summaryrefslogtreecommitdiffstats
path: root/drivers/net
Commit message (Collapse)AuthorAgeFilesLines
* bnx2x: Add missing afex codeYuval Mintz2012-09-132-11/+21
| | | | | | | | | | | | | | | Commit a334872224a67b614dc888460377862621f3dac7 added afex support but lacked several logical changes. This lack can cause afex to crash, and also have a slight effect on other flows (i.e., driver always assumes the Tx ring has less available buffers than what it actually has). This patch adds the missing segments, fixing said issues. Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Barak Witkowski <barak@broadcom.com> Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* bnx2x: fix registers dumpedDmitry Kravkov2012-09-132-18/+9
| | | | | | | | | | | Under traffic, there are several registers that when read (e.g., via 'ethtool -d') may cause the chip to stall. This patch corrects the registers read in such flows. Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com> Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* bnx2x: correct advertisement of pause capabilitiesYaniv Rosner2012-09-131-0/+6
| | | | | | | | | | | This patch propagates users' requested flow-control into the link layer, which will later be used to advertise this flow-control for auto-negotiation (until now these values were ignored). Signed-off-by: Yaniv Rosner <yaniv.rosner@broadcom.com> Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* bnx2x: display the correct duplex valueYaniv Rosner2012-09-131-4/+6
| | | | | | | | | | Prior to this fix, the driver reported the chip's active duplex state is always 'full', even if using half-duplex mode. Signed-off-by: Yaniv Rosner <yaniv.rosner@broadcom.com> Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* bnx2x: prevent timeouts when using PFCYaniv Rosner2012-09-131-3/+5
| | | | | | | | | | | Prevent updating the xmac PFC configuration when using a link speed slower than 10G -the umac block is responsible for 1G or slower connections, therefore it is possible the xmac block is reset when connection is slower. Signed-off-by: Yaniv Rosner <yaniv.rosner@broadcom.com> Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* bnx2x: fix stats copying logicYuval Mintz2012-09-131-2/+5
| | | | | | | | | | | | FW needs the driver statistics for management. Current logic is broken in that the function that gathers the port statistics does not copy its own statistics to a place where the FW can use it. This patch causes every function that can pass statistics to the FW to do so. Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* bnx2x: Avoid sending multiple statistics queriesDmitry Kravkov2012-09-131-3/+5
| | | | | | | | | | | | | During traffic when DCB is enabled, it is possible for multiple instances of statistics queries to be sent to the chip - this may cause the FW to assert. This patch prevents the sending of an additional instance of statistics query while the previous query hasn't completed. Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com> Signed-off-by: Yuval Mintz <yuvalmin@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: qmi_wwan: call subdriver with control intf onlyBjørn Mork2012-09-131-3/+4
| | | | | | | | | | | | | | | | | | | | | | | This fixes a hang on suspend due to calling wdm_suspend on the unregistered data interface. The hang should have been a NULL pointer reference had it not been for a logic error in the cdc_wdm code. commit 230718bd net: qmi_wwan: bind to both control and data interface changed qmi_wwan to use cdc_wdm as a subdriver for devices with a two-interface QMI/wwan function. The commit failed to update qmi_wwan_suspend and qmi_wwan_resume, which were written to handle either a single combined interface function, or no subdriver at all. The result was that we called into the subdriver both when the control interface was suspended and when the data interface was suspended. Calling the subdriver suspend function with an unregistered interface is not supported and will make the subdriver bug out. Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
