[PATCH] ovl: Prevent rw remount when it should be ro mount

[PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

Overlayfs should be mounted read-only when upper-fs is read-only or nonexistent.
But now it can be remounted read-write and this can cause kernel panic.
So we should prevent read-write remount when the above situation happens.

Signed-off-by: Seunghun Lee <waydi1@gmail.com>
---
 fs/overlayfs/super.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index 84f3144..8944651 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -522,10 +522,21 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
 	return 0;
 }
 
+static int ovl_remount(struct super_block *sb, int *flags, char *data)
+{
+	struct ovl_fs *ufs = sb->s_fs_info;
+
+	if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
+		*flags |= MS_RDONLY;
+
+	return 0;
+}
+
 static const struct super_operations ovl_super_operations = {
 	.put_super	= ovl_put_super,
 	.statfs		= ovl_statfs,
 	.show_options	= ovl_show_options,
+	.remount_fs	= ovl_remount,
 };
 
 enum {
-- 
2.1.3

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[hujianyang]

On 2015/1/3 1:26, Seunghun Lee wrote:
> Overlayfs should be mounted read-only when upper-fs is read-only or nonexistent.
> But now it can be remounted read-write and this can cause kernel panic.
> So we should prevent read-write remount when the above situation happens.
> 
> Signed-off-by: Seunghun Lee <waydi1@gmail.com>
> ---
>  fs/overlayfs/super.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)
> 
> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
> index 84f3144..8944651 100644
> --- a/fs/overlayfs/super.c
> +++ b/fs/overlayfs/super.c
> @@ -522,10 +522,21 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>  	return 0;
>  }
>  
> +static int ovl_remount(struct super_block *sb, int *flags, char *data)
> +{
> +	struct ovl_fs *ufs = sb->s_fs_info;
> +
> +	if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
> +		*flags |= MS_RDONLY;
> +
> +	return 0;
> +}
> +
>  static const struct super_operations ovl_super_operations = {
>  	.put_super	= ovl_put_super,
>  	.statfs		= ovl_statfs,
>  	.show_options	= ovl_show_options,
> +	.remount_fs	= ovl_remount,
>  };
>  
>  enum {
> 

I think this exporting of .remount_fs may allow people in userspace
have the ability to remount a filesystem with a new set of mounting
options. Your new adding function do nothing with the passing in
parameters.

I'm not sure if it could be competent for remount case.

Add Cc linux-unionfs.

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[hujianyang]

On 2015/1/3 1:26, Seunghun Lee wrote:
> Overlayfs should be mounted read-only when upper-fs is read-only or nonexistent.
> But now it can be remounted read-write and this can cause kernel panic.
> So we should prevent read-write remount when the above situation happens.
> 
> Signed-off-by: Seunghun Lee <waydi1@gmail.com>
> ---
>  fs/overlayfs/super.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)
> 
> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
> index 84f3144..8944651 100644
> --- a/fs/overlayfs/super.c
> +++ b/fs/overlayfs/super.c
> @@ -522,10 +522,21 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>  	return 0;
>  }
>  
> +static int ovl_remount(struct super_block *sb, int *flags, char *data)
> +{
> +	struct ovl_fs *ufs = sb->s_fs_info;
> +
> +	if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
> +		*flags |= MS_RDONLY;
> +
> +	return 0;
> +}
> +
>  static const struct super_operations ovl_super_operations = {
>  	.put_super	= ovl_put_super,
>  	.statfs		= ovl_statfs,
>  	.show_options	= ovl_show_options,
> +	.remount_fs	= ovl_remount,
>  };
>  
>  enum {
> 

I think this exporting of .remount_fs may allow people in userspace
have the ability to remount a filesystem with a new set of mounting
options. Your new adding function do nothing with the passing in
parameters.

I'm not sure if it could be competent for remount case.

Add Cc linux-unionfs.

