quota.c

linux 内核源代码

/*
 * Copyright (C) Sistina Software, Inc.  1997-2003 All rights reserved.
 * Copyright (C) 2004-2006 Red Hat, Inc.  All rights reserved.
 *
 * This copyrighted material is made available to anyone wishing to use,
 * modify, copy, or redistribute it subject to the terms and conditions
 * of the GNU General Public License version 2.
 */

/*
 * Quota change tags are associated with each transaction that allocates or
 * deallocates space.  Those changes are accumulated locally to each node (in a
 * per-node file) and then are periodically synced to the quota file.  This
 * avoids the bottleneck of constantly touching the quota file, but introduces
 * fuzziness in the current usage value of IDs that are being used on different
 * nodes in the cluster simultaneously.  So, it is possible for a user on
 * multiple nodes to overrun their quota, but that overrun is controlable.
 * Since quota tags are part of transactions, there is no need to a quota check
 * program to be run on node crashes or anything like that.
 *
 * There are couple of knobs that let the administrator manage the quota
 * fuzziness.  "quota_quantum" sets the maximum time a quota change can be
 * sitting on one node before being synced to the quota file.  (The default is
 * 60 seconds.)  Another knob, "quota_scale" controls how quickly the frequency
 * of quota file syncs increases as the user moves closer to their limit.  The
 * more frequent the syncs, the more accurate the quota enforcement, but that
 * means that there is more contention between the nodes for the quota file.
 * The default value is one.  This sets the maximum theoretical quota overrun
 * (with infinite node with infinite bandwidth) to twice the user's limit.  (In
 * practice, the maximum overrun you see should be much less.)  A "quota_scale"
 * number greater than one makes quota syncs more frequent and reduces the
 * maximum overrun.  Numbers less than one (but greater than zero) make quota
 * syncs less frequent.
 *
 * GFS quotas also use per-ID Lock Value Blocks (LVBs) to cache the contents of
 * the quota file, so it is not being constantly read.
 */

#include <linux/sched.h>
#include <linux/slab.h>
#include <linux/spinlock.h>
#include <linux/completion.h>
#include <linux/buffer_head.h>
#include <linux/sort.h>
#include <linux/fs.h>
#include <linux/bio.h>
#include <linux/gfs2_ondisk.h>
#include <linux/lm_interface.h>

#include "gfs2.h"
#include "incore.h"
#include "bmap.h"
#include "glock.h"
#include "glops.h"
#include "log.h"
#include "meta_io.h"
#include "quota.h"
#include "rgrp.h"
#include "super.h"
#include "trans.h"
#include "inode.h"
#include "ops_file.h"
#include "ops_address.h"
#include "util.h"

#define QUOTA_USER 1
#define QUOTA_GROUP 0

struct gfs2_quota_host {
	u64 qu_limit;
	u64 qu_warn;
	s64 qu_value;
	u32 qu_ll_next;
};

struct gfs2_quota_change_host {
	u64 qc_change;
	u32 qc_flags; /* GFS2_QCF_... */
	u32 qc_id;
};

static u64 qd2offset(struct gfs2_quota_data *qd)
{
	u64 offset;

	offset = 2 * (u64)qd->qd_id + !test_bit(QDF_USER, &qd->qd_flags);
	offset *= sizeof(struct gfs2_quota);

	return offset;
}

static int qd_alloc(struct gfs2_sbd *sdp, int user, u32 id,
		    struct gfs2_quota_data **qdp)
{
	struct gfs2_quota_data *qd;
	int error;

	qd = kzalloc(sizeof(struct gfs2_quota_data), GFP_KERNEL);
	if (!qd)
		return -ENOMEM;

	qd->qd_count = 1;
	qd->qd_id = id;
	if (user)
		set_bit(QDF_USER, &qd->qd_flags);
	qd->qd_slot = -1;

	error = gfs2_glock_get(sdp, 2 * (u64)id + !user,
			      &gfs2_quota_glops, CREATE, &qd->qd_gl);
	if (error)
		goto fail;

	error = gfs2_lvb_hold(qd->qd_gl);
	gfs2_glock_put(qd->qd_gl);
	if (error)
		goto fail;

