[PATCH 0/6] libxfrog: wrap version ioctl calls

[PATCH 0/6] libxfrog: wrap version ioctl calls

[Darrick J. Wong]

Hi all,

This series introduces into libfrog some wrapper functions for the XFS
geometry, bulkstat, and inumbers ioctls, then converts the tools to use
the new wrappers.  This is intended to smooth the transition to the new
v5 ioctls that were introduced in kernel 5.3 by providing standard
fallback code to the old ioctls on old kernels.

If you're going to start using this mess, you probably ought to just
pull from my git trees, which are linked below.

This is an extraordinary way to destroy everything.  Enjoy!
Comments and questions are, as always, welcome.

--D

xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=libfrog-refactor

[PATCH 1/6] libfrog: refactor online geometry queries

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Refactor all the open-coded XFS_IOC_FSGEOMETRY queries into a single
helper that we can use to standardize behaviors across mixed xfslibs
versions.  This is the prelude to introducing a new FSGEOMETRY version
in 5.2 and needing to fix the (relatively few) client programs.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 Makefile            |    1 +
 fsr/xfs_fsr.c       |   26 ++++----------------------
 growfs/xfs_growfs.c |   25 +++++++++----------------
 include/xfrog.h     |   22 ++++++++++++++++++++++
 io/bmap.c           |    3 ++-
 io/fsmap.c          |    3 ++-
 io/open.c           |    3 ++-
 io/stat.c           |    5 +++--
 libfrog/fsgeom.c    |   18 ++++++++++++++++++
 quota/free.c        |    6 +++---
 repair/xfs_repair.c |    5 +++--
 rtcp/Makefile       |    3 +++
 rtcp/xfs_rtcp.c     |    7 ++++---
 scrub/phase1.c      |    5 +++--
 spaceman/file.c     |    3 ++-
 spaceman/info.c     |   27 +++++++++------------------
 16 files changed, 90 insertions(+), 72 deletions(-)
 create mode 100644 include/xfrog.h


diff --git a/Makefile b/Makefile
index 9204bed8..0edc2700 100644
--- a/Makefile
+++ b/Makefile
@@ -107,6 +107,7 @@ copy: libxlog
 mkfs: libxcmd
 spaceman: libxcmd
 scrub: libhandle libxcmd
+rtcp: libfrog
 
 ifeq ($(HAVE_BUILDDEFS), yes)
 include $(BUILDRULES)
diff --git a/fsr/xfs_fsr.c b/fsr/xfs_fsr.c
index 1963a05e..1541917e 100644
--- a/fsr/xfs_fsr.c
+++ b/fsr/xfs_fsr.c
@@ -11,6 +11,7 @@
 #include "xfs_bmap_btree.h"
 #include "xfs_attr_sf.h"
 #include "path.h"
+#include "xfrog.h"
 
 #include <fcntl.h>
 #include <errno.h>
@@ -83,9 +84,8 @@ int cmp(const void *, const void *);
 static void tmp_init(char *mnt);
 static char * tmp_next(char *mnt);
 static void tmp_close(char *mnt);
-int xfs_getgeom(int , struct xfs_fsop_geom_v1 * );
 
-static struct xfs_fsop_geom_v1 fsgeom;	/* geometry of active mounted system */
+static struct xfs_fsop_geom fsgeom;	/* geometry of active mounted system */
 
 #define NMOUNT 64
 static int numfs;
@@ -102,12 +102,6 @@ static int	nfrags = 0;	/* Debug option: Coerse into specific number
 				 * of extents */
 static int	openopts = O_CREAT|O_EXCL|O_RDWR|O_DIRECT;
 
-static int
-xfs_fsgeometry(int fd, struct xfs_fsop_geom_v1 *geom)
-{
-    return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, geom);
-}
-
 static int
 xfs_bulkstat_single(int fd, xfs_ino_t *lastip, struct xfs_bstat *ubuffer)
 {
@@ -630,7 +624,7 @@ fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
 		return -1;
 	}
 
-	if (xfs_getgeom(fsfd, &fsgeom) < 0 ) {
+	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
 		fsrprintf(_("Skipping %s: could not get XFS geometry\n"),
 			  mntdir);
 		close(fsfd);
@@ -772,7 +766,7 @@ fsrfile(char *fname, xfs_ino_t ino)
 	}
 
 	/* Get the fs geometry */
-	if (xfs_getgeom(fsfd, &fsgeom) < 0 ) {
+	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
 		fsrprintf(_("Unable to get geom on fs for: %s\n"), fname);
 		goto out;
 	}
@@ -1612,18 +1606,6 @@ getnextents(int fd)
 	return(nextents);
 }
 
-/*
- * Get the fs geometry
- */
-int
-xfs_getgeom(int fd, struct xfs_fsop_geom_v1 *fsgeom)
-{
-	if (xfs_fsgeometry(fd, fsgeom) < 0) {
-		return -1;
-	}
-	return 0;
-}
-
 /*
  * Get xfs realtime space information
  */
diff --git a/growfs/xfs_growfs.c b/growfs/xfs_growfs.c
index 20089d2b..86b1d542 100644
--- a/growfs/xfs_growfs.c
+++ b/growfs/xfs_growfs.c
@@ -7,6 +7,7 @@
 #include "libxfs.h"
 #include "path.h"
 #include "fsgeom.h"
+#include "xfrog.h"
 
 static void
 usage(void)
@@ -165,22 +166,14 @@ main(int argc, char **argv)
 	}
 
 	/* get the current filesystem size & geometry */
-	if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY, &geo) < 0) {
-		/*
-		 * OK, new xfsctl barfed - back off and try earlier version
-		 * as we're probably running an older kernel version.
-		 * Only field added in the v2 geometry xfsctl is "logsunit"
-		 * so we'll zero that out for later display (as zero).
-		 */
-		geo.logsunit = 0;
-		if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY_V1, &geo) < 0) {
-			fprintf(stderr, _(
-				"%s: cannot determine geometry of filesystem"
-				" mounted at %s: %s\n"),
-				progname, fname, strerror(errno));
-			exit(1);
-		}
+	if (xfrog_geometry(ffd, &geo) < 0) {
+		fprintf(stderr, _(
+			"%s: cannot determine geometry of filesystem"
+			" mounted at %s: %s\n"),
+			progname, fname, strerror(errno));
+		exit(1);
 	}
+
 	isint = geo.logstart > 0;
 
 	/*
@@ -359,7 +352,7 @@ main(int argc, char **argv)
 		}
 	}
 
-	if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY_V1, &ngeo) < 0) {
+	if (xfrog_geometry(ffd, &ngeo) < 0) {
 		fprintf(stderr, _("%s: XFS_IOC_FSGEOMETRY xfsctl failed: %s\n"),
 			progname, strerror(errno));
 		exit(1);
diff --git a/include/xfrog.h b/include/xfrog.h
new file mode 100644
index 00000000..5420b47c
--- /dev/null
+++ b/include/xfrog.h
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2000-2002 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ */
+#ifndef __XFROG_H__
+#define __XFROG_H__
+
+/*
+ * XFS Filesystem Random Online Gluecode
+ * =====================================
+ *
+ * These support functions wrap the more complex xfs ioctls so that xfs
+ * utilities can take advantage of them without having to deal with graceful
+ * degradation in the face of new ioctls.  They will also provide higher level
+ * abstractions when possible.
+ */
+
+struct xfs_fsop_geom;
+int xfrog_geometry(int fd, struct xfs_fsop_geom *fsgeo);
+
+#endif	/* __XFROG_H__ */
diff --git a/io/bmap.c b/io/bmap.c
index d408826a..5f0d12ca 100644
--- a/io/bmap.c
+++ b/io/bmap.c
@@ -9,6 +9,7 @@
 #include "input.h"
 #include "init.h"
 #include "io.h"
+#include "xfrog.h"
 
 static cmdinfo_t bmap_cmd;
 
@@ -105,7 +106,7 @@ bmap_f(
 		bmv_iflags &= ~(BMV_IF_PREALLOC|BMV_IF_NO_DMAPI_READ);
 
 	if (vflag) {
-		c = xfsctl(file->name, file->fd, XFS_IOC_FSGEOMETRY_V1, &fsgeo);
+		c = xfrog_geometry(file->fd, &fsgeo);
 		if (c < 0) {
 			fprintf(stderr,
 				_("%s: can't get geometry [\"%s\"]: %s\n"),
diff --git a/io/fsmap.c b/io/fsmap.c
index 477c36fc..d3ff7dea 100644
--- a/io/fsmap.c
+++ b/io/fsmap.c
@@ -9,6 +9,7 @@
 #include "path.h"
 #include "io.h"
 #include "input.h"
+#include "xfrog.h"
 
 static cmdinfo_t	fsmap_cmd;
 static dev_t		xfs_data_dev;
@@ -447,7 +448,7 @@ fsmap_f(
 	}
 
 	if (vflag) {
-		c = ioctl(file->fd, XFS_IOC_FSGEOMETRY, &fsgeo);
+		c = xfrog_geometry(file->fd, &fsgeo);
 		if (c < 0) {
 			fprintf(stderr,
 				_("%s: can't get geometry [\"%s\"]: %s\n"),
diff --git a/io/open.c b/io/open.c
index c7f5248a..e70c8cb0 100644
--- a/io/open.c
+++ b/io/open.c
@@ -9,6 +9,7 @@
 #include "init.h"
 #include "io.h"
 #include "libxfs.h"
+#include "xfrog.h"
 
 #ifndef __O_TMPFILE
 #if defined __alpha__
@@ -118,7 +119,7 @@ openfile(
 	if (flags & IO_PATH) {
 		/* Can't call ioctl() on O_PATH fds */
 		memset(geom, 0, sizeof(*geom));
-	} else if (xfsctl(path, fd, XFS_IOC_FSGEOMETRY, geom) < 0) {
+	} else if (xfrog_geometry(fd, geom) < 0) {
 		perror("XFS_IOC_FSGEOMETRY");
 		close(fd);
 		return -1;
diff --git a/io/stat.c b/io/stat.c
index 37c0b2e8..f7bbe08b 100644
--- a/io/stat.c
+++ b/io/stat.c
@@ -12,6 +12,7 @@
 #include "io.h"
 #include "statx.h"
 #include "libxfs.h"
+#include "xfrog.h"
 
 #include <fcntl.h>
 
@@ -194,8 +195,8 @@ statfs_f(
 	}
 	if (file->flags & IO_FOREIGN)
 		return 0;
-	if ((xfsctl(file->name, file->fd, XFS_IOC_FSGEOMETRY_V1, &fsgeo)) < 0) {
-		perror("XFS_IOC_FSGEOMETRY_V1");
+	if (xfrog_geometry(file->fd, &fsgeo) < 0) {
+		perror("XFS_IOC_FSGEOMETRY");
 	} else {
 		printf(_("geom.bsize = %u\n"), fsgeo.blocksize);
 		printf(_("geom.agcount = %u\n"), fsgeo.agcount);
diff --git a/libfrog/fsgeom.c b/libfrog/fsgeom.c
index 8879d161..a6dab3a8 100644
--- a/libfrog/fsgeom.c
+++ b/libfrog/fsgeom.c
@@ -4,6 +4,7 @@
  */
 #include "libxfs.h"
 #include "fsgeom.h"
+#include "xfrog.h"
 
 void
 xfs_report_geom(
@@ -67,3 +68,20 @@ xfs_report_geom(
 		geo->rtextsize * geo->blocksize, (unsigned long long)geo->rtblocks,
 			(unsigned long long)geo->rtextents);
 }
+
+/* Try to obtain the xfs geometry. */
+int
+xfrog_geometry(
+	int			fd,
+	struct xfs_fsop_geom	*fsgeo)
+{
+	int			ret;
+
+	memset(fsgeo, 0, sizeof(*fsgeo));
+
+	ret = ioctl(fd, XFS_IOC_FSGEOMETRY, fsgeo);
+	if (!ret)
+		return 0;
+
+	return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, fsgeo);
+}
diff --git a/quota/free.c b/quota/free.c
index 1d13006e..5432b6a8 100644
--- a/quota/free.c
+++ b/quota/free.c
@@ -8,6 +8,7 @@
 #include "command.h"
 #include "init.h"
 #include "quota.h"
+#include "xfrog.h"
 
 static cmdinfo_t free_cmd;
 
@@ -67,9 +68,8 @@ mount_free_space_data(
 	}
 
 	if (!(mount->fs_flags & FS_FOREIGN)) {
-		if ((xfsctl(mount->fs_dir, fd, XFS_IOC_FSGEOMETRY_V1,
-							&fsgeo)) < 0) {
-			perror("XFS_IOC_FSGEOMETRY_V1");
+		if (xfrog_geometry(fd, &fsgeo) < 0) {
+			perror("XFS_IOC_FSGEOMETRY");
 			close(fd);
 			return 0;
 		}
diff --git a/repair/xfs_repair.c b/repair/xfs_repair.c
index ce70e2de..9f4f2611 100644
--- a/repair/xfs_repair.c
+++ b/repair/xfs_repair.c
@@ -22,6 +22,7 @@
 #include "dinode.h"
 #include "slab.h"
 #include "rmap.h"
+#include "xfrog.h"
 
 /*
  * option tables for getsubopt calls
@@ -636,11 +637,11 @@ check_fs_vs_host_sectsize(
 {
 	int	fd;
 	long	old_flags;
-	struct xfs_fsop_geom_v1 geom = { 0 };
+	struct xfs_fsop_geom	geom = { 0 };
 
 	fd = libxfs_device_to_fd(x.ddev);
 
-	if (ioctl(fd, XFS_IOC_FSGEOMETRY_V1, &geom) < 0) {
+	if (xfrog_geometry(fd, &geom) < 0) {
 		do_log(_("Cannot get host filesystem geometry.\n"
 	"Repair may fail if there is a sector size mismatch between\n"
 	"the image and the host filesystem.\n"));
diff --git a/rtcp/Makefile b/rtcp/Makefile
index 808b5378..264b4f27 100644
--- a/rtcp/Makefile
+++ b/rtcp/Makefile
@@ -9,6 +9,9 @@ LTCOMMAND = xfs_rtcp
 CFILES = xfs_rtcp.c
 LLDFLAGS = -static
 
+LLDLIBS = $(LIBFROG)
+LTDEPENDENCIES = $(LIBFROG)
+
 default: depend $(LTCOMMAND)
 
 include $(BUILDRULES)
diff --git a/rtcp/xfs_rtcp.c b/rtcp/xfs_rtcp.c
index 1027c913..0b37ee89 100644
--- a/rtcp/xfs_rtcp.c
+++ b/rtcp/xfs_rtcp.c
@@ -5,6 +5,7 @@
  */
 
 #include "libxfs.h"
+#include "xfrog.h"
 
 int rtcp(char *, char *, int);
 int xfsrtextsize(char *path);
@@ -368,8 +369,8 @@ rtcp( char *source, char *target, int fextsize)
 int
 xfsrtextsize( char *path)
 {
-	int fd, rval, rtextsize;
-	struct xfs_fsop_geom_v1 geo;
+	struct xfs_fsop_geom	geo;
+	int			fd, rval, rtextsize;
 
 	fd = open( path, O_RDONLY );
 	if ( fd < 0 ) {
@@ -377,7 +378,7 @@ xfsrtextsize( char *path)
 			progname, path, strerror(errno));
 		return -1;
 	}
-	rval = xfsctl( path, fd, XFS_IOC_FSGEOMETRY_V1, &geo );
+	rval = xfrog_geometry(fd, &geo);
 	close(fd);
 	if ( rval < 0 )
 		return -1;
diff --git a/scrub/phase1.c b/scrub/phase1.c
index 04a5f4a9..bd7c9fee 100644
--- a/scrub/phase1.c
+++ b/scrub/phase1.c
@@ -26,6 +26,7 @@
 #include "disk.h"
 #include "scrub.h"
 #include "repair.h"
+#include "xfrog.h"
 
 /* Phase 1: Find filesystem geometry (and clean up after) */
 
@@ -129,8 +130,8 @@ _("Does not appear to be an XFS filesystem!"));
 	}
 
 	/* Retrieve XFS geometry. */
-	error = ioctl(ctx->mnt_fd, XFS_IOC_FSGEOMETRY, &ctx->geo);
-	if (error) {
+	error = xfrog_geometry(ctx->mnt_fd, &ctx->geo);
+	if (error < 0) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
diff --git a/spaceman/file.c b/spaceman/file.c
index 7e33e07e..ef627fdb 100644
--- a/spaceman/file.c
+++ b/spaceman/file.c
@@ -6,6 +6,7 @@
  */
 
 #include "libxfs.h"
+#include "xfrog.h"
 #include <sys/mman.h>
 #include "command.h"
 #include "input.h"
@@ -56,7 +57,7 @@ openfile(
 		return -1;
 	}
 
-	if (ioctl(fd, XFS_IOC_FSGEOMETRY, geom) < 0) {
+	if (xfrog_geometry(fd, geom) < 0) {
 		if (errno == ENOTTY)
 			fprintf(stderr,
 _("%s: Not on a mounted XFS filesystem.\n"),
diff --git a/spaceman/info.c b/spaceman/info.c
index 01d0744a..5357b430 100644
--- a/spaceman/info.c
+++ b/spaceman/info.c
@@ -4,6 +4,7 @@
  * Author: Darrick J. Wong <darrick.wong@oracle.com>
  */
 #include "libxfs.h"
+#include "xfrog.h"
 #include "command.h"
 #include "init.h"
 #include "path.h"
@@ -37,24 +38,14 @@ info_f(
 	}
 
 	/* get the current filesystem size & geometry */
-	error = ioctl(file->fd, XFS_IOC_FSGEOMETRY, &geo);
-	if (error) {
-		/*
-		 * OK, new xfsctl barfed - back off and try earlier version
-		 * as we're probably running an older kernel version.
-		 * Only field added in the v2 geometry xfsctl is "logsunit"
-		 * so we'll zero that out for later display (as zero).
-		 */
-		geo.logsunit = 0;
-		error = ioctl(file->fd, XFS_IOC_FSGEOMETRY_V1, &geo);
-		if (error) {
-			fprintf(stderr, _(
-				"%s: cannot determine geometry of filesystem"
-				" mounted at %s: %s\n"),
-				progname, file->name, strerror(errno));
-			exitcode = 1;
-			return 0;
-		}
+	error = xfrog_geometry(file->fd, &geo);
+	if (error < 0) {
+		fprintf(stderr, _(
+			"%s: cannot determine geometry of filesystem"
+			" mounted at %s: %s\n"),
+			progname, file->name, strerror(errno));
+		exitcode = 1;
+		return 0;
 	}
 
 	xfs_report_geom(&geo, file->fs_path.fs_name, file->fs_path.fs_log,

Re: [PATCH 1/6] libfrog: refactor online geometry queries

[Dave Chinner]

On Tue, Aug 20, 2019 at 01:30:38PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Refactor all the open-coded XFS_IOC_FSGEOMETRY queries into a single
> helper that we can use to standardize behaviors across mixed xfslibs
> versions.  This is the prelude to introducing a new FSGEOMETRY version
> in 5.2 and needing to fix the (relatively few) client programs.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> ---
>  Makefile            |    1 +
>  fsr/xfs_fsr.c       |   26 ++++----------------------

Ok.

