NetBSD/sys/fs/nilfs/nilfs_vnops.c

1609 lines
39 KiB
C
Raw Normal View History

2014-10-15 13:05:46 +04:00
/* $NetBSD: nilfs_vnops.c,v 1.30 2014/10/15 09:05:46 hannken Exp $ */
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/*
* Copyright (c) 2008, 2009 Reinoud Zandijk
* 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.
*
* 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.
*
*/
#include <sys/cdefs.h>
#ifndef lint
2014-10-15 13:05:46 +04:00
__KERNEL_RCSID(0, "$NetBSD: nilfs_vnops.c,v 1.30 2014/10/15 09:05:46 hannken Exp $");
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
#endif /* not lint */
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/namei.h>
#include <sys/resourcevar.h> /* defines plimit structure in proc struct */
#include <sys/kernel.h>
#include <sys/file.h> /* define FWRITE ... */
#include <sys/stat.h>
#include <sys/buf.h>
#include <sys/proc.h>
#include <sys/mount.h>
#include <sys/vnode.h>
#include <sys/signalvar.h>
#include <sys/malloc.h>
#include <sys/dirent.h>
#include <sys/lockf.h>
#include <sys/kauth.h>
#include <miscfs/genfs/genfs.h>
#include <uvm/uvm_extern.h>
#include <fs/nilfs/nilfs_mount.h>
#include "nilfs.h"
#include "nilfs_subr.h"
#include "nilfs_bswap.h"
#define VTOI(vnode) ((struct nilfs_node *) (vnode)->v_data)
/* externs */
extern int prtactive;
/* implementations of vnode functions; table follows at end */
/* --------------------------------------------------------------------- */
int
nilfs_inactive(void *v)
{
struct vop_inactive_args /* {
struct vnode *a_vp;
bool *a_recycle;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct nilfs_node *nilfs_node = VTOI(vp);
DPRINTF(NODE, ("nilfs_inactive called for nilfs_node %p\n", VTOI(vp)));
if (nilfs_node == NULL) {
DPRINTF(NODE, ("nilfs_inactive: inactive NULL NILFS node\n"));
VOP_UNLOCK(vp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return 0;
}
/*
* Optionally flush metadata to disc. If the file has not been
* referenced anymore in a directory we ought to free up the resources
* on disc if applicable.
*/
VOP_UNLOCK(vp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return 0;
}
/* --------------------------------------------------------------------- */
int
nilfs_reclaim(void *v)
{
struct vop_reclaim_args /* {
struct vnode *a_vp;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct nilfs_node *nilfs_node = VTOI(vp);
DPRINTF(NODE, ("nilfs_reclaim called for node %p\n", nilfs_node));
if (prtactive && vp->v_usecount > 1)
vprint("nilfs_reclaim(): pushing active", vp);
if (nilfs_node == NULL) {
DPRINTF(NODE, ("nilfs_reclaim(): null nilfsnode\n"));
return 0;
}
/* update note for closure */
nilfs_update(vp, NULL, NULL, NULL, UPDATE_CLOSE);
2014-10-15 13:05:46 +04:00
/* remove from vnode cache. */
vcache_remove(vp->v_mount, &nilfs_node->ino, sizeof(nilfs_node->ino));
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* dispose all node knowledge */
genfs_node_destroy(vp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
nilfs_dispose_node(&nilfs_node);
vp->v_data = NULL;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return 0;
}
/* --------------------------------------------------------------------- */
int
nilfs_read(void *v)
{
struct vop_read_args /* {
struct vnode *a_vp;
struct uio *a_uio;
int a_ioflag;
kauth_cred_t a_cred;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct uio *uio = ap->a_uio;
int ioflag = ap->a_ioflag;
int advice = IO_ADV_DECODE(ap->a_ioflag);
struct uvm_object *uobj;
struct nilfs_node *nilfs_node = VTOI(vp);
uint64_t file_size;
vsize_t len;
int error;
DPRINTF(READ, ("nilfs_read called\n"));
/* can this happen? some filingsystems have this check */
if (uio->uio_offset < 0)
return EINVAL;
if (uio->uio_resid == 0)
return 0;
/* protect against rogue programs reading raw directories and links */
if ((ioflag & IO_ALTSEMANTICS) == 0) {
if (vp->v_type == VDIR)
return EISDIR;
/* all but regular files just give EINVAL */
if (vp->v_type != VREG)
return EINVAL;
}
assert(nilfs_node);
file_size = nilfs_rw64(nilfs_node->inode.i_size);
/* read contents using buffercache */
uobj = &vp->v_uobj;
error = 0;
while (uio->uio_resid > 0) {
/* reached end? */
if (file_size <= uio->uio_offset)
break;
/* maximise length to file extremity */
len = MIN(file_size - uio->uio_offset, uio->uio_resid);
if (len == 0)
break;
/* ubc, here we come, prepare to trap */
error = ubc_uiomove(uobj, uio, len, advice,
UBC_READ | UBC_PARTIALOK | UBC_UNMAP_FLAG(vp));
if (error)
break;
}
/* note access time unless not requested */
if (!(vp->v_mount->mnt_flag & MNT_NOATIME)) {
nilfs_node->i_flags |= IN_ACCESS;
if ((ioflag & IO_SYNC) == IO_SYNC)
error = nilfs_update(vp, NULL, NULL, NULL, UPDATE_WAIT);
}
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_write(void *v)
{
struct vop_write_args /* {
struct vnode *a_vp;
struct uio *a_uio;
int a_ioflag;
kauth_cred_t a_cred;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct uio *uio = ap->a_uio;
int ioflag = ap->a_ioflag;
int advice = IO_ADV_DECODE(ap->a_ioflag);
struct uvm_object *uobj;
struct nilfs_node *nilfs_node = VTOI(vp);
2013-10-18 23:57:28 +04:00
uint64_t file_size;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
vsize_t len;
