* [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs
@ 2021-11-30 4:04 zhanchengbin
2021-12-01 5:05 ` zhanchengbin
2021-12-21 20:57 ` Theodore Ts'o
0 siblings, 2 replies; 4+ messages in thread
From: zhanchengbin @ 2021-11-30 4:04 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: linux-ext4, liuzhiqiang26, linfeilong
If there is a tmpfs with the same name as the disk, and mount before the
disk,example:
/dev/sdd /root/tmp tmpfs rw,seclabel,relatime 0 0
/dev/sdd /root/mnt ext4 rw,seclabel,relatime 0 0
Create a hard link /dev/sdd-ln for the disk and resize2fs it.The items in
/proc/mounts are traversed, When you get to tmpfs,file!=mnt->mnt_fsname,
Therefore, the stat(mnt->mnt_fsname, &st_buf) branch is used,However, the
values of file_rdev and st_buf.st_rdev are the same.As a result, the system
mistakenly considers that disk is mounted to /root/tmp.As a result,resize2fs
fails.
mkdir /root/tmp
mkdir /root/mnt
mkfs.ext4 -F -b 1024 -E "resize=10000000" /dev/sdd 32768
mount -t tmpfs /dev/sdd /root/tmp
mount /dev/sdd /root/mnt
ln /dev/sdd /dev/sdd-ln
resize2fs /dev/sdd-ln 6G
Signed-off-by: zhanchengbin <zhanchengbin1@huawei.com>
Signed-off-by: guiyao@huawei.com
---
lib/ext2fs/ismounted.c | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/lib/ext2fs/ismounted.c b/lib/ext2fs/ismounted.c
index aee7d726..463a82a6 100644
--- a/lib/ext2fs/ismounted.c
+++ b/lib/ext2fs/ismounted.c
@@ -98,6 +98,7 @@ static errcode_t check_mntent_file(const char
*mtab_file, const char *file,
{
struct mntent *mnt;
struct stat st_buf;
+ struct stat dir_st_buf;
errcode_t retval = 0;
dev_t file_dev=0, file_rdev=0;
ino_t file_ino=0;
@@ -144,8 +145,12 @@ static errcode_t check_mntent_file(const char
*mtab_file, const char *file,
if (stat(mnt->mnt_fsname, &st_buf) == 0) {
if (ext2fsP_is_disk_device(st_buf.st_mode)) {
#ifndef __GNU__
- if (file_rdev && (file_rdev == st_buf.st_rdev))
- break;
+ if (file_rdev && (file_rdev == st_buf.st_rdev)) {
+ if (stat(mnt->mnt_dir, &dir_st_buf) != 0)
+ continue;
+ if (file_rdev == dir_st_buf.st_dev)
+ break;
+ }
if (check_loop_mounted(mnt->mnt_fsname,
st_buf.st_rdev, file_dev,
file_ino) == 1)
--
2.23.0
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs
2021-11-30 4:04 [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs zhanchengbin
@ 2021-12-01 5:05 ` zhanchengbin
2021-12-21 20:57 ` Theodore Ts'o
1 sibling, 0 replies; 4+ messages in thread
From: zhanchengbin @ 2021-12-01 5:05 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: linux-ext4, liuzhiqiang26, linfeilong
If there is a tmpfs with the same name as the disk, and mount before the
disk,example:
/dev/sdd /root/tmp tmpfs rw,seclabel,relatime 0 0
/dev/sdd /root/mnt ext4 rw,seclabel,relatime 0 0
Create a hard link /dev/sdd-ln for the disk and resize2fs it.The items in
/proc/mounts are traversed, When you get to tmpfs,file!=mnt->mnt_fsname,
Therefore, the stat(mnt->mnt_fsname, &st_buf) branch is used,However, the
values of file_rdev and st_buf.st_rdev are the same.As a result, the system
mistakenly considers that disk is mounted to /root/tmp.As a result,resize2fs
fails.
