/*
* trim a range of the filesystem.
*
* Note: the parameters passed from userspace are byte ranges into the
* filesystem which does not match to the format we use for filesystem block
* addressing. FSB addressing is sparse (AGNO|AGBNO), while the incoming format
* is a linear address range. Hence we need to use DADDR based conversions and
* comparisons for determining the correct offset and regions to trim.
*/
int
xfs_ioc_trim(
struct xfs_mount *mp,
struct fstrim_range __user *urange)
{
struct request_queue *q = bdev_get_queue(mp->m_ddev_targp->bt_bdev);
unsigned int granularity = q->limits.discard_granularity;
struct fstrim_range range;
xfs_daddr_t start, end, minlen;
xfs_agnumber_t start_agno, end_agno, agno;
__uint64_t blocks_trimmed = 0;
int error, last_error = 0;
if (!capable(CAP_SYS_ADMIN))
return -XFS_ERROR(EPERM);
if (!blk_queue_discard(q))
return -XFS_ERROR(EOPNOTSUPP);
if (copy_from_user(&range, urange, sizeof(range)))
return -XFS_ERROR(EFAULT);
/*
* Truncating down the len isn't actually quite correct, but using
* BBTOB would mean we trivially get overflows for values
* of ULLONG_MAX or slightly lower. And ULLONG_MAX is the default
* used by the fstrim application. In the end it really doesn't
* matter as trimming blocks is an advisory interface.
*/
if (range.start >= XFS_FSB_TO_B(mp, mp->m_sb.sb_dblocks) ||
range.minlen > XFS_FSB_TO_B(mp, XFS_ALLOC_AG_MAX_USABLE(mp)) ||
range.len < mp->m_sb.sb_blocksize)
return -XFS_ERROR(EINVAL);
start = BTOBB(range.start);
end = start + BTOBBT(range.len) - 1;
minlen = BTOBB(max_t(u64, granularity, range.minlen));
if (end > XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks) - 1)
end = XFS_FSB_TO_BB(mp, mp->m_sb.sb_dblocks)- 1;
start_agno = xfs_daddr_to_agno(mp, start);
end_agno = xfs_daddr_to_agno(mp, end);
for (agno = start_agno; agno <= end_agno; agno++) {
error = -xfs_trim_extents(mp, agno, start, end, minlen,
&blocks_trimmed);
if (error)
last_error = error;
}
if (last_error)
return last_error;
range.len = XFS_FSB_TO_B(mp, blocks_trimmed);
if (copy_to_user(urange, &range, sizeof(range)))
return -XFS_ERROR(EFAULT);
return 0;
}
/*
* Calculate the minimum valid log size for the given superblock configuration.
* Used to calculate the minimum log size at mkfs time, and to determine if
* the log is large enough or not at mount time. Returns the minimum size in
* filesystem block size units.
*/
int
xfs_log_calc_minimum_size(
struct xfs_mount *mp)
{
struct xfs_trans_res tres = {0};
int max_logres;
int min_logblks = 0;
int lsunit = 0;
xfs_log_get_max_trans_res(mp, &tres);
max_logres = xfs_log_calc_unit_res(mp, tres.tr_logres);
if (tres.tr_logcount > 1)
max_logres *= tres.tr_logcount;
if (xfs_sb_version_haslogv2(&mp->m_sb) && mp->m_sb.sb_logsunit > 1)
lsunit = BTOBB(mp->m_sb.sb_logsunit);
/*
* Two factors should be taken into account for calculating the minimum
* log space.
* 1) The fundamental limitation is that no single transaction can be
* larger than half size of the log.
*
* From mkfs.xfs, this is considered by the XFS_MIN_LOG_FACTOR
* define, which is set to 3. That means we can definitely fit
* maximally sized 2 transactions in the log. We'll use this same
* value here.
*
* 2) If the lsunit option is specified, a transaction requires 2 LSU
* for the reservation because there are two log writes that can
* require padding - the transaction data and the commit record which
* are written separately and both can require padding to the LSU.
* Consider that we can have an active CIL reservation holding 2*LSU,
* but the CIL is not over a push threshold, in this case, if we
* don't have enough log space for at one new transaction, which
* includes another 2*LSU in the reservation, we will run into dead
* loop situation in log space grant procedure. i.e.
* xlog_grant_head_wait().
*
* Hence the log size needs to be able to contain two maximally sized
* and padded transactions, which is (2 * (2 * LSU + maxlres)).
*
* Also, the log size should be a multiple of the log stripe unit, round
* it up to lsunit boundary if lsunit is specified.
*/
if (lsunit) {
min_logblks = roundup_64(BTOBB(max_logres), lsunit) +
2 * lsunit;
} else
min_logblks = BTOBB(max_logres) + 2 * BBSIZE;
min_logblks *= XFS_MIN_LOG_FACTOR;
return XFS_BB_TO_FSB(mp, min_logblks);
}
STATIC int
xfs_vn_fiemap(
struct inode *inode,
struct fiemap_extent_info *fieinfo,
u64 start,
u64 length)
{
xfs_inode_t *ip = XFS_I(inode);
struct getbmapx bm;
int error;
error = fiemap_check_flags(fieinfo, XFS_FIEMAP_FLAGS);
if (error)
return error;
/* Set up bmap header for xfs internal routine */
bm.bmv_offset = BTOBB(start);
/* Special case for whole file */
if (length == FIEMAP_MAX_OFFSET)
bm.bmv_length = -1LL;
else
bm.bmv_length = BTOBB(length);
/* We add one because in getbmap world count includes the header */
bm.bmv_count = !fieinfo->fi_extents_max ? MAXEXTNUM :
fieinfo->fi_extents_max + 1;
bm.bmv_count = min_t(__s32, bm.bmv_count,
(PAGE_SIZE * 16 / sizeof(struct getbmapx)));
bm.bmv_iflags = BMV_IF_PREALLOC | BMV_IF_NO_HOLES;
if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR)
bm.bmv_iflags |= BMV_IF_ATTRFORK;
if (!(fieinfo->fi_flags & FIEMAP_FLAG_SYNC))
bm.bmv_iflags |= BMV_IF_DELALLOC;
error = xfs_getbmap(ip, &bm, xfs_fiemap_format, fieinfo);
if (error)
return -error;
return 0;
}
STATIC int
xfs_vn_fiemap(
struct inode *inode,
struct fiemap_extent_info *fieinfo,
u64 start,
u64 length)
{
xfs_inode_t *ip = XFS_I(inode);
struct getbmapx bm;
int error;
error = fiemap_check_flags(fieinfo, XFS_FIEMAP_FLAGS);
if (error)
return error;
/* Set up bmap header for xfs internal routine */
bm.bmv_offset = BTOBB(start);
/* Special case for whole file */
if (length == FIEMAP_MAX_OFFSET)
bm.bmv_length = -1LL;
else
bm.bmv_length = BTOBB(length);
/* our formatter will tell xfs_getbmap when to stop. */
bm.bmv_count = MAXEXTNUM;
bm.bmv_iflags = BMV_IF_PREALLOC;
if (fieinfo->fi_flags & FIEMAP_FLAG_XATTR)
bm.bmv_iflags |= BMV_IF_ATTRFORK;
if (!(fieinfo->fi_flags & FIEMAP_FLAG_SYNC))
bm.bmv_iflags |= BMV_IF_DELALLOC;
error = xfs_getbmap(ip, &bm, xfs_fiemap_format, fieinfo);
if (error)
return -error;
return 0;
}
STATIC void
xfs_attr_rmtval_copyin(
struct xfs_mount *mp,
struct xfs_buf *bp,
xfs_ino_t ino,
int *offset,
int *valuelen,
__uint8_t **src)
{
char *dst = bp->b_addr;
xfs_daddr_t bno = bp->b_bn;
int len = BBTOB(bp->b_length);
int blksize = mp->m_attr_geo->blksize;
ASSERT(len >= blksize);
while (len > 0 && *valuelen > 0) {
int hdr_size;
int byte_cnt = XFS_ATTR3_RMT_BUF_SPACE(mp, blksize);
byte_cnt = min(*valuelen, byte_cnt);
hdr_size = xfs_attr3_rmt_hdr_set(mp, dst, ino, *offset,
byte_cnt, bno);
memcpy(dst + hdr_size, *src, byte_cnt);
/*
* If this is the last block, zero the remainder of it.
* Check that we are actually the last block, too.