* net: qmi_wwan: fix Gobi device probing for un2430Pierre Sauter2012-09-111-1/+1
| | | | | | | | | | | HP un2430 is a Gobi 3000 device. It was mistakenly treated as Gobi 1000 in patch b9f90eb2740203ff2592efe640409ad48335d1c2. I own this device and qmi_wwan works again with this fix. Signed-off-by: Pierre Sauter <pierre.sauter@gmail.com> Acked-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
* ixp4xx_hss: fix build failure due to missing linux/module.h inclusionFlorian Fainelli2012-09-101-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | Commit 36a1211970193ce215de50ed1e4e1272bc814df1 (netprio_cgroup.h: dont include module.h from other includes) made the following build error on ixp4xx_hss pop up: CC [M] drivers/net/wan/ixp4xx_hss.o drivers/net/wan/ixp4xx_hss.c:1412:20: error: expected ';', ',' or ')' before string constant drivers/net/wan/ixp4xx_hss.c:1413:25: error: expected ';', ',' or ')' before string constant drivers/net/wan/ixp4xx_hss.c:1414:21: error: expected ';', ',' or ')' before string constant drivers/net/wan/ixp4xx_hss.c:1415:19: error: expected ';', ',' or ')' before string constant make[8]: *** [drivers/net/wan/ixp4xx_hss.o] Error 1 This was previously hidden because ixp4xx_hss includes linux/hdlc.h which includes linux/netdevice.h which includes linux/netprio_cgroup.h which used to include linux/module.h. The real issue was actually present since the initial commit that added this driver since it uses macros from linux/module.h without including this file. Signed-off-by: Florian Fainelli <florian@openwrt.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* sierra_net: Endianess bug fix.Lennart Sorensen2012-09-101-1/+1
| | | | | | | | | I discovered I couldn't get sierra_net to work on a powerpc. Turns out the firmware attribute check assumes the system is little endian and hence fails because the attributes is a 16 bit value. Signed-off-by: Len Sorensen <lsorense@csclub.uwaterloo.ca> Signed-off-by: David S. Miller <davem@davemloft.net>
* Merge branch 'master' of ↵John W. Linville2012-09-0724-53/+260
|\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless into for-davem John W. Linville says: ==================== Please pull these fixes intended for 3.6. There are more commits here than I would like -- I got a bit behind while I was stalking Steven Rostedt in San Diego last week... I'll slow it down after this! There are a couple of pulls here. One is from Johannes: "Please pull (according to the below information) to get a few fixes. * a fix to properly disconnect in the driver when authentication or association fails * a fix to prevent invalid information about mesh paths being reported to userspace * a memory leak fix in an nl80211 error path" The other comes via Gustavo: "A few updates for the 3.6 kernel. There are two btusb patches to add more supported devices through the new USB_VENDOR_AND_INTEFACE_INFO() macro and another one that add a new device id for a Sony Vaio laptop, one fix for a user-after-free and, finally, two patches from Vinicius to fix a issue in SMP pairing." Along with those... Arend van Spriel provides a fix for a use-after-free bug in brcmfmac. Daniel Drake avoids a hang by not trying to touch the libertas hardware duing suspend if it is already powered-down. Felix Fietkau provides a batch of ath9k fixes that adress some potential problems with power settings, as well as a fix to avoid a potential interrupt storm. Gertjan van Wingerde provides a register-width fix for rt2x00, and a rt2x00 fix to prevent incorrectly detecting the rfkill status. He also provides a device ID patch. Hante Meuleman gives us three brcmfmac fixes, one that properly initializes a command structure, one that fixes a race condition that could lose usb requests, and one that removes some log spam. Marc Kleine-Budde offers an rt2x00 fix for a voltage setting on some specific devices. Mohammed Shafi Shajakhan sent an ath9k fix to avoid a crash related to using timers that aren't allocated when 2 wire bluetooth coexistence hardware is in use. Sergei Poselenov changes rt2800usb to do some validity checking for received packets, avoiding crashes on an ARM Soc. Stone Piao gives us an mwifiex fix for an incorrectly set skb length value for a command buffer. All of these are localized to their specific drivers, and relatively small. The power-related patches from Felix are bigger than I would like, but I merged them in consideration of their isolation to ath9k and the sensitive nature of power settings in wireless devices. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
| * libertas sdio: fix suspend when interface is downDaniel Drake2012-09-051-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | When the interface is down, the hardware is powered off. However, the suspend handler currently tries to send host sleep commands (when wakeup params are set) in this configuration, causing a system hang when going into suspend (the commands will never complete). Avoid this by detecting this situation and simply returning from the suspend handler without doing anything special. Signed-off-by: Daniel Drake <dsd@laptop.org> Acked-by: Dan Williams <dcbw@redhat.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rt2800usb: Added rx packet length validity checkSergei Poselenov2012-09-052-2/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On our system (ARM Cortex-M3 SOC running linux-2.6.33) frequent crashes were observed in the rt2800usb module because of the invalid length of the received packet (3392, 46920...). This patch adds the sanity check on the packet legth. Also, changed WARNING to ERROR in rt2x00lib_rxdone() so that the bad packet condition would be noticed. The fix was tested on the latest compat-wireless-3.5.1-1-snpc. Cc: stable@vger.kernel.org Signed-off-by: Sergei Poselenov <sposelenov@emcraft.com> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rt2x00: Fix rfkill polling prior to interface start.Gertjan van Wingerde2012-09-0511-0/+69
| | | | | | | | | | | | | | | | | | | | | | | | | | | | We need to program the rfkill switch GPIO pin direction to input at device initialization time, not only when the interface is brought up. Doing this only when the interface is brought up could lead to rfkill detecting the switch is turned on erroneously and inability to create the interface and bringing it up. Reported-and-tested-by: Andreas Messer <andi@bastelmap.de> Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Cc: <stable@vger.kernel.org> Acked-by: Ivo Van Doorn <ivdoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rt2x00: Fix word size of rt2500usb MAC_CSR19 register.Gertjan van Wingerde2012-09-052-9/+9
| | | | | | | | | | | | | | | | | | The register is 16 bits wide, not 32. Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Cc: <stable@vger.kernel.org> Acked-by: Ivo Van Doorn <ivdoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rt2x00: Identify ASUS USB-N53 device.Gertjan van Wingerde2012-09-051-1/+2
| | | | | | | | | | | | | | | | | | | | This is an RT3572 based device. Signed-off-by: Maximilian Engelhardt <maxi@daemonizer.de> Signed-off-by: Gertjan van Wingerde <gwingerde@gmail.com> Cc: <stable@vger.kernel.org> Acked-by: Ivo Van Doorn <ivdoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * brcmfmac: fix get rssi by clearing getvar struct.Hante Meuleman2012-09-051-7/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The function brcmf_cfg80211_get_station requests the RSSI from the device. The complete structure used needs to be cleared before sending the request to firmware. Otherwise the request fails filling the logs with "Could not get rssi (-2)" messages. Reviewed-by: Arend Van Spriel <arend@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * brcmfmac: fix race condition for rx and tx data.Hante Meuleman2012-09-051-10/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On both rx and tx there is was a race condition on the queueing of