>  growfs/xfs_growfs.c |   25 +++++++++----------------

Nit.

>  include/xfrog.h     |   22 ++++++++++++++++++++++

Copyright issue.

>  io/bmap.c           |    3 ++-
>  io/fsmap.c          |    3 ++-
>  io/open.c           |    3 ++-
>  io/stat.c           |    5 +++--

ok.

>  libfrog/fsgeom.c    |   18 ++++++++++++++++++

fallback question.

>  quota/free.c        |    6 +++---
>  repair/xfs_repair.c |    5 +++--
>  rtcp/Makefile       |    3 +++

Linker question

>  rtcp/xfs_rtcp.c     |    7 ++++---
>  scrub/phase1.c      |    5 +++--
>  spaceman/file.c     |    3 ++-
>  spaceman/info.c     |   27 +++++++++------------------

OK.


> diff --git a/growfs/xfs_growfs.c b/growfs/xfs_growfs.c
> index 20089d2b..86b1d542 100644
> --- a/growfs/xfs_growfs.c
> +++ b/growfs/xfs_growfs.c
> @@ -7,6 +7,7 @@
>  #include "libxfs.h"
>  #include "path.h"
>  #include "fsgeom.h"
> +#include "xfrog.h"
>  
>  static void
>  usage(void)
> @@ -165,22 +166,14 @@ main(int argc, char **argv)
>  	}
>  
>  	/* get the current filesystem size & geometry */
> -	if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY, &geo) < 0) {
> -		/*
> -		 * OK, new xfsctl barfed - back off and try earlier version
> -		 * as we're probably running an older kernel version.
> -		 * Only field added in the v2 geometry xfsctl is "logsunit"
> -		 * so we'll zero that out for later display (as zero).
> -		 */
> -		geo.logsunit = 0;
> -		if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY_V1, &geo) < 0) {
> -			fprintf(stderr, _(
> -				"%s: cannot determine geometry of filesystem"
> -				" mounted at %s: %s\n"),
> -				progname, fname, strerror(errno));
> -			exit(1);
> -		}
> +	if (xfrog_geometry(ffd, &geo) < 0) {
> +		fprintf(stderr, _(
> +			"%s: cannot determine geometry of filesystem"
> +			" mounted at %s: %s\n"),
> +			progname, fname, strerror(errno));
> +		exit(1);

Can you run the format string into a single line?

> diff --git a/include/xfrog.h b/include/xfrog.h
> new file mode 100644
> index 00000000..5420b47c
> --- /dev/null
> +++ b/include/xfrog.h
> @@ -0,0 +1,22 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (c) 2000-2002 Silicon Graphics, Inc.
> + * All Rights Reserved.
> + */

I don't think that copyright is valid/appropriate for a new header
file with new contents.

...

> @@ -67,3 +68,20 @@ xfs_report_geom(
>  		geo->rtextsize * geo->blocksize, (unsigned long long)geo->rtblocks,
>  			(unsigned long long)geo->rtextents);
>  }
> +
> +/* Try to obtain the xfs geometry. */
> +int
> +xfrog_geometry(
> +	int			fd,
> +	struct xfs_fsop_geom	*fsgeo)
> +{
> +	int			ret;
> +
> +	memset(fsgeo, 0, sizeof(*fsgeo));
> +
> +	ret = ioctl(fd, XFS_IOC_FSGEOMETRY, fsgeo);
> +	if (!ret)
> +		return 0;
> +
> +	return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, fsgeo);
> +}

Should this fall back to XFS_IOC_FSGEOMETRY_V4, then V1 if that
fails (which it shouldn't on any supported kernel)?


> diff --git a/rtcp/Makefile b/rtcp/Makefile
> index 808b5378..264b4f27 100644
> --- a/rtcp/Makefile
> +++ b/rtcp/Makefile
> @@ -9,6 +9,9 @@ LTCOMMAND = xfs_rtcp
>  CFILES = xfs_rtcp.c
>  LLDFLAGS = -static
>  
> +LLDLIBS = $(LIBFROG)
> +LTDEPENDENCIES = $(LIBFROG)

Does this work correctly given LLDFLAGS sets -static and not
-libtools-static-libs?

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

Re: [PATCH 1/6] libfrog: refactor online geometry queries

[Darrick J. Wong]

On Tue, Aug 27, 2019 at 04:30:54PM +1000, Dave Chinner wrote:
> On Tue, Aug 20, 2019 at 01:30:38PM -0700, Darrick J. Wong wrote:
> > From: Darrick J. Wong <darrick.wong@oracle.com>
> > 
> > Refactor all the open-coded XFS_IOC_FSGEOMETRY queries into a single
> > helper that we can use to standardize behaviors across mixed xfslibs
> > versions.  This is the prelude to introducing a new FSGEOMETRY version
> > in 5.2 and needing to fix the (relatively few) client programs.
> > 
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> > ---
> >  Makefile            |    1 +
> >  fsr/xfs_fsr.c       |   26 ++++----------------------
> 
> Ok.
> 
> >  growfs/xfs_growfs.c |   25 +++++++++----------------
> 
> Nit.
> 
> >  include/xfrog.h     |   22 ++++++++++++++++++++++
> 
> Copyright issue.
> 
> >  io/bmap.c           |    3 ++-
> >  io/fsmap.c          |    3 ++-
> >  io/open.c           |    3 ++-
> >  io/stat.c           |    5 +++--
> 
> ok.
> 
> >  libfrog/fsgeom.c    |   18 ++++++++++++++++++
> 
> fallback question.
> 
> >  quota/free.c        |    6 +++---
> >  repair/xfs_repair.c |    5 +++--
> >  rtcp/Makefile       |    3 +++
> 
> Linker question
> 
> >  rtcp/xfs_rtcp.c     |    7 ++++---
> >  scrub/phase1.c      |    5 +++--
> >  spaceman/file.c     |    3 ++-
> >  spaceman/info.c     |   27 +++++++++------------------
> 
> OK.
> 
> 
> > diff --git a/growfs/xfs_growfs.c b/growfs/xfs_growfs.c
> > index 20089d2b..86b1d542 100644
> > --- a/growfs/xfs_growfs.c
> > +++ b/growfs/xfs_growfs.c
> > @@ -7,6 +7,7 @@
> >  #include "libxfs.h"
> >  #include "path.h"
> >  #include "fsgeom.h"
> > +#include "xfrog.h"
> >  
> >  static void
> >  usage(void)
> > @@ -165,22 +166,14 @@ main(int argc, char **argv)
> >  	}
> >  
> >  	/* get the current filesystem size & geometry */
> > -	if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY, &geo) < 0) {
> > -		/*
> > -		 * OK, new xfsctl barfed - back off and try earlier version
> > -		 * as we're probably running an older kernel version.
> > -		 * Only field added in the v2 geometry xfsctl is "logsunit"
> > -		 * so we'll zero that out for later display (as zero).
> > -		 */
> > -		geo.logsunit = 0;
> > -		if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY_V1, &geo) < 0) {
> > -			fprintf(stderr, _(
> > -				"%s: cannot determine geometry of filesystem"
> > -				" mounted at %s: %s\n"),
> > -				progname, fname, strerror(errno));
> > -			exit(1);
> > -		}
> > +	if (xfrog_geometry(ffd, &geo) < 0) {
> > +		fprintf(stderr, _(
> > +			"%s: cannot determine geometry of filesystem"
> > +			" mounted at %s: %s\n"),
> > +			progname, fname, strerror(errno));
> > +		exit(1);
> 
> Can you run the format string into a single line?

ok.

> > diff --git a/include/xfrog.h b/include/xfrog.h
> > new file mode 100644
> > index 00000000..5420b47c
> > --- /dev/null
> > +++ b/include/xfrog.h
> > @@ -0,0 +1,22 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Copyright (c) 2000-2002 Silicon Graphics, Inc.
> > + * All Rights Reserved.
> > + */
> 
> I don't think that copyright is valid/appropriate for a new header
> file with new contents.

Hrmmm ok I'll uhh fix it.

> ...
> 
> > @@ -67,3 +68,20 @@ xfs_report_geom(
> >  		geo->rtextsize * geo->blocksize, (unsigned long long)geo->rtblocks,
> >  			(unsigned long long)geo->rtextents);
> >  }
> > +
> > +/* Try to obtain the xfs geometry. */
> > +int
> > +xfrog_geometry(
> > +	int			fd,
> > +	struct xfs_fsop_geom	*fsgeo)
> > +{
> > +	int			ret;
> > +
> > +	memset(fsgeo, 0, sizeof(*fsgeo));
> > +
> > +	ret = ioctl(fd, XFS_IOC_FSGEOMETRY, fsgeo);
> > +	if (!ret)
> > +		return 0;
> > +
> > +	return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, fsgeo);
> > +}
> 
> Should this fall back to XFS_IOC_FSGEOMETRY_V4, then V1 if that
> fails (which it shouldn't on any supported kernel)?

It should; there's a patch in the next series to do that, but I might as
well merge them (this series was originally targeting 5.1).

> 
> > diff --git a/rtcp/Makefile b/rtcp/Makefile
> > index 808b5378..264b4f27 100644
> > --- a/rtcp/Makefile
> > +++ b/rtcp/Makefile
> > @@ -9,6 +9,9 @@ LTCOMMAND = xfs_rtcp
> >  CFILES = xfs_rtcp.c
> >  LLDFLAGS = -static
> >  
> > +LLDLIBS = $(LIBFROG)
> > +LTDEPENDENCIES = $(LIBFROG)
> 
> Does this work correctly given LLDFLAGS sets -static and not
> -libtools-static-libs?

It seems to, at least on Ubuntu 18.04.

--D

> 
> Cheers,
> 
> Dave.
> -- 
> Dave Chinner
> david@fromorbit.com

[PATCH 2/6] libfrog: introduce xfs_fd to wrap an fd to a file on an xfs filesystem

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Introduce a new "xfs_fd" context structure where we can store a file
descriptor and all the runtime fs context (geometry, which ioctls work,
etc.) that goes with it.  We're going to create wrappers for the
bulkstat and inumbers ioctls in subsequent patches; and when we
introduce the v5 bulkstat/inumbers ioctls we'll need all that context to
downgrade gracefully on old kernels.  Start the transition by adopting
xfs_fd natively in scrub.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 include/xfrog.h    |   20 ++++++++++++++++++++
 libfrog/fsgeom.c   |   25 +++++++++++++++++++++++++
 scrub/fscounters.c |   22 ++++++++++++----------
 scrub/inodes.c     |   10 +++++-----
 scrub/phase1.c     |   41 ++++++++++++++++++++---------------------
 scrub/phase2.c     |    2 +-
 scrub/phase3.c     |    4 ++--
 scrub/phase4.c     |    8 ++++----
 scrub/phase5.c     |    2 +-
 scrub/phase6.c     |    6 +++---
 scrub/phase7.c     |    2 +-
 scrub/repair.c     |    4 ++--
 scrub/scrub.c      |   12 ++++++------
 scrub/spacemap.c   |   12 ++++++------
 scrub/vfs.c        |    2 +-
 scrub/xfs_scrub.h  |    7 ++++---
 16 files changed, 113 insertions(+), 66 deletions(-)


diff --git a/include/xfrog.h b/include/xfrog.h
index 5420b47c..f3808911 100644
--- a/include/xfrog.h
+++ b/include/xfrog.h
@@ -19,4 +19,24 @@
 struct xfs_fsop_geom;
 int xfrog_geometry(int fd, struct xfs_fsop_geom *fsgeo);
 
+/*
+ * Structure for recording whatever observations we want about the level of
+ * xfs runtime support for this fd.  Right now we only store the fd and fs
+ * geometry.
+ */
+struct xfs_fd {
+	/* ioctl file descriptor */
+	int			fd;
+
+	/* filesystem geometry */
+	struct xfs_fsop_geom	fsgeom;
+};
+
+/* Static initializers */
+#define XFS_FD_INIT(_fd)	{ .fd = (_fd), }
+#define XFS_FD_INIT_EMPTY	XFS_FD_INIT(-1)
+
+int xfrog_prepare_geometry(struct xfs_fd *xfd);
+int xfrog_close(struct xfs_fd *xfd);
+
 #endif	/* __XFROG_H__ */
diff --git a/libfrog/fsgeom.c b/libfrog/fsgeom.c
index a6dab3a8..debdece8 100644
--- a/libfrog/fsgeom.c
+++ b/libfrog/fsgeom.c
@@ -85,3 +85,28 @@ xfrog_geometry(
 
 	return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, fsgeo);
 }
+
+/*
+ * Prepare xfrog structure for future ioctl operations by computing the xfs
+ * geometry for @xfd->fd.
+ */
+int
+xfrog_prepare_geometry(
+	struct xfs_fd		*xfd)
+{
+	return xfrog_geometry(xfd->fd, &xfd->fsgeom);
+}
+
+/* Release any resources associated with this xfrog structure. */
+int
+xfrog_close(
+	struct xfs_fd		*xfd)
+{
+	int			ret = 0;
+
+	if (xfd->fd >= 0)
+		ret = close(xfd->fd);
+
+	xfd->fd = -1;
+	return ret;
+}
diff --git a/scrub/fscounters.c b/scrub/fscounters.c
index 9e93e2a6..f18d0e19 100644
--- a/scrub/fscounters.c
+++ b/scrub/fscounters.c
@@ -57,10 +57,10 @@ xfs_count_inodes_range(
 	igrpreq.ocount  = &igrplen;
 
 	igrp_ino = first_ino;
-	error = ioctl(ctx->mnt_fd, XFS_IOC_FSINUMBERS, &igrpreq);
+	error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
 	while (!error && igrplen && inogrp.xi_startino < last_ino) {
 		nr += inogrp.xi_alloccount;
-		error = ioctl(ctx->mnt_fd, XFS_IOC_FSINUMBERS, &igrpreq);
+		error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
 	}
 
 	if (error) {
@@ -113,7 +113,7 @@ xfs_count_all_inodes(
 	int			ret;
 
 	ci = calloc(1, sizeof(struct xfs_count_inodes) +
-			(ctx->geo.agcount * sizeof(uint64_t)));
+			(ctx->mnt.fsgeom.agcount * sizeof(uint64_t)));
 	if (!ci)
 		return false;
 	ci->moveon = true;
@@ -125,7 +125,7 @@ xfs_count_all_inodes(
 		str_info(ctx, ctx->mntpoint, _("Could not create workqueue."));
 		goto out_free;
 	}
-	for (agno = 0; agno < ctx->geo.agcount; agno++) {
+	for (agno = 0; agno < ctx->mnt.fsgeom.agcount; agno++) {
 		ret = workqueue_add(&wq, xfs_count_ag_inodes, agno, ci);
 		if (ret) {
 			moveon = false;
@@ -136,7 +136,7 @@ _("Could not queue AG %u icount work."), agno);
 	}
 	workqueue_destroy(&wq);
 
-	for (agno = 0; agno < ctx->geo.agcount; agno++)
+	for (agno = 0; agno < ctx->mnt.fsgeom.agcount; agno++)
 		*count += ci->counters[agno];
 	moveon = ci->moveon;
 
@@ -162,14 +162,14 @@ xfs_scan_estimate_blocks(
 	int				error;
 
 	/* Grab the fstatvfs counters, since it has to report accurately. */
-	error = fstatvfs(ctx->mnt_fd, &sfs);
+	error = fstatvfs(ctx->mnt.fd, &sfs);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
 
 	/* Fetch the filesystem counters. */
-	error = ioctl(ctx->mnt_fd, XFS_IOC_FSCOUNTS, &fc);
+	error = ioctl(ctx->mnt.fd, XFS_IOC_FSCOUNTS, &fc);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
@@ -179,14 +179,16 @@ xfs_scan_estimate_blocks(
 	 * XFS reserves some blocks to prevent hard ENOSPC, so add those
 	 * blocks back to the free data counts.
 	 */
-	error = ioctl(ctx->mnt_fd, XFS_IOC_GET_RESBLKS, &rb);
+	error = ioctl(ctx->mnt.fd, XFS_IOC_GET_RESBLKS, &rb);
 	if (error)
 		str_errno(ctx, ctx->mntpoint);
 	sfs.f_bfree += rb.resblks_avail;
 
