Aren/NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
953092f8c6
NetBSD/lib/libpuffs/puffs.h

672 lines
24 KiB
C
Raw Normal View History

Get rid of the "int *refs" parameter to inactive: the same can be accomplished now with puffs_setbacks.
2007-07-02 14:24:17 +04:00
/* $NetBSD: puffs.h,v 1.70 2007/07/02 10:24:18 pooka Exp $ */
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
/*
Introduce PUFFS_FLAG_HASHPATH, which for PUFFS_FLAG_BUILDPATH file servers creates a hash of the node's path into po->po_hash to avoid a full pathcmp every time.
2007-05-01 19:58:00 +04:00
* Copyright (c) 2005, 2006, 2007 Antti Kantee. All Rights Reserved.
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
*
* Development of this software was supported by the
* Google Summer of Code program and the Ulla Tuominen Foundation.
* The Google SoC project was mentored by Bill Studenmund.
*
* 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 OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
* SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
#ifndef _PUFFS_H_
#define _PUFFS_H_
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
#include <sys/param.h>
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
#include <sys/types.h>
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
#include <sys/mount.h>
Add preliminary code for a nullfs layer, which can be used to mount a directory hierarchy to another point, just like with the kernel nullfs. This is not really a layering scheme yet, but it should evolve into one. Currently it can just be used to do 1:1 mapping.
2007-01-11 04:01:55 +03:00
#include <sys/stat.h>
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
#include <sys/statvfs.h>
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
#include <sys/time.h>
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
#include <sys/vnode.h>
#include <fs/puffs/puffs_msgif.h>
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
#include <mntopts.h>
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
#include <string.h>
/* forwards */
struct puffs_cc;
struct puffs_getreq;
struct puffs_putreq;
make puffs_cred an opaque type
2007-07-01 19:30:15 +04:00
struct puffs_cred;
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
struct puffs_cid;
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
struct puffs_newinfo;
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
/* paths */
struct puffs_pathobj {
Introduce PUFFS_FLAG_HASHPATH, which for PUFFS_FLAG_BUILDPATH file servers creates a hash of the node's path into po->po_hash to avoid a full pathcmp every time.
2007-05-01 19:58:00 +04:00
void *po_path;
size_t po_len;
uint32_t po_hash;
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
};
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
/* for prefix rename */
struct puffs_pathinfo {
struct puffs_pathobj *pi_old;
struct puffs_pathobj *pi_new;
};
initial support for cacheops
2007-03-20 13:22:22 +03:00
/* describes one segment cached in the kernel */
struct puffs_kcache {
off_t pkc_start;
off_t pkc_end;
LIST_ENTRY(puffs_kcache) pkc_entries;
};
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
/* XXX: might disappear from here into a private header */
struct puffs_node {
off_t pn_size;
Introduce puffs_pn_remove() to signal that a node has been deleted and destroyed, but not yet reclaimed. This prevents puffs_pn_nodewalk() from returning stale entries. Make nullfs use this (some file systems are a bit too happy with recycling inode numbers).
2007-06-24 21:55:07 +04:00
int pn_flags;
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct vattr pn_va;
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
void *pn_data; /* private data */
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct puffs_pathobj pn_po; /* PUFFS_FLAG_BUILDPATH */
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct puffs_usermount *pn_mnt;
LIST_ENTRY(puffs_node) pn_entries;
initial support for cacheops
2007-03-20 13:22:22 +03:00
LIST_HEAD(,puffs_kcache)pn_cacheinfo; /* PUFFS_KFLAG_CACHE */
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
};
Introduce puffs_pn_remove() to signal that a node has been deleted and destroyed, but not yet reclaimed. This prevents puffs_pn_nodewalk() from returning stale entries. Make nullfs use this (some file systems are a bit too happy with recycling inode numbers).
2007-06-24 21:55:07 +04:00
#define PUFFS_NODE_REMOVED 0x01 /* not on entry list */
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
struct puffs_usermount;
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
/*
* megaXXX: these are values from inside _KERNEL
* need to work on the translation for ALL the necessary values.
*/
#define PUFFS_VNOVAL (-1)
#define PUFFS_IO_APPEND 0x20
implement counterpart of vaccess()
2007-03-20 21:28:08 +03:00
#define PUFFS_VEXEC 01
#define PUFFS_VWRITE 02
#define PUFFS_VREAD 04
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
implement fsync
2007-01-11 20:48:21 +03:00
#define PUFFS_FSYNC_DATAONLY 0x0002
#define PUFFS_FSYNC_CACHE 0x0100
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
/*
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
* Magic constants
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
*/
Be less shy about the default stack size, it's a only virtual memory space. TODO: put a guard page at the end
2007-01-10 23:11:04 +03:00
#define PUFFS_CC_STACKSIZE_DEFAULT (1024*1024)
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
struct puffs_cn {
struct puffs_kcn *pcn_pkcnp; /* kernel input */
make puffs_cred an opaque type
2007-07-01 19:30:15 +04:00
struct puffs_cred *pcn_cred; /* cred used for lookup */
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
struct puffs_cid *pcn_cid; /* the who called */
make puffs_cred an opaque type
2007-07-01 19:30:15 +04:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct puffs_pathobj pcn_po_full; /* PUFFS_FLAG_BUILDPATH */
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
};
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
#define pcn_nameiop pcn_pkcnp->pkcn_nameiop
#define pcn_flags pcn_pkcnp->pkcn_flags
#define pcn_pid pcn_pkcnp->pkcn_pid
#define pcn_name pcn_pkcnp->pkcn_name
#define pcn_namelen pcn_pkcnp->pkcn_namelen
Get rid of the "int *refs" parameter to inactive: the same can be accomplished now with puffs_setbacks.
