Aksal/nuttx
3
1
Fork 1
mirror of https://github.com/apache/nuttx.git synced 2026-08-01 20:28:58 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2
nuttx/fs/mount/fs_mount.c
Marco Casaroli e0cc0244c5 fs/xipfs: Add a contiguous execute-in-place file system. 
ROMFS is the usual way to carry executables on a NOMMU target with memory
mapped NOR flash: it can hand out a real flash pointer from mmap(), so the
NXFLAT loader maps a module's text in place instead of copying it into RAM.
But a ROMFS image is built on the host and is read only, so a module cannot
be downloaded onto the board at run time.

xipfs is a writable file system with the same in-place property.  Each file
is stored as one physically contiguous, erase-block aligned extent, so an
mmap() of it resolves to flash_base + extent_offset and a loader can execute
the file where it already lies.  This needs the underlying MTD driver to
answer BIOC_XIPBASE; on the RP2350 rp23xx_flash_mtd.c does.

Files are write once.  A file is created, its size is declared, it is
written sequentially, closed, and is thereafter immutable until it is
deleted.  That is the whole life cycle of a downloaded module, and it is
what licenses the design: the exact extent is reserved at create time, so
no file ever grows, moves, or fragments internally.  Random writes, appends
and truncation of a written file are not supported and are refused.

The only source of fragmentation is therefore free space holes left by
deletes.  Allocation fails with -ENOSPC when no single contiguous run is
large enough, and never defragments on its own; the caller decides whether
to compact and retry, through XIPFSIOC_DEFRAG.  Defragmentation is manual,
best effort and interruptible: it is a loop of atomic single-extent
relocations, each one copy, commit, erase, so every stop point -- a time
budget, a pinned extent, an erase error -- leaves a consistent layout that
is simply less compact.  It reports the largest contiguous run it achieved,
which is what tells the caller whether the retry will fit.

Metadata is committed power safely.  Two metadata block sets are used in
ping-pong, each generation carrying a sequence number and a CRC, and every
state change is ordered as write the new data, flip the metadata reference,
then erase what the old one referenced.  Mount scans both sets and selects
the last fully valid generation, so a torn write costs the interrupted
operation and nothing else.

A mapping takes a pin on the extent, and the pin lives on the extent rather
than on the file descriptor, so three running instances of one module hold
three pins and the extent becomes movable only when the last one goes.
Defragmentation skips pinned extents, which is what stops it relocating
code that is executing.  The pin is released by munmap() or by the task
teardown walk, so a task that dies without unmapping does not leak it.

Directories are records in that same generation, carrying their own identity
and the identity of the directory holding them; the root is implicit and owns
identity zero.  They are deliberately NOT objects in the data region, which
is what keeps the commit story in one piece: mkdir and rmdir add or remove a
record and commit one generation, exactly as create and unlink do, so there
is never a multi-object update to journal or an orphan to collect at mount.
An empty directory therefore exists, survives a remount, and costs one entry
out of the volume's fixed supply and no flash blocks at all.

A name is one path component; depth comes from the parent, so XIPFS_NAME_MAX
bounds a component, which is what statfs reports it as.  Mount rebuilds the
tree and checks that it is one: identities unique, names unique within a
directory, every parent a live directory, and following parents reaching the
root -- a cycle on the medium would otherwise hang a path walk rather than
merely answering wrongly.  '.' and '..' are refused as components, since an
entry stored under either could never be reached again.

The commands that act on the volume rather than on one file --
XIPFSIOC_DEFRAG and XIPFSIOC_LISTPINNED -- are reached through the ioctldir
method, on a descriptor for the mountpoint directory.  They are accepted on
a descriptor for a file inside the volume too, but that route holds the file
open for the duration and an open extent cannot be relocated, so a pass
asked for that way is obstructed by the act of asking.

mmap() falls back to the generic RAM copy for ordinary readers when the
media cannot be addressed directly.  A module loader must not silently get
a RAM copy, so MAP_XIP_STRICT is added: with it the mapping either resolves
in place or fails with -ENXIO, which the caller can turn into defragment
and retry.