*/
if (byte_cnt + hdr_size < blksize) {
ASSERT(*valuelen - byte_cnt == 0);
ASSERT(len == blksize);
memset(dst + hdr_size + byte_cnt, 0,
blksize - hdr_size - byte_cnt);
}
/* roll buffer forwards */
len -= blksize;
dst += blksize;
bno += BTOBB(blksize);
/* roll attribute data forwards */
*valuelen -= byte_cnt;
*src += byte_cnt;
*offset += byte_cnt;
}
}
/*
* Helper functions to copy attribute data in and out of the one disk extents
*/
STATIC int
xfs_attr_rmtval_copyout(
struct xfs_mount *mp,
struct xfs_buf *bp,
xfs_ino_t ino,
int *offset,
int *valuelen,
__uint8_t **dst)
{
char *src = bp->b_addr;
xfs_daddr_t bno = bp->b_bn;
int len = BBTOB(bp->b_length);
int blksize = mp->m_attr_geo->blksize;
ASSERT(len >= blksize);
while (len > 0 && *valuelen > 0) {
int hdr_size = 0;
int byte_cnt = XFS_ATTR3_RMT_BUF_SPACE(mp, blksize);
byte_cnt = min(*valuelen, byte_cnt);
if (xfs_sb_version_hascrc(&mp->m_sb)) {
if (!xfs_attr3_rmt_hdr_ok(src, ino, *offset,
byte_cnt, bno)) {
xfs_alert(mp,
"remote attribute header mismatch bno/off/len/owner (0x%llx/0x%x/Ox%x/0x%llx)",
bno, *offset, byte_cnt, ino);
return -EFSCORRUPTED;
}
hdr_size = sizeof(struct xfs_attr3_rmt_hdr);
}
memcpy(*dst, src + hdr_size, byte_cnt);
/* roll buffer forwards */
len -= blksize;
src += blksize;
bno += BTOBB(blksize);
/* roll attribute data forwards */
*valuelen -= byte_cnt;
*dst += byte_cnt;
*offset += byte_cnt;
}
return 0;
}
static void
xfs_attr3_rmt_write_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
xfs_failaddr_t fa;
int blksize = mp->m_attr_geo->blksize;
char *ptr;
int len;
xfs_daddr_t bno;
/* no verification of non-crc buffers */
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
ptr = bp->b_addr;
bno = bp->b_bn;
len = BBTOB(bp->b_length);
ASSERT(len >= blksize);
while (len > 0) {
struct xfs_attr3_rmt_hdr *rmt = (struct xfs_attr3_rmt_hdr *)ptr;
fa = xfs_attr3_rmt_verify(mp, ptr, blksize, bno);
if (fa) {
xfs_verifier_error(bp, -EFSCORRUPTED, fa);
return;
}
/*
* Ensure we aren't writing bogus LSNs to disk. See
* xfs_attr3_rmt_hdr_set() for the explanation.
*/
if (rmt->rm_lsn != cpu_to_be64(NULLCOMMITLSN)) {
xfs_verifier_error(bp, -EFSCORRUPTED, __this_address);
return;
}
xfs_update_cksum(ptr, blksize, XFS_ATTR3_RMT_CRC_OFF);
len -= blksize;
ptr += blksize;
bno += BTOBB(blksize);
}
if (len != 0)
xfs_verifier_error(bp, -EFSCORRUPTED, __this_address);
}
static int
__xfs_attr3_rmt_read_verify(
struct xfs_buf *bp,
bool check_crc,
xfs_failaddr_t *failaddr)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
char *ptr;
int len;
xfs_daddr_t bno;
int blksize = mp->m_attr_geo->blksize;
/* no verification of non-crc buffers */
if (!xfs_sb_version_hascrc(&mp->m_sb))
return 0;
ptr = bp->b_addr;
bno = bp->b_bn;
len = BBTOB(bp->b_length);
ASSERT(len >= blksize);
while (len > 0) {
if (check_crc &&
!xfs_verify_cksum(ptr, blksize, XFS_ATTR3_RMT_CRC_OFF)) {
*failaddr = __this_address;
return -EFSBADCRC;
}
*failaddr = xfs_attr3_rmt_verify(mp, ptr, blksize, bno);
if (*failaddr)
return -EFSCORRUPTED;
len -= blksize;
ptr += blksize;
bno += BTOBB(blksize);
}
if (len != 0) {
*failaddr = __this_address;
return -EFSCORRUPTED;
}
return 0;
}
static void
xfs_attr3_rmt_write_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
struct xfs_buf_log_item *bip = bp->b_fspriv;
char *ptr;
int len;
xfs_daddr_t bno;
int blksize = mp->m_attr_geo->blksize;
/* no verification of non-crc buffers */
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
ptr = bp->b_addr;
bno = bp->b_bn;
len = BBTOB(bp->b_length);
ASSERT(len >= blksize);
while (len > 0) {
if (!xfs_attr3_rmt_verify(mp, ptr, blksize, bno)) {
xfs_buf_ioerror(bp, -EFSCORRUPTED);
xfs_verifier_error(bp);
return;
}
if (bip) {
struct xfs_attr3_rmt_hdr *rmt;
rmt = (struct xfs_attr3_rmt_hdr *)ptr;
rmt->rm_lsn = cpu_to_be64(bip->bli_item.li_lsn);
}
xfs_update_cksum(ptr, blksize, XFS_ATTR3_RMT_CRC_OFF);
len -= blksize;
ptr += blksize;
bno += BTOBB(blksize);
}
ASSERT(len == 0);
}
static void
xfs_attr3_rmt_read_verify(
struct xfs_buf *bp)
{
struct xfs_mount *mp = bp->b_target->bt_mount;
char *ptr;
int len;
xfs_daddr_t bno;
int blksize = mp->m_attr_geo->blksize;
/* no verification of non-crc buffers */
if (!xfs_sb_version_hascrc(&mp->m_sb))
return;
ptr = bp->b_addr;
bno = bp->b_bn;
len = BBTOB(bp->b_length);
ASSERT(len >= blksize);
while (len > 0) {
if (!xfs_verify_cksum(ptr, blksize, XFS_ATTR3_RMT_CRC_OFF)) {
xfs_buf_ioerror(bp, -EFSBADCRC);
break;
}
if (!xfs_attr3_rmt_verify(mp, ptr, blksize, bno)) {
xfs_buf_ioerror(bp, -EFSCORRUPTED);
break;
}
len -= blksize;
ptr += blksize;
bno += BTOBB(blksize);
}
if (bp->b_error)
xfs_verifier_error(bp);
else
ASSERT(len == 0);
}
/* ----- Kernel only functions below ----- */
STATIC int
xfs_readlink_bmap(
struct xfs_inode *ip,
char *link)
{
struct xfs_mount *mp = ip->i_mount;
struct xfs_bmbt_irec mval[XFS_SYMLINK_MAPS];
struct xfs_buf *bp;
xfs_daddr_t d;
char *cur_chunk;
int pathlen = ip->i_d.di_size;
int nmaps = XFS_SYMLINK_MAPS;
int byte_cnt;
int n;
int error = 0;
int fsblocks = 0;
int offset;
fsblocks = xfs_symlink_blocks(mp, pathlen);
error = xfs_bmapi_read(ip, 0, fsblocks, mval, &nmaps, 0);
if (error)
goto out;
offset = 0;
for (n = 0; n < nmaps; n++) {
d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0,
&xfs_symlink_buf_ops);
if (!bp)
return XFS_ERROR(ENOMEM);
error = bp->b_error;
if (error) {
xfs_buf_ioerror_alert(bp, __func__);
xfs_buf_relse(bp);
/* bad CRC means corrupted metadata */
if (error == EFSBADCRC)
error = EFSCORRUPTED;
goto out;
}
byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
if (pathlen < byte_cnt)
byte_cnt = pathlen;
cur_chunk = bp->b_addr;
if (xfs_sb_version_hascrc(&mp->m_sb)) {
if (!xfs_symlink_hdr_ok(mp, ip->i_ino, offset,
byte_cnt, bp)) {
error = EFSCORRUPTED;
xfs_alert(mp,
"symlink header does not match required off/len/owner (0x%x/Ox%x,0x%llx)",
offset, byte_cnt, ip->i_ino);
xfs_buf_relse(bp);
goto out;
}
cur_chunk += sizeof(struct xfs_dsymlink_hdr);
}
memcpy(link + offset, bp->b_addr, byte_cnt);
pathlen -= byte_cnt;
offset += byte_cnt;
xfs_buf_relse(bp);
}
ASSERT(pathlen == 0);
link[ip->i_d.di_size] = '\0';
error = 0;
out:
return error;
}
int
xfs_symlink(
struct xfs_inode *dp,
struct xfs_name *link_name,
const char *target_path,
umode_t mode,
struct xfs_inode **ipp)
{
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans *tp = NULL;
struct xfs_inode *ip = NULL;
int error = 0;
int pathlen;
struct xfs_bmap_free free_list;
xfs_fsblock_t first_block;
bool unlock_dp_on_error = false;
uint cancel_flags;
int committed;
xfs_fileoff_t first_fsb;
xfs_filblks_t fs_blocks;
int nmaps;
struct xfs_bmbt_irec mval[XFS_SYMLINK_MAPS];
xfs_daddr_t d;
const char *cur_chunk;
int byte_cnt;
int n;
xfs_buf_t *bp;
prid_t prid;
struct xfs_dquot *udqp = NULL;
struct xfs_dquot *gdqp = NULL;
struct xfs_dquot *pdqp = NULL;
uint resblks;
*ipp = NULL;
trace_xfs_symlink(dp, link_name);
if (XFS_FORCED_SHUTDOWN(mp))
return XFS_ERROR(EIO);
/*
* Check component lengths of the target path name.
