diff mbox

[xfsprogs,v2,3/3] xfs_io: add a new 'log_writes' command

Message ID 20171205235651.17102-4-ross.zwisler@linux.intel.com (mailing list archive)
State Superseded
Headers show

Commit Message

Ross Zwisler Dec. 5, 2017, 11:56 p.m. UTC
Add a new 'log_writes' command to xfs_io so that we can add dm-log-writes
log marks.  It's helpful to allow users of xfs_io to adds these marks from
within xfs_io instead of waiting until after xfs_io exits because then they
are able to replay the dm-log-writes log up to immediately after another
xfs_io operation such as mwrite.  This isolates the log replay from other
operations that happen as part of xfs_io exiting (file handles being
closed, mmaps being torn down, etc.).  This also allows users to insert
multiple marks between different xfs_io commands.

Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Suggested-by: Dave Chinner <david@fromorbit.com>
---
 configure.ac            |   1 +
 debian/control          |   2 +-
 include/builddefs.in    |   2 +
 io/Makefile             |   6 +++
 io/init.c               |   1 +
 io/io.h                 |   6 +++
 io/log_writes.c         | 101 ++++++++++++++++++++++++++++++++++++++++++++++++
 m4/Makefile             |   1 +
 m4/package_devmapper.m4 |  11 ++++++
 man/man8/xfs_io.8       |  23 ++++++++++-
 10 files changed, 152 insertions(+), 2 deletions(-)
 create mode 100644 io/log_writes.c
 create mode 100644 m4/package_devmapper.m4

Comments

Dave Chinner Dec. 6, 2017, 12:29 a.m. UTC | #1
On Tue, Dec 05, 2017 at 04:56:51PM -0700, Ross Zwisler wrote:
> Add a new 'log_writes' command to xfs_io so that we can add dm-log-writes
> log marks.  It's helpful to allow users of xfs_io to adds these marks from
> within xfs_io instead of waiting until after xfs_io exits because then they
> are able to replay the dm-log-writes log up to immediately after another
> xfs_io operation such as mwrite.  This isolates the log replay from other
> operations that happen as part of xfs_io exiting (file handles being
> closed, mmaps being torn down, etc.).  This also allows users to insert
> multiple marks between different xfs_io commands.
> 
> Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
> Suggested-by: Dave Chinner <david@fromorbit.com>

having just looked at the xfs_io mess, this new command needs to set
"exitcode = 1" on failure. I haven't looked at anything else in the
patch.

Cheers,

Dave.
Ross Zwisler Dec. 6, 2017, 4:38 a.m. UTC | #2
On Wed, Dec 06, 2017 at 11:29:46AM +1100, Dave Chinner wrote:
> On Tue, Dec 05, 2017 at 04:56:51PM -0700, Ross Zwisler wrote:
> > Add a new 'log_writes' command to xfs_io so that we can add dm-log-writes
> > log marks.  It's helpful to allow users of xfs_io to adds these marks from
> > within xfs_io instead of waiting until after xfs_io exits because then they
> > are able to replay the dm-log-writes log up to immediately after another
> > xfs_io operation such as mwrite.  This isolates the log replay from other
> > operations that happen as part of xfs_io exiting (file handles being
> > closed, mmaps being torn down, etc.).  This also allows users to insert
> > multiple marks between different xfs_io commands.
> > 
> > Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
> > Suggested-by: Dave Chinner <david@fromorbit.com>
> 
> having just looked at the xfs_io mess, this new command needs to set
> "exitcode = 1" on failure. I haven't looked at anything else in the
> patch.

