From: Jiaying Zhang Subject: Re: Question on fallocate/ftruncate sequence Date: Wed, 23 Sep 2009 22:27:17 -0700 Message-ID: <5df78e1d0909232227y2cb52abew827d7732a3bc9040@mail.gmail.com> References: <1248389165.17459.3.camel@bobble.smo.corp.google.com> <5df78e1d0908281444x556a7c2ey763dc6233820abc6@mail.gmail.com> <20090828221432.GS4197@webber.adilger.int> <5df78e1d0908281740w7bc0f283x5004ca5b231b3af5@mail.gmail.com> <20090830025250.GA25768@mit.edu> <5df78e1d0908311240s3205b4bcrb65b2552b4ed579c@mail.gmail.com> <20090831215612.GG4197@webber.adilger.int> <5df78e1d0908311633k1f16a096t701e0cdab54b174c@mail.gmail.com> <20090902084134.GO4197@webber.adilger.int> <5df78e1d0909022220m1152b313o92f6cb7cc8858298@mail.gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: Theodore Tso , Frank Mayhar , Eric Sandeen , Curt Wohlgemuth , ext4 development To: Andreas Dilger Return-path: Received: from smtp-out.google.com ([216.239.45.13]:1452 "EHLO smtp-out.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752231AbZIXF1R convert rfc822-to-8bit (ORCPT ); Thu, 24 Sep 2009 01:27:17 -0400 Received: from wpaz1.hot.corp.google.com (wpaz1.hot.corp.google.com [172.24.198.65]) by smtp-out.google.com with ESMTP id n8O5RKdh029482 for ; Wed, 23 Sep 2009 22:27:20 -0700 Received: from pxi13 (pxi13.prod.google.com [10.243.27.13]) by wpaz1.hot.corp.google.com with ESMTP id n8O5QDcL027784 for ; Wed, 23 Sep 2009 22:27:17 -0700 Received: by pxi13 with SMTP id 13so1297750pxi.6 for ; Wed, 23 Sep 2009 22:27:17 -0700 (PDT) In-Reply-To: <5df78e1d0909022220m1152b313o92f6cb7cc8858298@mail.gmail.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: Sorry for taking so long to finish this. Here is the new patch based on Andreas's suggestions. Now the patch clears the EXT4_EOFBLOCKS_FL flag when we allocate beyond the maximum allocated block. I also made the EOFBLOCKS flag user visible and added the handling in ext4_ioctl as Andrea suggested. Index: linux-2.6.30.5/fs/ext4/inode.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.30.5.orig/fs/ext4/inode.c=A0=A0=A0 2009-08-31 12:08:10.00= 0000000 -0700 +++ linux-2.6.30.5/fs/ext4/inode.c=A0=A0=A0 2009-09-23 21:42:33.0000000= 00 -0700 @@ -3973,6 +3973,8 @@ void ext4_truncate(struct inode *inode) =A0=A0=A0=A0 if (!ext4_can_truncate(inode)) =A0=A0=A0=A0 =A0=A0=A0 return; +=A0=A0=A0 inode->i_flags &=3D ~EXT4_EOFBLOCKS_FL; + =A0=A0=A0=A0 if (inode->i_size =3D=3D 0 && !test_opt(inode->i_sb, NO_AU= TO_DA_ALLOC)) =A0=A0=A0=A0 =A0=A0=A0 ei->i_state |=3D EXT4_STATE_DA_ALLOC_CLOSE; @@ -4285,8 +4287,8 @@ void ext4_get_inode_flags(struct ext4_in =A0{ =A0=A0=A0=A0 unsigned int flags =3D ei->vfs_inode.i_flags; -=A0=A0=A0 ei->i_flags &=3D ~(EXT4_SYNC_FL|EXT4_APPEND_FL| -=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 EXT4_IMMUTABLE_FL|EXT4_NOATIME_FL|EXT4_D= IRSYNC_FL); +=A0=A0=A0 ei->i_flags &=3D ~(EXT4_SYNC_FL|EXT4_APPEND_FL|EXT4_IMMUTABL= E_FL| +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 EXT4_NOATIME_FL|EXT4_DIRSYNC_FL|EXT4_EOF= BLOCKS_FL); =A0=A0=A0=A0 if (flags & S_SYNC) =A0=A0=A0=A0 =A0=A0=A0 ei->i_flags |=3D EXT4_SYNC_FL; =A0=A0=A0=A0 