*/
pathlen = strlen(target_path);
if (pathlen >= MAXPATHLEN) /* total string too long */
return XFS_ERROR(ENAMETOOLONG);
udqp = gdqp = NULL;
if (dp->i_d.di_flags & XFS_DIFLAG_PROJINHERIT)
prid = xfs_get_projid(dp);
else
prid = XFS_PROJID_DEFAULT;
/*
* Make sure that we have allocated dquot(s) on disk.
*/
error = xfs_qm_vop_dqalloc(dp,
xfs_kuid_to_uid(current_fsuid()),
xfs_kgid_to_gid(current_fsgid()), prid,
XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT,
&udqp, &gdqp, &pdqp);
if (error)
goto std_return;
tp = xfs_trans_alloc(mp, XFS_TRANS_SYMLINK);
cancel_flags = XFS_TRANS_RELEASE_LOG_RES;
/*
* The symlink will fit into the inode data fork?
* There can't be any attributes so we get the whole variable part.
*/
if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
fs_blocks = 0;
else
fs_blocks = xfs_symlink_blocks(mp, pathlen);
resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, resblks, 0);
if (error == ENOSPC && fs_blocks == 0) {
resblks = 0;
error = xfs_trans_reserve(tp, &M_RES(mp)->tr_symlink, 0, 0);
}
if (error) {
cancel_flags = 0;
goto error_return;
}
xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
unlock_dp_on_error = true;
/*
* Check whether the directory allows new symlinks or not.
*/
if (dp->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) {
error = XFS_ERROR(EPERM);
goto error_return;
}
/*
* Reserve disk quota : blocks and inode.
*/
error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp,
pdqp, resblks, 1, 0);
if (error)
goto error_return;
/*
* Check for ability to enter directory entry, if no space reserved.
*/
error = xfs_dir_canenter(tp, dp, link_name, resblks);
if (error)
goto error_return;
/*
* Initialize the bmap freelist prior to calling either
* bmapi or the directory create code.
*/
xfs_bmap_init(&free_list, &first_block);
/*
* Allocate an inode for the symlink.
*/
error = xfs_dir_ialloc(&tp, dp, S_IFLNK | (mode & ~S_IFMT), 1, 0,
prid, resblks > 0, &ip, NULL);
if (error) {
if (error == ENOSPC)
goto error_return;
goto error1;
}
/*
* An error after we've joined dp to the transaction will result in the
* transaction cancel unlocking dp so don't do it explicitly in the
* error path.
*/
xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
unlock_dp_on_error = false;
/*
* Also attach the dquot(s) to it, if applicable.
*/
xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp, pdqp);
if (resblks)
resblks -= XFS_IALLOC_SPACE_RES(mp);
/*
* If the symlink will fit into the inode, write it inline.
*/
if (pathlen <= XFS_IFORK_DSIZE(ip)) {
xfs_idata_realloc(ip, pathlen, XFS_DATA_FORK);
memcpy(ip->i_df.if_u1.if_data, target_path, pathlen);
ip->i_d.di_size = pathlen;
/*
* The inode was initially created in extent format.
*/
ip->i_df.if_flags &= ~(XFS_IFEXTENTS | XFS_IFBROOT);
ip->i_df.if_flags |= XFS_IFINLINE;
ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
} else {
int offset;
first_fsb = 0;
nmaps = XFS_SYMLINK_MAPS;
error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
XFS_BMAPI_METADATA, &first_block, resblks,
mval, &nmaps, &free_list);
if (error)
goto error2;
if (resblks)
resblks -= fs_blocks;
ip->i_d.di_size = pathlen;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
cur_chunk = target_path;
offset = 0;
for (n = 0; n < nmaps; n++) {
char *buf;
d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
BTOBB(byte_cnt), 0);
if (!bp) {
error = ENOMEM;
goto error2;
}
bp->b_ops = &xfs_symlink_buf_ops;
byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
byte_cnt = min(byte_cnt, pathlen);
buf = bp->b_addr;
buf += xfs_symlink_hdr_set(mp, ip->i_ino, offset,
byte_cnt, bp);
memcpy(buf, cur_chunk, byte_cnt);
cur_chunk += byte_cnt;
pathlen -= byte_cnt;
offset += byte_cnt;
xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SYMLINK_BUF);
xfs_trans_log_buf(tp, bp, 0, (buf + byte_cnt - 1) -
(char *)bp->b_addr);
}
ASSERT(pathlen == 0);
}
/*
* Create the directory entry for the symlink.
*/
error = xfs_dir_createname(tp, dp, link_name, ip->i_ino,
&first_block, &free_list, resblks);
if (error)
goto error2;
xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
/*
* If this is a synchronous mount, make sure that the
* symlink transaction goes to disk before returning to
* the user.
*/
if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
xfs_trans_set_sync(tp);
}
error = xfs_bmap_finish(&tp, &free_list, &committed);
if (error) {
goto error2;
}
error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES);
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
*ipp = ip;
return 0;
error2:
IRELE(ip);
error1:
xfs_bmap_cancel(&free_list);
cancel_flags |= XFS_TRANS_ABORT;
error_return:
xfs_trans_cancel(tp, cancel_flags);
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
if (unlock_dp_on_error)
xfs_iunlock(dp, XFS_ILOCK_EXCL);
std_return:
return error;
}
int
xfs_symlink(
struct xfs_inode *dp,
struct xfs_name *link_name,
const char *target_path,
umode_t mode,
struct xfs_inode **ipp)
{
struct xfs_mount *mp = dp->i_mount;
struct xfs_trans *tp = NULL;
struct xfs_inode *ip = NULL;
int error = 0;
int pathlen;
struct xfs_defer_ops dfops;
xfs_fsblock_t first_block;
bool unlock_dp_on_error = false;
xfs_fileoff_t first_fsb;
xfs_filblks_t fs_blocks;
int nmaps;
struct xfs_bmbt_irec mval[XFS_SYMLINK_MAPS];
xfs_daddr_t d;
const char *cur_chunk;
int byte_cnt;
int n;
xfs_buf_t *bp;
prid_t prid;
struct xfs_dquot *udqp = NULL;
struct xfs_dquot *gdqp = NULL;
struct xfs_dquot *pdqp = NULL;
uint resblks;
*ipp = NULL;
trace_xfs_symlink(dp, link_name);
if (XFS_FORCED_SHUTDOWN(mp))
return -EIO;
/*
* Check component lengths of the target path name.
*/
pathlen = strlen(target_path);
if (pathlen >= MAXPATHLEN) /* total string too long */
return -ENAMETOOLONG;
udqp = gdqp = NULL;
prid = xfs_get_initial_prid(dp);
/*
* Make sure that we have allocated dquot(s) on disk.
*/
error = xfs_qm_vop_dqalloc(dp,
xfs_kuid_to_uid(current_fsuid()),
xfs_kgid_to_gid(current_fsgid()), prid,
XFS_QMOPT_QUOTALL | XFS_QMOPT_INHERIT,
&udqp, &gdqp, &pdqp);
if (error)
return error;
/*
* The symlink will fit into the inode data fork?
* There can't be any attributes so we get the whole variable part.
*/
if (pathlen <= XFS_LITINO(mp, dp->i_d.di_version))
fs_blocks = 0;
else
fs_blocks = xfs_symlink_blocks(mp, pathlen);
resblks = XFS_SYMLINK_SPACE_RES(mp, link_name->len, fs_blocks);
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_symlink, resblks, 0, 0, &tp);
if (error == -ENOSPC && fs_blocks == 0) {
resblks = 0;
error = xfs_trans_alloc(mp, &M_RES(mp)->tr_symlink, 0, 0, 0,
&tp);
}
if (error)
goto out_release_inode;
xfs_ilock(dp, XFS_ILOCK_EXCL | XFS_ILOCK_PARENT);
unlock_dp_on_error = true;
/*
* Check whether the directory allows new symlinks or not.
