summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* iio: st_sensors: make detection more helpfulLinus Walleij2015-05-071-1/+2
| | | | | | | | | | The ST sensors are detected by reading a WhoAmI register and matching the number found to a sensor name string. To make it easier to figure out what happens when things go wrong, print the WhoAmI value and the device name we're trying to match. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: st_sensors: make BDU optionalLinus Walleij2015-05-071-3/+5
| | | | | | | | | Not all sensors support BDU (block data update) and in fact a bunch of the in-kernel sensor settings do not specify the BDU address field. Make this optional. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: st_sensors: make interrupt optionalLinus Walleij2015-05-071-1/+11
| | | | | | | | | | | Some sensors such as magnetometers and pressure sensors doesn't have interrupts at all, and thus no DRDY setting applies. Make the assignment of an interrupt optional, and do not call st_sensors_set_drdy_int_pin() if there is no drdy (data ready) pin specified. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: st_sensors: print error when failing to get IRQLinus Walleij2015-05-071-1/+3
| | | | | | | | Print a proper error message if we're missing the trigger IRQ. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: kxcjk-1013: add some blank lines for readabilityAntonio Ospite2015-05-071-0/+3
| | | | | | | | | | Some extra blank lines between if checks don't hurt and improve readability. Signed-off-by: Antonio Ospite <ao2@ao2.it> Cc: Bastien Nocera <hadess@hadess.net> Reviewed-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: kxcjk-1013: add the "KXCJ9000" ACPI idAntonio Ospite2015-05-071-0/+1
| | | | | | | | | | | This id has been seen in the DSDT of the Teclast X98 Air 3G tablet based on Intel Bay Trail. Signed-off-by: Antonio Ospite <ao2@ao2.it> Cc: Bastien Nocera <hadess@hadess.net> Reviewed-by: Daniel Baluta <daniel.baluta@intel.com> Cc: <Stable@vger.kernel.org> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio:prox:sx9500 trivial simplification of return path in init function.Jonathan Cameron2015-05-071-5/+1
| | | | | | Signed-off-by: Jonathan Cameron <jic23@kernel.org> Reported-by: kbuild test robot <fengguang.wu@intel.com> Cc: Vlad Dogaru <vlad.dogaru@intel.com>
* staging:iio:light: Add some missing brackets to make sure code works as ↵Jonathan Cameron2015-05-071-1/+2
| | | | | | | | | intended. Note this is not a bug due to the fact the region cannot be reached without the sanity check passing. The autobuilder reported it as missaligned code which is kind of true as well. Signed-off-by: Jonathan Cameron <jic23@kernel.org> Cc: Roberta Dobrescu <roberta.dobrescu@gmail.com>
* iio:light:ltr501 bug in parameter sanity check.Jonathan Cameron2015-05-031-1/+1
| | | | | | | | | | | | | | Clearly the intent was to error if the value was not 0 or 1. As implemented we have (A != 0 || A != 1) which is always true as A is never both 0 and 1 at the same time. As the autobuilder suggested, && makes more sense for this error check. Reported-by: kbuild test robot <fengguang.wu@intel.com> Acked-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Cc: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* staging: iio: light: isl29018: Use standard sysfs attributes for scale and ↵Roberta Dobrescu2015-04-301-49/+177
| | | | | | | | | | | | | | | | | integration time This patch refactors the isl29018 driver code in order to use standard sysfs attributes for scale and integration time. ISL29018 light sensor uses four ranges and four ADC's resolutions which influence the calculated lux. Adc resolution is strongly connected to integration time and range should be controlled by scale. This patch introduces the usage of integration time and scale instead of adc resolution and range. Signed-off-by: Roberta Dobrescu <roberta.dobrescu@gmail.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* staging: iio: light: isl29018: Rename lux_scale to calibscaleRoberta Dobrescu2015-04-261-11/+11
| | | | | | | | | | | This patch renames lux_scale to calibscale and lux_uscale to ucalibscale. This is done in order to avoid confusion since these parameters are used for hardware applied calibration. Signed-off-by: Roberta Dobrescu <roberta.dobrescu@gmail.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* staging: iio: light: isl29018: Remove non-standard sysfs attributesRoberta Dobrescu2015-04-261-94/+0
| | | | | | | | | | | | This patch removes non-standard sysfs attributes range, range_available, adc_resolution and adc_resolution_available. It also removes the corresponding show and store functions. This is in preparation for using standard IIO attributes in order to move the code out of staging. Signed-off-by: Roberta Dobrescu <roberta.dobrescu@gmail.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: fix gpio bitnum init valueIrina Tirdea2015-04-261-1/+2
| | | | | | | | | | | | Initial value of gpio bitnum is set to -1, but the variable is declared as unsigned. Use a positive invalid value for initial gpio bitnum. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Suggested-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: refactor mma9553_read_rawIrina Tirdea2015-04-261-68/+33