2007-07-02 14:24:17 +04:00
#define pcn_consume pcn_pkcnp->pkcn_consume
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
/*
* Puffs options to mount
*/
/* kernel */
PUFFS_KFLAG_NOCACHE became two, so introduce the command line options -o nonamecache and -o nopagecache and adapt file systems where necessary.
2007-06-25 02:18:38 +04:00
#define PUFFSMOPT_NAMECACHE { "namecache", 1, PUFFS_KFLAG_NOCACHE_NAME, 1 }
#define PUFFSMOPT_PAGECACHE { "pagecache", 1, PUFFS_KFLAG_NOCACHE_PAGE, 1 }
Actually, keep PUFFS_KFLAG_NOCACHE and -o cache around as shorthand to neither page- nor namecache.
2007-06-25 02:25:49 +04:00
#define PUFFSMOPT_CACHE { "cache", 1, PUFFS_KFLAG_NOCACHE, 1 }
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
#define PUFFSMOPT_ALLOPS { "allops", 0, PUFFS_KFLAG_ALLOPS, 1 }
/* libpuffs */
#define PUFFSMOPT_DUMP { "dump", 0, PUFFS_FLAG_OPDUMP, 1 }
#define PUFFSMOPT_STD \
PUFFS_KFLAG_NOCACHE became two, so introduce the command line options -o nonamecache and -o nopagecache and adapt file systems where necessary.
2007-06-25 02:18:38 +04:00
PUFFSMOPT_NAMECACHE, \
PUFFSMOPT_PAGECACHE, \
Actually, keep PUFFS_KFLAG_NOCACHE and -o cache around as shorthand to neither page- nor namecache.
2007-06-25 02:25:49 +04:00
PUFFSMOPT_CACHE, \
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
PUFFSMOPT_ALLOPS, \
PUFFSMOPT_DUMP
extern const struct mntopt puffsmopts[]; /* puffs.c */
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
/* callbacks for operations */
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
struct puffs_ops {
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
int (*puffs_fs_unmount)(struct puffs_cc *, int,
const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_fs_statvfs)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
struct statvfs *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_fs_sync)(struct puffs_cc *, int,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
const struct puffs_cred *, const struct puffs_cid *);
* support nodetofh and fhtonode callbacks * use new signature of node_readdir
2007-04-12 01:04:51 +04:00
int (*puffs_fs_fhtonode)(struct puffs_cc *, void *, size_t,
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
struct puffs_newinfo *);
* support nodetofh and fhtonode callbacks * use new signature of node_readdir
2007-04-12 01:04:51 +04:00
int (*puffs_fs_nodetofh)(struct puffs_cc *, void *cookie,
void *, size_t *);
add interfaces for suspending the file system
2007-01-27 02:00:33 +03:00
void (*puffs_fs_suspend)(struct puffs_cc *, int);
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_lookup)(struct puffs_cc *,
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_create)(struct puffs_cc *,
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *,
const struct vattr *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_mknod)(struct puffs_cc *,
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *,
const struct vattr *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_open)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_close)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_access)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_getattr)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, struct vattr *, const struct puffs_cred *,
const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_setattr)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, const struct vattr *, const struct puffs_cred *,
const struct puffs_cid *);
int (*puffs_node_poll)(struct puffs_cc *, void *, int *,
const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_mmap)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_fsync)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, const struct puffs_cred *, int, off_t, off_t,
const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_seek)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, off_t, off_t, const struct puffs_cred *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_remove)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_link)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_rename)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *, void *, void *,
const struct puffs_cn *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_mkdir)(struct puffs_cc *,
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *,
const struct vattr *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_rmdir)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_symlink)(struct puffs_cc *,
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *,
const struct vattr *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
const char *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_readdir)(struct puffs_cc *,
* support nodetofh and fhtonode callbacks * use new signature of node_readdir
2007-04-12 01:04:51 +04:00
void *, struct dirent *, off_t *, size_t *,
const struct puffs_cred *, int *, off_t *, size_t *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_readlink)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, const struct puffs_cred *, char *, size_t *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_reclaim)(struct puffs_cc *,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_inactive)(struct puffs_cc *,
Get rid of the "int *refs" parameter to inactive: the same can be accomplished now with puffs_setbacks.
2007-07-02 14:24:17 +04:00
void *, const struct puffs_cid *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_print)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_pathconf)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, int, int *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_advlock)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, int, struct flock *, int);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_getextattr)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, struct puffs_vnreq_getextattr *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_setextattr)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, struct puffs_vnreq_setextattr *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_listextattr)(struct puffs_cc *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, struct puffs_vnreq_listextattr *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_read)(struct puffs_cc *, void *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
uint8_t *, off_t, size_t *, const struct puffs_cred *, int);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int (*puffs_node_write)(struct puffs_cc *, void *,
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
uint8_t *, off_t, size_t *, const struct puffs_cred *, int);
initial support for cacheops
2007-03-20 13:22:22 +03:00
int (*puffs_cache_write)(struct puffs_usermount *,
void *, size_t, struct puffs_cacherun *);
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
};
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
typedef int (*pu_pathbuild_fn)(struct puffs_usermount *,
const struct puffs_pathobj *,
const struct puffs_pathobj *, size_t,
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct puffs_pathobj *);
typedef int (*pu_pathtransform_fn)(struct puffs_usermount *,
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
const struct puffs_pathobj *,
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
const struct puffs_cn *,
struct puffs_pathobj *);
typedef int (*pu_pathcmp_fn)(struct puffs_usermount *, struct puffs_pathobj *,
Introduce puffs_path_walkcmp(), which is to be called from nodewalk and compares the path of the node against the given pathobject. Also make comparison method take a flag to indicate if it should check if the second path is a true prefix of the first. plus some namespace cleanup
2007-02-15 20:04:46 +03:00
struct puffs_pathobj *, size_t, int);
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
typedef void (*pu_pathfree_fn)(struct puffs_usermount *,
struct puffs_pathobj *);
typedef int (*pu_namemod_fn)(struct puffs_usermount *,
struct puffs_pathobj *, struct puffs_cn *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
enum {
Reflect kernel changes for supplying root node information in mount instead of puffs_start(). Get completely rid of puffs_start(), as everything it used to do is now handled by the mount routine. Introduce an optional pre-mount call puffs_setrootinfo() for setting non-default root node information. As the old puffs_mount() is now virtually useless, say byebye to it and rename the old puffs_domount() to puffs_mount(), but add a root cookie parameter to compensate for the late puffs_start().