Assisted-by: Claude Code:claude-opus-5
Signed-off-by: Marco Casaroli <marco.casaroli@gmail.com>
2026-07-29 07:49:03 -03:00

592 lines
16 KiB
C
Raw Permalink Blame History

/****************************************************************************
* fs/mount/fs_mount.c
*
* SPDX-License-Identifier: Apache-2.0
*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership. The
* ASF licenses this file to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance with the
* License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS, WITHOUT
* WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the
* License for the specific language governing permissions and limitations
* under the License.
*
****************************************************************************/
/****************************************************************************
* Included Files
****************************************************************************/
#include <nuttx/config.h>
#include <sys/mount.h>
#include <stdbool.h>
#include <string.h>
#include <errno.h>
#include <assert.h>
#include <nuttx/debug.h>
#include <nuttx/fs/fs.h>
#include "driver/driver.h"
#include "inode/inode.h"
#include "vfs/vfs.h"
/****************************************************************************
* Pre-processor Definitions
****************************************************************************/
/* Configuration ************************************************************/
/* In the canonical case, a file system is bound to a block driver. However,
* some less typical cases a block driver is not required. Examples are
* pseudo file systems (like BINFS or PROCFS) and MTD file systems (like
* NXFFS).
*
* These file systems all require block drivers:
*/
#if defined(CONFIG_FS_FAT) || defined(CONFIG_FS_ROMFS) || \
defined(CONFIG_FS_SMARTFS) || defined(CONFIG_FS_LITTLEFS)
# define BDFS_SUPPORT 1
#endif
/* These file systems require MTD drivers */
#if (defined(CONFIG_FS_SPIFFS) || defined(CONFIG_FS_LITTLEFS) || \
defined(CONFIG_FS_MNEMOFS) || defined(CONFIG_FS_XIPFS)) && \
defined(CONFIG_MTD)
# define MDFS_SUPPORT 1
#endif
/* These file systems do not require block or MTD drivers */
#if defined(CONFIG_FS_NXFFS) || defined(CONFIG_FS_BINFS) || \
defined(CONFIG_FS_PROCFS) || defined(CONFIG_NFS) || \
defined(CONFIG_FS_TMPFS) || defined(CONFIG_FS_USERFS) || \
defined(CONFIG_FS_CROMFS) || defined(CONFIG_FS_UNIONFS) || \
defined(CONFIG_FS_HOSTFS) || defined(CONFIG_FS_ZIPFS) || \
defined(CONFIG_FS_RPMSGFS) || defined(CONFIG_FS_V9FS)
# define NODFS_SUPPORT
#endif
/****************************************************************************
* Private Types
****************************************************************************/
struct fsmap_t
{
FAR const char *fs_filesystemtype;
FAR const struct mountpt_operations *fs_mops;
};
/****************************************************************************
* Private Data
****************************************************************************/
#ifdef BDFS_SUPPORT
/* File systems that require block drivers */
#ifdef CONFIG_FS_FAT
extern const struct mountpt_operations g_fat_operations;
#endif
#ifdef CONFIG_FS_ROMFS
extern const struct mountpt_operations g_romfs_operations;
#endif
#ifdef CONFIG_FS_SMARTFS
extern const struct mountpt_operations g_smartfs_operations;
#endif
#ifdef CONFIG_FS_LITTLEFS
extern const struct mountpt_operations g_littlefs_operations;
#endif
static const struct fsmap_t g_bdfsmap[] =
{
#ifdef CONFIG_FS_FAT
{ "vfat", &g_fat_operations },
#endif
#ifdef CONFIG_FS_ROMFS
{ "romfs", &g_romfs_operations },
#endif
#ifdef CONFIG_FS_SMARTFS
{ "smartfs", &g_smartfs_operations },
#endif
#ifdef CONFIG_FS_LITTLEFS
{ "littlefs", &g_littlefs_operations },
#endif
{ NULL, NULL },
};