int error, resid, extended;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
DPRINTF(WRITE, ("nilfs_write called\n"));
/* can this happen? some filingsystems have this check */
if (uio->uio_offset < 0)
return EINVAL;
if (uio->uio_resid == 0)
return 0;
/* protect against rogue programs writing raw directories or links */
if ((ioflag & IO_ALTSEMANTICS) == 0) {
if (vp->v_type == VDIR)
return EISDIR;
/* all but regular files just give EINVAL for now */
if (vp->v_type != VREG)
return EINVAL;
}
assert(nilfs_node);
panic("nilfs_write() called\n");
/* remember old file size */
assert(nilfs_node);
file_size = nilfs_rw64(nilfs_node->inode.i_size);
/* if explicitly asked to append, uio_offset can be wrong? */
if (ioflag & IO_APPEND)
uio->uio_offset = file_size;
#if 0
extended = (uio->uio_offset + uio->uio_resid > file_size);
if (extended) {
DPRINTF(WRITE, ("extending file from %"PRIu64" to %"PRIu64"\n",
file_size, uio->uio_offset + uio->uio_resid));
error = nilfs_grow_node(nilfs_node, uio->uio_offset + uio->uio_resid);
if (error)
return error;
file_size = uio->uio_offset + uio->uio_resid;
}
#endif
/* write contents using buffercache */
uobj = &vp->v_uobj;
resid = uio->uio_resid;
error = 0;
uvm_vnp_setwritesize(vp, file_size);
while (uio->uio_resid > 0) {
/* maximise length to file extremity */
len = MIN(file_size - uio->uio_offset, uio->uio_resid);
if (len == 0)
break;
/* ubc, here we come, prepare to trap */
error = ubc_uiomove(uobj, uio, len, advice,
UBC_WRITE | UBC_UNMAP_FLAG(vp));
if (error)
break;
}
uvm_vnp_setsize(vp, file_size);
/* mark node changed and request update */
nilfs_node->i_flags |= IN_CHANGE | IN_UPDATE;
if (vp->v_mount->mnt_flag & MNT_RELATIME)
nilfs_node->i_flags |= IN_ACCESS;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/*
* XXX TODO FFS has code here to reset setuid & setgid when we're not
* the superuser as a precaution against tampering.
*/
/* if we wrote a thing, note write action on vnode */
if (resid > uio->uio_resid)
VN_KNOTE(vp, NOTE_WRITE | (extended ? NOTE_EXTEND : 0));
if (error) {
/* bring back file size to its former size */
/* take notice of its errors? */
// (void) nilfs_chsize(vp, (u_quad_t) old_size, NOCRED);
/* roll back uio */
uio->uio_offset -= resid - uio->uio_resid;
uio->uio_resid = resid;
} else {
/* if we write and we're synchronous, update node */
if ((resid > uio->uio_resid) && ((ioflag & IO_SYNC) == IO_SYNC))
error = nilfs_update(vp, NULL, NULL, NULL, UPDATE_WAIT);
}
return error;
}
/* --------------------------------------------------------------------- */
/*
* bmap functionality that translates logical block numbers to the virtual
* block numbers to be stored on the vnode itself.
*
2013-05-08 14:39:17 +04:00
* Important alert!
*
* If runp is not NULL, the number of contiguous blocks __starting from the
* next block after the queried block__ will be returned in runp.
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
*/
int
nilfs_trivial_bmap(void *v)
{
struct vop_bmap_args /* {
struct vnode *a_vp;
daddr_t a_bn;
struct vnode **a_vpp;
daddr_t *a_bnp;
int *a_runp;
} */ *ap = v;
struct vnode *vp = ap->a_vp; /* our node */
struct vnode **vpp = ap->a_vpp; /* return node */
daddr_t *bnp = ap->a_bnp; /* translated */
daddr_t bn = ap->a_bn; /* origional */
int *runp = ap->a_runp;
struct nilfs_node *node = VTOI(vp);
uint64_t *l2vmap;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
uint32_t blocksize;
int blks, run, error;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
DPRINTF(TRANSLATE, ("nilfs_bmap() called\n"));
/* XXX could return `-1' to indicate holes/zero's */
blocksize = node->nilfsdev->blocksize;
blks = MAXPHYS / blocksize;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* get mapping memory */
l2vmap = malloc(sizeof(uint64_t) * blks, M_TEMP, M_WAITOK);
/* get virtual block numbers for the vnode's buffer span */
error = nilfs_btree_nlookup(node, bn, blks, l2vmap);
if (error) {
free(l2vmap, M_TEMP);
return error;
}
/* store virtual blocks on our own vp */
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
if (vpp)
*vpp = vp;
/* start at virt[0] */
*bnp = l2vmap[0];
/* get runlength */
run = 1;
while ((run < blks) && (l2vmap[run] == *bnp + run))
run++;
run--; /* see comment at start of function */
/* set runlength */
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
if (runp)
*runp = run;
DPRINTF(TRANSLATE, ("\tstart %"PRIu64" -> %"PRIu64" run %d\n",
bn, *bnp, run));
/* mark not translated on virtual block number 0 */
if (*bnp == 0)
*bnp = -1;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* return success */
free(l2vmap, M_TEMP);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return 0;
}
/* --------------------------------------------------------------------- */
static void
nilfs_read_filebuf(struct nilfs_node *node, struct buf *bp)
{
struct nilfs_device *nilfsdev = node->nilfsdev;
struct buf *nbp;
uint64_t *l2vmap, *v2pmap;
uint64_t from, blks;
uint32_t blocksize, buf_offset;
uint8_t *buf_pos;
int blk2dev = nilfsdev->blocksize / DEV_BSIZE;
int i, error;
/*
* Translate all the block sectors into a series of buffers to read
* asynchronously from the nilfs device. Note that this lookup may
* induce readin's too.
