/*
* This routine is called to handle zeroing any space in the last block of the
* file that is beyond the EOF. We do this since the size is being increased
* without writing anything to that block and we don't want to read the
* garbage on the disk.
*/
STATIC int /* error (positive) */
xfs_zero_last_block(
struct xfs_inode *ip,
xfs_fsize_t offset,
xfs_fsize_t isize)
{
struct xfs_mount *mp = ip->i_mount;
xfs_fileoff_t last_fsb = XFS_B_TO_FSBT(mp, isize);
int zero_offset = XFS_B_FSB_OFFSET(mp, isize);
int zero_len;
int nimaps = 1;
int error = 0;
struct xfs_bmbt_irec imap;
xfs_ilock(ip, XFS_ILOCK_EXCL);
error = xfs_bmapi_read(ip, last_fsb, 1, &imap, &nimaps, 0);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
if (error)
return error;
ASSERT(nimaps > 0);
/*
* If the block underlying isize is just a hole, then there
* is nothing to zero.
*/
if (imap.br_startblock == HOLESTARTBLOCK)
return 0;
zero_len = mp->m_sb.sb_blocksize - zero_offset;
if (isize + zero_len > offset)
zero_len = offset - isize;
return xfs_iozero(ip, isize, zero_len);
}
/*
* This routine is called to handle zeroing any space in the last
* block of the file that is beyond the EOF. We do this since the
* size is being increased without writing anything to that block
* and we don't want anyone to read the garbage on the disk.
*/
STATIC int /* error (positive) */
xfs_zero_last_block(
xfs_vnode_t *vp,
xfs_iocore_t *io,
xfs_fsize_t isize,
xfs_fsize_t end_size)
{
xfs_fileoff_t last_fsb;
xfs_mount_t *mp;
int nimaps;
int zero_offset;
int zero_len;
int error = 0;
xfs_bmbt_irec_t imap;
xfs_off_t loff;
ASSERT(ismrlocked(io->io_lock, MR_UPDATE) != 0);
mp = io->io_mount;
zero_offset = XFS_B_FSB_OFFSET(mp, isize);
if (zero_offset == 0) {
/*
* There are no extra bytes in the last block on disk to
* zero, so return.
*/
return 0;
}
last_fsb = XFS_B_TO_FSBT(mp, isize);
nimaps = 1;
error = XFS_BMAPI(mp, NULL, io, last_fsb, 1, 0, NULL, 0, &imap,
&nimaps, NULL, NULL);
if (error) {
return error;
}
ASSERT(nimaps > 0);
/*
* If the block underlying isize is just a hole, then there
* is nothing to zero.
*/
if (imap.br_startblock == HOLESTARTBLOCK) {
return 0;
}
/*
* Zero the part of the last block beyond the EOF, and write it
* out sync. We need to drop the ilock while we do this so we
* don't deadlock when the buffer cache calls back to us.
*/
XFS_IUNLOCK(mp, io, XFS_ILOCK_EXCL| XFS_EXTSIZE_RD);
loff = XFS_FSB_TO_B(mp, last_fsb);
zero_len = mp->m_sb.sb_blocksize - zero_offset;
error = xfs_iozero(vp, loff + zero_offset, zero_len, end_size);
XFS_ILOCK(mp, io, XFS_ILOCK_EXCL|XFS_EXTSIZE_RD);
ASSERT(error >= 0);
return error;
}
/*
* This routine is called to handle zeroing any space in the last
* block of the file that is beyond the EOF. We do this since the
* size is being increased without writing anything to that block
* and we don't want anyone to read the garbage on the disk.
*/
STATIC int /* error (positive) */
xfs_zero_last_block(
xfs_inode_t *ip,
xfs_fsize_t offset,
xfs_fsize_t isize)
{
xfs_fileoff_t last_fsb;
xfs_mount_t *mp = ip->i_mount;
int nimaps;
int zero_offset;
int zero_len;
int error = 0;
xfs_bmbt_irec_t imap;
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
zero_offset = XFS_B_FSB_OFFSET(mp, isize);
if (zero_offset == 0) {
/*
* There are no extra bytes in the last block on disk to
* zero, so return.
*/
return 0;
}
last_fsb = XFS_B_TO_FSBT(mp, isize);
nimaps = 1;
error = xfs_bmapi(NULL, ip, last_fsb, 1, 0, NULL, 0, &imap,
&nimaps, NULL, NULL);
if (error) {
return error;
}
ASSERT(nimaps > 0);
/*
* If the block underlying isize is just a hole, then there
* is nothing to zero.
*/
if (imap.br_startblock == HOLESTARTBLOCK) {
return 0;
}
/*
* Zero the part of the last block beyond the EOF, and write it
* out sync. We need to drop the ilock while we do this so we
* don't deadlock when the buffer cache calls back to us.