Ah, yea, I did think it was odd that when hitting errors in xfs_io it always
still returned 0.  I'll fix up for the next version.  Is someone else working
on making the rest of the error cases in xfs_io return 1 as well, or should I
throw that at the front of the series?
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Ross Zwisler Dec. 6, 2017, 4:41 a.m. UTC | #3
On Tue, Dec 05, 2017 at 09:38:13PM -0700, Ross Zwisler wrote:
> On Wed, Dec 06, 2017 at 11:29:46AM +1100, Dave Chinner wrote:
> > On Tue, Dec 05, 2017 at 04:56:51PM -0700, Ross Zwisler wrote:
> > > Add a new 'log_writes' command to xfs_io so that we can add dm-log-writes
> > > log marks.  It's helpful to allow users of xfs_io to adds these marks from
> > > within xfs_io instead of waiting until after xfs_io exits because then they
> > > are able to replay the dm-log-writes log up to immediately after another
> > > xfs_io operation such as mwrite.  This isolates the log replay from other
> > > operations that happen as part of xfs_io exiting (file handles being
> > > closed, mmaps being torn down, etc.).  This also allows users to insert
> > > multiple marks between different xfs_io commands.
> > > 
> > > Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
> > > Suggested-by: Dave Chinner <david@fromorbit.com>
> > 
> > having just looked at the xfs_io mess, this new command needs to set
> > "exitcode = 1" on failure. I haven't looked at anything else in the
> > patch.
> 
> Ah, yea, I did think it was odd that when hitting errors in xfs_io it always
> still returned 0.  I'll fix up for the next version.  Is someone else working
> on making the rest of the error cases in xfs_io return 1 as well, or should I
> throw that at the front of the series?

Ah, nevermind, I found your patch that sets exitcode throughout xfs_io.  I'll
fix up for the next version, thanks for the review.
--
To unsubscribe from this list: send the line "unsubscribe linux-xfs" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Dave Chinner Dec. 6, 2017, 5:43 a.m. UTC | #4
On Tue, Dec 05, 2017 at 09:38:13PM -0700, Ross Zwisler wrote:
> On Wed, Dec 06, 2017 at 11:29:46AM +1100, Dave Chinner wrote:
> > On Tue, Dec 05, 2017 at 04:56:51PM -0700, Ross Zwisler wrote:
> > > Add a new 'log_writes' command to xfs_io so that we can add dm-log-writes
> > > log marks.  It's helpful to allow users of xfs_io to adds these marks from
> > > within xfs_io instead of waiting until after xfs_io exits because then they
> > > are able to replay the dm-log-writes log up to immediately after another
> > > xfs_io operation such as mwrite.  This isolates the log replay from other
> > > operations that happen as part of xfs_io exiting (file handles being
> > > closed, mmaps being torn down, etc.).  This also allows users to insert
> > > multiple marks between different xfs_io commands.
> > > 
> > > Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
> > > Suggested-by: Dave Chinner <david@fromorbit.com>
> > 
> > having just looked at the xfs_io mess, this new command needs to set
> > "exitcode = 1" on failure. I haven't looked at anything else in the
> > patch.
> 
> Ah, yea, I did think it was odd that when hitting errors in xfs_io it always
> still returned 0.  I'll fix up for the next version.  Is someone else working
> on making the rest of the error cases in xfs_io return 1 as well, or should I
> throw that at the front of the series?

I sent a patch earlier today to do exactly that.

https://marc.info/?l=fstests&m=151252000511167&w=2

Cheers,

Dave.
diff mbox

Patch

diff --git a/configure.ac b/configure.ac
index f3325aa0..f83d5817 100644
--- a/configure.ac
+++ b/configure.ac
@@ -164,6 +164,7 @@  AC_NEED_INTERNAL_FSXATTR
 AC_HAVE_GETFSMAP
 AC_HAVE_STATFS_FLAGS
 AC_HAVE_MAP_SYNC
+AC_HAVE_DEVMAPPER
 
 if test "$enable_blkid" = yes; then
 AC_HAVE_BLKID_TOPO
diff --git a/debian/control b/debian/control
index ad816622..5c26e3d7 100644
--- a/debian/control
+++ b/debian/control
@@ -3,7 +3,7 @@  Section: admin
 Priority: optional
 Maintainer: XFS Development Team <linux-xfs@vger.kernel.org>
 Uploaders: Nathan Scott <nathans@debian.org>, Anibal Monsalve Salazar <anibal@debian.org>
-Build-Depends: uuid-dev, dh-autoreconf, debhelper (>= 5), gettext, libtool, libreadline-gplv2-dev | libreadline5-dev, libblkid-dev (>= 2.17), linux-libc-dev
+Build-Depends: uuid-dev, dh-autoreconf, debhelper (>= 5), gettext, libtool, libreadline-gplv2-dev | libreadline5-dev, libblkid-dev (>= 2.17), linux-libc-dev, libdevmapper-dev
 Standards-Version: 3.9.1
 Homepage: http://xfs.org/
 
