^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Copyright (C) 2010 Red Hat, Inc.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) * All Rights Reserved.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) #include "xfs.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) #include "xfs_shared.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) #include "xfs_format.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) #include "xfs_log_format.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) #include "xfs_trans_resv.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) #include "xfs_sb.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include "xfs_mount.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #include "xfs_btree.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) #include "xfs_alloc_btree.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include "xfs_alloc.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) #include "xfs_discard.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) #include "xfs_error.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) #include "xfs_extent_busy.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) #include "xfs_trace.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) #include "xfs_log.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) STATIC int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) xfs_trim_extents(
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) struct xfs_mount *mp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) xfs_agnumber_t agno,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) xfs_daddr_t start,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) xfs_daddr_t end,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) xfs_daddr_t minlen,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) uint64_t *blocks_trimmed)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) struct block_device *bdev = mp->m_ddev_targp->bt_bdev;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) struct xfs_btree_cur *cur;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) struct xfs_buf *agbp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) struct xfs_agf *agf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) struct xfs_perag *pag;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) int error;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) pag = xfs_perag_get(mp, agno);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) * Force out the log. This means any transactions that might have freed
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) * space before we take the AGF buffer lock are now on disk, and the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) * volatile disk cache is flushed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) xfs_log_force(mp, XFS_LOG_SYNC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) error = xfs_alloc_read_agf(mp, NULL, agno, 0, &agbp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) if (error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) goto out_put_perag;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) agf = agbp->b_addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) cur = xfs_allocbt_init_cursor(mp, NULL, agbp, agno, XFS_BTNUM_CNT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) * Look up the longest btree in the AGF and start with it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) error = xfs_alloc_lookup_ge(cur, 0, be32_to_cpu(agf->agf_longest), &i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) if (error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) * Loop until we are done with all extents that are large
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) * enough to be worth discarding.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) while (i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) xfs_agblock_t fbno;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) xfs_extlen_t flen;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) xfs_daddr_t dbno;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) xfs_extlen_t dlen;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) error = xfs_alloc_get_rec(cur, &fbno, &flen, &i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) if (error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) if (XFS_IS_CORRUPT(mp, i != 1)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) error = -EFSCORRUPTED;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) ASSERT(flen <= be32_to_cpu(agf->agf_longest));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) * use daddr format for all range/len calculations as that is
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) * the format the range/len variables are supplied in by
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) * userspace.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) dbno = XFS_AGB_TO_DADDR(mp, agno, fbno);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) dlen = XFS_FSB_TO_BB(mp, flen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) * Too small? Give up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) if (dlen < minlen) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) trace_xfs_discard_toosmall(mp, agno, fbno, flen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) * If the extent is entirely outside of the range we are
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) * supposed to discard skip it. Do not bother to trim
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) * down partially overlapping ranges for now.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) if (dbno + dlen < start || dbno > end) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) trace_xfs_discard_exclude(mp, agno, fbno, flen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) goto next_extent;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) * If any blocks in the range are still busy, skip the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) * discard and try again the next time.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) if (xfs_extent_busy_search(mp, agno, fbno, flen)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) trace_xfs_discard_busy(mp, agno, fbno, flen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) goto next_extent;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) trace_xfs_discard_extent(mp, agno, fbno, flen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) error = blkdev_issue_discard(bdev, dbno, dlen, GFP_NOFS, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) if (error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) *blocks_trimmed += flen;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) next_extent:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) error = xfs_btree_decrement(cur, 0, &i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) if (error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) if (fatal_signal_pending(current)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) error = -ERESTARTSYS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) goto out_del_cursor;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) out_del_cursor:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) xfs_btree_del_cursor(cur, error);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) xfs_buf_relse(agbp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) out_put_perag:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) xfs_perag_put(pag);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) return error;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) * trim a range of the filesystem.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) * Note: the parameters passed from userspace are byte ranges into the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) * filesystem which does not match to the format we use for filesystem block
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) * addressing. FSB addressing is sparse (AGNO|AGBNO), while the incoming format
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) * is a linear address range. Hence we need to use DADDR based conversions and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) * comparisons for determining the correct offset and regions to trim.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) xfs_ioc_trim(
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) struct xfs_mount *mp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) struct fstrim_range __user *urange)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) struct request_queue *q = bdev_get_queue(mp->m_ddev_targp->bt_bdev);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) unsigned int granularity = q->limits.discard_granularity;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) struct fstrim_range range;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) xfs_daddr_t start, end, minlen;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) xfs_agnumber_t start_agno, end_agno, agno;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) uint64_t blocks_trimmed = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) int error, last_error = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) if (!capable(CAP_SYS_ADMIN))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) return -EPERM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) if (!blk_queue_discard(q))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) return -EOPNOTSUPP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) * We haven't recovered the log, so we cannot use our bnobt-guided
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) * storage zapping commands.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) if (mp->m_flags & XFS_MOUNT_NORECOVERY)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) return -EROFS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) if (copy_from_user(&range, urange, sizeof(range)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) return -EFAULT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) range.minlen = max_t(u64, granularity, range.minlen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) minlen = BTOBB(range.minlen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) * Truncating down the len isn't actually quite correct, but using
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) * BBTOB would mean we trivially get overflows for values
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) * of ULLONG_MAX or slightly lower. And ULLONG_MAX is the default
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) * used by the fstrim application. In the end it really doesn't
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) * matter as trimming blocks is an advisory interface.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) if (range.start >= XFS_FSB_TO_B(mp, mp->m_sb.sb_dblocks) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) range.minlen > XFS_FSB_TO_B(mp, mp->m_ag_max_usable) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) range.len < mp->m_sb.sb_blocksize)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) return -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) start = BTOBB(range.start);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) end = start + BTOBBT(range.len) - 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) if (end > XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks) - 1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) end = XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)- 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) start_agno = xfs_daddr_to_agno(mp, start);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) end_agno = xfs_daddr_to_agno(mp, end);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) for (agno = start_agno; agno <= end_agno; agno++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) error = xfs_trim_extents(mp, agno, start, end, minlen,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) &blocks_trimmed);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) if (error) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) last_error = error;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) if (error == -ERESTARTSYS)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) if (last_error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) return last_error;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) range.len = XFS_FSB_TO_B(mp, blocks_trimmed);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) if (copy_to_user(urange, &range, sizeof(range)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) return -EFAULT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) }