samba-mirror/source3/modules/vfs_commit.c

/*
 * Copyright (c) James Peach 2006, 2007
 * Copyright (c) David Losada Carballo 2007
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation; either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, see <http://www.gnu.org/licenses/>.
 */

#include "includes.h"
#include "system/filesys.h"
#include "smbd/smbd.h"
#include "lib/util/tevent_unix.h"

/* Commit data module.
 *
 * The purpose of this module is to flush data to disk at regular intervals,
 * just like the NFS commit operation. There's two rationales for this. First,
 * it minimises the data loss in case of a power outage without incurring
 * the poor performance of synchronous I/O. Second, a steady flush rate
 * can produce better throughput than suddenly dumping massive amounts of
 * writes onto a disk.
 *
 * Tunables:
 *
 *  commit: dthresh         Amount of dirty data that can accumulate
 *                          before we commit (sync) it.
 *
 *  commit: debug           Debug level at which to emit messages.
 *
 *  commit: eof mode        String. Tunes how the module tries to guess when
 *                          the client has written the last bytes of the file.
 *                          Possible values (default = hinted):
 *
 *     (*)  = hinted        Some clients (i.e. Windows Explorer) declare the
 *                          size of the file before transferring it. With this
 *                          option, we remember that hint, and commit after
 *                          writing in that file position. If the client
 *                          doesn't declare the size of file, commiting on EOF 
 *                          is not triggered.
 *
 *          = growth        Commits after a write operation has made the file
 *                          size grow. If the client declares a file size, it
 *                          refrains to commit until the file has reached it.
 *                          Useful for defeating writeback on NFS shares.
 *
 */

#define MODULE "commit"

static int module_debug;

enum eof_mode
{
    EOF_NONE = 0x0000,
    EOF_HINTED = 0x0001,
    EOF_GROWTH = 0x0002
};

struct commit_info
{
        /* For chunk-based commits */
        off_t dbytes;	/* Dirty (uncommitted) bytes */
        off_t dthresh;	/* Dirty data threshold */
        /* For commits on EOF */
        enum eof_mode on_eof;
        off_t eof;		/* Expected file size */
};

static int commit_do(
        struct commit_info *            c,
        int                             fd)
{
        int result;

	DEBUG(module_debug,
		("%s: flushing %lu dirty bytes\n",
		 MODULE, (unsigned long)c->dbytes));

#if defined(HAVE_FDATASYNC)
        result = fdatasync(fd);
#elif defined(HAVE_FSYNC)
        result = fsync(fd);
#else
	DEBUG(0, ("%s: WARNING: no commit support on this platform\n",
		MODULE));
	result = 0
#endif
        if (result == 0) {
                c->dbytes = 0;	/* on success, no dirty bytes */
        }
        return result;
}

static int commit_all(
        struct vfs_handle_struct *	handle,
        files_struct *		        fsp)
{
        struct commit_info *c;

        if ((c = (struct commit_info *)VFS_FETCH_FSP_EXTENSION(handle, fsp))) {
                if (c->dbytes) {
                        DEBUG(module_debug,
                                ("%s: flushing %lu dirty bytes\n",
                                 MODULE, (unsigned long)c->dbytes));

                        return commit_do(c, fsp_get_io_fd(fsp));
                }
        }
        return 0;
}

static int commit(
        struct vfs_handle_struct *	handle,
        files_struct *		        fsp,
	off_t			offset,
        ssize_t			        last_write)
{
        struct commit_info *c;

        if ((c = (struct commit_info *)VFS_FETCH_FSP_EXTENSION(handle, fsp))
	    == NULL) {
		return 0;
	}

	c->dbytes += last_write;	/* dirty bytes always counted */

	if (c->dthresh && (c->dbytes > c->dthresh)) {
		return commit_do(c, fsp_get_io_fd(fsp));
	}

	/* Return if we are not in EOF mode or if we have temporarily opted
	 * out of it.
	 */
	if (c->on_eof == EOF_NONE || c->eof < 0) {
		return 0;
	}

	/* This write hit or went past our cache the file size. */
	if ((offset + last_write) >= c->eof) {
		if (commit_do(c, fsp_get_io_fd(fsp)) == -1) {
			return -1;
		}

		/* Hinted mode only commits the first time we hit EOF. */
		if (c->on_eof == EOF_HINTED) {
		    c->eof = -1;
		} else if (c->on_eof == EOF_GROWTH) {
		    c->eof = offset + last_write;
		}
	}

        return 0;
}

static int commit_connect(
        struct vfs_handle_struct *  handle,
        const char *                service,
        const char *                user)
{
	int ret = SMB_VFS_NEXT_CONNECT(handle, service, user);

	if (ret < 0) {
		return ret;
	}

        module_debug = lp_parm_int(SNUM(handle->conn), MODULE, "debug", 100);
        return 0;
}

static int commit_openat(struct vfs_handle_struct *handle,
			 const struct files_struct *dirfsp,
			 const struct smb_filename *smb_fname,
			 files_struct *fsp,
			 const struct vfs_open_how *how)
{
        off_t dthresh;
	const char *eof_mode;
        struct commit_info *c = NULL;
        int fd;

        /* Don't bother with read-only files. */
        if ((how->flags & O_ACCMODE) == O_RDONLY) {
                return SMB_VFS_NEXT_OPENAT(handle,
					   dirfsp,
					   smb_fname,
					   fsp,
					   how);
        }