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

On 01/04/2015 11:59 AM, hujianyang wrote:
> I think this exporting of .remount_fs may allow people in userspace have the ability to remount a filesystem with a new set of mounting options. Your new adding function do nothing with the passing in parameters. I'm not sure if it could be competent for remount case. Add Cc linux-unionfs. 
Hi hujianyang,

I think it makes no difference whether .remount_fs is exported or not,
except in the read-write mount case.

And the patch's use case is below:

Before patch:
root@qemux86:~# mount -t overlay overlay -olowerdir=lower:lower2 merged
mount: warning: merged seems to be mounted read-only.
root@qemux86:~# mount | grep overlay
overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
root@qemux86:~# mount -o remount,rw merged
root@qemux86:~# mount | grep overlay
overlay on /home/root/merged type overlay (rw,relatime,lowerdir=lower:lower2)
root@qemux86:~# echo hi > merged/hi

[   95.906739] BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
[   95.907172] IP: [<ffffffff8117e0b6>] mnt_want_write+0x16/0x50
[   95.907172] PGD 1e34e067 PUD 1e20a067 PMD 0
[   95.907172] Oops: 0000 [#1] SMP
[   95.907172] Modules linked in:
[   95.907172] CPU: 0 PID: 1358 Comm: sh Not tainted 3.19.0-rc2-next-20141231+ #5
[   95.907172] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011
[   95.907172] task: ffff88001dad2d00 ti: ffff88001d4dc000 task.ti: ffff88001d4dc000
[   95.907172] RIP: 0010:[<ffffffff8117e0b6>]  [<ffffffff8117e0b6>] mnt_want_write+0x16/0x50
[   95.907172] RSP: 0000:ffff88001d4dfbf8  EFLAGS: 00000292
[   95.907172] RAX: ffff88001da4ac80 RBX: 0000000000000000 RCX: 0000000000000000
[   95.907172] RDX: 0000000000000001 RSI: 0000000000000001 RDI: 0000000000000000
[   95.907172] RBP: ffff88001d4dfc18 R08: 0000000000000000 R09: 0000000000000000
[   95.907172] R10: 0000000000000000 R11: ffff88000080fb40 R12: 00000000000081a4
[   95.907172] R13: 0000000000000000 R14: 0000000000000000 R15: 0000000000000000
[   95.907172] FS:  0000000000000000(0003) GS:ffff88001fc00000(0063) knlGS:00000000f7755b40
[   95.907172] CS:  0010 DS: 002b ES: 002b CR0: 000000008005003b
[   95.907172] CR2: 0000000000000008 CR3: 000000001d4bc000 CR4: 00000000000006f0
[   95.907172] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[   95.907172] DR3: 0000000000000000 DR6: 0000000000000000 DR7: 0000000000000000
[   95.907172] Stack:
[   95.907172]  0000000000000000 00000003000081a4 ffffffff0000000a ffff88000080fb40
[   95.907172]  ffff88001d4dfc28 ffffffff8128ca97 ffff88001d4dfc68 ffffffff8128f1e0
[   95.907172]  ffff88001dd55c70 ffff88000080fb40 0000000000000000 ffff88001dd55c70
[   95.907172] Call Trace:
[   95.907172]  [<ffffffff8128ca97>] ovl_want_write+0x17/0x20
[   95.907172]  [<ffffffff8128f1e0>] ovl_create_object+0x20/0x60
[   95.907172]  [<ffffffff8128f2be>] ovl_create+0x1e/0x20
[   95.907172]  [<ffffffff8116aabd>] vfs_create+0xcd/0x130
[   95.907172]  [<ffffffff8116d572>] do_last+0x962/0x1110
[   95.907172]  [<ffffffff8116b97c>] ? path_init+0xbc/0x450
[   95.907172]  [<ffffffff8116dd9f>] path_openat+0x7f/0x620
[   95.907172]  [<ffffffff81136aa0>] ? handle_mm_fault+0x5e0/0xa30
[   95.907172]  [<ffffffff8116fd05>] do_filp_open+0x35/0x90
[   95.907172]  [<ffffffff8116ecda>] ? getname_flags+0x4a/0x1a0
[   95.907172]  [<ffffffff8117bcdd>] ? __alloc_fd+0x7d/0x120
[   95.907172]  [<ffffffff8115ea23>] do_sys_open+0x123/0x220
[   95.907172]  [<ffffffff811aadc6>] compat_SyS_open+0x16/0x20
[   95.907172]  [<ffffffff8184ea89>] ia32_do_call+0x13/0x13
[   95.907172] Code: c3 0f 1f 40 00 48 8b 47 28 65 ff 48 04 b8 e2 ff ff ff 5d c3 90 55 ba 01 00 00 00 be 01 00 00 00 48 89 e5 53 48 89 fb 48 83 ec 18 <48> 8b 7f 08 e8 31 3c fe ff 48 89 df e8 79 ff ff ff 85 c0 74 14
[   95.907172] RIP  [<ffffffff8117e0b6>] mnt_want_write+0x16/0x50
[   95.907172]  RSP <ffff88001d4dfbf8>
[   95.907172] CR2: 0000000000000008
[   95.919596] ---[ end trace 770a329b637fe67d ]---

