This is an automated email from the ASF dual-hosted git repository.
xiaoxiang pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/nuttx.git
The following commit(s) were added to refs/heads/master by this push:
new 786701c752 fs/mnemofs: Add master node and r/w methods Adds master
node and r/w methods to mnemofs
786701c752 is described below
commit 786701c75273242f6222985b25a67fbce19c22d0
Author: Saurav Pal <[email protected]>
AuthorDate: Mon Jul 15 13:51:22 2024 +0000
fs/mnemofs: Add master node and r/w methods
Adds master node and r/w methods to mnemofs
Signed-off-by: Saurav Pal <[email protected]>
---
fs/mnemofs/mnemofs.h | 44 ++++++++
fs/mnemofs/mnemofs_master.c | 249 +++++++++++++++++++++++++++++++++++++++++++-
fs/mnemofs/mnemofs_rw.c | 34 +++++-
3 files changed, 321 insertions(+), 6 deletions(-)
diff --git a/fs/mnemofs/mnemofs.h b/fs/mnemofs/mnemofs.h
index 7511d22203..df1d1040b3 100644
--- a/fs/mnemofs/mnemofs.h
+++ b/fs/mnemofs/mnemofs.h
@@ -1303,6 +1303,27 @@ void mfs_lru_updatedsz(FAR struct mfs_sb_s * const sb,
/* mnemofs_master.c */
+/****************************************************************************
+ * Name: mfs_mn_init
+ *
+ * Description:
+ * Initialize master node by reading from the flash.
+ *
+ * Input Parameters:
+ * sb - Superblock instance of the device.
+ * jrnl_blk - First block of the journal.
+ *
+ * Returned Value:
+ * 0 - OK
+ * < 0 - Error
+ *
+ * Assumptions/Limitations:
+ * The journal will have to be initialized before this.
+ *
+ ****************************************************************************/
+
+int mfs_mn_init(FAR struct mfs_sb_s * const sb, const mfs_t jrnl_blk);
+
/****************************************************************************
* Name: mfs_mn_fmt
*
@@ -1324,6 +1345,29 @@ void mfs_lru_updatedsz(FAR struct mfs_sb_s * const sb,
int mfs_mn_fmt(FAR struct mfs_sb_s * const sb, const mfs_t jrnl_blk);
+/****************************************************************************
+ * Name: mfs_mn_move
+ *
+ * Description:
+ * Move the master node.
+ *
+ * Input Parameters:
+ * sb - Superblock instance of the device.
+ * root - New location of the root of the file system.
+ * root_sz - New size of the CTZ list of the root of the file syste.
+ *
+ * Returned Value:
+ * 0 - OK
+ * < 0 - Error
+ *
+ * Assumptions/Limitations:
+ * This is called when the root is updated to a new location.
+ *
+ ****************************************************************************/
+
+int mfs_mn_move(FAR struct mfs_sb_s * const sb, struct mfs_ctz_s root,
+ const mfs_t root_sz);
+
/* mnemofs_fsobj.c */
/****************************************************************************
diff --git a/fs/mnemofs/mnemofs_master.c b/fs/mnemofs/mnemofs_master.c
index b971671b1a..3f94720aa8 100644
--- a/fs/mnemofs/mnemofs_master.c
+++ b/fs/mnemofs/mnemofs_master.c
@@ -97,6 +97,11 @@
* Private Function Prototypes
****************************************************************************/
+static FAR char *ser_mn(const struct mfs_mn_s mn,
+ FAR char * const out);
+static FAR const char *deser_mn(FAR const char * const in,
+ FAR struct mfs_mn_s *mn, FAR uint8_t *hash);
+
/****************************************************************************
* Private Data
****************************************************************************/
@@ -109,13 +114,253 @@
* Private Functions
****************************************************************************/
+/****************************************************************************
+ * Name: ser_mn
+ *
+ * Description:
+ * Serialize master node.
+ *
+ * Input Parameters:
+ * mn - Master node.
+ * out - Out buffer.
+ *
+ * Returned Value:
+ * Pointer to the end of the serialized data in `out`.
+ *
+ * Assumptions/Limitations:
+ * Out should contain enough space for `mn` and 1 byte extra for the hash.
+ *
+ ****************************************************************************/
+
+static FAR char *ser_mn(const struct mfs_mn_s mn, FAR char * const out)
+{
+ FAR char *tmp = out;
+
+ tmp = mfs_ser_mfs(mn.jrnl_blk, tmp);
+ tmp = mfs_ser_mfs(mn.mblk_idx, tmp);
+ tmp = mfs_ser_ctz(&mn.root_ctz, tmp);
+ tmp = mfs_ser_mfs(mn.root_sz, tmp);
+ tmp = mfs_ser_timespec(&mn.ts, tmp);
+ tmp = mfs_ser_8(mfs_arrhash(out, tmp - out), tmp);
+
+ return tmp;
+}
+
+/****************************************************************************
+ * Name: ser_mn
+ *
+ * Description:
+ * Deserialize master node.