2007-05-17 18:03:13 +04:00
PUFFS_STATE_BEFOREMOUNT, PUFFS_STATE_RUNNING,
Instead of keeping on adding parameters to puffs_mount(), make it only take the bare essentials, which currently means removing "maxreqlen" from the argument list (all current callers I'm aware of set it as 0 anyway). Introduce puffs_init(), which provides a context for setting various parameters and puffs_domount(), which can be used to mount the file system. Keep puffs_mount() as a shortcut for the above two for simple file systems. Bump development ABI version to 13. After all, it's Friday the 13th. Watch out! Bad things can happen on Friday the 13th. --No carrier--
2007-04-13 17:35:46 +04:00
PUFFS_STATE_UNMOUNTING, PUFFS_STATE_UNMOUNTED
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
};
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
#define PUFFS_FLAG_BUILDPATH 0x80000000 /* node paths in pnode */
#define PUFFS_FLAG_OPDUMP 0x40000000 /* dump all operations */
Introduce PUFFS_FLAG_HASHPATH, which for PUFFS_FLAG_BUILDPATH file servers creates a hash of the node's path into po->po_hash to avoid a full pathcmp every time.
2007-05-01 19:58:00 +04:00
#define PUFFS_FLAG_HASHPATH 0x20000000 /* speedup: hash paths */
#define PUFFS_FLAG_MASK 0xe0000000
use more accurate flag masking
2007-04-22 22:02:44 +04:00
#define PUFFS_FLAG_KERN(a) ((a) & PUFFS_KFLAG_MASK)
#define PUFFS_FLAG_LIB(a) ((a) & PUFFS_FLAG_MASK)
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
summon daemon(3) in mainloop unless the nodaemon flag is given
2006-11-30 08:37:48 +03:00
/* blocking mode argument */
* catch up with kernel changes * better error handling when mounting
2006-11-09 16:11:01 +03:00
#define PUFFSDEV_BLOCK 0
#define PUFFSDEV_NONBLOCK 1
summon daemon(3) in mainloop unless the nodaemon flag is given
2006-11-30 08:37:48 +03:00
/* mainloop flags */
#define PUFFSLOOP_NODAEMON 0x01
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
#define DENT_DOT 0
#define DENT_DOTDOT 1
#define DENT_ADJ(a) ((a)-2) /* nth request means dir's n-2th */
outline & export code for interfacing with kernel GET/PUT requests
2006-12-08 02:15:20 +03:00
/*
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
* protos
outline & export code for interfacing with kernel GET/PUT requests
2006-12-08 02:15:20 +03:00
*/
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
#define PUFFSOP_PROTOS(fsname) \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
int fsname##_fs_unmount(struct puffs_cc *, int, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_fs_statvfs(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
struct statvfs *, const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_fs_sync(struct puffs_cc *, int, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
const struct puffs_cred *cred, const struct puffs_cid *); \
* support nodetofh and fhtonode callbacks * use new signature of node_readdir
2007-04-12 01:04:51 +04:00
int fsname##_fs_fhtonode(struct puffs_cc *, void *, size_t, \
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
struct puffs_newinfo *); \
* support nodetofh and fhtonode callbacks * use new signature of node_readdir
2007-04-12 01:04:51 +04:00
int fsname##_fs_nodetofh(struct puffs_cc *, void *cookie, \
void *, size_t *); \
add interfaces for suspending the file system
2007-01-27 02:00:33 +03:00
void fsname##_fs_suspend(struct puffs_cc *, int); \
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
\
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_lookup(struct puffs_cc *, \
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_create(struct puffs_cc *, \
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
const struct vattr *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_mknod(struct puffs_cc *, \
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
const struct vattr *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_open(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_close(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_access(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_getattr(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, struct vattr *, const struct puffs_cred *, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_setattr(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, const struct vattr *, const struct puffs_cred *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
const struct puffs_cid *); \
int fsname##_node_poll(struct puffs_cc *, void *, int *, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_mmap(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, int, const struct puffs_cred *, \
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_fsync(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, const struct puffs_cred *, int, off_t, off_t, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_seek(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, off_t, off_t, const struct puffs_cred *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_remove(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_link(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_rename(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *, void *, void *, \
const struct puffs_cn *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_mkdir(struct puffs_cc *, \
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
const struct vattr *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_rmdir(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, const struct puffs_cn *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_symlink(struct puffs_cc *, \
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void *, struct puffs_newinfo *, const struct puffs_cn *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
const struct vattr *, const char *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_readdir(struct puffs_cc *, \
* support nodetofh and fhtonode callbacks * use new signature of node_readdir
2007-04-12 01:04:51 +04:00
void *, struct dirent *, off_t *, size_t *, \
const struct puffs_cred *, int *, off_t *, size_t *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_readlink(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, const struct puffs_cred *, char *, size_t *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_reclaim(struct puffs_cc *, \
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
void *, const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_inactive(struct puffs_cc *, \
Get rid of the "int *refs" parameter to inactive: the same can be accomplished now with puffs_setbacks.
