Subversion Repositories HelenOS

Compare Revisions

Ignore whitespace Rev 2770 → Rev 2782

/tags/0.3.0/uspace/srv/fs/tmpfs/tmpfs.h
0,0 → 1,73
/*
* Copyright (c) 2008 Jakub Jermar
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* - Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* - 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.
* - The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
*/
 
/** @addtogroup fs
* @{
*/
 
#ifndef TMPFS_TMPFS_H_
#define TMPFS_TMPFS_H_
 
#include <ipc/ipc.h>
#include <libfs.h>
#include <atomic.h>
#include <sys/types.h>
#include <bool.h>
#include <libadt/hash_table.h>
 
#define dprintf(...) printf(__VA_ARGS__)
 
typedef struct tmpfs_dentry {
fs_index_t index; /**< TMPFS node index. */
link_t dh_link; /**< Dentries hash table link. */
struct tmpfs_dentry *sibling;
struct tmpfs_dentry *child;
hash_table_t names; /**< All names linking to this TMPFS node. */
enum {
TMPFS_NONE,
TMPFS_FILE,
TMPFS_DIRECTORY
} type;
unsigned lnkcnt; /**< Link count. */
size_t size; /**< File size if type is TMPFS_FILE. */
void *data; /**< File content's if type is TMPFS_FILE. */
} tmpfs_dentry_t;
 
extern fs_reg_t tmpfs_reg;
 
extern void tmpfs_lookup(ipc_callid_t, ipc_call_t *);
extern void tmpfs_read(ipc_callid_t, ipc_call_t *);
extern void tmpfs_write(ipc_callid_t, ipc_call_t *);
extern void tmpfs_truncate(ipc_callid_t, ipc_call_t *);
extern void tmpfs_destroy(ipc_callid_t, ipc_call_t *);
 
#endif
 
/**
* @}
*/
/tags/0.3.0/uspace/srv/fs/tmpfs/tmpfs_ops.c
0,0 → 1,587
/*
* Copyright (c) 2008 Jakub Jermar
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* - Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* - 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.
* - The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
*/
 
/** @addtogroup fs
* @{
*/
 
/**
* @file tmpfs_ops.c
* @brief Implementation of VFS operations for the TMPFS file system
* server.
*/
 
#include "tmpfs.h"
#include "../../vfs/vfs.h"
#include <ipc/ipc.h>
#include <async.h>
#include <errno.h>
#include <atomic.h>
#include <stdlib.h>
#include <string.h>
#include <stdio.h>
#include <assert.h>
#include <sys/types.h>
#include <libadt/hash_table.h>
#include <as.h>
#include <libfs.h>
 
#define min(a, b) ((a) < (b) ? (a) : (b))
#define max(a, b) ((a) > (b) ? (a) : (b))
 
#define DENTRIES_BUCKETS 256
 
#define NAMES_BUCKETS 4
 
/*
* For now, we don't distinguish between different dev_handles/instances. All
* requests resolve to the only instance, rooted in the following variable.
*/
static tmpfs_dentry_t *root;
 
/*
* Implementation of the libfs interface.
*/
 
/* Forward declarations of static functions. */
static bool tmpfs_match(void *, void *, const char *);
static void *tmpfs_node_get(fs_handle_t, dev_handle_t, fs_index_t);
static void *tmpfs_create_node(int);
static bool tmpfs_link_node(void *, void *, const char *);
static int tmpfs_unlink_node(void *, void *);
static void tmpfs_destroy_node(void *);
 
/* Implementation of helper functions. */
static fs_index_t tmpfs_index_get(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->index;
}
 
static size_t tmpfs_size_get(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->size;
}
 
static unsigned tmpfs_lnkcnt_get(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->lnkcnt;
}
 
static void *tmpfs_child_get(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->child;
}
 
static void *tmpfs_sibling_get(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->sibling;
}
 
static void *tmpfs_root_get(void)
{
return root;
}
 
static char tmpfs_plb_get_char(unsigned pos)
{
return tmpfs_reg.plb_ro[pos % PLB_SIZE];
}
 
static bool tmpfs_is_directory(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->type == TMPFS_DIRECTORY;
}
 
static bool tmpfs_is_file(void *nodep)
{
return ((tmpfs_dentry_t *) nodep)->type == TMPFS_FILE;
}
 