*/
blocksize = nilfsdev->blocksize;
from = bp->b_blkno;
blks = bp->b_bcount / blocksize;
DPRINTF(READ, ("\tread in from inode %"PRIu64" blkno %"PRIu64" "
"+ %"PRIu64" blocks\n", node->ino, from, blks));
DPRINTF(READ, ("\t\tblkno %"PRIu64" "
"+ %d bytes\n", bp->b_blkno, bp->b_bcount));
/* get mapping memory */
l2vmap = malloc(sizeof(uint64_t) * blks, M_TEMP, M_WAITOK);
v2pmap = malloc(sizeof(uint64_t) * blks, M_TEMP, M_WAITOK);
/* get virtual block numbers for the vnode's buffer span */
for (i = 0; i < blks; i++)
l2vmap[i] = from + i;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* translate virtual block numbers to physical block numbers */
error = nilfs_nvtop(node, blks, l2vmap, v2pmap);
if (error)
goto out;
/* issue translated blocks */
bp->b_resid = bp->b_bcount;
for (i = 0; i < blks; i++) {
DPRINTF(READ, ("read_filebuf : ino %"PRIu64" blk %d -> "
"%"PRIu64" -> %"PRIu64"\n",
node->ino, i, l2vmap[i], v2pmap[i]));
buf_offset = i * blocksize;
buf_pos = (uint8_t *) bp->b_data + buf_offset;
/* note virtual block 0 marks not mapped */
if (l2vmap[i] == 0) {
memset(buf_pos, 0, blocksize);
nestiobuf_done(bp, blocksize, 0);
continue;
}
/* nest iobuf */
nbp = getiobuf(NULL, true);
nestiobuf_setup(bp, nbp, buf_offset, blocksize);
KASSERT(nbp->b_vp == node->vnode);
/* nbp is B_ASYNC */
nbp->b_lblkno = i;
nbp->b_blkno = v2pmap[i] * blk2dev; /* in DEV_BSIZE */
nbp->b_rawblkno = nbp->b_blkno;
VOP_STRATEGY(nilfsdev->devvp, nbp);
}
if ((bp->b_flags & B_ASYNC) == 0)
biowait(bp);
out:
free(l2vmap, M_TEMP);
free(v2pmap, M_TEMP);
if (error) {
bp->b_error = EIO;
biodone(bp);
}
}
static void
nilfs_write_filebuf(struct nilfs_node *node, struct buf *bp)
{
/* TODO pass on to segment collector */
panic("nilfs_strategy writing called\n");
}
int
nilfs_vfsstrategy(void *v)
{
struct vop_strategy_args /* {
struct vnode *a_vp;
struct buf *a_bp;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct buf *bp = ap->a_bp;
struct nilfs_node *node = VTOI(vp);
DPRINTF(STRATEGY, ("nilfs_strategy called\n"));
/* check if we ought to be here */
if (vp->v_type == VBLK || vp->v_type == VCHR)
panic("nilfs_strategy: spec");
/* translate if needed and pass on */
if (bp->b_flags & B_READ) {
nilfs_read_filebuf(node, bp);
return bp->b_error;
}
/* send to segment collector */
nilfs_write_filebuf(node, bp);
return bp->b_error;
}
/* --------------------------------------------------------------------- */
int
nilfs_readdir(void *v)
{
struct vop_readdir_args /* {
struct vnode *a_vp;
struct uio *a_uio;
kauth_cred_t a_cred;
int *a_eofflag;
off_t **a_cookies;
int *a_ncookies;
} */ *ap = v;
struct uio *uio = ap->a_uio;
struct vnode *vp = ap->a_vp;
struct nilfs_node *node = VTOI(vp);
struct nilfs_dir_entry *ndirent;
struct dirent dirent;
struct buf *bp;
uint64_t file_size, diroffset, transoffset, blkoff;
uint64_t blocknr;
uint32_t blocksize = node->nilfsdev->blocksize;
uint8_t *pos, name_len;
int error;
DPRINTF(READDIR, ("nilfs_readdir called\n"));
if (vp->v_type != VDIR)
return ENOTDIR;
file_size = nilfs_rw64(node->inode.i_size);
/* we are called just as long as we keep on pushing data in */
error = 0;
if ((uio->uio_offset < file_size) &&
(uio->uio_resid >= sizeof(struct dirent))) {
diroffset = uio->uio_offset;
transoffset = diroffset;
blocknr = diroffset / blocksize;
blkoff = diroffset % blocksize;
error = nilfs_bread(node, blocknr, NOCRED, 0, &bp);
if (error)
return EIO;
while (diroffset < file_size) {
DPRINTF(READDIR, ("readdir : offset = %"PRIu64"\n",
diroffset));
if (blkoff >= blocksize) {
blkoff = 0; blocknr++;
brelse(bp, BC_AGE);
error = nilfs_bread(node, blocknr, NOCRED, 0,
&bp);
if (error)
return EIO;
}
/* read in one dirent */
pos = (uint8_t *) bp->b_data + blkoff;
ndirent = (struct nilfs_dir_entry *) pos;
name_len = ndirent->name_len;
memset(&dirent, 0, sizeof(struct dirent));
dirent.d_fileno = nilfs_rw64(ndirent->inode);
dirent.d_type = ndirent->file_type; /* 1:1 ? */
dirent.d_namlen = name_len;
strncpy(dirent.d_name, ndirent->name, name_len);
dirent.d_reclen = _DIRENT_SIZE(&dirent);
DPRINTF(READDIR, ("copying `%*.*s`\n", name_len,
name_len, dirent.d_name));
/*
* If there isn't enough space in the uio to return a
* whole dirent, break off read
*/
if (uio->uio_resid < _DIRENT_SIZE(&dirent))
break;
/* transfer */
if (name_len)
uiomove(&dirent, _DIRENT_SIZE(&dirent), uio);
/* advance */
diroffset += nilfs_rw16(ndirent->rec_len);
blkoff += nilfs_rw16(ndirent->rec_len);
/* remember the last entry we transfered */
transoffset = diroffset;
}
brelse(bp, BC_AGE);
/* pass on last transfered offset */
uio->uio_offset = transoffset;
}
if (ap->a_eofflag)
*ap->a_eofflag = (uio->uio_offset >= file_size);
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_lookup(void *v)
{
struct vop_lookup_v2_args /* {
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode **vpp = ap->a_vpp;
struct componentname *cnp = ap->a_cnp;
struct mount *mp = dvp->v_mount;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
uint64_t ino;
const char *name;
int namelen, nameiop, islastcn, mounted_ro;
int vnodetp;
int error, found;
*vpp = NULL;
DPRINTF(LOOKUP, ("nilfs_lookup called\n"));
/* simplify/clarification flags */
nameiop = cnp->cn_nameiop;
islastcn = cnp->cn_flags & ISLASTCN;
mounted_ro = mp->mnt_flag & MNT_RDONLY;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* check exec/dirread permissions first */
error = VOP_ACCESS(dvp, VEXEC, cnp->cn_cred);
if (error)
return error;
DPRINTF(LOOKUP, ("\taccess ok\n"));
/*
* If requesting a modify on the last path element on a read-only
* filingsystem, reject lookup; XXX why is this repeated in every FS ?
