linux-kernel - Re: [PATCH] Rearrange i_flags to be consistent with FS_IOC

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 05 Jul 2010 10:54:07 -0500
From:	Eric Sandeen <sandeen@...hat.com>
To:	David Howells <dhowells@...hat.com>
CC:	viro@...IV.linux.org.uk, linux-fsdevel@...r.kernel.org,
	linux-kernel@...r.kernel.org, linux-ext4@...r.kernel.org
Subject: Re: [PATCH] Rearrange i_flags to be consistent with FS_IOC_GETFLAGS

David Howells wrote:
> Rearrange the constituent bits of i_flags to be consistent with the flag values
> returned by the FS_IOC_GETFLAGS ioctl where flags with common meanings occur.
> Otherwise pack the bits of i_flags as low as possible so that RISC CPUs can
> use small instructions.
> 
> This allows those filesystems that use i_flags (Ext2/3/4) to simplify their
> get/set flags routines.
> 
> Signed-off-by: David Howells <dhowells@...hat.com>

Seems like a good idea to me, although:

> diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c
> index 3675088..da5fd2d 100644
> --- a/fs/ext2/inode.c
> +++ b/fs/ext2/inode.c
> @@ -1259,38 +1259,17 @@ Egdp:
>  
>  void ext2_set_inode_flags(struct inode *inode)
>  {
> -	unsigned int flags = EXT2_I(inode)->i_flags;
> -
> -	inode->i_flags &= ~(S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC);
> -	if (flags & EXT2_SYNC_FL)
> -		inode->i_flags |= S_SYNC;
> -	if (flags & EXT2_APPEND_FL)
> -		inode->i_flags |= S_APPEND;
> -	if (flags & EXT2_IMMUTABLE_FL)
> -		inode->i_flags |= S_IMMUTABLE;
> -	if (flags & EXT2_NOATIME_FL)
> -		inode->i_flags |= S_NOATIME;
> -	if (flags & EXT2_DIRSYNC_FL)
> -		inode->i_flags |= S_DIRSYNC;

Maybe a stern comment should be here about the relationship between
this and the flag definitions...

> +	unsigned int flags = EXT2_I(inode)->i_flags & S_FS_IOC_FLAGS;
> +
> +	inode->i_flags = (inode->i_flags & ~S_FS_IOC_FLAGS) | flags;
>  }
>  

...

> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index 48616db..5808b9b 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h

...

> +/*
> + * Inode flags - they have no relation to superblock flags now
> + *
> + * Where applicable, flags that are also in FS_IOC_GETFLAGS have the same bit
> + * position
> + */

... a stern comment here about not re-ordering since other fs/*/* code depends
strongly on this order?

"where applicable" sounds like a wish, but now these can't be changed, right?

If somebody rearranges these bad things will happen.

-Eric

> +#define S_DEAD		0x00000001	/* removed, but still open directory */
> +#define S_NOQUOTA	0x00000002	/* Inode is not counted to quota */
> +#define S_NOCMTIME	0x00000004	/* Do not update file c/mtime */
> +#define S_SYNC		0x00000008	/* [FS_SYNC_FL] Writes are synced at once */
> +#define S_IMMUTABLE	0x00000010	/* [FS_IMMUTABLE_FL] Immutable file */
> +#define S_APPEND	0x00000020	/* [FS_APPEND_FL] Append-only file */
> +#define S_SWAPFILE	0x00000040	/* Do not truncate: swapon got its bmaps */
> +#define S_NOATIME	0x00000080	/* [FS_NOATIME_FL] Do not update access times */
> +#define S_PRIVATE	0x00000100	/* Inode is fs-internal */
> +#define S_DIRSYNC	0x00010000	/* [FS_DIRSYNC_FL] Directory mods are synchronous */
> +
> +#define S_FS_IOC_FLAGS	0x000100B8	/* the S_xxx flags that are also FS_xxx_FL flags */
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/