Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Tue, 5 Jul 2022 15:19:21 -0700
From:   "Darrick J. Wong" <djwong@kernel.org>
To:     Zorro Lang <zlang@redhat.com>
Cc:     Jeremy Bongio <bongiojp@gmail.com>, Ted Tso <tytso@mit.edu>,
        linux-ext4@vger.kernel.org, fstests@vger.kernel.org
Subject: Re: [PATCH v2] ext4/056: add a check to make sure ext4 uuid ioctls
 get/set during fsstress.
Message-ID: <YsS46dI6XQcJB/2B@magnolia>
References: <20220701201332.183711-1-bongiojp@gmail.com>
 <20220705152507.wsnqr56o7h3hzfxn@zlang-mailbox>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220705152507.wsnqr56o7h3hzfxn@zlang-mailbox>
Precedence: bulk

On Tue, Jul 05, 2022 at 11:25:07PM +0800, Zorro Lang wrote:
> On Fri, Jul 01, 2022 at 01:13:32PM -0700, Jeremy Bongio wrote:
> > Adds a utility to set/get uuid through ext4 ioctl. Executes the ioctls
> > while running fsstress. These ioctls are used by tune2fs to safely change
> > the uuid without racing other filesystem modifications.
> > 
> > Signed-off-by: Jeremy Bongio <bongiojp@gmail.com>
> > ---
> >  .gitignore         |   1 +
> >  src/Makefile       |   4 +-
> >  src/uuid_ioctl.c   | 106 +++++++++++++++++++++++++++++++++++++++++++++
> >  tests/ext4/056     |  55 +++++++++++++++++++++++
> >  tests/ext4/056.out |   2 +
> >  5 files changed, 166 insertions(+), 2 deletions(-)
> >  create mode 100644 src/uuid_ioctl.c
> >  create mode 100755 tests/ext4/056
> >  create mode 100644 tests/ext4/056.out
> > 
> > diff --git a/.gitignore b/.gitignore
> > index ba0c572b..dab24d68 100644
> > --- a/.gitignore
> > +++ b/.gitignore
> > @@ -169,6 +169,7 @@ tags
> >  /src/unwritten_mmap
> >  /src/unwritten_sync
> >  /src/usemem
> > +/src/uuid_ioctl
> >  /src/writemod
> >  /src/writev_on_pagefault
> >  /src/xfsctl
> > diff --git a/src/Makefile b/src/Makefile
> > index 111ce1d9..e33e04de 100644
> > --- a/src/Makefile
> > +++ b/src/Makefile
> > @@ -31,14 +31,14 @@ LINUX_TARGETS = xfsctl bstat t_mtab getdevicesize preallo_rw_pattern_reader \
> >  	dio-invalidate-cache stat_test t_encrypted_d_revalidate \
> >  	attr_replace_test swapon mkswap t_attr_corruption t_open_tmpfiles \
> >  	fscrypt-crypt-util bulkstat_null_ocount splice-test chprojid_fail \
> > -	detached_mounts_propagation ext4_resize
> > +	detached_mounts_propagation ext4_resize uuid_ioctl
> >  
> >  EXTRA_EXECS = dmerror fill2attr fill2fs fill2fs_check scaleread.sh \
> >  	      btrfs_crc32c_forged_name.py
> >  
> >  SUBDIRS = log-writes perf
> >  
> > -LLDLIBS = $(LIBHANDLE) $(LIBACL) -lpthread -lrt
> > +LLDLIBS = $(LIBHANDLE) $(LIBACL) -lpthread -lrt -luuid
> >  
> >  ifeq ($(HAVE_XLOG_ASSIGN_LSN), true)
> >  LINUX_TARGETS += loggen
> > diff --git a/src/uuid_ioctl.c b/src/uuid_ioctl.c
> > new file mode 100644
> > index 00000000..a4937478
> > --- /dev/null
> > +++ b/src/uuid_ioctl.c
> > @@ -0,0 +1,106 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +
> > +/*
> > + * Test program which uses the raw ext4 set_fsuuid ioctl directly.
> > + * SYNOPSIS:
> > + *   $0 COMMAND MOUNT_POINT [UUID]
> > + *
> > + * COMMAND must be either "get" or "set".
> > + * The UUID must be a 16 octet represented as 32 hexadecimal digits in canonical
> > + * textual representation, e.g. output from `uuidgen`.
> > + *
> > + */
> > +
> > +#include <stdio.h>
> > +#include <fcntl.h>
> > +#include <errno.h>
> > +#include <unistd.h>
> > +#include <stdint.h>
> > +#include <stdlib.h>
> > +#include <string.h>
> > +#include <sys/ioctl.h>
> > +#include <uuid/uuid.h>
> > +#include <linux/fs.h>
> > +
> > +struct fsuuid {
> > +	__u32   fu_len;
> > +	__u32   fu_flags;
> > +	__u8    fu_uuid[];
> > +};
> > +
> > +#define EXT4_IOC_SETFSUUID_FLAG_BLOCKING 0x1
> > +
> > +#ifndef EXT4_IOC_GETFSUUID
> > +#define EXT4_IOC_GETFSUUID      _IOR('f', 44, struct fsuuid)
> > +#endif
> > +
> > +#ifndef EXT4_IOC_SETFSUUID
> > +#define EXT4_IOC_SETFSUUID      _IOW('f', 44, struct fsuuid)
> > +#endif
> 
> Will this be ext4 only ioctl? If this C program is only used for ext4 ioctl, I'd
> like to rename it as "ext4_uuid_ioctl.c" (or something else make it not likes a
> common test program). I think XFS doesn't has this ioctl, for now at least, it
> get/set uuid by another way.