/** libfs operations */
libfs_ops_t tmpfs_libfs_ops = {
.match = tmpfs_match,
.node_get = tmpfs_node_get,
.create = tmpfs_create_node,
.destroy = tmpfs_destroy_node,
.link = tmpfs_link_node,
.unlink = tmpfs_unlink_node,
.index_get = tmpfs_index_get,
.size_get = tmpfs_size_get,
.lnkcnt_get = tmpfs_lnkcnt_get,
.child_get = tmpfs_child_get,
.sibling_get = tmpfs_sibling_get,
.root_get = tmpfs_root_get,
.plb_get_char = tmpfs_plb_get_char,
.is_directory = tmpfs_is_directory,
.is_file = tmpfs_is_file
};
 
/** Hash table of all directory entries. */
hash_table_t dentries;
 
/* Implementation of hash table interface for the dentries hash table. */
static hash_index_t dentries_hash(unsigned long *key)
{
return *key % DENTRIES_BUCKETS;
}
 
static int dentries_compare(unsigned long *key, hash_count_t keys,
link_t *item)
{
tmpfs_dentry_t *dentry = hash_table_get_instance(item, tmpfs_dentry_t,
dh_link);
return dentry->index == *key;
}
 
static void dentries_remove_callback(link_t *item)
{
}
 
/** TMPFS dentries hash table operations. */
hash_table_operations_t dentries_ops = {
.hash = dentries_hash,
.compare = dentries_compare,
.remove_callback = dentries_remove_callback
};
 
fs_index_t tmpfs_next_index = 1;
 
typedef struct {
char *name;
tmpfs_dentry_t *parent;
link_t link;
} tmpfs_name_t;
 
/* Implementation of hash table interface for the names hash table. */
static hash_index_t names_hash(unsigned long *key)
{
tmpfs_dentry_t *dentry = (tmpfs_dentry_t *) *key;
return dentry->index % NAMES_BUCKETS;
}
 
static int names_compare(unsigned long *key, hash_count_t keys, link_t *item)
{
tmpfs_dentry_t *dentry = (tmpfs_dentry_t *) *key;
tmpfs_name_t *namep = hash_table_get_instance(item, tmpfs_name_t,
link);
return dentry == namep->parent;
}
 
static void names_remove_callback(link_t *item)
{
tmpfs_name_t *namep = hash_table_get_instance(item, tmpfs_name_t,
link);
free(namep->name);
free(namep);
}
 
/** TMPFS node names hash table operations. */
static hash_table_operations_t names_ops = {
.hash = names_hash,
.compare = names_compare,
.remove_callback = names_remove_callback
};
 
static void tmpfs_name_initialize(tmpfs_name_t *namep)
{
namep->name = NULL;
namep->parent = NULL;
link_initialize(&namep->link);
}
 
static bool tmpfs_dentry_initialize(tmpfs_dentry_t *dentry)
{
dentry->index = 0;
dentry->sibling = NULL;
dentry->child = NULL;
dentry->type = TMPFS_NONE;
dentry->lnkcnt = 0;
dentry->size = 0;
dentry->data = NULL;
link_initialize(&dentry->dh_link);
return (bool)hash_table_create(&dentry->names, NAMES_BUCKETS, 1,
&names_ops);
}
 
static bool tmpfs_init(void)
{
if (!hash_table_create(&dentries, DENTRIES_BUCKETS, 1, &dentries_ops))
return false;
root = (tmpfs_dentry_t *) tmpfs_create_node(L_DIRECTORY);
if (!root) {
hash_table_destroy(&dentries);
return false;
}
root->lnkcnt = 1;
return true;
}
 
