xfs_trans_buf.c

linux 内核源代码

			}
		}
		/*
		 * We never locked this buf ourselves, so we shouldn't
		 * brelse it either. Just get out.
		 */
		if (XFS_FORCED_SHUTDOWN(mp)) {
			xfs_buftrace("READ_BUF_INCORE XFSSHUTDN", bp);
			*bpp = NULL;
			return XFS_ERROR(EIO);
		}


		bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
		bip->bli_recur++;

		ASSERT(atomic_read(&bip->bli_refcount) > 0);
		xfs_buf_item_trace("READ RECUR", bip);
		*bpp = bp;
		return 0;
	}

	/*
	 * We always specify the BUF_BUSY flag within a transaction so
	 * that get_buf does not try to push out a delayed write buffer
	 * which might cause another transaction to take place (if the
	 * buffer was delayed alloc).  Such recursive transactions can
	 * easily deadlock with our current transaction as well as cause
	 * us to run out of stack space.
	 */
	bp = xfs_buf_read_flags(target, blkno, len, flags | BUF_BUSY);
	if (bp == NULL) {
		*bpp = NULL;
		return 0;
	}
	if (XFS_BUF_GETERROR(bp) != 0) {
	    XFS_BUF_SUPER_STALE(bp);
		xfs_buftrace("READ ERROR", bp);
		error = XFS_BUF_GETERROR(bp);

		xfs_ioerror_alert("xfs_trans_read_buf", mp,
				  bp, blkno);
		if (tp->t_flags & XFS_TRANS_DIRTY)
			xfs_force_shutdown(tp->t_mountp, SHUTDOWN_META_IO_ERROR);
		xfs_buf_relse(bp);
		return error;
	}
#ifdef DEBUG
	if (xfs_do_error && !(tp->t_flags & XFS_TRANS_DIRTY)) {
		if (xfs_error_target == target) {
			if (((xfs_req_num++) % xfs_error_mod) == 0) {
				xfs_force_shutdown(tp->t_mountp,
						   SHUTDOWN_META_IO_ERROR);
				xfs_buf_relse(bp);
				cmn_err(CE_DEBUG, "Returning trans error!\n");
				return XFS_ERROR(EIO);
			}
		}
	}
#endif
	if (XFS_FORCED_SHUTDOWN(mp))
		goto shutdown_abort;

	/*
	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
	 * it doesn't have one yet, then allocate one and initialize it.
	 * The checks to see if one is there are in xfs_buf_item_init().
	 */
	xfs_buf_item_init(bp, tp->t_mountp);

	/*
	 * Set the recursion count for the buffer within this transaction
	 * to 0.
	 */
	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t*);
	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
	bip->bli_recur = 0;

	/*
	 * Take a reference for this transaction on the buf item.
	 */
	atomic_inc(&bip->bli_refcount);

	/*
	 * Get a log_item_desc to point at the new item.
	 */
	(void) xfs_trans_add_item(tp, (xfs_log_item_t*)bip);

	/*
	 * Initialize b_fsprivate2 so we can find it with incore_match()
	 * above.
	 */
	XFS_BUF_SET_FSPRIVATE2(bp, tp);

	xfs_buftrace("TRANS READ", bp);
	xfs_buf_item_trace("READ", bip);
	*bpp = bp;
	return 0;

shutdown_abort:
	/*
	 * the theory here is that buffer is good but we're
	 * bailing out because the filesystem is being forcibly
	 * shut down.  So we should leave the b_flags alone since
	 * the buffer's not staled and just get out.
	 */
#if defined(DEBUG)
	if (XFS_BUF_ISSTALE(bp) && XFS_BUF_ISDELAYWRITE(bp))
		cmn_err(CE_NOTE, "about to pop assert, bp == 0x%p", bp);
#endif
	ASSERT((XFS_BUF_BFLAGS(bp) & (XFS_B_STALE|XFS_B_DELWRI)) !=
						(XFS_B_STALE|XFS_B_DELWRI));

	xfs_buftrace("READ_BUF XFSSHUTDN", bp);
	xfs_buf_relse(bp);
	*bpp = NULL;
	return XFS_ERROR(EIO);
}