*/
if (dp->i_d.di_flags & XFS_DIFLAG_NOSYMLINKS) {
error = -EPERM;
goto out_trans_cancel;
}
/*
* Reserve disk quota : blocks and inode.
*/
error = xfs_trans_reserve_quota(tp, mp, udqp, gdqp,
pdqp, resblks, 1, 0);
if (error)
goto out_trans_cancel;
/*
* Check for ability to enter directory entry, if no space reserved.
*/
if (!resblks) {
error = xfs_dir_canenter(tp, dp, link_name);
if (error)
goto out_trans_cancel;
}
/*
* Initialize the bmap freelist prior to calling either
* bmapi or the directory create code.
*/
xfs_defer_init(&dfops, &first_block);
/*
* Allocate an inode for the symlink.
*/
error = xfs_dir_ialloc(&tp, dp, S_IFLNK | (mode & ~S_IFMT), 1, 0,
prid, resblks > 0, &ip, NULL);
if (error)
goto out_trans_cancel;
/*
* Now we join the directory inode to the transaction. We do not do it
* earlier because xfs_dir_ialloc might commit the previous transaction
* (and release all the locks). An error from here on will result in
* the transaction cancel unlocking dp so don't do it explicitly in the
* error path.
*/
xfs_trans_ijoin(tp, dp, XFS_ILOCK_EXCL);
unlock_dp_on_error = false;
/*
* Also attach the dquot(s) to it, if applicable.
*/
xfs_qm_vop_create_dqattach(tp, ip, udqp, gdqp, pdqp);
if (resblks)
resblks -= XFS_IALLOC_SPACE_RES(mp);
/*
* If the symlink will fit into the inode, write it inline.
*/
if (pathlen <= XFS_IFORK_DSIZE(ip)) {
xfs_init_local_fork(ip, XFS_DATA_FORK, target_path, pathlen);
ip->i_d.di_size = pathlen;
ip->i_d.di_format = XFS_DINODE_FMT_LOCAL;
xfs_trans_log_inode(tp, ip, XFS_ILOG_DDATA | XFS_ILOG_CORE);
} else {
int offset;
first_fsb = 0;
nmaps = XFS_SYMLINK_MAPS;
error = xfs_bmapi_write(tp, ip, first_fsb, fs_blocks,
XFS_BMAPI_METADATA, &first_block, resblks,
mval, &nmaps, &dfops);
if (error)
goto out_bmap_cancel;
if (resblks)
resblks -= fs_blocks;
ip->i_d.di_size = pathlen;
xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE);
cur_chunk = target_path;
offset = 0;
for (n = 0; n < nmaps; n++) {
char *buf;
d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock);
byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount);
bp = xfs_trans_get_buf(tp, mp->m_ddev_targp, d,
BTOBB(byte_cnt), 0);
if (!bp) {
error = -ENOMEM;
goto out_bmap_cancel;
}
bp->b_ops = &xfs_symlink_buf_ops;
byte_cnt = XFS_SYMLINK_BUF_SPACE(mp, byte_cnt);
byte_cnt = min(byte_cnt, pathlen);
buf = bp->b_addr;
buf += xfs_symlink_hdr_set(mp, ip->i_ino, offset,
byte_cnt, bp);
memcpy(buf, cur_chunk, byte_cnt);
cur_chunk += byte_cnt;
pathlen -= byte_cnt;
offset += byte_cnt;
xfs_trans_buf_set_type(tp, bp, XFS_BLFT_SYMLINK_BUF);
xfs_trans_log_buf(tp, bp, 0, (buf + byte_cnt - 1) -
(char *)bp->b_addr);
}
ASSERT(pathlen == 0);
}
/*
* Create the directory entry for the symlink.
*/
error = xfs_dir_createname(tp, dp, link_name, ip->i_ino,
&first_block, &dfops, resblks);
if (error)
goto out_bmap_cancel;
xfs_trans_ichgtime(tp, dp, XFS_ICHGTIME_MOD | XFS_ICHGTIME_CHG);
xfs_trans_log_inode(tp, dp, XFS_ILOG_CORE);
/*
* If this is a synchronous mount, make sure that the
* symlink transaction goes to disk before returning to
* the user.
*/
if (mp->m_flags & (XFS_MOUNT_WSYNC|XFS_MOUNT_DIRSYNC)) {
xfs_trans_set_sync(tp);
}
error = xfs_defer_finish(&tp, &dfops, NULL);
if (error)
goto out_bmap_cancel;
error = xfs_trans_commit(tp);
if (error)
goto out_release_inode;
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
*ipp = ip;
return 0;
out_bmap_cancel:
xfs_defer_cancel(&dfops);
out_trans_cancel:
xfs_trans_cancel(tp);
out_release_inode:
/*
* Wait until after the current transaction is aborted to finish the
* setup of the inode and release the inode. This prevents recursive
* transactions and deadlocks from xfs_inactive.
*/
if (ip) {
xfs_finish_inode_setup(ip);
IRELE(ip);
}
xfs_qm_dqrele(udqp);
xfs_qm_dqrele(gdqp);
xfs_qm_dqrele(pdqp);
if (unlock_dp_on_error)
xfs_iunlock(dp, XFS_ILOCK_EXCL);
return error;
}
static void
zero_log(
struct xfs_mount *mp)
{
int error;
xfs_daddr_t head_blk;
xfs_daddr_t tail_blk;
struct xlog *log = mp->m_log;
memset(log, 0, sizeof(struct xlog));
x.logBBsize = XFS_FSB_TO_BB(mp, mp->m_sb.sb_logblocks);
x.logBBstart = XFS_FSB_TO_DADDR(mp, mp->m_sb.sb_logstart);
x.lbsize = BBSIZE;
if (xfs_sb_version_hassector(&mp->m_sb))
x.lbsize <<= (mp->m_sb.sb_logsectlog - BBSHIFT);
log->l_dev = mp->m_logdev_targp;
log->l_logBBsize = x.logBBsize;
log->l_logBBstart = x.logBBstart;
log->l_sectBBsize = BTOBB(x.lbsize);
log->l_mp = mp;
if (xfs_sb_version_hassector(&mp->m_sb)) {
log->l_sectbb_log = mp->m_sb.sb_logsectlog - BBSHIFT;
ASSERT(log->l_sectbb_log <= mp->m_sectbb_log);
/* for larger sector sizes, must have v2 or external log */
ASSERT(log->l_sectbb_log == 0 ||
log->l_logBBstart == 0 ||
xfs_sb_version_haslogv2(&mp->m_sb));
ASSERT(mp->m_sb.sb_logsectlog >= BBSHIFT);
}
log->l_sectbb_mask = (1 << log->l_sectbb_log) - 1;
/*
* Find the log head and tail and alert the user to the situation if the
* log appears corrupted or contains data. In either case, we do not
* proceed past this point unless the user explicitly requests to zap
* the log.
*/
error = xlog_find_tail(log, &head_blk, &tail_blk);
if (error) {
do_warn(
_("zero_log: cannot find log head/tail (xlog_find_tail=%d)\n"),
error);
if (!no_modify && !zap_log)
do_error(_(
"ERROR: The log head and/or tail cannot be discovered. Attempt to mount the\n"
"filesystem to replay the log or use the -L option to destroy the log and\n"
"attempt a repair.\n"));
} else {
if (verbose) {
do_warn(
_("zero_log: head block %" PRId64 " tail block %" PRId64 "\n"),
head_blk, tail_blk);
}
if (!no_modify && head_blk != tail_blk) {
if (zap_log) {
do_warn(_(
"ALERT: The filesystem has valuable metadata changes in a log which is being\n"
"destroyed because the -L option was used.\n"));
} else {
do_warn(_(
"ERROR: The filesystem has valuable metadata changes in a log which needs to\n"
"be replayed. Mount the filesystem to replay the log, and unmount it before\n"
"re-running xfs_repair. If you are unable to mount the filesystem, then use\n"
"the -L option to destroy the log and attempt a repair.\n"
"Note that destroying the log may cause corruption -- please attempt a mount\n"
"of the filesystem before doing this.\n"));
exit(2);
}
}
}
/*
* Only clear the log when explicitly requested. Doing so is unnecessary
* unless something is wrong. Further, this resets the current LSN of
* the filesystem and creates more work for repair of v5 superblock
* filesystems.