-	*d_blocks = sfs.f_blocks + (ctx->geo.logstart ? ctx->geo.logblocks : 0);
+	*d_blocks = sfs.f_blocks;
+	if (ctx->mnt.fsgeom.logstart > 0)
+		*d_blocks += ctx->mnt.fsgeom.logblocks;
 	*d_bfree = sfs.f_bfree;
-	*r_blocks = ctx->geo.rtblocks;
+	*r_blocks = ctx->mnt.fsgeom.rtblocks;
 	*r_bfree = fc.freertx;
 	*f_files = sfs.f_files;
 	*f_free = sfs.f_ffree;
diff --git a/scrub/inodes.c b/scrub/inodes.c
index 442a5978..08f3d847 100644
--- a/scrub/inodes.c
+++ b/scrub/inodes.c
@@ -72,7 +72,7 @@ xfs_iterate_inodes_range_check(
 		/* Load the one inode. */
 		oneino = inogrp->xi_startino + i;
 		onereq.ubuffer = bs;
-		error = ioctl(ctx->mnt_fd, XFS_IOC_FSBULKSTAT_SINGLE,
+		error = ioctl(ctx->mnt.fd, XFS_IOC_FSBULKSTAT_SINGLE,
 				&onereq);
 		if (error || bs->bs_ino != inogrp->xi_startino + i) {
 			memset(bs, 0, sizeof(struct xfs_bstat));
@@ -134,7 +134,7 @@ xfs_iterate_inodes_range(
 
 	/* Find the inode chunk & alloc mask */
 	igrp_ino = first_ino;
-	error = ioctl(ctx->mnt_fd, XFS_IOC_FSINUMBERS, &igrpreq);
+	error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
 	while (!error && igrplen) {
 		/* Load the inodes. */
 		ino = inogrp.xi_startino - 1;
@@ -145,7 +145,7 @@ xfs_iterate_inodes_range(
 		 */
 		if (inogrp.xi_alloccount == 0)
 			goto igrp_retry;
-		error = ioctl(ctx->mnt_fd, XFS_IOC_FSBULKSTAT, &bulkreq);
+		error = ioctl(ctx->mnt.fd, XFS_IOC_FSBULKSTAT, &bulkreq);
 		if (error)
 			str_info(ctx, descr, "%s", strerror_r(errno,
 						buf, DESCR_BUFSZ));
@@ -190,7 +190,7 @@ _("Changed too many times during scan; giving up."));
 
 		stale_count = 0;
 igrp_retry:
-		error = ioctl(ctx->mnt_fd, XFS_IOC_FSINUMBERS, &igrpreq);
+		error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
 	}
 
 err:
@@ -260,7 +260,7 @@ xfs_scan_all_inodes(
 		return false;
 	}
 
-	for (agno = 0; agno < ctx->geo.agcount; agno++) {
+	for (agno = 0; agno < ctx->mnt.fsgeom.agcount; agno++) {
 		ret = workqueue_add(&wq, xfs_scan_ag_inodes, agno, &si);
 		if (ret) {
 			si.moveon = false;
diff --git a/scrub/phase1.c b/scrub/phase1.c
index bd7c9fee..2b8e69d2 100644
--- a/scrub/phase1.c
+++ b/scrub/phase1.c
@@ -39,7 +39,7 @@ xfs_shutdown_fs(
 
 	flag = XFS_FSOP_GOING_FLAGS_LOGFLUSH;
 	str_info(ctx, ctx->mntpoint, _("Shutting down filesystem!"));
-	if (ioctl(ctx->mnt_fd, XFS_IOC_GOINGDOWN, &flag))
+	if (ioctl(ctx->mnt.fd, XFS_IOC_GOINGDOWN, &flag))
 		str_errno(ctx, ctx->mntpoint);
 }
 
@@ -60,11 +60,9 @@ xfs_cleanup_fs(
 	if (ctx->datadev)
 		disk_close(ctx->datadev);
 	fshandle_destroy();
-	if (ctx->mnt_fd >= 0) {
-		error = close(ctx->mnt_fd);
-		if (error)
-			str_errno(ctx, _("closing mountpoint fd"));
-	}
+	error = xfrog_close(&ctx->mnt);
+	if (error)
+		str_errno(ctx, _("closing mountpoint fd"));
 	fs_table_destroy();
 
 	return true;
@@ -86,8 +84,8 @@ xfs_setup_fs(
 	 * CAP_SYS_ADMIN, which we probably need to do anything fancy
 	 * with the (XFS driver) kernel.
 	 */
-	ctx->mnt_fd = open(ctx->mntpoint, O_RDONLY | O_NOATIME | O_DIRECTORY);
-	if (ctx->mnt_fd < 0) {
+	ctx->mnt.fd = open(ctx->mntpoint, O_RDONLY | O_NOATIME | O_DIRECTORY);
+	if (ctx->mnt.fd < 0) {
 		if (errno == EPERM)
 			str_info(ctx, ctx->mntpoint,
 _("Must be root to run scrub."));
@@ -96,23 +94,23 @@ _("Must be root to run scrub."));
 		return false;
 	}
 
-	error = fstat(ctx->mnt_fd, &ctx->mnt_sb);
+	error = fstat(ctx->mnt.fd, &ctx->mnt_sb);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
-	error = fstatvfs(ctx->mnt_fd, &ctx->mnt_sv);
+	error = fstatvfs(ctx->mnt.fd, &ctx->mnt_sv);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
-	error = fstatfs(ctx->mnt_fd, &ctx->mnt_sf);
+	error = fstatfs(ctx->mnt.fd, &ctx->mnt_sf);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
 
-	if (!platform_test_xfs_fd(ctx->mnt_fd)) {
+	if (!platform_test_xfs_fd(ctx->mnt.fd)) {
 		str_info(ctx, ctx->mntpoint,
 _("Does not appear to be an XFS filesystem!"));
 		return false;
@@ -123,27 +121,28 @@ _("Does not appear to be an XFS filesystem!"));
 	 * This seems to reduce the incidence of stale file handle
 	 * errors when we open things by handle.
 	 */
-	error = syncfs(ctx->mnt_fd);
+	error = syncfs(ctx->mnt.fd);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
 
-	/* Retrieve XFS geometry. */
-	error = xfrog_geometry(ctx->mnt_fd, &ctx->geo);
+	/* Set up xfrog and compute XFS geometry. */
+	error = xfrog_prepare_geometry(&ctx->mnt);
 	if (error < 0) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
 	}
 
-	if (!xfs_action_lists_alloc(ctx->geo.agcount, &ctx->action_lists)) {
+	if (!xfs_action_lists_alloc(ctx->mnt.fsgeom.agcount,
+				&ctx->action_lists)) {
 		str_error(ctx, ctx->mntpoint, _("Not enough memory."));
 		return false;
 	}
 
-	ctx->agblklog = log2_roundup(ctx->geo.agblocks);
-	ctx->blocklog = highbit32(ctx->geo.blocksize);
-	ctx->inodelog = highbit32(ctx->geo.inodesize);
+	ctx->agblklog = log2_roundup(ctx->mnt.fsgeom.agblocks);
+	ctx->blocklog = highbit32(ctx->mnt.fsgeom.blocksize);
+	ctx->inodelog = highbit32(ctx->mnt.fsgeom.inodesize);
 	ctx->inopblog = ctx->blocklog - ctx->inodelog;
 
 	error = path_to_fshandle(ctx->mntpoint, &ctx->fshandle,
@@ -171,12 +170,12 @@ _("Kernel metadata repair facility is not available.  Use -n to scrub."));
 	}
 
 	/* Did we find the log and rt devices, if they're present? */
-	if (ctx->geo.logstart == 0 && ctx->fsinfo.fs_log == NULL) {
+	if (ctx->mnt.fsgeom.logstart == 0 && ctx->fsinfo.fs_log == NULL) {
 		str_info(ctx, ctx->mntpoint,
 _("Unable to find log device path."));
 		return false;
 	}
-	if (ctx->geo.rtblocks && ctx->fsinfo.fs_rt == NULL) {
+	if (ctx->mnt.fsgeom.rtblocks && ctx->fsinfo.fs_rt == NULL) {
 		str_info(ctx, ctx->mntpoint,
 _("Unable to find realtime device path."));
 		return false;
diff --git a/scrub/phase2.c b/scrub/phase2.c
index 653f666c..a80da7fd 100644
--- a/scrub/phase2.c
+++ b/scrub/phase2.c
@@ -141,7 +141,7 @@ xfs_scan_metadata(
 	if (!moveon)
 		goto out;
 
-	for (agno = 0; moveon && agno < ctx->geo.agcount; agno++) {
+	for (agno = 0; moveon && agno < ctx->mnt.fsgeom.agcount; agno++) {
 		ret = workqueue_add(&wq, xfs_scan_ag_metadata, agno, &moveon);
 		if (ret) {
 			moveon = false;
diff --git a/scrub/phase3.c b/scrub/phase3.c
index 4963d675..a42d8213 100644
--- a/scrub/phase3.c
+++ b/scrub/phase3.c
@@ -33,7 +33,7 @@ xfs_scrub_fd(
 	struct xfs_bstat	*bs,
 	struct xfs_action_list	*alist)
 {
-	return fn(ctx, bs->bs_ino, bs->bs_gen, ctx->mnt_fd, alist);
+	return fn(ctx, bs->bs_ino, bs->bs_gen, ctx->mnt.fd, alist);
 }
 
 struct scrub_inode_ctx {
@@ -115,7 +115,7 @@ xfs_scrub_inode(
 	if (S_ISLNK(bstat->bs_mode)) {
 		/* Check symlink contents. */
 		moveon = xfs_scrub_symlink(ctx, bstat->bs_ino,
-				bstat->bs_gen, ctx->mnt_fd, &alist);
+				bstat->bs_gen, ctx->mnt.fd, &alist);
 	} else if (S_ISDIR(bstat->bs_mode)) {
 		/* Check the directory entries. */
 		moveon = xfs_scrub_fd(ctx, xfs_scrub_dir, bstat, &alist);
diff --git a/scrub/phase4.c b/scrub/phase4.c
index 79248326..49f00723 100644
--- a/scrub/phase4.c
+++ b/scrub/phase4.c
@@ -40,7 +40,7 @@ xfs_repair_ag(
 
 	/* Repair anything broken until we fail to make progress. */
 	do {
-		moveon = xfs_action_list_process(ctx, ctx->mnt_fd, alist, flags);
+		moveon = xfs_action_list_process(ctx, ctx->mnt.fd, alist, flags);
 		if (!moveon) {
 			*pmoveon = false;
 			return;
@@ -56,7 +56,7 @@ xfs_repair_ag(
 
 	/* Try once more, but this time complain if we can't fix things. */
 	flags |= ALP_COMPLAIN_IF_UNFIXED;
-	moveon = xfs_action_list_process(ctx, ctx->mnt_fd, alist, flags);
+	moveon = xfs_action_list_process(ctx, ctx->mnt.fd, alist, flags);
 	if (!moveon)
 		*pmoveon = false;
 }
@@ -77,7 +77,7 @@ xfs_process_action_items(
 		str_error(ctx, ctx->mntpoint, _("Could not create workqueue."));
 		return false;
 	}
-	for (agno = 0; agno < ctx->geo.agcount; agno++) {
+	for (agno = 0; agno < ctx->mnt.fsgeom.agcount; agno++) {
 		if (xfs_action_list_length(&ctx->action_lists[agno]) > 0) {
 			ret = workqueue_add(&wq, xfs_repair_ag, agno, &moveon);
 			if (ret) {
@@ -121,7 +121,7 @@ xfs_estimate_repair_work(
 	xfs_agnumber_t		agno;
 	size_t			need_fixing = 0;
 
-	for (agno = 0; agno < ctx->geo.agcount; agno++)
+	for (agno = 0; agno < ctx->mnt.fsgeom.agcount; agno++)
 		need_fixing += xfs_action_list_length(&ctx->action_lists[agno]);
 	need_fixing++;
 	*items = need_fixing;
diff --git a/scrub/phase5.c b/scrub/phase5.c
index 1743119d..748885d4 100644
--- a/scrub/phase5.c
+++ b/scrub/phase5.c
@@ -306,7 +306,7 @@ xfs_scrub_fs_label(
 		return false;
 
 	/* Retrieve label; quietly bail if we don't support that. */
-	error = ioctl(ctx->mnt_fd, FS_IOC_GETFSLABEL, &label);
+	error = ioctl(ctx->mnt.fd, FS_IOC_GETFSLABEL, &label);
 	if (error) {
 		if (errno != EOPNOTSUPP && errno != ENOTTY) {
 			moveon = false;
diff --git a/scrub/phase6.c b/scrub/phase6.c
index 66e6451c..e5a0b3c1 100644
--- a/scrub/phase6.c
+++ b/scrub/phase6.c
@@ -468,7 +468,7 @@ xfs_scan_blocks(
 	}
 
 	vs.rvp_data = read_verify_pool_init(ctx, ctx->datadev,
-			ctx->geo.blocksize, xfs_check_rmap_ioerr,
+			ctx->mnt.fsgeom.blocksize, xfs_check_rmap_ioerr,
 			scrub_nproc(ctx));
 	if (!vs.rvp_data) {
 		str_info(ctx, ctx->mntpoint,
@@ -477,7 +477,7 @@ _("Could not create data device media verifier."));
 	}
 	if (ctx->logdev) {
 		vs.rvp_log = read_verify_pool_init(ctx, ctx->logdev,
-				ctx->geo.blocksize, xfs_check_rmap_ioerr,
+				ctx->mnt.fsgeom.blocksize, xfs_check_rmap_ioerr,
 				scrub_nproc(ctx));
 		if (!vs.rvp_log) {
 			str_info(ctx, ctx->mntpoint,
@@ -487,7 +487,7 @@ _("Could not create data device media verifier."));
 	}
 	if (ctx->rtdev) {
 		vs.rvp_realtime = read_verify_pool_init(ctx, ctx->rtdev,
-				ctx->geo.blocksize, xfs_check_rmap_ioerr,
+				ctx->mnt.fsgeom.blocksize, xfs_check_rmap_ioerr,
 				scrub_nproc(ctx));
 		if (!vs.rvp_realtime) {
 			str_info(ctx, ctx->mntpoint,
diff --git a/scrub/phase7.c b/scrub/phase7.c
index 0c3202e4..13959ca8 100644
--- a/scrub/phase7.c
+++ b/scrub/phase7.c
@@ -111,7 +111,7 @@ xfs_scan_summary(
 	int			error;
 
 	/* Flush everything out to disk before we start counting. */
-	error = syncfs(ctx->mnt_fd);
+	error = syncfs(ctx->mnt.fd);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
diff --git a/scrub/repair.c b/scrub/repair.c
index 4ed3c09a..45450d8c 100644
--- a/scrub/repair.c
+++ b/scrub/repair.c
@@ -262,7 +262,7 @@ xfs_action_list_defer(
 	xfs_agnumber_t			agno,
 	struct xfs_action_list		*alist)
 {
-	ASSERT(agno < ctx->geo.agcount);
+	ASSERT(agno < ctx->mnt.fsgeom.agcount);
 
 	xfs_action_list_splice(&ctx->action_lists[agno], alist);
 }
@@ -276,7 +276,7 @@ xfs_action_list_process_or_defer(
 {
 	bool				moveon;
 
-	moveon = xfs_action_list_process(ctx, ctx->mnt_fd, alist,
+	moveon = xfs_action_list_process(ctx, ctx->mnt.fd, alist,
 			ALP_REPAIR_ONLY | ALP_NOPROGRESS);
 	if (!moveon)
 		return moveon;
diff --git a/scrub/scrub.c b/scrub/scrub.c
index 0f0c9639..136ed529 100644
--- a/scrub/scrub.c
+++ b/scrub/scrub.c
@@ -363,7 +363,7 @@ xfs_scrub_metadata(
 		background_sleep();
 
 		/* Check the item. */
-		fix = xfs_check_metadata(ctx, ctx->mnt_fd, &meta, false);
+		fix = xfs_check_metadata(ctx, ctx->mnt.fd, &meta, false);
 		progress_add(1);
 		switch (fix) {
 		case CHECK_ABORT:
@@ -399,7 +399,7 @@ xfs_scrub_primary_super(
 	enum check_outcome		fix;
 
 	/* Check the item. */
-	fix = xfs_check_metadata(ctx, ctx->mnt_fd, &meta, false);
+	fix = xfs_check_metadata(ctx, ctx->mnt.fd, &meta, false);
 	switch (fix) {
 	case CHECK_ABORT:
 		return false;
@@ -460,7 +460,7 @@ xfs_scrub_estimate_ag_work(
 		switch (sc->type) {
 		case ST_AGHEADER:
 		case ST_PERAG:
-			estimate += ctx->geo.agcount;
+			estimate += ctx->mnt.fsgeom.agcount;
 			break;
 		case ST_FS:
 			estimate++;
@@ -605,9 +605,9 @@ __xfs_scrub_test(
 	if (debug_tweak_on("XFS_SCRUB_NO_KERNEL"))
 		return false;
 	if (debug_tweak_on("XFS_SCRUB_FORCE_REPAIR") && !injected) {
-		inject.fd = ctx->mnt_fd;
+		inject.fd = ctx->mnt.fd;
 		inject.errtag = XFS_ERRTAG_FORCE_SCRUB_REPAIR;
-		error = ioctl(ctx->mnt_fd, XFS_IOC_ERROR_INJECTION, &inject);
+		error = ioctl(ctx->mnt.fd, XFS_IOC_ERROR_INJECTION, &inject);
 		if (error == 0)
 			injected = true;
 	}
@@ -615,7 +615,7 @@ __xfs_scrub_test(
 	meta.sm_type = type;
 	if (repair)
 		meta.sm_flags |= XFS_SCRUB_IFLAG_REPAIR;
-	error = ioctl(ctx->mnt_fd, XFS_IOC_SCRUB_METADATA, &meta);
+	error = ioctl(ctx->mnt.fd, XFS_IOC_SCRUB_METADATA, &meta);
 	if (!error)
 		return true;
 	switch (errno) {
diff --git a/scrub/spacemap.c b/scrub/spacemap.c
index d547a041..c3621a3a 100644
--- a/scrub/spacemap.c
+++ b/scrub/spacemap.c
@@ -56,7 +56,7 @@ xfs_iterate_fsmap(
 	memcpy(head->fmh_keys, keys, sizeof(struct fsmap) * 2);
 	head->fmh_count = FSMAP_NR;
 
