posix/ctime: Fix race during lookup ctime xattr heal

Problem:
Ctime heals the ctime xattr ("trusted.glusterfs.mdata") in lookup
if it's not present. In a multi client scenario, there is a race
which results in updating the ctime xattr to older value.

e.g. Let c1 and c2 be two clients and file1 be the file which
doesn't have the ctime xattr. Let the ctime of file1 be t1.
(from backend, ctime heals time attributes from backend when not present).

Now following operations are done on mount
c1 -> ls -l /mnt/file1  |   c2 -> ls -l /mnt/file1;echo "append" >> /mnt/file1;

The race is that the both c1 and c2 didn't fetch the ctime xattr in lookup,
so both of them tries to heal ctime to time 't1'. If c2 wins the race and
appends the file before c1 heals it, it sets the time to 't1' and updates
it to 't2' (because of append). Now c1 proceeds to heal and sets it to 't1'
which is incorrect.

Solution:
Compare the times during heal and only update the larger time. This is the
general approach used in ctime feature but got missed with healing legacy
files.

Backport of:
 > Patch: https://review.gluster.org/23131
 > BUG: 1734299
 > Change-Id: I930bda192c64c3d49d0aed431ce23d3bc57e51b7
 > Signed-off-by: Kotresh HR <khiremat@redhat.com>

fixes: bz#1737745
Change-Id: I930bda192c64c3d49d0aed431ce23d3bc57e51b7
Signed-off-by: Kotresh HR <khiremat@redhat.com>

1 files changed, 58 insertions, 18 deletions

diff --git a/xlators/storage/posix/src/posix-metadata.c b/xlators/storage/posix/src/posix-metadata.c
index 647c0bb3a65..57791fa9536 100644
--- a/xlators/storage/posix/src/posix-metadata.c
+++ b/xlators/storage/posix/src/posix-metadata.c
@@ -344,33 +344,73 @@ posix_set_mdata_xattr_legacy_files(xlator_t *this, inode_t *inode,
                                    struct mdata_iatt *mdata_iatt, int *op_errno)
 {
     posix_mdata_t *mdata = NULL;
+    posix_mdata_t imdata = {
+        0,
+    };
     int ret = 0;
+    gf_boolean_t mdata_already_set = _gf_false;
 
     GF_VALIDATE_OR_GOTO("posix", this, out);
     GF_VALIDATE_OR_GOTO(this->name, inode, out);
 
     LOCK(&inode->lock);
     {
-        mdata = GF_CALLOC(1, sizeof(posix_mdata_t), gf_posix_mt_mdata_attr);
-        if (!mdata) {
-            gf_msg(this->name, GF_LOG_ERROR, ENOMEM, P_MSG_NOMEM,
-                   "Could not allocate mdata. gfid: %s",
-                   uuid_utoa(inode->gfid));
-            ret = -1;
-            *op_errno = ENOMEM;
-            goto unlock;
-        }
+        ret = __inode_ctx_get1(inode, this, (uint64_t *)&mdata);
+        if (ret == 0 && mdata) {
+            mdata_already_set = _gf_true;
+        } else if (ret == -1 || !mdata) {
+            mdata = GF_CALLOC(1, sizeof(posix_mdata_t), gf_posix_mt_mdata_attr);
+            if (!mdata) {
+                gf_msg(this->name, GF_LOG_ERROR, ENOMEM, P_MSG_NOMEM,
+                       "Could not allocate mdata. gfid: %s",
+                       uuid_utoa(inode->gfid));
+                ret = -1;
+                *op_errno = ENOMEM;
+                goto unlock;
+            }
+
+            ret = posix_fetch_mdata_xattr(this, NULL, -1, inode, (void *)mdata,
+                                          op_errno);
+            if (ret == 0) {
+                /* Got mdata from disk. This is a race, another client
+                 * has healed the xattr during lookup. So set it in inode
+                 * ctx */
+                __inode_ctx_set1(inode, this, (uint64_t *)&mdata);
+                mdata_already_set = _gf_true;
+            } else {
+                *op_errno = 0;
+                mdata->version = 1;
+                mdata->flags = 0;
+                mdata->ctime.tv_sec = mdata_iatt->ia_ctime;
+                mdata->ctime.tv_nsec = mdata_iatt->ia_ctime_nsec;
+                mdata->atime.tv_sec = mdata_iatt->ia_atime;
+                mdata->atime.tv_nsec = mdata_iatt->ia_atime_nsec;
+                mdata->mtime.tv_sec = mdata_iatt->ia_mtime;
+                mdata->mtime.tv_nsec = mdata_iatt->ia_mtime_nsec;
 
-        mdata->version = 1;
-        mdata->flags = 0;
-        mdata->ctime.tv_sec = mdata_iatt->ia_ctime;
-        mdata->ctime.tv_nsec = mdata_iatt->ia_ctime_nsec;
-        mdata->atime.tv_sec = mdata_iatt->ia_atime;
-        mdata->atime.tv_nsec = mdata_iatt->ia_atime_nsec;
-        mdata->mtime.tv_sec = mdata_iatt->ia_mtime;
-        mdata->mtime.tv_nsec = mdata_iatt->ia_mtime_nsec;
+                __inode_ctx_set1(inode, this, (uint64_t *)&mdata);
+            }
+        }
 
-        __inode_ctx_set1(inode, this, (uint64_t *)&mdata);
+        if (mdata_already_set) {
+            /* Compare and update the larger time */
+            imdata.ctime.tv_sec = mdata_iatt->ia_ctime;
+            imdata.ctime.tv_nsec = mdata_iatt->ia_ctime_nsec;
+            imdata.atime.tv_sec = mdata_iatt->ia_atime;
+            imdata.atime.tv_nsec = mdata_iatt->ia_atime_nsec;
+            imdata.mtime.tv_sec = mdata_iatt->ia_mtime;
+            imdata.mtime.tv_nsec = mdata_iatt->ia_mtime_nsec;
+
+            if (posix_compare_timespec(&imdata.ctime, &mdata->ctime) > 0) {
+                mdata->ctime = imdata.ctime;
+            }
+            if (posix_compare_timespec(&imdata.mtime, &mdata->mtime) > 0) {
+                mdata->mtime = imdata.mtime;
+            }
+            if (posix_compare_timespec(&imdata.atime, &mdata->atime) > 0) {
+                mdata->atime = imdata.atime;
+            }
+        }
 
         ret = posix_store_mdata_xattr(this, NULL, -1, inode, mdata);
         if (ret) {