*/
if (!no_modify && zap_log) {
libxfs_log_clear(log->l_dev, NULL,
XFS_FSB_TO_DADDR(mp, mp->m_sb.sb_logstart),
(xfs_extlen_t)XFS_FSB_TO_BB(mp, mp->m_sb.sb_logblocks),
&mp->m_sb.sb_uuid,
xfs_sb_version_haslogv2(&mp->m_sb) ? 2 : 1,
mp->m_sb.sb_logsunit, XLOG_FMT, XLOG_INIT_CYCLE, true);
/* update the log data structure with new state */
error = xlog_find_tail(log, &head_blk, &tail_blk);
if (error || head_blk != tail_blk)
do_error(_("failed to clear log"));
}
/*
* Finally, seed the max LSN from the current state of the log if this
* is a v5 filesystem.
*/
if (xfs_sb_version_hascrc(&mp->m_sb))
libxfs_max_lsn = log->l_last_sync_lsn;
}
int
xlog_print_record(
struct xlog *log,
int fd,
int num_ops,
int len,
int *read_type,
char **partial_buf,
xlog_rec_header_t *rhead,
xlog_rec_ext_header_t *xhdrs,
int bad_hdr_warn)
{
char *buf, *ptr;
int read_len, skip, lost_context = 0;
int ret, n, i, j, k;
if (print_no_print)
return NO_ERROR;
if (!len) {
printf("\n");
return NO_ERROR;
}
/* read_len must read up to some block boundary */
read_len = (int) BBTOB(BTOBB(len));
/* read_type => don't malloc() new buffer, use old one */
if (*read_type == FULL_READ) {
if ((ptr = buf = malloc(read_len)) == NULL) {
fprintf(stderr, _("%s: xlog_print_record: malloc failed\n"), progname);
exit(1);
}
} else {
read_len -= *read_type;
buf = (char *)((intptr_t)(*partial_buf) + (intptr_t)(*read_type));
ptr = *partial_buf;
}
if ((ret = (int) read(fd, buf, read_len)) == -1) {
fprintf(stderr, _("%s: xlog_print_record: read error\n"), progname);
exit(1);
}
/* Did we overflow the end? */
if (*read_type == FULL_READ &&
BLOCK_LSN(be64_to_cpu(rhead->h_lsn)) + BTOBB(read_len) >=
logBBsize) {
*read_type = BBTOB(logBBsize - BLOCK_LSN(be64_to_cpu(rhead->h_lsn))-1);
*partial_buf = buf;
return PARTIAL_READ;
}
/* Did we read everything? */
if ((ret == 0 && read_len != 0) || ret != read_len) {
*read_type = ret;
*partial_buf = buf;
return PARTIAL_READ;
}
if (*read_type != FULL_READ)
read_len += *read_type;
/* Everything read in. Start from beginning of buffer
* Unpack the data, by putting the saved cycle-data back
* into the first word of each BB.
* Do some checks.
*/
buf = ptr;
for (i = 0; ptr < buf + read_len; ptr += BBSIZE, i++) {
xlog_rec_header_t *rechead = (xlog_rec_header_t *)ptr;
/* sanity checks */
if (be32_to_cpu(rechead->h_magicno) == XLOG_HEADER_MAGIC_NUM) {
/* data should not have magicno as first word
* as it should by cycle#
*/
free(buf);
return -1;
} else {
/* verify cycle#
* FIXME: cycle+1 should be a macro pv#900369
*/
if (be32_to_cpu(rhead->h_cycle) !=
be32_to_cpu(*(__be32 *)ptr)) {
if ((*read_type == FULL_READ) ||
(be32_to_cpu(rhead->h_cycle) + 1 !=
be32_to_cpu(*(__be32 *)ptr))) {
free(buf);
return -1;
}
}
}
/* copy back the data from the header */
if (i < XLOG_HEADER_CYCLE_SIZE / BBSIZE) {
/* from 1st header */
*(__be32 *)ptr = rhead->h_cycle_data[i];
}
else {
ASSERT(xhdrs != NULL);
/* from extra headers */
j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE);
*(__be32 *)ptr = xhdrs[j-1].xh_cycle_data[k];
}
}
ptr = buf;
for (i=0; i<num_ops; i++) {
int continued;
xlog_op_header_t *op_head = (xlog_op_header_t *)ptr;
print_xlog_op_line();
xlog_print_op_header(op_head, i, &ptr);
continued = ((op_head->oh_flags & XLOG_WAS_CONT_TRANS) ||
(op_head->oh_flags & XLOG_CONTINUE_TRANS));
if (continued && be32_to_cpu(op_head->oh_len) == 0)
continue;
if (print_no_data) {
for (n = 0; n < be32_to_cpu(op_head->oh_len); n++) {
printf("0x%02x ", (unsigned int)*ptr);
if (n % 16 == 15)
printf("\n");
ptr++;
}
printf("\n");
continue;
}
/* print transaction data */
if (xlog_print_find_tid(be32_to_cpu(op_head->oh_tid),
op_head->oh_flags & XLOG_WAS_CONT_TRANS)) {
printf(_("Left over region from split log item\n"));
/* Skip this leftover bit */
ptr += be32_to_cpu(op_head->oh_len);
/* We've lost context; don't complain if next one looks bad too */
lost_context = 1;
continue;
}
if (be32_to_cpu(op_head->oh_len) != 0) {
if (*(uint *)ptr == XFS_TRANS_HEADER_MAGIC) {
skip = xlog_print_trans_header(&ptr,
be32_to_cpu(op_head->oh_len));
} else {
switch (*(unsigned short *)ptr) {
case XFS_LI_BUF: {
skip = xlog_print_trans_buffer(&ptr,
be32_to_cpu(op_head->oh_len),
&i, num_ops);
break;
}
case XFS_LI_ICREATE: {
skip = xlog_print_trans_icreate(&ptr,
be32_to_cpu(op_head->oh_len),
&i, num_ops);
break;
}
case XFS_LI_INODE: {
skip = xlog_print_trans_inode(log, &ptr,
be32_to_cpu(op_head->oh_len),
&i, num_ops, continued);
break;
}
case XFS_LI_DQUOT: {
skip = xlog_print_trans_dquot(&ptr,
be32_to_cpu(op_head->oh_len),
&i, num_ops);
break;
}
case XFS_LI_EFI: {
skip = xlog_print_trans_efi(&ptr,
be32_to_cpu(op_head->oh_len),
continued);
break;
}
case XFS_LI_EFD: {
skip = xlog_print_trans_efd(&ptr,
be32_to_cpu(op_head->oh_len));
break;
}
case XFS_LI_QUOTAOFF: {
skip = xlog_print_trans_qoff(&ptr,
be32_to_cpu(op_head->oh_len));
break;
}
case XLOG_UNMOUNT_TYPE: {
printf(_("Unmount filesystem\n"));
skip = 0;
break;
}
default: {
if (bad_hdr_warn && !lost_context) {
fprintf(stderr,
_("%s: unknown log operation type (%x)\n"),
progname, *(unsigned short *)ptr);
if (print_exit) {
free(buf);
return BAD_HEADER;
}
} else {
printf(
_("Left over region from split log item\n"));
}
skip = 0;
ptr += be32_to_cpu(op_head->oh_len);
lost_context = 0;
}
} /* switch */
} /* else */
if (skip != 0)
xlog_print_add_to_trans(be32_to_cpu(op_head->oh_tid), skip);
}
}
printf("\n");
free(buf);
return NO_ERROR;
} /* xlog_print_record */
/*
* This code is gross and needs to be rewritten.
*/
void xfs_log_print(struct xlog *log,
int fd,
int print_block_start)
{
char hbuf[XLOG_HEADER_SIZE];
xlog_rec_header_t *hdr = (xlog_rec_header_t *)&hbuf[0];
xlog_rec_ext_header_t *xhdrs = NULL;
int num_ops, len, num_hdrs = 1;
xfs_daddr_t block_end = 0, block_start, blkno, error;
xfs_daddr_t zeroed_blkno = 0, cleared_blkno = 0;
int read_type = FULL_READ;
char *partial_buf;
int zeroed = 0;
int cleared = 0;
int first_hdr_found = 0;
logBBsize = log->l_logBBsize;
/*
* Normally, block_start and block_end are the same value since we
* are printing the entire log. However, if the start block is given,
* we still end at the end of the logical log.