usb requests. When for example frame gets submitted it is possible that complete function gets called even before usb_submit_urb() returns. As a result it is possible that usb requests get losts, which was noticed on OMAP4 pandaboard platform. This patch fixes the race condition. Reviewed-by: Arend Van Spriel <arend@broadcom.com> Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * brcmfmac: dont use ZERO flag for usb INHante Meuleman2012-09-051-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | URB_ZERO_PACKET should only be set or bulk OUT and this condition is checked with a WARN_ON in usb_submit_urb(). This patch fixes brcmfmac to get rid of this warning filling the logs. Reviewed-by: Franky (Zhenhui) Lin <frankyl@broadcom.com> Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Signed-off-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * brcmfmac: fix use of dev_kfree_skb() in irq contextArend van Spriel2012-09-051-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The USB part of the brcmfmac did a dev_kfree_skb() that resulted in a warning in net/core/skbuff.c: Jul 11 04:53:33 lb-bun-10 kernel: [53282.667745] WARNING: at net/core/skbuff.c:490 skb_release_head_state+0xcc/0xe0() The brcmutil modules provides brcmu_pkt_buf_free_skb() which takes the context into account. This patch makes use of this function instead of dev_kfree_skb(). Reviewed-by: Pieter-Paul Giesberts <pieterpg@broadcom.com> Reviewed-by: Hante Meuleman <meuleman@broadcom.com> Signed-off-by: Arend van Spriel <arend@broadcom.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k: Fix a crash in 2 WIRE btcoex chipsetsMohammed Shafi Shajakhan2012-09-051-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Generic timers for BTCOEX functionality is applicable only for 3 WIRE BTCOEX (and MCI) chipsets. Hence btcoex->no_stomp_timer is allocated only 3 WIRE btcoex chipsets and in all the other cases its NULL. Make sure we stop the generic timer only if 'btcoex->hw_timer_enabled' is true(only if its up and running) Fixes the following crash [68757.020454] BUG: unable to handle kernel NULL pointer dereference at 0000000c [68757.020916] IP: [<f9b055c3>] ath9k_hw_gen_timer_stop+0x13/0x80 [ath9k_hw] [68757.021251] *pde = 00000000 [68757.024384] EIP: 0060:[<f9b055c3>] EFLAGS: 00010082 CPU: 0 [68757.024384] EIP is at ath9k_hw_gen_timer_stop+0x13/0x80 [ath9k_hw] [68757.024384] EAX: d32d0000 EBX: d32d0000 ECX: 00000000 EDX: 00000000 [68757.024384] ESI: e67c24c0 EDI: 00000296 EBP: e137be2c ESP: e137be20 [68757.024384] DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068 [68757.024384] CR0: 8005003b CR2: 0000000c CR3: 00b99000 CR4: 000407d0 [68757.024384] DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000 [68757.024384] DR6: ffff0ff0 DR7: 00000400 [68757.024384] Process kworker/u:2 (pid: 8917, ti=e137a000 task=ea7a6860 task.ti=e137a000) [68757.024384] Stack: [68757.024384] c06c4676 d32d0000 e67c24c0 e137be38 f81c9590 e67c1ca0 e137be40 f81c95d9 [68757.024384] e137be64 f81cd1c5 00000246 00000002 d32d0000 e67c05e0 e67c1ca0 e67c05e0 [68757.024384] 00000000 e137beac f81cdfa0 e137be84 00000246 00000246 e67c1ca0 e67c1ca0 [68757.024384] Call Trace: [68757.024384] [<c06c4676>] ? _raw_spin_lock_irqsave+0x86/0xa0 [68757.024384] [<f81c9590>] ath9k_gen_timer_stop+0x10/0x40 [ath9k] [68757.024384] [<f81c95d9>] ath9k_btcoex_stop_gen_timer+0x19/0x20 [ath9k] [68757.024384] [<f81cd1c5>] ath9k_ps_restore+0x85/0x110 [ath9k] [68757.024384] [<f81cdfa0>] ath9k_config+0x220/0x520 [ath9k] [68757.024384] [<f81cd47d>] ? ath9k_flush+0x15d/0x1b0 [ath9k] [68757.024384] [<f85c7ca5>] ieee80211_hw_config+0x135/0x2c0 [mac80211] [68757.024384] [<f860e3c8>] ieee80211_dynamic_ps_enable_work+0x198/0x5f0 [mac80211] Cc: Rajkumar Manoharan <rmanohar@qca.qualcomm.com> Cc: Bala Shanmugam <bkamatch@qca.qualcomm.com> Signed-off-by: Mohammed Shafi Shajakhan <mohammed@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k_hw: enable PA linearizationFelix Fietkau2012-09-051-3/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This feature had been disabled in ath9k because the code to support it was incomplete, but now the code is in sync with the internal QCA codebase, so it's time to enable it. On many newer devices, the calibration is assumed to be done with PA linearization enabled. Tests with a particular AR933x device showed that the signal emitted at full power was highly distorted and unreliable with PA linearization disabled. With this patch, the signal becomes clear and stability is improved. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k: fix PA linearization calibration related crashFelix Fietkau2012-09-054-8/+7
| | | | | | | | | | | | | | | | | | | | | | Before PAPRD training can run, the card needs to have sent a packet for thermal calibration. Sending a dummy packet with the PAPRD training flag set causes a crash under some circumstance. Fix the code by replacing the dummy tx with a delay that waits for a real packet tx to have occurred. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k_hw: disable PA linearization for AR9462Felix Fietkau2012-09-051-1/+2
| | | | | | | | | | | | | | Support for it is incomplete Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k_hw: calibrate PA input for PA predistortionFelix Fietkau2012-09-053-1/+111
| | | | | | | | | | | | | | | | Re-train if the calibrated PA linearization curve is out of bounds (affects AR933x and AR9485). Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k_hw: clear the AM2PM predistortion mask on AR933xFelix Fietkau2012-09-051-1/+5
| | | | | | | | | | | | | | That predistortion type is not supported Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * ath9k_hw: do not enable the MIB interrupt in the interrupt mask registerFelix Fietkau2012-09-051-3/+0
| | | | | | | | | | | | | | | | | | The interrupt is no longer handling it. While it shouldn't fire (wraparound is highly unlikely), the consequences would be fatal (interrupt storm). Disable the interrupt to prevent that from happening. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * rt2x00: fix voltage setting for RT3572/RT3592Marc Kleine-Budde2012-09-051-0/+1
| | | | | | | | | | | | | | | | | | | | | | According to the vendor driver v2.6.0.1, during the rf register init the SRAM voltage should be increased to 1.35V and after 1ms decreased back to 1.2V. This patch adds the field setting of LDO_CFG0_LDO_CORE_VLEVEL accordingly. Cc: Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by: Marc Kleine-Budde <mkl@blackshift.org> Acked-by: Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mwifiex: fix skb length issue when send a command to firmwareStone Piao2012-09-051-1/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When we send a command to firmware, we assumed that cmd_size will be always less than or equal to the structure size of host_cmd_ds_command. However, this is no longer true after we added AP support. There are some AP commands that Custom IE TLVs are included in command buffer, hence the cmd_size gets enlarged by the TLV data. We need to increase the skb length for the extra data. Signed-off-by: Stone Piao <piaoyun@marvell.com> Signed-off-by: Avinash Patil <patila@marvell.com> Signed-off-by: Amitkumar Karwar <akarwar@marvell.com> Signed-off-by: Bing Zhao <bzhao@marvell.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
* | |PATCH] seeq: Add missing spinlock initJean Delvare2012-09-071-0/+1
| | | | | | | | | | | | | | It doesn't seem this spinlock was properly initialized. Signed-off-by: Jean Delvare <khali@linux-fr.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net/mlx4_core: Return the error value in case of command initialization failureEugenia Emantayev2012-09-071-1/+2
| | | | | | | | | | | | | | | | | | If mlx4_cmd_init() failed, the init_one function returned success, although no resources were opened. Signed-off-by: Eugenia Emantayev <eugenia@mellanox.com> Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net/mlx4_core: Fixing error flow in case of QUERY_FW failureAviad Yehezkel2012-09-071-6/+7