diff --git a/include/builddefs.in b/include/builddefs.in
index 126f7e95..66bdbfa2 100644
--- a/include/builddefs.in
+++ b/include/builddefs.in
@@ -35,6 +35,7 @@  LIBTERMCAP = @libtermcap@
 LIBEDITLINE = @libeditline@
 LIBREADLINE = @libreadline@
 LIBBLKID = @libblkid@
+LIBDEVMAPPER = @libdevmapper@
 LIBXFS = $(TOPDIR)/libxfs/libxfs.la
 LIBXCMD = $(TOPDIR)/libxcmd/libxcmd.la
 LIBXLOG = $(TOPDIR)/libxlog/libxlog.la
@@ -116,6 +117,7 @@  NEED_INTERNAL_FSXATTR = @need_internal_fsxattr@
 HAVE_GETFSMAP = @have_getfsmap@
 HAVE_STATFS_FLAGS = @have_statfs_flags@
 HAVE_MAP_SYNC = @have_map_sync@
+HAVE_DEVMAPPER = @have_devmapper@
 
 GCCFLAGS = -funsigned-char -fno-strict-aliasing -Wall
 #	   -Wbitwise -Wno-transparent-union -Wno-old-initializer -Wno-decl
diff --git a/io/Makefile b/io/Makefile
index 2987ee11..b7585a1b 100644
--- a/io/Makefile
+++ b/io/Makefile
@@ -107,6 +107,12 @@  ifeq ($(HAVE_MAP_SYNC),yes)
 LCFLAGS += -DHAVE_MAP_SYNC
 endif
 