*/
if ((error = xlog_print_find_oldest(log, &block_end))) {
fprintf(stderr, _("%s: problem finding oldest LR\n"), progname);
return;
}
if (print_block_start == -1)
block_start = block_end;
else
block_start = print_block_start;
xlog_print_lseek(log, fd, block_start, SEEK_SET);
blkno = block_start;
for (;;) {
if (read(fd, hbuf, 512) == 0) {
printf(_("%s: physical end of log\n"), progname);
print_xlog_record_line();
break;
}
if (print_only_data) {
printf(_("BLKNO: %lld\n"), (long long)blkno);
xlog_recover_print_data(hbuf, 512);
blkno++;
goto loop;
}
num_ops = xlog_print_rec_head(hdr, &len, first_hdr_found);
blkno++;
if (zeroed && num_ops != ZEROED_LOG) {
printf(_("%s: after %d zeroed blocks\n"), progname, zeroed);
/* once we find zeroed blocks - that's all we expect */
print_xlog_bad_zeroed(blkno-1);
/* reset count since we're assuming previous zeroed blocks
* were bad
*/
zeroed = 0;
}
if (num_ops == ZEROED_LOG ||
num_ops == CLEARED_BLKS ||
num_ops == BAD_HEADER) {
if (num_ops == ZEROED_LOG) {
if (zeroed == 0)
zeroed_blkno = blkno-1;
zeroed++;
}
else if (num_ops == CLEARED_BLKS) {
if (cleared == 0)
cleared_blkno = blkno-1;
cleared++;
} else {
if (!first_hdr_found)
block_start = blkno;
else
print_xlog_bad_header(blkno-1, hbuf);
}
goto loop;
}
if (be32_to_cpu(hdr->h_version) == 2) {
if (xlog_print_extended_headers(fd, len, &blkno, hdr, &num_hdrs, &xhdrs) != 0)
break;
}
error = xlog_print_record(log, fd, num_ops, len, &read_type, &partial_buf,
hdr, xhdrs, first_hdr_found);
first_hdr_found++;
switch (error) {
case 0: {
blkno += BTOBB(len);
if (print_block_start != -1 &&
blkno >= block_end) /* If start specified, we */
goto end; /* end early */
break;
}
case -1: {
print_xlog_bad_data(blkno-1);
if (print_block_start != -1 &&
blkno >= block_end) /* If start specified, */
goto end; /* we end early */
xlog_print_lseek(log, fd, blkno, SEEK_SET);
goto loop;
}
case PARTIAL_READ: {
print_xlog_record_line();
printf(_("%s: physical end of log\n"), progname);
print_xlog_record_line();
blkno = 0;
xlog_print_lseek(log, fd, 0, SEEK_SET);
/*
* We may have hit the end of the log when we started at 0.
* In this case, just end.
*/
if (block_start == 0)
goto end;
goto partial_log_read;
}
default: xlog_panic(_("illegal value"));
}
print_xlog_record_line();
loop:
if (blkno >= logBBsize) {
if (cleared) {
printf(_("%s: skipped %d cleared blocks in range: %lld - %lld\n"),
progname, cleared,
(long long)(cleared_blkno),
(long long)(cleared + cleared_blkno - 1));
if (cleared == logBBsize)
printf(_("%s: totally cleared log\n"), progname);
cleared=0;
}
if (zeroed) {
printf(_("%s: skipped %d zeroed blocks in range: %lld - %lld\n"),
progname, zeroed,
(long long)(zeroed_blkno),
(long long)(zeroed + zeroed_blkno - 1));
if (zeroed == logBBsize)
printf(_("%s: totally zeroed log\n"), progname);
zeroed=0;
}
printf(_("%s: physical end of log\n"), progname);
print_xlog_record_line();
break;
}
}
/* Do we need to print the first part of physical log? */
if (block_start != 0) {
blkno = 0;
xlog_print_lseek(log, fd, 0, SEEK_SET);
for (;;) {
if (read(fd, hbuf, 512) == 0) {
xlog_panic(_("xlog_find_head: bad read"));
}
if (print_only_data) {
printf(_("BLKNO: %lld\n"), (long long)blkno);
xlog_recover_print_data(hbuf, 512);
blkno++;
goto loop2;
}
num_ops = xlog_print_rec_head(hdr, &len, first_hdr_found);
blkno++;
if (num_ops == ZEROED_LOG ||
num_ops == CLEARED_BLKS ||
num_ops == BAD_HEADER) {
/* we only expect zeroed log entries or cleared log
* entries at the end of the _physical_ log,
* so treat them the same as bad blocks here
*/
print_xlog_bad_header(blkno-1, hbuf);
if (blkno >= block_end)
break;
continue;
}
if (be32_to_cpu(hdr->h_version) == 2) {
if (xlog_print_extended_headers(fd, len, &blkno, hdr, &num_hdrs, &xhdrs) != 0)
break;
}
partial_log_read:
error= xlog_print_record(log, fd, num_ops, len, &read_type,
&partial_buf, (xlog_rec_header_t *)hbuf,
xhdrs, first_hdr_found);
if (read_type != FULL_READ)
len -= read_type;
read_type = FULL_READ;
if (!error)
blkno += BTOBB(len);
else {
print_xlog_bad_data(blkno-1);
xlog_print_lseek(log, fd, blkno, SEEK_SET);
goto loop2;
}
print_xlog_record_line();
loop2:
if (blkno >= block_end)
break;
}
}
end:
printf(_("%s: logical end of log\n"), progname);
print_xlog_record_line();
}
/* for V2 logs read each extra hdr and print it out */
static int
xlog_print_extended_headers(
int fd,
int len,
xfs_daddr_t *blkno,
xlog_rec_header_t *hdr,
int *ret_num_hdrs,
xlog_rec_ext_header_t **ret_xhdrs)
{
int i, j;
int coverage_bb;
int num_hdrs;
int num_required;
char xhbuf[XLOG_HEADER_SIZE];
xlog_rec_ext_header_t *x;
num_required = howmany(len, XLOG_HEADER_CYCLE_SIZE);
num_hdrs = be32_to_cpu(hdr->h_size) / XLOG_HEADER_CYCLE_SIZE;
if (be32_to_cpu(hdr->h_size) % XLOG_HEADER_CYCLE_SIZE)
num_hdrs++;
if (num_required > num_hdrs) {
print_xlog_bad_reqd_hdrs((*blkno)-1, num_required, num_hdrs);
}
if (num_hdrs == 1) {
free(*ret_xhdrs);
*ret_xhdrs = NULL;
*ret_num_hdrs = 1;
return 0;
}
if (*ret_xhdrs == NULL || num_hdrs > *ret_num_hdrs) {
xlog_reallocate_xhdrs(num_hdrs, ret_xhdrs);
}
*ret_num_hdrs = num_hdrs;
/* don't include 1st header */
for (i = 1, x = *ret_xhdrs; i < num_hdrs; i++, (*blkno)++, x++) {
/* read one extra header blk */
if (read(fd, xhbuf, 512) == 0) {
printf(_("%s: physical end of log\n"), progname);
print_xlog_record_line();
/* reached the end so return 1 */
return 1;
}
if (print_only_data) {
printf(_("BLKNO: %lld\n"), (long long)*blkno);
xlog_recover_print_data(xhbuf, 512);
}
else {
if (i == num_hdrs - 1) {
/* last header */
coverage_bb = BTOBB(len) %
(XLOG_HEADER_CYCLE_SIZE / BBSIZE);
}
else {
/* earliear header */
coverage_bb = XLOG_HEADER_CYCLE_SIZE / BBSIZE;
}
xlog_print_rec_xhead((xlog_rec_ext_header_t*)xhbuf, coverage_bb);
}
/* Copy from buffer into xhdrs array for later.
* Could endian convert here but then code later on
* will look asymmetric with the 1 hdr normal case
* which does endian coversion on access.
*/
x->xh_cycle = ((xlog_rec_ext_header_t*)xhbuf)->xh_cycle;
for (j = 0; j < XLOG_HEADER_CYCLE_SIZE / BBSIZE; j++) {
x->xh_cycle_data[j] =
((xlog_rec_ext_header_t*)xhbuf)->xh_cycle_data[j];
}
}
return 0;
}
extern int
HsmInitFileContext(
hsm_f_ctxt_t *contextp,
const xfs_bstat_t *statp)
{
dmf_f_ctxt_t *dmf_f_ctxtp = (dmf_f_ctxt_t *)contextp;
XFSattrvalue1_t *dmfattrp;
int attrlen;
void *hanp;
size_t hlen;
dm_ino_t ino;
dm_igen_t igen;
int fd;
int error;
dmf_f_ctxtp->candidate = 0; /* assume file will NOT be of interest */
/* Try and rule out a dualstate inode by doing some quick tests. */
if ((statp->bs_mode & S_IFMT) != S_IFREG) {
return 0; /* not a regular file */
}
if ((statp->bs_xflags & XFS_XFLAG_HASATTR) == 0) {
return 0; /* no DMF attribute exists */
}
if ((statp->bs_dmevmask & DUALSTATE_BITS) != DUALSTATE_BITS) {
return 0; /* non-dualstate managed region bits */
}
/* We have a likely candidate, so we have to pay the price and look
for the DMF attribute. (It could be in a disk block separate from
the inode.)
