From c9a98553d513dfc82cdce869970d5662c1f22c68 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 1 Jan 2009 14:21:16 -0500 Subject: [PATCH 01/10] [XFS] pass XFS_IGET_BULKSTAT to xfs_iget for handle operations NFS clients or users of the handle ioctls can pass us arbitrary inode numbers through the exportfs interface. Make sure we use the XFS_IGET_BULKSTAT so that these don't cause shutdowns due to the corruption checks. Also translate the EINVAL we get back for invalid inode clusters into an ESTALE which is more appropinquate, and remove the useless check for a NULL inode on a successfull xfs_iget return. I have a testcase to reproduce this using the handle interface which I will submit to xfsqa. Reported-by: Mario Becroft Signed-off-by: Christoph Hellwig Reviewed-by: Eric Sandeen Signed-off-by: Lachlan McIlroy --- fs/xfs/linux-2.6/xfs_export.c | 23 +++++++++++++++++++---- 1 file changed, 19 insertions(+), 4 deletions(-) diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c index 595751f78350..87b8cbd23d4b 100644 --- a/fs/xfs/linux-2.6/xfs_export.c +++ b/fs/xfs/linux-2.6/xfs_export.c @@ -126,11 +126,26 @@ xfs_nfs_get_inode( if (ino == 0) return ERR_PTR(-ESTALE); - error = xfs_iget(mp, NULL, ino, 0, XFS_ILOCK_SHARED, &ip, 0); - if (error) + /* + * The XFS_IGET_BULKSTAT means that an invalid inode number is just + * fine and not an indication of a corrupted filesystem. Because + * clients can send any kind of invalid file handle, e.g. after + * a restore on the server we have to deal with this case gracefully. + */ + error = xfs_iget(mp, NULL, ino, XFS_IGET_BULKSTAT, + XFS_ILOCK_SHARED, &ip, 0); + if (error) { + /* + * EINVAL means the inode cluster doesn't exist anymore. + * This implies the filehandle is stale, so we should + * translate it here. + * We don't use ESTALE directly down the chain to not + * confuse applications using bulkstat that expect EINVAL. + */ + if (error == EINVAL) + error = ESTALE; return ERR_PTR(-error); - if (!ip) - return ERR_PTR(-EIO); + } if (ip->i_d.di_gen != generation) { xfs_iput_new(ip, XFS_ILOCK_SHARED); From 9800b550355e99c9bcaba7ec6540751dce0823d7 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 1 Jan 2009 16:40:10 -0600 Subject: [PATCH 02/10] [XFS] Remove several unused typedefs. Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/xfs/linux-2.6/xfs_aops.h | 2 -- fs/xfs/xfs_acl.h | 1 - fs/xfs/xfs_types.h | 2 -- 3 files changed, 5 deletions(-) diff --git a/fs/xfs/linux-2.6/xfs_aops.h b/fs/xfs/linux-2.6/xfs_aops.h index 7b26f5ff9692..1dd528849755 100644 --- a/fs/xfs/linux-2.6/xfs_aops.h +++ b/fs/xfs/linux-2.6/xfs_aops.h @@ -21,8 +21,6 @@ extern struct workqueue_struct *xfsdatad_workqueue; extern mempool_t *xfs_ioend_pool; -typedef void (*xfs_ioend_func_t)(void *); - /* * xfs_ioend struct manages large extent writes for XFS. * It can manage several multi-page bio's at once. diff --git a/fs/xfs/xfs_acl.h b/fs/xfs/xfs_acl.h index a4e293b93efa..642f1db4def4 