+ifeq ($(HAVE_DEVMAPPER),yes)
+CFILES += log_writes.c
+LLDLIBS += $(LIBDEVMAPPER)
+LCFLAGS += -DHAVE_DEVMAPPER
+endif
+
 # On linux we get fsmap from the system or define it ourselves
 # so include this based on platform type.  If this reverts to only
 # the autoconf check w/o local definition, change to testing HAVE_GETFSMAP
diff --git a/io/init.c b/io/init.c
index 20d5f80d..34d87b5f 100644
--- a/io/init.c
+++ b/io/init.c
@@ -72,6 +72,7 @@  init_commands(void)
 	help_init();
 	imap_init();
 	inject_init();
+	log_writes_init();
 	madvise_init();
 	mincore_init();
 	mmap_init();
diff --git a/io/io.h b/io/io.h
index 8b2753b3..9349cc75 100644
--- a/io/io.h
+++ b/io/io.h
@@ -186,3 +186,9 @@  extern void		fsmap_init(void);
 #else
 # define fsmap_init()	do { } while (0)
 #endif
+
+#ifdef HAVE_DEVMAPPER
+extern void		log_writes_init(void);
+#else
+#define log_writes_init()	do { } while (0)
+#endif
diff --git a/io/log_writes.c b/io/log_writes.c
new file mode 100644
index 00000000..37c0024f
--- /dev/null
+++ b/io/log_writes.c
@@ -0,0 +1,101 @@ 
+/*
+ * Copyright (c) 2017 Intel Corporation.
+ * All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it would be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write the Free Software Foundation,
+ * Inc.,  51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+
+#include "platform_defs.h"
+#include <libdevmapper.h>
+#include "command.h"
+#include "init.h"
+
+static cmdinfo_t log_writes_cmd;
+
+static void
+mark_log(const char *device, const char *mark)
+{
+	struct dm_task *dmt;
+	const int size = 80;
+	char message[size];
+	int len;
+
+	len = snprintf(message, size, "mark %s", mark);
+	if (len >= size) {
+		printf("mark '%s' is too long\n", mark);
+		return;
+	}
+
+	if (!(dmt = dm_task_create(DM_DEVICE_TARGET_MSG)))
+		return;
+
+	if (!dm_task_set_name(dmt, device))
+		goto out;
+
+	if (!dm_task_set_sector(dmt, 0))
+		goto out;
+
+	if (!dm_task_set_message(dmt, message))
+		goto out;
+
+	dm_task_run(dmt);
+out:
+	dm_task_destroy(dmt);
+}
+
+static int
+log_writes_f(
+	int			argc,
+	char			**argv)
+{
+	const char *device = NULL;
+	const char *mark = NULL;
+	int c;
+
+	while ((c = getopt(argc, argv, "d:m:")) != EOF) {
+		switch (c) {
+		case 'd':
+			device = optarg;
+			break;
+		case 'm':
+			mark = optarg;
+			break;
+		default:
+			return command_usage(&log_writes_cmd);
+		}
+	}
+
+	if (device == NULL || mark == NULL)
+		return command_usage(&log_writes_cmd);
+
+	mark_log(device, mark);
+	return 0;
+}
+
+void
+log_writes_init(void)
+{
+	log_writes_cmd.name = "log_writes";
+	log_writes_cmd.altname = "lw";
+	log_writes_cmd.cfunc = log_writes_f;
+	log_writes_cmd.flags = CMD_NOMAP_OK | CMD_NOFILE_OK | CMD_FOREIGN_OK
+				| CMD_FLAG_ONESHOT;
+	log_writes_cmd.argmin = 0;
+	log_writes_cmd.argmax = -1;
+	log_writes_cmd.args = _("-d device -m mark");
+	log_writes_cmd.oneline =
+		_("create mark <mark> in the dm-log-writes log specified by <device>");
+
+	add_command(&log_writes_cmd);
+}
diff --git a/m4/Makefile b/m4/Makefile
index 4706121b..77f2edda 100644
--- a/m4/Makefile
+++ b/m4/Makefile
@@ -15,6 +15,7 @@  CONFIGURE = \
 LSRCFILES = \
 	manual_format.m4 \
 	package_blkid.m4 \
+	package_devmapper.m4 \
 	package_globals.m4 \
 	package_libcdev.m4 \
 	package_pthread.m4 \
diff --git a/m4/package_devmapper.m4 b/m4/package_devmapper.m4
new file mode 100644
index 00000000..821f283c
--- /dev/null
+++ b/m4/package_devmapper.m4
@@ -0,0 +1,11 @@ 
+#
+# See if libdevmapper is available on the system.
+#
+AC_DEFUN([AC_HAVE_DEVMAPPER],
+[ AC_SEARCH_LIBS([dm_task_create], [devmapper],
+        libdevmapper="-ldevmapper"
+        have_devmapper=yes,
+        have_devmapper=no,)
+    AC_SUBST(have_devmapper)
+    AC_SUBST(libdevmapper)
+])
diff --git a/man/man8/xfs_io.8 b/man/man8/xfs_io.8
index 1693f7f1..4fce6d39 100644
--- a/man/man8/xfs_io.8
+++ b/man/man8/xfs_io.8
@@ -1123,7 +1123,27 @@  version of policy structure (numeric)
 .BR get_encpolicy
 On filesystems that support encryption, display the encryption policy of the
 current file.
+.RE
+.PD
+.TP
+.BI "log_writes \-d " device " \-m "  mark
+Create a mark named
+.I mark
+in the dm-log-writes log specified by
+.I device.
+This is intended to be equivalent to the shell command:
 
+.B dmsetup message
+.I device
+.B 0 mark
+.I mark
+.PD
+.RE
+.TP
+.B lw
+See the
+.B log_writes
+command.
 .SH SEE ALSO
 .BR mkfs.xfs (8),
 .BR xfsctl (3),
@@ -1141,4 +1161,5 @@  current file.
 .BR open (2),
 .BR pread (2),
 .BR pwrite (2),
-.BR readdir (3).
+.BR readdir (3),
+.BR dmsetup (8).