if (flags & S_APPEND) @@ -4297,6 +4299,8 @@ void ext4_get_inode_flags(struct ext4_in =A0=A0=A0=A0 =A0=A0=A0 ei->i_flags |=3D EXT4_NOATIME_FL; =A0=A0=A0=A0 if (flags & S_DIRSYNC) =A0=A0=A0=A0 =A0=A0=A0 ei->i_flags |=3D EXT4_DIRSYNC_FL; +=A0=A0=A0 if (flags & FS_EOFBLOCKS_FL) +=A0=A0=A0 =A0=A0=A0 ei->i_flags |=3D EXT4_EOFBLOCKS_FL; =A0} =A0static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 struct ext4_inode_= info *ei) @@ -4807,7 +4811,9 @@ int ext4_setattr(struct dentry *dentry, =A0=A0=A0=A0 } =A0=A0=A0=A0 if (S_ISREG(inode->i_mode) && -=A0=A0=A0 =A0=A0=A0 attr->ia_valid & ATTR_SIZE && attr->ia_size < inod= e->i_size) { +=A0=A0=A0 =A0=A0=A0 attr->ia_valid & ATTR_SIZE && +=A0=A0=A0 =A0=A0=A0 (attr->ia_size < inode->i_size || +=A0=A0=A0 =A0=A0=A0=A0 (inode->i_flags & EXT4_EOFBLOCKS_FL))) { =A0=A0=A0=A0 =A0=A0=A0 handle_t *handle; =A0=A0=A0=A0 =A0=A0=A0 handle =3D ext4_journal_start(inode, 3); @@ -4838,6 +4844,11 @@ int ext4_setattr(struct dentry *dentry, =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 goto err_out; =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 } =A0=A0=A0=A0 =A0=A0=A0 } +=A0=A0=A0 =A0=A0=A0 if ((inode->i_flags & EXT4_EOFBLOCKS_FL)) { +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 rc =3D vmtruncate(inode, attr->ia_size); +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 if (rc) +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 goto err_out; +=A0=A0=A0 =A0=A0=A0 } =A0=A0=A0=A0 } =A0=A0=A0=A0 rc =3D inode_setattr(inode, attr); Index: linux-2.6.30.5/include/linux/fs.h =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.30.5.orig/include/linux/fs.h=A0=A0=A0 2009-08-31 12:08:10.000000000 -0700 +++ linux-2.6.30.5/include/linux/fs.h=A0=A0=A0 2009-09-10 21:27:30.0000= 00000 -0700 @@ -343,9 +343,10 @@ struct inodes_stat_t { =A0#define FS_TOPDIR_FL=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 0x00020000 /* Top = of directory hierarchies*/ =A0#define FS_EXTENT_FL=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 0x00080000 /* Exte= nts */ =A0#define FS_DIRECTIO_FL=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 0x00100000 /* Us= e direct i/o */ +#define FS_EOFBLOCKS_FL=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 0x00200000 /* Blo= cks allocated beyond EOF */ =A0#define FS_RESERVED_FL=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 0x80000000 /* re= served for ext2 lib */ -#define FS_FL_USER_VISIBLE=A0=A0=A0 =A0=A0=A0 0x0003DFFF /* User visib= le flags */ +#define FS_FL_USER_VISIBLE=A0=A0=A0 =A0=A0=A0 0x0023DFFF /* User visib= le flags */ =A0#define FS_FL_USER_MODIFIABLE=A0=A0=A0 =A0=A0=A0 0x000380FF /* User = modifiable flags */ Index: linux-2.6.30.5/fs/ext4/ext4.h =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.30.5.orig/fs/ext4/ext4.h=A0=A0=A0 2009-08-31 12:08:10.000= 000000 -0700 +++ linux-2.6.30.5/fs/ext4/ext4.h=A0=A0=A0 2009-09-10 21:28:14.00000000= 0 -0700 @@ -235,9 +235,10 @@ struct flex_groups { =A0#define EXT4_HUGE_FILE_FL=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0=A0 = 0x00040000 /* Set to each huge file */ =A0#define