*/
ino = (dm_ino_t)statp->bs_ino;
igen = (dm_igen_t)statp->bs_gen;
if (dm_make_handle(&dmf_f_ctxtp->fsys.fsid, &ino, &igen, &hanp, &hlen) != 0) {
return 0; /* can't make a proper handle */
}
/* The following code should eventually be replaced with the
attr_multif-by-handle call when it is available.
*/
fd = open_by_handle(hanp, hlen, O_RDONLY);
dm_handle_free(hanp, hlen);
if (fd < 0) {
return 0;
}
attrlen = sizeof(dmf_f_ctxtp->attrval);
error = attr_getf(fd, DMF_ATTR_NAME, dmf_f_ctxtp->attrval,
&attrlen, ATTR_ROOT);
(void)close(fd);
if (error) {
return 0;
}
if (attrlen != DMF_ATTR_LEN) {
return 0; /* not the right length to be the attribute */
}
dmfattrp = (XFSattrvalue1_t *)dmf_f_ctxtp->attrval;
if (dmfattrp->version != XFS_ATTR_VERSION_1) {
return 0; /* we don't support this version */
}
if (dmfattrp->state[0] != '\0') {
return 0; /* should be zero */
}
if (dmfattrp->state[1] != DMF_ST_DUALSTATE &&
dmfattrp->state[1] != DMF_ST_UNMIGRATING) {
return 0;
}
/* We have a DMF dual state file. */
dmf_f_ctxtp->candidate = 1;
dmf_f_ctxtp->filesize = BTOBB(statp->bs_size);
return 0;
}
/*
* Zero file bytes between startoff and endoff inclusive.
* The iolock is held exclusive and no blocks are buffered.
*
* This function is used by xfs_free_file_space() to zero
* partial blocks when the range to free is not block aligned.
* When unreserving space with boundaries that are not block
* aligned we round up the start and round down the end
* boundaries and then use this function to zero the parts of
* the blocks that got dropped during the rounding.
*/
STATIC int
xfs_zero_remaining_bytes(
xfs_inode_t *ip,
xfs_off_t startoff,
xfs_off_t endoff)
{
xfs_bmbt_irec_t imap;
xfs_fileoff_t offset_fsb;
xfs_off_t lastoffset;
xfs_off_t offset;
xfs_buf_t *bp;
xfs_mount_t *mp = ip->i_mount;
int nimap;
int error = 0;
/*
* Avoid doing I/O beyond eof - it's not necessary
* since nothing can read beyond eof. The space will
* be zeroed when the file is extended anyway.
*/
if (startoff >= XFS_ISIZE(ip))
return 0;
if (endoff > XFS_ISIZE(ip))
endoff = XFS_ISIZE(ip);
bp = xfs_buf_get_uncached(XFS_IS_REALTIME_INODE(ip) ?
mp->m_rtdev_targp : mp->m_ddev_targp,
BTOBB(mp->m_sb.sb_blocksize), 0);
if (!bp)
return XFS_ERROR(ENOMEM);
xfs_buf_unlock(bp);
for (offset = startoff; offset <= endoff; offset = lastoffset + 1) {
uint lock_mode;
offset_fsb = XFS_B_TO_FSBT(mp, offset);
nimap = 1;
lock_mode = xfs_ilock_data_map_shared(ip);
error = xfs_bmapi_read(ip, offset_fsb, 1, &imap, &nimap, 0);
xfs_iunlock(ip, lock_mode);
if (error || nimap < 1)
break;
ASSERT(imap.br_blockcount >= 1);
ASSERT(imap.br_startoff == offset_fsb);
lastoffset = XFS_FSB_TO_B(mp, imap.br_startoff + 1) - 1;
if (lastoffset > endoff)
lastoffset = endoff;
if (imap.br_startblock == HOLESTARTBLOCK)
continue;
ASSERT(imap.br_startblock != DELAYSTARTBLOCK);
if (imap.br_state == XFS_EXT_UNWRITTEN)
continue;
XFS_BUF_UNDONE(bp);
XFS_BUF_UNWRITE(bp);
XFS_BUF_READ(bp);
XFS_BUF_SET_ADDR(bp, xfs_fsb_to_db(ip, imap.br_startblock));
if (XFS_FORCED_SHUTDOWN(mp)) {
error = XFS_ERROR(EIO);
break;
}
xfs_buf_iorequest(bp);
error = xfs_buf_iowait(bp);
if (error) {
xfs_buf_ioerror_alert(bp,
"xfs_zero_remaining_bytes(read)");
break;
}
memset(bp->b_addr +
(offset - XFS_FSB_TO_B(mp, imap.br_startoff)),
0, lastoffset - offset + 1);
XFS_BUF_UNDONE(bp);
XFS_BUF_UNREAD(bp);
XFS_BUF_WRITE(bp);
if (XFS_FORCED_SHUTDOWN(mp)) {
error = XFS_ERROR(EIO);
break;
}
xfs_buf_iorequest(bp);
error = xfs_buf_iowait(bp);
if (error) {
xfs_buf_ioerror_alert(bp,
"xfs_zero_remaining_bytes(write)");
break;
}
}
xfs_buf_free(bp);
return error;
}
/* copy out attributes from cache entry */
int
afs_CopyOutAttrs(struct vcache *avc, struct vattr *attrs)
{
struct volume *tvp;
struct cell *tcell;
#if defined(AFS_FBSD_ENV) || defined(AFS_DFBSD_ENV)
struct vnode *vp = AFSTOV(avc);
#endif
int fakedir = 0;
AFS_STATCNT(afs_CopyOutAttrs);
if (afs_fakestat_enable && avc->mvstat == AFS_MVSTAT_MTPT)
fakedir = 1;
attrs->va_type = fakedir ? VDIR : vType(avc);
#if defined(AFS_SGI_ENV) || defined(AFS_AIX32_ENV) || defined(AFS_SUN5_ENV) || defined(AFS_DARWIN_ENV)
attrs->va_mode = fakedir ? S_IFDIR | 0755 : (mode_t) (avc->f.m.Mode & 0xffff);
#else
attrs->va_mode = fakedir ? VDIR | 0755 : avc->f.m.Mode;
#endif
if (avc->f.m.Mode & (VSUID | VSGID)) {
/* setuid or setgid, make sure we're allowed to run them from this cell */
tcell = afs_GetCell(avc->f.fid.Cell, 0);
if (tcell && (tcell->states & CNoSUID))
attrs->va_mode &= ~(VSUID | VSGID);
}
#if defined(AFS_DARWIN_ENV)
{
if (!afs_darwin_realmodes) {
/* Mac OS X uses the mode bits to determine whether a file or
* directory is accessible, and believes them, even though under
* AFS they're almost assuredly wrong, especially if the local uid
* does not match the AFS ID. So we set the mode bits
* conservatively.
*/
if (S_ISDIR(attrs->va_mode)) {
/* all access bits need to be set for directories, since even
* a mode 0 directory can still be used normally.
*/
attrs->va_mode |= ACCESSPERMS;
} else {
/* for other files, replicate the user bits to group and other */
mode_t ubits = (attrs->va_mode & S_IRWXU) >> 6;
attrs->va_mode |= ubits | (ubits << 3);
}
}
}
#endif /* AFS_DARWIN_ENV */
attrs->va_uid = fakedir ? 0 : avc->f.m.Owner;
attrs->va_gid = fakedir ? 0 : avc->f.m.Group; /* yeah! */
#if defined(AFS_SUN5_ENV)
attrs->va_fsid = avc->v.v_vfsp->vfs_fsid.val[0];
#elif defined(AFS_DARWIN80_ENV)
VATTR_RETURN(attrs, va_fsid, vfs_statfs(vnode_mount(AFSTOV(avc)))->f_fsid.val[0]);
#elif defined(AFS_DARWIN_ENV)
attrs->va_fsid = avc->v->v_mount->mnt_stat.f_fsid.val[0];
#else /* ! AFS_DARWIN_ENV */
attrs->va_fsid = 1;
#endif
if (avc->mvstat == AFS_MVSTAT_ROOT) {
tvp = afs_GetVolume(&avc->f.fid, 0, READ_LOCK);
/* The mount point's vnode. */
if (tvp) {
attrs->va_nodeid =
afs_calc_inum(tvp->mtpoint.Cell,
tvp->mtpoint.Fid.Volume,
tvp->mtpoint.Fid.Vnode);
if (FidCmp(&afs_rootFid, &avc->f.fid) && !attrs->va_nodeid)
attrs->va_nodeid = 2;
afs_PutVolume(tvp, READ_LOCK);
} else
attrs->va_nodeid = 2;
} else
attrs->va_nodeid =
afs_calc_inum(avc->f.fid.Cell,
avc->f.fid.Fid.Volume,
avc->f.fid.Fid.Vnode);
attrs->va_nodeid &= 0x7fffffff; /* Saber C hates negative inode #s! */
attrs->va_nlink = fakedir ? 100 : avc->f.m.LinkCount;
attrs->va_size = fakedir ? 4096 : avc->f.m.Length;
#if defined(AFS_FBSD_ENV) || defined(AFS_DFBSD_ENV)
vnode_pager_setsize(vp, (u_long) attrs->va_size);
#endif
attrs->va_atime.tv_sec = attrs->va_mtime.tv_sec = attrs->va_ctime.tv_sec =
fakedir ? 0 : (int)avc->f.m.Date;
/* set microseconds to be dataversion # so that we approximate NFS-style
* use of mtime as a dataversion #. We take it mod 512K because
* microseconds *must* be less than a million, and 512K is the biggest
* power of 2 less than such. DataVersions are typically pretty small
* anyway, so the difference between 512K and 1000000 shouldn't matter
* much, and "&" is a lot faster than "%".