/** Compare one component of path to a directory entry.
*
* @param prnt Node from which we descended.
* @param chld Node to compare the path component with.
* @param component Array of characters holding component name.
*
* @return True on match, false otherwise.
*/
bool tmpfs_match(void *prnt, void *chld, const char *component)
{
tmpfs_dentry_t *parentp = (tmpfs_dentry_t *) prnt;
tmpfs_dentry_t *childp = (tmpfs_dentry_t *) chld;
 
unsigned long key = (unsigned long) parentp;
link_t *hlp = hash_table_find(&childp->names, &key);
assert(hlp);
tmpfs_name_t *namep = hash_table_get_instance(hlp, tmpfs_name_t, link);
 
return !strcmp(namep->name, component);
}
 
void *
tmpfs_node_get(fs_handle_t fs_handle, dev_handle_t dev_handle, fs_index_t index)
{
unsigned long key = index;
link_t *lnk = hash_table_find(&dentries, &key);
if (!lnk)
return NULL;
return hash_table_get_instance(lnk, tmpfs_dentry_t, dh_link);
}
 
void *tmpfs_create_node(int lflag)
{
assert((lflag & L_FILE) ^ (lflag & L_DIRECTORY));
 
tmpfs_dentry_t *node = malloc(sizeof(tmpfs_dentry_t));
if (!node)
return NULL;
 
if (!tmpfs_dentry_initialize(node)) {
free(node);
return NULL;
}
node->index = tmpfs_next_index++;
if (lflag & L_DIRECTORY)
node->type = TMPFS_DIRECTORY;
else
node->type = TMPFS_FILE;
 
/* Insert the new node into the dentry hash table. */
unsigned long key = node->index;
hash_table_insert(&dentries, &key, &node->dh_link);
return (void *) node;
}
 
bool tmpfs_link_node(void *prnt, void *chld, const char *nm)
{
tmpfs_dentry_t *parentp = (tmpfs_dentry_t *) prnt;
tmpfs_dentry_t *childp = (tmpfs_dentry_t *) chld;
 
assert(parentp->type == TMPFS_DIRECTORY);
 
tmpfs_name_t *namep = malloc(sizeof(tmpfs_name_t));
if (!namep)
return false;
tmpfs_name_initialize(namep);
size_t len = strlen(nm);
namep->name = malloc(len + 1);
if (!namep->name) {
free(namep);
return false;
}
strcpy(namep->name, nm);
namep->parent = parentp;
childp->lnkcnt++;
 
unsigned long key = (unsigned long) parentp;
hash_table_insert(&childp->names, &key, &namep->link);
 
/* Insert the new node into the namespace. */
if (parentp->child) {
tmpfs_dentry_t *tmp = parentp->child;
while (tmp->sibling)
tmp = tmp->sibling;
tmp->sibling = childp;
} else {
parentp->child = childp;
}
 
return true;
}
 
int tmpfs_unlink_node(void *prnt, void *chld)
{
tmpfs_dentry_t *parentp = (tmpfs_dentry_t *)prnt;
tmpfs_dentry_t *childp = (tmpfs_dentry_t *)chld;
 
if (!parentp)
return EBUSY;
 
if (childp->child)
return ENOTEMPTY;
 
if (parentp->child == childp) {
parentp->child = childp->sibling;
} else {
/* TODO: consider doubly linked list for organizing siblings. */
tmpfs_dentry_t *tmp = parentp->child;
while (tmp->sibling != childp)
tmp = tmp->sibling;
tmp->sibling = childp->sibling;
}
childp->sibling = NULL;
 
unsigned long key = (unsigned long) parentp;
hash_table_remove(&childp->names, &key, 1);
 
childp->lnkcnt--;
 
return EOK;
}
 
void tmpfs_destroy_node(void *nodep)
{
tmpfs_dentry_t *dentry = (tmpfs_dentry_t *) nodep;
assert(!dentry->lnkcnt);
assert(!dentry->child);
assert(!dentry->sibling);
 
unsigned long key = dentry->index;
hash_table_remove(&dentries, &key, 1);
 
hash_table_destroy(&dentry->names);
 
if (dentry->type == TMPFS_FILE)
free(dentry->data);
free(dentry);
}
 
void tmpfs_lookup(ipc_callid_t rid, ipc_call_t *request)
{
/* Initialize TMPFS. */
if (!root && !tmpfs_init()) {
ipc_answer_0(rid, ENOMEM);
return;
}
libfs_lookup(&tmpfs_libfs_ops, tmpfs_reg.fs_handle, rid, request);
}
 
void tmpfs_read(ipc_callid_t rid, ipc_call_t *request)
{
dev_handle_t dev_handle = (dev_handle_t)IPC_GET_ARG1(*request);
fs_index_t index = (fs_index_t)IPC_GET_ARG2(*request);
off_t pos = (off_t)IPC_GET_ARG3(*request);
 
