kernel: fs: Introduce ctlfs

ctlfs (Control FS) is a filesystem aimed to provide configuration of
devices separate from their respective devices files. Most UNIX-like
systems rely on the ioctl() mechanism for sending commands to device
drivers, while this certainly works, it is important to realize how
messy it can get having massive switch cases per driver. ctlfs fixes
this by giving each device a "node" in ctlfs that contains several
control files.

For example, imagine we have a SCSI device (/dev/sd1). Its ctlfs
counterpart would be a directory at /ctl/sd1/. For an hdd this may
contain files such as:

/ctl/sd1/bsize   # Block size if read (mode [rwx]: -w-)
/ctl/sd1/cmd     # Command if written, status if read (mode [rwx]: rw-)
... et cetera

Signed-off-by: Ian Moffett <ian@osmora.org>

4 files changed, 472 insertions, 1 deletions

diff --git a/sys/fs/ctlfs.c b/sys/fs/ctlfs.c
new file mode 100644
index 0000000..efbf448
--- /dev/null
+++ b/sys/fs/ctlfs.c
@@ -0,0 +1,406 @@
+/*
+ * Copyright (c) 2023-2025 Ian Marco Moffett and the Osmora Team.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of Hyra nor the names of its
+ *    contributors may be used to endorse or promote products derived from
+ *    this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include <sys/types.h>
+#include <sys/errno.h>
+#include <sys/vnode.h>
+#include <sys/device.h>
+#include <sys/syslog.h>
+#include <sys/mount.h>
+#include <sys/queue.h>
+#include <fs/ctlfs.h>
+#include <vm/dynalloc.h>
+#include <string.h>
+
+#define CTLFS_MPNAME "ctl"
+#define CTLFS_ENTRY_MAG 0x43454E54UL  /* 'CENT' */
+#define CTLFS_NODE_MAG  0x43544C4EUL  /* 'CTLN' */
+
+#define pr_trace(fmt, ...) kprintf("ctlfs: " fmt, ##__VA_ARGS__)
+#define pr_error(...) pr_trace(__VA_ARGS__)
+
+static const struct vops ctlfs_vops;
+
+struct ctlfs_hdr {
+    uint32_t magic;
+    char *name;
+};
+
+/*
+ * Control fs entry, represents a control
+ * file within a ctlfs node.
+ * -- HDR START --
+ * @magic: Magic number [MUST BE FIRST] (CTLFS_ENTRY_MAG)
+ * @name: Entry name    [MUST BE SECOND]
+ * -- HDR END --
+ * @parent: Parent (ctlfs_node)
+ * @io: Ctlfs operations.
+ * @mode: Access flags.
+ * @link: TAILQ link.
+ */
+struct ctlfs_entry {
+    uint32_t magic;
+    char *name;
+    struct ctlfs_node *parent;
+    const struct ctlops *io;
+    mode_t mode;
+    TAILQ_ENTRY(ctlfs_entry) link;
+};
+
+/*
+ * Control fs node, represents a directory
+ * within ctlfs. These directories represent
+ * devices, each device directory contains
+ * control files.
+ *
+ *              For example:
+ *
+ *             /ctl/sd1/bsize  # Block size
+ *             /ctl/sd1/health # Health
+ *             [et cetera]
+ *
+ * @magic: Magic number [MUST BE FIRST] (CTLFS_NODE_MAG)
+ * @name: Name of node  [MUST BE SECOND]
+ * @mode: Access flags.
+ * @major: Device major number.
+ * @minor: Device major number.
+ * @eq: Entries for this ctlfs node.
+ */
+struct ctlfs_node {
+    uint32_t magic;
+    char *name;
+    mode_t mode;
+    TAILQ_HEAD(, ctlfs_entry) eq;
+    TAILQ_ENTRY(ctlfs_node) link;
+};
+
+static TAILQ_HEAD(, ctlfs_node) nodeq;
+
+/*
+ * Look up entries within a ctlfs
+ * node by name.
+ */