2007-07-02 14:24:17 +04:00
void *, const struct puffs_cid *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_print(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_pathconf(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, int, int *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_advlock(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, void *, int, struct flock *, int); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_getextattr(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, struct puffs_vnreq_getextattr *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_setextattr(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, struct puffs_vnreq_setextattr *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_listextattr(struct puffs_cc *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
void *, struct puffs_vnreq_listextattr *); \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_read(struct puffs_cc *, void *, \
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
uint8_t *, off_t, size_t *, const struct puffs_cred *, int);\
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int fsname##_node_write(struct puffs_cc *, void *, \
initial support for cacheops
2007-03-20 13:22:22 +03:00
uint8_t *, off_t, size_t *, const struct puffs_cred *, int);\
\
int fsname##_cache_write(struct puffs_usermount *, void *, \
size_t, struct puffs_cacheinfo *);
add libpuffs - the userspace programming interface for puffs This work was initially started and completed for Google SoC 2005 and tweaked to work a bit better in the past few weeks. While being far from complete, it is functional enough to be able and stable to host a fairly general-purpose in-memory file system in userspace. Even so, puffs should be considered experimental and no binary compatibility for interfaces or crash-freedom or zero security implications should be relied upon just yet. The GSoC project was mentored by William Studenmund and the final review for the code was done by Christos.
2006-10-23 02:52:21 +04:00
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
#define PUFFSOP_INIT(ops) \
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
ops = malloc(sizeof(struct puffs_ops)); \
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
memset(ops, 0, sizeof(struct puffs_ops))
#define PUFFSOP_SET(ops, fsname, fsornode, opname) \
(ops)->puffs_##fsornode##_##opname = fsname##_##fsornode##_##opname
#define PUFFSOP_SETFSNOP(ops, opname) \
(ops)->puffs_fs_##opname = puffs_fsnop_##opname
Introduce puffs_null_setops(), which initializes the ops vector with puffs nullfs ops.
2007-06-24 22:42:25 +04:00
PUFFSOP_PROTOS(puffs_null) /* XXX */
Get rid of the "int *refs" parameter to inactive: the same can be accomplished now with puffs_setbacks.
2007-07-02 14:24:17 +04:00
#define PUFFS_DEVEL_LIBVERSION 22
Instead of keeping on adding parameters to puffs_mount(), make it only take the bare essentials, which currently means removing "maxreqlen" from the argument list (all current callers I'm aware of set it as 0 anyway). Introduce puffs_init(), which provides a context for setting various parameters and puffs_domount(), which can be used to mount the file system. Keep puffs_mount() as a shortcut for the above two for simple file systems. Bump development ABI version to 13. After all, it's Friday the 13th. Watch out! Bad things can happen on Friday the 13th. --No carrier--
2007-04-13 17:35:46 +04:00
#define puffs_init(a,b,c,d) \
_puffs_init(PUFFS_DEVEL_LIBVERSION,a,b,c,d)
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
#define PNPATH(pnode) ((pnode)->pn_po.po_path)
#define PNPLEN(pnode) ((pnode)->pn_po.po_len)
#define PCNPATH(pcnode) ((pcnode)->pcn_po_full.po_path)
#define PCNPLEN(pcnode) ((pcnode)->pcn_po_full.po_len)
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
#define PCNISDOTDOT(pcnode) \
((pcnode)->pcn_namelen == 2 && strcmp((pcnode)->pcn_name, "..") == 0)
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
Introduce PUFFS_STORE_DCOOKIE, which stores a directory cookie for a given offset iff the directory read in question wants cookies.
2007-04-17 15:42:14 +04:00
#define PUFFS_STORE_DCOOKIE(cp, ncp, off) \
if (cp) { \
*((cp)++) = off; \
(*(ncp))++; \
}
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
/* mainloop */
typedef void (*puffs_ml_loop_fn)(struct puffs_usermount *);
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
/* framebuf stuff */
struct puffs_framebuf;
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
typedef int (*puffs_framev_readframe_fn)(struct puffs_usermount *,
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
struct puffs_framebuf *,
int, int *);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
typedef int (*puffs_framev_writeframe_fn)(struct puffs_usermount *,
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
struct puffs_framebuf *,
int, int *);
Change the signature of puffs_respcmp to return -1,0,1 for smaller, equal, larger, respectively instead of 0/1 for non/equal. This will allow sorting the buffers for faster matching in libpuffs. While here, change the name from respcmp to framecmp, as that better reflects the purpose. NOTE! there is no obvious way to make compilation fail for file systems which may already be using this feature (although I don't think there are any outside our tree, as the feature is two weeks old). Nevertheless, non-updated file systems will fail very quickly.
2007-05-20 20:24:37 +04:00
typedef int (*puffs_framev_cmpframe_fn)(struct puffs_usermount *,
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
struct puffs_framebuf *,
struct puffs_framebuf *);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
typedef void (*puffs_framev_fdnotify_fn)(struct puffs_usermount *, int, int);
typedef void (*puffs_framev_cb)(struct puffs_usermount *,
struct puffs_framebuf *,
Add error parameter to puffs_framev_cb to signal if there is a result of if the function should just do resource cleanup.
