Merge branch 'xfs-for-linus-v3.13-rc5' into for-next

7 files changed, 92 insertions, 40 deletions

diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c
index 8401f11f378f..3b2c14b6f0fb 100644
--- a/fs/xfs/xfs_bmap.c
+++ b/fs/xfs/xfs_bmap.c
@@ -3648,10 +3648,19 @@ xfs_bmap_btalloc(
         int             isaligned;
         int             tryagain;
         int             error;
+        int             stripe_align;
 
         ASSERT(ap->length);
 
         mp = ap->ip->i_mount;
+
+        /* stripe alignment for allocation is determined by mount parameters */
+        stripe_align = 0;
+        if (mp->m_swidth && (mp->m_flags & XFS_MOUNT_SWALLOC))
+                stripe_align = mp->m_swidth;
+        else if (mp->m_dalign)
+                stripe_align = mp->m_dalign;
+
         align = ap->userdata ? xfs_get_extsz_hint(ap->ip) : 0;
         if (unlikely(align)) {
                 error = xfs_bmap_extsize_align(mp, &ap->got, &ap->prev,
@@ -3660,6 +3669,8 @@ xfs_bmap_btalloc(
                 ASSERT(!error);
                 ASSERT(ap->length);
         }
+
+
         nullfb = *ap->firstblock == NULLFSBLOCK;
         fb_agno = nullfb ? NULLAGNUMBER : XFS_FSB_TO_AGNO(mp, *ap->firstblock);
         if (nullfb) {
@@ -3735,7 +3746,7 @@ xfs_bmap_btalloc(
          */
         if (!ap->flist->xbf_low && ap->aeof) {
                 if (!ap->offset) {
-                        args.alignment = mp->m_dalign;
+                        args.alignment = stripe_align;
                         atype = args.type;
                         isaligned = 1;
                         /*
@@ -3760,13 +3771,13 @@ xfs_bmap_btalloc(
                          * of minlen+alignment+slop doesn't go up
                          * between the calls.
                          */
-                        if (blen > mp->m_dalign && blen <= args.maxlen)
-                                nextminlen = blen - mp->m_dalign;
+                        if (blen > stripe_align && blen <= args.maxlen)
+                                nextminlen = blen - stripe_align;
                         else
                                 nextminlen = args.minlen;
-                        if (nextminlen + mp->m_dalign > args.minlen + 1)
+                        if (nextminlen + stripe_align > args.minlen + 1)
                                 args.minalignslop =
-                                        nextminlen + mp->m_dalign -
+                                        nextminlen + stripe_align -
                                         args.minlen - 1;
                         else
                                 args.minalignslop = 0;
@@ -3788,7 +3799,7 @@ xfs_bmap_btalloc(
                  */
                 args.type = atype;
                 args.fsbno = ap->blkno;
-                args.alignment = mp->m_dalign;
+                args.alignment = stripe_align;
                 args.minlen = nextminlen;
                 args.minalignslop = 0;
                 isaligned = 1;
diff --git a/fs/xfs/xfs_bmap_util.c b/fs/xfs/xfs_bmap_util.c
index 5887e41c0323..1394106ed22d 100644
--- a/fs/xfs/xfs_bmap_util.c
+++ b/fs/xfs/xfs_bmap_util.c
@@ -1187,7 +1187,12 @@ xfs_zero_remaining_bytes(
                 XFS_BUF_UNWRITE(bp);
                 XFS_BUF_READ(bp);
                 XFS_BUF_SET_ADDR(bp, xfs_fsb_to_db(ip, imap.br_startblock));
-                xfsbdstrat(mp, 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,
@@ -1200,7 +1205,12 @@ xfs_zero_remaining_bytes(
                 XFS_BUF_UNDONE(bp);
                 XFS_BUF_UNREAD(bp);
                 XFS_BUF_WRITE(bp);
-                xfsbdstrat(mp, 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,
diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c
index ce01c1a17cc1..9fccfb594291 100644
--- a/fs/xfs/xfs_buf.c
+++ b/fs/xfs/xfs_buf.c
@@ -698,7 +698,11 @@ xfs_buf_read_uncached(
         bp->b_flags |= XBF_READ;
         bp->b_ops = ops;
 
-        xfsbdstrat(target->bt_mount, bp);
+        if (XFS_FORCED_SHUTDOWN(target->bt_mount)) {
+                xfs_buf_relse(bp);
+                return NULL;
+        }
+        xfs_buf_iorequest(bp);
         xfs_buf_iowait(bp);
         return bp;
 }
@@ -1089,7 +1093,7 @@ xfs_bioerror(
  * This is meant for userdata errors; metadata bufs come with
  * iodone functions attached, so that we can track down errors.
  */
-STATIC int
+int
 xfs_bioerror_relse(
         struct xfs_buf  *bp)
 {
@@ -1152,7 +1156,7 @@ xfs_bwrite(
         ASSERT(xfs_buf_islocked(bp));
 
         bp->b_flags |= XBF_WRITE;
-        bp->b_flags &= ~(XBF_ASYNC | XBF_READ | _XBF_DELWRI_Q);
+        bp->b_flags &= ~(XBF_ASYNC | XBF_READ | _XBF_DELWRI_Q | XBF_WRITE_FAIL);
 
         xfs_bdstrat_cb(bp);
 