/*
* Lookup the respective dentry.
*/
link_t *hlp;
unsigned long key = index;
hlp = hash_table_find(&dentries, &key);
if (!hlp) {
ipc_answer_0(rid, ENOENT);
return;
}
tmpfs_dentry_t *dentry = hash_table_get_instance(hlp, tmpfs_dentry_t,
dh_link);
 
/*
* Receive the read request.
*/
ipc_callid_t callid;
size_t len;
if (!ipc_data_read_receive(&callid, &len)) {
ipc_answer_0(callid, EINVAL);
ipc_answer_0(rid, EINVAL);
return;
}
 
size_t bytes;
if (dentry->type == TMPFS_FILE) {
bytes = max(0, min(dentry->size - pos, len));
(void) ipc_data_read_finalize(callid, dentry->data + pos,
bytes);
} else {
int i;
tmpfs_dentry_t *cur;
assert(dentry->type == TMPFS_DIRECTORY);
/*
* Yes, we really use O(n) algorithm here.
* If it bothers someone, it could be fixed by introducing a
* hash table.
*/
for (i = 0, cur = dentry->child; i < pos && cur; i++,
cur = cur->sibling)
;
 
if (!cur) {
ipc_answer_0(callid, ENOENT);
ipc_answer_1(rid, ENOENT, 0);
return;
}
 
unsigned long key = (unsigned long) dentry;
link_t *hlp = hash_table_find(&cur->names, &key);
assert(hlp);
tmpfs_name_t *namep = hash_table_get_instance(hlp, tmpfs_name_t,
link);
 
(void) ipc_data_read_finalize(callid, namep->name,
strlen(namep->name) + 1);
bytes = 1;
}
 
/*
* Answer the VFS_READ call.
*/
ipc_answer_1(rid, EOK, bytes);
}
 
void tmpfs_write(ipc_callid_t rid, ipc_call_t *request)
{
dev_handle_t dev_handle = (dev_handle_t)IPC_GET_ARG1(*request);
fs_index_t index = (fs_index_t)IPC_GET_ARG2(*request);
off_t pos = (off_t)IPC_GET_ARG3(*request);
 
/*
* Lookup the respective dentry.
*/
link_t *hlp;
unsigned long key = index;
hlp = hash_table_find(&dentries, &key);
if (!hlp) {
ipc_answer_0(rid, ENOENT);
return;
}
tmpfs_dentry_t *dentry = hash_table_get_instance(hlp, tmpfs_dentry_t,
dh_link);
 
/*
* Receive the write request.
*/
ipc_callid_t callid;
size_t len;
if (!ipc_data_write_receive(&callid, &len)) {
ipc_answer_0(callid, EINVAL);
ipc_answer_0(rid, EINVAL);
return;
}
 
/*
* Check whether the file needs to grow.
*/
if (pos + len <= dentry->size) {
/* The file size is not changing. */
(void) ipc_data_write_finalize(callid, dentry->data + pos, len);
ipc_answer_2(rid, EOK, len, dentry->size);
return;
}
size_t delta = (pos + len) - dentry->size;
/*
* At this point, we are deliberately extremely straightforward and
* simply realloc the contents of the file on every write that grows the
* file. In the end, the situation might not be as bad as it may look:
* our heap allocator can save us and just grow the block whenever
* possible.
*/
void *newdata = realloc(dentry->data, dentry->size + delta);
if (!newdata) {
ipc_answer_0(callid, ENOMEM);
ipc_answer_2(rid, EOK, 0, dentry->size);
return;
}
/* Clear any newly allocated memory in order to emulate gaps. */
memset(newdata + dentry->size, 0, delta);
dentry->size += delta;
dentry->data = newdata;
(void) ipc_data_write_finalize(callid, dentry->data + pos, len);
ipc_answer_2(rid, EOK, len, dentry->size);
}
 
void tmpfs_truncate(ipc_callid_t rid, ipc_call_t *request)
{
dev_handle_t dev_handle = (dev_handle_t)IPC_GET_ARG1(*request);
fs_index_t index = (fs_index_t)IPC_GET_ARG2(*request);
size_t size = (off_t)IPC_GET_ARG3(*request);
 
