linux-kernel - Re: [PATCH 4/67] aufs configuration

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:	Fri, 16 May 2008 17:28:16 +0200 (CEST)
From:	Jan Engelhardt <jengelh@...ozas.de>
To:	Junjiro Okajima <hooanon05@...oo.co.jp>
cc:	linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 4/67] aufs configuration


On Friday 2008-05-16 16:32, hooanon05@...oo.co.jp wrote:
>+
>+if AUFS
>+config AUFS_FAKE_DM
>+	bool
>+	depends on AUFS
>+	default n

When you have "if AUFS" you do not need all the "depends on AUFS".

>+config AUFS_ROBR
>+	bool "Aufs as an readonly branch of another aufs"
>+	depends on AUFS
>+	default n
>+	help
>+	If you want make your aufs to be a part of another aufs, then
>+	enable this option. In other words, you can specify your aufs
>+	path in 'br:' mount option for another aufs, but cannot
>+	specify 'rw' as the branch permission.
>+	It will damage the performance.
>+	See detail in aufs.5.

IIRC, unionfs did not need such an option, so why does AUFS?

>+config AUFS_DLGT
>+	bool "Delegate the internal branch access the kernel thread"
>+	depends on AUFS
>+	default n

I do not think you need "default n" either as that should be the
implicit default.

>+config AUFS_SPLICE_PATCH
>+	bool
>+	depends on AUFS
>+	default y

This should just go away if it is always selected anyway.

>+config AUFS_BR_NFS
>+	bool
>+	depends on AUFS
>+	depends on NFS_FS
>+	depends on !AUFS_FAKE_DM
>+	default n if (!AUFS_LHASH_PATCH || AUFS_FAKE_DM)
>+	default y

One AUFS_FAKE_DM is redundant here. It should probably be just

	depends on NFS_FS && !AUFS_FAKE_DM && AUFS_LHASH_PATCH

>+config AUFS_BR_XFS
>+	bool
>+	depends on AUFS
>+	depends on XFS_FS
>+	default y

Same here, just depends on XFS_FS.

>+config AUFS_WORKAROUND_FUSE
>+	bool "Special handling for FUSE-based filesystem"
>+	depends on AUFS && FUSE_FS
>+	default n
>+	help
>+	A FUSE-based filesystem may not initialize its inode
>+	attributes and the FUSE developer thinks the inode attributes
>+	in a positive dentry which is returned by VFS lookup operation
>+	are not reliable.
>+	If you use a FUSE-based filesystem as an aufs branch, and it
>+	customizes the inode attribute on it without overriding
>+	fuse_lowlevel_ops.lookup, probably you need to enable this
>+	configuration.
>+	If you enable this configuration, aufs calls getattr operation
>+	in every lookup and revalidate operation for the FUSE-based
>+	filesystem branch.
>+	It will damage the performance even if you don't use a
>+	FUSE-based filesystem branch.

Performance is not "damaged" but "hurt"  :-) [or just "will negatively
impact performance"]

>+config AUFS_MAGIC_SYSRQ
>+	bool
>+	depends on AUFS
>+	depends on AUFS_DEBUG
>+	depends on MAGIC_SYSRQ
>+	default y

Since I see this for the first time, a bit of help text would not be
bad, even if it is not visible in menuconfig.

>+config AUFS_COMPAT
>+	bool "Compatibility with Unionfs (obsolete)"

If it is obsolete, it should not be merged.
--
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/