| | | | | | | | | | | | | | | | The order of operations was wrong on the teardown flow. Signed-off-by: Aviad Yehezkel <aviadye@mellanox.com> Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.co.il> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net/mlx4_core: Looking for promiscuous entries on the correct portAviad Yehezkel2012-09-071-8/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The search for promisc entries was always done on the first port, While the addition is done on the correct port. This lead to resource leackage of promisc entries on the second port and brought to a state where we could no longer enter to promiscuous mode after enough iterations of "ifconfig promisc" on the second port. Fix that by using the correct port when searching. Reported-by: Marcelo Ricardo Leitner <mleitner@redhat.com> Signed-off-by: Aviad Yehezkel <aviadye@mellanox.com> Signed-off-by: Yevgeny Petrilin <yevgenyp@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net/mlx4_core: Add security check / enforcement for flow steering rules set ↵Hadar Hen Zion2012-09-071-0/+116
| | | | | | | | | | | | | | | | | | | | | | | | | | for VMs Since VFs may be mapped to VMs which aren't trusted entities, flow steering rules attached through the wrapper on behalf of VFs must be checked to make sure that their L2 specification relate to MAC address assigned to that VF, and add L2 specification if its missing. Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net/mlx4_core: Put Firmware flow steering structures in common header filesHadar Hen Zion2012-09-072-83/+83
| | | | | | | | | | | | | | | | | | To allow for usage of the flow steering Firmware structures in more locations over the driver, such as the resource tracker, move them from mcg.c to common header files. Signed-off-by: Hadar Hen Zion <hadarh@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | ibmveth: Fix alignment of rx queue bugSantiago Leon2012-09-051-17/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch fixes a bug found by Nish Aravamudan (https://lkml.org/lkml/2012/5/15/220) where the driver is not following the spec (it is not aligning the rx buffer on a 16-byte boundary) and the hypervisor aborts the registration, making the device unusable. The fix follows BenH's recommendation (https://lkml.org/lkml/2012/7/20/461) to replace the kmalloc+map for a single call to dma_alloc_coherent() because that function always aligns to a 16-byte boundary. The stable trees will run into this bug whenever the rx buffer kmalloc call returns something not aligned on a 16-byte boundary. Cc: <stable@vger.kernel.org> Signed-off-by: Santiago Leon <santil@linux.vnet.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | Merge branch 'fixes-for-3.6' of git://gitorious.org/linux-can/linux-canDavid S. Miller2012-09-051-1/+10
|\ \ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Marc Kleine-Budde says: ==================== this patch is for the v3.6 release cycle. Benoît Locher fixed a repeated frame bug in the mcp251x driver. He implemented the workaround suggested by the errata sheet. ==================== Signed-off-by: David S. Miller <davem@davemloft.net>
| * | can: mcp251x: avoid repeated frame bugBenoît Locher2012-09-031-1/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The MCP2515 has a silicon bug causing repeated frame transmission, see section 5 of MCP2515 Rev. B Silicon Errata Revision G (March 2007). Basically, setting TXBnCTRL.TXREQ in either SPI mode (00 or 11) will eventually cause the bug. The workaround proposed by Microchip is to use mode 00 and send a RTS command on the SPI bus to initiate the transmission. Cc: <stable@vger.kernel.org> Signed-off-by: Benoît Locher <Benoit.Locher@skf.com> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