	*qdp = qd;

	return 0;

fail:
	kfree(qd);
	return error;
}

static int qd_get(struct gfs2_sbd *sdp, int user, u32 id, int create,
		  struct gfs2_quota_data **qdp)
{
	struct gfs2_quota_data *qd = NULL, *new_qd = NULL;
	int error, found;

	*qdp = NULL;

	for (;;) {
		found = 0;
		spin_lock(&sdp->sd_quota_spin);
		list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) {
			if (qd->qd_id == id &&
			    !test_bit(QDF_USER, &qd->qd_flags) == !user) {
				qd->qd_count++;
				found = 1;
				break;
			}
		}

		if (!found)
			qd = NULL;

		if (!qd && new_qd) {
			qd = new_qd;
			list_add(&qd->qd_list, &sdp->sd_quota_list);
			atomic_inc(&sdp->sd_quota_count);
			new_qd = NULL;
		}

		spin_unlock(&sdp->sd_quota_spin);

		if (qd || !create) {
			if (new_qd) {
				gfs2_lvb_unhold(new_qd->qd_gl);
				kfree(new_qd);
			}
			*qdp = qd;
			return 0;
		}

		error = qd_alloc(sdp, user, id, &new_qd);
		if (error)
			return error;
	}
}

static void qd_hold(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;

	spin_lock(&sdp->sd_quota_spin);
	gfs2_assert(sdp, qd->qd_count);
	qd->qd_count++;
	spin_unlock(&sdp->sd_quota_spin);
}

static void qd_put(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;
	spin_lock(&sdp->sd_quota_spin);
	gfs2_assert(sdp, qd->qd_count);
	if (!--qd->qd_count)
		qd->qd_last_touched = jiffies;
	spin_unlock(&sdp->sd_quota_spin);
}

static int slot_get(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;
	unsigned int c, o = 0, b;
	unsigned char byte = 0;

	spin_lock(&sdp->sd_quota_spin);

	if (qd->qd_slot_count++) {
		spin_unlock(&sdp->sd_quota_spin);
		return 0;
	}

	for (c = 0; c < sdp->sd_quota_chunks; c++)
		for (o = 0; o < PAGE_SIZE; o++) {
			byte = sdp->sd_quota_bitmap[c][o];
			if (byte != 0xFF)
				goto found;
		}

	goto fail;

found:
	for (b = 0; b < 8; b++)
		if (!(byte & (1 << b)))
			break;
	qd->qd_slot = c * (8 * PAGE_SIZE) + o * 8 + b;

	if (qd->qd_slot >= sdp->sd_quota_slots)
		goto fail;

	sdp->sd_quota_bitmap[c][o] |= 1 << b;

	spin_unlock(&sdp->sd_quota_spin);

	return 0;

fail:
	qd->qd_slot_count--;
	spin_unlock(&sdp->sd_quota_spin);
	return -ENOSPC;
}

static void slot_hold(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;

	spin_lock(&sdp->sd_quota_spin);
	gfs2_assert(sdp, qd->qd_slot_count);
	qd->qd_slot_count++;
	spin_unlock(&sdp->sd_quota_spin);
}

static void slot_put(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;

	spin_lock(&sdp->sd_quota_spin);
	gfs2_assert(sdp, qd->qd_slot_count);
	if (!--qd->qd_slot_count) {
		gfs2_icbit_munge(sdp, sdp->sd_quota_bitmap, qd->qd_slot, 0);
		qd->qd_slot = -1;
	}
	spin_unlock(&sdp->sd_quota_spin);
}

static int bh_get(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;
	struct gfs2_inode *ip = GFS2_I(sdp->sd_qc_inode);
	unsigned int block, offset;
	struct buffer_head *bh;
	int error;
	struct buffer_head bh_map = { .b_state = 0, .b_blocknr = 0 };

	mutex_lock(&sdp->sd_quota_mutex);