After patch:
root@qemux86:~# mount -t overlay overlay -olowerdir=lower:lower2 merged
mount: warning: merged seems to be mounted read-only.
root@qemux86:~# mount | grep overlay
overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
root@qemux86:~# mount -o remount,rw merged
mount: warning: /home/root/merged seems to be mounted read-only.
root@qemux86:~# mount | grep overlay
overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
root@qemux86:~# echo hi > merged/hi
-sh: merged/hi: Read-only file system
root@qemux86:~#

If what I think is incorrect, please let me know.

Thanks.

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[hujianyang]

Hi,

There maybe some misunderstandings here. I think your patch really
fix an important problem, but not in correct way.

On 2015/1/6 22:02, Seunghun Lee wrote:
> 
> After patch:
> root@qemux86:~# mount -t overlay overlay -olowerdir=lower:lower2 merged
> mount: warning: merged seems to be mounted read-only.
> root@qemux86:~# mount | grep overlay
> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
> root@qemux86:~# mount -o remount,rw merged
> mount: warning: /home/root/merged seems to be mounted read-only.
> root@qemux86:~# mount | grep overlay
> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
> root@qemux86:~# echo hi > merged/hi
> -sh: merged/hi: Read-only file system
> root@qemux86:~#
> 

If users want a rw mount, can we give them a ro mount? I think it's
wrong, .remount_fs should refuse this request.

So I think your .remount_fs should check both what users in userpace
want and what kernel can offer, then realize legal requests and
refuse illegal requests. Not changing the requests from users.

Further more, can we replace upper/lower/work directories or mount
point by this .remount_fs?

If you want to export a new function, I think you should considering
more about these.

Thanks,
Hu

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[hujianyang]

Hi,

There maybe some misunderstandings here. I think your patch really
fix an important problem, but not in correct way.

On 2015/1/6 22:02, Seunghun Lee wrote:
> 
> After patch:
> root@qemux86:~# mount -t overlay overlay -olowerdir=lower:lower2 merged
> mount: warning: merged seems to be mounted read-only.
> root@qemux86:~# mount | grep overlay
> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
> root@qemux86:~# mount -o remount,rw merged
> mount: warning: /home/root/merged seems to be mounted read-only.
> root@qemux86:~# mount | grep overlay
> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
> root@qemux86:~# echo hi > merged/hi
> -sh: merged/hi: Read-only file system
> root@qemux86:~#
> 

If users want a rw mount, can we give them a ro mount? I think it's
wrong, .remount_fs should refuse this request.

So I think your .remount_fs should check both what users in userpace
want and what kernel can offer, then realize legal requests and
refuse illegal requests. Not changing the requests from users.

Further more, can we replace upper/lower/work directories or mount
point by this .remount_fs?

If you want to export a new function, I think you should considering
more about these.