* | | i825xx: fix paging fault on znet_probe()Wu Fengguang2012-09-041-4/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In znet_probe(), strncmp() may access beyond 0x100000 and trigger the below oops in kvm. Fix it by limiting the loop under 0x100000-8. I suspect the limit could be further decreased to 0x100000-sizeof(struct netidblk), however no datasheet at hand.. [ 3.744312] BUG: unable to handle kernel paging request at 80100000 [ 3.746145] IP: [<8119d12a>] strncmp+0xc/0x20 [ 3.747446] *pde = 01d10067 *pte = 00100160 [ 3.747493] Oops: 0000 [#1] DEBUG_PAGEALLOC [ 3.747493] Pid: 1, comm: swapper Not tainted 3.6.0-rc1-00018-g57bfc0a #73 Bochs Bochs [ 3.747493] EIP: 0060:[<8119d12a>] EFLAGS: 00010206 CPU: 0 [ 3.747493] EIP is at strncmp+0xc/0x20 [ 3.747493] EAX: 800fff4e EBX: 00000006 ECX: 00000006 EDX: 814d2bb9 [ 3.747493] ESI: 80100000 EDI: 814d2bba EBP: 8e03dfa0 ESP: 8e03df98 [ 3.747493] DS: 007b ES: 007b FS: 0000 GS: 00e0 SS: 0068 [ 3.747493] CR0: 8005003b CR2: 80100000 CR3: 016f7000 CR4: 00000690 [ 3.747493] DR0: 00000000 DR1: 00000000 DR2: 00000000 DR3: 00000000 [ 3.747493] DR6: ffff0ff0 DR7: 00000400 [ 3.747493] Process swapper (pid: 1, ti=8e03c000 task=8e040000 task.ti=8e03c000) [ 3.747493] Stack: [ 3.747493] 800fffff 00000000 8e03dfb4 816a1376 00000006 816a134a 00000000 8e03dfd0 [ 3.747493] 816819b5 816ed1c0 8e03dfe4 00000006 00000123 816ed604 8e03dfe4 81681b29 [ 3.747493] 00000000 81681a5b 00000000 00000000 8134e542 00000000 00000000 00000000 [ 3.747493] Call Trace: [ 3.747493] [<816a1376>] znet_probe+0x2c/0x26b [ 3.747493] [<816a134a>] ? dnet_driver_init+0xf/0xf [ 3.747493] [<816819b5>] do_one_initcall+0x6a/0x110 [ 3.747493] [<81681b29>] kernel_init+0xce/0x14b Signed-off-by: Fengguang Wu <fengguang.wu@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | net: usbnet: fix softirq storm on suspendBjørn Mork2012-09-031-4/+12
|/ / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Suspending an open usbnet device results in constant rescheduling of usbnet_bh. commit 65841fd5 "usbnet: handle remote wakeup asap" refactored the usbnet_bh code to allow sharing the urb allocate and submit code with usbnet_resume. In this process, a test for, and immediate return on, ENOLINK from rx_submit was unintentionally dropped. The rx queue will not grow if rx_submit fails, making usbnet_bh reschedule itself. This results in a softirq storm if the error is persistent. rx_submit translates the usb_submit_urb error EHOSTUNREACH into ENOLINK, so this is an expected and persistent error for a suspended device. The old code tested for this condition and avoided rescheduling. Putting this test back. Cc: <stable@vger.kernel.org> # v3.5 Cc: Ming Lei <ming.lei@canonical.com> Cc: Oliver Neukum <oneukum@suse.de> Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: qmi_wwan: add several new Gobi devicesBjørn Mork2012-09-011-0/+3
| | | | | | | | | | | | | | | | | | | | | | Gobi devices are composite, needing both the qcserial and qmi_wwan drivers to support all functions. Re-syncing the list of supported devices with qcserial. Cc: Aleksander Morgado <aleksander@lanedo.com> Cc: Thomas Tuttle <ttuttle@chromium.org> Signed-off-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@tempietto.lan>
* | fddi: 64 bit bug in smt_add_para()Dan Carpenter2012-09-011-1/+1
| | | | | | | | | | | | | | | | | | The intent was to set 4 bytes of data so that's why the sp_len is set to 4 on the next line. The cast to u_long pointer clears 8 bytes on 64 bit arches. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@tempietto.lan>
* | net: ethernet: fix kernel OOPS when remove davinci_mdio moduleBin Liu2012-08-311-1/+3
| | | | | | | | | | | | | | | | | | | | davinci mdio device is not unregistered from mdiobus when removing the module, which causes BUG_ON() when free the device from mdiobus. Calling mdiobus_unregister() before mdiobus_free() fixes the issue. Signed-off-by: Bin Liu <b-liu@ti.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | net: qmi_wwan: new device: Foxconn/Novatel E396Aleksander Morgado2012-08-311-0/+1