+static struct ctlfs_entry *
+entry_lookup(struct ctlfs_node *cnp, const char *name)
+{
+    struct ctlfs_entry *ep;
+
+    TAILQ_FOREACH(ep, &cnp->eq, link) {
+        if (strcmp(ep->name, name) == 0) {
+            return ep;
+        }
+    }
+
+    return NULL;
+}
+
+/*
+ * Lookup a ctlfs entry by name.
+ */
+static struct ctlfs_node *
+node_lookup(const char *name)
+{
+    struct ctlfs_node *cnp;
+
+    TAILQ_FOREACH(cnp, &nodeq, link) {
+        if (strcmp(cnp->name, name) == 0) {
+            return cnp;
+        }
+    }
+
+    return NULL;
+}
+
+static int
+ctlfs_init(struct fs_info *fip)
+{
+    struct vnode *vp;
+    struct mount *mp;
+    int error;
+
+    if ((error = vfs_alloc_vnode(&vp, VDIR)) != 0) {
+        pr_error("failed to alloc vnode\n");
+        return error;
+    }
+
+    vp->vops = &ctlfs_vops;
+    if ((mp = vfs_alloc_mount(vp, fip)) == NULL) {
+        pr_trace("failed to alloc mountpoint\n");
+        return -ENOMEM;
+    }
+
+    error = vfs_name_mount(mp, CTLFS_MPNAME);
+    if (error != 0) {
+        pr_trace("failed to mount @ /%s\n", CTLFS_MPNAME);
+        return error;
+    }
+
+    TAILQ_INSERT_TAIL(&g_mountlist, mp, mnt_list);
+    TAILQ_INIT(&nodeq);
+    return 0;
+}
+
+static int
+ctlfs_lookup(struct vop_lookup_args *args)
+{
+    int error;
+    const char *name = args->name;
+    struct vnode *vp, *dirvp;
+    struct ctlfs_node *cnp = NULL;
+    struct ctlfs_entry *enp = NULL;
+
+    if (name == NULL) {
+        return -EINVAL;
+    }
+
+    dirvp = args->dirvp;
+    if (dirvp == NULL) {
+        return -EIO;
+    }
+
+    /*
+     * If we already have data within this vnode
+     * it *might* be a control node but we'll have
+     * to verify its magic number...
+     */
+    if (dirvp->data != NULL) {
+        cnp = (struct ctlfs_node *)dirvp->data;
+        if (cnp->magic != CTLFS_NODE_MAG) {
+            pr_error("bad `cnp' magic (name=%s)\n", name);
+            return -EIO;
+        }
+    }
+
+    /*
+     * Handle cases where we are looking up
+     * relative to a control node.
+     */
+    if (cnp != NULL) {
+        enp = entry_lookup(cnp, name);
+        if (enp == NULL) {
+            return -ENOENT;
+        }
+
+        /* Create a vnode for this enp */
+        error = vfs_alloc_vnode(&vp, VCHR);
+        if (error != 0) {
+            return error;
+        }
+
+        vp->data = (void *)enp;
+        vp->vops = &ctlfs_vops;
+        *args->vpp = vp;
+        return 0;
+    }
+
+    /* Does this entry exist? */
+    if ((cnp = node_lookup(name)) == NULL) {
+        return -ENOENT;
+    }
+
+    if ((error = vfs_alloc_vnode(&vp, VDIR)) != 0) {
+        return error;
+    }
+
+    vp->data = cnp;
+    vp->vops = &ctlfs_vops;
+    *args->vpp = vp;
+    return 0;
+}
+
+/*
+ * Create a ctlfs node (directory) within the
+ * root fs.
+ *
+ * @name: Node name (e.g., "sd1" for "/ctl/sd1/")
+ * @dp: Device related arguments (see ctlfs_dev)
+ *      Args used:
+ *      - mode (access flags)
+ *
+ */
+int
+ctlfs_create_node(const char *name, const struct ctlfs_dev *dp)
+{
+    struct ctlfs_node *cnp;
+    size_t namelen;
+
+    if (name == NULL || dp == NULL) {
+        return -EINVAL;
+    }