-	while ((error = ioctl(ctx->mnt_fd, FS_IOC_GETFSMAP, head)) == 0) {
+	while ((error = ioctl(ctx->mnt.fd, FS_IOC_GETFSMAP, head)) == 0) {
 		for (i = 0, p = head->fmh_recs;
 		     i < head->fmh_entries;
 		     i++, p++) {
@@ -107,8 +107,8 @@ xfs_scan_ag_blocks(
 	off64_t			bperag;
 	bool			moveon;
 
-	bperag = (off64_t)ctx->geo.agblocks *
-		 (off64_t)ctx->geo.blocksize;
+	bperag = (off64_t)ctx->mnt.fsgeom.agblocks *
+		 (off64_t)ctx->mnt.fsgeom.blocksize;
 
 	snprintf(descr, DESCR_BUFSZ, _("dev %d:%d AG %u fsmap"),
 				major(ctx->fsinfo.fs_datadev),
@@ -205,7 +205,7 @@ xfs_scan_all_spacemaps(
 	}
 	if (ctx->fsinfo.fs_rt) {
 		ret = workqueue_add(&wq, xfs_scan_rt_blocks,
-				ctx->geo.agcount + 1, &sbx);
+				ctx->mnt.fsgeom.agcount + 1, &sbx);
 		if (ret) {
 			sbx.moveon = false;
 			str_info(ctx, ctx->mntpoint,
@@ -215,7 +215,7 @@ _("Could not queue rtdev fsmap work."));
 	}
 	if (ctx->fsinfo.fs_log) {
 		ret = workqueue_add(&wq, xfs_scan_log_blocks,
-				ctx->geo.agcount + 2, &sbx);
+				ctx->mnt.fsgeom.agcount + 2, &sbx);
 		if (ret) {
 			sbx.moveon = false;
 			str_info(ctx, ctx->mntpoint,
@@ -223,7 +223,7 @@ _("Could not queue logdev fsmap work."));
 			goto out;
 		}
 	}
-	for (agno = 0; agno < ctx->geo.agcount; agno++) {
+	for (agno = 0; agno < ctx->mnt.fsgeom.agcount; agno++) {
 		ret = workqueue_add(&wq, xfs_scan_ag_blocks, agno, &sbx);
 		if (ret) {
 			sbx.moveon = false;
diff --git a/scrub/vfs.c b/scrub/vfs.c
index 8bcc4e79..7b0b5bcd 100644
--- a/scrub/vfs.c
+++ b/scrub/vfs.c
@@ -232,7 +232,7 @@ fstrim(
 	int			error;
 
 	range.len = ULLONG_MAX;
-	error = ioctl(ctx->mnt_fd, FITRIM, &range);
+	error = ioctl(ctx->mnt.fd, FITRIM, &range);
 	if (error && errno != EOPNOTSUPP && errno != ENOTTY)
 		perror(_("fstrim"));
 }
diff --git a/scrub/xfs_scrub.h b/scrub/xfs_scrub.h
index a459e4b5..28eae6fe 100644
--- a/scrub/xfs_scrub.h
+++ b/scrub/xfs_scrub.h
@@ -6,6 +6,8 @@
 #ifndef XFS_SCRUB_XFS_SCRUB_H_
 #define XFS_SCRUB_XFS_SCRUB_H_
 
+#include <xfrog.h>
+
 extern char *progname;
 
 #define _PATH_PROC_MOUNTS	"/proc/mounts"
@@ -53,14 +55,13 @@ struct scrub_ctx {
 	/* How does the user want us to react to errors? */
 	enum error_action	error_action;
 
-	/* fd to filesystem mount point */
-	int			mnt_fd;
+	/* xfrog context for the mount point */
+	struct xfs_fd		mnt;
 
 	/* Number of threads for metadata scrubbing */
 	unsigned int		nr_io_threads;
 
 	/* XFS specific geometry */
-	struct xfs_fsop_geom	geo;
 	struct fs_path		fsinfo;
 	unsigned int		agblklog;
 	unsigned int		blocklog;

Re: [PATCH 2/6] libfrog: introduce xfs_fd to wrap an fd to a file on an xfs filesystem

[Dave Chinner]

On Tue, Aug 20, 2019 at 01:30:44PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Introduce a new "xfs_fd" context structure where we can store a file
> descriptor and all the runtime fs context (geometry, which ioctls work,
> etc.) that goes with it.  We're going to create wrappers for the
> bulkstat and inumbers ioctls in subsequent patches; and when we
> introduce the v5 bulkstat/inumbers ioctls we'll need all that context to
> downgrade gracefully on old kernels.  Start the transition by adopting
> xfs_fd natively in scrub.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> ---
>  include/xfrog.h    |   20 ++++++++++++++++++++

naming.

>  libfrog/fsgeom.c   |   25 +++++++++++++++++++++++++
>  scrub/fscounters.c |   22 ++++++++++++----------
>  scrub/inodes.c     |   10 +++++-----
>  scrub/phase1.c     |   41 ++++++++++++++++++++---------------------
>  scrub/phase2.c     |    2 +-
>  scrub/phase3.c     |    4 ++--
>  scrub/phase4.c     |    8 ++++----
>  scrub/phase5.c     |    2 +-
>  scrub/phase6.c     |    6 +++---
>  scrub/phase7.c     |    2 +-
>  scrub/repair.c     |    4 ++--
>  scrub/scrub.c      |   12 ++++++------
>  scrub/spacemap.c   |   12 ++++++------
>  scrub/vfs.c        |    2 +-
>  scrub/xfs_scrub.h  |    7 ++++---
>  16 files changed, 113 insertions(+), 66 deletions(-)
> 
> 
> diff --git a/include/xfrog.h b/include/xfrog.h
> index 5420b47c..f3808911 100644
> --- a/include/xfrog.h
> +++ b/include/xfrog.h
> @@ -19,4 +19,24 @@
>  struct xfs_fsop_geom;
>  int xfrog_geometry(int fd, struct xfs_fsop_geom *fsgeo);
>  
> +/*
> + * Structure for recording whatever observations we want about the level of
> + * xfs runtime support for this fd.  Right now we only store the fd and fs
> + * geometry.
> + */
> +struct xfs_fd {
> +	/* ioctl file descriptor */
> +	int			fd;
> +
> +	/* filesystem geometry */
> +	struct xfs_fsop_geom	fsgeom;
> +};
> +
> +/* Static initializers */
> +#define XFS_FD_INIT(_fd)	{ .fd = (_fd), }
> +#define XFS_FD_INIT_EMPTY	XFS_FD_INIT(-1)
> +
> +int xfrog_prepare_geometry(struct xfs_fd *xfd);
> +int xfrog_close(struct xfs_fd *xfd);

I'd much prefer to see these named xfd_prepare_geometry() and
xfd_close()...

>  	ci = calloc(1, sizeof(struct xfs_count_inodes) +
> -			(ctx->geo.agcount * sizeof(uint64_t)));
> +			(ctx->mnt.fsgeom.agcount * sizeof(uint64_t)));

This gets a bit verbose, but.... <shrug> .... doesn't make that much
of a mess...

Ok, apart from the naming thing, this looks ok.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

Re: [PATCH 2/6] libfrog: introduce xfs_fd to wrap an fd to a file on an xfs filesystem

[Darrick J. Wong]

On Tue, Aug 27, 2019 at 04:41:57PM +1000, Dave Chinner wrote:
> On Tue, Aug 20, 2019 at 01:30:44PM -0700, Darrick J. Wong wrote:
> > From: Darrick J. Wong <darrick.wong@oracle.com>
> > 
> > Introduce a new "xfs_fd" context structure where we can store a file
> > descriptor and all the runtime fs context (geometry, which ioctls work,
> > etc.) that goes with it.  We're going to create wrappers for the
> > bulkstat and inumbers ioctls in subsequent patches; and when we
> > introduce the v5 bulkstat/inumbers ioctls we'll need all that context to
> > downgrade gracefully on old kernels.  Start the transition by adopting
> > xfs_fd natively in scrub.
> > 
> > Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> > ---
> >  include/xfrog.h    |   20 ++++++++++++++++++++
> 
> naming.
> 
> >  libfrog/fsgeom.c   |   25 +++++++++++++++++++++++++
> >  scrub/fscounters.c |   22 ++++++++++++----------
> >  scrub/inodes.c     |   10 +++++-----
> >  scrub/phase1.c     |   41 ++++++++++++++++++++---------------------
> >  scrub/phase2.c     |    2 +-
> >  scrub/phase3.c     |    4 ++--
> >  scrub/phase4.c     |    8 ++++----
> >  scrub/phase5.c     |    2 +-
> >  scrub/phase6.c     |    6 +++---
> >  scrub/phase7.c     |    2 +-
> >  scrub/repair.c     |    4 ++--
> >  scrub/scrub.c      |   12 ++++++------
> >  scrub/spacemap.c   |   12 ++++++------
> >  scrub/vfs.c        |    2 +-
> >  scrub/xfs_scrub.h  |    7 ++++---
> >  16 files changed, 113 insertions(+), 66 deletions(-)
> > 
> > 
> > diff --git a/include/xfrog.h b/include/xfrog.h
> > index 5420b47c..f3808911 100644
> > --- a/include/xfrog.h
> > +++ b/include/xfrog.h
> > @@ -19,4 +19,24 @@
> >  struct xfs_fsop_geom;
> >  int xfrog_geometry(int fd, struct xfs_fsop_geom *fsgeo);
> >  
> > +/*
> > + * Structure for recording whatever observations we want about the level of
> > + * xfs runtime support for this fd.  Right now we only store the fd and fs
> > + * geometry.
> > + */
> > +struct xfs_fd {
> > +	/* ioctl file descriptor */
> > +	int			fd;
> > +
> > +	/* filesystem geometry */
> > +	struct xfs_fsop_geom	fsgeom;
> > +};
> > +
> > +/* Static initializers */
> > +#define XFS_FD_INIT(_fd)	{ .fd = (_fd), }
> > +#define XFS_FD_INIT_EMPTY	XFS_FD_INIT(-1)
> > +
> > +int xfrog_prepare_geometry(struct xfs_fd *xfd);
> > +int xfrog_close(struct xfs_fd *xfd);
> 
> I'd much prefer to see these named xfd_prepare_geometry() and
> xfd_close()...

Done.

> >  	ci = calloc(1, sizeof(struct xfs_count_inodes) +
> > -			(ctx->geo.agcount * sizeof(uint64_t)));
> > +			(ctx->mnt.fsgeom.agcount * sizeof(uint64_t)));
> 
> This gets a bit verbose, but.... <shrug> .... doesn't make that much
> of a mess...
> 
> Ok, apart from the naming thing, this looks ok.

Ok.

--D

> Cheers,
> 
> Dave.
> -- 
> Dave Chinner
> david@fromorbit.com

[PATCH 3/6] libfrog: store more inode and block geometry in struct xfs_fd

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Move the extra AG geometry fields out of scrub and into the libfrog code
so that we can consolidate the geoemtry code in one place.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 include/xfrog.h    |   12 ++++++++++++
 libfrog/fsgeom.c   |   13 ++++++++++++-
 scrub/fscounters.c |    4 ++--
 scrub/inodes.c     |    4 ++--
 scrub/phase1.c     |    5 -----
 scrub/phase3.c     |    6 +++---
 scrub/phase5.c     |    4 ++--
 scrub/phase6.c     |    2 +-
 scrub/phase7.c     |    6 +++---
 scrub/xfs_scrub.h  |    4 ----
 10 files changed, 37 insertions(+), 23 deletions(-)


diff --git a/include/xfrog.h b/include/xfrog.h
index f3808911..d11317f7 100644
--- a/include/xfrog.h
+++ b/include/xfrog.h
@@ -30,6 +30,18 @@ struct xfs_fd {
 
 	/* filesystem geometry */
 	struct xfs_fsop_geom	fsgeom;
+
+	/* log2 of sb_agblocks (rounded up) */
+	unsigned int		agblklog;
+
+	/* log2 of sb_blocksize */
+	unsigned int		blocklog;
+
+	/* log2 of sb_inodesize */
+	unsigned int		inodelog;
+
+	/* log2 of sb_inopblock */
+	unsigned int		inopblog;
 };
 
 /* Static initializers */
diff --git a/libfrog/fsgeom.c b/libfrog/fsgeom.c
index debdece8..bc6158fb 100644
--- a/libfrog/fsgeom.c
+++ b/libfrog/fsgeom.c
@@ -5,6 +5,7 @@
 #include "libxfs.h"
 #include "fsgeom.h"
 #include "xfrog.h"
+#include "libfrog.h"
 
 void
 xfs_report_geom(
@@ -94,7 +95,17 @@ int
 xfrog_prepare_geometry(
 	struct xfs_fd		*xfd)
 {
-	return xfrog_geometry(xfd->fd, &xfd->fsgeom);
+	int			ret;
+
+	ret = xfrog_geometry(xfd->fd, &xfd->fsgeom);
+	if (ret)
+		return ret;
+
+	xfd->agblklog = log2_roundup(xfd->fsgeom.agblocks);
+	xfd->blocklog = highbit32(xfd->fsgeom.blocksize);
+	xfd->inodelog = highbit32(xfd->fsgeom.inodesize);
+	xfd->inopblog = xfd->blocklog - xfd->inodelog;
+	return 0;
 }
 
 /* Release any resources associated with this xfrog structure. */
diff --git a/scrub/fscounters.c b/scrub/fscounters.c
index f18d0e19..ac898764 100644
--- a/scrub/fscounters.c
+++ b/scrub/fscounters.c
@@ -91,8 +91,8 @@ xfs_count_ag_inodes(
 				minor(ctx->fsinfo.fs_datadev),
 				agno);
 
-	ag_ino = (__u64)agno << (ctx->inopblog + ctx->agblklog);
-	next_ag_ino = (__u64)(agno + 1) << (ctx->inopblog + ctx->agblklog);
+	ag_ino = (__u64)agno << (ctx->mnt.inopblog + ctx->mnt.agblklog);
+	next_ag_ino = (__u64)(agno + 1) << (ctx->mnt.inopblog + ctx->mnt.agblklog);
 
 	moveon = xfs_count_inodes_range(ctx, descr, ag_ino, next_ag_ino - 1,
 			&ci->counters[agno]);
diff --git a/scrub/inodes.c b/scrub/inodes.c
index 08f3d847..873ad425 100644
--- a/scrub/inodes.c
+++ b/scrub/inodes.c
@@ -228,8 +228,8 @@ xfs_scan_ag_inodes(
 				minor(ctx->fsinfo.fs_datadev),
 				agno);
 
-	ag_ino = (__u64)agno << (ctx->inopblog + ctx->agblklog);
-	next_ag_ino = (__u64)(agno + 1) << (ctx->inopblog + ctx->agblklog);
+	ag_ino = (__u64)agno << (ctx->mnt.inopblog + ctx->mnt.agblklog);
+	next_ag_ino = (__u64)(agno + 1) << (ctx->mnt.inopblog + ctx->mnt.agblklog);
 
 	moveon = xfs_iterate_inodes_range(ctx, descr, ctx->fshandle, ag_ino,
 			next_ag_ino - 1, si->fn, si->arg);
diff --git a/scrub/phase1.c b/scrub/phase1.c
index 2b8e69d2..d052a21c 100644
--- a/scrub/phase1.c
+++ b/scrub/phase1.c
@@ -140,11 +140,6 @@ _("Does not appear to be an XFS filesystem!"));
 		return false;
 	}
 
-	ctx->agblklog = log2_roundup(ctx->mnt.fsgeom.agblocks);
-	ctx->blocklog = highbit32(ctx->mnt.fsgeom.blocksize);
-	ctx->inodelog = highbit32(ctx->mnt.fsgeom.inodesize);
-	ctx->inopblog = ctx->blocklog - ctx->inodelog;
-
 	error = path_to_fshandle(ctx->mntpoint, &ctx->fshandle,
 			&ctx->fshandle_len);
 	if (error) {
diff --git a/scrub/phase3.c b/scrub/phase3.c
index a42d8213..579e08c3 100644
--- a/scrub/phase3.c
+++ b/scrub/phase3.c
@@ -52,8 +52,8 @@ xfs_scrub_inode_vfs_error(
 	xfs_agino_t		agino;
 	int			old_errno = errno;
 
-	agno = bstat->bs_ino / (1ULL << (ctx->inopblog + ctx->agblklog));
-	agino = bstat->bs_ino % (1ULL << (ctx->inopblog + ctx->agblklog));
+	agno = bstat->bs_ino / (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
+	agino = bstat->bs_ino % (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
 	snprintf(descr, DESCR_BUFSZ, _("inode %"PRIu64" (%u/%u)"),
 			(uint64_t)bstat->bs_ino, agno, agino);
 	errno = old_errno;
@@ -77,7 +77,7 @@ xfs_scrub_inode(
 	int			error;
 
 	xfs_action_list_init(&alist);
-	agno = bstat->bs_ino / (1ULL << (ctx->inopblog + ctx->agblklog));
+	agno = bstat->bs_ino / (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
 	background_sleep();
 
 	/* Try to open the inode to pin it. */
diff --git a/scrub/phase5.c b/scrub/phase5.c
index 748885d4..36ec27b3 100644
--- a/scrub/phase5.c
+++ b/scrub/phase5.c
@@ -239,8 +239,8 @@ xfs_scrub_connections(
 	int			fd = -1;
 	int			error;
 
-	agno = bstat->bs_ino / (1ULL << (ctx->inopblog + ctx->agblklog));
-	agino = bstat->bs_ino % (1ULL << (ctx->inopblog + ctx->agblklog));
+	agno = bstat->bs_ino / (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
+	agino = bstat->bs_ino % (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
 	snprintf(descr, DESCR_BUFSZ, _("inode %"PRIu64" (%u/%u)"),
 			(uint64_t)bstat->bs_ino, agno, agino);
 	background_sleep();
diff --git a/scrub/phase6.c b/scrub/phase6.c
index e5a0b3c1..48971270 100644
--- a/scrub/phase6.c
+++ b/scrub/phase6.c
@@ -547,7 +547,7 @@ xfs_estimate_verify_work(
 	if (!moveon)
 		return moveon;
 
-	*items = ((d_blocks - d_bfree) + (r_blocks - r_bfree)) << ctx->blocklog;
+	*items = ((d_blocks - d_bfree) + (r_blocks - r_bfree)) << ctx->mnt.blocklog;
 	*nr_threads = disk_heads(ctx->datadev);
 	*rshift = 20;
 	return moveon;
diff --git a/scrub/phase7.c b/scrub/phase7.c
index 13959ca8..41a77356 100644
--- a/scrub/phase7.c
+++ b/scrub/phase7.c
@@ -148,11 +148,11 @@ xfs_scan_summary(
 	 * filesystem treats them as "free", but since we scanned
 	 * them, we'll consider them used.
 	 */
-	d_bfree -= totalcount.agbytes >> ctx->blocklog;
+	d_bfree -= totalcount.agbytes >> ctx->mnt.blocklog;
 
 	/* Report on what we found. */
-	used_data = (d_blocks - d_bfree) << ctx->blocklog;
-	used_rt = (r_blocks - r_bfree) << ctx->blocklog;
+	used_data = (d_blocks - d_bfree) << ctx->mnt.blocklog;
+	used_rt = (r_blocks - r_bfree) << ctx->mnt.blocklog;
 	used_files = f_files - f_free;
 	stat_data = totalcount.dbytes;
 	stat_rt = totalcount.rbytes;
diff --git a/scrub/xfs_scrub.h b/scrub/xfs_scrub.h
index 28eae6fe..fb34c587 100644
--- a/scrub/xfs_scrub.h
+++ b/scrub/xfs_scrub.h
@@ -63,10 +63,6 @@ struct scrub_ctx {
 
 	/* XFS specific geometry */
 	struct fs_path		fsinfo;
-	unsigned int		agblklog;
-	unsigned int		blocklog;
-	unsigned int		inodelog;
-	unsigned int		inopblog;
 	void			*fshandle;
 	size_t			fshandle_len;
 

Re: [PATCH 3/6] libfrog: store more inode and block geometry in struct xfs_fd

[Dave Chinner]

On Tue, Aug 20, 2019 at 01:30:51PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Move the extra AG geometry fields out of scrub and into the libfrog code
> so that we can consolidate the geoemtry code in one place.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>

Looks OK. Plenty of scope for ino_to_agno/ino_to_agbno scope helpers
here....

