glusterfs.git/xlators/storage, branch v3.10.1

storage/posix: Use granular mutex locks for pgfid update syscalls

2017-03-13T14:21:20+00:00

Backport of
> Change-Id: Ie5d635951c483d858dc4be2a90fb24b8b5f4f02d
> BUG: 1421938
> Reviewed-on: https://review.gluster.org/16869
> (cherry picked from commit fc97b1dc38ad46302d51a988cda18335f88514a2)

Change-Id: Ie5d635951c483d858dc4be2a90fb24b8b5f4f02d
BUG: 1427387
Signed-off by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/16891
Tested-by: Krutika Dhananjay 
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

storage/posix: Use more granular mutex locks for atomic writes

2017-03-06T16:19:18+00:00

        Backport of: https://review.gluster.org/16785

Change-Id: If80b1543c4fe16cb81e2ae92ccca695f3bf450d7
BUG: 1427387
Signed-off-by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/16841
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Shyamsundar Ranganathan

storage/posix: Execute syscalls in xattrop under different locks

2017-03-02T17:54:23+00:00

        Backport of: https://review.gluster.org/16462 and
                     https://review.gluster.org/16792

... and not inode->lock. This is to prevent the epoll thread from
*potentially* being blocked on this lock in the worst case for
extended period elsewhere in the brick stack, while the syscalls
in xattrop are being performed under the same lock by a different
thread. This could potentially lead to ping-timeout, if the only
available epoll thread is busy waiting on the inode->lock, thereby
preventing it from picking up the ping request from the client(s).

Also removed some unused functions.

>Change-Id: I2054a06701ecab11aed1c04e80ee57bbe2e52564
>BUG: 1421938
>Signed-off-by: Krutika Dhananjay 
>Reviewed-on: https://review.gluster.org/16462
>Smoke: Gluster Build System 
>NetBSD-regression: NetBSD Build System 
>Reviewed-by: Raghavendra G 
>CentOS-regression: Gluster Build System 
(cherry picked from commit b5c26a462caf97bfc5380c81092f5c331ccaf1ae)

Change-Id: I2054a06701ecab11aed1c04e80ee57bbe2e52564
BUG: 1427387
Signed-off-by: Krutika Dhananjay 
Reviewed-on: https://review.gluster.org/16776
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Raghavendra G 
Reviewed-by: Shyamsundar Ranganathan

posix: Fix creation of files with S_ISVTX on FreeBSD

2017-02-20T15:42:51+00:00

On FreeBSD the S_ISVTX flag is completely ignored when creating a
regular file. Since gluster needs to create files with this flag set,
specialy for DHT link files, it's necessary to force the flag.

This fix does this by calling fchmod() after creating a file that
must have this flag set.

> Change-Id: I51eecfe4642974df6106b9084a0b144835a4997a
> BUG: 1411228
> Signed-off-by: Xavier Hernandez 
> Reviewed-on: https://review.gluster.org/16417
> Smoke: Gluster Build System 
> NetBSD-regression: NetBSD Build System 
> CentOS-regression: Gluster Build System 
> Reviewed-by: Raghavendra Bhat 
> Reviewed-by: Kaleb KEITHLEY 

Change-Id: I2087516383bd132c59bbab98eda8f2243a2163fe
BUG: 1424973
Signed-off-by: Xavier Hernandez 
Reviewed-on: https://review.gluster.org/16686
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Raghavendra Bhat 
Reviewed-by: Shyamsundar Ranganathan

dht : file rename operation is successful but log has error File exists

2017-01-10T06:27:50+00:00

Problem: file rename operation is successful but log has error
         'key:trusted.glusterfs.dht.linkto error:File exists'

Solution: The error comes in log only when file already has set this xattr
          (trusted.glusterfs.dht.linkto) before run rename operation and
          in that case sys_lsetxattr throws this kind of message.
          To avoid the error message in logs update the condition in
          posix_handle_pair.

BUG: 1404905
Change-Id: Iafd8cb45f9d7f4fe247e297a6ef0af978a8d0b30
Signed-off-by: Mohit Agrawal 
Reviewed-on: http://review.gluster.org/16185
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Raghavendra G

posix: make sure atime and mtime are set when calling lutimes()

2017-01-06T17:33:10+00:00

When overwriting an existing file with O_TRUNC, the 'atime' was set to
0, meaning the Epoch (01-Jan-1970 UTC). However, the 'mtime' gets
updated correcty.

In case 'atime' or 'mtime' is not passed in the 'struct iatt', the time
values passed to the systemcall are taken from the current values are
returned by lstat().

Change-Id: I7021b7161dcd6c9a3e515d98f6d4847533c434b3
BUG: 1401777
Reported-by: Eivind Sarto 
Signed-off-by: Niels de Vos 
Reviewed-on: http://review.gluster.org/16034
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Raghavendra Bhat 
CentOS-regression: Gluster Build System 
Reviewed-by: Kaleb KEITHLEY

storage/posix: log the error when locking the export directory fails

2017-01-03T13:42:40+00:00

posix xlator tries to prevent external unmount of the brick export directory
by locking it. It locks the export directory by opening it (i.e. doing a opendir).

But if opendir fails it errors out resulting in the brick process going down. The
error due to which opendir failed is not logged.

Change-Id: I28648382b7f0b88fb1aeb36152e5ab5d812374c9
BUG: 1401095
Signed-off-by: Raghavendra Bhat 
Reviewed-on: http://review.gluster.org/16018
Smoke: Gluster Build System 
Tested-by: Vijay Bellur 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Vijay Bellur

Fixes GlusterFS process crashes on deep directory hierarchy

2016-12-22T06:42:30+00:00

alloca() calls to recusive function posix_make_ancestryfromgfid
consumes the entire stack leading to segfault.
Changed the function to  iterative and reused the same linkname
buffer across iterations.

Change-Id: If065c8b9f7f85219bdd2d23259a115fe66f5f60d
BUG: 1405775
Signed-off-by: Sanoj Unnikrishnan 
Reviewed-on: http://review.gluster.org/16192
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Raghavendra G 
CentOS-regression: Gluster Build System

storage/posix: Do not create a directory which already exist

2016-12-22T05:41:03+00:00

Problem: While brick process starts, it tries to create trashcan
and internal_op directory with fixed gfids. For the first time
these directories will not be present. However, stop and start
of volume again tries to create these directories, which are
already present.
This throws a gfid exist warning and EEXIST error in logs.

Solution: Check if the directory exist on brick and gfid of
this dir and the requested gfid matches or not.

Change-Id: Ie4966a550ed44046e0f2d10e9dbd540182ea97d2
BUG: 1406348
Signed-off-by: Ashish Pandey 
Reviewed-on: http://review.gluster.org/16212
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
CentOS-regression: Gluster Build System 
Reviewed-by: Pranith Kumar Karampuri

events: Add FMT_WARN for gf_event

2016-11-18T09:56:39+00:00

Raghavendra G found that posix is trying to print %s
but passing an int when HEALTH_CHECK fails in posix.
These are the kind of bugs that should be caught
at compilation itself.
Also fixed the problematic gf_event() callers.

BUG: 1386097
Change-Id: Id7bd6d9a9690237cec3ca1aefa2aac085e8a1270
Signed-off-by: Pranith Kumar K 
Reviewed-on: http://review.gluster.org/15671
Smoke: Gluster Build System 
NetBSD-regression: NetBSD Build System 
Reviewed-by: Atin Mukherjee 
CentOS-regression: Gluster Build System