*/
if (islastcn && mounted_ro && (nameiop == DELETE || nameiop == RENAME))
return EROFS;
DPRINTF(LOOKUP, ("\tlooking up cnp->cn_nameptr '%s'\n",
cnp->cn_nameptr));
/* look in the namecache */
if (cache_lookup(dvp, cnp->cn_nameptr, cnp->cn_namelen,
cnp->cn_nameiop, cnp->cn_flags, NULL, vpp)) {
return *vpp == NULLVP ? ENOENT : 0;
}
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
DPRINTF(LOOKUP, ("\tNOT found in cache\n"));
/*
* Obviously, the file is not (anymore) in the namecache, we have to
* search for it. There are three basic cases: '.', '..' and others.
*
* Following the guidelines of VOP_LOOKUP manpage and tmpfs.
*/
error = 0;
if ((cnp->cn_namelen == 1) && (cnp->cn_nameptr[0] == '.')) {
DPRINTF(LOOKUP, ("\tlookup '.'\n"));
/* special case 1 '.' */
vref(dvp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
*vpp = dvp;
/* done */
} else if (cnp->cn_flags & ISDOTDOT) {
/* special case 2 '..' */
DPRINTF(LOOKUP, ("\tlookup '..'\n"));
/* get our node */
name = "..";
namelen = 2;
error = nilfs_lookup_name_in_dir(dvp, name, namelen,
&ino, &found);
if (error)
goto out;
if (!found)
error = ENOENT;
if (error == 0) {
DPRINTF(LOOKUP, ("\tfound '..'\n"));
/* try to create/reuse the node */
2014-10-15 13:05:46 +04:00
error = vcache_get(mp, &ino, sizeof(ino), vpp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
if (!error) {
DPRINTF(LOOKUP,
("\tnode retrieved/created OK\n"));
}
}
} else {
DPRINTF(LOOKUP, ("\tlookup file\n"));
/* all other files */
/* lookup filename in the directory returning its inode */
name = cnp->cn_nameptr;
namelen = cnp->cn_namelen;
error = nilfs_lookup_name_in_dir(dvp, name, namelen,
&ino, &found);
if (error)
goto out;
if (!found) {
DPRINTF(LOOKUP, ("\tNOT found\n"));
/*
* UGH, didn't find name. If we're creating or
* renaming on the last name this is OK and we ought
* to return EJUSTRETURN if its allowed to be created.
*/
error = ENOENT;
if (islastcn &&
(nameiop == CREATE || nameiop == RENAME))
error = 0;
if (!error) {
error = VOP_ACCESS(dvp, VWRITE, cnp->cn_cred);
if (!error) {
error = EJUSTRETURN;
}
}
/* done */
} else {
/* try to create/reuse the node */
2014-10-15 13:05:46 +04:00
error = vcache_get(mp, &ino, sizeof(ino), vpp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
if (!error) {
/*
* If we are not at the last path component
* and found a non-directory or non-link entry
* (which may itself be pointing to a
* directory), raise an error.
*/
vnodetp = (*vpp)->v_type;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
if ((vnodetp != VDIR) && (vnodetp != VLNK)) {
if (!islastcn) {
vrele(*vpp);
*vpp = NULL;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
error = ENOTDIR;
}
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
}
}
}
}
out:
/*
* Store result in the cache if requested. If we are creating a file,
* the file might not be found and thus putting it into the namecache
* might be seen as negative caching.
*/
if (error == 0 && nameiop != CREATE)
cache_enter(dvp, *vpp, cnp->cn_nameptr, cnp->cn_namelen,
cnp->cn_flags);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
DPRINTFIF(LOOKUP, error, ("nilfs_lookup returing error %d\n", error));
if (error)
return error;
return 0;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
}
/* --------------------------------------------------------------------- */
static void
nilfs_ctime_to_timespec(struct timespec *ts, uint64_t ctime)
{
ts->tv_sec = ctime;
ts->tv_nsec = 0;
}
int
nilfs_getattr(void *v)
{
struct vop_getattr_args /* {
struct vnode *a_vp;
struct vattr *a_vap;
kauth_cred_t a_cred;
struct lwp *a_l;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct vattr *vap = ap->a_vap;
struct nilfs_node *node = VTOI(vp);
struct nilfs_inode *inode = &node->inode;
DPRINTF(VFSCALL, ("nilfs_getattr called\n"));
/* basic info */
vattr_null(vap);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
vap->va_type = vp->v_type;
vap->va_mode = nilfs_rw16(inode->i_mode) & ALLPERMS;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
vap->va_nlink = nilfs_rw16(inode->i_links_count);
vap->va_uid = nilfs_rw32(inode->i_uid);
vap->va_gid = nilfs_rw32(inode->i_gid);
vap->va_fsid = vp->v_mount->mnt_stat.f_fsidx.__fsid_val[0];
vap->va_fileid = node->ino;
vap->va_size = nilfs_rw64(inode->i_size);
vap->va_blocksize = node->nilfsdev->blocksize;
/* times */
nilfs_ctime_to_timespec(&vap->va_atime, nilfs_rw64(inode->i_mtime));
nilfs_ctime_to_timespec(&vap->va_mtime, nilfs_rw64(inode->i_mtime));
nilfs_ctime_to_timespec(&vap->va_ctime, nilfs_rw64(inode->i_ctime));
nilfs_ctime_to_timespec(&vap->va_birthtime, nilfs_rw64(inode->i_ctime));
vap->va_gen = nilfs_rw32(inode->i_generation);
vap->va_flags = 0; /* vattr flags */
vap->va_bytes = nilfs_rw64(inode->i_blocks) * vap->va_blocksize;
vap->va_filerev = vap->va_gen; /* XXX file revision? same as gen? */
vap->va_vaflags = 0; /* XXX chflags flags */
return 0;
}
/* --------------------------------------------------------------------- */
#if 0
static int
nilfs_chown(struct vnode *vp, uid_t new_uid, gid_t new_gid,
kauth_cred_t cred)
{
return EINVAL;
}
static int
nilfs_chmod(struct vnode *vp, mode_t mode, kauth_cred_t cred)
{
return EINVAL;
}
/* exported */
int
nilfs_chsize(struct vnode *vp, u_quad_t newsize, kauth_cred_t cred)
{
return EINVAL;
}
static int
nilfs_chflags(struct vnode *vp, mode_t mode, kauth_cred_t cred)
{
return EINVAL;
}
static int
nilfs_chtimes(struct vnode *vp,
struct timespec *atime, struct timespec *mtime,
struct timespec *birthtime, int setattrflags,
kauth_cred_t cred)
{
return EINVAL;
}
#endif
int
nilfs_setattr(void *v)
{
struct vop_setattr_args /* {
struct vnode *a_vp;
struct vattr *a_vap;
kauth_cred_t a_cred;
struct lwp *a_l;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
vp = vp;
DPRINTF(VFSCALL, ("nilfs_setattr called\n"));
return EINVAL;
}
/* --------------------------------------------------------------------- */
/*
* Return POSIX pathconf information for NILFS file systems.