mkdir /root/tmp
mkdir /root/mnt
mkfs.ext4 -F -b 1024 -E "resize=10000000" /dev/sdd 32768
mount -t tmpfs /dev/sdd /root/tmp
mount /dev/sdd /root/mnt
ln /dev/sdd /dev/sdd-ln
resize2fs /dev/sdd-ln 6G
Signed-off-by: zhanchengbin <zhanchengbin1@huawei.com>
Signed-off-by: guiyao <guiyao@huawei.com>
---
lib/ext2fs/ismounted.c | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/lib/ext2fs/ismounted.c b/lib/ext2fs/ismounted.c
index aee7d726..463a82a6 100644
--- a/lib/ext2fs/ismounted.c
+++ b/lib/ext2fs/ismounted.c
@@ -98,6 +98,7 @@ static errcode_t check_mntent_file(const char
*mtab_file, const char *file,
{
struct mntent *mnt;
struct stat st_buf;
+ struct stat dir_st_buf;
errcode_t retval = 0;
dev_t file_dev=0, file_rdev=0;
ino_t file_ino=0;
@@ -144,8 +145,12 @@ static errcode_t check_mntent_file(const char
*mtab_file, const char *file,
if (stat(mnt->mnt_fsname, &st_buf) == 0) {
if (ext2fsP_is_disk_device(st_buf.st_mode)) {
#ifndef __GNU__
- if (file_rdev && (file_rdev == st_buf.st_rdev))
- break;
+ if (file_rdev && (file_rdev == st_buf.st_rdev)) {
+ if (stat(mnt->mnt_dir, &dir_st_buf) != 0)
+ continue;
+ if (file_rdev == dir_st_buf.st_dev)
+ break;
+ }
if (check_loop_mounted(mnt->mnt_fsname,
st_buf.st_rdev, file_dev,
file_ino) == 1)
--
2.23.0
在 2021/11/30 12:04, zhanchengbin 写道:
> If there is a tmpfs with the same name as the disk, and mount before the
> disk,example:
> /dev/sdd /root/tmp tmpfs rw,seclabel,relatime 0 0
> /dev/sdd /root/mnt ext4 rw,seclabel,relatime 0 0
>
> Create a hard link /dev/sdd-ln for the disk and resize2fs it.The items in
> /proc/mounts are traversed, When you get to tmpfs,file!=mnt->mnt_fsname,
> Therefore, the stat(mnt->mnt_fsname, &st_buf) branch is used,However, the
> values of file_rdev and st_buf.st_rdev are the same.As a result, the system
> mistakenly considers that disk is mounted to /root/tmp.As a
> result,resize2fs
> fails.
>
> mkdir /root/tmp
> mkdir /root/mnt
> mkfs.ext4 -F -b 1024 -E "resize=10000000" /dev/sdd 32768
> mount -t tmpfs /dev/sdd /root/tmp
> mount /dev/sdd /root/mnt
> ln /dev/sdd /dev/sdd-ln
> resize2fs /dev/sdd-ln 6G
>
> Signed-off-by: zhanchengbin <zhanchengbin1@huawei.com>
> Signed-off-by: guiyao@huawei.com
> ---
> lib/ext2fs/ismounted.c | 9 +++++++--
> 1 file changed, 7 insertions(+), 2 deletions(-)
>
> diff --git a/lib/ext2fs/ismounted.c b/lib/ext2fs/ismounted.c
> index aee7d726..463a82a6 100644
> --- a/lib/ext2fs/ismounted.c
> +++ b/lib/ext2fs/ismounted.c
> @@ -98,6 +98,7 @@ static errcode_t check_mntent_file(const char
> *mtab_file, const char *file,
> {
> struct mntent *mnt;
> struct stat st_buf;
> + struct stat dir_st_buf;
> errcode_t retval = 0;
> dev_t file_dev=0, file_rdev=0;
> ino_t file_ino=0;
> @@ -144,8 +145,12 @@ static errcode_t check_mntent_file(const char
> *mtab_file, const char *file,
> if (stat(mnt->mnt_fsname, &st_buf) == 0) {
> if (ext2fsP_is_disk_device(st_buf.st_mode)) {
> #ifndef __GNU__
> - if (file_rdev && (file_rdev == st_buf.st_rdev))
> - break;
> + if (file_rdev && (file_rdev == st_buf.st_rdev)) {
> + if (stat(mnt->mnt_dir, &dir_st_buf) != 0)
> + continue;
> + if (file_rdev == dir_st_buf.st_dev)
> + break;
> + }
> if (check_loop_mounted(mnt->mnt_fsname,
> st_buf.st_rdev, file_dev,
> file_ino) == 1)
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs
2021-11-30 4:04 [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs zhanchengbin
2021-12-01 5:05 ` zhanchengbin
@ 2021-12-21 20:57 ` Theodore Ts'o
2021-12-31 8:15 ` zhanchengbin
1 sibling, 1 reply; 4+ messages in thread
From: Theodore Ts'o @ 2021-12-21 20:57 UTC (permalink / raw)
To: zhanchengbin; +Cc: linux-ext4, liuzhiqiang26, linfeilong
On Tue, Nov 30, 2021 at 12:04:48PM +0800, zhanchengbin wrote:
> If there is a tmpfs with the same name as the disk, and mount before the
> disk,example:
> /dev/sdd /root/tmp tmpfs rw,seclabel,relatime 0 0
> /dev/sdd /root/mnt ext4 rw,seclabel,relatime 0 0
This should already be fixed e2fsprogs 1.45.5+ via this commit:
commit ea4d53b7b9079fd6e2cc34cf569a993a183bfbd2
Author: Theodore Ts'o <tytso@mit.edu>
Date: Sun Nov 10 12:11:49 2019 -0500
libext2fs/ismounted.c: check device id in advance to skip false device names
If there is a trickster which tries to use device names as the mount
device for pseudo-file systems, the resulting /proc/mounts can confuse
ext2fs_check_mount_point(). (So far as I can tell, there's no good
reason to do this, but sysadmins do the darnest things.)
