm.com> Signed-off-by:
-
Felix Fietkau authored
If that flag stays set for a buffer that already ran through the tx path once, it might cause issues in tx completion processing. Better clear it early to ensure that this does not happen Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Signed-off-by:
-
Felix Fietkau authored
ath_tid_drain is only called when a station entry is being removed, so there is no point in still tracking BAW state. Remove some unnecessary code and a bogus TODO comment related to this. Signed-off-by:
-
Felix Fietkau authored
They are not implemented, and accessing them might trigger errors Cc: stable@vger.kernel.org Signed-off-by:
-
Felix Fietkau authored
Writing to that file is unnecessary and quirky, the antenna API should be used instead. Use debugfs_create_u8 to allow reading the values. Signed-off-by:
-
Felix Fietkau authored
mac80211 uses debugfs_remove_recursive, so there's no need for the driver to do an explicit cleanup of its sta debugfs entry. Signed-off-by:
-
Felix Fietkau authored
Also reduce the size of a few fields where possible Signed-off-by:
-
Felix Fietkau authored
Similar to a race condition that exists in the tx path, the hardware might re-read the 'next' pointer of a descriptor of the last completed frame. This only affects non-EDMA (pre-AR93xx) devices. To deal with this race, defer clearing and re-linking a completed rx descriptor until the next one has been processed. Cc: stable@vger.kernel.org Signed-off-by:
-
Franky Lin authored
Host tx glomming require an extended hardware sdio bus header to store information for dongle. Introduce a variable in struct brcmf_sdio to replace macro SDPCM_HDRLEN Reviewed-by:
Hante Meuleman <meuleman@broadcom.com> Reviewed-by:
Pieter-Paul Giesberts <pieterpg@broadcom.com> Reviewed-by:
Franky Lin <frankyl@broadcom.com> Signed-off-by:
-
Franky Lin authored
Streamlining sdio bus specific header related code as preparation for host tx glomming Reviewed-by:
-
Franky Lin authored
remove align from brcmf_bus since it is only used by sdio bus layer internally Reviewed-by:
-
Franky Lin authored
Abstract brcmf_sdio_txpkt_prep and brcmf_sdio_txpkt_postp as a preparation of chained tx packets for host side tx glomming. Reviewed-by:
-
Arend van Spriel authored
Implement the .tdls_oper() callback and indicate TDLS support in the wiphy flags. Reviewed-by:
-
Arend van Spriel authored
Not every IF event from the firmware needs to result in a related interface, netdev or wdev, on the host. This is indicated in the event message. Handle that flag and effectively ignore the firmware event. Reviewed-by:
-
Hante Meuleman authored
FWS uses locking to protect its data while being called from various entries. On bus_txdata the lock was kept resulting in unnecessary long locking, but also creating possibility for deadlock. This update changes the locking to release lock when bus_txdata is called. Reviewed-by:
-
Hante Meuleman authored
When fw signalling is disabled tx is sent immediately. Using queues and worker thread allows usb to do synchronous autopm. This patch makes fws use queues and worker thread even if signalling is not supported by FW or not enabled. Reviewed-by:
-
Arend van Spriel authored
This feature moves the responsibility of collecting all MPDUs in an AMPDU session in the correct order from the firmware to the host driver. This reduces buffer requirement on the firmware side. Reviewed-by:
-
Arend van Spriel authored
With firmware-signalling the packet handed to the bus specific driver layer should not be discarded with brcmf_txcomplete() in the failure path. Instead only an error is returned and the caller decides what to do with the packet. Reviewed-by:
-
Arend van Spriel authored
Firmware-signalling needs transmit to firmware to be atomic and uses a spinlock with irq disabled. Therefor, brcmf_sdbrcm_txdata() should not use spin_unlock_bh() as it would enable the interrupts. Reviewed-by:
-