*/
int
nilfs_pathconf(void *v)
{
struct vop_pathconf_args /* {
struct vnode *a_vp;
int a_name;
register_t *a_retval;
} */ *ap = v;
uint32_t bits;
DPRINTF(VFSCALL, ("nilfs_pathconf called\n"));
switch (ap->a_name) {
case _PC_LINK_MAX:
*ap->a_retval = (1<<16)-1; /* 16 bits */
return 0;
case _PC_NAME_MAX:
2011-09-27 05:34:41 +04:00
*ap->a_retval = NILFS_MAXNAMLEN;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return 0;
case _PC_PATH_MAX:
*ap->a_retval = PATH_MAX;
return 0;
case _PC_PIPE_BUF:
*ap->a_retval = PIPE_BUF;
return 0;
case _PC_CHOWN_RESTRICTED:
*ap->a_retval = 1;
return 0;
case _PC_NO_TRUNC:
*ap->a_retval = 1;
return 0;
case _PC_SYNC_IO:
*ap->a_retval = 0; /* synchronised is off for performance */
return 0;
case _PC_FILESIZEBITS:
/* 64 bit file offsets -> 2+floor(2log(2^64-1)) = 2 + 63 = 65 */
bits = 64; /* XXX ought to deliver 65 */
#if 0
if (nilfs_node)
bits = 64 * vp->v_mount->mnt_dev_bshift;
#endif
*ap->a_retval = bits;
return 0;
}
return EINVAL;
}
/* --------------------------------------------------------------------- */
int
nilfs_open(void *v)
{
struct vop_open_args /* {
struct vnode *a_vp;
int a_mode;
kauth_cred_t a_cred;
struct proc *a_p;
} */ *ap = v;
int flags;
DPRINTF(VFSCALL, ("nilfs_open called\n"));
/*
* Files marked append-only must be opened for appending.
*/
flags = 0;
if ((flags & APPEND) && (ap->a_mode & (FWRITE | O_APPEND)) == FWRITE)
return (EPERM);
return 0;
}
/* --------------------------------------------------------------------- */
int
nilfs_close(void *v)
{
struct vop_close_args /* {
struct vnode *a_vp;
int a_fflag;
kauth_cred_t a_cred;
struct proc *a_p;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct nilfs_node *nilfs_node = VTOI(vp);
DPRINTF(VFSCALL, ("nilfs_close called\n"));
nilfs_node = nilfs_node; /* shut up gcc */
mutex_enter(vp->v_interlock);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
if (vp->v_usecount > 1)
nilfs_itimes(nilfs_node, NULL, NULL, NULL);
mutex_exit(vp->v_interlock);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return 0;
}
/* --------------------------------------------------------------------- */
static int
nilfs_check_possible(struct vnode *vp, struct vattr *vap, mode_t mode)
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
{
int flags;
/* check if we are allowed to write */
switch (vap->va_type) {
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
case VDIR:
case VLNK:
case VREG:
/*
* normal nodes: check if we're on a read-only mounted
* filingsystem and bomb out if we're trying to write.
*/
if ((mode & VWRITE) && (vp->v_mount->mnt_flag & MNT_RDONLY))
return EROFS;
break;
case VBLK:
case VCHR:
case VSOCK:
case VFIFO:
/*
* special nodes: even on read-only mounted filingsystems
* these are allowed to be written to if permissions allow.