2007-05-16 14:04:08 +04:00
void *, int);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
#define PUFFS_FBGONE_READ 0x01
#define PUFFS_FBGONE_WRITE 0x02
#define PUFFS_FBGONE_BOTH(a) ((a)==(PUFFS_FBGONE_READ|PUFFS_FBGONE_WRITE))
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
__BEGIN_DECLS
Instead of keeping on adding parameters to puffs_mount(), make it only take the bare essentials, which currently means removing "maxreqlen" from the argument list (all current callers I'm aware of set it as 0 anyway). Introduce puffs_init(), which provides a context for setting various parameters and puffs_domount(), which can be used to mount the file system. Keep puffs_mount() as a shortcut for the above two for simple file systems. Bump development ABI version to 13. After all, it's Friday the 13th. Watch out! Bad things can happen on Friday the 13th. --No carrier--
2007-04-13 17:35:46 +04:00
struct puffs_usermount *_puffs_init(int, struct puffs_ops *pops, const char *,
void *, uint32_t);
Reflect kernel changes for supplying root node information in mount instead of puffs_start(). Get completely rid of puffs_start(), as everything it used to do is now handled by the mount routine. Introduce an optional pre-mount call puffs_setrootinfo() for setting non-default root node information. As the old puffs_mount() is now virtually useless, say byebye to it and rename the old puffs_domount() to puffs_mount(), but add a root cookie parameter to compensate for the late puffs_start().
2007-05-17 18:03:13 +04:00
int puffs_mount(struct puffs_usermount *, const char *, int, void*);
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
int puffs_exit(struct puffs_usermount *, int);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
int puffs_mainloop(struct puffs_usermount *, int);
int puffs_getselectable(struct puffs_usermount *);
int puffs_setblockingmode(struct puffs_usermount *, int);
int puffs_getstate(struct puffs_usermount *);
void puffs_setstacksize(struct puffs_usermount *, size_t);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
void puffs_ml_setloopfn(struct puffs_usermount *, puffs_ml_loop_fn);
void puffs_ml_settimeout(struct puffs_usermount *, struct timespec *);
Hide struct puffs_usermount from the rest of the world and provide accessors for interesting data in it. Namely, you can now get pu->pu_privdata with puffs_getspecific(), pu->pu_pn_root with puffs_set/getroot() and pu->pu_maxreqlen with puffs_getmaxreqlen().
2007-04-12 19:09:00 +04:00
void puffs_setroot(struct puffs_usermount *,
struct puffs_node *);
struct puffs_node *puffs_getroot(struct puffs_usermount *);
Reflect kernel changes for supplying root node information in mount instead of puffs_start(). Get completely rid of puffs_start(), as everything it used to do is now handled by the mount routine. Introduce an optional pre-mount call puffs_setrootinfo() for setting non-default root node information. As the old puffs_mount() is now virtually useless, say byebye to it and rename the old puffs_domount() to puffs_mount(), but add a root cookie parameter to compensate for the late puffs_start().
2007-05-17 18:03:13 +04:00
void puffs_setrootinfo(struct puffs_usermount *,
enum vtype, vsize_t, dev_t);
Hide struct puffs_usermount from the rest of the world and provide accessors for interesting data in it. Namely, you can now get pu->pu_privdata with puffs_getspecific(), pu->pu_pn_root with puffs_set/getroot() and pu->pu_maxreqlen with puffs_getmaxreqlen().
2007-04-12 19:09:00 +04:00
void *puffs_getspecific(struct puffs_usermount *);
some routines for setting mount parameters
2007-04-16 12:28:55 +04:00
void puffs_setmaxreqlen(struct puffs_usermount *, size_t);
Hide struct puffs_usermount from the rest of the world and provide accessors for interesting data in it. Namely, you can now get pu->pu_privdata with puffs_getspecific(), pu->pu_pn_root with puffs_set/getroot() and pu->pu_maxreqlen with puffs_getmaxreqlen().
2007-04-12 19:09:00 +04:00
size_t puffs_getmaxreqlen(struct puffs_usermount *);
alloc to set flags when setting fhsize
2007-04-16 17:04:49 +04:00
void puffs_setfhsize(struct puffs_usermount *, size_t, int);
some routines for setting mount parameters
2007-04-16 12:28:55 +04:00
void puffs_setncookiehash(struct puffs_usermount *, int);
Hide struct puffs_usermount from the rest of the world and provide accessors for interesting data in it. Namely, you can now get pu->pu_privdata with puffs_getspecific(), pu->pu_pn_root with puffs_set/getroot() and pu->pu_maxreqlen with puffs_getmaxreqlen().
2007-04-12 19:09:00 +04:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct puffs_pathobj *puffs_getrootpathobj(struct puffs_usermount *);
Add puffs_setback(), which can be used to set setbacks for nodes before the call returns to the kernel. Currently it can be called only from open(), mmap(), remove() and rmdir(). (But that might change)
2007-05-07 21:16:07 +04:00
void puffs_setback(struct puffs_cc *, int);
Add preliminary code for a nullfs layer, which can be used to mount a directory hierarchy to another point, just like with the kernel nullfs. This is not really a layering scheme yet, but it should evolve into one. Currently it can just be used to do 1:1 mapping.
2007-01-11 04:01:55 +03:00
struct puffs_node *puffs_pn_new(struct puffs_usermount *, void *);
Introduce puffs_pn_remove() to signal that a node has been deleted and destroyed, but not yet reclaimed. This prevents puffs_pn_nodewalk() from returning stale entries. Make nullfs use this (some file systems are a bit too happy with recycling inode numbers).
2007-06-24 21:55:07 +04:00
void puffs_pn_remove(struct puffs_node *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
void puffs_pn_put(struct puffs_node *);
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
Instead of requesting various pieces of information from node-creating methods using "please fill me"-pointers, pass "struct puffs_newinfo *" and provide methods to fill in new node info.
2007-07-01 22:39:39 +04:00
void puffs_newinfo_setcookie(struct puffs_newinfo *, void *);
void puffs_newinfo_setvtype(struct puffs_newinfo *, enum vtype);
void puffs_newinfo_setsize(struct puffs_newinfo *, voff_t);
void puffs_newinfo_setrdev(struct puffs_newinfo *, dev_t);
Hide struct puffs_usermount from the rest of the world and provide accessors for interesting data in it. Namely, you can now get pu->pu_privdata with puffs_getspecific(), pu->pu_pn_root with puffs_set/getroot() and pu->pu_maxreqlen with puffs_getmaxreqlen().