Thanks,
Hu

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

Hi Hu,

On 01/07/2015 11:56 AM, hujianyang wrote:
> Hi,
>
> There maybe some misunderstandings here. I think your patch really
> fix an important problem, but not in correct way.
>
> On 2015/1/6 22:02, Seunghun Lee wrote:
>> After patch:
>> root@qemux86:~# mount -t overlay overlay -olowerdir=lower:lower2 merged
>> mount: warning: merged seems to be mounted read-only.
>> root@qemux86:~# mount | grep overlay
>> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
>> root@qemux86:~# mount -o remount,rw merged
>> mount: warning: /home/root/merged seems to be mounted read-only.
>> root@qemux86:~# mount | grep overlay
>> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
>> root@qemux86:~# echo hi > merged/hi
>> -sh: merged/hi: Read-only file system
>> root@qemux86:~#
>>
> If users want a rw mount, can we give them a ro mount? I think it's
> wrong, .remount_fs should refuse this request.
>
> So I think your .remount_fs should check both what users in userpace
> want and what kernel can offer, then realize legal requests and
> refuse illegal requests. Not changing the requests from users.
Many file systems just change flags when user requests read-write remount.
(romfs, squashfs, sysv...)
I thought this case is similar above filesystems.
> Further more, can we replace upper/lower/work directories or mount
> point by this .remount_fs?
>
> If you want to export a new function, I think you should considering
> more about these.
>
> Thanks,
> Hu
>
Yes, you are right. However, this patch is a minimal support to
prevent kernel panic when file system is remounted to read-write mode.
And many file systems have remount_fs function of this kind.

I think what you mentioned is can be added later if it is necessary.

Thanks.

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Miklos Szeredi]

On Wed, Jan 7, 2015 at 4:04 PM, Seunghun Lee <waydi1@gmail.com> wrote:
>> On 2015/1/6 22:02, Seunghun Lee wrote:
>>> After patch:
>>> root@qemux86:~# mount -t overlay overlay -olowerdir=lower:lower2 merged
>>> mount: warning: merged seems to be mounted read-only.
>>> root@qemux86:~# mount | grep overlay
>>> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
>>> root@qemux86:~# mount -o remount,rw merged
>>> mount: warning: /home/root/merged seems to be mounted read-only.
>>> root@qemux86:~# mount | grep overlay
>>> overlay on /home/root/merged type overlay (ro,relatime,lowerdir=lower:lower2)
>>> root@qemux86:~# echo hi > merged/hi
>>> -sh: merged/hi: Read-only file system
>>> root@qemux86:~#
>>>
>> If users want a rw mount, can we give them a ro mount? I think it's
>> wrong, .remount_fs should refuse this request.

Yeah.  Applied fixed patch.

Thanks,
Mikos

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

On 01/02/2015 03:33 AM, Sedat Dilek wrote:
> On Thu, Jan 1, 2015 at 4:38 PM, Seunghun Lee <waydi1@gmail.com> wrote:
>> Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
>> But now it can be remounted read-write and this can causes kernel panic.
>> So we should prevent read-write remount when the above situation.
>>
> Cannot say much to the code, but you have some typos in your commit-message.
> Here some corrections... wording, style etc.
>
> ...when upper-fs (with a dash) is read-only (if you write read-write
> below) or nonexisten*t*...
> ...this can cause (without s at the end)...
> ...when the above situation *happens* (missing word)...
>
> - Sedat -
>
>> Signed-off-by: Seunghun Lee <waydi1@gmail.com>
>> ---
>>  fs/overlayfs/super.c | 11 +++++++++++
>>  1 file changed, 11 insertions(+)
>>
>> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
>> index 84f3144..8944651 100644
>> --- a/fs/overlayfs/super.c
>> +++ b/fs/overlayfs/super.c
>> @@ -522,10 +522,21 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>>         return 0;
>>  }
>>
>> +static int ovl_remount(struct super_block *sb, int *flags, char *data)
>> +{
>> +       struct ovl_fs *ufs = sb->s_fs_info;
>> +
>> +       if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
>> +               *flags |= MS_RDONLY;
>> +
>> +       return 0;
>> +}
>> +
>>  static const struct super_operations ovl_super_operations = {
>>         .put_super      = ovl_put_super,
>>         .statfs         = ovl_statfs,
>>         .show_options   = ovl_show_options,
>> +       .remount_fs     = ovl_remount,
>>  };
>>
>>  enum {
>> --
>> 2.1.3
>>
Thank you for your correction!