/*
* Lookup the respective dentry.
*/
link_t *hlp;
unsigned long key = index;
hlp = hash_table_find(&dentries, &key);
if (!hlp) {
ipc_answer_0(rid, ENOENT);
return;
}
tmpfs_dentry_t *dentry = hash_table_get_instance(hlp, tmpfs_dentry_t,
dh_link);
 
if (size == dentry->size) {
ipc_answer_0(rid, EOK);
return;
}
 
void *newdata = realloc(dentry->data, size);
if (!newdata) {
ipc_answer_0(rid, ENOMEM);
return;
}
if (size > dentry->size) {
size_t delta = size - dentry->size;
memset(newdata + dentry->size, 0, delta);
}
dentry->size = size;
dentry->data = newdata;
ipc_answer_0(rid, EOK);
}
 
void tmpfs_destroy(ipc_callid_t rid, ipc_call_t *request)
{
dev_handle_t dev_handle = (dev_handle_t)IPC_GET_ARG1(*request);
fs_index_t index = (fs_index_t)IPC_GET_ARG2(*request);
 
link_t *hlp;
unsigned long key = index;
hlp = hash_table_find(&dentries, &key);
if (!hlp) {
ipc_answer_0(rid, ENOENT);
return;
}
tmpfs_dentry_t *dentry = hash_table_get_instance(hlp, tmpfs_dentry_t,
dh_link);
tmpfs_destroy_node(dentry);
ipc_answer_0(rid, EOK);
}
 
/**
* @}
*/
/tags/0.3.0/uspace/srv/fs/tmpfs/tmpfs.c
0,0 → 1,158
/*
* Copyright (c) 2006 Martin Decky
* Copyright (c) 2008 Jakub Jermar
* All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
*
* - Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* - 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.
* - The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
*/
 
/** @addtogroup fs
* @{
*/
 
/**
* @file tmpfs.c
* @brief File system driver for in-memory file system.
*
* Every instance of tmpfs exists purely in memory and has neither a disk layout
* nor any permanent storage (e.g. disk blocks). With each system reboot, data
* stored in a tmpfs file system is lost.
*/
 
#include "tmpfs.h"
#include <ipc/ipc.h>
#include <ipc/services.h>
#include <async.h>
#include <errno.h>
#include <unistd.h>
#include <stdio.h>
#include <libfs.h>
#include "../../vfs/vfs.h"
 
 
vfs_info_t tmpfs_vfs_info = {
.name = "tmpfs",
.ops = {
[IPC_METHOD_TO_VFS_OP(VFS_LOOKUP)] = VFS_OP_DEFINED,
[IPC_METHOD_TO_VFS_OP(VFS_READ)] = VFS_OP_DEFINED,
[IPC_METHOD_TO_VFS_OP(VFS_WRITE)] = VFS_OP_DEFINED,
[IPC_METHOD_TO_VFS_OP(VFS_TRUNCATE)] = VFS_OP_DEFINED,
[IPC_METHOD_TO_VFS_OP(VFS_MOUNT)] = VFS_OP_NULL,
[IPC_METHOD_TO_VFS_OP(VFS_UNMOUNT)] = VFS_OP_NULL,
[IPC_METHOD_TO_VFS_OP(VFS_DESTROY)] = VFS_OP_DEFINED,
}
};
 
fs_reg_t tmpfs_reg;
 