EXT4_EXTENTS_FL=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 0x00080000 /* I= node uses extents */ =A0#define EXT4_EXT_MIGRATE=A0=A0=A0 =A0=A0=A0 0x00100000 /* Inode is m= igrating */ +#define EXT4_EOFBLOCKS_FL=A0=A0=A0 =A0=A0=A0 0x00200000 /* Blocks allo= cated beyond EOF (bit reserved in fs.h) */ =A0#define EXT4_RESERVED_FL=A0=A0=A0 =A0=A0=A0 0x80000000 /* reserved f= or ext4 lib */ -#define EXT4_FL_USER_VISIBLE=A0=A0=A0 =A0=A0=A0 0x000BDFFF /* User vis= ible flags */ +#define EXT4_FL_USER_VISIBLE=A0=A0=A0 =A0=A0=A0 0x002BDFFF /* User vis= ible flags */ =A0#define EXT4_FL_USER_MODIFIABLE=A0=A0=A0 =A0=A0=A0 0x000B80FF /* Use= r modifiable flags */ =A0/* Flags that should be inherited by new inodes from their parent. *= / Index: linux-2.6.30.5/fs/ext4/extents.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.30.5.orig/fs/ext4/extents.c=A0=A0=A0 2009-09-01 18:14:58.= 000000000 -0700 +++ linux-2.6.30.5/fs/ext4/extents.c=A0=A0=A0 2009-09-23 22:12:22.00000= 0000 -0700 @@ -2788,7 +2788,7 @@ int ext4_ext_get_blocks(handle_t *handle =A0{ =A0=A0=A0=A0 struct ext4_ext_path *path =3D NULL; =A0=A0=A0=A0 struct ext4_extent_header *eh; -=A0=A0=A0 struct ext4_extent newex, *ex; +=A0=A0=A0 struct ext4_extent newex, *ex, *last_ex; =A0=A0=A0=A0 ext4_fsblk_t newblock; =A0=A0=A0=A0 int err =3D 0, depth, ret, cache_type; =A0=A0=A0=A0 unsigned int allocated =3D 0; @@ -2968,6 +2968,14 @@ int ext4_ext_get_blocks(handle_t *handle =A0=A0=A0=A0 newex.ee_len =3D cpu_to_le16(ar.len); =A0=A0=A0=A0 if (create =3D=3D EXT4_CREATE_UNINITIALIZED_EXT)=A0 /* Mar= k uninitialized */ =A0=A0=A0=A0 =A0=A0=A0 ext4_ext_mark_uninitialized(&newex); + +=A0=A0=A0 if (unlikely(inode->i_flags & EXT4_EOFBLOCKS_FL)) { +=A0=A0=A0 =A0=A0=A0 BUG_ON(!eh->eh_entries); +=A0=A0=A0 =A0=A0=A0 last_ex =3D EXT_LAST_EXTENT(eh); +=A0=A0=A0 =A0=A0=A0 if (iblock + ar.len > le32_to_cpu(last_ex->ee_bloc= k) +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 + ext4_ext_get_actua= l_len(last_ex)) +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 inode->i_flags &=3D ~EXT4_EOFBLOCKS_FL; +=A0=A0=A0 } =A0=A0=A0=A0 err =3D ext4_ext_insert_extent(handle, inode, path, &newex= ); =A0=A0=A0=A0 if (err) { =A0=A0=A0=A0 =A0=A0=A0 /* free data blocks we just allocated */ @@ -3095,6 +3103,13 @@ static void ext4_falloc_update_inode(str =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 i_size_write(inode, new_size); =A0=A0=A0=A0 =A0=A0=A0 if (new_size > EXT4_I(inode)->i_disksize) =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 ext4_update_i_disksize(inode, new_size= ); +=A0=A0=A0 } else { +=A0=A0=A0 =A0=A0=A0 /* +=A0=A0=A0 =A0=A0=A0 =A0* Mark that we allocate beyond EOF so the subse= quent truncate +=A0=A0=A0 =A0=A0=A0 =A0* can proceed even if the new size is the same = as i_size. +=A0=A0=A0 =A0=A0=A0 =A0*/ +=A0=A0=A0 =A0=A0=A0 if (new_size > i_size_read(inode)) +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 inode->i_flags |=3D EXT4_EOFBLOCKS_FL; =A0=A0=A0=A0 } =A0} Index: linux-2.6.30.5/fs/ext4/ioctl.