linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ingo Molnar <mingo@elte.hu>
To: Ken Sumrall <ksumrall@android.com>,
	Linus Torvalds <torvalds@linux-foundation.org>
Cc: linux-kernel@vger.kernel.org,
	Alexander Viro <viro@zeniv.linux.org.uk>,
	Christoph Hellwig <hch@lst.de>,
	Andrew Morton <akpm@linux-foundation.org>,
	Jan Kara <jack@suse.cz>, Jens Axboe <axboe@kernel.dk>,
	Matthew Wilcox <matthew@wil.cx>, Eric Paris <eparis@redhat.com>,
	Dave Young <hidave.darkstar@gmail.com>,
	Jiri Slaby <jslaby@suse.cz>, James Morris <jmorris@namei.org>,
	linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH] Syscalls: reboot: Add options to the reboot syscall to remount filesystems ro
Date: Thu, 3 Mar 2011 09:18:41 +0100	[thread overview]
Message-ID: <20110303081841.GA14854@elte.hu> (raw)
In-Reply-To: <1299137483-10975-1-git-send-email-ksumrall@android.com>


* Ken Sumrall <ksumrall@android.com> wrote:

> Add 4 new commands to the reboot system call, that do the same thing as the
> RESTART, HALT, POWER_OFF, and RESTART2 commands, but also remount writable
> filesystems as read-only just before doing what the command normally does.
> Now that Android is using EXT4, and since we don't have a standard init
> setup to unmount filesystems before rebooting, this allows the system to
> reboot with clean filesystems, and also improves boot time as the journal
> does not need to be replayed when mounting the filesystem.
> 
> Signed-off-by: Ken Sumrall <ksumrall@android.com>
> ---
>  fs/super.c             |    9 +++++++++
>  include/linux/fs.h     |    1 +
>  include/linux/reboot.h |    4 ++++
>  kernel/sys.c           |   12 ++++++++++++
>  4 files changed, 26 insertions(+), 0 deletions(-)
> 
> diff --git a/fs/super.c b/fs/super.c
> index 8819e3a..3f39a16 100644
> --- a/fs/super.c
> +++ b/fs/super.c
> @@ -638,6 +638,15 @@ void emergency_remount(void)
>  	}
>  }
>  
> +void emergency_remount_synchronous(void)
> +{
> +	struct work_struct *work;
> +
> +	work = kmalloc(sizeof(*work), GFP_ATOMIC);
> +	if (work)
> +		do_emergency_remount(work);
> +}
> +
>  /*
>   * Unnamed block devices are dummy devices used by virtual
>   * filesystems which don't use real block-devices.  -- jrs
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 63d069b..e48ef0d 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -2110,6 +2110,7 @@ extern int generic_write_sync(struct file *file, loff_t pos, loff_t count);
>  extern void sync_supers(void);
>  extern void emergency_sync(void);
>  extern void emergency_remount(void);
> +extern void emergency_remount_synchronous(void);
>  #ifdef CONFIG_BLOCK
>  extern sector_t bmap(struct inode *, sector_t);
>  #endif
> diff --git a/include/linux/reboot.h b/include/linux/reboot.h
> index 3005d5a..24b185d 100644
> --- a/include/linux/reboot.h
> +++ b/include/linux/reboot.h
> @@ -26,11 +26,15 @@
>   */
>  
>  #define	LINUX_REBOOT_CMD_RESTART	0x01234567
> +#define	LINUX_REBOOT_CMD_RMNT_RESTART	0x12345670
>  #define	LINUX_REBOOT_CMD_HALT		0xCDEF0123
> +#define	LINUX_REBOOT_CMD_RMNT_HALT	0xDEF0123C
>  #define	LINUX_REBOOT_CMD_CAD_ON		0x89ABCDEF
>  #define	LINUX_REBOOT_CMD_CAD_OFF	0x00000000
>  #define	LINUX_REBOOT_CMD_POWER_OFF	0x4321FEDC
> +#define	LINUX_REBOOT_CMD_RMNT_POWER_OFF	0x321FEDC4
>  #define	LINUX_REBOOT_CMD_RESTART2	0xA1B2C3D4
> +#define	LINUX_REBOOT_CMD_RMNT_RESTART2	0x1B2C3D4A
>  #define	LINUX_REBOOT_CMD_SW_SUSPEND	0xD000FCE2
>  #define	LINUX_REBOOT_CMD_KEXEC		0x45584543
>  
> diff --git a/kernel/sys.c b/kernel/sys.c
> index 7f5a0cd..3f474e6 100644
> --- a/kernel/sys.c
> +++ b/kernel/sys.c
> @@ -392,6 +392,9 @@ SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd,
>  	mutex_lock(&reboot_mutex);
>  	switch (cmd) {
>  	case LINUX_REBOOT_CMD_RESTART:
> +	case LINUX_REBOOT_CMD_RMNT_RESTART:
> +		if (cmd == LINUX_REBOOT_CMD_RMNT_RESTART)
> +			emergency_remount_synchronous();
>  		kernel_restart(NULL);
>  		break;
>  
> @@ -404,22 +407,31 @@ SYSCALL_DEFINE4(reboot, int, magic1, int, magic2, unsigned int, cmd,
>  		break;
>  
>  	case LINUX_REBOOT_CMD_HALT:
> +	case LINUX_REBOOT_CMD_RMNT_HALT:
> +		if (cmd == LINUX_REBOOT_CMD_RMNT_HALT)
> +			emergency_remount_synchronous();
>  		kernel_halt();
>  		do_exit(0);
>  		panic("cannot halt");
>  
>  	case LINUX_REBOOT_CMD_POWER_OFF:
> +	case LINUX_REBOOT_CMD_RMNT_POWER_OFF:
> +		if (cmd == LINUX_REBOOT_CMD_RMNT_POWER_OFF)
> +			emergency_remount_synchronous();
>  		kernel_power_off();
>  		do_exit(0);
>  		break;
>  
>  	case LINUX_REBOOT_CMD_RESTART2:
> +	case LINUX_REBOOT_CMD_RMNT_RESTART2:
>  		if (strncpy_from_user(&buffer[0], arg, sizeof(buffer) - 1) < 0) {
>  			ret = -EFAULT;
>  			break;
>  		}
>  		buffer[sizeof(buffer) - 1] = '\0';
>  
> +		if (cmd == LINUX_REBOOT_CMD_RMNT_RESTART2)
> +			emergency_remount_synchronous();
>  		kernel_restart(buffer);
>  		break;