| | | | | | | | Refactor code for simplicity and clarity. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Suggested-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: prefix naming fixesIrina Tirdea2015-04-261-7/+7
| | | | | | | | Add mma9553_ prefix to all local functions/declarations. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Suggested-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: use GENMASKIrina Tirdea2015-04-261-2/+2
| | | | | | | | | Use GENMASK instead of BIT or direct value to define a mask. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Suggested-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: comment and error message fixesIrina Tirdea2015-04-261-9/+10
| | | | | | | | | | | Use "GPIO" instead of "gpio" and "ACPI" instead of "acpi". Includes a couple of small style fixes in comments (missing full stop, whitespace, paranthesis). Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Suggested-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: return 0 as indication of successIrina Tirdea2015-04-261-1/+1
| | | | | | | | | Use return 0 instead of return ret to mark clearly the success return path. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Suggested-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9553: check for error in reading initial activity and stepcntIrina Tirdea2015-04-261-3/+6
| | | | | | | | | | | | When configuring gpio, we need to read initial values for activity and step count. This function may fail due to i2c read errors. Check the error code returned by mma9553_read_activity_stepcnt and return the appropriate error in gpio config function. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Reported-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9551_core: typo fix in RSC APP IDIrina Tirdea2015-04-262-2/+2
| | | | | | | | Fix typo in Reset/Suspend/Clear Application ID definition. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Reported-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: mma9551_core: wrong doc fixesIrina Tirdea2015-04-261-3/+3
| | | | | | | | Fix docummentation for mma9553_read_* functions. Signed-off-by: Irina Tirdea <irina.tirdea@intel.com> Reported-by: Hartmut Knaack <knaack.h@gmx.de> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add support for ltr301 chipDaniel Baluta2015-04-262-5/+73
| | | | | | | | | | | | | | | Added support for Liteon 301 Ambient light sensor. Since LTR-301 and LTR-501 are register compatible(and even have same part id), LTR-501 driver has been extended to support both devices. LTR-501 is similar to LTR-301 in ALS sensing, But the only difference is, LTR-501 also supports proximity sensing. LTR-501 - ALS + Proximity combo LTR-301 - ALS sensor. Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Signed-off-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add support for ltr559 chipDaniel Baluta2015-04-262-35/+186
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This device is register compatible with LTR501, with a minor difference for ALS control register as showed below: ALS Control register for LTR501: 7 6 5 4 3 2 1 0 +------+------+------+------+------+------+------+------+ | | | | | | Reserved | Gain | SW | ALS Mode | | | | Reset| | +------+------+------+------+------+------+------+------+ ALS Control register for LTR559: 7 6 5 4 3 2 1 0 +------+------+------+------+------+------+------+------+ | | | | | | Reserved | Gain | SW | ALS | | | | Reset| Mode | +------+------+------+------+------+------+------+------+ We handle this difference by introducing ltr501_chip_info. Datasheet for LTR559 is at: http://optoelectronics.liteon.com/upload/download/DS86-2013-0003/S_110_LTR-559ALS-01_DS_V1.pdf Signed-off-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add ACPI enumeration supportKuppuswamy Sathyanarayanan2015-04-191-0/+8
| | | | | | | Added ACPI enumeration support for LTR501 chip. Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add interrupt rate control supportKuppuswamy Sathyanarayanan2015-04-191-7/+406
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Added rate control support for ALS and proximity threshold interrupts.Also, Added support to modify and read ALS & proximity sensor sampling frequency. LTR-501 supports interrupt rate control using persistence register settings. Writing <n> to persistence register would generate interrupt only if there are <n> consecutive data values outside the threshold range. Since we don't have any existing ABI's to directly control the persistence register count, we have implemented the rate control using IIO_EV_INFO_PERIOD. _period event attribute represents the amount of time in seconds an event should be true for the device to generate the interrupt. So using _period value and device frequency, persistence count is calculated in driver using following logic. count = period / measurement_rate If the given period is not a multiple of measurement rate then we round up the value to next multiple. This patch also handles change to persistence count whenever there is change in frequency. Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add interrupt supportKuppuswamy Sathyanarayanan2015-04-191-6/+313
| | | | | | | | | | | This patch adds interrupt support for Liteon 501 chip. Interrupt will be generated whenever ALS or proximity data exceeds values given in upper and lower threshold register settings. Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add integration time supportKuppuswamy Sathyanarayanan2015-04-191-1/+89