100644 --- a/fs/xfs/xfs_acl.h +++ b/fs/xfs/xfs_acl.h @@ -22,7 +22,6 @@ * Access Control Lists */ typedef __uint16_t xfs_acl_perm_t; -typedef __int32_t xfs_acl_type_t; typedef __int32_t xfs_acl_tag_t; typedef __int32_t xfs_acl_id_t; diff --git a/fs/xfs/xfs_types.h b/fs/xfs/xfs_types.h index 0f5191644ab2..baedbd14dc21 100644 --- a/fs/xfs/xfs_types.h +++ b/fs/xfs/xfs_types.h @@ -111,8 +111,6 @@ typedef __uint64_t xfs_fileoff_t; /* block number in a file */ typedef __int64_t xfs_sfiloff_t; /* signed block number in a file */ typedef __uint64_t xfs_filblks_t; /* number of blocks in a file */ -typedef __uint8_t xfs_arch_t; /* architecture of an xfs fs */ - /* * Null values for the types. */ From c9fb86a917640d66ba2e0613a12f3a76eda8a30f Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 1 Jan 2009 16:40:11 -0600 Subject: [PATCH 03/10] [XFS] Remove macro-to-function indirections in attr code Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/xfs/xfs_attr_leaf.c | 72 +++++++++++++++++++++--------------------- fs/xfs/xfs_attr_leaf.h | 12 ------- 2 files changed, 36 insertions(+), 48 deletions(-) diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index 79da6b2ea99e..6c323f8a4cd1 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -736,7 +736,7 @@ xfs_attr_shortform_allfit(xfs_dabuf_t *bp, xfs_inode_t *dp) continue; /* don't copy partial entries */ if (!(entry->flags & XFS_ATTR_LOCAL)) return(0); - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, i); + name_loc = xfs_attr_leaf_name_local(leaf, i); if (name_loc->namelen >= XFS_ATTR_SF_ENTSIZE_MAX) return(0); if (be16_to_cpu(name_loc->valuelen) >= XFS_ATTR_SF_ENTSIZE_MAX) @@ -823,7 +823,7 @@ xfs_attr_leaf_to_shortform(xfs_dabuf_t *bp, xfs_da_args_t *args, int forkoff) if (!entry->nameidx) continue; ASSERT(entry->flags & XFS_ATTR_LOCAL); - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, i); + name_loc = xfs_attr_leaf_name_local(leaf, i); nargs.name = (char *)name_loc->nameval; nargs.namelen = name_loc->namelen; nargs.value = (char *)&name_loc->nameval[nargs.namelen]; @@ -1141,14 +1141,14 @@ xfs_attr_leaf_add_work(xfs_dabuf_t *bp, xfs_da_args_t *args, int mapindex) * as part of this transaction (a split operation for example). */ if (entry->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, args->index); + name_loc = xfs_attr_leaf_name_local(leaf, args->index); name_loc->namelen = args->namelen; name_loc->valuelen = cpu_to_be16(args->valuelen); memcpy((char *)name_loc->nameval, args->name, args->namelen); memcpy((char *)&name_loc->nameval[args->namelen], args->value, be16_to_cpu(name_loc->valuelen)); } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf, args->index); name_rmt->namelen = args->namelen; memcpy((char *)name_rmt->name, args->name, args->namelen); entry->flags |= XFS_ATTR_INCOMPLETE; @@ -1159,7 +1159,7 @@ xfs_attr_leaf_add_work(xfs_dabuf_t *bp, xfs_da_args_t *args, int mapindex) args->rmtblkcnt = XFS_B_TO_FSB(mp, args->valuelen); } xfs_da_log_buf(args->trans, bp, - XFS_DA_LOGRANGE(leaf, XFS_ATTR_LEAF_NAME(leaf, args->index), + XFS_DA_LOGRANGE(leaf, xfs_attr_leaf_name(leaf, args->index), xfs_attr_leaf_entsize(leaf, args->index))); /* @@ -1749,10 +1749,10 @@ xfs_attr_leaf_remove(xfs_dabuf_t *bp, xfs_da_args_t *args) /* * Compress the remaining entries and zero out the removed stuff. */ - memset(XFS_ATTR_LEAF_NAME(leaf, args->index), 0, entsize); + memset(xfs_attr_leaf_name(leaf, args->index), 0, entsize); be16_add_cpu(&hdr->usedbytes, -entsize); xfs_da_log_buf(args->trans, bp, - XFS_DA_LOGRANGE(leaf, XFS_ATTR_LEAF_NAME(leaf, args->index), + XFS_DA_LOGRANGE(leaf, xfs_attr_leaf_name(leaf, args->index), entsize)); tmp = (be16_to_cpu(hdr->count) - args->index) @@ -1985,7 +1985,7 @@ xfs_attr_leaf_lookup_int(xfs_dabuf_t *bp, xfs_da_args_t *args) continue; } if (entry->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, probe); + name_loc = xfs_attr_leaf_name_local(leaf, probe); if (name_loc->namelen != args->namelen) continue; if (memcmp(args->name, (char *)name_loc->nameval, args->namelen) != 0) @@ -1995,7 +1995,7 @@ xfs_attr_leaf_lookup_int(xfs_dabuf_t *bp, xfs_da_args_t *args) args->index = probe; return(XFS_ERROR(EEXIST)); } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, probe); + name_rmt = xfs_attr_leaf_name_remote(leaf, probe); if (name_rmt->namelen != args->namelen) continue; if (memcmp(args->name, (char *)name_rmt->name, @@ -2035,7 +2035,7 @@ xfs_attr_leaf_getvalue(xfs_dabuf_t *bp, xfs_da_args_t *args) entry = &leaf->entries[args->index]; if (entry->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, args->index); + name_loc = xfs_attr_leaf_name_local(leaf, args->index); ASSERT(name_loc->namelen == args->namelen); ASSERT(memcmp(args->name, name_loc->nameval, args->namelen) == 0); valuelen = be16_to_cpu(name_loc->valuelen); @@ -2050,7 +2050,7 @@ xfs_attr_leaf_getvalue(xfs_dabuf_t *bp, xfs_da_args_t *args) args->valuelen = valuelen; memcpy(args->value, &name_loc->nameval[args->namelen], valuelen); } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf, args->index); ASSERT(name_rmt->namelen == args->namelen); ASSERT(memcmp(args->name, name_rmt->name, args->namelen) == 0); valuelen = be32_to_cpu(name_rmt->valuelen); @@ -2143,7 +2143,7 @@ xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s, * off for 6.2, should be revisited later. */ if (entry_s->flags & XFS_ATTR_INCOMPLETE) { /* skip partials? */ - memset(XFS_ATTR_LEAF_NAME(leaf_s, start_s + i), 0, tmp); + memset(xfs_attr_leaf_name(leaf_s, start_s + i), 0, tmp); be16_add_cpu(&hdr_s->usedbytes, -tmp); be16_add_cpu(&hdr_s->count, -1); entry_d--; /* to compensate for ++ in loop hdr */ @@ -2160,11 +2160,11 @@ xfs_attr_leaf_moveents(xfs_attr_leafblock_t *leaf_s, int start_s, entry_d->flags = entry_s->flags; ASSERT(be16_to_cpu(entry_d->nameidx) + tmp <= XFS_LBSIZE(mp)); - memmove(XFS_ATTR_LEAF_NAME(leaf_d, desti), - XFS_ATTR_LEAF_NAME(leaf_s, start_s + i), tmp); + memmove(xfs_attr_leaf_name(leaf_d, desti), + xfs_attr_leaf_name(leaf_s, start_s + i), tmp); ASSERT(be16_to_cpu(entry_s->nameidx) + tmp <= XFS_LBSIZE(mp)); - memset(XFS_ATTR_LEAF_NAME(leaf_s, start_s + i), 0, tmp); + memset(xfs_attr_leaf_name(leaf_s, start_s + i), 0, tmp); be16_add_cpu(&hdr_s->usedbytes, -tmp); be16_add_cpu(&hdr_d->usedbytes, tmp); be16_add_cpu(&hdr_s->count, -1); @@ -2276,12 +2276,12 @@ xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index) ASSERT(be16_to_cpu(leaf->hdr.info.magic) == XFS_ATTR_LEAF_MAGIC); if (leaf->entries[index].flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, index); - size = XFS_ATTR_LEAF_ENTSIZE_LOCAL(name_loc->namelen, + name_loc = xfs_attr_leaf_name_local(leaf, index); + size = xfs_attr_leaf_entsize_local(name_loc->namelen, be16_to_cpu(name_loc->valuelen)); } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, index); - size = XFS_ATTR_LEAF_ENTSIZE_REMOTE(name_rmt->namelen); + name_rmt = xfs_attr_leaf_name_remote(leaf, index); + size = xfs_attr_leaf_entsize_remote(name_rmt->namelen); } return(size); } @@ -2297,13 +2297,13 @@ xfs_attr_leaf_newentsize(int namelen, int valuelen, int blocksize, int *local) { int size; - size = XFS_ATTR_LEAF_ENTSIZE_LOCAL(namelen, valuelen); - if (size < XFS_ATTR_LEAF_ENTSIZE_LOCAL_MAX(blocksize)) { + size = xfs_attr_leaf_entsize_local(namelen, valuelen); + if (size < xfs_attr_leaf_entsize_local_max(blocksize)) { if (local) { *local = 1; } } else { - size = XFS_ATTR_LEAF_ENTSIZE_REMOTE(namelen); + size = xfs_attr_leaf_entsize_remote(namelen); if (local) { *local = 0; } @@ -2372,7 +2372,7 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context) if (entry->flags & XFS_ATTR_LOCAL) { xfs_attr_leaf_name_local_t *name_loc = - XFS_ATTR_LEAF_NAME_LOCAL(leaf, i); + xfs_attr_leaf_name_local(leaf, i); retval = context->put_listent(context, entry->flags, @@ -2384,7 +2384,7 @@ xfs_attr_leaf_list_int(xfs_dabuf_t *bp, xfs_attr_list_context_t *context) return retval; } else { xfs_attr_leaf_name_remote_t *name_rmt = - XFS_ATTR_LEAF_NAME_REMOTE(leaf, i); + xfs_attr_leaf_name_remote(leaf, i); int valuelen = be32_to_cpu(name_rmt->valuelen); @@ -2468,11 +2468,11 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args) #ifdef DEBUG if (entry->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf, args->index); + name_loc = xfs_attr_leaf_name_local(leaf, args->index); namelen = name_loc->namelen; name = (char *)name_loc->nameval; } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf, args->index); namelen = name_rmt->namelen; name = (char *)name_rmt->name; } @@ -2487,7 +2487,7 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args) if (args->rmtblkno) { ASSERT((entry->flags & XFS_ATTR_LOCAL) == 0); - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf, args->index); name_rmt->valueblk = cpu_to_be32(args->rmtblkno); name_rmt->valuelen = cpu_to_be32(args->valuelen); xfs_da_log_buf(args->trans, bp, @@ -2534,7 +2534,7 @@ xfs_attr_leaf_setflag(xfs_da_args_t *args) xfs_da_log_buf(args->trans, bp, XFS_DA_LOGRANGE(leaf, entry, sizeof(*entry))); if ((entry->flags & XFS_ATTR_LOCAL) == 0) { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf, args->index); name_rmt->valueblk = 0; name_rmt->valuelen = 0; xfs_da_log_buf(args->trans, bp, @@ -2607,20 +2607,20 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args) #ifdef DEBUG if (entry1->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf1, args->index); + name_loc = xfs_attr_leaf_name_local(leaf1, args->index); namelen1 = name_loc->namelen; name1 = (char *)name_loc->nameval; } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf1, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf1, args->index); namelen1 = name_rmt->namelen; name1 = (char *)name_rmt->name; } if (entry2->flags & XFS_ATTR_LOCAL) { - name_loc = XFS_ATTR_LEAF_NAME_LOCAL(leaf2, args->index2); + name_loc = xfs_attr_leaf_name_local(leaf2, args->index2); namelen2 = name_loc->namelen; name2 = (char *)name_loc->nameval; } else { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf2, args->index2); + name_rmt = xfs_attr_leaf_name_remote(leaf2, args->index2); namelen2 = name_rmt->namelen; name2 = (char *)name_rmt->name; } @@ -2637,7 +2637,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args) XFS_DA_LOGRANGE(leaf1, entry1, sizeof(*entry1))); if (args->rmtblkno) { ASSERT((entry1->flags & XFS_ATTR_LOCAL) == 0); - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf1, args->index); + name_rmt = xfs_attr_leaf_name_remote(leaf1, args->index); name_rmt->valueblk = cpu_to_be32(args->rmtblkno); name_rmt->valuelen = cpu_to_be32(args->valuelen); xfs_da_log_buf(args->trans, bp1, @@ -2648,7 +2648,7 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args) xfs_da_log_buf(args->trans, bp2, XFS_DA_LOGRANGE(leaf2, entry2, sizeof(*entry2))); if ((entry2->flags & XFS_ATTR_LOCAL) == 0) { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf2, args->index2); + name_rmt = xfs_attr_leaf_name_remote(leaf2, args->index2); name_rmt->valueblk = 0; name_rmt->valuelen = 0; xfs_da_log_buf(args->trans, bp2, @@ -2855,7 +2855,7 @@ xfs_attr_leaf_inactive(xfs_trans_t **trans, xfs_inode_t *dp, xfs_dabuf_t *bp) for (i = 0; i < be16_to_cpu(leaf->hdr.count); entry++, i++) { if (be16_to_cpu(entry->nameidx) && ((entry->flags & XFS_ATTR_LOCAL) == 0)) { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, i); + name_rmt = xfs_attr_leaf_name_remote(leaf, i); if (name_rmt->valueblk) count++; } @@ -2883,7 +2883,7 @@ xfs_attr_leaf_inactive(xfs_trans_t **trans, xfs_inode_t *dp, xfs_dabuf_t *bp) for (i = 0; i < be16_to_cpu(leaf->hdr.count); entry++, i++) { if (be16_to_cpu(entry->nameidx) && ((entry->flags & XFS_ATTR_LOCAL) == 0)) { - name_rmt = XFS_ATTR_LEAF_NAME_REMOTE(leaf, i); + name_rmt = xfs_attr_leaf_name_remote(leaf, i); if (name_rmt->valueblk) { lp->valueblk = be32_to_cpu(name_rmt->valueblk); lp->valuelen = XFS_B_TO_FSB(dp->i_mount, diff --git