*/
break;
default:
/* no idea what this is */
return EINVAL;
}
/* noone may write immutable files */
/* TODO: get chflags(2) flags */
flags = 0;
if ((mode & VWRITE) && (flags & IMMUTABLE))
return EPERM;
return 0;
}
static int
nilfs_check_permitted(struct vnode *vp, struct vattr *vap, mode_t mode,
kauth_cred_t cred)
{
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* ask the generic genfs_can_access to advice on security */
return kauth_authorize_vnode(cred, KAUTH_ACCESS_ACTION(mode,
vp->v_type, vap->va_mode), vp, NULL, genfs_can_access(vp->v_type,
vap->va_mode, vap->va_uid, vap->va_gid, mode, cred));
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
}
int
nilfs_access(void *v)
{
struct vop_access_args /* {
struct vnode *a_vp;
int a_mode;
kauth_cred_t a_cred;
struct proc *a_p;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
mode_t mode = ap->a_mode;
kauth_cred_t cred = ap->a_cred;
/* struct nilfs_node *nilfs_node = VTOI(vp); */
struct vattr vap;
int error;
DPRINTF(VFSCALL, ("nilfs_access called\n"));
error = VOP_GETATTR(vp, &vap, NULL);
if (error)
return error;
error = nilfs_check_possible(vp, &vap, mode);
if (error)
return error;
error = nilfs_check_permitted(vp, &vap, mode, cred);
return error;
}
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* --------------------------------------------------------------------- */
int
nilfs_create(void *v)
{
struct vop_create_v3_args /* {
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
struct vattr *a_vap;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode **vpp = ap->a_vpp;
struct vattr *vap = ap->a_vap;
struct componentname *cnp = ap->a_cnp;
int error;
DPRINTF(VFSCALL, ("nilfs_create called\n"));
error = nilfs_create_node(dvp, vpp, vap, cnp);
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_mknod(void *v)
{
struct vop_mknod_v3_args /* {
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
struct vattr *a_vap;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode **vpp = ap->a_vpp;
struct vattr *vap = ap->a_vap;
struct componentname *cnp = ap->a_cnp;
int error;
DPRINTF(VFSCALL, ("nilfs_mknod called\n"));
error = nilfs_create_node(dvp, vpp, vap, cnp);
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_mkdir(void *v)
{
struct vop_mkdir_v3_args /* {
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
struct vattr *a_vap;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode **vpp = ap->a_vpp;
struct vattr *vap = ap->a_vap;
struct componentname *cnp = ap->a_cnp;
int error;
DPRINTF(VFSCALL, ("nilfs_mkdir called\n"));
error = nilfs_create_node(dvp, vpp, vap, cnp);
return error;
}
/* --------------------------------------------------------------------- */
static int
nilfs_do_link(struct vnode *dvp, struct vnode *vp, struct componentname *cnp)
{
struct nilfs_node *nilfs_node, *dir_node;
struct vattr vap;
int error;
DPRINTF(VFSCALL, ("nilfs_link called\n"));
KASSERT(dvp != vp);
KASSERT(vp->v_type != VDIR);
KASSERT(dvp->v_mount == vp->v_mount);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* lock node */
error = vn_lock(vp, LK_EXCLUSIVE);
if (error)
return error;
/* get attributes */
dir_node = VTOI(dvp);
nilfs_node = VTOI(vp);
error = VOP_GETATTR(vp, &vap, FSCRED);
if (error) {
VOP_UNLOCK(vp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return error;
}
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
/* check link count overflow */
if (vap.va_nlink >= (1<<16)-1) { /* uint16_t */
VOP_UNLOCK(vp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
return EMLINK;
}
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
error = nilfs_dir_attach(dir_node->ump, dir_node, nilfs_node,
&vap, cnp);
if (error)
VOP_UNLOCK(vp);
return error;
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
}
int
nilfs_link(void *v)
{
struct vop_link_args /* {
struct vnode *a_dvp;
struct vnode *a_vp;
struct componentname *a_cnp;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode *vp = ap->a_vp;
struct componentname *cnp = ap->a_cnp;
int error;
error = nilfs_do_link(dvp, vp, cnp);
if (error)
VOP_ABORTOP(dvp, cnp);
VN_KNOTE(vp, NOTE_LINK);
VN_KNOTE(dvp, NOTE_WRITE);
vput(dvp);
return error;
}
/* --------------------------------------------------------------------- */
static int
nilfs_do_symlink(struct nilfs_node *nilfs_node, char *target)
{
return EROFS;
}
int
nilfs_symlink(void *v)
{
struct vop_symlink_v3_args /* {
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct vnode *a_dvp;
struct vnode **a_vpp;
struct componentname *a_cnp;
struct vattr *a_vap;
char *a_target;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode **vpp = ap->a_vpp;
struct vattr *vap = ap->a_vap;
struct componentname *cnp = ap->a_cnp;
struct nilfs_node *dir_node;
struct nilfs_node *nilfs_node;
int error;
DPRINTF(VFSCALL, ("nilfs_symlink called\n"));
DPRINTF(VFSCALL, ("\tlinking to `%s`\n", ap->a_target));
error = nilfs_create_node(dvp, vpp, vap, cnp);
KASSERT(((error == 0) && (*vpp != NULL)) || ((error && (*vpp == NULL))));
if (!error) {
dir_node = VTOI(dvp);
nilfs_node = VTOI(*vpp);
KASSERT(nilfs_node);
error = nilfs_do_symlink(nilfs_node, ap->a_target);
if (error) {
/* remove node */
nilfs_shrink_node(nilfs_node, 0);
nilfs_dir_detach(nilfs_node->ump, dir_node, nilfs_node, cnp);
}
}
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_readlink(void *v)
{
struct vop_readlink_args /* {
struct vnode *a_vp;
struct uio *a_uio;
kauth_cred_t a_cred;
} */ *ap = v;
#if 0
struct vnode *vp = ap->a_vp;
struct uio *uio = ap->a_uio;
kauth_cred_t cred = ap->a_cred;
struct nilfs_node *nilfs_node;
struct pathcomp pathcomp;
struct vattr vattr;