I will fix it.

Thanks.

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Sedat Dilek]

On Thu, Jan 1, 2015 at 4:38 PM, Seunghun Lee <waydi1@gmail.com> wrote:
> Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
> But now it can be remounted read-write and this can causes kernel panic.
> So we should prevent read-write remount when the above situation.
>

Cannot say much to the code, but you have some typos in your commit-message.
Here some corrections... wording, style etc.

...when upper-fs (with a dash) is read-only (if you write read-write
below) or nonexisten*t*...
...this can cause (without s at the end)...
...when the above situation *happens* (missing word)...

- Sedat -

> Signed-off-by: Seunghun Lee <waydi1@gmail.com>
> ---
>  fs/overlayfs/super.c | 11 +++++++++++
>  1 file changed, 11 insertions(+)
>
> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
> index 84f3144..8944651 100644
> --- a/fs/overlayfs/super.c
> +++ b/fs/overlayfs/super.c
> @@ -522,10 +522,21 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>         return 0;
>  }
>
> +static int ovl_remount(struct super_block *sb, int *flags, char *data)
> +{
> +       struct ovl_fs *ufs = sb->s_fs_info;
> +
> +       if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
> +               *flags |= MS_RDONLY;
> +
> +       return 0;
> +}
> +
>  static const struct super_operations ovl_super_operations = {
>         .put_super      = ovl_put_super,
>         .statfs         = ovl_statfs,
>         .show_options   = ovl_show_options,
> +       .remount_fs     = ovl_remount,
>  };
>
>  enum {
> --
> 2.1.3
>

[PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
But now it can be remounted read-write and this can causes kernel panic.
So we should prevent read-write remount when the above situation.

Signed-off-by: Seunghun Lee <waydi1@gmail.com>
---
 fs/overlayfs/super.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index 84f3144..8944651 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -522,10 +522,21 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
 	return 0;
 }
 
+static int ovl_remount(struct super_block *sb, int *flags, char *data)
+{
+	struct ovl_fs *ufs = sb->s_fs_info;
+
+	if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
+		*flags |= MS_RDONLY;
+
+	return 0;
+}
+
 static const struct super_operations ovl_super_operations = {
 	.put_super	= ovl_put_super,
 	.statfs		= ovl_statfs,
 	.show_options	= ovl_show_options,
+	.remount_fs	= ovl_remount,
 };
 
 enum {
-- 
2.1.3

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

On 01/02/2015 12:24 AM, Sedat Dilek wrote:
> On Thu, Jan 1, 2015 at 4:13 PM, Richard Weinberger
> <richard.weinberger@gmail.com> wrote:
>> On Thu, Jan 1, 2015 at 6:07 AM, Seunghun Lee <waydi1@gmail.com> wrote:
>>> Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
>>> But now it can be remounted read-write and this can causes kernel panic.
>>> So we should prevent read-write remount when the above situation.
>>>
>>> Signed-off-by: Seunghun Lee <waydi1@gmail.com>
>> This patch makes zero sense, did you use the wrong diff?
>>
> Hmm, was also my 1st thoughts when I saw squashfs_remount() in overlayfs code.
>
> - Sedat -
>
>>> ---
>>>  fs/overlayfs/super.c | 8 ++++++++
>>>  1 file changed, 8 insertions(+)
>>>
>>> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
>>> index 84f3144..1faa5e5 100644
>>> --- a/fs/overlayfs/super.c
>>> +++ b/fs/overlayfs/super.c
>>> @@ -522,6 +522,14 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>>>         return 0;
>>>  }
>>>
>>> +static int squashfs_remount(struct super_block *sb, int *flags, char *data)
>>> +{
>>> +       if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
>>> +               *flags |= MS_RDONLY;
>>> +
>>> +       return 0;
>>> +}
>>> +
>>>  static const struct super_operations ovl_super_operations = {
>>>         .put_super      = ovl_put_super,
>>>         .statfs         = ovl_statfs,
>>> --
>>> 2.1.3
>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>> Please read the FAQ at  http://www.tux.org/lkml/
>>
>>
>> --
>> Thanks,
>> //richard
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
Sorry, I sent wrong patch.