2007-04-12 19:09:00 +04:00
void *puffs_pn_getmntspecific(struct puffs_node *);
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
typedef void * (*puffs_nodewalk_fn)(struct puffs_usermount *,
struct puffs_node *, void *);
void *puffs_pn_nodewalk(struct puffs_usermount *,
puffs_nodewalk_fn, void *);
Add preliminary code for a nullfs layer, which can be used to mount a directory hierarchy to another point, just like with the kernel nullfs. This is not really a layering scheme yet, but it should evolve into one. Currently it can just be used to do 1:1 mapping.
2007-01-11 04:01:55 +03:00
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
void puffs_setvattr(struct vattr *, const struct vattr *);
void puffs_vattr_null(struct vattr *);
Introduce puffs_null_setops(), which initializes the ops vector with puffs nullfs ops.
2007-06-24 22:42:25 +04:00
void puffs_null_setops(struct puffs_ops *);
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
/*
* generic/dummy routines applicable for some file systems
*/
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
int puffs_fsnop_unmount(struct puffs_cc *, int, const struct puffs_cid *);
int puffs_fsnop_statvfs(struct puffs_cc *, struct statvfs *,
const struct puffs_cid *);
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
void puffs_zerostatvfs(struct statvfs *);
int puffs_fsnop_sync(struct puffs_cc *, int waitfor,
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
const struct puffs_cred *, const struct puffs_cid *);
int puffs_genfs_node_reclaim(struct puffs_cc *, void *,
const struct puffs_cid *);
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
Add preliminary code for a nullfs layer, which can be used to mount a directory hierarchy to another point, just like with the kernel nullfs. This is not really a layering scheme yet, but it should evolve into one. Currently it can just be used to do 1:1 mapping.
2007-01-11 04:01:55 +03:00
/*
* Subroutine stuff
*/
int puffs_gendotdent(struct dirent **, ino_t, int, size_t *);
int puffs_nextdent(struct dirent **, const char *, ino_t,
uint8_t, size_t *);
int puffs_vtype2dt(enum vtype);
enum vtype puffs_mode2vt(mode_t);
void puffs_stat2vattr(struct vattr *va, const struct stat *);
va->va_mode doesn't contain the full argument to the mknod() system call, so introduce puffs_addvtype2mode() and use that in null.c to generate the proper syscall argument
2007-03-16 11:14:49 +03:00
mode_t puffs_addvtype2mode(mode_t, enum vtype);
add routiens which check against ufs semantics for permission to chown, chmod and change a/mtime
2007-03-21 22:55:55 +03:00
/*
* credentials & permissions
*/
/* Credential fetch */
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
int puffs_cred_getuid(const struct puffs_cred *, uid_t *);
int puffs_cred_getgid(const struct puffs_cred *, gid_t *);
int puffs_cred_getgroups(const struct puffs_cred *, gid_t *, short *);
add routiens which check against ufs semantics for permission to chown, chmod and change a/mtime
2007-03-21 22:55:55 +03:00
/* Credential check */
remove variable names from header
2007-03-22 17:22:08 +03:00
int puffs_cred_isuid(const struct puffs_cred *, uid_t);
int puffs_cred_hasgroup(const struct puffs_cred *, gid_t);
* return -1 and set errno instead of returning errno directly * introduce puffs_cred_isregular() which checks if the cred structure has regular uid/gid (i.e. "person" credentials and not kernel creds)
2007-03-22 18:32:22 +03:00
int puffs_cred_isregular(const struct puffs_cred *);
remove variable names from header
2007-03-22 17:22:08 +03:00
int puffs_cred_iskernel(const struct puffs_cred *);
int puffs_cred_isfs(const struct puffs_cred *);
int puffs_cred_isjuggernaut(const struct puffs_cred *);
add routiens which check against ufs semantics for permission to chown, chmod and change a/mtime
2007-03-21 22:55:55 +03:00
Instead of supplying a plain pid, supply an abstract struct puffs_cid *, which can currently be used to query the pid and lwpid.
2007-07-01 21:22:13 +04:00
/* Caller ID */
int puffs_cid_getpid(const struct puffs_cid *, pid_t *);
int puffs_cid_getlwpid(const struct puffs_cid *, lwpid_t *);
add routiens which check against ufs semantics for permission to chown, chmod and change a/mtime
2007-03-21 22:55:55 +03:00
/* misc */
int puffs_access(enum vtype, mode_t, uid_t, gid_t, mode_t,
const struct puffs_cred *);
unify the location of *pcr in puffs_access_*
2007-03-22 19:57:27 +03:00
int puffs_access_chown(uid_t, gid_t, uid_t, gid_t,
const struct puffs_cred *);
int puffs_access_chmod(uid_t, gid_t, enum vtype, mode_t,
const struct puffs_cred *);
int puffs_access_times(uid_t, gid_t, mode_t, int,
const struct puffs_cred *);
add routiens which check against ufs semantics for permission to chown, chmod and change a/mtime
2007-03-21 22:55:55 +03:00
Add preliminary code for a nullfs layer, which can be used to mount a directory hierarchy to another point, just like with the kernel nullfs. This is not really a layering scheme yet, but it should evolve into one. Currently it can just be used to do 1:1 mapping.