#endif /* BDFS_SUPPORT */
#ifdef MDFS_SUPPORT
/* File systems that require MTD drivers */
#ifdef CONFIG_FS_SPIFFS
extern const struct mountpt_operations g_spiffs_operations;
#endif
#ifdef CONFIG_FS_LITTLEFS
extern const struct mountpt_operations g_littlefs_operations;
#endif
#ifdef CONFIG_FS_MNEMOFS
extern const struct mountpt_operations g_mnemofs_operations;
#endif
#ifdef CONFIG_FS_XIPFS
extern const struct mountpt_operations g_xipfs_operations;
#endif
static const struct fsmap_t g_mdfsmap[] =
{
#ifdef CONFIG_FS_SPIFFS
{ "spiffs", &g_spiffs_operations },
#endif
#ifdef CONFIG_FS_LITTLEFS
{ "littlefs", &g_littlefs_operations },
#endif
#ifdef CONFIG_FS_MNEMOFS
{ "mnemofs", &g_mnemofs_operations },
#endif
#ifdef CONFIG_FS_XIPFS
{ "xipfs", &g_xipfs_operations },
#endif
{ NULL, NULL },
};
#endif /* MDFS_SUPPORT */
#ifdef NODFS_SUPPORT
/* File systems that require neither block nor MTD drivers */
#ifdef CONFIG_FS_NXFFS
extern const struct mountpt_operations g_nxffs_operations;
#endif
#ifdef CONFIG_FS_TMPFS
extern const struct mountpt_operations g_tmpfs_operations;
#endif
#ifdef CONFIG_NFS
extern const struct mountpt_operations g_nfs_operations;
#endif
#ifdef CONFIG_FS_BINFS
extern const struct mountpt_operations g_binfs_operations;
#endif
#ifdef CONFIG_FS_PROCFS
extern const struct mountpt_operations g_procfs_operations;
#endif
#ifdef CONFIG_FS_USERFS
extern const struct mountpt_operations g_userfs_operations;
#endif
#ifdef CONFIG_FS_HOSTFS
extern const struct mountpt_operations g_hostfs_operations;
#endif
#ifdef CONFIG_FS_CROMFS
extern const struct mountpt_operations g_cromfs_operations;
#endif
#ifdef CONFIG_FS_UNIONFS
extern const struct mountpt_operations g_unionfs_operations;
#endif
#ifdef CONFIG_FS_RPMSGFS
extern const struct mountpt_operations g_rpmsgfs_operations;
#endif
#ifdef CONFIG_FS_ZIPFS
extern const struct mountpt_operations g_zipfs_operations;
#endif
#ifdef CONFIG_FS_V9FS
extern const struct mountpt_operations g_v9fs_operations;
#endif
static const struct fsmap_t g_nonbdfsmap[] =
{
#ifdef CONFIG_FS_NXFFS
{ "nxffs", &g_nxffs_operations },
#endif
#ifdef CONFIG_FS_TMPFS
{ "tmpfs", &g_tmpfs_operations },
#endif
#ifdef CONFIG_NFS
{ "nfs", &g_nfs_operations },
#endif
#ifdef CONFIG_FS_BINFS
{ "binfs", &g_binfs_operations },
#endif
#ifdef CONFIG_FS_PROCFS
{ "procfs", &g_procfs_operations },
#endif
#ifdef CONFIG_FS_USERFS
{ "userfs", &g_userfs_operations },
#endif
#ifdef CONFIG_FS_HOSTFS
{ "hostfs", &g_hostfs_operations },
#endif
#ifdef CONFIG_FS_CROMFS
{ "cromfs", &g_cromfs_operations },
#endif
#ifdef CONFIG_FS_UNIONFS
{ "unionfs", &g_unionfs_operations },
#endif
#ifdef CONFIG_FS_RPMSGFS
{ "rpmsgfs", &g_rpmsgfs_operations },
#endif
#ifdef CONFIG_FS_ZIPFS
{ "zipfs", &g_zipfs_operations},
#endif
#ifdef CONFIG_FS_V9FS
{ "v9fs", &g_v9fs_operations},
#endif
{ NULL, NULL },
};
#endif /* NODFS_SUPPORT */
/****************************************************************************
* Private Functions
****************************************************************************/
/****************************************************************************
* Name: mount_findfs
*
* Description:
* find the specified filesystem
*
****************************************************************************/
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT) || defined(NODFS_SUPPORT)
static FAR const struct mountpt_operations *
mount_findfs(FAR const struct fsmap_t *fstab, FAR const char *filesystemtype)
{
FAR const struct fsmap_t *fsmap;
for (fsmap = fstab; fsmap->fs_filesystemtype; fsmap++)
{
if (strcmp(filesystemtype, fsmap->fs_filesystemtype) == 0)
{
return fsmap->fs_mops;
}
}
return NULL;
}
#endif
/****************************************************************************
* Public Functions