| | | | | | | | | | | | | | | | | | | | | | Foxconn-branded Novatel E396, Gobi3k modem. Cc: Dan Williams <dcbw@redhat.com> Cc: Bjørn Mork <bjorn@mork.no> Cc: Ben Chan <benchan@google.com> Signed-off-by: Aleksander Morgado <aleksander@lanedo.com> Acked-by: Bjørn Mork <bjorn@mork.no> Signed-off-by: David S. Miller <davem@davemloft.net>
* | usbnet: fix deadlock in resumeOliver Neukum2012-08-311-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | A usbnet device can share a multifunction device with a storage device. If the storage device is autoresumed the usbnet devices also needs to be autoresumed. Allocating memory with GFP_KERNEL can deadlock in this case. This should go back into all kernels that have commit 65841fd5132c3941cdf5df09e70df3ed28323212 That is 3.5 Signed-off-by: Oliver Neukum <oneukum@suse.de> CC: stable@kernel.org Signed-off-by: David S. Miller <davem@davemloft.net>
* | cs89x0 : packet reception not workingJaccon Bastiaansen2012-08-311-5/+5
| | | | | | | | | | | | | | | | | | The RxCFG register of the CS89x0 could be configured incorrectly (because of misplaced parentheses), resulting in the disabling of packet reception. Signed-off-by: Jaccon Bastiaansen <jaccon.bastiaansen@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | bnx2x: Correct the ndo_poll_controller callMerav Sicron2012-08-301-3/+5
| | | | | | | | | | | | | | | | | | | | This patch correct poll_bnx2x (ndo_poll_controller call) which was not functioning well with MSI-X. Signed-off-by: Merav Sicron <meravs@broadcom.com> Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | bnx2x: Move netif_napi_add to the open callMerav Sicron2012-08-305-13/+10
| | | | | | | | | | | | | | | | | | | | | | | | Move netif_napi_add for all queues from the probe call to the open call, to avoid the case that napi objects are added for queues that may eventually not be initialized and activated. With the former behavior, the driver could crash when netpoll was calling ndo_poll_controller. Signed-off-by: Merav Sicron <meravs@broadcom.com> Signed-off-by: Dmitry Kravkov <dmitry@broadcom.com> Signed-off-by: Eilon Greenstein <eilong@broadcom.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | e1000e: DoS while TSO enabled caused by link partner with small MSSBruce Allan2012-08-302-25/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | With a low enough MSS on the link partner and TSO enabled locally, the networking stack can periodically send a very large (e.g. 64KB) TCP message for which the driver will attempt to use more Tx descriptors than are available by default in the Tx ring. This is due to a workaround in the code that imposes a limit of only 4 MSS-sized segments per descriptor which appears to be a carry-over from the older e1000 driver and may be applicable only to some older PCI or PCIx parts which are not supported in e1000e. When the driver gets a message that is too large to fit across the configured number of Tx descriptors, it stops the upper stack from queueing any more and gets stuck in this state. After a timeout, the upper stack assumes the adapter is hung and calls the driver to reset it. Remove the unnecessary limitation of using up to only 4 MSS-sized segments per Tx descriptor, and put in a hard failure test to catch when attempting to check for message sizes larger than would fit in the whole Tx ring. Refactor the remaining logic that limits the size of data per Tx descriptor from a seemingly arbitrary 8KB to a limit based on the dynamic size of the Tx packet buffer as described in the hardware specification. Also, fix the logic in the check for space in the Tx ring for the next largest possible packet after the current one has been successfully queued for transmit, and use the appropriate defines for default ring sizes in e1000_probe instead of magic values. This issue goes back to the introduction of e1000e in 2.6.24 when it was split off from e1000. Reported-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: Bruce Allan <bruce.w.allan@intel.com> Cc: Stable <stable@vger.kernel.org> [2.6.24+] Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
OpenPOWER on IntegriCloud