2007-01-11 04:01:55 +03:00
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
/*
* Requests
*/
be consistent in naming request handlers
2007-01-20 16:52:14 +03:00
struct puffs_getreq *puffs_req_makeget(struct puffs_usermount *,
size_t, int);
int puffs_req_loadget(struct puffs_getreq *);
struct puffs_req *puffs_req_get(struct puffs_getreq *);
int puffs_req_remainingget(struct puffs_getreq *);
void puffs_req_setmaxget(struct puffs_getreq *, int);
void puffs_req_destroyget(struct puffs_getreq *);
struct puffs_putreq *puffs_req_makeput(struct puffs_usermount *);
void puffs_req_put(struct puffs_putreq *,struct puffs_req *);
void puffs_req_putcc(struct puffs_putreq *,struct puffs_cc*);
int puffs_req_putput(struct puffs_putreq *);
void puffs_req_resetput(struct puffs_putreq *);
void puffs_req_destroyput(struct puffs_putreq *);
Remove the puffs_usermount argument from puffs_req_handle(). It can be found from the other arguments and this way puffs_req_handle() is more symmetric with the other puffs_req() routines.
2007-05-09 22:36:52 +04:00
int puffs_req_handle(struct puffs_getreq *,
be consistent in naming request handlers
2007-01-20 16:52:14 +03:00
struct puffs_putreq *, int);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
/*
* Call Context interfaces relevant for user.
*/
void puffs_cc_yield(struct puffs_cc *);
void puffs_cc_continue(struct puffs_cc *);
struct puffs_usermount *puffs_cc_getusermount(struct puffs_cc *);
add puffs_cc_getspecific(pcc), which is shorthand for: puffs_getspecific(puffs_cc_getusermount(pcc))
2007-04-19 18:45:03 +04:00
void *puffs_cc_getspecific(struct puffs_cc *);
* get rid of the mount callback; it's no great surprise to the server that it needs to mount the file system backend if it wants to call mount * provide some options for getmntopts(), assume that callers will parse command line (or fstab) args * reorganize the puffs_cc interface just a bit, preparing for a bigger revamp later
2007-01-06 21:22:09 +03:00
struct puffs_cc *puffs_cc_create(struct puffs_usermount *);
void puffs_cc_destroy(struct puffs_cc *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
/*
* Execute or continue a request
*/
Be consistent with the rest of the framework about the argument order of puffs_docc and puffs_dopreq
2007-01-20 17:37:06 +03:00
int puffs_dopreq(struct puffs_usermount *, struct puffs_req *,
struct puffs_putreq *);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
void puffs_docc(struct puffs_cc *, struct puffs_putreq *);
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
lib interface for invalidation routines
2007-01-09 21:15:08 +03:00
/*
* Flushing / invalidation routines
*/
rename name cache invalidation interface to be in sync with the kernel operation names, i.e. contain "namecache" instead of just "name"
2007-01-09 21:19:01 +03:00
int puffs_inval_namecache_dir(struct puffs_usermount *, void *);
int puffs_inval_namecache_all(struct puffs_usermount *);
lib interface for invalidation routines
2007-01-09 21:15:08 +03:00
initial support for cacheops
2007-03-20 13:22:22 +03:00
int puffs_inval_pagecache_node(struct puffs_usermount *, void *);
implement puffs_inval_pagecache_node_range()
2007-04-06 20:38:03 +04:00
int puffs_inval_pagecache_node_range(struct puffs_usermount *, void *,
off_t, off_t);
support page cache flushing
2007-04-06 21:49:11 +04:00
int puffs_flush_pagecache_node(struct puffs_usermount *, void *);
int puffs_flush_pagecache_node_range(struct puffs_usermount *, void *,
off_t, off_t);
initial support for cacheops
2007-03-20 13:22:22 +03:00
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
/*
* Path constructicons
*/
Introduce puffs_path_walkcmp(), which is to be called from nodewalk and compares the path of the node against the given pathobject. Also make comparison method take a flag to indicate if it should check if the second path is a true prefix of the first. plus some namespace cleanup
2007-02-15 20:04:46 +03:00
int puffs_stdpath_buildpath(struct puffs_usermount *,
* get rid of the holy kernel-given ISDOTDOT, and instead define a macro which does strcmp against ".." and (the untranslated) componentname * make PUFFS_FLAG_BUILDPATH build paths also if dotdot is the case, and adapt the regular path objects to this * make nullfs lookup readable because we can now get rid of dotdot processing there
2007-02-15 15:51:45 +03:00
const struct puffs_pathobj *,
const struct puffs_pathobj *, size_t,
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
struct puffs_pathobj *);
Introduce puffs_path_walkcmp(), which is to be called from nodewalk and compares the path of the node against the given pathobject. Also make comparison method take a flag to indicate if it should check if the second path is a true prefix of the first. plus some namespace cleanup
2007-02-15 20:04:46 +03:00
int puffs_stdpath_cmppath(struct puffs_usermount *, struct puffs_pathobj *,
struct puffs_pathobj *, size_t, int);
void puffs_stdpath_freepath(struct puffs_usermount *,struct puffs_pathobj *);
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
Introduce puffs_path_walkcmp(), which is to be called from nodewalk and compares the path of the node against the given pathobject. Also make comparison method take a flag to indicate if it should check if the second path is a true prefix of the first. plus some namespace cleanup
2007-02-15 20:04:46 +03:00
void *puffs_path_walkcmp(struct puffs_usermount *,
struct puffs_node *, void *);
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
void *puffs_path_prefixadj(struct puffs_usermount *,
struct puffs_node *, void *);
int puffs_path_pcnbuild(struct puffs_usermount *,
struct puffs_cn *, void *);
Introduce PUFFS_FLAG_HASHPATH, which for PUFFS_FLAG_BUILDPATH file servers creates a hash of the node's path into po->po_hash to avoid a full pathcmp every time.