*/
xfs_iunlock(ip, XFS_ILOCK_EXCL);
zero_len = mp->m_sb.sb_blocksize - zero_offset;
if (isize + zero_len > offset)
zero_len = offset - isize;
error = xfs_iozero(ip, isize, zero_len);
xfs_ilock(ip, XFS_ILOCK_EXCL);
ASSERT(error >= 0);
return error;
}
STATIC int
xfs_zero_last_block(
xfs_inode_t *ip,
xfs_fsize_t offset,
xfs_fsize_t isize)
{
xfs_fileoff_t last_fsb;
xfs_mount_t *mp = ip->i_mount;
int nimaps;
int zero_offset;
int zero_len;
int error = 0;
xfs_bmbt_irec_t imap;
ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL));
zero_offset = XFS_B_FSB_OFFSET(mp, isize);
if (zero_offset == 0) {
return 0;
}
last_fsb = XFS_B_TO_FSBT(mp, isize);
nimaps = 1;
error = xfs_bmapi_read(ip, last_fsb, 1, &imap, &nimaps, 0);
if (error)
return error;
ASSERT(nimaps > 0);
if (imap.br_startblock == HOLESTARTBLOCK) {
return 0;
}
xfs_iunlock(ip, XFS_ILOCK_EXCL);
zero_len = mp->m_sb.sb_blocksize - zero_offset;
if (isize + zero_len > offset)
zero_len = offset - isize;
error = xfs_iozero(ip, isize, zero_len);
xfs_ilock(ip, XFS_ILOCK_EXCL);
ASSERT(error >= 0);
return error;
}
/*
* Zero any on disk space between the current EOF and the new, larger EOF.
*
* This handles the normal case of zeroing the remainder of the last block in
* the file and the unusual case of zeroing blocks out beyond the size of the
* file. This second case only happens with fixed size extents and when the
* system crashes before the inode size was updated but after blocks were
* allocated.
*
* Expects the iolock to be held exclusive, and will take the ilock internally.
*/
int /* error (positive) */
xfs_zero_eof(
struct xfs_inode *ip,
xfs_off_t offset, /* starting I/O offset */
xfs_fsize_t isize) /* current inode size */
{
struct xfs_mount *mp = ip->i_mount;
xfs_fileoff_t start_zero_fsb;
xfs_fileoff_t end_zero_fsb;
xfs_fileoff_t zero_count_fsb;
xfs_fileoff_t last_fsb;
xfs_fileoff_t zero_off;
xfs_fsize_t zero_len;
int nimaps;
int error = 0;
struct xfs_bmbt_irec imap;
ASSERT(xfs_isilocked(ip, XFS_IOLOCK_EXCL));
ASSERT(offset > isize);
/*
* First handle zeroing the block on which isize resides.
*
* We only zero a part of that block so it is handled specially.
*/
if (XFS_B_FSB_OFFSET(mp, isize) != 0) {
error = xfs_zero_last_block(ip, offset, isize);
if (error)
return error;
}
/*
* Calculate the range between the new size and the old where blocks
* needing to be zeroed may exist.
*
* To get the block where the last byte in the file currently resides,
* we need to subtract one from the size and truncate back to a block
* boundary. We subtract 1 in case the size is exactly on a block
* boundary.
*/
last_fsb = isize ? XFS_B_TO_FSBT(mp, isize - 1) : (xfs_fileoff_t)-1;
start_zero_fsb = XFS_B_TO_FSB(mp, (xfs_ufsize_t)isize);
end_zero_fsb = XFS_B_TO_FSBT(mp, offset - 1);
ASSERT((xfs_sfiloff_t)last_fsb < (xfs_sfiloff_t)start_zero_fsb);
if (last_fsb == end_zero_fsb) {
/*
* The size was only incremented on its last block.
* We took care of that above, so just return.
*/
return 0;
}
ASSERT(start_zero_fsb <= end_zero_fsb);
while (start_zero_fsb <= end_zero_fsb) {
nimaps = 1;
zero_count_fsb = end_zero_fsb - start_zero_fsb + 1;
xfs_ilock(ip, XFS_ILOCK_EXCL);
error = xfs_bmapi_read(ip, start_zero_fsb, zero_count_fsb,
&imap, &nimaps, 0);
xfs_iunlock(ip, XFS_ILOCK_EXCL);
if (error)
return error;
ASSERT(nimaps > 0);
if (imap.br_state == XFS_EXT_UNWRITTEN ||
imap.br_startblock == HOLESTARTBLOCK) {
start_zero_fsb = imap.br_startoff + imap.br_blockcount;
ASSERT(start_zero_fsb <= (end_zero_fsb + 1));
continue;
}
/*
* There are blocks we need to zero.
