user/sven/linux.git/drivers, branch v5.0.12

leds: trigger: netdev: use memcpy in device_name_store

2019-05-04T07:21:23Z

[ Upstream commit 909346433064b8d840dc82af26161926b8d37558 ] If userspace doesn't end the input with a newline (which can easily happen if the write happens from a C program that does write(fd, iface, strlen(iface))), we may end up including garbage from a previous, longer value in the device_name. For example # cat device_name # printf 'eth12' > device_name # cat device_name eth12 # printf 'eth3' > device_name # cat device_name eth32 I highly doubt anybody is relying on this behaviour, so switch to simply copying the bytes (we've already checked that size is < IFNAMSIZ) and unconditionally zero-terminate it; of course, we also still have to strip a trailing newline. This is also preparation for future patches. Fixes: 06f502f57d0d ("leds: trigger: Introduce a NETDEV trigger") Signed-off-by: Rasmus Villemoes Acked-by: Pavel Machek Signed-off-by: Jacek Anaszewski Signed-off-by: Sasha Levin (Microsoft)

leds: pca9532: fix a potential NULL pointer dereference

2019-05-04T07:21:23Z

[ Upstream commit 0aab8e4df4702b31314a27ec4b0631dfad0fae0a ] In case of_match_device cannot find a match, return -EINVAL to avoid NULL pointer dereference. Fixes: fa4191a609f2 ("leds: pca9532: Add device tree support") Signed-off-by: Kangjie Lu Signed-off-by: Jacek Anaszewski Signed-off-by: Sasha Levin (Microsoft)

iommu/amd: Reserve exclusion range in iova-domain

2019-05-04T07:21:22Z

[ Upstream commit 8aafaaf2212192012f5bae305bb31cdf7681d777 ] If a device has an exclusion range specified in the IVRS table, this region needs to be reserved in the iova-domain of that device. This hasn't happened until now and can cause data corruption on data transfered with these devices. Treat exclusion ranges as reserved regions in the iommu-core to fix the problem. Fixes: be2a022c0dd0 ('x86, AMD IOMMU: add functions to parse IOMMU memory mapping requirements for devices') Signed-off-by: Joerg Roedel Reviewed-by: Gary R Hook Signed-off-by: Sasha Levin (Microsoft)

nvmet: fix error flow during ns enable

2019-05-04T07:21:22Z

[ Upstream commit a536b49785759bf99465fdf6e248d34322123fcd ] In case we fail to enable p2pmem on the current namespace, disable the backing store device before exiting. Cc: Stephen Bates Signed-off-by: Max Gurtovoy Signed-off-by: Christoph Hellwig Signed-off-by: Sasha Levin (Microsoft)

nvmet: fix building bvec from sg list

2019-05-04T07:21:22Z

[ Upstream commit 02db99548d3608a625cf481cff2bb7b626829b3f ] There are two mistakes for building bvec from sg list for file backed ns: - use request data length to compute number of io vector, this way doesn't consider sg->offset, and the result may be smaller than required io vectors - bvec->bv_len isn't capped by sg->length This patch fixes this issue by building bvec from sg directly, given the whole IO stack is ready for multi-page bvec. Reported-by: Yi Zhang Fixes: 3a85a5de29ea ("nvme-loop: add a NVMe loopback host driver") Signed-off-by: Ming Lei Signed-off-by: Christoph Hellwig Signed-off-by: Sasha Levin (Microsoft)

nvme-multipath: relax ANA state check

2019-05-04T07:21:21Z

[ Upstream commit cc2278c413c3a06a93c23ee8722e4dd3d621de12 ] When undergoing state transitions I/O might be requeued, hence we should always call nvme_mpath_set_live() to schedule requeue_work whenever the nvme device is live, independent on whether the old state was live or not. Signed-off-by: Martin George Signed-off-by: Gargi Srinivas Signed-off-by: Hannes Reinecke Signed-off-by: Christoph Hellwig Signed-off-by: Sasha Levin (Microsoft)

gpio: of: Fix of_gpiochip_add() error path

2019-05-04T07:21:21Z

[ Upstream commit f7299d441a4da8a5088e651ea55023525a793a13 ] If the call to of_gpiochip_scan_gpios() in of_gpiochip_add() fails, no error handling is performed. This lead to the need of callers to call of_gpiochip_remove() on failure, which causes "BAD of_node_put() on ..." if the failure happened before the call to of_node_get(). Fix this by adding proper error handling. Note that calling gpiochip_remove_pin_ranges() multiple times causes no harm: subsequent calls are a no-op. Fixes: dfbd379ba9b7431e ("gpio: of: Return error if gpio hog configuration failed") Signed-off-by: Geert Uytterhoeven Reviewed-by: Mukesh Ojha Signed-off-by: Linus Walleij Signed-off-by: Sasha Levin (Microsoft)

gpio: of: Check for "spi-cs-high" in child instead of parent node

2019-05-04T07:21:20Z

[ Upstream commit 7ce40277bf848391705011ba37eac2e377cbd9e6 ] "spi-cs-high" is going to be specified in child node of an SPI controller's representing attached SPI device, so change the code to look for it there, instead of checking parent node. Signed-off-by: Andrey Smirnov Cc: Linus Walleij Cc: Bartosz Golaszewski Cc: Chris Healy Cc: linux-gpio@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Linus Walleij Signed-off-by: Sasha Levin (Microsoft)

gpio: of: Check propname before applying "cs-gpios" quirks

2019-05-04T07:21:20Z

[ Upstream commit e5545c94e43b8f6599ffc01df8d1aedf18ee912a ] SPI GPIO device has more than just "cs-gpio" property in its node and would request those GPIOs as a part of its initialization. To avoid applying CS-specific quirk to all of them add a check to make sure that propname is "cs-gpios". Signed-off-by: Andrey Smirnov Cc: Linus Walleij Cc: Bartosz Golaszewski Cc: Chris Healy Cc: linux-gpio@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Linus Walleij Signed-off-by: Sasha Levin (Microsoft)

libata: fix using DMA buffers on stack

2019-05-04T07:21:20Z

[ Upstream commit dd08a8d9a66de4b54575c294a92630299f7e0fe7 ] When CONFIG_VMAP_STACK=y, __pa() returns incorrect physical address for a stack virtual address. Stack DMA buffers must be avoided. Signed-off-by: raymond pang Signed-off-by: Jens Axboe Signed-off-by: Sasha Levin (Microsoft)