2007-05-01 19:58:00 +04:00
void puffs_path_buildhash(struct puffs_usermount *, struct puffs_pathobj *);
Checkpoint some more work in progress: for the benefit those file system backends which operate purely based on paths, push out more path management into the library and make path management more abstract: enable a file system to define a bunch of path management callbacks, which are used by the framework. Management of normal /this/is/a/path type paths is provided by the library.
2007-01-15 03:39:02 +03:00
void puffs_set_pathbuild(struct puffs_usermount *, pu_pathbuild_fn);
void puffs_set_pathtransform(struct puffs_usermount *, pu_pathtransform_fn);
void puffs_set_pathcmp(struct puffs_usermount *, pu_pathcmp_fn);
void puffs_set_pathfree(struct puffs_usermount *, pu_pathfree_fn);
void puffs_set_namemod(struct puffs_usermount *, pu_namemod_fn);
add interfaces for suspending the file system
2007-01-27 02:00:33 +03:00
/*
* Suspension
*/
int puffs_fs_suspend(struct puffs_usermount *);
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
/*
* Frame buffering
*/
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
void puffs_framev_init(struct puffs_usermount *,
puffs_framev_readframe_fn,
puffs_framev_writeframe_fn,
Change the signature of puffs_respcmp to return -1,0,1 for smaller, equal, larger, respectively instead of 0/1 for non/equal. This will allow sorting the buffers for faster matching in libpuffs. While here, change the name from respcmp to framecmp, as that better reflects the purpose. NOTE! there is no obvious way to make compilation fail for file systems which may already be using this feature (although I don't think there are any outside our tree, as the feature is two weeks old). Nevertheless, non-updated file systems will fail very quickly.
2007-05-20 20:24:37 +04:00
puffs_framev_cmpframe_fn,
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
puffs_framev_fdnotify_fn);
Step 2 of general-purpose eventloop: get rid of puffs_framebuf_eventloop(). puffs_mainloop() now handles that functionality as well.
2007-05-12 01:27:13 +04:00
Add puffs "frame buffers", which are an abstraction of the buffering and event handling mechanisms required in file servers with blocking I/O backends. puffs_framebuf is built on the concept of puffs_cc and uses those to multiplex execution where needed. File systems are required to implement three methods: * read frame * write frame * compare if frame is a response to the given one Memory management is provided by puffs_framebuf, but the file systems must still, of course, interpret the protocol and do e.g. byte order conversion. As always, puffs_framebuf is work in progress. Current users are mount_psshfs and mount_9p.
2007-05-05 19:48:18 +04:00
struct puffs_framebuf *puffs_framebuf_make(void);
void puffs_framebuf_destroy(struct puffs_framebuf *);
void puffs_framebuf_recycle(struct puffs_framebuf *);
int puffs_framebuf_reserve_space(struct puffs_framebuf *,
size_t);
int puffs_framebuf_putdata(struct puffs_framebuf *, const void *, size_t);
int puffs_framebuf_putdata_atoff(struct puffs_framebuf *, size_t,
const void *, size_t);
int puffs_framebuf_getdata(struct puffs_framebuf *, void *, size_t);
int puffs_framebuf_getdata_atoff(struct puffs_framebuf *, size_t,
void *, size_t);
size_t puffs_framebuf_telloff(struct puffs_framebuf *);
size_t puffs_framebuf_tellsize(struct puffs_framebuf *);
size_t puffs_framebuf_remaining(struct puffs_framebuf *);
int puffs_framebuf_seekset(struct puffs_framebuf *, size_t);
int puffs_framebuf_getwindow(struct puffs_framebuf *, size_t,
void **, size_t *);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
int puffs_framev_enqueue_cc(struct puffs_cc *, int,
struct puffs_framebuf *);
int puffs_framev_enqueue_cb(struct puffs_usermount *, int,
struct puffs_framebuf *,
puffs_framev_cb, void *);
int puffs_framev_enqueue_justsend(struct puffs_usermount *, int,
struct puffs_framebuf *, int);
Introduce puffs_framev_framebuf_ccpromote, which can be used to turn a reference to puffs_framebuf in the file system from a cb/justsend operation to a cc wait, should the file system find itself desiring the result.
2007-05-20 23:56:56 +04:00
int puffs_framev_framebuf_ccpromote(struct puffs_framebuf *,
struct puffs_cc *);
Step 3 and currently final step of a more general purpose event loop: support removal and addition of i/o file descriptors on the fly. * detect closed file descriptors * automatically free waiters of a dead file descriptor * give the file server the possibility to specify a callback which notifies of a dead file descriptor * move loop function to be a property of the mainloop instead of framebuf (doesn't change effective behaviour) * add the possibility to configure a timespec parameter which attempts to call the loop function periodically * move the event loop functions from the puffs_framebuf namespace to puffs_framev to differential between pure memory management functions
2007-05-15 17:44:46 +04:00
int puffs_framev_addfd(struct puffs_usermount *pu, int);
int puffs_framev_removefd(struct puffs_usermount *pu, int, int);
void puffs_framev_unmountonclose(struct puffs_usermount *pu, int, int);
Step 1 in moving towards a general-purpose eventloop: allow the framebuf event loop to take n i/o fd's as parameters and also allow dynamic add/remove of fd's. (not tested except for one fd, but more changes coming soon)
2007-05-11 20:22:38 +04:00
checkpoint some experimental work-in-progress, namely: Add support for having multiple outstanding operations. This is done by exposing enough interfaces so that it is convenient to have the main event loop in the implementation itself and by providing a continuation framework for convinient blocking and rescheduling. works fine, but will undergo further cleanup & development
2006-12-29 18:28:11 +03:00
__END_DECLS
rototill the operation callbacks: unify the ops under struct puffs_ops and namespace them
2006-12-07 13:53:21 +03:00
#endif /* _PUFFS_H_ */
Reference in New Issue Copy Permalink