- Oct 13, 2021
-
-
Srinivas Kandagatla authored
Add .cell_post_process callback for imx-ocotp to deal with MAC address, since MAC address need to be reversed byte for some i.MX SoCs. Tested-by:
Joakim Zhang <qiangqing.zhang@nxp.com> Signed-off-by:
Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Link: https://lore.kernel.org/r/20211013131957.30271-4-srinivas.kandagatla@linaro.org Signed-off-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-
Srinivas Kandagatla authored
Some NVMEM providers have certain nvmem cells encoded, which requires post processing before actually using it. For example mac-address is stored in either in ascii or delimited or reverse-order. Having a post-process callback hook to provider drivers would enable them to do this vendor specific post processing before nvmem consumers see it. Tested-by:
-
Srinivas Kandagatla authored
In the existing design, we do not create a instance per nvmem cell consumer but we directly refer cell from nvmem cell list that are added to provider. However this design has some limitations when consumers want to assign name or connection id the nvmem cell instance, ex: via "nvmem-cell-names" or id in nvmem_cell_get(id). Having a name associated with nvmem cell consumer instance will help provider drivers in performing post processing of nvmem cell data if required before data is seen by the consumers. This is pretty normal with some vendors storing nvmem cells like mac-address in a vendor specific data layouts that are not directly usable by the consumer drivers. With this patch nvmem cell will be created dynamically during nvmem_cell_get and destroyed in nvmem_cell_put, allowing consumers to associate name with nvmem cell consumer instance. With this patch a new struct nvmem_cell_entry replaces struct nvmem_cell for storing nvmem cell information within the core. This patch does not change nvmem-consumer interface based on nvmem_cell. Tested-by:
-
Stephen Boyd authored
If a cell has 'nbits' equal to a multiple of BITS_PER_BYTE the logic *p &= GENMASK((cell->nbits%BITS_PER_BYTE) - 1, 0); will become undefined behavior because nbits modulo BITS_PER_BYTE is 0, and we subtract one from that making a large number that is then shifted more than the number of bits that fit into an unsigned long. UBSAN reports this problem: UBSAN: shift-out-of-bounds in drivers/nvmem/core.c:1386:8 shift exponent 64 is too large for 64-bit type 'unsigned long' CPU: 6 PID: 7 Comm: kworker/u16:0 Not tainted 5.15.0-rc3+ #9 Hardware name: Google Lazor (rev3+) with KB Backlight (DT) Workqueue: events_unbound deferred_probe_work_func Call trace: dump_backtrace+0x0/0x170 show_stack+0x24/0x30 dump_stack_lvl+0x64/0x7c dump_stack+0x18/0x38 ubsan_epilogue+0x10/0x54 __ubsan_handle_shift_out_of_bounds+0x180/0x194 __nvmem_cell_read+0x1ec/0x21c nvmem_cell_read+0x58/0x94 nvmem_cell_read_variable_common+0x4c/0xb0 nvmem_cell_read_variable_le_u32+0x40/0x100 a6xx_gpu_init+0x170/0x2f4 adreno_bind+0x174/0x284 component_bind_all+0xf0/0x264 msm_drm_bind+0x1d8/0x7a0 try_to_bring_up_master+0x164/0x1ac __component_add+0xbc/0x13c component_add+0x20/0x2c dp_display_probe+0x340/0x384 platform_probe+0xc0/0x100 really_probe+0x110/0x304 __driver_probe_device+0xb8/0x120 driver_probe_device+0x4c/0xfc __device_attach_driver+0xb0/0x128 bus_for_each_drv+0x90/0xdc __device_attach+0xc8/0x174 device_initial_probe+0x20/0x2c bus_probe_device+0x40/0xa4 deferred_probe_work_func+0x7c/0xb8 process_one_work+0x128/0x21c process_scheduled_works+0x40/0x54 worker_thread+0x1ec/0x2a8 kthread+0x138/0x158 ret_from_fork+0x10/0x20 Fix it by making sure there are any bits to mask out. Fixes: 69aba794 ("nvmem: Add a simple NVMEM framework for consumers") Cc: Douglas Anderson <dianders@chromium.org> Cc: stable@vger.kernel.org Signed-off-by:
Stephen Boyd <swboyd@chromium.org> Signed-off-by:
-
- Sep 21, 2021
-
-
Geert Uytterhoeven authored