Right now XFS only supports offline changes to UUIDs, but in principle
we could implement this pair of ioctls some day.

> > +
> > +int main(int argc, char **argv)
> > +{
> > +	int error, fd;
> > +	struct fsuuid *fsuuid = NULL;
> > +
> > +	if (argc < 3) {
> > +		fprintf(stderr, "Invalid arguments\n");
> > +		return 1;
> > +	}
> > +
> > +	fd = open(argv[2], O_RDONLY);
> > +	if (!fd) {
> > +		perror(argv[2]);
> > +		return 1;
> > +	}
> > +
> > +	fsuuid = malloc(sizeof(*fsuuid) + sizeof(uuid_t));
> > +	if (!fsuuid) {
> > +		perror("malloc");
> > +		return 1;
> > +	}
> > +	fsuuid->fu_len = sizeof(uuid_t);
> > +
> > +	if (strcmp(argv[1], "get") == 0) {
> > +		uuid_t uuid;
> > +		char uuid_str[37];
> > +
> > +		if (ioctl(fd, EXT4_IOC_GETFSUUID, fsuuid)) {
> > +			fprintf(stderr, "%s while trying to get fs uuid\n",
> > +				strerror(errno));
> > +			return 1;
> > +		}
> > +
> > +		memcpy(&uuid, fsuuid->fu_uuid, sizeof(uuid));
> > +		uuid_unparse(uuid, uuid_str);
> > +		printf("%s", uuid_str);
> > +	} else if (strcmp(argv[1], "set") == 0) {
> > +		uuid_t uuid;
> > +
> > +		if (argc != 4) {
> > +			fprintf(stderr, "UUID argument missing.\n");
> > +			return 1;
> > +		}
> > +
> > +		error = uuid_parse(argv[3], uuid);
> > +		if (error < 0) {
> > +			fprintf(stderr, "Invalid UUID. The UUID should be in "
> > +				"canonical format. Example: "
> > +				"8c628557-6987-42b2-ba16-b7cc79ddfb43\n");
> > +			return 1;
> > +		}
> > +
> > +		memcpy(&fsuuid->fu_uuid, uuid, sizeof(uuid));
> > +		fsuuid->fu_flags = EXT4_IOC_SETFSUUID_FLAG_BLOCKING;
> > +		if (ioctl(fd, EXT4_IOC_SETFSUUID, fsuuid)) {
> > +			fprintf(stderr, "%s while trying to set fs uuid\n",
> > +				strerror(errno));
> > +			return 1;
> > +		}
> > +	} else {
> > +		fprintf(stderr, "Invalid command\n");
> > +		return 1;
> > +	}
> > +
> > +	return 0;
> > +}
> > diff --git a/tests/ext4/056 b/tests/ext4/056
> > new file mode 100755
> > index 00000000..ebefb136
> > --- /dev/null
> > +++ b/tests/ext4/056
> > @@ -0,0 +1,55 @@
> > +#! /bin/bash
> > +# SPDX-License-Identifier: GPL-2.0
> > +#
> > +# Test the set/get UUID ioctl.
> > +#
> > +
> > +. ./common/preamble
> > +_begin_fstest auto ioctl
> > +
> > +# Import common functions.
> > +. ./common/filter
> > +
> > +# real QA test starts here
> > +_supported_fs ext4
> > +_require_scratch
> > +_require_test_program uuid_ioctl
> > +_require_command $UUIDGEN_PROG uuidgen