Reviewed-by: Dave Chinner <dchinner@redhat.com>

-- 
Dave Chinner
david@fromorbit.com

[PATCH 4/6] libfrog: create online fs geometry converters

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Create helper functions to perform unit conversions against a runtime
filesystem, then remove the open-coded versions in scrub.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 include/xfrog.h    |   49 +++++++++++++++++++++++++++++++++++++++++++++++++
 libfrog/fsgeom.c   |    1 +
 scrub/fscounters.c |    4 ++--
 scrub/inodes.c     |    4 ++--
 scrub/phase3.c     |    6 +++---
 scrub/phase5.c     |    4 ++--
 scrub/phase6.c     |    3 ++-
 scrub/phase7.c     |    6 +++---
 8 files changed, 64 insertions(+), 13 deletions(-)


diff --git a/include/xfrog.h b/include/xfrog.h
index d11317f7..091edebb 100644
--- a/include/xfrog.h
+++ b/include/xfrog.h
@@ -42,6 +42,9 @@ struct xfs_fd {
 
 	/* log2 of sb_inopblock */
 	unsigned int		inopblog;
+
+	/* bits for agino in inum */
+	unsigned int		aginolog;
 };
 
 /* Static initializers */
@@ -51,4 +54,50 @@ struct xfs_fd {
 int xfrog_prepare_geometry(struct xfs_fd *xfd);
 int xfrog_close(struct xfs_fd *xfd);
 
+/* Convert AG number and AG inode number into fs inode number. */
+static inline uint64_t
+xfrog_agino_to_ino(
+	const struct xfs_fd	*xfd,
+	uint32_t		agno,
+	uint32_t		agino)
+{
+	return ((uint64_t)agno << xfd->aginolog) + agino;
+}
+
+/* Convert fs inode number into AG number. */
+static inline uint32_t
+xfrog_ino_to_agno(
+	const struct xfs_fd	*xfd,
+	uint64_t		ino)
+{
+	return ino >> xfd->aginolog;
+}
+
+/* Convert fs inode number into AG inode number. */
+static inline uint32_t
+xfrog_ino_to_agino(
+	const struct xfs_fd	*xfd,
+	uint64_t		ino)
+{
+	return ino & ((1ULL << xfd->aginolog) - 1);
+}
+
+/* Convert fs block number into bytes */
+static inline uint64_t
+xfrog_fsb_to_b(
+	const struct xfs_fd	*xfd,
+	uint64_t		fsb)
+{
+	return fsb << xfd->blocklog;
+}
+
+/* Convert bytes into (rounded down) fs block number */
+static inline uint64_t
+xfrog_b_to_fsbt(
+	const struct xfs_fd	*xfd,
+	uint64_t		bytes)
+{
+	return bytes >> xfd->blocklog;
+}
+
 #endif	/* __XFROG_H__ */
diff --git a/libfrog/fsgeom.c b/libfrog/fsgeom.c
index bc6158fb..a3b748f8 100644
--- a/libfrog/fsgeom.c
+++ b/libfrog/fsgeom.c
@@ -105,6 +105,7 @@ xfrog_prepare_geometry(
 	xfd->blocklog = highbit32(xfd->fsgeom.blocksize);
 	xfd->inodelog = highbit32(xfd->fsgeom.inodesize);
 	xfd->inopblog = xfd->blocklog - xfd->inodelog;
+	xfd->aginolog = xfd->agblklog + xfd->inopblog;
 	return 0;
 }
 
diff --git a/scrub/fscounters.c b/scrub/fscounters.c
index ac898764..adb79b50 100644
--- a/scrub/fscounters.c
+++ b/scrub/fscounters.c
@@ -91,8 +91,8 @@ xfs_count_ag_inodes(
 				minor(ctx->fsinfo.fs_datadev),
 				agno);
 
-	ag_ino = (__u64)agno << (ctx->mnt.inopblog + ctx->mnt.agblklog);
-	next_ag_ino = (__u64)(agno + 1) << (ctx->mnt.inopblog + ctx->mnt.agblklog);
+	ag_ino = xfrog_agino_to_ino(&ctx->mnt, agno, 0);
+	next_ag_ino = xfrog_agino_to_ino(&ctx->mnt, agno + 1, 0);
 
 	moveon = xfs_count_inodes_range(ctx, descr, ag_ino, next_ag_ino - 1,
 			&ci->counters[agno]);
diff --git a/scrub/inodes.c b/scrub/inodes.c
index 873ad425..a9000218 100644
--- a/scrub/inodes.c
+++ b/scrub/inodes.c
@@ -228,8 +228,8 @@ xfs_scan_ag_inodes(
 				minor(ctx->fsinfo.fs_datadev),
 				agno);
 
-	ag_ino = (__u64)agno << (ctx->mnt.inopblog + ctx->mnt.agblklog);
-	next_ag_ino = (__u64)(agno + 1) << (ctx->mnt.inopblog + ctx->mnt.agblklog);
+	ag_ino = xfrog_agino_to_ino(&ctx->mnt, agno, 0);
+	next_ag_ino = xfrog_agino_to_ino(&ctx->mnt, agno + 1, 0);
 
 	moveon = xfs_iterate_inodes_range(ctx, descr, ctx->fshandle, ag_ino,
 			next_ag_ino - 1, si->fn, si->arg);
diff --git a/scrub/phase3.c b/scrub/phase3.c
index 579e08c3..def9a0de 100644
--- a/scrub/phase3.c
+++ b/scrub/phase3.c
@@ -52,8 +52,8 @@ xfs_scrub_inode_vfs_error(
 	xfs_agino_t		agino;
 	int			old_errno = errno;
 
-	agno = bstat->bs_ino / (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
-	agino = bstat->bs_ino % (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
+	agno = xfrog_ino_to_agno(&ctx->mnt, bstat->bs_ino);
+	agino = xfrog_ino_to_agino(&ctx->mnt, bstat->bs_ino);
 	snprintf(descr, DESCR_BUFSZ, _("inode %"PRIu64" (%u/%u)"),
 			(uint64_t)bstat->bs_ino, agno, agino);
 	errno = old_errno;
@@ -77,7 +77,7 @@ xfs_scrub_inode(
 	int			error;
 
 	xfs_action_list_init(&alist);
-	agno = bstat->bs_ino / (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
+	agno = xfrog_ino_to_agno(&ctx->mnt, bstat->bs_ino);
 	background_sleep();
 
 	/* Try to open the inode to pin it. */
diff --git a/scrub/phase5.c b/scrub/phase5.c
index 36ec27b3..2189c9e4 100644
--- a/scrub/phase5.c
+++ b/scrub/phase5.c
@@ -239,8 +239,8 @@ xfs_scrub_connections(
 	int			fd = -1;
 	int			error;
 
-	agno = bstat->bs_ino / (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
-	agino = bstat->bs_ino % (1ULL << (ctx->mnt.inopblog + ctx->mnt.agblklog));
+	agno = xfrog_ino_to_agno(&ctx->mnt, bstat->bs_ino);
+	agino = xfrog_ino_to_agino(&ctx->mnt, bstat->bs_ino);
 	snprintf(descr, DESCR_BUFSZ, _("inode %"PRIu64" (%u/%u)"),
 			(uint64_t)bstat->bs_ino, agno, agino);
 	background_sleep();
diff --git a/scrub/phase6.c b/scrub/phase6.c
index 48971270..630d15b0 100644
--- a/scrub/phase6.c
+++ b/scrub/phase6.c
@@ -547,7 +547,8 @@ xfs_estimate_verify_work(
 	if (!moveon)
 		return moveon;
 
-	*items = ((d_blocks - d_bfree) + (r_blocks - r_bfree)) << ctx->mnt.blocklog;
+	*items = xfrog_fsb_to_b(&ctx->mnt,
+			(d_blocks - d_bfree) + (r_blocks - r_bfree));
 	*nr_threads = disk_heads(ctx->datadev);
 	*rshift = 20;
 	return moveon;
diff --git a/scrub/phase7.c b/scrub/phase7.c
index 41a77356..1c459dfc 100644
--- a/scrub/phase7.c
+++ b/scrub/phase7.c
@@ -148,11 +148,11 @@ xfs_scan_summary(
 	 * filesystem treats them as "free", but since we scanned
 	 * them, we'll consider them used.
 	 */
-	d_bfree -= totalcount.agbytes >> ctx->mnt.blocklog;
+	d_bfree -= xfrog_b_to_fsbt(&ctx->mnt, totalcount.agbytes);
 
 	/* Report on what we found. */
-	used_data = (d_blocks - d_bfree) << ctx->mnt.blocklog;
-	used_rt = (r_blocks - r_bfree) << ctx->mnt.blocklog;
+	used_data = xfrog_fsb_to_b(&ctx->mnt, d_blocks - d_bfree);
+	used_rt = xfrog_fsb_to_b(&ctx->mnt, r_blocks - r_bfree);
 	used_files = f_files - f_free;
 	stat_data = totalcount.dbytes;
 	stat_rt = totalcount.rbytes;

Re: [PATCH 4/6] libfrog: create online fs geometry converters

[Dave Chinner]

On Tue, Aug 20, 2019 at 01:30:57PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Create helper functions to perform unit conversions against a runtime
> filesystem, then remove the open-coded versions in scrub.

.... and there they are...

> +/* Convert fs block number into bytes */
> +static inline uint64_t
> +xfrog_fsb_to_b(
> +	const struct xfs_fd	*xfd,
> +	uint64_t		fsb)
> +{
> +	return fsb << xfd->blocklog;
> +}

FWIW, this is for converting linear offsets in fsb /units/, not the
sparse fsbno (= agno | agbno) to bytes. I've always found it a bit
nasty that this distinction is not clearly made in the core FSB
conversion macros.

perhaps off_fsb_to_b?

> +/* Convert bytes into (rounded down) fs block number */
> +static inline uint64_t
> +xfrog_b_to_fsbt(
> +	const struct xfs_fd	*xfd,
> +	uint64_t		bytes)
> +{
> +	return bytes >> xfd->blocklog;
> +}

Ditto.

Otherwise looks ok.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

Re: [PATCH 4/6] libfrog: create online fs geometry converters

[Darrick J. Wong]

On Tue, Aug 27, 2019 at 05:11:06PM +1000, Dave Chinner wrote:
> On Tue, Aug 20, 2019 at 01:30:57PM -0700, Darrick J. Wong wrote:
> > From: Darrick J. Wong <darrick.wong@oracle.com>
> > 
> > Create helper functions to perform unit conversions against a runtime
> > filesystem, then remove the open-coded versions in scrub.
> 
> .... and there they are...
> 
> > +/* Convert fs block number into bytes */
> > +static inline uint64_t
> > +xfrog_fsb_to_b(
> > +	const struct xfs_fd	*xfd,
> > +	uint64_t		fsb)
> > +{
> > +	return fsb << xfd->blocklog;
> > +}
> 
> FWIW, this is for converting linear offsets in fsb /units/, not the
> sparse fsbno (= agno | agbno) to bytes. I've always found it a bit
> nasty that this distinction is not clearly made in the core FSB
> conversion macros.
> 
> perhaps off_fsb_to_b?

Ok to both.

--D

> > +/* Convert bytes into (rounded down) fs block number */
> > +static inline uint64_t
> > +xfrog_b_to_fsbt(
> > +	const struct xfs_fd	*xfd,
> > +	uint64_t		bytes)
> > +{
> > +	return bytes >> xfd->blocklog;
> > +}
> 
> Ditto.
> 
> Otherwise looks ok.
> 
> Cheers,
> 
> Dave.
> -- 
> Dave Chinner
> david@fromorbit.com

[PATCH 5/6] libfrog: refactor open-coded bulkstat calls

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Refactor the BULKSTAT_SINGLE and BULKSTAT ioctl callsites into helper
functions.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 fsr/xfs_fsr.c      |  104 ++++++++++++++++++++++------------------------------
 include/xfrog.h    |    7 ++++
 io/open.c          |   70 ++++++++++++++++++-----------------
 io/swapext.c       |   19 ++--------
 libfrog/Makefile   |    1 +
 libfrog/bulkstat.c |   44 ++++++++++++++++++++++
 quota/quot.c       |   27 ++++++--------
 scrub/inodes.c     |   28 ++++----------
 8 files changed, 154 insertions(+), 146 deletions(-)
 create mode 100644 libfrog/bulkstat.c


diff --git a/fsr/xfs_fsr.c b/fsr/xfs_fsr.c
index 1541917e..ceee9576 100644
--- a/fsr/xfs_fsr.c
+++ b/fsr/xfs_fsr.c
@@ -102,31 +102,6 @@ static int	nfrags = 0;	/* Debug option: Coerse into specific number
 				 * of extents */
 static int	openopts = O_CREAT|O_EXCL|O_RDWR|O_DIRECT;
 
-static int
-xfs_bulkstat_single(int fd, xfs_ino_t *lastip, struct xfs_bstat *ubuffer)
-{
-    struct xfs_fsop_bulkreq  bulkreq;
-
-    bulkreq.lastip = (__u64 *)lastip;
-    bulkreq.icount = 1;
-    bulkreq.ubuffer = ubuffer;
-    bulkreq.ocount = NULL;
-    return ioctl(fd, XFS_IOC_FSBULKSTAT_SINGLE, &bulkreq);
-}
-
-static int
-xfs_bulkstat(int fd, xfs_ino_t *lastip, int icount,
-                    struct xfs_bstat *ubuffer, __s32 *ocount)
-{
-    struct xfs_fsop_bulkreq  bulkreq;
-
-    bulkreq.lastip = (__u64 *)lastip;
-    bulkreq.icount = icount;
-    bulkreq.ubuffer = ubuffer;
-    bulkreq.ocount = ocount;
-    return ioctl(fd, XFS_IOC_FSBULKSTAT, &bulkreq);
-}
-
 static int
 xfs_swapext(int fd, xfs_swapext_t *sx)
 {
@@ -596,11 +571,11 @@ fsrall_cleanup(int timeout)
 static int
 fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
 {
-
-	int	fsfd, fd;
+	struct xfs_fd	fsxfd = XFS_FD_INIT_EMPTY;
+	int	fd;
 	int	count = 0;
 	int	ret;
-	__s32	buflenout;
+	uint32_t buflenout;
 	struct xfs_bstat buf[GRABSZ];
 	char	fname[64];
 	char	*tname;
@@ -617,25 +592,27 @@ fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
 		return -1;
 	}
 
-	if ((fsfd = open(mntdir, O_RDONLY)) < 0) {
+	if ((fsxfd.fd = open(mntdir, O_RDONLY)) < 0) {
 		fsrprintf(_("unable to open: %s: %s\n"),
 		          mntdir, strerror( errno ));
 		free(fshandlep);
 		return -1;
 	}
 
-	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
+	ret = xfrog_prepare_geometry(&fsxfd);
+	if (ret) {
 		fsrprintf(_("Skipping %s: could not get XFS geometry\n"),
 			  mntdir);
-		close(fsfd);
+		xfrog_close(&fsxfd);
 		free(fshandlep);
 		return -1;
 	}
+	memcpy(&fsgeom, &fsxfd.fsgeom, sizeof(fsgeom));
 
 	tmp_init(mntdir);
 
-	while ((ret = xfs_bulkstat(fsfd,
-				&lastino, GRABSZ, &buf[0], &buflenout)) == 0) {
+	while ((ret = xfrog_bulkstat(&fsxfd, &lastino, GRABSZ, &buf[0],
+				&buflenout)) == 0) {
 		struct xfs_bstat *p;
 		struct xfs_bstat *endp;
 