| | | | | | | Added support to modify and read ALS integration time. Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio:tmp006: Prefix #defines with TMP006_Peter Meerwald2015-04-181-3/+3
| | | | | | | just cleanup, no functional change Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio:tsl2563: Use tsl2563_ prefix for driver's functionsPeter Meerwald2015-04-181-18/+18
| | | | | | | just cleanup, no functional change Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio:tsl4531: Fix leftover TCS3472_ prefix in tsl4531 driverPeter Meerwald2015-04-181-5/+5
| | | | | | | just cleanup, no functional change Signed-off-by: Peter Meerwald <pmeerw@pmeerw.net> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: sx9500: add GPIO reset pinVlad Dogaru2015-04-181-2/+20
| | | | | | | | If a GPIO reset pin is listed in ACPI or Device Tree, use it to reset the device on initialization. Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: sx9500: refactor GPIO interrupt codeVlad Dogaru2015-04-181-18/+11
| | | | | Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: sx9500: optimize power usageVlad Dogaru2015-04-181-54/+306
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In the interest of lowering power usage, we only activate the proximity channels and interrupts that we are currently using. For raw reads, we activate the corresponding channel and the data ready interrupt and wait for the interrupt to trigger. If no interrupt is available, we wait for the documented scan period, as specified in the datasheet. The following types of usage patterns may overlap: * raw proximity reads (need a single data ready interrupt) * trigger usage (needs data ready interrupts as long as active) * proximity events (need near/far interrupts) * triggered buffer reads (don't need any interrupts, but are usually coupled with our own trigger. To mitigate all possible patterns, we implement usage counting for all the resources used: data ready interrupts, near/far interrupts and individual channels. The device enters sleep mode as documented in the data sheet when its buffer, trigger and events are disabled, and no raw reads are currently running. Because of this new usage pattern, it is important that we give the device a chance to perform an initial compensation for all its channels at probe time. Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: mlx90614: Fix duplicate const warningVianney le Clément de Saint-Marcq2015-04-181-1/+1
| | | | | | | | Fix a typo triggering a duplicate const warning on some compilers. Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement@essensium.com> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: ltr501: Add regmap support.Kuppuswamy Sathyanarayanan2015-04-181-43/+86
| | | | | | | | Added regmap support. It will be useful to handle bitwise updates to als & ps control registers. Signed-off-by: Kuppuswamy Sathyanarayanan <sathyanarayanan.kuppuswamy@linux.intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: sx9500: fix formattingVlad Dogaru2015-04-181-1/+1
| | | | | Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: sx9500: rename GPIO interrupt pinVlad Dogaru2015-04-181-1/+2
| | | | | Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: sx9500: add power managementVlad Dogaru2015-04-091-0/+47
| | | | | Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: light: ltr501: Fix alignment to match open parenthesisDaniel Baluta2015-04-091-19/+23
| | | | | | | | | | | This makes ltr501 code consistent with the coding style adopted for the new drivers added to IIO. We prepare the path for adding support for LTR559 chip. Reported by checkpatch.pl Signed-off-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: adc: ti_am335x_adc: refactor DT parsing into a functionVignesh R2015-04-091-10/+19
| | | | | | | | | | Refactor DT parsing into a separate function from probe() to help addition of more DT parameters later. No functional changes. Signed-off-by: Vignesh R <vigneshr@ti.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: mlx90614: Check for errors in read valuesVianney le Clément de Saint-Marcq2015-04-091-0/+5
| | | | | | | | | The device uses the MSB of the returned temperature value as an error flag. Return a read error when this bit is set. Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement@essensium.com> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: mlx90614: Add power managementVianney le Clément de Saint-Marcq2015-04-092-2/+253
| | | | | | | | | | | | | | | | | | | Add support for system sleep and runtime power management. To wake up the device, the SDA line should be held low for at least 33ms while SCL is high. As this is not possible using the i2c API (and not supported by all i2c adapters), a GPIO connected to the SDA line is needed. The GPIO is named "wakeup" and can be specified in a device tree with the "wakeup-gpios" binding. If the wake-up GPIO is not given, disable power management for the device. Entering sleep requires an SMBus byte access, hence power management is also disabled if byte access is not supported by the adapter. Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement@essensium.com> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: mlx90614: Add emissivity settingVianney le Clément de Saint-Marcq2015-04-091-3/+103