/**
* This connection fibril processes VFS requests from VFS.
*
* In order to support simultaneous VFS requests, our design is as follows.
* The connection fibril accepts VFS requests from VFS. If there is only one
* instance of the fibril, VFS will need to serialize all VFS requests it sends
* to FAT. To overcome this bottleneck, VFS can send TMPFS the
* IPC_M_CONNECT_ME_TO call. In that case, a new connection fibril will be
* created, which in turn will accept the call. Thus, a new phone will be
* opened for VFS.
*
* There are few issues with this arrangement. First, VFS can run out of
* available phones. In that case, VFS can close some other phones or use one
* phone for more serialized requests. Similarily, TMPFS can refuse to duplicate
* the connection. VFS should then just make use of already existing phones and
* route its requests through them. To avoid paying the fibril creation price
* upon each request, TMPFS might want to keep the connections open after the
* request has been completed.
*/
static void tmpfs_connection(ipc_callid_t iid, ipc_call_t *icall)
{
if (iid) {
/*
* This only happens for connections opened by
* IPC_M_CONNECT_ME_TO calls as opposed to callback connections
* created by IPC_M_CONNECT_TO_ME.
*/
ipc_answer_0(iid, EOK);
}
dprintf("VFS-TMPFS connection established.\n");
while (1) {
ipc_callid_t callid;
ipc_call_t call;
callid = async_get_call(&call);
switch (IPC_GET_METHOD(call)) {
case VFS_LOOKUP:
tmpfs_lookup(callid, &call);
break;
case VFS_READ:
tmpfs_read(callid, &call);
break;
case VFS_WRITE:
tmpfs_write(callid, &call);
break;
case VFS_TRUNCATE:
tmpfs_truncate(callid, &call);
break;
case VFS_DESTROY:
tmpfs_destroy(callid, &call);
break;
default:
ipc_answer_0(callid, ENOTSUP);
break;
}
}
}
 
int main(int argc, char **argv)
{
int vfs_phone;
 
printf("TMPFS: HelenOS TMPFS file system server.\n");
 
vfs_phone = ipc_connect_me_to(PHONE_NS, SERVICE_VFS, 0, 0);
while (vfs_phone < EOK) {
usleep(10000);
vfs_phone = ipc_connect_me_to(PHONE_NS, SERVICE_VFS, 0, 0);
}
int rc;
rc = fs_register(vfs_phone, &tmpfs_reg, &tmpfs_vfs_info,
tmpfs_connection);
if (rc != EOK) {
printf("Failed to register the TMPFS file system (%d)\n", rc);
return rc;
}
dprintf("TMPFS filesystem registered, fs_handle=%d.\n",
tmpfs_reg.fs_handle);
 
async_manager();
/* not reached */
return 0;
}
 
/**
* @}
*/
/tags/0.3.0/uspace/srv/fs/tmpfs/Makefile
0,0 → 1,76
#
# Copyright (c) 2006 Martin Decky
# All rights reserved.
#
# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions
# are met:
#
# - Redistributions of source code must retain the above copyright
# notice, this list of conditions and the following disclaimer.
# - 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.
# - The name of the author may not be used to endorse or promote products
# derived from this software without specific prior written permission.
#
# THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 AUTHOR 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.
#
 
## Setup toolchain
#
 
LIBC_PREFIX = ../../../lib/libc
LIBFS_PREFIX = ../../../lib/libfs
SOFTINT_PREFIX = ../../../lib/softint
include $(LIBC_PREFIX)/Makefile.toolchain
 
CFLAGS += -I $(LIBFS_PREFIX)
 
LIBS = $(LIBC_PREFIX)/libc.a $(LIBFS_PREFIX)/libfs.a
 
## Sources
#
 
OUTPUT = tmpfs
SOURCES = \
tmpfs.c \
tmpfs_ops.c
 
OBJECTS := $(addsuffix .o,$(basename $(SOURCES)))
 
.PHONY: all clean depend disasm
 
all: $(OUTPUT) disasm
 
-include Makefile.depend
 
clean:
-rm -f $(OUTPUT) $(OUTPUT).map $(OUTPUT).disasm Makefile.depend
 
depend:
$(CC) $(DEFS) $(CFLAGS) -M $(SOURCES) > Makefile.depend
 
$(OUTPUT): $(OBJECTS) $(LIBS)
$(LD) -T $(LIBC_PREFIX)/arch/$(ARCH)/_link.ld $(OBJECTS) $(LIBS) $(LFLAGS) -o $@ -Map $(OUTPUT).map
 
disasm:
$(OBJDUMP) -d $(OUTPUT) >$(OUTPUT).disasm
 
%.o: %.S
$(CC) $(DEFS) $(AFLAGS) $(CFLAGS) -D__ASM__ -c $< -o $@
 
%.o: %.s
$(AS) $(AFLAGS) $< -o $@
 
%.o: %.c
$(CC) $(DEFS) $(CFLAGS) -c $< -o $@