@@ -1164,25 +1168,6 @@ xfs_bwrite(
         return error;
 }
 
-/*
- * Wrapper around bdstrat so that we can stop data from going to disk in case
- * we are shutting down the filesystem.  Typically user data goes thru this
- * path; one of the exceptions is the superblock.
- */
-void
-xfsbdstrat(
-        struct xfs_mount        *mp,
-        struct xfs_buf          *bp)
-{
-        if (XFS_FORCED_SHUTDOWN(mp)) {
-                trace_xfs_bdstrat_shut(bp, _RET_IP_);
-                xfs_bioerror_relse(bp);
-                return;
-        }
-
-        xfs_buf_iorequest(bp);
-}
-
 STATIC void
 _xfs_buf_ioend(
         xfs_buf_t               *bp,
@@ -1516,6 +1501,12 @@ xfs_wait_buftarg(
                         struct xfs_buf *bp;
                         bp = list_first_entry(&dispose, struct xfs_buf, b_lru);
                         list_del_init(&bp->b_lru);
+                        if (bp->b_flags & XBF_WRITE_FAIL) {
+                                xfs_alert(btp->bt_mount,
+"Corruption Alert: Buffer at block 0x%llx had permanent write failures!\n"
+"Please run xfs_repair to determine the extent of the problem.",
+                                        (long long)bp->b_bn);
+                        }
                         xfs_buf_rele(bp);
                 }
                 if (loop++ != 0)
@@ -1789,7 +1780,7 @@ __xfs_buf_delwri_submit(
 
         blk_start_plug(&plug);
         list_for_each_entry_safe(bp, n, io_list, b_list) {
-                bp->b_flags &= ~(_XBF_DELWRI_Q | XBF_ASYNC);
+                bp->b_flags &= ~(_XBF_DELWRI_Q | XBF_ASYNC | XBF_WRITE_FAIL);
                 bp->b_flags |= XBF_WRITE;
 
                 if (!wait) {
diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h
index e65683361017..1cf21a4a9f22 100644
--- a/fs/xfs/xfs_buf.h
+++ b/fs/xfs/xfs_buf.h
@@ -45,6 +45,7 @@ typedef enum {
 #define XBF_ASYNC        (1 << 4) /* initiator will not wait for completion */
 #define XBF_DONE         (1 << 5) /* all pages in the buffer uptodate */
 #define XBF_STALE        (1 << 6) /* buffer has been staled, do not find it */
+#define XBF_WRITE_FAIL   (1 << 24)/* async writes have failed on this buffer */
 
 /* I/O hints for the BIO layer */
 #define XBF_SYNCIO       (1 << 10)/* treat this buffer as synchronous I/O */
@@ -70,6 +71,7 @@ typedef unsigned int xfs_buf_flags_t;
         { XBF_ASYNC,            "ASYNC" }, \
         { XBF_DONE,             "DONE" }, \
         { XBF_STALE,            "STALE" }, \
+        { XBF_WRITE_FAIL,       "WRITE_FAIL" }, \
         { XBF_SYNCIO,           "SYNCIO" }, \
         { XBF_FUA,              "FUA" }, \
         { XBF_FLUSH,            "FLUSH" }, \
@@ -80,6 +82,7 @@ typedef unsigned int xfs_buf_flags_t;
         { _XBF_DELWRI_Q,        "DELWRI_Q" }, \
         { _XBF_COMPOUND,        "COMPOUND" }
 
+
 /*
  * Internal state flags.
  */
@@ -269,9 +272,6 @@ extern void xfs_buf_unlock(xfs_buf_t *);
 
 /* Buffer Read and Write Routines */
 extern int xfs_bwrite(struct xfs_buf *bp);
-
-extern void xfsbdstrat(struct xfs_mount *, struct xfs_buf *);
-
 extern void xfs_buf_ioend(xfs_buf_t *,  int);
 extern void xfs_buf_ioerror(xfs_buf_t *, int);
 extern void xfs_buf_ioerror_alert(struct xfs_buf *, const char *func);
@@ -282,6 +282,8 @@ extern void xfs_buf_iomove(xfs_buf_t *, size_t, size_t, void *,
 #define xfs_buf_zero(bp, off, len) \
             xfs_buf_iomove((bp), (off), (len), NULL, XBRW_ZERO)
 
+extern int xfs_bioerror_relse(struct xfs_buf *);
+
 static inline int xfs_buf_geterror(xfs_buf_t *bp)
 {
         return bp ? bp->b_error : ENOMEM;
@@ -301,7 +303,8 @@ extern void xfs_buf_terminate(void);
 