| | | | | | | | | | | | | | | | The mapping from the 16-bit EEPROM value to the decimal 0-1 range is approximate. A special case ensures 0xFFFF shows as 1.0 instead of 0.999998565. Writing to EEPROM requires an explicit erase by writing zero. In addition, it takes 20ms for the erase/write to complete. During this time no EEPROM register should be accessed. Therefore, two msleep()s are added to the write function and a mutex protects against concurrent access. Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement@essensium.com> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: mlx90614: Add devicetree bindings documentationVianney le Clément de Saint-Marcq2015-04-092-0/+16
| | | | | | | | Also introduce "melexis" as a vendor prefix for device tree bindings. Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement@essensium.com> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: core: Introduce IIO_CHAN_INFO_CALIBEMISSIVITYVianney le Clément de Saint-Marcq2015-04-093-0/+13
| | | | | | | | | | | | | | | | | Contact-less IR temperature sensors measure the temperature of an object by using its thermal radiation. Surfaces with different emissivity ratios emit different amounts of energy at the same temperature. IIO_CHAN_INFO_CALIBEMISSIVITY allows the user to inform the sensor of the emissivity of the object in front of it, in order to effectively measure its temperature. A device providing such setting is Melexis's MLX90614: http://melexis.com/Assets/IR-sensor-thermometer-MLX90614-Datasheet-5152.aspx. Signed-off-by: Vianney le Clément de Saint-Marcq <vianney.leclement@essensium.com> Cc: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* iio: accel: kxcjk1013: allow using an external triggerVlad Dogaru2015-04-091-15/+23
| | | | | | | | | | | | | | | | In its present state, the driver mandates that its buffer only be triggered by one of the device's own triggers (data ready or any motion). This is not always desirable, for example because the interrupt pins may not be wired in. Patch the driver to be able to accept using an external trigger, such as one based on hrtimer. When using such a trigger, we need to ensure that the device is powered on when the buffer is started. We do that by setting setup_ops for the buffer. Signed-off-by: Vlad Dogaru <vlad.dogaru@intel.com> Reviewed-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* tools: iio: Add iio targets in tools MakefileRoberta Dobrescu2015-04-091-3/+4
| | | | | | | | | | | | | | | This patch adds targets for building and cleaning iio tools to tools/Makefile. To build iio tools from the toplevel kernel directory one should call: $ make -C tools iio and for cleaning it $ make -C tools iio_clean Signed-off-by: Roberta Dobrescu <roberta.dobrescu@gmail.com> Reviewed-by: Daniel Baluta <daniel.baluta@intel.com> Signed-off-by: Jonathan Cameron <jic23@kernel.org>
* Merge 4.0-rc7 into staging-nextGreg Kroah-Hartman2015-04-07304-1381/+2831
|\ | | | | | | | | | | | | We want those fixes (iio primarily) into the -next branch to help with merge and testing issues. Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
| * Linux 4.0-rc7v4.0-rc7Linus Torvalds2015-04-071-1/+1
| |
| * Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/netLinus Torvalds2015-04-0721-71/+78
| |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Pull networking fixes from David Miller: 1) In TCP, don't register an FRTO for cumulatively ACK'd data that was previously SACK'd, from Neal Cardwell. 2) Need to hold RNL mutex in ipv4 multicast code namespace cleanup, from Cong WANG. 3) Similarly we have to hold RNL mutex for fib_rules_unregister(), also from Cong WANG. 4) Revert and rework netns nsid allocation fix, from Nicolas Dichtel. 5) When we encapsulate for a tunnel device, skb->sk still points to the user socket. So this leads to cases where we retraverse the ipv4/ipv6 output path with skb->sk being of some other address family (f.e. AF_PACKET). This can cause things to crash since the ipv4 output path is dereferencing an AF_PACKET socket as if it were an ipv4 one. The short term fix for 'net' and -stable is to elide these socket checks once we've entered an encapsulation sequence by testing xmit_recursion. Longer term we have a better solution wherein we pass the tunnel's socket down through the output paths, but that is way too invasive for 'net' and -stable. From Hannes Frederic Sowa. 6) l2tp_init() failure path forgets to unregister per-net ops, from Cong WANG. * git://git.kernel.org/pub/scm/linux/kernel/git/davem/net: net/mlx4_core: Fix error message deprecation for ConnectX-2 cards net: dsa: fix filling routing table from OF description l2tp: unregister l2tp_net_ops on failure path mvneta: dont call mvneta_adjust_link() manually ipv6: protect skb->sk accesses from recursive dereference inside the stack netns: don't allocate an id for dead netns Revert "netns: don't clear nsid too early on removal" ip6mr: call del_timer_sync() in ip6mr_free_table() net: move fib_rules_unregister() under rtnl lock ipv4: take rtnl_lock and mark mrt table as freed on namespace cleanup tcp: fix FRTO undo on cumulative ACK of SACKed range xen-netfront: transmit fully GSO-sized packets
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-12 04:23:51 +0100