The Nintendo Wii and Wii U OTP is only present on Nintendo Wii and Wii U consoles. Hence add a dependency on WII, to prevent asking the user about this driver when configuring a kernel without Nintendo Wii and Wii U console support. Fixes: 3683b761 ("nvmem: nintendo-otp: Add new driver for the Wii and Wii U OTP") Reviewed-by:
Emmanuel Gil Peyrot <linkmauve@linkmauve.fr> Signed-off-by:
Geert Uytterhoeven <geert+renesas@glider.be> Link: https://lore.kernel.org/r/01318920709dddc4d85fe895e2083ca0eee234d8.1631611652.git.geert+renesas@glider.be Signed-off-by:
-
- Aug 13, 2021
-
-
Link Mauve authored
This OTP is read-only and contains various keys used by the console to decrypt, encrypt or verify various pieces of storage. Its size depends on the console, it is 128 bytes on the Wii and 1024 bytes on the Wii U (split into eight 128 bytes banks). It can be used directly by writing into one register and reading from the other one, without any additional synchronisation. This driver was written based on reversed documentation, see: https://wiiubrew.org/wiki/Hardware/OTP Tested-by: Jonathan Neuschäfer <j.ne@posteo.net> # on Wii Tested-by: Emmanuel Gil Peyrot <linkmauve@linkmauve.fr> # on Wii U Signed-off-by:
-
- Aug 06, 2021
-
-
Srinivas Kandagatla authored
Current error path on failure of validating keepout regions is calling put_device, eventhough the device is not even registered at that point. Fix this by adding proper error handling of freeing ida and nvmem. Fixes: fd3bb8f5 ("nvmem: core: Add support for keepout regions") Cc: <stable@vger.kernel.org> Signed-off-by:
-
Rajendra Nayak authored
On sc7280, to reliably blow fuses, we need an additional vote on max performance state of 'MX' power-domain. Add support for power-domain performance state voting in the driver. Reviewed-by:
Douglas Anderson <dianders@chromium.org> Signed-off-by:
Rajendra Nayak <rnayak@codeaurora.org> Signed-off-by:
-
Rajendra Nayak authored
qfprom_disable_fuse_blowing() disables a bunch of resources, and then does a few register writes in the 'conf' address space. It works perhaps because the resources are needed only for the 'raw' register space writes, and that the 'conf' space allows read/writes regardless. However that makes the code look confusing, so just move the register writes before turning off the resources in the function. Reviewed-by:
-
- Jun 11, 2021
-
-
Christophe JAILLET authored
'for_each_child_of_node' performs an of_node_get on each iteration, so a return from the middle of the loop requires an of_node_put. Fixes: e888d445 ("nvmem: resolve cells from DT at registration time") Cc: <stable@vger.kernel.org> Signed-off-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by:
-
Jiri Prchal authored
Added enum and string for FRAM (ferroelectric RAM) to expose it as file named "fram". Added documentation of sysfs file. Signed-off-by:
Jiri Prchal <jiri.prchal@aksignal.cz> Link: https://lore.kernel.org/r/20210611094601.95131-2-jiri.prchal@aksignal.cz Signed-off-by:
-
Douglas Anderson authored
In review feedback Joe Perches found the existing comment confusing. Let's use something based on the wording proposed by Joe. Suggested-by:
Joe Perches <joe@perches.com> Signed-off-by:
-
Douglas Anderson authored
The caller doesn't modify the memory pointed to by the pointer so it can be const. Suggested-by:
-
Rajendra Nayak authored
Fix a missed newline, change an 'if' to 'else if' and update a comment which is stale after the merge of '5a1bea2a: nvmem: qfprom: Add support for fuseblowing on sc7280' Signed-off-by:
-
Samuel Holland authored
This device currently reports an "Unknown" type in sysfs. Since it is an eFuse hardware device, set its type to OTP. Signed-off-by:
Samuel Holland <samuel@sholland.org> Acked-by:
Chen-Yu Tsai <wens@csie.org> Signed-off-by:
-
- May 14, 2021
-
-
Christophe JAILLET authored
'ret' is known to be 0 here. The expected error status is stored in 'status', so use it instead. Also change %d in %u, because status is an u32, not a int. Fixes: 096030e7 ("nvmem: sprd: Add Spreadtrum SoCs eFuse support") Acked-by:
Chunyan Zhang <zhang.lyra@gmail.com> Signed-off-by:
-
- May 10, 2021
-
-
Michael Walle authored
Until now, the of_node of the parent device is used. Some devices provide more than just the nvmem provider. To avoid name space clashes, add a way to allow specifying the nvmem cells in subnodes. Consider the following example: flash@0 { compatible = "jedec,spi-nor"; partitions { compatible = "fixed-partitions"; #address-cells = <1>; #size-cells = <1>; partition@0 { reg = <0x000000 0x010000>; }; }; otp { compatible = "user-otp"; #address-cells = <1>; #size-cells = <1>; serial-number@0 { reg = <0x0 0x8>; }; }; }; There the nvmem provider might be the MTD partition or the OTP region of the flash. Add a new config->of_node parameter, which if set, will be used instead of the parent's of_node. Signed-off-by:Michael Walle <michael@walle.cc> Acked-by:
Miquel Raynal <miquel.raynal@bootlin.com> Link: https://lore.kernel.org/linux-mtd/20210424110608.15748-2-michael@walle.cc
-
- Apr 02, 2021
-
-
Rajendra Nayak authored
Handle the differences across LDO voltage needed for blowing fuses, and the blow timer value, identified using a minor version of 15 on sc7280. Signed-off-by:
Ravi Kumar Bokka <rbokka@codeaurora.org> Signed-off-by:
-
Srinivas Kandagatla authored
Fix below error reporte by kernel test robot rmem.c:(.text+0x14e): undefined reference to memremap s390x-linux-gnu-ld: rmem.c:(.text+0x1b6): undefined reference to memunmap Fixes: 5a3fa75a ("nvmem: Add driver to expose reserved memory as nvmem") Reported-by:
kernel test robot <lkp@intel.com> Reviewed-by:
Nicolas Saenz Julienne <nsaenzjulienne@suse.de> Signed-off-by:
-
Colin Ian King authored
The shifting of the u8 integer buf[3] by 24 bits to the left will be promoted to a 32 bit signed int and then sign-extended to a u64. In the event that the top bit of buf[3] is set then all then all the upper 32 bits of the u64 end up as also being set because of the sign-extension. Fix this by casting buf[i] to a u64 before the shift. Fixes: a28e824f ("nvmem: core: Add functions to make number reading easy") Reviewed-by:
Colin Ian King <colin.king@canonical.com> Signed-off-by:
-
Douglas Anderson authored
Sometimes the clients of nvmem just want to get a number out of nvmem. They don't want to think about exactly how many bytes the nvmem cell took up. They just want the number. Let's make it easy. In general this concept is useful because nvmem space is precious and usually the fewest bits are allocated that will hold a given value on a given system. However, even though small numbers might be fine on one system that doesn't mean that logically the number couldn't be bigger. Imagine nvmem containing a max frequency for a component. On one system perhaps that fits in 16 bits. On another system it might fit in 32 bits. The code reading this number doesn't care--it just wants the number. We'll provide two functions: nvmem_cell_read_variable_le_u32() and nvmem_cell_read_variable_le_u64(). Comparing these to the existing functions like nvmem_cell_read_u32(): * These new functions have no problems if the value was stored in nvmem in fewer bytes. It's OK to use these function as long as the value stored will fit in 32-bits (or 64-bits). * These functions avoid problems that the earlier APIs had with bit offsets. For instance, you can't use nvmem_cell_read_u32() to read a value has nbits=32 and bit_offset=4 because the nvmem cell must be at least 5 bytes big to hold this value. The new API accounts for this and works fine. * These functions make it very explicit that they assume that the number was stored in little endian format. The old functions made this assumption whenever bit_offset was non-zero (see nvmem_shift_read_buffer_in_place()) but didn't whenever the bit_offset was zero. NOTE: it's assumed that we don't need an 8-bit or 16-bit version of this function. The 32-bit version of the function can be used to read 8-bit or 16-bit data. At the moment, I'm only adding the "unsigned" versions of these functions, but if it ends up being useful someone could add a "signed" version that did 2's complement sign extension. At the moment, I'm only adding the "little endian" versions of these functions. Adding the "big endian" version would require adding "big endian" support to nvmem_shift_read_buffer_in_place(). Signed-off-by:
-
Rafał Miłecki authored
This driver provides access to Broadcom's NVRAM. Signed-off-by:
Rafał Miłecki <rafal@milecki.pl> Signed-off-by:
-
Ravi Kumar Bokka authored
QFPROM controller hardware requires 1.8V min for fuse blowing. So, this change sets the voltage to 1.8V, required to blow the fuse for qfprom-efuse controller. To disable fuse blowing, we set the voltage to 0V since this may be a shared rail and may be able to run at a lower rate when we're not blowing fuses. Fixes: 93b4e49f ("nvmem: qfprom: Add fuse blowing support") Reported-by:
-
Zheng Yongjun authored
Replace a comma between expression statements by a semicolon. Reviewed-by:
Bjorn Andersson <bjorn.andersson@linaro.org> Signed-off-by:
Zheng Yongjun <zhengyongjun3@huawei.com> Signed-off-by:
-
- Feb 05, 2021
-
-
Subbaraman Narayanamurthy authored
"sdam->pdev" is uninitialized and it is used to print error logs. Fix it. Since device pointer can be used from sdam_config, use it directly thereby removing pdev pointer. Fixes: 40ce9798 ("nvmem: add QTI SDAM driver") Cc: stable@vger.kernel.org Signed-off-by:
Subbaraman Narayanamurthy <subbaram@codeaurora.org> Signed-off-by:
-
Nicolas Saenz Julienne authored
s/drivers/driver/ as the configuration selects a single driver. Suggested-by:
Randy Dunlap <rdunlap@infradead.org> Acked-by:
-
- Feb 04, 2021
-
-
Ahmad Fatoum authored
The nvmem cell binding applies to all eeprom child nodes matching "^.*@[0-9a-f]+$" without taking a compatible into account. Linux drivers, like at24, are even more extensive and assume _all_ at24 eeprom child nodes to be nvmem cells since e888d445 ("nvmem: resolve cells from DT at registration time"). Since df5f3b6f ("dt-bindings: nvmem: stm32: new property for data access"), the additionalProperties: True means it's Ok to have other properties as long as they don't match "^.*@[0-9a-f]+$". The barebox bootloader extends the MTD partitions binding to EEPROM and can fix up following device tree node: &eeprom { partitions { compatible = "fixed-partitions"; }; }; This is allowed binding-wise, but drivers using nvmem_register() like at24 will fail to parse because the function expects all child nodes to have a reg property present. This results in the whole EEPROM driver probe failing despite the device tree being correct. Fix this by skipping nodes lacking a reg property instead of returning an error. This effectively makes the drivers adhere to the binding because all nodes with a unit address must have a reg property and vice versa. Fixes: e888d445 ("nvmem: resolve cells from DT at registration time"). Signed-off-by:
Ahmad Fatoum <a.fatoum@pengutronix.de> Signed-off-by:
-
Nicolas Saenz Julienne authored
Firmware/co-processors might use reserved memory areas in order to pass data stemming from an nvmem device otherwise non accessible to Linux. For example an EEPROM memory only physically accessible to firmware, or data only accessible early at boot time. In order to expose this data to other drivers and user-space, the driver models the reserved memory area as an nvmem device. Tested-by:
Tim Gover <tim.gover@raspberrypi.com> Reviewed-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Fabio Estevam authored
The retrieval of driver data via of_device_get_match_data() can make the code simpler. Use of_device_get_match_data() to simplify the code. Signed-off-by:
Fabio Estevam <festevam@gmail.com> Signed-off-by:
-
Dan Carpenter authored
This doesn't call of_node_put() on the error path so it leads to a memory leak. Fixes: 0749aa25 ("nvmem: core: fix regression in of_nvmem_cell_get()") Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
- Nov 27, 2020