****************************************************************************/
/****************************************************************************
* Name: nx_mount
*
* Description:
* nx_mount() is similar to the standard 'mount' interface except that is
* not a cancellation point and it does not modify the errno variable.
*
* nx_mount() is an internal NuttX interface and should not be called from
* applications.
*
* Returned Value:
* Zero is returned on success; a negated value is returned on any failure.
*
****************************************************************************/
int nx_mount(FAR const char *source, FAR const char *target,
FAR const char *filesystemtype, unsigned long mountflags,
FAR const void *data)
{
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT) || defined(NODFS_SUPPORT)
FAR struct inode *drvr_inode = NULL;
FAR struct inode *mountpt_inode = NULL;
FAR const struct mountpt_operations *mops = NULL;
#ifndef CONFIG_DISABLE_PSEUDOFS_OPERATIONS
struct inode_search_s desc;
#endif
FAR void *fshandle = NULL;
int ret;
/* Verify required pointer arguments */
DEBUGASSERT(target && filesystemtype);
/* Find the specified filesystem. Try the block driver filesystems first */
if (source != NULL && source[0] != '\0' &&
find_blockdriver(source, mountflags, &drvr_inode) >= 0)
{
/* Find the block based file system */
#ifdef BDFS_SUPPORT
mops = mount_findfs(g_bdfsmap, filesystemtype);
#endif /* BDFS_SUPPORT */
if (mops == NULL)
{
ferr("ERROR: Failed to find block based file system %s\n",
filesystemtype);
ret = -ENODEV;
goto errout_with_inode;
}
}
else if (source != NULL && source[0] != '\0' &&
(ret = find_mtddriver(source, &drvr_inode)) >= 0)
{
/* Find the MTD based file system */
#ifdef MDFS_SUPPORT
mops = mount_findfs(g_mdfsmap, filesystemtype);
#endif /* MDFS_SUPPORT */
if (mops == NULL)
{
#ifdef BDFS_SUPPORT
mops = mount_findfs(g_bdfsmap, filesystemtype);
#endif /* BDFS_SUPPORT */
if (mops == NULL)
{
ferr("ERROR: Failed to find MTD based file system %s\n",
filesystemtype);
ret = -ENODEV;
goto errout_with_inode;
}
#ifdef CONFIG_MTD
else
{
inode_release(drvr_inode);
ret = mtd_proxy(source, mountflags, &drvr_inode);
if (ret < 0)
{
goto errout_with_inode;
}
}
#endif
}
}
else
#ifdef NODFS_SUPPORT
if ((mops = mount_findfs(g_nonbdfsmap, filesystemtype)) != NULL)
{
finfo("found %s\n", filesystemtype);
}
else
#endif /* NODFS_SUPPORT */
{
ferr("ERROR: Failed to find block driver %s\n", source);
ret = -ENOTBLK;
goto errout;
}
inode_lock();
#ifndef CONFIG_DISABLE_PSEUDOFS_OPERATIONS
/* Check if the inode already exists */
SETUP_SEARCH(&desc, target, false);
ret = inode_find(&desc);
if (ret >= 0)
{
/* Successfully found. The reference count on the inode has been
* incremented.
*/
mountpt_inode = desc.node;
DEBUGASSERT(mountpt_inode != NULL);
/* But is it a directory node (i.e., not a driver or other special
* node)?
*/
if (!INODE_IS_PSEUDODIR(mountpt_inode))
{
ferr("ERROR: target %s exists and is a special node\n", target);
ret = -ENOTDIR;
inode_release(mountpt_inode);
goto errout_with_lock;
}
}
#endif
/* Bind the block driver to an instance of the file system. The file
* system returns a reference to some opaque, fs-dependent structure
* that encapsulates this binding.
*/
if (mops->bind == NULL)
{
/* The filesystem does not support the bind operation ??? */
ferr("ERROR: Filesystem does not support bind\n");
ret = -EINVAL;
goto errout_with_lock;
}
/* Increment reference count for the reference we pass to the file system */
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT)
#ifdef NODFS_SUPPORT
if (drvr_inode != NULL)
#endif
{
atomic_fetch_add(&drvr_inode->i_crefs, 1);
}
#endif
inode_unlock();
/* On failure, the bind method returns -errorcode */
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT)