	if (qd->qd_bh_count++) {
		mutex_unlock(&sdp->sd_quota_mutex);
		return 0;
	}

	block = qd->qd_slot / sdp->sd_qc_per_block;
	offset = qd->qd_slot % sdp->sd_qc_per_block;;

	bh_map.b_size = 1 << ip->i_inode.i_blkbits;
	error = gfs2_block_map(&ip->i_inode, block, 0, &bh_map);
	if (error)
		goto fail;
	error = gfs2_meta_read(ip->i_gl, bh_map.b_blocknr, DIO_WAIT, &bh);
	if (error)
		goto fail;
	error = -EIO;
	if (gfs2_metatype_check(sdp, bh, GFS2_METATYPE_QC))
		goto fail_brelse;

	qd->qd_bh = bh;
	qd->qd_bh_qc = (struct gfs2_quota_change *)
		(bh->b_data + sizeof(struct gfs2_meta_header) +
		 offset * sizeof(struct gfs2_quota_change));

	mutex_unlock(&sdp->sd_quota_mutex);

	return 0;

fail_brelse:
	brelse(bh);
fail:
	qd->qd_bh_count--;
	mutex_unlock(&sdp->sd_quota_mutex);
	return error;
}

static void bh_put(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;

	mutex_lock(&sdp->sd_quota_mutex);
	gfs2_assert(sdp, qd->qd_bh_count);
	if (!--qd->qd_bh_count) {
		brelse(qd->qd_bh);
		qd->qd_bh = NULL;
		qd->qd_bh_qc = NULL;
	}
	mutex_unlock(&sdp->sd_quota_mutex);
}

static int qd_fish(struct gfs2_sbd *sdp, struct gfs2_quota_data **qdp)
{
	struct gfs2_quota_data *qd = NULL;
	int error;
	int found = 0;

	*qdp = NULL;

	if (sdp->sd_vfs->s_flags & MS_RDONLY)
		return 0;

	spin_lock(&sdp->sd_quota_spin);

	list_for_each_entry(qd, &sdp->sd_quota_list, qd_list) {
		if (test_bit(QDF_LOCKED, &qd->qd_flags) ||
		    !test_bit(QDF_CHANGE, &qd->qd_flags) ||
		    qd->qd_sync_gen >= sdp->sd_quota_sync_gen)
			continue;

		list_move_tail(&qd->qd_list, &sdp->sd_quota_list);

		set_bit(QDF_LOCKED, &qd->qd_flags);
		gfs2_assert_warn(sdp, qd->qd_count);
		qd->qd_count++;
		qd->qd_change_sync = qd->qd_change;
		gfs2_assert_warn(sdp, qd->qd_slot_count);
		qd->qd_slot_count++;
		found = 1;

		break;
	}

	if (!found)
		qd = NULL;

	spin_unlock(&sdp->sd_quota_spin);

	if (qd) {
		gfs2_assert_warn(sdp, qd->qd_change_sync);
		error = bh_get(qd);
		if (error) {
			clear_bit(QDF_LOCKED, &qd->qd_flags);
			slot_put(qd);
			qd_put(qd);
			return error;
		}
	}

	*qdp = qd;

	return 0;
}

static int qd_trylock(struct gfs2_quota_data *qd)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;

	if (sdp->sd_vfs->s_flags & MS_RDONLY)
		return 0;

	spin_lock(&sdp->sd_quota_spin);

	if (test_bit(QDF_LOCKED, &qd->qd_flags) ||
	    !test_bit(QDF_CHANGE, &qd->qd_flags)) {
		spin_unlock(&sdp->sd_quota_spin);
		return 0;
	}

	list_move_tail(&qd->qd_list, &sdp->sd_quota_list);

	set_bit(QDF_LOCKED, &qd->qd_flags);
	gfs2_assert_warn(sdp, qd->qd_count);
	qd->qd_count++;
	qd->qd_change_sync = qd->qd_change;
	gfs2_assert_warn(sdp, qd->qd_slot_count);
	qd->qd_slot_count++;

	spin_unlock(&sdp->sd_quota_spin);

