Remove CONFIG_SDMMC_VOLUME_NAME, and set the disk name based on the
``disk-name`` property. This aligns with other disk drivers, and allows
for multiple instances of the sdmmc_subsys disk driver to be registered.
Add disk-name properties for all in tree definitions for the
sdmmc-subsys disk driver, and change all in tree usage of the disk name
Fixes#75004
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
This resolves some addressed comments in this PR
https://github.com/zephyrproject-rtos/zephyr/pull/77930
It adds as well a section in the documentation about some
recommendations to increase ZMS performance.
Signed-off-by: Riadh Ghaddab <rghaddab@baylibre.com>
The cache_size and lookahead_size are set at compile time using the
CONFIG_FS_LITTLEFS_CACHE_SIZE and CONFIG_FS_LITTLEFS_LOOKAHEAD_SIZE values
from Kconfig, or from the cache-size and lookahead-size properties in a
'zephyr,fstab,littlefs' compatible in the devicetree. Those values are
also used to statically allocate buffers that are pointed at by the
read_buffer, prog_buffer, and lookahead_buffer members of the lfs_config
structure.
At runtime, when using a block device, the cache_size and lookahead_size
are updated to be multiples of the underlying block device's block_size,
which may make them bigger than the original size used to allocate the
static buffers. Log an error and fail the operation when this occurs.
Signed-off-by: Phil Hindman <phindman@xes-inc.com>
ZMS is the abreviation of Zephyr Memory Storage.
It is a storage developed to target especially the non erasable devices.
The new memory storage system inherit from the NVS storage multiple
features and introduce new ones :
* Inherited features :
- light key-value based storage
- cache for entries
- Wear Leveling of flash memory
- Resilience to power failures
* New features :
- cycle counter for non erasable devices (instead of erase emulation)
- Keys up to 32-bit
- Built-in support of CRC32 for data
- Small size data (<= 8 bytes) integrated within entries
Signed-off-by: Riadh Ghaddab <rghaddab@baylibre.com>
Clean up if the mount command fails, such as due to a typo in the mount
point name, so that it can be retried.
Signed-off-by: Phil Hindman <phindman@xes-inc.com>
Streamline the `rc` return logic in `nvs_flash_wrt_entry()`, `nvs_gc()`
by directly returning `rc`, as the redundant conditional check and
explicit return of 0 were unnecessary.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Property "block-cycles" is required for node "zephyr,fstab,littlefs",
but source code did not get the value from dts file, now add it.
Additionally correct the wrong description of property "block-cycles"
in binding file.
Signed-off-by: Paul He <pawpawhe@gmail.com>
Add an option that links in the underlying file system libraries as
usual, but doesn't compile in the Zephyr file system abstraction layer.
This can be useful to avoid linking in the entire filesystem
implementation through the `fs_file_system_t` API struct if only a
subset of the functions are used.
Signed-off-by: Jordan Yates <jordan@embeint.com>
Reading superblock didn't work with DMAs that only support aligned reads
(e.g. STM32 DMA).
Co-authored-by: Maciej Sobkowski <msobkowski@antmicro.com>
Signed-off-by: Dominik Lau <dlau@internships.antmicro.com>
There is asignment from size_t variable to uint8_t variable in
nvs_al_size function that can trigger static analysis error.
The problem is fixed by changing the variable to size_t type
instead of uint8_t.
Signed-off-by: Andrej Trnkóci <andrej.trnkoci@nxp.com>
This option switches support for multiple volumes on the physical drive. By
default (0), each logical drive number is bound to the same physical drive
number and only a first FAT volume found on the physical drive will be
mounted.
When this function is enabled (1), each logical drive number can be bound
to arbitrary physical drive and partition listed in the VolToPart[].
The VolToPart[] is expected to be provided by Zephyr application.
For example, 2 FAT partition on SD disk ("SD" index 3) in terms of Zephyr:
{3, 1} - mount point "/0:"
{3, 2} - mount point "/1:"
The mount points have to be numbered in this case.
Code example of mounting a second FATFS partition places on SD-card:
static FATFS fat_fs;
static struct fs_mount_t mp = {
.type = FS_FATFS,
.fs_data = &fat_fs,
.mnt_point = "/1:
};
/*
* 2 FAT partition on SD disk
* PARTITION.pd - Physical drive number. "SD" has index 3 in terms of
* Zephyr (see FF_VOLUME_STRS)
* PARTITION.pt - Partition (0:Auto detect, 1-4:Forced partition). So 1 for
* the first FATFS partition and 2 - for second.