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D --- linux-2.6.30.5.orig/fs/ext4/ioctl.c=A0=A0=A0 2009-08-16 14:19:38.00= 0000000 -0700 +++ linux-2.6.30.5/fs/ext4/ioctl.c=A0=A0=A0 2009-09-23 22:04:47.0000000= 00 -0700 @@ -92,6 +92,16 @@ long ext4_ioctl(struct file *filp, unsig =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 flags &=3D ~EXT4_EXTENTS_FL; =A0=A0=A0=A0 =A0=A0=A0 } +=A0=A0=A0 =A0=A0=A0 if (flags & EXT4_EOFBLOCKS_FL) { +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 /* we don't support adding EOFBLOCKS fla= g */ +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 if (!(oldflags & EXT4_EOFBLOCKS_FL)) { +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 err =3D -EOPNOTSUPP; +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 =A0=A0=A0 goto flags_out; +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 } +=A0=A0=A0 =A0=A0=A0 } else if (oldflags & EXT4_EOFBLOCKS_FL) +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 /* free the space reserved with fallocat= e KEEPSIZE */ +=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 vmtruncate(inode, inode->i_size); + =A0=A0=A0=A0 =A0=A0=A0 handle =3D ext4_journal_start(inode, 1); =A0=A0=A0=A0 =A0=A0=A0 if (IS_ERR(handle)) { =A0=A0=A0=A0 =A0=A0=A0 =A0=A0=A0 err =3D PTR_ERR(handle); Jiaying On Wed, Sep 2, 2009 at 10:20 PM, Jiaying Zhang wr= ote: > > On Wed, Sep 2, 2009 at 1:41 AM, Andreas Dilger wrote= : > > On Aug 31, 2009 =A016:33 -0700, Jiaying Zhang wrote: > >> > EXT4_KEEPSIZE_FL should only be cleared if there were writes to > >> > the end of the fallocated space. =A0In that regard, I think the = name > >> > of this flag should be changed to something like "EXT4_EOFBLOCKS= _FL" > >> > to indicate that blocks are allocated beyond the end of file (i_= size). > >> > >> Thanks for catching this! I changed the patch to only clear the fl= ag > >> when the new_size is larger than i_size and changed the flag name > >> as you suggested. It would be nice if we only clear the flag when = we > >> write beyond the fallocated space, but this seems hard to detect > >> because we no longer have the allocated size once that keepsize > >> fallocate call returns. > > > > The problem is that if e2fsck depends on the EXT4_EOFBLOCKS_FL set > > for fallocate-beyond-EOF then it is worse to clear it than to leave > > it set. =A0At worst, leaving the flag set results in too many trunc= ates > > on the file. =A0Clearing the flag when not correct may result in us= er > > visible data corruption if the file size is extended... > > > >> Here is the new patch: > >> > >> --- .pc/fallocate_keepsizse.patch/fs/ext4/extents.c =A0 2009-08-31 > >> 12:08:10.000000000 -0700 > >> +++ fs/ext4/extents.c 2009-08-31 15:51:13.000000000 -0700 > >> @@ -3091,11 +3091,19 @@ static void ext4_falloc_update_inode(str > >> =A0 =A0 =A0 =A0* the file size. > >> =A0 =A0 =A0 =A0*/ > >> =A0 =A0 =A0 if (!