@@ -684,16 +661,16 @@ fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
 		}
 		if (endtime && endtime < time(NULL)) {
 			tmp_close(mntdir);
-			close(fsfd);
+			xfrog_close(&fsxfd);
 			fsrall_cleanup(1);
 			exit(1);
 		}
 	}
 	if (ret < 0)
-		fsrprintf(_("%s: xfs_bulkstat: %s\n"), progname, strerror(errno));
+		fsrprintf(_("%s: xfrog_bulkstat: %s\n"), progname, strerror(errno));
 out0:
 	tmp_close(mntdir);
-	close(fsfd);
+	xfrog_close(&fsxfd);
 	free(fshandlep);
 	return 0;
 }
@@ -726,13 +703,16 @@ fsrdir(char *dirname)
  * an open on the file and passes this all to fsrfile_common.
  */
 static int
-fsrfile(char *fname, xfs_ino_t ino)
+fsrfile(
+	char			*fname,
+	xfs_ino_t		ino)
 {
-	struct xfs_bstat statbuf;
-	jdm_fshandle_t	*fshandlep;
-	int	fd = -1, fsfd = -1;
-	int	error = -1;
-	char	*tname;
+	struct xfs_fd		fsxfd = XFS_FD_INIT_EMPTY;
+	struct xfs_bstat	statbuf;
+	jdm_fshandle_t		*fshandlep;
+	int			fd = -1;
+	int			error = -1;
+	char			*tname;
 
 	fshandlep = jdm_getfshandle(getparent (fname) );
 	if (!fshandlep) {
@@ -745,14 +725,21 @@ fsrfile(char *fname, xfs_ino_t ino)
 	 * Need to open something on the same filesystem as the
 	 * file.  Open the parent.
 	 */
-	fsfd = open(getparent(fname), O_RDONLY);
-	if (fsfd < 0) {
+	fsxfd.fd = open(getparent(fname), O_RDONLY);
+	if (fsxfd.fd < 0) {
 		fsrprintf(_("unable to open sys handle for %s: %s\n"),
 			fname, strerror(errno));
 		goto out;
 	}
 
-	if ((xfs_bulkstat_single(fsfd, &ino, &statbuf)) < 0) {
+	error = xfrog_prepare_geometry(&fsxfd);
+	if (error) {
+		fsrprintf(_("Unable to get geom on fs for: %s\n"), fname);
+		goto out;
+	}
+
+	error = xfrog_bulkstat_single(&fsxfd, ino, &statbuf);
+	if (error < 0) {
 		fsrprintf(_("unable to get bstat on %s: %s\n"),
 			fname, strerror(errno));
 		goto out;
@@ -765,11 +752,8 @@ fsrfile(char *fname, xfs_ino_t ino)
 		goto out;
 	}
 
-	/* Get the fs geometry */
-	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
-		fsrprintf(_("Unable to get geom on fs for: %s\n"), fname);
-		goto out;
-	}
+	/* Stash the fs geometry for general use. */
+	memcpy(&fsgeom, &fsxfd.fsgeom, sizeof(fsgeom));
 
 	tname = gettmpname(fname);
 
@@ -777,8 +761,7 @@ fsrfile(char *fname, xfs_ino_t ino)
 		error = fsrfile_common(fname, tname, NULL, fd, &statbuf);
 
 out:
-	if (fsfd >= 0)
-		close(fsfd);
+	xfrog_close(&fsxfd);
 	if (fd >= 0)
 		close(fd);
 	free(fshandlep);
@@ -945,6 +928,7 @@ fsr_setup_attr_fork(
 	struct xfs_bstat *bstatp)
 {
 #ifdef HAVE_FSETXATTR
+	struct xfs_fd	txfd = XFS_FD_INIT(tfd);
 	struct stat	tstatbuf;
 	int		i;
 	int		diff = 0;
@@ -962,7 +946,7 @@ fsr_setup_attr_fork(
 	if (!(fsgeom.flags & XFS_FSOP_GEOM_FLAGS_ATTR2) ||
 	    bstatp->bs_forkoff == 0) {
 		/* attr1 */
-		ret = fsetxattr(tfd, "user.X", "X", 1, XATTR_CREATE);
+		ret = fsetxattr(txfd.fd, "user.X", "X", 1, XATTR_CREATE);
 		if (ret) {
 			fsrprintf(_("could not set ATTR\n"));
 			return -1;
@@ -972,7 +956,7 @@ fsr_setup_attr_fork(
 
 	/* attr2 w/ fork offsets */
 
-	if (fstat(tfd, &tstatbuf) < 0) {
+	if (fstat(txfd.fd, &tstatbuf) < 0) {
 		fsrprintf(_("unable to stat temp file: %s\n"),
 					strerror(errno));
 		return -1;
@@ -981,16 +965,16 @@ fsr_setup_attr_fork(
 	i = 0;
 	do {
 		struct xfs_bstat tbstat;
-		xfs_ino_t	ino;
 		char		name[64];
+		int		ret;
 
 		/*
 		 * bulkstat the temp inode to see what the forkoff is.  Use
 		 * this to compare against the target and determine what we
 		 * need to do.
 		 */
-		ino = tstatbuf.st_ino;
-		if ((xfs_bulkstat_single(tfd, &ino, &tbstat)) < 0) {
+		ret = xfrog_bulkstat_single(&txfd, tstatbuf.st_ino, &tbstat);
+		if (ret < 0) {
 			fsrprintf(_("unable to get bstat on temp file: %s\n"),
 						strerror(errno));
 			return -1;
@@ -1012,7 +996,7 @@ fsr_setup_attr_fork(
 		 */
 		if (!tbstat.bs_forkoff) {
 			ASSERT(i == 0);
-			ret = fsetxattr(tfd, name, "XX", 2, XATTR_CREATE);
+			ret = fsetxattr(txfd.fd, name, "XX", 2, XATTR_CREATE);
 			if (ret) {
 				fsrprintf(_("could not set ATTR\n"));
 				return -1;
@@ -1048,7 +1032,7 @@ fsr_setup_attr_fork(
 			if (diff < 0) {
 				char val[2048];
 				memset(val, 'X', 2048);
-				if (fsetxattr(tfd, name, val, 2048, 0)) {
+				if (fsetxattr(txfd.fd, name, val, 2048, 0)) {
 					fsrprintf(_("big ATTR set failed\n"));
 					return -1;
 				}
@@ -1092,7 +1076,7 @@ fsr_setup_attr_fork(
 		}
 
 		/* we need to grow the attr fork, so create another attr */
-		ret = fsetxattr(tfd, name, "XX", 2, XATTR_CREATE);
+		ret = fsetxattr(txfd.fd, name, "XX", 2, XATTR_CREATE);
 		if (ret) {
 			fsrprintf(_("could not set ATTR\n"));
 			return -1;
diff --git a/include/xfrog.h b/include/xfrog.h
index 091edebb..a7c32d87 100644
--- a/include/xfrog.h
+++ b/include/xfrog.h
@@ -100,4 +100,11 @@ xfrog_b_to_fsbt(
 	return bytes >> xfd->blocklog;
 }
 
+/* Bulkstat wrappers */
+struct xfs_bstat;
+int xfrog_bulkstat_single(struct xfs_fd *xfd, uint64_t ino,
+		struct xfs_bstat *ubuffer);
+int xfrog_bulkstat(struct xfs_fd *xfd, uint64_t *lastino, uint32_t icount,
+		struct xfs_bstat *ubuffer, uint32_t *ocount);
+
 #endif	/* __XFROG_H__ */
diff --git a/io/open.c b/io/open.c
index e70c8cb0..67976f7f 100644
--- a/io/open.c
+++ b/io/open.c
@@ -713,19 +713,18 @@ get_last_inode(void)
 
 static int
 inode_f(
-	  int			argc,
-	  char			**argv)
+	int			argc,
+	char			**argv)
 {
-	__s32			count = 0;
-	__u64			result_ino = 0;
-	__u64			userino = NULLFSINO;
+	struct xfs_bstat	bstat;
+	uint32_t		count = 0;
+	uint64_t		result_ino = 0;
+	uint64_t		userino = NULLFSINO;
 	char			*p;
 	int			c;
 	int			verbose = 0;
 	int			ret_next = 0;
-	int			cmd = 0;
-	struct xfs_fsop_bulkreq	bulkreq;
-	struct xfs_bstat	bstat;
+	int			ret;
 
 	while ((c = getopt(argc, argv, "nv")) != EOF) {
 		switch (c) {
@@ -767,35 +766,36 @@ inode_f(
 			exitcode = 1;
 			return 0;
 		}
+	} else if (ret_next) {
+		struct xfs_fd	xfd = XFS_FD_INIT(file->fd);
+
+		/* get next inode */
+		ret = xfrog_bulkstat(&xfd, &userino, 1, &bstat, &count);
+		if (ret) {
+			perror("xfsctl");
+			exitcode = 1;
+			return 0;
+		}
+
+		/* The next inode in use, or 0 if none */
+		if (count)
+			result_ino = bstat.bs_ino;
+		else
+			result_ino = 0;
 	} else {
-		if (ret_next)	/* get next inode */
-			cmd = XFS_IOC_FSBULKSTAT;
-		else		/* get this inode */
-			cmd = XFS_IOC_FSBULKSTAT_SINGLE;
-
-		bulkreq.lastip = &userino;
-		bulkreq.icount = 1;
-		bulkreq.ubuffer = &bstat;
-		bulkreq.ocount = &count;
-
-		if (xfsctl(file->name, file->fd, cmd, &bulkreq)) {
-			if (!ret_next && errno == EINVAL) {
-				/* Not in use */
-				result_ino = 0;
-			} else {
-				perror("xfsctl");
-				exitcode = 1;
-				return 0;
-			}
-		} else if (ret_next) {
-			/* The next inode in use, or 0 if none */
-			if (*bulkreq.ocount)
-				result_ino = bstat.bs_ino;
-			else
-				result_ino = 0;
+		struct xfs_fd	xfd = XFS_FD_INIT(file->fd);
+
+		/* get this inode */
+		ret = xfrog_bulkstat_single(&xfd, userino, &bstat);
+		if (ret && errno == EINVAL) {
+			/* Not in use */
+			result_ino = 0;
+		} else if (ret) {
+			perror("bulkstat_single");
+			exitcode = 1;
+			return 0;
 		} else {
-			/* The inode we asked about */
-			result_ino = userino;
+			result_ino = bstat.bs_ino;
 		}
 	}
 
diff --git a/io/swapext.c b/io/swapext.c
index d360c221..e8432e7d 100644
--- a/io/swapext.c
+++ b/io/swapext.c
@@ -8,6 +8,7 @@
 #include "input.h"
 #include "init.h"
 #include "io.h"
+#include "xfrog.h"
 
 static cmdinfo_t swapext_cmd;
 
@@ -20,26 +21,12 @@ swapext_help(void)
 "\n"));
 }
 
-static int
-xfs_bulkstat_single(
-	int			fd,
-	xfs_ino_t		*lastip,
-	struct xfs_bstat	*ubuffer)
-{
-	struct xfs_fsop_bulkreq	bulkreq;
-
-	bulkreq.lastip = (__u64 *)lastip;
-	bulkreq.icount = 1;
-	bulkreq.ubuffer = ubuffer;
-	bulkreq.ocount = NULL;
-	return ioctl(fd, XFS_IOC_FSBULKSTAT_SINGLE, &bulkreq);
-}
-
 static int
 swapext_f(
 	int			argc,
 	char			**argv)
 {
+	struct xfs_fd		fxfd = XFS_FD_INIT(file->fd);
 	int			fd;
 	int			error;
 	struct xfs_swapext	sx;
@@ -60,7 +47,7 @@ swapext_f(
 		goto out;
 	}
 
-	error = xfs_bulkstat_single(file->fd, &stat.st_ino, &sx.sx_stat);
+	error = xfrog_bulkstat_single(&fxfd, stat.st_ino, &sx.sx_stat);
 	if (error) {
 		perror("bulkstat");
 		goto out;
diff --git a/libfrog/Makefile b/libfrog/Makefile
index f5a0539b..05c6f701 100644
--- a/libfrog/Makefile
+++ b/libfrog/Makefile
@@ -13,6 +13,7 @@ LT_AGE = 0
 CFILES = \
 avl64.c \
 bitmap.c \
+bulkstat.c \
 convert.c \
 crc32.c \
 fsgeom.c \
diff --git a/libfrog/bulkstat.c b/libfrog/bulkstat.c
new file mode 100644
index 00000000..5db34eba
--- /dev/null
+++ b/libfrog/bulkstat.c
@@ -0,0 +1,44 @@
+// SPDX-License-Identifier: GPL-2.0+
+/*
+ * Copyright (C) 2019 Oracle.  All Rights Reserved.
+ * Author: Darrick J. Wong <darrick.wong@oracle.com>
+ */
+#include "xfs.h"
+#include "xfrog.h"
+
+/* Bulkstat a single inode. */
+int
+xfrog_bulkstat_single(
+	struct xfs_fd		*xfd,
+	uint64_t		ino,
+	struct xfs_bstat	*ubuffer)
+{
+	__u64			i = ino;
+	struct xfs_fsop_bulkreq	bulkreq = {
+		.lastip		= &i,
+		.icount		= 1,
+		.ubuffer	= ubuffer,
+		.ocount		= NULL,
+	};
+
+	return ioctl(xfd->fd, XFS_IOC_FSBULKSTAT_SINGLE, &bulkreq);
+}
+
+/* Bulkstat a bunch of inodes. */
+int
+xfrog_bulkstat(
+	struct xfs_fd		*xfd,
+	uint64_t		*lastino,
+	uint32_t		icount,
+	struct xfs_bstat	*ubuffer,
+	uint32_t		*ocount)
+{
+	struct xfs_fsop_bulkreq	bulkreq = {
+		.lastip		= (__u64 *)lastino,
+		.icount		= icount,
+		.ubuffer	= ubuffer,
+		.ocount		= (__s32 *)ocount,
+	};
+
+	return ioctl(xfd->fd, XFS_IOC_FSBULKSTAT, &bulkreq);
+}
diff --git a/quota/quot.c b/quota/quot.c
index 6bc91171..3455636c 100644
--- a/quota/quot.c
+++ b/quota/quot.c
@@ -11,6 +11,7 @@
 #include <grp.h>
 #include "init.h"
 #include "quota.h"
+#include "xfrog.h"
 
 typedef struct du {
 	struct du	*next;
@@ -124,13 +125,13 @@ quot_bulkstat_add(
 static void
 quot_bulkstat_mount(
 	char			*fsdir,
-	uint			flags)
+	unsigned int		flags)
 {
-	struct xfs_fsop_bulkreq	bulkreq;
+	struct xfs_fd		fsxfd = XFS_FD_INIT_EMPTY;
 	struct xfs_bstat	*buf;
-	__u64			last = 0;
-	__s32			count;
-	int			i, sts, fsfd;
+	uint64_t		last = 0;
+	uint32_t		count;
+	int			i, sts;
 	du_t			**dp;
 
 	/*
@@ -145,8 +146,8 @@ quot_bulkstat_mount(
 			*dp = NULL;
 	ndu[0] = ndu[1] = ndu[2] = 0;
 
-	fsfd = open(fsdir, O_RDONLY);
-	if (fsfd < 0) {
+	fsxfd.fd = open(fsdir, O_RDONLY);
+	if (fsxfd.fd < 0) {
 		perror(fsdir);
 		return;
 	}
@@ -154,16 +155,12 @@ quot_bulkstat_mount(
 	buf = (struct xfs_bstat *)calloc(NBSTAT, sizeof(struct xfs_bstat));
 	if (!buf) {
 		perror("calloc");
-		close(fsfd);
+		xfrog_close(&fsxfd);
 		return;
 	}
 
-	bulkreq.lastip = &last;
-	bulkreq.icount = NBSTAT;
-	bulkreq.ubuffer = buf;
-	bulkreq.ocount = &count;
-
-	while ((sts = xfsctl(fsdir, fsfd, XFS_IOC_FSBULKSTAT, &bulkreq)) == 0) {
+	while ((sts = xfrog_bulkstat(&fsxfd, &last, NBSTAT, buf,
+				&count)) == 0) {
 		if (count == 0)
 			break;
 		for (i = 0; i < count; i++)
@@ -172,7 +169,7 @@ quot_bulkstat_mount(
 	if (sts < 0)
 		perror("XFS_IOC_FSBULKSTAT"),
 	free(buf);
-	close(fsfd);
+	xfrog_close(&fsxfd);
 }
 
 static int
diff --git a/scrub/inodes.c b/scrub/inodes.c
index a9000218..09dd0055 100644
--- a/scrub/inodes.c
+++ b/scrub/inodes.c
@@ -17,6 +17,7 @@
 #include "xfs_scrub.h"
 #include "common.h"
 #include "inodes.h"
+#include "xfrog.h"
 
 /*
  * Iterate a range of inodes.
@@ -50,17 +51,10 @@ xfs_iterate_inodes_range_check(
 	struct xfs_inogrp	*inogrp,
 	struct xfs_bstat	*bstat)
 {
-	struct xfs_fsop_bulkreq	onereq = {NULL};
 	struct xfs_bstat	*bs;
-	__u64			oneino;
-	__s32			onelen = 0;
 	int			i;
 	int			error;
 
-	onereq.lastip  = &oneino;
-	onereq.icount  = 1;
-	onereq.ocount  = &onelen;
-
 	for (i = 0, bs = bstat; i < XFS_INODES_PER_CHUNK; i++) {
 		if (!(inogrp->xi_allocmask & (1ULL << i)))
 			continue;
@@ -70,10 +64,8 @@ xfs_iterate_inodes_range_check(
 		}
 
