michaelh/zephyr
1
0
Fork 0
Code Releases Activity

fs: littlefs: Rename *area (struct flash_area*) to *backend void pointer

Browse source 
The struct flash_area *area pointer has been renamed to void *backend
pointer.

This change is enabling further rework of the littlefs subsystem to work
with other backend devices (like block ones - i.e. SD card).

Signed-off-by: Lukasz Majewski <lukma@denx.de>
This commit is contained in:
Lukasz Majewski 2022-01-10 11:52:38 +01:00 • committed by Anas Nashif
commit 6fa771f2d0
2 changed files with 21 additions and 18 deletions
Download patch file Download diff file Expand all files Collapse all files

2
include/fs/littlefs.h
View file

@ -35,7 +35,7 @@ struct fs_littlefs {
/* These structures are filled automatically at mount. */
struct lfs lfs;
const struct flash_area *area;
void *backend;
struct k_mutex mutex;
};

37
subsys/fs/littlefs_fs.c
View file

@ -578,7 +578,7 @@ static int littlefs_mount(struct fs_mount_t *mountp)
return -ENOTSUP;
}
if (fs->area) {
if (fs->backend) {
return -EBUSY;
}
@ -587,19 +587,22 @@ static int littlefs_mount(struct fs_mount_t *mountp)
fs_lock(fs);
/* Open flash area */
ret = flash_area_open(area_id, &fs->area);
if ((ret < 0) || (fs->area == NULL)) {
const struct flash_area **fap = (const struct flash_area **)&fs->backend;
ret = flash_area_open(area_id, fap);
if ((ret < 0) || (*fap == NULL)) {
LOG_ERR("can't open flash area %d", area_id);
ret = -ENODEV;
goto out;
}
LOG_DBG("FS area %u at 0x%x for %u bytes",
area_id, (uint32_t)fs->area->fa_off,
(uint32_t)fs->area->fa_size);
dev = flash_area_get_device(fs->area);
LOG_DBG("FS area %u at 0x%x for %u bytes", area_id,
(uint32_t)(*fap)->fa_off, (uint32_t)(*fap)->fa_size);
dev = flash_area_get_device(*fap);
if (dev == NULL) {
LOG_ERR("can't get flash device: %s", log_strdup(fs->area->fa_dev_name));
LOG_ERR("can't get flash device: %s",
log_strdup((*fap)->fa_dev_name));
ret = -ENODEV;
goto out;
}
@ -632,7 +635,7 @@ static int littlefs_mount(struct fs_mount_t *mountp)
lfs_size_t block_size = lcp->block_size;
if (block_size == 0) {
block_size = get_block_size(fs->area);
block_size = get_block_size(*fap);
}
if (block_size == 0) {
__ASSERT_NO_MSG(block_size != 0);
@ -664,11 +667,11 @@ static int littlefs_mount(struct fs_mount_t *mountp)
/* No, you don't get to override this. */
lfs_size_t block_count = fs->area->fa_size / block_size;
lfs_size_t block_count = (*fap)->fa_size / block_size;
LOG_INF("FS at %s:0x%x is %u 0x%x-byte blocks with %u cycle",
log_strdup(dev->name), (uint32_t)fs->area->fa_off,
block_count, block_size, block_cycles);
log_strdup(dev->name), (uint32_t)(*fap)->fa_off, block_count,
block_size, block_cycles);
LOG_INF("sizes: rd %u ; pr %u ; ca %u ; la %u",
read_size, prog_size, cache_size, lookahead_size);
@ -677,7 +680,7 @@ static int littlefs_mount(struct fs_mount_t *mountp)
__ASSERT_NO_MSG(cache_size != 0);
__ASSERT_NO_MSG(block_size != 0);
__ASSERT((fs->area->fa_size % block_size) == 0,
__ASSERT(((*fap)->fa_size % block_size) == 0,
"partition size must be multiple of block size");
__ASSERT((block_size % prog_size) == 0,
"erase size must be multiple of write size");
@ -685,7 +688,7 @@ static int littlefs_mount(struct fs_mount_t *mountp)
"cache size incompatible with block size");
/* Set the validated/defaulted values. */
lcp->context = (void *)fs->area;
lcp->context = fs->backend;
lcp->read = lfs_api_read;
lcp->prog = lfs_api_prog;
lcp->erase = lfs_api_erase;
@ -728,7 +731,7 @@ static int littlefs_mount(struct fs_mount_t *mountp)
out:
if (ret < 0) {
fs->area = NULL;
fs->backend = NULL;
}
fs_unlock(fs);
@ -743,8 +746,8 @@ static int littlefs_unmount(struct fs_mount_t *mountp)
fs_lock(fs);
lfs_unmount(&fs->lfs);
flash_area_close(fs->area);
fs->area = NULL;
flash_area_close(fs->backend);
fs->backend = NULL;
fs_unlock(fs);