-
-
Peng Fan authored
When offset is not 4 bytes aligned, directly shift righty by 2 bits will cause reading out wrong data. Since imx ocotp only supports 4 bytes reading once, we need handle offset is not 4 bytes aligned and enlarge the bytes to 4 bytes aligned. After reading finished, copy the needed data from buffer to caller and free buffer. Signed-off-by:
Peng Fan <peng.fan@nxp.com> Signed-off-by:
-
Evan Green authored
Some fuse ranges are protected by the XPU such that the AP cannot access them. Attempting to do so causes an SError. Use the newly introduced per-soc compatible string, and the newly introduced nvmem keepout support to attach the set of regions we should not access. Reviewed-by:
Evan Green <evgreen@chromium.org> Signed-off-by:
-
Evan Green authored
Introduce support into the nvmem core for arrays of register ranges that should not result in actual device access. For these regions a constant byte (repeated) is returned instead on read, and writes are quietly ignored and returned as successful. This is useful for instance if certain efuse regions are protected from access by Linux because they contain secret info to another part of the system (like an integrated modem). Signed-off-by:
-
- Sep 27, 2020
-
-
Vadym Kochan authored
Fix missing 'kfree_const(cell->name)' when call to nvmem_cell_info_to_nvmem_cell() in several places: * after nvmem_cell_info_to_nvmem_cell() failed during nvmem_add_cells() * during nvmem_device_cell_{read,write} when cell->name is kstrdup'ed() without calling kfree_const() at the end, but really there is no reason to do that 'dup, because the cell instance is allocated on the stack for some short period to be read/write without exposing it to the caller. So the new nvmem_cell_info_to_nvmem_cell_nodup() helper is introduced which is used to convert cell_info -> cell without name duplication as a lighweight version of nvmem_cell_info_to_nvmem_cell(). Fixes: e2a5402e ("nvmem: Add nvmem_device based consumer apis.") Reviewed-by:Vadym Kochan <vadym.kochan@plvision.eu> Link: https://lore.kernel.org/r/20200923204456.14032-1-vadym.kochan@plvision.eu Signed-off-by:
-
- Sep 17, 2020
-
-
Vadym Kochan authored
of_parse_phandle() returns device_node with incremented ref count which needs to be decremented by of_node_put() when device_node is not used. Fixes: e2a5402e ("nvmem: Add nvmem_device based consumer apis.") Signed-off-by:
-
Bartosz Golaszewski authored
We don't need to specify any ranges when allocating IDs so we can switch to ida_alloc() and ida_free() instead of the ida_simple_ counterparts. ida_simple_get(ida, 0, 0, gfp) is equivalent to ida_alloc_range(ida, 0, UINT_MAX, gfp) which is equivalent to ida_alloc(ida, gfp). Note: IDR will never actually allocate an ID larger than INT_MAX. Signed-off-by:
Bartosz Golaszewski <bgolaszewski@baylibre.com> Signed-off-by:
-
Tian Tao authored
Use kobj_to_dev() instead of container_of() Signed-off-by:
Tian Tao <tiantao6@hisilicon.com> Signed-off-by:
-
Chih-En Hsu authored
This patch is to remove function "mtk_reg_write" since Mediatek EFUSE hardware only supports read functionality for NVMEM consumers. Fixes: ba360fd0 ("nvmem: mtk-efuse: remove nvmem regmap dependency") Acked-by:
Andrew-CT Chen <andrew-ct.chen@mediatek.com> Signed-off-by:
Chih-En Hsu <chih-en.hsu@mediatek.com> Signed-off-by:
-
- Jul 29, 2020
-
-
Guru Das Srinagesh authored
Using pdev->id as the nvmem's config ID (which, by default, is NVMEM_DEVID_NONE) prevents multiple instances of this driver from probing because of the following error: sysfs: cannot create duplicate filename '/bus/nvmem/devices/spmi_sdam' Use NVMEM_DEVID_AUTO as the NVMEM config ID to fix the issue. Signed-off-by:
Guru Das Srinagesh <gurus@codeaurora.org> Signed-off-by:
-
Matteo Croce authored
nvmem can't be built as module anymore, update its Kconfig description. Fixes: 2a37ce25 ("nvmem: disallow modular CONFIG_NVMEM") Signed-off-by:
Matteo Croce <mcroce@microsoft.com> Signed-off-by:
-