/*
 * Release the buffer bp which was previously acquired with one of the
 * xfs_trans_... buffer allocation routines if the buffer has not
 * been modified within this transaction.  If the buffer is modified
 * within this transaction, do decrement the recursion count but do
 * not release the buffer even if the count goes to 0.  If the buffer is not
 * modified within the transaction, decrement the recursion count and
 * release the buffer if the recursion count goes to 0.
 *
 * If the buffer is to be released and it was not modified before
 * this transaction began, then free the buf_log_item associated with it.
 *
 * If the transaction pointer is NULL, make this just a normal
 * brelse() call.
 */
void
xfs_trans_brelse(xfs_trans_t	*tp,
		 xfs_buf_t	*bp)
{
	xfs_buf_log_item_t	*bip;
	xfs_log_item_t		*lip;
	xfs_log_item_desc_t	*lidp;

	/*
	 * Default to a normal brelse() call if the tp is NULL.
	 */
	if (tp == NULL) {
		ASSERT(XFS_BUF_FSPRIVATE2(bp, void *) == NULL);
		/*
		 * If there's a buf log item attached to the buffer,
		 * then let the AIL know that the buffer is being
		 * unlocked.
		 */
		if (XFS_BUF_FSPRIVATE(bp, void *) != NULL) {
			lip = XFS_BUF_FSPRIVATE(bp, xfs_log_item_t *);
			if (lip->li_type == XFS_LI_BUF) {
				bip = XFS_BUF_FSPRIVATE(bp,xfs_buf_log_item_t*);
				xfs_trans_unlocked_item(
						bip->bli_item.li_mountp,
						lip);
			}
		}
		xfs_buf_relse(bp);
		return;
	}

	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
	ASSERT(bip->bli_item.li_type == XFS_LI_BUF);
	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
	ASSERT(atomic_read(&bip->bli_refcount) > 0);

	/*
	 * Find the item descriptor pointing to this buffer's
	 * log item.  It must be there.
	 */
	lidp = xfs_trans_find_item(tp, (xfs_log_item_t*)bip);
	ASSERT(lidp != NULL);

	/*
	 * If the release is just for a recursive lock,
	 * then decrement the count and return.
	 */
	if (bip->bli_recur > 0) {
		bip->bli_recur--;
		xfs_buf_item_trace("RELSE RECUR", bip);
		return;
	}

	/*
	 * If the buffer is dirty within this transaction, we can't
	 * release it until we commit.
	 */
	if (lidp->lid_flags & XFS_LID_DIRTY) {
		xfs_buf_item_trace("RELSE DIRTY", bip);
		return;
	}

	/*
	 * If the buffer has been invalidated, then we can't release
	 * it until the transaction commits to disk unless it is re-dirtied
	 * as part of this transaction.  This prevents us from pulling
	 * the item from the AIL before we should.
	 */
	if (bip->bli_flags & XFS_BLI_STALE) {
		xfs_buf_item_trace("RELSE STALE", bip);
		return;
	}

	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));
	xfs_buf_item_trace("RELSE", bip);

	/*
	 * Free up the log item descriptor tracking the released item.
	 */
	xfs_trans_free_item(tp, lidp);

	/*
	 * Clear the hold flag in the buf log item if it is set.
	 * We wouldn't want the next user of the buffer to
	 * get confused.
	 */
	if (bip->bli_flags & XFS_BLI_HOLD) {
		bip->bli_flags &= ~XFS_BLI_HOLD;
	}

	/*
	 * Drop our reference to the buf log item.
	 */
	atomic_dec(&bip->bli_refcount);

	/*
	 * If the buf item is not tracking data in the log, then
	 * we must free it before releasing the buffer back to the
	 * free pool.  Before releasing the buffer to the free pool,
	 * clear the transaction pointer in b_fsprivate2 to dissolve
	 * its relation to this transaction.
	 */
	if (!xfs_buf_item_dirty(bip)) {
/***
		ASSERT(bp->b_pincount == 0);
***/
		ASSERT(atomic_read(&bip->bli_refcount) == 0);
		ASSERT(!(bip->bli_item.li_flags & XFS_LI_IN_AIL));
		ASSERT(!(bip->bli_flags & XFS_BLI_INODE_ALLOC_BUF));
		xfs_buf_item_relse(bp);
		bip = NULL;
	}
	XFS_BUF_SET_FSPRIVATE2(bp, NULL);