	gfs2_assert_warn(sdp, qd->qd_change_sync);
	if (bh_get(qd)) {
		clear_bit(QDF_LOCKED, &qd->qd_flags);
		slot_put(qd);
		qd_put(qd);
		return 0;
	}

	return 1;
}

static void qd_unlock(struct gfs2_quota_data *qd)
{
	gfs2_assert_warn(qd->qd_gl->gl_sbd,
			 test_bit(QDF_LOCKED, &qd->qd_flags));
	clear_bit(QDF_LOCKED, &qd->qd_flags);
	bh_put(qd);
	slot_put(qd);
	qd_put(qd);
}

static int qdsb_get(struct gfs2_sbd *sdp, int user, u32 id, int create,
		    struct gfs2_quota_data **qdp)
{
	int error;

	error = qd_get(sdp, user, id, create, qdp);
	if (error)
		return error;

	error = slot_get(*qdp);
	if (error)
		goto fail;

	error = bh_get(*qdp);
	if (error)
		goto fail_slot;

	return 0;

fail_slot:
	slot_put(*qdp);
fail:
	qd_put(*qdp);
	return error;
}

static void qdsb_put(struct gfs2_quota_data *qd)
{
	bh_put(qd);
	slot_put(qd);
	qd_put(qd);
}

int gfs2_quota_hold(struct gfs2_inode *ip, u32 uid, u32 gid)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct gfs2_alloc *al = &ip->i_alloc;
	struct gfs2_quota_data **qd = al->al_qd;
	int error;

	if (gfs2_assert_warn(sdp, !al->al_qd_num) ||
	    gfs2_assert_warn(sdp, !test_bit(GIF_QD_LOCKED, &ip->i_flags)))
		return -EIO;

	if (sdp->sd_args.ar_quota == GFS2_QUOTA_OFF)
		return 0;

	error = qdsb_get(sdp, QUOTA_USER, ip->i_inode.i_uid, CREATE, qd);
	if (error)
		goto out;
	al->al_qd_num++;
	qd++;

	error = qdsb_get(sdp, QUOTA_GROUP, ip->i_inode.i_gid, CREATE, qd);
	if (error)
		goto out;
	al->al_qd_num++;
	qd++;

	if (uid != NO_QUOTA_CHANGE && uid != ip->i_inode.i_uid) {
		error = qdsb_get(sdp, QUOTA_USER, uid, CREATE, qd);
		if (error)
			goto out;
		al->al_qd_num++;
		qd++;
	}

	if (gid != NO_QUOTA_CHANGE && gid != ip->i_inode.i_gid) {
		error = qdsb_get(sdp, QUOTA_GROUP, gid, CREATE, qd);
		if (error)
			goto out;
		al->al_qd_num++;
		qd++;
	}

out:
	if (error)
		gfs2_quota_unhold(ip);
	return error;
}

void gfs2_quota_unhold(struct gfs2_inode *ip)
{
	struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode);
	struct gfs2_alloc *al = &ip->i_alloc;
	unsigned int x;

	gfs2_assert_warn(sdp, !test_bit(GIF_QD_LOCKED, &ip->i_flags));

	for (x = 0; x < al->al_qd_num; x++) {
		qdsb_put(al->al_qd[x]);
		al->al_qd[x] = NULL;
	}
	al->al_qd_num = 0;
}

static int sort_qd(const void *a, const void *b)
{
	const struct gfs2_quota_data *qd_a = *(const struct gfs2_quota_data **)a;
	const struct gfs2_quota_data *qd_b = *(const struct gfs2_quota_data **)b;

	if (!test_bit(QDF_USER, &qd_a->qd_flags) !=
	    !test_bit(QDF_USER, &qd_b->qd_flags)) {
		if (test_bit(QDF_USER, &qd_a->qd_flags))
			return -1;
		else
			return 1;
	}
	if (qd_a->qd_id < qd_b->qd_id)
		return -1;
	if (qd_a->qd_id > qd_b->qd_id)
		return 1;