+ *
+ * Input Parameters:
+ * in - In buffer.
+ * mn - Master node to populate.
+ * hash - Stored hash (of serialized data) to populate.
+ *
+ * Returned Value:
+ * Pointer to the end of the serialized data in `in`.
+ *
+ * Assumptions/Limitations:
+ * In should contain enough space for `mn` and 1 byte extra for the hash.
+ *
+ ****************************************************************************/
+
+static FAR const char *deser_mn(FAR const char * const in,
+ FAR struct mfs_mn_s *mn, FAR uint8_t *hash)
+{
+ FAR const char *tmp = in;
+
+ tmp = mfs_deser_mfs(tmp, &mn->jrnl_blk);
+ tmp = mfs_deser_mfs(tmp, &mn->mblk_idx);
+ tmp = mfs_deser_ctz(tmp, &mn->root_ctz);
+ tmp = mfs_deser_mfs(tmp, &mn->root_sz);
+ tmp = mfs_deser_timespec(tmp, &mn->ts);
+ tmp = mfs_deser_8(tmp, hash);
+
+ return tmp;
+}
+
/****************************************************************************
* Public Functions
****************************************************************************/
+int mfs_mn_init(FAR struct mfs_sb_s * const sb, const mfs_t jrnl_blk)
+{
+ int ret = OK;
+ mfs_t i = 0;
+ mfs_t mblk1;
+ mfs_t mblk2;
+ mfs_t jrnl_blk_tmp;
+ bool found = false;
+ uint8_t hash;
+ struct mfs_mn_s mn;
+ const mfs_t sz = sizeof(struct mfs_mn_s) - sizeof(mn.pg);
+ char buftmp[4];
+ char buf[sz + 1];
+
+ mblk1 = mfs_jrnl_blkidx2blk(sb, MFS_JRNL(sb).n_blks);
+ mblk2 = mfs_jrnl_blkidx2blk(sb, MFS_JRNL(sb).n_blks + 1);
+
+ mn.jrnl_blk = mn.jrnl_blk;
+ mn.mblk_idx = 0;
+ mn.pg = MFS_BLK2PG(sb, mblk1);
+
+ for (i = 0; i < MFS_PGINBLK(sb); i++)
+ {
+ mfs_read_page(sb, buftmp, 4, mn.pg, 0);
+ mfs_deser_mfs(buftmp, &jrnl_blk_tmp);
+
+ if (jrnl_blk_tmp == 0)
+ {
+ break;
+ }
+
+ if (jrnl_blk_tmp != jrnl_blk)
+ {
+ break;
+ }
+
+ found = true;
+ mn.mblk_idx++;
+ mn.pg++;
+ }
+
+ if (found == false)
+ {
+ ret = -EINVAL;
+ goto errout;
+ }
+
+ if (i == MFS_PGINBLK(sb))
+ {
+ ret = -ENOSPC;
+ goto errout;
+ }
+ else
+ {
+ mn.mblk_idx--;
+ mn.pg--;
+ }
+
+ mfs_read_page(sb, buf, sz + 1, mn.pg, 0);
+
+ /* Deserialize. */
+
+ deser_mn(buf, &mn, &hash);
+ if (hash != mfs_arrhash(buf, sz))
+ {
+ ret = -EINVAL;
+ goto errout;
+ }
+
+ /* FUTURE TODO: Recovery in case of hash not matching, or page not
+ * readable.