*/
zero_off = XFS_FSB_TO_B(mp, start_zero_fsb);
zero_len = XFS_FSB_TO_B(mp, imap.br_blockcount);
if ((zero_off + zero_len) > offset)
zero_len = offset - zero_off;
error = xfs_iozero(ip, zero_off, zero_len);
if (error)
return error;
start_zero_fsb = imap.br_startoff + imap.br_blockcount;
ASSERT(start_zero_fsb <= (end_zero_fsb + 1));
}
return 0;
}
int
xfs_write_file(xfs_inode_t *xip, struct uio *uio, int ioflag)
{
struct buf *bp;
//struct thread *td;
daddr_t lbn;
off_t osize = 0;
off_t offset= 0;
int blkoffset, error, resid, xfersize;
int fsblocksize;
int seqcount;
xfs_iomap_t iomap;
int maps = 1;
xfs_vnode_t *xvp = XFS_ITOV(xip);
struct vnode *vp = xvp->v_vnode;
xfs_mount_t *mp = (&xip->i_iocore)->io_mount;
seqcount = ioflag >> IO_SEQSHIFT;
memset(&iomap,0,sizeof(xfs_iomap_t));
/*
* Maybe this should be above the vnode op call, but so long as
* file servers have no limits, I don't think it matters.
*/
#if 0
td = uio->uio_td;
if (vp->v_type == VREG && td != NULL) {
PROC_LOCK(td->td_proc);
if (uio->uio_offset + uio->uio_resid >
lim_cur(td->td_proc, RLIMIT_FSIZE)) {
psignal(td->td_proc, SIGXFSZ);
PROC_UNLOCK(td->td_proc);
return (EFBIG);
}
PROC_UNLOCK(td->td_proc);
}
#endif
resid = uio->uio_resid;
offset = uio->uio_offset;
osize = xip->i_d.di_size;
/* xfs bmap wants bytes for both offset and size */
XVOP_BMAP(xvp,
uio->uio_offset,
uio->uio_resid,
BMAPI_WRITE|BMAPI_DIRECT,
&iomap, &maps, error);
if(error) {
printf("XVOP_BMAP failed\n");
goto error;
}
for (error = 0; uio->uio_resid > 0;) {
lbn = XFS_B_TO_FSBT(mp, offset);
blkoffset = XFS_B_FSB_OFFSET(mp, offset);
xfersize = mp->m_sb.sb_blocksize - blkoffset;
fsblocksize = mp->m_sb.sb_blocksize;
if (uio->uio_resid < xfersize)
xfersize = uio->uio_resid;
/*
* getblk sets buf by blkno * bo->bo_bsize
* bo_bsize is set from the mnt point fsize
* so we call getblk in the case using fsblocks
* not basic blocks
*/
bp = getblk(vp, lbn, fsblocksize, 0, 0, 0);
if(!bp) {
printf("getblk failed\n");
error = EINVAL;
break;
}
if (!(bp->b_flags & B_CACHE) && fsblocksize > xfersize)
vfs_bio_clrbuf(bp);
if (offset + xfersize > xip->i_d.di_size) {
xip->i_d.di_size = offset + xfersize;
vnode_pager_setsize(vp, offset + fsblocksize);
}
/* move the offset for the next itteration of the loop */
offset += xfersize;
error = uiomove((char *)bp->b_data + blkoffset, xfersize, uio);
if ((ioflag & IO_VMIO) &&
(LIST_FIRST(&bp->b_dep) == NULL)) /* in ext2fs? */
bp->b_flags |= B_RELBUF;
/* force to full direct for now */
bp->b_flags |= B_DIRECT;
/* and sync ... the delay path is not pushing data out */
ioflag |= IO_SYNC;
if (ioflag & IO_SYNC) {
(void)bwrite(bp);
} else if (0 /* RMC xfersize + blkoffset == fs->s_frag_size */) {
if ((vp->v_mount->mnt_flag & MNT_NOCLUSTERW) == 0) {
bp->b_flags |= B_CLUSTEROK;
cluster_write(vp, bp, osize, seqcount);
} else {
bawrite(bp);
}
} else {
bp->b_flags |= B_CLUSTEROK;
bdwrite(bp);
}
if (error || xfersize == 0)
break;
}
/*
* If we successfully wrote any data, and we are not the superuser
* we clear the setuid and setgid bits as a precaution against
* tampering.