I will resend it.

Thanks.

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Sedat Dilek]

On Thu, Jan 1, 2015 at 4:13 PM, Richard Weinberger
<richard.weinberger@gmail.com> wrote:
> On Thu, Jan 1, 2015 at 6:07 AM, Seunghun Lee <waydi1@gmail.com> wrote:
>> Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
>> But now it can be remounted read-write and this can causes kernel panic.
>> So we should prevent read-write remount when the above situation.
>>
>> Signed-off-by: Seunghun Lee <waydi1@gmail.com>
>
> This patch makes zero sense, did you use the wrong diff?
>

Hmm, was also my 1st thoughts when I saw squashfs_remount() in overlayfs code.

- Sedat -

>> ---
>>  fs/overlayfs/super.c | 8 ++++++++
>>  1 file changed, 8 insertions(+)
>>
>> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
>> index 84f3144..1faa5e5 100644
>> --- a/fs/overlayfs/super.c
>> +++ b/fs/overlayfs/super.c
>> @@ -522,6 +522,14 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>>         return 0;
>>  }
>>
>> +static int squashfs_remount(struct super_block *sb, int *flags, char *data)
>> +{
>> +       if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
>> +               *flags |= MS_RDONLY;
>> +
>> +       return 0;
>> +}
>> +
>>  static const struct super_operations ovl_super_operations = {
>>         .put_super      = ovl_put_super,
>>         .statfs         = ovl_statfs,
>> --
>> 2.1.3
>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> Please read the FAQ at  http://www.tux.org/lkml/
>
>
>
> --
> Thanks,
> //richard
> --
> To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH] ovl: Prevent rw remount when it should be ro mount

[Richard Weinberger]

On Thu, Jan 1, 2015 at 6:07 AM, Seunghun Lee <waydi1@gmail.com> wrote:
> Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
> But now it can be remounted read-write and this can causes kernel panic.
> So we should prevent read-write remount when the above situation.
>
> Signed-off-by: Seunghun Lee <waydi1@gmail.com>

This patch makes zero sense, did you use the wrong diff?

> ---
>  fs/overlayfs/super.c | 8 ++++++++
>  1 file changed, 8 insertions(+)
>
> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
> index 84f3144..1faa5e5 100644
> --- a/fs/overlayfs/super.c
> +++ b/fs/overlayfs/super.c
> @@ -522,6 +522,14 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
>         return 0;
>  }
>
> +static int squashfs_remount(struct super_block *sb, int *flags, char *data)
> +{
> +       if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
> +               *flags |= MS_RDONLY;
> +
> +       return 0;
> +}
> +
>  static const struct super_operations ovl_super_operations = {
>         .put_super      = ovl_put_super,
>         .statfs         = ovl_statfs,
> --
> 2.1.3
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/



-- 
Thanks,
//richard

[PATCH] ovl: Prevent rw remount when it should be ro mount

[Seunghun Lee]

Overlayfs should be mounted read-only when upper fs is r/o or nonexistend.
But now it can be remounted read-write and this can causes kernel panic.
So we should prevent read-write remount when the above situation.

Signed-off-by: Seunghun Lee <waydi1@gmail.com>
---
 fs/overlayfs/super.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c
index 84f3144..1faa5e5 100644
--- a/fs/overlayfs/super.c
+++ b/fs/overlayfs/super.c
@@ -522,6 +522,14 @@ static int ovl_show_options(struct seq_file *m, struct dentry *dentry)
 	return 0;
 }
 
+static int squashfs_remount(struct super_block *sb, int *flags, char *data)
+{
+	if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))
+		*flags |= MS_RDONLY;
+
+	return 0;
+}
+
 static const struct super_operations ovl_super_operations = {
 	.put_super	= ovl_put_super,
 	.statfs		= ovl_statfs,
-- 
2.1.3