ret = mops->bind(drvr_inode, data, &fshandle);
#else
ret = mops->bind(NULL, data, &fshandle);
#endif
inode_lock();
if (ret < 0)
{
/* The inode is unhappy with the driver for some reason. Back out
* the count for the reference we failed to pass and exit with an
* error.
*/
ferr("ERROR: Bind method failed: %d\n", ret);
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT)
#ifdef NODFS_SUPPORT
if (drvr_inode != NULL)
#endif
{
atomic_fetch_sub(&drvr_inode->i_crefs, 1);
}
#endif
goto errout_with_lock;
}
/* Insert a dummy node -- we need to hold the inode semaphore
* to do this because we will have a momentarily bad structure.
* NOTE that the new inode will be created with an initial reference
* count of zero.
*/
if (mountpt_inode == NULL)
{
ret = inode_reserve(target, 0777, &mountpt_inode);
if (ret < 0)
{
/* inode_reserve can fail for a couple of reasons, but the most
* likely one is that the inode already exists. inode_reserve may
* return:
*
* -EINVAL - 'path' is invalid for this operation
* -EEXIST - An inode already exists at 'path'
* -ENOMEM - Failed to allocate in-memory resources for the
* operation
*/
ferr("ERROR: Failed to reserve inode for target %s\n", target);
goto errout_with_bind;
}
}
/* We have it, now populate it with driver specific information. */
INODE_SET_MOUNTPT(mountpt_inode);
mountpt_inode->u.i_mops = mops;
mountpt_inode->i_private = fshandle;
inode_unlock();
/* We can release our reference to the blkdrver_inode, if the filesystem
* wants to retain the blockdriver inode (which it should), then it must
* have called inode_addref(). There is one reference on mountpt_inode
* that will persist until umount2() is called.
*/
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT)
#ifdef NODFS_SUPPORT
if (drvr_inode != NULL)
#endif
{
inode_release(drvr_inode);
}
#endif
#ifndef CONFIG_DISABLE_PSEUDOFS_OPERATIONS
RELEASE_SEARCH(&desc);
#endif
#ifdef CONFIG_FS_NOTIFY
notify_create(target);
#endif
return OK;
/* A lot of goto's! But they make the error handling much simpler */
errout_with_bind:
if (mops->unbind != NULL)
{
mops->unbind(fshandle, &drvr_inode, 0);
}
errout_with_lock:
inode_unlock();
#ifndef CONFIG_DISABLE_PSEUDOFS_OPERATIONS
RELEASE_SEARCH(&desc);
#endif
errout_with_inode:
#if defined(BDFS_SUPPORT) || defined(MDFS_SUPPORT)
if (drvr_inode != NULL)
{
inode_release(drvr_inode);
}
#endif
errout:
return ret;
#else
ferr("ERROR: No filesystems enabled\n");
return -ENOSYS;
#endif /* BDFS_SUPPORT || MDFS_SUPPORT || NODFS_SUPPORT */
}
/****************************************************************************
* Name: mount
*
* Description:
* mount() attaches the filesystem specified by the 'source' block device
* name into the root file system at the path specified by 'target.'
*
* Returned Value:
* Zero is returned on success; -1 is returned on an error and errno is
* set appropriately:
*
* EACCES A component of a path was not searchable or mounting a read-only
* filesystem was attempted without giving the MS_RDONLY flag.
* EBUSY 'source' is already mounted.
* EFAULT One of the pointer arguments points outside the user address
* space.
* EINVAL 'source' had an invalid superblock.
* ENODEV 'filesystemtype' not configured
* ENOENT A pathname was empty or had a nonexistent component.
* ENOMEM Could not allocate a memory to copy filenames or data into.
* ENOTBLK 'source' is not a block device
*
****************************************************************************/
int mount(FAR const char *source, FAR const char *target,
FAR const char *filesystemtype, unsigned long mountflags,
FAR const void *data)
{
int ret;
ret = nx_mount(source, target, filesystemtype, mountflags, data);
if (ret < 0)
{
set_errno(-ret);
ret = ERROR;
}
return ret;
}
Reference in a new issue View git blame Copy permalink