*/
PARTITION VolToPart[FF_VOLUMES] = {
[0] = {3, 1}, /* "0:" ==> 1st partition on the pd#0 */
[1] = {3, 2}, /* "1:" ==> 2nd partition on the pd#0 */
[2] = {0, -1}, /* "2:" ==> 3rd partition on the pd#0 */
[3] = {0, -1},
[4] = {0, -1},
[5] = {0, -1},
[6] = {0, -1},
[7] = {0, -1},
};
fs_mount(&mp);
Signed-off-by: Grygorii Strashko <grygorii_strashko@epam.com>
The Kconfig option enables support for 64-bit LBA, which also allows
to enable GUID Partition Table (GPT) support.
Signed-off-by: Grygorii Strashko <grygorii_strashko@epam.com>
Add missing braces to comply with MISRA C:2012 Rule 15.6 and
also following Zephyr's style guideline.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
set _XOPEN_SOURCE appropriately to avoid compilation errors
due to missing S_IFxxx macros on some systems.
Signed-off-by: Benjamin Cabé <benjamin@zephyrproject.org>
Add an API function allowing the application to determine the amount of
free bytes in the current sector.
Add a second API function allowing the application to switch to the next
NVS sector, calling the garbage collector on such sector.
The goal is togive more granularity to the application to control when
the garbage collector is triggered.
Signed-off-by: Adrien Ricciardi <aricciardi@baylibre.com>
Even just making minor edits to fcb, warnings are triggered
from Coding Guidelines of the form
Violation to rule 5.7 (Tag name should be unique) tag: fcb
Rename pointer variable names to fcbp.
Signed-off-by: Chris Friedt <cfriedt@tenstorrent.com>
Make sure that the ATE metadata does not overflow the sector size.
Take into account the data length and also the mandatory reserved ATEs
in each sector.
Signed-off-by: Adrien Ricciardi <aricciardi@baylibre.com>
The nvs_calc_free_space() function does not return 0 when the NVS
is considered full, because some special ATEs are not taken into account.
This commit takes into account the ATE that is reserved for deletion in
each sector, in addition of the 'GC done' ATE when present.
Signed-off-by: Adrien Ricciardi <aricciardi@baylibre.com>
If filesystem relies on zfp flags being set after open, this can lead to
inconsistency with the current implementation of FS_O_TRUNC.
Move the assignment of zfp->flags before any other call to fs api to avoid
such issues.
Signed-off-by: Pierrick Guillaume <pierguill@gmail.com>
Allow to protect the data part of each NVS item with a 32-bit CRC.
This uses 4 more bytes per NVS item.
Signed-off-by: Adrien Ricciardi <aricciardi@baylibre.com>
Make call to de-initialize disk in fatfs_unmount(). This will permit the
disk to be reinitialized when it is mounted with fatfs_mount().
Signed-off-by: Daniel DeGrasse <daniel.degrasse@nxp.com>
Introduced temporary variables `buf0_xor` and `buf1_xor` to store
XOR results and added an explicit cast to `uint16_t` for
the `*len` variable.
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
Correct the boundary condition in `fcb_put_len` function to
properly include `FCB_MAX_LEN` and change the #define to address
the potential flaw where `CHAR_MAX` might be treated as unsigned by
the compiler flag `-funsigned-char`, which would yield `FCB_MAX_LEN`
to 0x7fff instead of 0x3fff.
Fixes#73868
Signed-off-by: Pisit Sawangvonganan <pisit@ndrsolution.com>
The NVS currently requires explict erase capability of
a device to work, so usage of flash_erase has been replaced
with flash_flatten.
There has been additional LOG_WRN added to warn user that
NVS may not efficiently work with device that do not really
have erase.
Currently NVS relies on devices that require erase.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
LittleFS is based on program-erase devices and require that
characteristics to work; the commit replaces flash_area_erase
with flash_area_flatten to allow LFS work on devices that
do not provide erase callback.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
This commit has added new flag FS_O_TRUNC to support truncation
during file open. Modified fs_open to handle truncation based on
provided flags. Included unit tests for flag behavior with common
filesystems.