+ */
+
+ MFS_MN(sb) = mn;
+
+errout:
+ return ret;
+}
+
int mfs_mn_fmt(FAR struct mfs_sb_s * const sb, const mfs_t jrnl_blk)
{
- /* TODO */
+ int ret = OK;
+ mfs_t pg;
+ mfs_t mblk1;
+ mfs_t mblk2;
+ struct mfs_mn_s mn;
+ const mfs_t sz = sizeof(struct mfs_mn_s) - sizeof(mn.pg);
+ char buf[sz + 1];
+ struct timespec ts;
+
+ clock_gettime(CLOCK_REALTIME, &ts);
+
+ memset(buf, 0, sz + 1);
+
+ mblk1 = mfs_jrnl_blkidx2blk(sb, MFS_JRNL(sb).n_blks);
+ mblk2 = mfs_jrnl_blkidx2blk(sb, MFS_JRNL(sb).n_blks + 1);
+
+ pg = mfs_ba_getpg(sb);
+ if (predict_false(pg == 0))
+ {
+ ret = -ENOSPC;
+ goto errout;
+ }
+
+ finfo("Root formatted to be at Page %u", pg);
+
+ mn.root_ctz.idx_e = 0;
+ mn.root_ctz.pg_e = pg;
+ mn.jrnl_blk = jrnl_blk;
+ mn.mblk_idx = 0;
+ mn.pg = MFS_BLK2PG(sb, mblk1);
+ mn.root_sz = 0;
+ mn.ts = ts;
+ mn.root_st_atim = ts;
+ mn.root_st_ctim = ts;
+ mn.root_st_mtim = ts;
+ mn.root_mode = 0777 | S_IFDIR;
+
+ /* Serialize. */
+
+ ser_mn(mn, buf);
+
+ ret = mfs_write_page(sb, buf, sz, MFS_BLK2PG(sb, mblk1), 0);
+ if (predict_false(ret < 0))
+ {
+ goto errout;
+ }
+
+ ret = mfs_write_page(sb, buf, sz, MFS_BLK2PG(sb, mblk2), 0);
+ if (predict_false(ret < 0))
+ {
+ goto errout;
+ }
+
+ MFS_MN(sb) = mn;
+ finfo("Master node written. Now at page %d, timestamp %lld.%.9ld.",
+ MFS_MN(sb).pg, (long long)MFS_MN(sb).ts.tv_sec,
+ MFS_MN(sb).ts.tv_nsec);
+
+errout:
+ return ret;
+}
+
+int mfs_mn_move(FAR struct mfs_sb_s * const sb, struct mfs_ctz_s root,
+ const mfs_t root_sz)
+{
+ int ret = OK;
+ mfs_t mblk1;
+ mfs_t mblk2;
+ struct mfs_mn_s mn;
+ const mfs_t sz = sizeof(struct mfs_mn_s) - sizeof(mn.pg);
+ char buf[sz + 1];
+
+ if (MFS_MN(sb).mblk_idx == MFS_PGINBLK(sb) - 1)
+ {
+ /* TODO: Move journal. Master blocks are full. */
+ }
+
+ mblk1 = mfs_jrnl_blkidx2blk(sb, MFS_JRNL(sb).n_blks);
+ mblk2 = mfs_jrnl_blkidx2blk(sb, MFS_JRNL(sb).n_blks + 1);
+ mn = MFS_MN(sb);
+
+ mn.root_ctz = root;
+ mn.root_sz = root_sz;
+ mn.mblk_idx++;
+ mn.pg++;
+
+ ser_mn(mn, buf);
+
+ ret = mfs_write_page(sb, buf, sz + 1, mn.pg, 0);
+ if (predict_false(ret < 0))
+ {
+ goto errout;
+ }
+
+ MFS_MN(sb) = mn;
- return OK;
+errout:
+ return ret;
}
diff --git a/fs/mnemofs/mnemofs_rw.c b/fs/mnemofs/mnemofs_rw.c
index 306051d5b2..4a01e8aec9 100644
--- a/fs/mnemofs/mnemofs_rw.c
+++ b/fs/mnemofs/mnemofs_rw.c
@@ -93,22 +93,48 @@ int mfs_markbadblk(FAR const struct mfs_sb_s * const sb,
mfs_t blk)
return MTD_ISBAD(MFS_MTD(sb), blk);
}
+/* NOTE: These functions do not update the block allocator's state nor do
+ * they enforce it.
+ */
+
ssize_t mfs_write_page(FAR const struct mfs_sb_s * const sb,
FAR const char *data, const mfs_t datalen,
const off_t page, const mfs_t pgoff)
{
- /* TODO */
+ int ret = OK;
+
+ mempcpy(MFS_RWBUF(sb) + pgoff, data, datalen);
+
+ ret = MTD_BWRITE(MFS_MTD(sb), page, 1, MFS_RWBUF(sb));
+ if (ret < 0)
+ {
+ goto errout_with_reset;
+ }
+
+errout_with_reset:
+ memset(MFS_RWBUF(sb), 0, MFS_PGSZ(sb));
- return OK;
+ return ret;
}
ssize_t mfs_read_page(FAR const struct mfs_sb_s * const sb,
FAR char *data, const mfs_t datalen, const off_t page,
const mfs_t pgoff)
{
- /* TODO */
+ int ret = OK;
+
+ ret = MTD_BREAD(MFS_MTD(sb), page, 1, MFS_RWBUF(sb));
+ if (ret < 0)
+ {
+ goto errout_with_reset;
+ }
+
+ memcpy(data, MFS_RWBUF(sb) + pgoff, datalen);
+
+errout_with_reset:
+ memset(MFS_RWBUF(sb), 0, MFS_PGSZ(sb));
- return OK;
+ return ret;
}
int mfs_erase_blk(FAR const struct mfs_sb_s * const sb, const off_t blk)