(mode & FALLOC_FL_KEEP_SIZE)) { > >> + =A0 =A0 =A0 =A0 =A0 =A0 if (new_size > i_size_read(inode)) { > >> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 i_size_write(inode, ne= w_size); > >> + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 inode->i_flags &=3D ~EXT= 4_EOFBLOCKS_FL; > > > > This again isn't quite correct, since the EOFBLOCKS_FL shouldn't > > be cleared unless new_size is beyond the allocated size. =A0The > > allocation code itself might be a better place to clear this, > > since it knows whether there were new blocks being added beyond > > the current max allocated block. > > We were thinking to clear this flag when we need to allocate new > blocks, but I was not sure how to get the current max allocated > block -- that is mostly because I just started working on the ext4 > code. After digging into the ext4 allocation code today, I think we > can put the check&clear in ext4_ext_get_blocks: > > @@ -2968,6 +2968,14 @@ int ext4_ext_get_blocks(handle_t *handle > =A0 =A0 =A0 =A0newex.ee_len =3D cpu_to_le16(ar.len); > =A0 =A0 =A0 =A0if (create =3D=3D EXT4_CREATE_UNINITIALIZED_EXT) =A0/*= Mark uninitialized */ > =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0ext4_ext_mark_uninitialized(&newex); > + > + =A0 =A0 =A0 if (unlikely(inode->i_flags & EXT4_EOFBLOCKS_FL)) { > + =A0 =A0 =A0 =A0 =A0 =A0 =A0 BUG_ON(!eh->eh_entries); > + =A0 =A0 =A0 =A0 =A0 =A0 =A0 last_ex =3D EXT_LAST_EXTENT(eh); > + =A0 =A0 =A0 =A0 =A0 =A0 =A0 if (iblock + max_blocks > le32_to_cpu(l= ast_ex->ee_block) > + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0= =A0 =A0 + ext4_ext_get_actual_len(last_ex)) > + =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 inode->i_flags &=3D ~EX= T4_EOFBLOCKS_FL; > + =A0 =A0 =A0 } > =A0 =A0 =A0 =A0err =3D ext4_ext_insert_extent(handle, inode, path, &n= ewex); > =A0 =A0 =A0 =A0if (err) { > =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0/* free data blocks we just allocated = */ > > Again, I just started looking at this part of code, so please let me = know > if I am in the right direction. > > Another thing I am not sure is whether we can allocate a non-data blo= ck, > like extended attributes, beyond the current max block without changi= ng > the i_size. In that case, clearing the EOFBLOCKS flag will be wrong. > > >> =A0#define FS_FL_USER_VISIBLE =A0 =A0 =A0 =A0 =A0 0x0003DFFF /* Us= er visible flags */ > > > > It probably isn't a bad idea to make this flag user-visible, since = it > > would allow scanning for files that have excess space reserved (e.g= =2E > > if the filesystem is getting full). =A0Making it user-settable (i.e= =2E > > clearable) would essentially mean truncating the file to i_size wit= hout > > updating the timestamps so that the reserved space is discarded. =A0= I > > don't think there is any value in allowing a user to turn this flag= on > > for a file. > > So to make it user-settable, we need to add the handling in ext4_ioct= l > that calls vmtruncate when the flag to be cleared. But how can we get > the right size to truncate in that case? Can we just set that to the > max initialized block shift with block size? But that may also trunca= te > the blocks reserved without the KEEP_SIZE flag. > > Jiaying > > > > > Cheers, Andreas > > -- > > Andreas Dilger > > Sr. Staff Engineer, Lustre Group > > Sun Microsystems of Canada, Inc. > > > > -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html