+
+    cnp = dynalloc(sizeof(*cnp));
+    if (cnp == NULL) {
+        return -ENOMEM;
+    }
+
+    namelen = strlen(name);
+    cnp->name = dynalloc(namelen + 1);
+    if (cnp->name == NULL) {
+        dynfree(cnp);
+        return -ENOMEM;
+    }
+
+    memcpy(cnp->name, name, namelen);
+    cnp->name[namelen] = '\0';
+    cnp->mode = dp->mode;
+    cnp->magic = CTLFS_NODE_MAG;
+    TAILQ_INSERT_TAIL(&nodeq, cnp, link);
+    TAILQ_INIT(&cnp->eq);
+    return 0;
+}
+
+/*
+ * Create a ctlfs entry within a specific node.
+ *
+ * @name: Name e.g., "/health" for "/ctl/xxx/health".
+ * @dp: Device related arguments (see ctlfs_dev)
+ *      Args used:
+ *      - devname (name of device)
+ *      - mode (access flags)
+ *      - ops (operations vector)
+ */
+int
+ctlfs_create_entry(const char *name, const struct ctlfs_dev *dp)
+{
+    struct ctlfs_entry *enp;
+    struct ctlfs_node *parent;
+    size_t namelen;
+
+    if (name == NULL || dp == NULL) {
+        return -EINVAL;
+    }
+    if (dp->devname == NULL) {
+        return -EINVAL;
+    }
+    if (dp->ops == NULL) {
+        return -EINVAL;
+    }
+
+    parent = node_lookup(dp->devname);
+    if (parent == NULL) {
+        pr_trace("could not find %s\n", dp->devname);
+        return -ENOENT;
+    }
+
+    enp = dynalloc(sizeof(*enp));
+    if (enp == NULL) {
+        return -ENOMEM;
+    }
+
+    namelen = strlen(name);
+    enp->name = dynalloc(namelen + 1);
+    if (enp->name == NULL) {
+        dynfree(enp);
+        return -ENOMEM;
+    }
+
+    memcpy(enp->name, name, namelen);
+    enp->name[namelen] = '\0';
+    enp->io = dp->ops;
+    enp->magic = CTLFS_ENTRY_MAG;
+    enp->mode = dp->mode;
+    enp->parent = parent;
+    TAILQ_INSERT_TAIL(&parent->eq, enp, link);
+    return 0;
+}
+
+/*
+ * Read a control file
+ *
+ * Args passed to driver:
+ *   - ctlfs_dev.ctlname
+ *   - ctlfs_dev.iop
+ *   - ctlfs_dev.mode
+ */
+static int
+ctlfs_read(struct vnode *vp, struct sio_txn *sio)
+{
+    const struct ctlops *iop;
+    struct ctlfs_entry *enp;
+    struct ctlfs_dev dev;
+
+    if ((enp = vp->data) == NULL) {
+        pr_error("no vnode data for ctlfs entry\n");
+        return -EIO;
+    }
+    if (enp->magic != CTLFS_ENTRY_MAG) {
+        pr_error("ctlfs entry has bad magic\n");
+        return -EIO;
+    }
+    if ((iop = enp->io) == NULL) {
+        pr_error("no i/o ops for ctlfs entry\n");
+        return -EIO;
+    }
+    if (iop->read == NULL) {
+        pr_trace("no read op for ctlfs entry\n");
+        return -EIO;
+    }
+
+    dev.ctlname = enp->name;
+    dev.ops = iop;
+    dev.mode = enp->mode;
+    return iop->read(&dev, sio);
+}
+
+static int
+ctlfs_reclaim(struct vnode *vp)
+{
+    struct ctlfs_hdr *hp;
+
+    if (vp->data == NULL) {
+        return 0;
+    }
+
+    /* Ensure the magic is correct */
+    hp = vp->data;
+    switch (hp->magic) {
+    case CTLFS_NODE_MAG:
+    case CTLFS_ENTRY_MAG:
+        dynfree(hp->name);
+        break;
+    default:
+        pr_error("reclaim: bad magic in vp\n");
+        break;
+    }
+
+    dynfree(vp->data);
+    return 0;
+}
+
+static const struct vops ctlfs_vops = {
+    .lookup = ctlfs_lookup,
+    .read = ctlfs_read,