	/*
	 * If we've still got a buf log item on the buffer, then
	 * tell the AIL that the buffer is being unlocked.
	 */
	if (bip != NULL) {
		xfs_trans_unlocked_item(bip->bli_item.li_mountp,
					(xfs_log_item_t*)bip);
	}

	xfs_buf_relse(bp);
	return;
}

/*
 * Add the locked buffer to the transaction.
 * The buffer must be locked, and it cannot be associated with any
 * transaction.
 *
 * If the buffer does not yet have a buf log item associated with it,
 * then allocate one for it.  Then add the buf item to the transaction.
 */
void
xfs_trans_bjoin(xfs_trans_t	*tp,
		xfs_buf_t	*bp)
{
	xfs_buf_log_item_t	*bip;

	ASSERT(XFS_BUF_ISBUSY(bp));
	ASSERT(XFS_BUF_FSPRIVATE2(bp, void *) == NULL);

	/*
	 * The xfs_buf_log_item pointer is stored in b_fsprivate.  If
	 * it doesn't have one yet, then allocate one and initialize it.
	 * The checks to see if one is there are in xfs_buf_item_init().
	 */
	xfs_buf_item_init(bp, tp->t_mountp);
	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
	ASSERT(!(bip->bli_flags & XFS_BLI_LOGGED));

	/*
	 * Take a reference for this transaction on the buf item.
	 */
	atomic_inc(&bip->bli_refcount);

	/*
	 * Get a log_item_desc to point at the new item.
	 */
	(void) xfs_trans_add_item(tp, (xfs_log_item_t *)bip);

	/*
	 * Initialize b_fsprivate2 so we can find it with incore_match()
	 * in xfs_trans_get_buf() and friends above.
	 */
	XFS_BUF_SET_FSPRIVATE2(bp, tp);

	xfs_buf_item_trace("BJOIN", bip);
}

/*
 * Mark the buffer as not needing to be unlocked when the buf item's
 * IOP_UNLOCK() routine is called.  The buffer must already be locked
 * and associated with the given transaction.
 */
/* ARGSUSED */
void
xfs_trans_bhold(xfs_trans_t	*tp,
		xfs_buf_t	*bp)
{
	xfs_buf_log_item_t	*bip;

	ASSERT(XFS_BUF_ISBUSY(bp));
	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);

	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
	ASSERT(atomic_read(&bip->bli_refcount) > 0);
	bip->bli_flags |= XFS_BLI_HOLD;
	xfs_buf_item_trace("BHOLD", bip);
}

/*
 * Cancel the previous buffer hold request made on this buffer
 * for this transaction.
 */
void
xfs_trans_bhold_release(xfs_trans_t	*tp,
			xfs_buf_t	*bp)
{
	xfs_buf_log_item_t	*bip;

	ASSERT(XFS_BUF_ISBUSY(bp));
	ASSERT(XFS_BUF_FSPRIVATE2(bp, xfs_trans_t *) == tp);
	ASSERT(XFS_BUF_FSPRIVATE(bp, void *) != NULL);

	bip = XFS_BUF_FSPRIVATE(bp, xfs_buf_log_item_t *);
	ASSERT(!(bip->bli_flags & XFS_BLI_STALE));
	ASSERT(!(bip->bli_format.blf_flags & XFS_BLI_CANCEL));
	ASSERT(atomic_read(&bip->bli_refcount) > 0);
	ASSERT(bip->bli_flags & XFS_BLI_HOLD);
	bip->bli_flags &= ~XFS_BLI_HOLD;
	xfs_buf_item_trace("BHOLD RELEASE", bip);
}

/*
 * This is called to mark bytes first through last inclusive of the given
 * buffer as needing to be logged when the transaction is committed.
 * The buffer must already be associated with the given transaction.
 *
 * First and last are numbers relative to the beginning of this buffer,
 * so the first byte in the buffer is numbered 0 regardless of the
 * value of b_blkno.
 */
void
xfs_trans_log_buf(xfs_trans_t	*tp,