Wouldnt it be *much* simpler to add it as a magic1 variant:

 #define LINUX_REBOOT_UMOUNT_RO   0xfee1deaf

Used as a 'also remount ro please' flag?

That way the whole patch would literally be 2 lines:

	if (magic1 == LINUX_REBOOT_UMOUNT_RO)
		emergency_remount_synchronous();

Note that this also has the advantage that both kexec and suspend reboots could be 
done with emergency umounts, if so desired - and any future reboot variant would be 
supported as well. Your patch left out those other reboot methods.

Thanks,

	Ingo

  reply	other threads:[~2011-03-03  8:19 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-03-03  7:31 [PATCH] Syscalls: reboot: Add options to the reboot syscall to remount filesystems ro Ken Sumrall
2011-03-03  8:18 ` Ingo Molnar [this message]
2011-03-03  8:46 ` Dave Young
2011-03-03  8:49   ` Dave Young
2011-03-03 14:00 ` Christoph Hellwig
2011-03-03 17:45 ` Linus Torvalds
2011-03-03 18:17   ` Linus Torvalds
2011-03-03 18:29     ` Mark Lord
2011-03-03 23:23       ` Ken Sumrall
2011-03-07 11:00     ` Dimitris Papastamos
2011-03-03 18:28   ` Mark Lord
2011-03-03 23:36     ` Linus Torvalds
2011-03-04  2:00       ` Ken Sumrall
2011-03-04  2:33         ` Linus Torvalds
2011-03-04  2:48           ` Mark Lord
2011-03-04  2:55             ` Scott James Remnant
2011-03-05  2:45               ` Mark Lord
2011-03-04  6:55       ` Artem Bityutskiy

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20110303081841.GA14854@elte.hu \
    --to=mingo@elte.hu \
    --cc=akpm@linux-foundation.org \
    --cc=axboe@kernel.dk \
    --cc=eparis@redhat.com \
    --cc=hch@lst.de \
    --cc=hidave.darkstar@gmail.com \
    --cc=jack@suse.cz \
    --cc=jmorris@namei.org \
    --cc=jslaby@suse.cz \
    --cc=ksumrall@android.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=matthew@wil.cx \
    --cc=torvalds@linux-foundation.org \
    --cc=viro@zeniv.linux.org.uk \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).