*/
#if 0
if (resid > uio->uio_resid && ap->a_cred && ap->a_cred->cr_uid != 0)
ip->i_mode &= ~(ISUID | ISGID);
#endif
if (error) {
if (ioflag & IO_UNIT) {
#if 0
(void)ext2_truncate(vp, osize,
ioflag & IO_SYNC, ap->a_cred, uio->uio_td);
#endif
uio->uio_offset -= resid - uio->uio_resid;
uio->uio_resid = resid;
}
} else if (resid > uio->uio_resid && (ioflag & IO_SYNC)) {
/* Update the vnode here? */
}
error:
return error;
}
int
xfs_read_file(xfs_mount_t *mp, xfs_inode_t *ip, struct uio *uio, int ioflag)
{
xfs_fileoff_t lbn, nextlbn;
xfs_fsize_t bytesinfile;
long size, xfersize, blkoffset;
struct buf *bp;
struct vnode *vp;
int error, orig_resid;
int seqcount;
seqcount = ioflag >> IO_SEQSHIFT;
orig_resid = uio->uio_resid;
if (orig_resid <= 0)
return (0);
vp = XFS_ITOV(ip)->v_vnode;
/*
* Ok so we couldn't do it all in one vm trick...
* so cycle around trying smaller bites..
*/
for (error = 0, bp = NULL; uio->uio_resid > 0; bp = NULL) {
if ((bytesinfile = ip->i_d.di_size - uio->uio_offset) <= 0)
break;
lbn = XFS_B_TO_FSBT(mp, uio->uio_offset);
nextlbn = lbn + 1;
/*
* size of buffer. The buffer representing the
* end of the file is rounded up to the size of
* the block type ( fragment or full block,
* depending ).
*/
size = mp->m_sb.sb_blocksize;
blkoffset = XFS_B_FSB_OFFSET(mp, uio->uio_offset);
/*
* The amount we want to transfer in this iteration is
* one FS block less the amount of the data before
* our startpoint (duh!)
*/
xfersize = mp->m_sb.sb_blocksize - blkoffset;
/*
* But if we actually want less than the block,
* or the file doesn't have a whole block more of data,
* then use the lesser number.
*/
if (uio->uio_resid < xfersize)
xfersize = uio->uio_resid;
if (bytesinfile < xfersize)
xfersize = bytesinfile;
if (XFS_FSB_TO_B(mp, nextlbn) >= ip->i_d.di_size ) {
/*
* Don't do readahead if this is the end of the file.
*/
error = bread(vp, lbn, size, NOCRED, &bp);
} else if ((vp->v_mount->mnt_flag & MNT_NOCLUSTERR) == 0) {
/*
* Otherwise if we are allowed to cluster,
* grab as much as we can.
*
* XXX This may not be a win if we are not
* doing sequential access.
*/
error = cluster_read(vp, ip->i_d.di_size, lbn,
size, NOCRED, uio->uio_resid, seqcount, &bp);
} else if (seqcount > 1) {
/*
* If we are NOT allowed to cluster, then
* if we appear to be acting sequentially,
* fire off a request for a readahead
* as well as a read. Note that the 4th and 5th
* arguments point to arrays of the size specified in
* the 6th argument.
*/
int nextsize = mp->m_sb.sb_blocksize;
error = breadn(vp, lbn,
size, &nextlbn, &nextsize, 1, NOCRED, &bp);
} else {
/*
* Failing all of the above, just read what the
* user asked for. Interestingly, the same as
* the first option above.
*/
error = bread(vp, lbn, size, NOCRED, &bp);
}
if (error) {
brelse(bp);
bp = NULL;
break;
}
/*
* If IO_DIRECT then set B_DIRECT for the buffer. This
* will cause us to attempt to release the buffer later on
* and will cause the buffer cache to attempt to free the
* underlying pages.
*/
if (ioflag & IO_DIRECT)
bp->b_flags |= B_DIRECT;
/*
* We should only get non-zero b_resid when an I/O error
* has occurred, which should cause us to break above.
* However, if the short read did not cause an error,
* then we want to ensure that we do not uiomove bad
* or uninitialized data.
*/
size -= bp->b_resid;
if (size < xfersize) {
if (size == 0)
break;
xfersize = size;
}
/*
* otherwise use the general form
*/
error = uiomove((char *)bp->b_data + blkoffset,
(int)xfersize, uio);
if (error)
break;
if (ioflag & (IO_VMIO|IO_DIRECT) ) {
/*
* If there are no dependencies, and it's VMIO,
* then we don't need the buf, mark it available
* for freeing. The VM has the data.
*/
bp->b_flags |= B_RELBUF;
brelse(bp);
} else {
/*
* Otherwise let whoever
* made the request take care of
* freeing it. We just queue
* it onto another list.
*/
bqrelse(bp);
}
}
/*
* This can only happen in the case of an error
* because the loop above resets bp to NULL on each iteration
* and on normal completion has not set a new value into it.
* so it must have come from a 'break' statement
*/
if (bp != NULL) {
if (ioflag & (IO_VMIO|IO_DIRECT)) {
bp->b_flags |= B_RELBUF;
brelse(bp);
} else
bqrelse(bp);
}
return (error);
}