 #define XFS_BUF_ZEROFLAGS(bp) \
         ((bp)->b_flags &= ~(XBF_READ|XBF_WRITE|XBF_ASYNC| \
-                            XBF_SYNCIO|XBF_FUA|XBF_FLUSH))
+                            XBF_SYNCIO|XBF_FUA|XBF_FLUSH| \
+                            XBF_WRITE_FAIL))
 
 void xfs_buf_stale(struct xfs_buf *bp);
 #define XFS_BUF_UNSTALE(bp)     ((bp)->b_flags &= ~XBF_STALE)
diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c
index 764117305438..33149113e333 100644
--- a/fs/xfs/xfs_buf_item.c
+++ b/fs/xfs/xfs_buf_item.c
@@ -495,6 +495,14 @@ xfs_buf_item_unpin(
         }
 }
 
+/*
+ * Buffer IO error rate limiting. Limit it to no more than 10 messages per 30
+ * seconds so as to not spam logs too much on repeated detection of the same
+ * buffer being bad..
+ */
+
+DEFINE_RATELIMIT_STATE(xfs_buf_write_fail_rl_state, 30 * HZ, 10);
+
 STATIC uint
 xfs_buf_item_push(
         struct xfs_log_item     *lip,
@@ -523,6 +531,14 @@ xfs_buf_item_push(
 
         trace_xfs_buf_item_push(bip);
 
+        /* has a previous flush failed due to IO errors? */
+        if ((bp->b_flags & XBF_WRITE_FAIL) &&
+            ___ratelimit(&xfs_buf_write_fail_rl_state, "XFS:")) {
+                xfs_warn(bp->b_target->bt_mount,
+"Detected failing async write on buffer block 0x%llx. Retrying async write.\n",
+                         (long long)bp->b_bn);
+        }
+
         if (!xfs_buf_delwri_queue(bp, buffer_list))
                 rval = XFS_ITEM_FLUSHING;
         xfs_buf_unlock(bp);
@@ -1095,8 +1111,9 @@ xfs_buf_iodone_callbacks(
 
                 xfs_buf_ioerror(bp, 0); /* errno of 0 unsets the flag */
 
-                if (!XFS_BUF_ISSTALE(bp)) {
-                        bp->b_flags |= XBF_WRITE | XBF_ASYNC | XBF_DONE;
+                if (!(bp->b_flags & (XBF_STALE|XBF_WRITE_FAIL))) {
+                        bp->b_flags |= XBF_WRITE | XBF_ASYNC |
+                                       XBF_DONE | XBF_WRITE_FAIL;
                         xfs_buf_iorequest(bp);
                 } else {
                         xfs_buf_relse(bp);
diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c
index 22b6f35765c1..bce53ac81096 100644
--- a/fs/xfs/xfs_log_recover.c
+++ b/fs/xfs/xfs_log_recover.c
@@ -193,7 +193,10 @@ xlog_bread_noalign(
         bp->b_io_length = nbblks;
         bp->b_error = 0;
 
-        xfsbdstrat(log->l_mp, bp);
+        if (XFS_FORCED_SHUTDOWN(log->l_mp))
+                return XFS_ERROR(EIO);
+
+        xfs_buf_iorequest(bp);
         error = xfs_buf_iowait(bp);
         if (error)
                 xfs_buf_ioerror_alert(bp, __func__);
@@ -4408,7 +4411,13 @@ xlog_do_recover(
         XFS_BUF_READ(bp);
         XFS_BUF_UNASYNC(bp);
         bp->b_ops = &xfs_sb_buf_ops;
-        xfsbdstrat(log->l_mp, bp);
+
+        if (XFS_FORCED_SHUTDOWN(log->l_mp)) {
+                xfs_buf_relse(bp);
+                return XFS_ERROR(EIO);
+        }
+
+        xfs_buf_iorequest(bp);
         error = xfs_buf_iowait(bp);
         if (error) {
                 xfs_buf_ioerror_alert(bp, __func__);
diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c
index c035d11b7734..647b6f1d8923 100644
--- a/fs/xfs/xfs_trans_buf.c
+++ b/fs/xfs/xfs_trans_buf.c
@@ -314,7 +314,18 @@ xfs_trans_read_buf_map(
                         ASSERT(bp->b_iodone == NULL);
                         XFS_BUF_READ(bp);
                         bp->b_ops = ops;
-                        xfsbdstrat(tp->t_mountp, bp);
+
+                        /*
+                         * XXX(hch): clean up the error handling here to be less
+                         * of a mess..
+                         */
+                        if (XFS_FORCED_SHUTDOWN(mp)) {
+                                trace_xfs_bdstrat_shut(bp, _RET_IP_);
+                                xfs_bioerror_relse(bp);
+                        } else {
+                                xfs_buf_iorequest(bp);
+                        }
+
                         error = xfs_buf_iowait(bp);
                         if (error) {
                                 xfs_buf_ioerror_alert(bp, __func__);