+    .getattr = NULL,
+    .write = NULL,
+    .reclaim = ctlfs_reclaim
+};
+
+const struct vfsops g_ctlfs_vfsops = {
+    .init = ctlfs_init
+};
diff --git a/sys/include/fs/ctlfs.h b/sys/include/fs/ctlfs.h
new file mode 100644
index 0000000..582421c
--- /dev/null
+++ b/sys/include/fs/ctlfs.h
@@ -0,0 +1,62 @@
+/*
+ * Copyright (c) 2023-2025 Ian Marco Moffett and the Osmora Team.
+ * All rights reserved.
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice,
+ *    this list of conditions and the following disclaimer.
+ * 2. Redistributions in binary form must reproduce the above copyright
+ *    notice, this list of conditions and the following disclaimer in the
+ *    documentation and/or other materials provided with the distribution.
+ * 3. Neither the name of Hyra nor the names of its
+ *    contributors may be used to endorse or promote products derived from
+ *    this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+ * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+ * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
+ * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE
+ * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
+ * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
+ * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
+ * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
+ * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
+ * POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#ifndef _FS_CTLFS_H_
+#define _FS_CTLFS_H_
+
+struct ctlfs_dev;
+
+struct ctlops {
+    int(*read)(struct ctlfs_dev *cdp, struct sio_txn *sio);
+    int(*write)(struct ctlfs_dev *cdp, struct sio_txn *sio);
+};
+
+/*
+ * Ctlfs op arguments
+ *
+ * @devname: Device name.
+ * @major: Device major
+ * @minor: Device minor.
+ * @mode: Access flags.
+ * @devname [1]: Device name (node name)
+ * @ctlname: [1]: Control name (node entry name)
+ */
+struct ctlfs_dev {
+    union {
+        const char *devname;
+        const char *ctlname;
+    };
+    const struct ctlops *ops;
+    mode_t mode;
+};
+
+int ctlfs_create_node(const char *name, const struct ctlfs_dev *dp);
+int ctlfs_create_entry(const char *name, const struct ctlfs_dev *dp);
+
+#endif  /* !_FS_CTLFS_H_ */
diff --git a/sys/include/sys/mount.h b/sys/include/sys/mount.h
index 3b5d89e..1fcdbfa 100644
--- a/sys/include/sys/mount.h
+++ b/sys/include/sys/mount.h
@@ -46,6 +46,7 @@
  */
 #define MOUNT_RAMFS "initramfs"
 #define MOUNT_DEVFS "devfs"
+#define MOUNT_CTLFS "ctlfs"
 
 struct vfsops;
 struct mount;
@@ -57,6 +58,7 @@ extern mountlist_t g_mountlist;
 /* Filesystem operations */
 extern const struct vfsops g_initramfs_vfsops;
 extern const struct vfsops g_devfs_vfsops;
+extern const struct vfsops g_ctlfs_vfsops;
 
 struct mount {
     char *name;
diff --git a/sys/kern/vfs_init.c b/sys/kern/vfs_init.c
index caa0766..8c1bc74 100644
--- a/sys/kern/vfs_init.c
+++ b/sys/kern/vfs_init.c
@@ -36,7 +36,8 @@
 struct vnode *g_root_vnode = NULL;
 static struct fs_info fs_list[] = {
     {MOUNT_RAMFS, &g_initramfs_vfsops, 0, 0},
-    {MOUNT_DEVFS, &g_devfs_vfsops, 0, 0}
+    {MOUNT_DEVFS, &g_devfs_vfsops, 0, 0},
+    {MOUNT_CTLFS, &g_ctlfs_vfsops, 0, 0}
 };
 
 void