uint8_t *pathbuf, *targetbuf, *tmpname;
uint8_t *pathpos, *targetpos;
char *mntonname;
int pathlen, targetlen, namelen, mntonnamelen, len, l_ci;
int first, error;
#endif
ap = ap;
DPRINTF(VFSCALL, ("nilfs_readlink called\n"));
return EROFS;
}
/* --------------------------------------------------------------------- */
/* note: i tried to follow the logics of the tmpfs rename code */
int
nilfs_rename(void *v)
{
struct vop_rename_args /* {
struct vnode *a_fdvp;
struct vnode *a_fvp;
struct componentname *a_fcnp;
struct vnode *a_tdvp;
struct vnode *a_tvp;
struct componentname *a_tcnp;
} */ *ap = v;
struct vnode *tvp = ap->a_tvp;
struct vnode *tdvp = ap->a_tdvp;
struct vnode *fvp = ap->a_fvp;
struct vnode *fdvp = ap->a_fdvp;
struct componentname *tcnp = ap->a_tcnp;
struct componentname *fcnp = ap->a_fcnp;
struct nilfs_node *fnode, *fdnode, *tnode, *tdnode;
struct vattr fvap, tvap;
int error;
DPRINTF(VFSCALL, ("nilfs_rename called\n"));
/* disallow cross-device renames */
if (fvp->v_mount != tdvp->v_mount ||
(tvp != NULL && fvp->v_mount != tvp->v_mount)) {
error = EXDEV;
goto out_unlocked;
}
fnode = VTOI(fvp);
fdnode = VTOI(fdvp);
tnode = (tvp == NULL) ? NULL : VTOI(tvp);
tdnode = VTOI(tdvp);
/* lock our source dir */
if (fdnode != tdnode) {
error = vn_lock(fdvp, LK_EXCLUSIVE | LK_RETRY);
if (error != 0)
goto out_unlocked;
}
/* get info about the node to be moved */
vn_lock(fvp, LK_SHARED | LK_RETRY);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
error = VOP_GETATTR(fvp, &fvap, FSCRED);
VOP_UNLOCK(fvp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
KASSERT(error == 0);
/* check when to delete the old already existing entry */
if (tvp) {
/* get info about the node to be moved to */
error = VOP_GETATTR(tvp, &tvap, FSCRED);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
KASSERT(error == 0);
/* if both dirs, make sure the destination is empty */
if (fvp->v_type == VDIR && tvp->v_type == VDIR) {
if (tvap.va_nlink > 2) {
error = ENOTEMPTY;
goto out;
}
}
/* if moving dir, make sure destination is dir too */
if (fvp->v_type == VDIR && tvp->v_type != VDIR) {
error = ENOTDIR;
goto out;
}
/* if we're moving a non-directory, make sure dest is no dir */
if (fvp->v_type != VDIR && tvp->v_type == VDIR) {
error = EISDIR;
goto out;
}
}
/* dont allow renaming directories acros directory for now */
if (fdnode != tdnode) {
if (fvp->v_type == VDIR) {
error = EINVAL;
goto out;
}
}
/* remove existing entry if present */
if (tvp)
nilfs_dir_detach(tdnode->ump, tdnode, tnode, tcnp);
/* create new directory entry for the node */
error = nilfs_dir_attach(tdnode->ump, tdnode, fnode, &fvap, tcnp);
if (error)
goto out;
/* unlink old directory entry for the node, if failing, unattach new */
error = nilfs_dir_detach(tdnode->ump, fdnode, fnode, fcnp);
if (error)
nilfs_dir_detach(tdnode->ump, tdnode, fnode, tcnp);
out:
if (fdnode != tdnode)
VOP_UNLOCK(fdvp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
out_unlocked:
VOP_ABORTOP(tdvp, tcnp);
if (tdvp == tvp)
vrele(tdvp);
else
vput(tdvp);
if (tvp)
vput(tvp);
VOP_ABORTOP(fdvp, fcnp);
/* release source nodes. */
vrele(fdvp);
vrele(fvp);
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_remove(void *v)
{
struct vop_remove_args /* {
struct vnode *a_dvp;
struct vnode *a_vp;
struct componentname *a_cnp;
} */ *ap = v;
struct vnode *dvp = ap->a_dvp;
struct vnode *vp = ap->a_vp;
struct componentname *cnp = ap->a_cnp;
2010-01-05 16:30:10 +03:00
struct nilfs_node *dir_node = VTOI(dvp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct nilfs_node *nilfs_node = VTOI(vp);
struct nilfs_mount *ump = dir_node->ump;
int error;
DPRINTF(VFSCALL, ("nilfs_remove called\n"));
if (vp->v_type != VDIR) {
error = nilfs_dir_detach(ump, dir_node, nilfs_node, cnp);
DPRINTFIF(NODE, error, ("\tgot error removing file\n"));
} else {
DPRINTF(NODE, ("\tis a directory: perm. denied\n"));
error = EPERM;
}
if (error == 0) {
VN_KNOTE(vp, NOTE_DELETE);
VN_KNOTE(dvp, NOTE_WRITE);
}
if (dvp == vp)
vrele(vp);
else
vput(vp);
vput(dvp);
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_rmdir(void *v)
{
struct vop_rmdir_args /* {
struct vnode *a_dvp;
struct vnode *a_vp;
struct componentname *a_cnp;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct vnode *dvp = ap->a_dvp;
struct componentname *cnp = ap->a_cnp;
2010-01-05 16:30:10 +03:00
struct nilfs_node *dir_node = VTOI(dvp);
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
struct nilfs_node *nilfs_node = VTOI(vp);
struct nilfs_mount *ump = dir_node->ump;
int refcnt, error;
DPRINTF(NOTIMPL, ("nilfs_rmdir called\n"));
/* don't allow '.' to be deleted */
if (dir_node == nilfs_node) {
vrele(dvp);
vput(vp);
return EINVAL;
}
/* check to see if the directory is empty */
error = 0;
refcnt = 2; /* XXX */
if (refcnt > 1) {
/* NOT empty */
vput(dvp);
vput(vp);
return ENOTEMPTY;
}
/* detach the node from the directory */
error = nilfs_dir_detach(ump, dir_node, nilfs_node, cnp);
if (error == 0) {
cache_purge(vp);
// cache_purge(dvp); /* XXX from msdosfs, why? */
VN_KNOTE(vp, NOTE_DELETE);
}
DPRINTFIF(NODE, error, ("\tgot error removing file\n"));
/* unput the nodes and exit */
vput(dvp);
vput(vp);
return error;
}
/* --------------------------------------------------------------------- */
int
nilfs_fsync(void *v)
{
struct vop_fsync_args /* {
struct vnode *a_vp;
kauth_cred_t a_cred;
int a_flags;
off_t offlo;
off_t offhi;
struct proc *a_p;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
// struct nilfs_node *nilfs_node = VTOI(vp);
// int error, flags, wait;
DPRINTF(STRATEGY, ("nilfs_fsync called : %s, %s\n",