An example of this might be the following /proc/mounts excerpt:
/dev/sdb /mnt2 tmpfs rw,relatime 0 0
/dev/sdb /mnt ext4 rw,relatime 0 0
This is created via "mount -t tmpfs /dev/sdb /mnt2" followed via
"mount -t ext4 /dev/sdb /mnt". (Normally, a sane mount of tmpfs would
use something like "mount -t tmpfs tmpfs /mnt2".)
Fix this by double checking the st_rdev of the claimed mountpoint and
match it with the dev_t of the device. (Note that the GNU HURD
doesn't support st_rdev, so we can't solve this problem for the HURD.)
Reported-by: GuiYao <guiyao@huawei.com>
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
I've tested via tst_ismounted and I can't replicate the issue you've described.
% cd /build/e2fsprogs-maint/lib/ext2fs
% make tst_ismounted
% sudo ./tst_ismounted /dev/dm-7
Bogus entry in /proc/mounts! (/dev/dm-7 is not mounted on /root/tmp)
Device /dev/dm-7 reports flags 11
/dev/dm-7 is apparently in use.
/dev/dm-7 is mounted.
/dev/dm-7 is mounted on /root/mnt.
Cheers,
- Ted
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs
2021-12-21 20:57 ` Theodore Ts'o
@ 2021-12-31 8:15 ` zhanchengbin
0 siblings, 0 replies; 4+ messages in thread
From: zhanchengbin @ 2021-12-31 8:15 UTC (permalink / raw)
To: Theodore Ts'o; +Cc: linux-ext4, liuzhiqiang26, linfeilong
To clarify, commit ea4d53b7b9079fd6e2cc34cf569a993a183bfbd2 does
solve the problem for mounting a disk with the same name as file
system "tmpfs". However, after mounting it, "ln" command (creating
a hard link) and "resize2fs" command still report an error.
example:
dev_name="/dev/sdc" (ps: a disk in you self)
mkdir /root/tmp
mkdir /root/mnt
mkfs.ext4 -F -b 1024 -E "resize=10000000" "${dev_name}" 32768
mount -t tmpfs "${dev_name}" /root/tmp
mount "${dev_name}" /root/mnt
ln "${dev_name}" "${dev_name}"-ln
resize2fs "${dev_name}"-ln 6G
You can see the disk is mounted on /root/tmp, but it should be
mounted on /root/mnt actually. I will modify the commit message
and submit the v2 patch to solve this problem.
在 2021/12/22 4:57, Theodore Ts'o 写道:
> On Tue, Nov 30, 2021 at 12:04:48PM +0800, zhanchengbin wrote:
>> If there is a tmpfs with the same name as the disk, and mount before the
>> disk,example:
>> /dev/sdd /root/tmp tmpfs rw,seclabel,relatime 0 0
>> /dev/sdd /root/mnt ext4 rw,seclabel,relatime 0 0
>
> This should already be fixed e2fsprogs 1.45.5+ via this commit:
>
> commit ea4d53b7b9079fd6e2cc34cf569a993a183bfbd2
> Author: Theodore Ts'o <tytso@mit.edu>
> Date: Sun Nov 10 12:11:49 2019 -0500
>
> libext2fs/ismounted.c: check device id in advance to skip false device names
>
> If there is a trickster which tries to use device names as the mount
> device for pseudo-file systems, the resulting /proc/mounts can confuse
> ext2fs_check_mount_point(). (So far as I can tell, there's no good
> reason to do this, but sysadmins do the darnest things.)
>
> An example of this might be the following /proc/mounts excerpt:
>
> /dev/sdb /mnt2 tmpfs rw,relatime 0 0
> /dev/sdb /mnt ext4 rw,relatime 0 0
>
> This is created via "mount -t tmpfs /dev/sdb /mnt2" followed via
> "mount -t ext4 /dev/sdb /mnt". (Normally, a sane mount of tmpfs would
> use something like "mount -t tmpfs tmpfs /mnt2".)
>
> Fix this by double checking the st_rdev of the claimed mountpoint and
> match it with the dev_t of the device. (Note that the GNU HURD
> doesn't support st_rdev, so we can't solve this problem for the HURD.)
>
> Reported-by: GuiYao <guiyao@huawei.com>
> Signed-off-by: Theodore Ts'o <tytso@mit.edu>
>
> I've tested via tst_ismounted and I can't replicate the issue you've described.
>
> % cd /build/e2fsprogs-maint/lib/ext2fs
> % make tst_ismounted
> % sudo ./tst_ismounted /dev/dm-7
> Bogus entry in /proc/mounts! (/dev/dm-7 is not mounted on /root/tmp)
> Device /dev/dm-7 reports flags 11
> /dev/dm-7 is apparently in use.
> /dev/dm-7 is mounted.
> /dev/dm-7 is mounted on /root/mnt.
>
> Cheers,
>
> - Ted
> .
>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2021-12-31 8:15 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-11-30 4:04 [PATCH] resize2fs : resize2fs failed due to the same name of tmpfs zhanchengbin
2021-12-01 5:05 ` zhanchengbin
2021-12-21 20:57 ` Theodore Ts'o
2021-12-31 8:15 ` zhanchengbin
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.