Signed-off-by: RAJAGOPALAN GANGADHARAN <g.raju2000@gmail.com>
This commit fixes the incorrect include path for the fuse library.
pkg_search_module populates the <XXX>_INCLUDE_DIRS cmake var.
Signed-off-by: Russ Webber <russ@rw.id.au>
Spell checking tools do not recognize "iff", replace with "if and only if".
See https://en.wikipedia.org/wiki/If_and_only_if
Signed-off-by: Pieter De Gendt <pieter.degendt@basalte.be>
This commit fixes removing indirect blocks (marking them as 0)
in the inode structure. Previous version of the code was removing
the top-level blocks only when the first removed block was
one of the first 12 direct blocks. However, when the first removed
block is the first block in the referenced block list, its parent
(indirect block) should also be removed.
Signed-off-by: Franciszek Pindel <fpindel@antmicro.com>
First sector starts at CONFIG_EXT2_DISK_STARTING_SECTOR.
This commit fixes calculating free space, based on that value.
Signed-off-by: Franciszek Pindel <fpindel@antmicro.com>
The commit adds dependency on Kconfig FLASH_PAGE_LAYOUT to subsystems
that really require it:
FCB, NVS, LittleFS
and removes direct selection from '*.conf' files where no longer
needed.
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
Users should be able to call fs_mkfs() manually even if FS_FATFS_MOUNT_MKFS
is disabled.
Signed-off-by: Armin Brauns <armin.brauns@embedded-solutions.at>
Replace the use of a memcmp of a temp buffer filled with
the comparison value with a direct comparison of
the read buffer contents.
Improves speed of the function (no more memset() call)
and reduces stack usage.
Signed-off-by: Mike J. Chen <mjchen@google.com>
Add warning that file opened without R/W flags will have no read/write
access.
Remove suggestion for using fs_open to check if file exists.
Clarify -ENOENT return reason.
Fixes#64030
Signed-off-by: Dominik Ermel <dominik.ermel@nordicsemi.no>
Non-configurable init priority POST_KERNEL/99 might
be too late for certain uses cases (e.g. fetching data from
the file system in driver initialization). Chabge it to
be configurable so that applications can mount the
file systems earlies if they want.
Signed-off-by: Miika Karanki <miika.karanki@vaisala.com>
fs_init was just initializing mutex and dlist. Those can be
done statically. Not having the SYS_INIT function helps
in ordering the init functions in case e.g. file system
mounts are wanted to happen early stage on the system
startup.
Signed-off-by: Miika Karanki <miika.karanki@vaisala.com>
For native_sim we need to have the fuse library linked with the
native simulator runner, not with the embedded code.
Let's fix it.
Signed-off-by: Alberto Escolar Piedras <alberto.escolar.piedras@nordicsemi.no>
rand32.h does not make much sense, since the random subsystem
provides more APIs than just getting a random 32 bits value.
Rename it to random.h and get consistently with other
subsystems.
Signed-off-by: Flavio Ceolin <flavio.ceolin@intel.com>
NVS lookup cache currently uses CRC8/16 as a hash function
to determine the cache position, which is not ideal choice.
For example, when NVS lookup cache size is 512 and 256
subsequent NVS IDs are written (that is, 0, 1.., 255), this
results in 128 cache collisions.
It is better to use a dedicated integer hash function. This
PR uses one of the 16-bit integer hash functions discovered
with https://github.com/skeeto/hash-prospector project. The
hash function was additionally tested in the context of NVS
lookup cache using simple NVS ID allocation patterns as well
as using real device NVS dump.
Also, add a test case to verify that the hash function is
not extremely bad.
Signed-off-by: Damian Krolik <damian.krolik@nordicsemi.no>
Modify the signature of the k_mem_slab_free() function with a new one,
replacing the old void **mem with void *mem as a parameter.
The following function:
void k_mem_slab_free(struct k_mem_slab *slab, void **mem);
has the wrong signature. mem is only used as a regular pointer, so there
is no need to use a double-pointer. The correct signature should be:
void k_mem_slab_free(struct k_mem_slab *slab, void *mem);
The issue with the current signature, although functional, is that it is
extremely confusing. I myself, a veteran Zephyr developer, was confused
by this parameter when looking at it recently.
All in-tree uses of the function have been adapted.
Fixes#61888.
Signed-off-by: Carles Cufi <carles.cufi@nordicsemi.no>