        /* Read and check module configuration */
        dthresh = conv_str_size(lp_parm_const_string(SNUM(handle->conn),
                                        MODULE, "dthresh", NULL));

	eof_mode = lp_parm_const_string(SNUM(handle->conn),
                                        MODULE, "eof mode", "none");

        if (dthresh > 0 || !strequal(eof_mode, "none")) {
                c = VFS_ADD_FSP_EXTENSION(
			handle, fsp, struct commit_info, NULL);
                /* Process main tunables */
                if (c) {
                        c->dthresh = dthresh;
                        c->dbytes = 0;
                        c->on_eof = EOF_NONE;
                        c->eof = 0;
                }
        }
        /* Process eof_mode tunable */
        if (c) {
                if (strequal(eof_mode, "hinted")) {
                        c->on_eof = EOF_HINTED;
                } else if (strequal(eof_mode, "growth")) {
                        c->on_eof = EOF_GROWTH;
                }
        }

	fd = SMB_VFS_NEXT_OPENAT(handle, dirfsp, smb_fname, fsp, how);
	if (fd == -1) {
		VFS_REMOVE_FSP_EXTENSION(handle, fsp);
		return fd;
	}

        /* EOF commit modes require us to know the initial file size. */
        if (c && (c->on_eof != EOF_NONE)) {
                SMB_STRUCT_STAT st;
		/*
		 * Setting the fd of the FSP is a hack
		 * but also practiced elsewhere -
		 * needed for calling the VFS.
		 */
		fsp_set_fd(fsp, fd);
		if (SMB_VFS_FSTAT(fsp, &st) == -1) {
			int saved_errno = errno;
			SMB_VFS_CLOSE(fsp);
			fsp_set_fd(fsp, -1);
			errno = saved_errno;
                        return -1;
                }
		c->eof = st.st_ex_size;
        }

        return fd;
}

static ssize_t commit_pwrite(
        vfs_handle_struct * handle,
        files_struct *      fsp,
        const void *        data,
        size_t              count,
	off_t	    offset)
{
        ssize_t ret;

        ret = SMB_VFS_NEXT_PWRITE(handle, fsp, data, count, offset);
        if (ret > 0) {
                if (commit(handle, fsp, offset, ret) == -1) {
                        return -1;
                }
        }

        return ret;
}

struct commit_pwrite_state {
	struct vfs_handle_struct *handle;
	struct files_struct *fsp;
	ssize_t ret;
	struct vfs_aio_state vfs_aio_state;
};

static void commit_pwrite_written(struct tevent_req *subreq);

static struct tevent_req *commit_pwrite_send(struct vfs_handle_struct *handle,
					     TALLOC_CTX *mem_ctx,
					     struct tevent_context *ev,
					     struct files_struct *fsp,
					     const void *data,
					     size_t n, off_t offset)
{
	struct tevent_req *req, *subreq;
	struct commit_pwrite_state *state;

	req = tevent_req_create(mem_ctx, &state, struct commit_pwrite_state);
	if (req == NULL) {
		return NULL;
	}
	state->handle = handle;
	state->fsp = fsp;

	subreq = SMB_VFS_NEXT_PWRITE_SEND(state, ev, handle, fsp, data,
					  n, offset);
	if (tevent_req_nomem(subreq, req)) {
		return tevent_req_post(req, ev);
	}
	tevent_req_set_callback(subreq, commit_pwrite_written, req);
	return req;
}

static void commit_pwrite_written(struct tevent_req *subreq)
{
	struct tevent_req *req = tevent_req_callback_data(
		subreq, struct tevent_req);
	struct commit_pwrite_state *state = tevent_req_data(
		req, struct commit_pwrite_state);
	int commit_ret;

	state->ret = SMB_VFS_PWRITE_RECV(subreq, &state->vfs_aio_state);
	TALLOC_FREE(subreq);

	if (state->ret <= 0) {
		tevent_req_done(req);
		return;
	}

	/*
	 * Ok, this is a sync fake. We should make the sync async as well, but
	 * I'm too lazy for that right now -- vl
	 */
	commit_ret = commit(state->handle,
			    state->fsp,
			    fh_get_pos(state->fsp->fh),
			    state->ret);

	if (commit_ret == -1) {
		state->ret = -1;
	}

	tevent_req_done(req);
}

static ssize_t commit_pwrite_recv(struct tevent_req *req,
				  struct vfs_aio_state *vfs_aio_state)
{
	struct commit_pwrite_state *state =
		tevent_req_data(req, struct commit_pwrite_state);

	if (tevent_req_is_unix_error(req, &vfs_aio_state->error)) {
		return -1;
	}
	*vfs_aio_state = state->vfs_aio_state;
	return state->ret;
}

static int commit_close(
        vfs_handle_struct * handle,
        files_struct *      fsp)
{
        /* Commit errors not checked, close() will find them again */
        commit_all(handle, fsp);
        return SMB_VFS_NEXT_CLOSE(handle, fsp);
}

static int commit_ftruncate(
        vfs_handle_struct * handle,
        files_struct *      fsp,
        off_t           len)
{
        int result;

        result = SMB_VFS_NEXT_FTRUNCATE(handle, fsp, len);
        if (result == 0) {
		struct commit_info *c;
		if ((c = (struct commit_info *)VFS_FETCH_FSP_EXTENSION(
			     handle, fsp))) {
			commit(handle, fsp, len, 0);
			c->eof = len;
		}
        }

        return result;
}

static struct vfs_fn_pointers vfs_commit_fns = {
        .openat_fn = commit_openat,
        .close_fn = commit_close,
        .pwrite_fn = commit_pwrite,
        .pwrite_send_fn = commit_pwrite_send,
        .pwrite_recv_fn = commit_pwrite_recv,
        .connect_fn = commit_connect,
        .ftruncate_fn = commit_ftruncate
};

static_decl_vfs;
NTSTATUS vfs_commit_init(TALLOC_CTX *ctx)
{
	return smb_register_vfs(SMB_VFS_INTERFACE_VERSION, MODULE,
				&vfs_commit_fns);
}