(ap->a_flags & FSYNC_WAIT) ? "wait":"no wait",
(ap->a_flags & FSYNC_DATAONLY) ? "data_only":"complete"));
vp = vp;
return 0;
}
/* --------------------------------------------------------------------- */
int
nilfs_advlock(void *v)
{
struct vop_advlock_args /* {
struct vnode *a_vp;
void *a_id;
int a_op;
struct flock *a_fl;
int a_flags;
} */ *ap = v;
struct vnode *vp = ap->a_vp;
struct nilfs_node *nilfs_node = VTOI(vp);
uint64_t file_size;
DPRINTF(LOCKING, ("nilfs_advlock called\n"));
assert(nilfs_node);
file_size = nilfs_rw64(nilfs_node->inode.i_size);
return lf_advlock(ap, &nilfs_node->lockf, file_size);
}
/* --------------------------------------------------------------------- */
/* Global vfs vnode data structures for nilfss */
int (**nilfs_vnodeop_p) __P((void *));
const struct vnodeopv_entry_desc nilfs_vnodeop_entries[] = {
{ &vop_default_desc, vn_default_error },
{ &vop_lookup_desc, nilfs_lookup }, /* lookup */
{ &vop_create_desc, nilfs_create }, /* create */
{ &vop_mknod_desc, nilfs_mknod }, /* mknod */ /* TODO */
{ &vop_open_desc, nilfs_open }, /* open */
{ &vop_close_desc, nilfs_close }, /* close */
{ &vop_access_desc, nilfs_access }, /* access */
{ &vop_getattr_desc, nilfs_getattr }, /* getattr */
{ &vop_setattr_desc, nilfs_setattr }, /* setattr */ /* TODO chflags */
{ &vop_read_desc, nilfs_read }, /* read */
{ &vop_write_desc, nilfs_write }, /* write */ /* WRITE */
{ &vop_fallocate_desc, genfs_eopnotsupp }, /* fallocate */
{ &vop_fdiscard_desc, genfs_eopnotsupp }, /* fdiscard */
Import read-only part of the NiLFS (v2) implementation for NetBSD. It has been tested with a DEBUG+DIAGNOSTIC+LOCKDEBUG kernel. To summerise NiLFS, i'll repeat my posting to tech-kern here: NiLFS stands for New implementation of Logging File System; LFS done right they claim :) It is at version 2 now and is being developed by NTT, the Japanese telecom company and recently put into the linux source tree. See http://www.nilfs.org. The on-disc format is not completely frozen and i expect at least one minor revision to come in time. The benefits of NiLFS are build-in fine-grained checkpointing, persistent snapshots, multiple mounts and very large file and media support. Every checkpoint can be transformed into a snapshot and v.v. It is said to perform very well on flash media since it is not overwriting pieces apart from a incidental update of the superblock, but that might change. It is accompanied by a cleaner to clean up the segments and recover lost space. My work is not a port of the linux code; its a new implementation. Porting the code would be more work since its very linux oriented and never written to be ported outside linux. The goal is to be fully interchangable. The code is non intrusive to other parts of the kernel. It is also very light-weight. The current state of the code is read-only access to both clean and dirty NiLFS partitions. On mounting a dirty partition it rolls forward the log to the last checkpoint. Full read-write support is however planned! Just as the linux code, mount_nilfs allows for the `head' to be mounted read/write and allows multiple read-only snapshots/checkpoint mounts next to it. By allowing the RW mount at a different snapshot for read-write it should be possible eventually to revert back to a previous state; i.e. try to upgrade a system and being able to revert to the exact state prior to the upgrade. Compared to other FS's its pretty light-weight, suitable for embedded use and on flash media. The read-only code is currently 17kb object code on NetBSD/i386. I doubt the read-write code will surpass the 50 or 60. Compared this to FFS being 156kb, UDF being 84 kb and NFS being 130kb. Run-time memory usage is most likely not very different from other uses though maybe a bit higher than FFS.
2009-07-18 20:31:41 +04:00
{ &vop_fcntl_desc, genfs_fcntl }, /* fcntl */ /* TODO? */
{ &vop_ioctl_desc, genfs_enoioctl }, /* ioctl */ /* TODO? */
{ &vop_poll_desc, genfs_poll }, /* poll */ /* TODO/OK? */
{ &vop_kqfilter_desc, genfs_kqfilter }, /* kqfilter */ /* ? */
{ &vop_revoke_desc, genfs_revoke }, /* revoke */ /* TODO? */
{ &vop_mmap_desc, genfs_mmap }, /* mmap */ /* OK? */
{ &vop_fsync_desc, nilfs_fsync }, /* fsync */
{ &vop_seek_desc, genfs_seek }, /* seek */
{ &vop_remove_desc, nilfs_remove }, /* remove */
{ &vop_link_desc, nilfs_link }, /* link */ /* TODO */
{ &vop_rename_desc, nilfs_rename }, /* rename */ /* TODO */
{ &vop_mkdir_desc, nilfs_mkdir }, /* mkdir */
{ &vop_rmdir_desc, nilfs_rmdir }, /* rmdir */
{ &vop_symlink_desc, nilfs_symlink }, /* symlink */ /* TODO */
{ &vop_readdir_desc, nilfs_readdir }, /* readdir */
{ &vop_readlink_desc, nilfs_readlink }, /* readlink */ /* TEST ME */
{ &vop_abortop_desc, genfs_abortop }, /* abortop */ /* TODO/OK? */
{ &vop_inactive_desc, nilfs_inactive }, /* inactive */
{ &vop_reclaim_desc, nilfs_reclaim }, /* reclaim */
{ &vop_lock_desc, genfs_lock }, /* lock */
{ &vop_unlock_desc, genfs_unlock }, /* unlock */
{ &vop_bmap_desc, nilfs_trivial_bmap }, /* bmap */ /* 1:1 bmap */
{ &vop_strategy_desc, nilfs_vfsstrategy },/* strategy */
/* { &vop_print_desc, nilfs_print }, */ /* print */
{ &vop_islocked_desc, genfs_islocked }, /* islocked */
{ &vop_pathconf_desc, nilfs_pathconf }, /* pathconf */
{ &vop_advlock_desc, nilfs_advlock }, /* advlock */ /* TEST ME */
{ &vop_bwrite_desc, vn_bwrite }, /* bwrite */ /* ->strategy */
{ &vop_getpages_desc, genfs_getpages }, /* getpages */
{ &vop_putpages_desc, genfs_putpages }, /* putpages */
{ NULL, NULL }
};
const struct vnodeopv_desc nilfs_vnodeop_opv_desc = {
&nilfs_vnodeop_p, nilfs_vnodeop_entries
};