 		/* Load the one inode. */
-		oneino = inogrp->xi_startino + i;
-		onereq.ubuffer = bs;
-		error = ioctl(ctx->mnt.fd, XFS_IOC_FSBULKSTAT_SINGLE,
-				&onereq);
+		error = xfrog_bulkstat_single(&ctx->mnt,
+				inogrp->xi_startino + i, bs);
 		if (error || bs->bs_ino != inogrp->xi_startino + i) {
 			memset(bs, 0, sizeof(struct xfs_bstat));
 			bs->bs_ino = inogrp->xi_startino + i;
@@ -99,7 +91,6 @@ xfs_iterate_inodes_range(
 	void			*arg)
 {
 	struct xfs_fsop_bulkreq	igrpreq = {NULL};
-	struct xfs_fsop_bulkreq	bulkreq = {NULL};
 	struct xfs_handle	handle;
 	struct xfs_inogrp	inogrp;
 	struct xfs_bstat	bstat[XFS_INODES_PER_CHUNK];
@@ -107,8 +98,8 @@ xfs_iterate_inodes_range(
 	char			buf[DESCR_BUFSZ];
 	struct xfs_bstat	*bs;
 	__u64			igrp_ino;
-	__u64			ino;
-	__s32			bulklen = 0;
+	uint64_t		ino;
+	uint32_t		bulklen = 0;
 	__s32			igrplen = 0;
 	bool			moveon = true;
 	int			i;
@@ -117,10 +108,6 @@ xfs_iterate_inodes_range(
 
 
 	memset(bstat, 0, XFS_INODES_PER_CHUNK * sizeof(struct xfs_bstat));
-	bulkreq.lastip  = &ino;
-	bulkreq.icount  = XFS_INODES_PER_CHUNK;
-	bulkreq.ubuffer = &bstat;
-	bulkreq.ocount  = &bulklen;
 
 	igrpreq.lastip  = &igrp_ino;
 	igrpreq.icount  = 1;
@@ -138,14 +125,15 @@ xfs_iterate_inodes_range(
 	while (!error && igrplen) {
 		/* Load the inodes. */
 		ino = inogrp.xi_startino - 1;
-		bulkreq.icount = inogrp.xi_alloccount;
+
 		/*
 		 * We can have totally empty inode chunks on filesystems where
 		 * there are more than 64 inodes per block.  Skip these.
 		 */
 		if (inogrp.xi_alloccount == 0)
 			goto igrp_retry;
-		error = ioctl(ctx->mnt.fd, XFS_IOC_FSBULKSTAT, &bulkreq);
+		error = xfrog_bulkstat(&ctx->mnt, &ino, inogrp.xi_alloccount,
+				bstat, &bulklen);
 		if (error)
 			str_info(ctx, descr, "%s", strerror_r(errno,
 						buf, DESCR_BUFSZ));

Re: [PATCH 5/6] libfrog: refactor open-coded bulkstat calls

[Dave Chinner]

On Tue, Aug 20, 2019 at 01:31:03PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Refactor the BULKSTAT_SINGLE and BULKSTAT ioctl callsites into helper
> functions.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
.....
> ---
> @@ -617,25 +592,27 @@ fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
>  		return -1;
>  	}
>  
> -	if ((fsfd = open(mntdir, O_RDONLY)) < 0) {
> +	if ((fsxfd.fd = open(mntdir, O_RDONLY)) < 0) {
>  		fsrprintf(_("unable to open: %s: %s\n"),
>  		          mntdir, strerror( errno ));
>  		free(fshandlep);
>  		return -1;
>  	}
>  
> -	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
> +	ret = xfrog_prepare_geometry(&fsxfd);
> +	if (ret) {
>  		fsrprintf(_("Skipping %s: could not get XFS geometry\n"),
>  			  mntdir);
> -		close(fsfd);
> +		xfrog_close(&fsxfd);
>  		free(fshandlep);
>  		return -1;
>  	}

/me wonders if this would be better wrapped as xfd_open(&xfd, path, flags,
mode) ?

> @@ -684,16 +661,16 @@ fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
>  		}
>  		if (endtime && endtime < time(NULL)) {
>  			tmp_close(mntdir);
> -			close(fsfd);
> +			xfrog_close(&fsxfd);
>  			fsrall_cleanup(1);
>  			exit(1);
>  		}
>  	}
>  	if (ret < 0)
> -		fsrprintf(_("%s: xfs_bulkstat: %s\n"), progname, strerror(errno));
> +		fsrprintf(_("%s: xfrog_bulkstat: %s\n"), progname, strerror(errno));

It'd change this to just "bulkstat" - it's a user facing error, and
they aren't going to know what "xfrog" means...

> @@ -745,14 +725,21 @@ fsrfile(char *fname, xfs_ino_t ino)
>  	 * Need to open something on the same filesystem as the
>  	 * file.  Open the parent.
>  	 */
> -	fsfd = open(getparent(fname), O_RDONLY);
> -	if (fsfd < 0) {
> +	fsxfd.fd = open(getparent(fname), O_RDONLY);
> +	if (fsxfd.fd < 0) {
>  		fsrprintf(_("unable to open sys handle for %s: %s\n"),
>  			fname, strerror(errno));
>  		goto out;
>  	}
>  
> -	if ((xfs_bulkstat_single(fsfd, &ino, &statbuf)) < 0) {
> +	error = xfrog_prepare_geometry(&fsxfd);
> +	if (error) {
> +		fsrprintf(_("Unable to get geom on fs for: %s\n"), fname);
> +		goto out;
> +	}
> +

same xfd_open() code here.

> diff --git a/io/open.c b/io/open.c
> index e70c8cb0..67976f7f 100644
> --- a/io/open.c
> +++ b/io/open.c
....
> @@ -767,35 +766,36 @@ inode_f(
>  			exitcode = 1;
>  			return 0;
>  		}
> +	} else if (ret_next) {
> +		struct xfs_fd	xfd = XFS_FD_INIT(file->fd);
> +
> +		/* get next inode */
> +		ret = xfrog_bulkstat(&xfd, &userino, 1, &bstat, &count);
> +		if (ret) {
> +			perror("xfsctl");

perror("bulkstat");

....

Otherwise looks ok.

Cheers,

Dave.
-- 
Dave Chinner
david@fromorbit.com

[PATCH 6/6] libfrog: refactor open-coded INUMBERS calls

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Refactor all the INUMBERS ioctl callsites into helper functions.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 include/xfrog.h    |    4 ++++
 io/imap.c          |   32 +++++++++++++++-----------------
 io/open.c          |   20 ++++++++------------
 libfrog/bulkstat.c |   19 +++++++++++++++++++
 scrub/fscounters.c |   18 +++++++-----------
 scrub/inodes.c     |   15 +++++----------
 6 files changed, 58 insertions(+), 50 deletions(-)


diff --git a/include/xfrog.h b/include/xfrog.h
index a7c32d87..6b605dcf 100644
--- a/include/xfrog.h
+++ b/include/xfrog.h
@@ -107,4 +107,8 @@ int xfrog_bulkstat_single(struct xfs_fd *xfd, uint64_t ino,
 int xfrog_bulkstat(struct xfs_fd *xfd, uint64_t *lastino, uint32_t icount,
 		struct xfs_bstat *ubuffer, uint32_t *ocount);
 
+struct xfs_inogrp;
+int xfrog_inumbers(struct xfs_fd *xfd, uint64_t *lastino, uint32_t icount,
+		struct xfs_inogrp *ubuffer, uint32_t *ocount);
+
 #endif	/* __XFROG_H__ */
diff --git a/io/imap.c b/io/imap.c
index 9667289a..9a3d8965 100644
--- a/io/imap.c
+++ b/io/imap.c
@@ -8,18 +8,20 @@
 #include "input.h"
 #include "init.h"
 #include "io.h"
+#include "xfrog.h"
 
 static cmdinfo_t imap_cmd;
 
 static int
 imap_f(int argc, char **argv)
 {
-	int		count;
-	int		nent;
-	int		i;
-	__u64		last = 0;
-	struct xfs_inogrp *t;
-	struct xfs_fsop_bulkreq bulkreq;
+	struct xfs_fd		xfd = XFS_FD_INIT(file->fd);
+	struct xfs_inogrp	*t;
+	uint64_t		last = 0;
+	uint32_t		count;
+	uint32_t		nent;
+	int			i;
+	int			error;
 
 	if (argc != 2)
 		nent = 1;
@@ -30,14 +32,8 @@ imap_f(int argc, char **argv)
 	if (!t)
 		return 0;
 
-	bulkreq.lastip  = &last;
-	bulkreq.icount  = nent;
-	bulkreq.ubuffer = (void *)t;
-	bulkreq.ocount  = &count;
-
-	while (xfsctl(file->name, file->fd, XFS_IOC_FSINUMBERS, &bulkreq) == 0) {
-		if (count == 0)
-			goto out_free;
+	while ((error = xfrog_inumbers(&xfd, &last, nent, t, &count)) == 0 &&
+	       count > 0) {
 		for (i = 0; i < count; i++) {
 			printf(_("ino %10llu count %2d mask %016llx\n"),
 				(unsigned long long)t[i].xi_startino,
@@ -45,9 +41,11 @@ imap_f(int argc, char **argv)
 				(unsigned long long)t[i].xi_allocmask);
 		}
 	}
-	perror("xfsctl(XFS_IOC_FSINUMBERS)");
-	exitcode = 1;
-out_free:
+
+	if (error) {
+		perror("xfsctl(XFS_IOC_FSINUMBERS)");
+		exitcode = 1;
+	}
 	free(t);
 	return 0;
 }
diff --git a/io/open.c b/io/open.c
index 67976f7f..968a9d9e 100644
--- a/io/open.c
+++ b/io/open.c
@@ -669,24 +669,20 @@ inode_help(void)
 "\n"));
 }
 
+#define IGROUP_NR	(1024)
 static __u64
 get_last_inode(void)
 {
-	__u64			lastip = 0;
-	__u64			lastgrp = 0;
-	__s32			ocount = 0;
+	struct xfs_fd		xfd = XFS_FD_INIT(file->fd);
+	uint64_t		lastip = 0;
+	uint32_t		lastgrp = 0;
+	uint32_t		ocount = 0;
 	__u64			last_ino;
-	struct xfs_inogrp	igroup[1024];
-	struct xfs_fsop_bulkreq	bulkreq;
-
-	bulkreq.lastip = &lastip;
-	bulkreq.ubuffer = &igroup;
-	bulkreq.icount = sizeof(igroup) / sizeof(struct xfs_inogrp);
-	bulkreq.ocount = &ocount;
+	struct xfs_inogrp	igroup[IGROUP_NR];
 
 	for (;;) {
-		if (xfsctl(file->name, file->fd, XFS_IOC_FSINUMBERS,
-				&bulkreq)) {
+		if (xfrog_inumbers(&xfd, &lastip, IGROUP_NR, igroup,
+					&ocount)) {
 			perror("XFS_IOC_FSINUMBERS");
 			return 0;
 		}
diff --git a/libfrog/bulkstat.c b/libfrog/bulkstat.c
index 5db34eba..6632ffbb 100644
--- a/libfrog/bulkstat.c
+++ b/libfrog/bulkstat.c
@@ -42,3 +42,22 @@ xfrog_bulkstat(
 
 	return ioctl(xfd->fd, XFS_IOC_FSBULKSTAT, &bulkreq);
 }
+
+/* Query inode allocation bitmask information. */
+int
+xfrog_inumbers(
+	struct xfs_fd		*xfd,
+	uint64_t		*lastino,
+	uint32_t		icount,
+	struct xfs_inogrp	*ubuffer,
+	uint32_t		*ocount)
+{
+	struct xfs_fsop_bulkreq	bulkreq = {
+		.lastip		= (__u64 *)lastino,
+		.icount		= icount,
+		.ubuffer	= ubuffer,
+		.ocount		= (__s32 *)ocount,
+	};
+
+	return ioctl(xfd->fd, XFS_IOC_FSINUMBERS, &bulkreq);
+}
diff --git a/scrub/fscounters.c b/scrub/fscounters.c
index adb79b50..cd216b30 100644
--- a/scrub/fscounters.c
+++ b/scrub/fscounters.c
@@ -15,6 +15,7 @@
 #include "xfs_scrub.h"
 #include "common.h"
 #include "fscounters.h"
+#include "xfrog.h"
 
 /*
  * Filesystem counter collection routines.  We can count the number of
@@ -41,26 +42,21 @@ xfs_count_inodes_range(
 	uint64_t		last_ino,
 	uint64_t		*count)
 {
-	struct xfs_fsop_bulkreq	igrpreq = {NULL};
 	struct xfs_inogrp	inogrp;
-	__u64			igrp_ino;
+	uint64_t		igrp_ino;
 	uint64_t		nr = 0;
-	__s32			igrplen = 0;
+	uint32_t		igrplen = 0;
 	int			error;
 
 	ASSERT(!(first_ino & (XFS_INODES_PER_CHUNK - 1)));
 	ASSERT((last_ino & (XFS_INODES_PER_CHUNK - 1)));
 
-	igrpreq.lastip  = &igrp_ino;
-	igrpreq.icount  = 1;
-	igrpreq.ubuffer = &inogrp;
-	igrpreq.ocount  = &igrplen;
-
 	igrp_ino = first_ino;
-	error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
-	while (!error && igrplen && inogrp.xi_startino < last_ino) {
+	while (!(error = xfrog_inumbers(&ctx->mnt, &igrp_ino, 1, &inogrp,
+			&igrplen))) {
+		if (igrplen == 0 || inogrp.xi_startino >= last_ino)
+			break;
 		nr += inogrp.xi_alloccount;
-		error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
 	}
 
 	if (error) {
diff --git a/scrub/inodes.c b/scrub/inodes.c
index 09dd0055..06d1245d 100644
--- a/scrub/inodes.c
+++ b/scrub/inodes.c
@@ -90,17 +90,16 @@ xfs_iterate_inodes_range(
 	xfs_inode_iter_fn	fn,
 	void			*arg)
 {
-	struct xfs_fsop_bulkreq	igrpreq = {NULL};
 	struct xfs_handle	handle;
 	struct xfs_inogrp	inogrp;
 	struct xfs_bstat	bstat[XFS_INODES_PER_CHUNK];
 	char			idescr[DESCR_BUFSZ];
 	char			buf[DESCR_BUFSZ];
 	struct xfs_bstat	*bs;
-	__u64			igrp_ino;
+	uint64_t		igrp_ino;
 	uint64_t		ino;
 	uint32_t		bulklen = 0;
-	__s32			igrplen = 0;
+	uint32_t		igrplen = 0;
 	bool			moveon = true;
 	int			i;
 	int			error;
@@ -109,11 +108,6 @@ xfs_iterate_inodes_range(
 
 	memset(bstat, 0, XFS_INODES_PER_CHUNK * sizeof(struct xfs_bstat));
 
-	igrpreq.lastip  = &igrp_ino;
-	igrpreq.icount  = 1;
-	igrpreq.ubuffer = &inogrp;
-	igrpreq.ocount  = &igrplen;
-
 	memcpy(&handle.ha_fsid, fshandle, sizeof(handle.ha_fsid));
 	handle.ha_fid.fid_len = sizeof(xfs_fid_t) -
 			sizeof(handle.ha_fid.fid_len);
@@ -121,7 +115,7 @@ xfs_iterate_inodes_range(
 
 	/* Find the inode chunk & alloc mask */
 	igrp_ino = first_ino;
-	error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
+	error = xfrog_inumbers(&ctx->mnt, &igrp_ino, 1, &inogrp, &igrplen);
 	while (!error && igrplen) {
 		/* Load the inodes. */
 		ino = inogrp.xi_startino - 1;
@@ -178,7 +172,8 @@ _("Changed too many times during scan; giving up."));
 
 		stale_count = 0;
 igrp_retry:
-		error = ioctl(ctx->mnt.fd, XFS_IOC_FSINUMBERS, &igrpreq);
+		error = xfrog_inumbers(&ctx->mnt, &igrp_ino, 1, &inogrp,
+				&igrplen);
 	}
 
 err:

Re: [PATCH 6/6] libfrog: refactor open-coded INUMBERS calls

[Dave Chinner]

On Tue, Aug 20, 2019 at 01:31:09PM -0700, Darrick J. Wong wrote:
> From: Darrick J. Wong <darrick.wong@oracle.com>
> 
> Refactor all the INUMBERS ioctl callsites into helper functions.
> 
> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
> ---
>  include/xfrog.h    |    4 ++++
>  io/imap.c          |   32 +++++++++++++++-----------------
>  io/open.c          |   20 ++++++++------------
>  libfrog/bulkstat.c |   19 +++++++++++++++++++
>  scrub/fscounters.c |   18 +++++++-----------
>  scrub/inodes.c     |   15 +++++----------
>  6 files changed, 58 insertions(+), 50 deletions(-)

looks fine.

Reviewed-by: Dave Chinner <dchinner@redhat.com>
-- 
Dave Chinner
david@fromorbit.com

[PATCH v2 0/6] xfsprogs-5.1: fix various problems

[Darrick J. Wong]

Hi all,

Latest respin of patches fixing various problems in xfsprogs.

Patches 1-3 refactor all utilities to use common libfrog functions to
retrieve the filesystem geometry, bulkstat, and inumbers.  The helpers
will make it easier for newer userspace to fall back to older versions
of ioctls.

Patch 4 strengthens mkfs's log alignment checking code.

Patch 5 fixes xfs_io repair command error reporting.

Patch 6 fixes libxfs-diff to handle files that are in libxfs/ in xfsprogs
but still in fs/xfs/ in the kernel.

If you're going to start using this mess, you probably ought to just
pull from my git trees, which are linked below.

This is an extraordinary way to destroy everything.  Enjoy!
Comments and questions are, as always, welcome.

--D

xfsprogs git tree:
https://git.kernel.org/cgit/linux/kernel/git/djwong/xfsprogs-dev.git/log/?h=xfsprogs-5.1-fixes

[PATCH 1/6] libfrog: refactor online geometry queries

[Darrick J. Wong]

From: Darrick J. Wong <darrick.wong@oracle.com>

Refactor all the open-coded XFS_IOC_FSGEOMETRY queries into a single
helper that we can use to standardize behaviors across mixed xfslibs
versions.  This is the prelude to introducing a new FSGEOMETRY version
in 5.2 and needing to fix the (relatively few) client programs.

Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com>
---
 Makefile            |    1 +
 fsr/xfs_fsr.c       |   26 ++++----------------------
 growfs/xfs_growfs.c |   25 +++++++++----------------
 include/xfrog.h     |   22 ++++++++++++++++++++++
 io/bmap.c           |    3 ++-
 io/fsmap.c          |    3 ++-
 io/open.c           |    3 ++-
 io/stat.c           |    5 +++--
 libfrog/fsgeom.c    |   18 ++++++++++++++++++
 quota/free.c        |    6 +++---
 repair/xfs_repair.c |    5 +++--
 rtcp/Makefile       |    3 +++
 rtcp/xfs_rtcp.c     |    7 ++++---
 scrub/phase1.c      |    3 ++-
 spaceman/file.c     |    3 ++-
 spaceman/info.c     |   25 ++++++++-----------------
 16 files changed, 88 insertions(+), 70 deletions(-)
 create mode 100644 include/xfrog.h


diff --git a/Makefile b/Makefile
index 9204bed8..0edc2700 100644
--- a/Makefile
+++ b/Makefile
@@ -107,6 +107,7 @@ copy: libxlog
 mkfs: libxcmd
 spaceman: libxcmd
 scrub: libhandle libxcmd
+rtcp: libfrog
 
 ifeq ($(HAVE_BUILDDEFS), yes)
 include $(BUILDRULES)
diff --git a/fsr/xfs_fsr.c b/fsr/xfs_fsr.c
index fef6262c..0bfecf37 100644
--- a/fsr/xfs_fsr.c
+++ b/fsr/xfs_fsr.c
@@ -11,6 +11,7 @@
 #include "xfs_bmap_btree.h"
 #include "xfs_attr_sf.h"
 #include "path.h"
+#include "xfrog.h"
 
 #include <fcntl.h>
 #include <errno.h>
@@ -83,9 +84,8 @@ int cmp(const void *, const void *);
 static void tmp_init(char *mnt);
 static char * tmp_next(char *mnt);
 static void tmp_close(char *mnt);
-int xfs_getgeom(int , xfs_fsop_geom_v1_t * );
 
-static xfs_fsop_geom_v1_t fsgeom;	/* geometry of active mounted system */
+static struct xfs_fsop_geom fsgeom;	/* geometry of active mounted system */
 
 #define NMOUNT 64
 static int numfs;
@@ -102,12 +102,6 @@ static int	nfrags = 0;	/* Debug option: Coerse into specific number
 				 * of extents */
 static int	openopts = O_CREAT|O_EXCL|O_RDWR|O_DIRECT;
 
-static int
-xfs_fsgeometry(int fd, xfs_fsop_geom_v1_t *geom)
-{
-    return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, geom);
-}
-
 static int
 xfs_bulkstat_single(int fd, xfs_ino_t *lastip, xfs_bstat_t *ubuffer)
 {
@@ -630,7 +624,7 @@ fsrfs(char *mntdir, xfs_ino_t startino, int targetrange)
 		return -1;
 	}
 
-	if (xfs_getgeom(fsfd, &fsgeom) < 0 ) {
+	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
 		fsrprintf(_("Skipping %s: could not get XFS geometry\n"),
 			  mntdir);
 		close(fsfd);
@@ -772,7 +766,7 @@ fsrfile(char *fname, xfs_ino_t ino)
 	}
 
 	/* Get the fs geometry */
-	if (xfs_getgeom(fsfd, &fsgeom) < 0 ) {
+	if (xfrog_geometry(fsfd, &fsgeom) < 0 ) {
 		fsrprintf(_("Unable to get geom on fs for: %s\n"), fname);
 		goto out;
 	}
@@ -1612,18 +1606,6 @@ getnextents(int fd)
 	return(nextents);
 }
 
-/*
- * Get the fs geometry
- */
-int
-xfs_getgeom(int fd, xfs_fsop_geom_v1_t * fsgeom)
-{
-	if (xfs_fsgeometry(fd, fsgeom) < 0) {
-		return -1;
-	}
-	return 0;
-}
-
 /*
  * Get xfs realtime space information
  */
diff --git a/growfs/xfs_growfs.c b/growfs/xfs_growfs.c
index 20089d2b..28dd6d84 100644
--- a/growfs/xfs_growfs.c
+++ b/growfs/xfs_growfs.c
@@ -7,6 +7,7 @@
 #include "libxfs.h"
 #include "path.h"
 #include "fsgeom.h"
+#include "xfrog.h"
 
 static void
 usage(void)
@@ -165,22 +166,14 @@ main(int argc, char **argv)
 	}
 
 	/* get the current filesystem size & geometry */
-	if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY, &geo) < 0) {
-		/*
-		 * OK, new xfsctl barfed - back off and try earlier version
-		 * as we're probably running an older kernel version.
-		 * Only field added in the v2 geometry xfsctl is "logsunit"
-		 * so we'll zero that out for later display (as zero).
-		 */
-		geo.logsunit = 0;
-		if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY_V1, &geo) < 0) {
-			fprintf(stderr, _(
-				"%s: cannot determine geometry of filesystem"
-				" mounted at %s: %s\n"),
-				progname, fname, strerror(errno));
-			exit(1);
-		}
+	if (xfrog_geometry(ffd, &geo)) {
+		fprintf(stderr, _(
+			"%s: cannot determine geometry of filesystem"
+			" mounted at %s: %s\n"),
+			progname, fname, strerror(errno));
+		exit(1);
 	}
+
 	isint = geo.logstart > 0;
 
 	/*
@@ -359,7 +352,7 @@ main(int argc, char **argv)
 		}
 	}
 
-	if (xfsctl(fname, ffd, XFS_IOC_FSGEOMETRY_V1, &ngeo) < 0) {
+	if (xfrog_geometry(ffd, &ngeo)) {
 		fprintf(stderr, _("%s: XFS_IOC_FSGEOMETRY xfsctl failed: %s\n"),
 			progname, strerror(errno));
 		exit(1);
diff --git a/include/xfrog.h b/include/xfrog.h
new file mode 100644
index 00000000..b683cf63
--- /dev/null
+++ b/include/xfrog.h
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (c) 2000-2002 Silicon Graphics, Inc.
+ * All Rights Reserved.
+ */
+#ifndef __XFROG_H__
+#define __XFROG_H__
+
+/*
+ * XFS Filesystem Runtime Online Gluecode
+ * ======================================
+ *
+ * These support functions wrap the more complex xfs ioctls so that xfs
+ * utilities can take advantage of them without having to deal with graceful
+ * degradation in the face of new ioctls.  They will also provide higher level
+ * abstractions when possible.
+ */
+
+struct xfs_fsop_geom;
+int xfrog_geometry(int fd, struct xfs_fsop_geom *fsgeo);
+
+#endif	/* __XFROG_H__ */
diff --git a/io/bmap.c b/io/bmap.c
index d408826a..5f0d12ca 100644
--- a/io/bmap.c
+++ b/io/bmap.c
@@ -9,6 +9,7 @@
 #include "input.h"
 #include "init.h"
 #include "io.h"
+#include "xfrog.h"
 
 static cmdinfo_t bmap_cmd;
 
@@ -105,7 +106,7 @@ bmap_f(
 		bmv_iflags &= ~(BMV_IF_PREALLOC|BMV_IF_NO_DMAPI_READ);
 
 	if (vflag) {
-		c = xfsctl(file->name, file->fd, XFS_IOC_FSGEOMETRY_V1, &fsgeo);
+		c = xfrog_geometry(file->fd, &fsgeo);
 		if (c < 0) {
 			fprintf(stderr,
 				_("%s: can't get geometry [\"%s\"]: %s\n"),
diff --git a/io/fsmap.c b/io/fsmap.c
index 477c36fc..d3ff7dea 100644
--- a/io/fsmap.c
+++ b/io/fsmap.c
@@ -9,6 +9,7 @@
 #include "path.h"
 #include "io.h"
 #include "input.h"
+#include "xfrog.h"
 
 static cmdinfo_t	fsmap_cmd;
 static dev_t		xfs_data_dev;
@@ -447,7 +448,7 @@ fsmap_f(
 	}
 
 	if (vflag) {
-		c = ioctl(file->fd, XFS_IOC_FSGEOMETRY, &fsgeo);
+		c = xfrog_geometry(file->fd, &fsgeo);
 		if (c < 0) {
 			fprintf(stderr,
 				_("%s: can't get geometry [\"%s\"]: %s\n"),
diff --git a/io/open.c b/io/open.c
index c7f5248a..997df119 100644
--- a/io/open.c
+++ b/io/open.c
@@ -9,6 +9,7 @@
 #include "init.h"
 #include "io.h"
 #include "libxfs.h"
+#include "xfrog.h"
 
 #ifndef __O_TMPFILE
 #if defined __alpha__
@@ -118,7 +119,7 @@ openfile(
 	if (flags & IO_PATH) {
 		/* Can't call ioctl() on O_PATH fds */
 		memset(geom, 0, sizeof(*geom));
-	} else if (xfsctl(path, fd, XFS_IOC_FSGEOMETRY, geom) < 0) {
+	} else if (xfrog_geometry(fd, geom)) {
 		perror("XFS_IOC_FSGEOMETRY");
 		close(fd);
 		return -1;
diff --git a/io/stat.c b/io/stat.c
index 37c0b2e8..26b4eb68 100644
--- a/io/stat.c
+++ b/io/stat.c
@@ -12,6 +12,7 @@
 #include "io.h"
 #include "statx.h"
 #include "libxfs.h"
+#include "xfrog.h"
 
 #include <fcntl.h>
 
@@ -194,8 +195,8 @@ statfs_f(
 	}
 	if (file->flags & IO_FOREIGN)
 		return 0;
-	if ((xfsctl(file->name, file->fd, XFS_IOC_FSGEOMETRY_V1, &fsgeo)) < 0) {
-		perror("XFS_IOC_FSGEOMETRY_V1");
+	if (xfrog_geometry(file->fd, &fsgeo)) {
+		perror("XFS_IOC_FSGEOMETRY");
 	} else {
 		printf(_("geom.bsize = %u\n"), fsgeo.blocksize);
 		printf(_("geom.agcount = %u\n"), fsgeo.agcount);
diff --git a/libfrog/fsgeom.c b/libfrog/fsgeom.c
index 8879d161..a6dab3a8 100644
--- a/libfrog/fsgeom.c
+++ b/libfrog/fsgeom.c
@@ -4,6 +4,7 @@
  */
 #include "libxfs.h"
 #include "fsgeom.h"
+#include "xfrog.h"
 
 void
 xfs_report_geom(
@@ -67,3 +68,20 @@ xfs_report_geom(
 		geo->rtextsize * geo->blocksize, (unsigned long long)geo->rtblocks,
 			(unsigned long long)geo->rtextents);
 }
+
+/* Try to obtain the xfs geometry. */
+int
+xfrog_geometry(
+	int			fd,
+	struct xfs_fsop_geom	*fsgeo)
+{
+	int			ret;
+
+	memset(fsgeo, 0, sizeof(*fsgeo));
+
+	ret = ioctl(fd, XFS_IOC_FSGEOMETRY, fsgeo);
+	if (!ret)
+		return 0;
+
+	return ioctl(fd, XFS_IOC_FSGEOMETRY_V1, fsgeo);
+}
diff --git a/quota/free.c b/quota/free.c
index 1d13006e..f47001bf 100644
--- a/quota/free.c
+++ b/quota/free.c
@@ -8,6 +8,7 @@
 #include "command.h"
 #include "init.h"
 #include "quota.h"
+#include "xfrog.h"
 
 static cmdinfo_t free_cmd;
 
@@ -67,9 +68,8 @@ mount_free_space_data(
 	}
 
 	if (!(mount->fs_flags & FS_FOREIGN)) {
-		if ((xfsctl(mount->fs_dir, fd, XFS_IOC_FSGEOMETRY_V1,
-							&fsgeo)) < 0) {
-			perror("XFS_IOC_FSGEOMETRY_V1");
+		if (xfrog_geometry(fd, &fsgeo)) {
+			perror("XFS_IOC_FSGEOMETRY");
 			close(fd);
 			return 0;
 		}
diff --git a/repair/xfs_repair.c b/repair/xfs_repair.c
index 9657503f..e6717d3c 100644
--- a/repair/xfs_repair.c
+++ b/repair/xfs_repair.c
@@ -22,6 +22,7 @@
 #include "dinode.h"
 #include "slab.h"
 #include "rmap.h"
+#include "xfrog.h"
 
 /*
  * option tables for getsubopt calls
@@ -636,11 +637,11 @@ check_fs_vs_host_sectsize(
 {
 	int	fd;
 	long	old_flags;
-	struct xfs_fsop_geom_v1 geom = { 0 };
+	struct xfs_fsop_geom	geom = { 0 };
 
 	fd = libxfs_device_to_fd(x.ddev);
 
-	if (ioctl(fd, XFS_IOC_FSGEOMETRY_V1, &geom) < 0) {
+	if (xfrog_geometry(fd, &geom)) {
 		do_log(_("Cannot get host filesystem geometry.\n"
 	"Repair may fail if there is a sector size mismatch between\n"
 	"the image and the host filesystem.\n"));
diff --git a/rtcp/Makefile b/rtcp/Makefile
index 808b5378..264b4f27 100644
--- a/rtcp/Makefile
+++ b/rtcp/Makefile
@@ -9,6 +9,9 @@ LTCOMMAND = xfs_rtcp
 CFILES = xfs_rtcp.c
 LLDFLAGS = -static
 
+LLDLIBS = $(LIBFROG)
+LTDEPENDENCIES = $(LIBFROG)
+
 default: depend $(LTCOMMAND)
 
 include $(BUILDRULES)
diff --git a/rtcp/xfs_rtcp.c b/rtcp/xfs_rtcp.c
index f928a86a..0b37ee89 100644
--- a/rtcp/xfs_rtcp.c
+++ b/rtcp/xfs_rtcp.c
@@ -5,6 +5,7 @@
  */
 
 #include "libxfs.h"
+#include "xfrog.h"
 
 int rtcp(char *, char *, int);
 int xfsrtextsize(char *path);
@@ -368,8 +369,8 @@ rtcp( char *source, char *target, int fextsize)
 int
 xfsrtextsize( char *path)
 {
-	int fd, rval, rtextsize;
-	xfs_fsop_geom_v1_t geo;
+	struct xfs_fsop_geom	geo;
+	int			fd, rval, rtextsize;
 
 	fd = open( path, O_RDONLY );
 	if ( fd < 0 ) {
@@ -377,7 +378,7 @@ xfsrtextsize( char *path)
 			progname, path, strerror(errno));
 		return -1;
 	}
-	rval = xfsctl( path, fd, XFS_IOC_FSGEOMETRY_V1, &geo );
+	rval = xfrog_geometry(fd, &geo);
 	close(fd);
 	if ( rval < 0 )
 		return -1;
diff --git a/scrub/phase1.c b/scrub/phase1.c
index 04a5f4a9..5ab2a4fe 100644
--- a/scrub/phase1.c
+++ b/scrub/phase1.c
@@ -26,6 +26,7 @@
 #include "disk.h"
 #include "scrub.h"
 #include "repair.h"
+#include "xfrog.h"
 
 /* Phase 1: Find filesystem geometry (and clean up after) */
 
@@ -129,7 +130,7 @@ _("Does not appear to be an XFS filesystem!"));
 	}
 
 	/* Retrieve XFS geometry. */
-	error = ioctl(ctx->mnt_fd, XFS_IOC_FSGEOMETRY, &ctx->geo);
+	error = xfrog_geometry(ctx->mnt_fd, &ctx->geo);
 	if (error) {
 		str_errno(ctx, ctx->mntpoint);
 		return false;
diff --git a/spaceman/file.c b/spaceman/file.c
index 7e33e07e..690dd760 100644
--- a/spaceman/file.c
+++ b/spaceman/file.c
@@ -12,6 +12,7 @@
 #include "init.h"
 #include "path.h"
 #include "space.h"
+#include "xfrog.h"
 
 static cmdinfo_t print_cmd;
 
@@ -56,7 +57,7 @@ openfile(
 		return -1;
 	}
 
-	if (ioctl(fd, XFS_IOC_FSGEOMETRY, geom) < 0) {
+	if (xfrog_geometry(fd, geom)) {
 		if (errno == ENOTTY)
 			fprintf(stderr,
 _("%s: Not on a mounted XFS filesystem.\n"),
diff --git a/spaceman/info.c b/spaceman/info.c
index 01d0744a..714d715c 100644
--- a/spaceman/info.c
+++ b/spaceman/info.c
@@ -9,6 +9,7 @@
 #include "path.h"
 #include "space.h"
 #include "fsgeom.h"
+#include "xfrog.h"
 
 static void
 info_help(void)
@@ -37,24 +38,14 @@ info_f(
 	}
 
 	/* get the current filesystem size & geometry */
-	error = ioctl(file->fd, XFS_IOC_FSGEOMETRY, &geo);
+	error = xfrog_geometry(file->fd, &geo);
 	if (error) {
-		/*
-		 * OK, new xfsctl barfed - back off and try earlier version
-		 * as we're probably running an older kernel version.
-		 * Only field added in the v2 geometry xfsctl is "logsunit"
-		 * so we'll zero that out for later display (as zero).
-		 */
-		geo.logsunit = 0;
-		error = ioctl(file->fd, XFS_IOC_FSGEOMETRY_V1, &geo);
-		if (error) {
-			fprintf(stderr, _(
-				"%s: cannot determine geometry of filesystem"
-				" mounted at %s: %s\n"),
-				progname, file->name, strerror(errno));
-			exitcode = 1;
-			return 0;
-		}
+		fprintf(stderr, _(
+			"%s: cannot determine geometry of filesystem"
+			" mounted at %s: %s\n"),
+			progname, file->name, strerror(errno));
+		exitcode = 1;
+		return 0;
 	}
 
 	xfs_report_geom(&geo, file->fs_path.fs_name, file->fs_path.fs_log,