> > +
> > +UUID_IOCTL=$here/src/uuid_ioctl
> > +
> > +# if the ioctl is not supported by the kernel, then skip test.
> > +current_uuid=$($UUID_IOCTL get $SCRATCH_MNT 2>&1)
> > +if [[ "$current_uuid" =~ ^Inappropriate[[:space:]]ioctl ]]; then
> > +        _notrun "UUID ioctls are not supported by kernel."
> > +fi
> > +
> > +# Create filesystem and mount
> 
> Rather than this such obvious comment description, explain why the below
> "metadata_csum_seed" is necessary?
> 
> > +_scratch_mkfs_ext4 -O metadata_csum_seed >> $seqres.full 2>&1
> > +_scratch_mount >> $seqres.full
> > +
> > +# Begin fsstress while modifying UUID
> > +fsstress_args=$(_scale_fsstress_args -d $SCRATCH_MNT -p 15 -n 999999)
> > +$FSSTRESS_PROG $fsstress_args > /dev/null 2>&1 &
> > +fsstress_pid=$!
> > +
> > +for n in $(seq 1 20); do
> > +        new_uuid=$($UUIDGEN_PROG)
> > +
> > +        echo "Setting UUID to ${new_uuid}" >> $seqres.full 2>&1
> > +        $UUID_IOCTL set $SCRATCH_MNT $new_uuid
> > +
> > +        current_uuid=$($UUID_IOCTL get $SCRATCH_MNT)
> > +        echo "$UUID_IOCTL get $SCARTCH_MNT: $current_uuid" >> $seqres.full 2>&1
> > +        if [[ "$current_uuid" != "$new_uuid" ]]; then
> > +                echo "Current UUID ($current_uuid) does not equal what "
> > +                "was sent with the ioctl ($new_uuid)"
> 
> Can these two lines really be printed?

Let's hope not.

> > +        fi
> > +done
> > +
> > +# success, all done
> > +echo "Silence is golden"
> > +kill $fsstress_pid >/dev/null 2>&1
> > +wait
> 
> We can't make sure the testing will exit normally, so better to make sure the
> fsstress processes are killed in _cleanup(). You can refer to generic/390.

Might be time to turn that (run fsstress/fsx in the background and kill it
later) into a set of common helper functions?

(Or just implement this fix now and we'll revisit it whenever I get
around to submitting the "fsstress vs. online fsck" series[1].)

--D

[1] https://git.kernel.org/pub/scm/linux/kernel/git/djwong/xfstests-dev.git/log/?h=refactor-fsmap-stress

> Thanks,
> Zorro
> 
> > +status=0
> > +exit
> > diff --git a/tests/ext4/056.out b/tests/ext4/056.out
> > new file mode 100644
> > index 00000000..6142fcd2
> > --- /dev/null
> > +++ b/tests/ext4/056.out
> > @@ -0,0 +1,2 @@
> > +QA output created by 056
> > +Silence is golden
> > -- 
> > 2.37.0.rc0.161.g10f37bed90-goog
> > 
>