	return 0;
}

static void do_qc(struct gfs2_quota_data *qd, s64 change)
{
	struct gfs2_sbd *sdp = qd->qd_gl->gl_sbd;
	struct gfs2_inode *ip = GFS2_I(sdp->sd_qc_inode);
	struct gfs2_quota_change *qc = qd->qd_bh_qc;
	s64 x;

	mutex_lock(&sdp->sd_quota_mutex);
	gfs2_trans_add_bh(ip->i_gl, qd->qd_bh, 1);

	if (!test_bit(QDF_CHANGE, &qd->qd_flags)) {
		qc->qc_change = 0;
		qc->qc_flags = 0;
		if (test_bit(QDF_USER, &qd->qd_flags))
			qc->qc_flags = cpu_to_be32(GFS2_QCF_USER);
		qc->qc_id = cpu_to_be32(qd->qd_id);
	}

	x = be64_to_cpu(qc->qc_change) + change;
	qc->qc_change = cpu_to_be64(x);

	spin_lock(&sdp->sd_quota_spin);
	qd->qd_change = x;
	spin_unlock(&sdp->sd_quota_spin);

	if (!x) {
		gfs2_assert_warn(sdp, test_bit(QDF_CHANGE, &qd->qd_flags));
		clear_bit(QDF_CHANGE, &qd->qd_flags);
		qc->qc_flags = 0;
		qc->qc_id = 0;
		slot_put(qd);
		qd_put(qd);
	} else if (!test_and_set_bit(QDF_CHANGE, &qd->qd_flags)) {
		qd_hold(qd);
		slot_hold(qd);
	}

	mutex_unlock(&sdp->sd_quota_mutex);
}

static void gfs2_quota_in(struct gfs2_quota_host *qu, const void *buf)
{
	const struct gfs2_quota *str = buf;

	qu->qu_limit = be64_to_cpu(str->qu_limit);
	qu->qu_warn = be64_to_cpu(str->qu_warn);
	qu->qu_value = be64_to_cpu(str->qu_value);
	qu->qu_ll_next = be32_to_cpu(str->qu_ll_next);
}

static void gfs2_quota_out(const struct gfs2_quota_host *qu, void *buf)
{
	struct gfs2_quota *str = buf;

	str->qu_limit = cpu_to_be64(qu->qu_limit);
	str->qu_warn = cpu_to_be64(qu->qu_warn);
	str->qu_value = cpu_to_be64(qu->qu_value);
	str->qu_ll_next = cpu_to_be32(qu->qu_ll_next);
	memset(&str->qu_reserved, 0, sizeof(str->qu_reserved));
}

/**
 * gfs2_adjust_quota
 *
 * This function was mostly borrowed from gfs2_block_truncate_page which was
 * in turn mostly borrowed from ext3
 */
static int gfs2_adjust_quota(struct gfs2_inode *ip, loff_t loc,
			     s64 change, struct gfs2_quota_data *qd)
{
	struct inode *inode = &ip->i_inode;
	struct address_space *mapping = inode->i_mapping;
	unsigned long index = loc >> PAGE_CACHE_SHIFT;
	unsigned offset = loc & (PAGE_CACHE_SIZE - 1);
	unsigned blocksize, iblock, pos;
	struct buffer_head *bh;
	struct page *page;
	void *kaddr;
	char *ptr;
	struct gfs2_quota_host qp;
	s64 value;
	int err = -EIO;

	if (gfs2_is_stuffed(ip)) {
		struct gfs2_alloc *al = NULL;
		al = gfs2_alloc_get(ip);
		/* just request 1 blk */
		al->al_requested = 1;
		gfs2_inplace_reserve(ip);
		gfs2_unstuff_dinode(ip, NULL);
		gfs2_inplace_release(ip);
		gfs2_alloc_put(ip);
	}
	page = grab_cache_page(mapping, index);
	if (!page)
		return -ENOMEM;

	blocksize = inode->i_sb->s_blocksize;
	iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits);

	if (!page_has_buffers(page))
		create_empty_buffers(page, blocksize, 0);

	bh = page_buffers(page);
	pos = blocksize;
	while (offset >= pos) {
		bh = bh->b_this_page;