*/
#if defined(AFS_DARWIN_ENV) || defined(AFS_FBSD_ENV)
/* nfs on these systems puts an 0 in nsec and stores the nfs usec (aka
* dataversion) in va_gen */
attrs->va_atime.tv_nsec = attrs->va_mtime.tv_nsec =
attrs->va_ctime.tv_nsec = 0;
attrs->va_gen = hgetlo(avc->f.m.DataVersion);
#elif defined(AFS_SGI_ENV) || defined(AFS_SUN5_ENV) || defined(AFS_AIX41_ENV) || defined(AFS_OBSD_ENV) || defined(AFS_NBSD_ENV)
attrs->va_atime.tv_nsec = attrs->va_mtime.tv_nsec =
attrs->va_ctime.tv_nsec =
(hgetlo(avc->f.m.DataVersion) & 0x7ffff) * 1000;
#else
attrs->va_atime.tv_usec = attrs->va_mtime.tv_usec =
attrs->va_ctime.tv_usec = (hgetlo(avc->f.m.DataVersion) & 0x7ffff);
#endif
#if defined(AFS_DARWIN_ENV) || defined(AFS_XBSD_ENV)
attrs->va_flags = 0;
#endif
#if defined(AFS_SGI_ENV) || defined(AFS_SUN5_ENV)
attrs->va_blksize = AFS_BLKSIZE; /* XXX Was 8192 XXX */
#else
attrs->va_blocksize = AFS_BLKSIZE; /* XXX Was 8192 XXX */
#endif
attrs->va_rdev = 1;
#if defined(AFS_HPUX110_ENV)
if (afs_globalVFS)
attrs->va_fstype = afs_globalVFS->vfs_mtype;
#endif
/*
* Below return 0 (and not 1) blocks if the file is zero length. This conforms
* better with the other filesystems that do return 0.
*/
#if defined(AFS_DARWIN_ENV) || defined(AFS_XBSD_ENV)
attrs->va_bytes = (attrs->va_size ? (attrs->va_size + 1023) : 1024);
#ifdef va_bytes_rsv
attrs->va_bytes_rsv = -1;
#endif
#elif defined(AFS_HPUX_ENV)
attrs->va_blocks = (attrs->va_size ? ((attrs->va_size + 1023)>>10) : 0);
#elif defined(AFS_SGI_ENV)
attrs->va_blocks = BTOBB(attrs->va_size);
#elif defined(AFS_SUN5_ENV)
attrs->va_nblocks = (attrs->va_size ? ((attrs->va_size + 1023)>>10)<<1:0);
#else /* everything else */
attrs->va_blocks = (attrs->va_size ? ((attrs->va_size + 1023)>>10)<<1:0);
#endif
return 0;
}
int
xlog_print_rec_head(xlog_rec_header_t *head, int *len, int bad_hdr_warn)
{
int i;
char uub[64];
int datalen,bbs;
if (print_no_print)
return be32_to_cpu(head->h_num_logops);
if (!head->h_magicno)
return ZEROED_LOG;
if (be32_to_cpu(head->h_magicno) != XLOG_HEADER_MAGIC_NUM) {
if (bad_hdr_warn)
printf(_("Header 0x%x wanted 0x%x\n"),
be32_to_cpu(head->h_magicno),
XLOG_HEADER_MAGIC_NUM);
return BAD_HEADER;
}
/* check for cleared blocks written by xlog_clear_stale_blocks() */
if (!head->h_len && !head->h_crc && !head->h_prev_block &&
!head->h_num_logops && !head->h_size)
return CLEARED_BLKS;
datalen=be32_to_cpu(head->h_len);
bbs=BTOBB(datalen);
printf(_("cycle: %d version: %d "),
be32_to_cpu(head->h_cycle),
be32_to_cpu(head->h_version));
print_lsn(" lsn", &head->h_lsn);
print_lsn(" tail_lsn", &head->h_tail_lsn);
printf("\n");
printf(_("length of Log Record: %d prev offset: %d num ops: %d\n"),
datalen,
be32_to_cpu(head->h_prev_block),
be32_to_cpu(head->h_num_logops));
if (print_overwrite) {
printf(_("cycle num overwrites: "));
for (i=0; i< MIN(bbs, XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++)
printf("%d - 0x%x ",
i,
be32_to_cpu(head->h_cycle_data[i]));
printf("\n");
}
platform_uuid_unparse(&head->h_fs_uuid, uub);
printf(_("uuid: %s format: "), uub);
switch (be32_to_cpu(head->h_fmt)) {
case XLOG_FMT_UNKNOWN:
printf(_("unknown\n"));
break;
case XLOG_FMT_LINUX_LE:
printf(_("little endian linux\n"));
break;
case XLOG_FMT_LINUX_BE:
printf(_("big endian linux\n"));
break;
case XLOG_FMT_IRIX_BE:
printf(_("big endian irix\n"));
break;
default:
printf("? (%d)\n", be32_to_cpu(head->h_fmt));
break;
}
printf(_("h_size: %d\n"), be32_to_cpu(head->h_size));
*len = be32_to_cpu(head->h_len);
return(be32_to_cpu(head->h_num_logops));
} /* xlog_print_rec_head */
/*
* The data and realtime block counts returned (count, used, and
* free) are all in basic block units.
*/
static int
mount_free_space_data(
struct fs_path *mount,
__uint64_t *bcount,
__uint64_t *bused,
__uint64_t *bfree,
__uint64_t *icount,
__uint64_t *iused,
__uint64_t *ifree,
__uint64_t *rcount,
__uint64_t *rused,
__uint64_t *rfree)
{
struct xfs_fsop_counts fscounts;
struct xfs_fsop_geom fsgeo;
struct statfs st;
__uint64_t logsize, count, free;
int fd;
if ((fd = open(mount->fs_dir, O_RDONLY)) < 0) {
exitcode = 1;
fprintf(stderr, "%s: cannot open %s: %s\n",
progname, mount->fs_dir, strerror(errno));
return 0;
}
if (platform_fstatfs(fd, &st) < 0) {
perror("fstatfs");
close(fd);
return 0;
}
if ((xfsctl(mount->fs_dir, fd, XFS_IOC_FSGEOMETRY_V1, &fsgeo)) < 0) {
perror("XFS_IOC_FSGEOMETRY_V1");
close(fd);
return 0;
}
if ((xfsctl(mount->fs_dir, fd, XFS_IOC_FSCOUNTS, &fscounts)) < 0) {
perror("XFS_IOC_FSCOUNTS");
close(fd);
return 0;
}
logsize = fsgeo.logstart ? fsgeo.logblocks : 0;
count = (fsgeo.datablocks - logsize) * fsgeo.blocksize;
free = fscounts.freedata * fsgeo.blocksize;
*bcount = BTOBB(count);
*bfree = BTOBB(free);
*bused = BTOBB(count - free);
*icount = st.f_files;
*ifree = st.f_ffree;
*iused = st.f_files - st.f_ffree;
count = fsgeo.rtextents * fsgeo.rtextsize * fsgeo.blocksize;
free = fscounts.freertx * fsgeo.rtextsize * fsgeo.blocksize;
*rcount = BTOBB(count);
*rfree = BTOBB(free);
*rused = BTOBB(count - free);
close(fd);
return 1;
}