a/fs/xfs/xfs_attr_leaf.h b/fs/xfs/xfs_attr_leaf.h index 83e9af417ca2..9c7d22fdcf4d 100644 --- a/fs/xfs/xfs_attr_leaf.h +++ b/fs/xfs/xfs_attr_leaf.h @@ -151,8 +151,6 @@ typedef struct xfs_attr_leafblock { /* * Cast typed pointers for "local" and "remote" name/value structs. */ -#define XFS_ATTR_LEAF_NAME_REMOTE(leafp,idx) \ - xfs_attr_leaf_name_remote(leafp,idx) static inline xfs_attr_leaf_name_remote_t * xfs_attr_leaf_name_remote(xfs_attr_leafblock_t *leafp, int idx) { @@ -160,8 +158,6 @@ xfs_attr_leaf_name_remote(xfs_attr_leafblock_t *leafp, int idx) &((char *)leafp)[be16_to_cpu(leafp->entries[idx].nameidx)]; } -#define XFS_ATTR_LEAF_NAME_LOCAL(leafp,idx) \ - xfs_attr_leaf_name_local(leafp,idx) static inline xfs_attr_leaf_name_local_t * xfs_attr_leaf_name_local(xfs_attr_leafblock_t *leafp, int idx) { @@ -169,8 +165,6 @@ xfs_attr_leaf_name_local(xfs_attr_leafblock_t *leafp, int idx) &((char *)leafp)[be16_to_cpu(leafp->entries[idx].nameidx)]; } -#define XFS_ATTR_LEAF_NAME(leafp,idx) \ - xfs_attr_leaf_name(leafp,idx) static inline char *xfs_attr_leaf_name(xfs_attr_leafblock_t *leafp, int idx) { return &((char *)leafp)[be16_to_cpu(leafp->entries[idx].nameidx)]; @@ -181,24 +175,18 @@ static inline char *xfs_attr_leaf_name(xfs_attr_leafblock_t *leafp, int idx) * a "local" name/value structure, a "remote" name/value structure, and * a pointer which might be either. */ -#define XFS_ATTR_LEAF_ENTSIZE_REMOTE(nlen) \ - xfs_attr_leaf_entsize_remote(nlen) static inline int xfs_attr_leaf_entsize_remote(int nlen) { return ((uint)sizeof(xfs_attr_leaf_name_remote_t) - 1 + (nlen) + \ XFS_ATTR_LEAF_NAME_ALIGN - 1) & ~(XFS_ATTR_LEAF_NAME_ALIGN - 1); } -#define XFS_ATTR_LEAF_ENTSIZE_LOCAL(nlen,vlen) \ - xfs_attr_leaf_entsize_local(nlen,vlen) static inline int xfs_attr_leaf_entsize_local(int nlen, int vlen) { return ((uint)sizeof(xfs_attr_leaf_name_local_t) - 1 + (nlen) + (vlen) + XFS_ATTR_LEAF_NAME_ALIGN - 1) & ~(XFS_ATTR_LEAF_NAME_ALIGN - 1); } -#define XFS_ATTR_LEAF_ENTSIZE_LOCAL_MAX(bsize) \ - xfs_attr_leaf_entsize_local_max(bsize) static inline int xfs_attr_leaf_entsize_local_max(int bsize) { return (((bsize) >> 1) + ((bsize) >> 2)); From fb82557f16f3700ae4961a4ce599bdaff6a10b1c Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Fri, 9 Jan 2009 15:53:54 +1100 Subject: [PATCH 04/10] [XFS] Remove macro-to-function indirections in the mask code Signed-off-by: Eric Sandeen Reviewed-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/xfs/xfs_ag.h | 2 +- fs/xfs/xfs_bit.h | 10 +---- fs/xfs/xfs_bmap_btree.c | 84 ++++++++++++++++++++--------------------- 3 files changed, 44 insertions(+), 52 deletions(-) diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index f2e21817a226..d3b3cf742999 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -231,7 +231,7 @@ typedef struct xfs_perag #define XFS_FSB_TO_AGNO(mp,fsbno) \ ((xfs_agnumber_t)((fsbno) >> (mp)->m_sb.sb_agblklog)) #define XFS_FSB_TO_AGBNO(mp,fsbno) \ - ((xfs_agblock_t)((fsbno) & XFS_MASK32LO((mp)->m_sb.sb_agblklog))) + ((xfs_agblock_t)((fsbno) & xfs_mask32lo((mp)->m_sb.sb_agblklog))) #define XFS_AGB_TO_DADDR(mp,agno,agbno) \ ((xfs_daddr_t)XFS_FSB_TO_BB(mp, \ (xfs_fsblock_t)(agno) * (mp)->m_sb.sb_agblocks + (agbno))) diff --git a/fs/xfs/xfs_bit.h b/fs/xfs/xfs_bit.h index bca7b243c319..f1e3c907044d 100644 --- a/fs/xfs/xfs_bit.h +++ b/fs/xfs/xfs_bit.h @@ -23,24 +23,16 @@ */ /* - * masks with n high/low bits set, 32-bit values & 64-bit values + * masks with n high/low bits set, 64-bit values */ -#define XFS_MASK32HI(n) xfs_mask32hi(n) -static inline __uint32_t xfs_mask32hi(int n) -{ - return (__uint32_t)-1 << (32 - (n)); -} -#define XFS_MASK64HI(n) xfs_mask64hi(n) static inline __uint64_t xfs_mask64hi(int n) { return (__uint64_t)-1 << (64 - (n)); } -#define XFS_MASK32LO(n) xfs_mask32lo(n) static inline __uint32_t xfs_mask32lo(int n) { return ((__uint32_t)1 << (n)) - 1; } -#define XFS_MASK64LO(n) xfs_mask64lo(n) static inline __uint64_t xfs_mask64lo(int n) { return ((__uint64_t)1 << (n)) - 1; diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c index 8f1ec73725d3..ba6b08c2fb02 100644 --- a/fs/xfs/xfs_bmap_btree.c +++ b/fs/xfs/xfs_bmap_btree.c @@ -110,16 +110,16 @@ __xfs_bmbt_get_all( ext_flag = (int)(l0 >> (64 - BMBT_EXNTFLAG_BITLEN)); s->br_startoff = ((xfs_fileoff_t)l0 & - XFS_MASK64LO(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; + xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; #if XFS_BIG_BLKNOS - s->br_startblock = (((xfs_fsblock_t)l0 & XFS_MASK64LO(9)) << 43) | + s->br_startblock = (((xfs_fsblock_t)l0 & xfs_mask64lo(9)) << 43) | (((xfs_fsblock_t)l1) >> 21); #else #ifdef DEBUG { xfs_dfsbno_t b; - b = (((xfs_dfsbno_t)l0 & XFS_MASK64LO(9)) << 43) | + b = (((xfs_dfsbno_t)l0 & xfs_mask64lo(9)) << 43) | (((xfs_dfsbno_t)l1) >> 21); ASSERT((b >> 32) == 0 || ISNULLDSTARTBLOCK(b)); s->br_startblock = (xfs_fsblock_t)b; @@ -128,7 +128,7 @@ __xfs_bmbt_get_all( s->br_startblock = (xfs_fsblock_t)(((xfs_dfsbno_t)l1) >> 21); #endif /* DEBUG */ #endif /* XFS_BIG_BLKNOS */ - s->br_blockcount = (xfs_filblks_t)(l1 & XFS_MASK64LO(21)); + s->br_blockcount = (xfs_filblks_t)(l1 & xfs_mask64lo(21)); /* This is xfs_extent_state() in-line */ if (ext_flag) { ASSERT(s->br_blockcount != 0); /* saved for DMIG */ @@ -153,7 +153,7 @@ xfs_filblks_t xfs_bmbt_get_blockcount( xfs_bmbt_rec_host_t *r) { - return (xfs_filblks_t)(r->l1 & XFS_MASK64LO(21)); + return (xfs_filblks_t)(r->l1 & xfs_mask64lo(21)); } /* @@ -164,13 +164,13 @@ xfs_bmbt_get_startblock( xfs_bmbt_rec_host_t *r) { #if XFS_BIG_BLKNOS - return (((xfs_fsblock_t)r->l0 & XFS_MASK64LO(9)) << 43) | + return (((xfs_fsblock_t)r->l0 & xfs_mask64lo(9)) << 43) | (((xfs_fsblock_t)r->l1) >> 21); #else #ifdef DEBUG xfs_dfsbno_t b; - b = (((xfs_dfsbno_t)r->l0 & XFS_MASK64LO(9)) << 43) | + b = (((xfs_dfsbno_t)r->l0 & xfs_mask64lo(9)) << 43) | (((xfs_dfsbno_t)r->l1) >> 21); ASSERT((b >> 32) == 0 || ISNULLDSTARTBLOCK(b)); return (xfs_fsblock_t)b; @@ -188,7 +188,7 @@ xfs_bmbt_get_startoff( xfs_bmbt_rec_host_t *r) { return ((xfs_fileoff_t)r->l0 & - XFS_MASK64LO(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; + xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; } xfs_exntst_t @@ -219,7 +219,7 @@ xfs_filblks_t xfs_bmbt_disk_get_blockcount( xfs_bmbt_rec_t *r) { - return (xfs_filblks_t)(be64_to_cpu(r->l1) & XFS_MASK64LO(21)); + return (xfs_filblks_t)(be64_to_cpu(r->l1) & xfs_mask64lo(21)); } /* @@ -230,7 +230,7 @@ xfs_bmbt_disk_get_startoff( xfs_bmbt_rec_t *r) { return ((xfs_fileoff_t)be64_to_cpu(r->l0) & - XFS_MASK64LO(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; + xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN)) >> 9; } @@ -248,33 +248,33 @@ xfs_bmbt_set_allf( int extent_flag = (state == XFS_EXT_NORM) ? 0 : 1; ASSERT(state == XFS_EXT_NORM || state == XFS_EXT_UNWRITTEN); - ASSERT((startoff & XFS_MASK64HI(64-BMBT_STARTOFF_BITLEN)) == 0); - ASSERT((blockcount & XFS_MASK64HI(64-BMBT_BLOCKCOUNT_BITLEN)) == 0); + ASSERT((startoff & xfs_mask64hi(64-BMBT_STARTOFF_BITLEN)) == 0); + ASSERT((blockcount & xfs_mask64hi(64-BMBT_BLOCKCOUNT_BITLEN)) == 0); #if XFS_BIG_BLKNOS - ASSERT((startblock & XFS_MASK64HI(64-BMBT_STARTBLOCK_BITLEN)) == 0); + ASSERT((startblock & xfs_mask64hi(64-BMBT_STARTBLOCK_BITLEN)) == 0); r->l0 = ((xfs_bmbt_rec_base_t)extent_flag << 63) | ((xfs_bmbt_rec_base_t)startoff << 9) | ((xfs_bmbt_rec_base_t)startblock >> 43); r->l1 = ((xfs_bmbt_rec_base_t)startblock << 21) | ((xfs_bmbt_rec_base_t)blockcount & - (xfs_bmbt_rec_base_t)XFS_MASK64LO(21)); + (xfs_bmbt_rec_base_t)xfs_mask64lo(21)); #else /* !XFS_BIG_BLKNOS */ if (ISNULLSTARTBLOCK(startblock)) { r->l0 = ((xfs_bmbt_rec_base_t)extent_flag << 63) | ((xfs_bmbt_rec_base_t)startoff << 9) | - (xfs_bmbt_rec_base_t)XFS_MASK64LO(9); - r->l1 = XFS_MASK64HI(11) | + (xfs_bmbt_rec_base_t)xfs_mask64lo(9); + r->l1 = xfs_mask64hi(11) | ((xfs_bmbt_rec_base_t)startblock << 21) | ((xfs_bmbt_rec_base_t)blockcount & - (xfs_bmbt_rec_base_t)XFS_MASK64LO(21)); + (xfs_bmbt_rec_base_t)xfs_mask64lo(21)); } else { r->l0 = ((xfs_bmbt_rec_base_t)extent_flag << 63) | ((xfs_bmbt_rec_base_t)startoff << 9); r->l1 = ((xfs_bmbt_rec_base_t)startblock << 21) | ((xfs_bmbt_rec_base_t)blockcount & - (xfs_bmbt_rec_base_t)XFS_MASK64LO(21)); + (xfs_bmbt_rec_base_t)xfs_mask64lo(21)); } #endif /* XFS_BIG_BLKNOS */ } @@ -306,11 +306,11 @@ xfs_bmbt_disk_set_allf( int extent_flag = (state == XFS_EXT_NORM) ? 0 : 1; ASSERT(state == XFS_EXT_NORM || state == XFS_EXT_UNWRITTEN); - ASSERT((startoff & XFS_MASK64HI(64-BMBT_STARTOFF_BITLEN)) == 0); - ASSERT((blockcount & XFS_MASK64HI(64-BMBT_BLOCKCOUNT_BITLEN)) == 0); + ASSERT((startoff & xfs_mask64hi(64-BMBT_STARTOFF_BITLEN)) == 0); + ASSERT((blockcount & xfs_mask64hi(64-BMBT_BLOCKCOUNT_BITLEN)) == 0); #if XFS_BIG_BLKNOS - ASSERT((startblock & XFS_MASK64HI(64-BMBT_STARTBLOCK_BITLEN)) == 0); + ASSERT((startblock & xfs_mask64hi(64-BMBT_STARTBLOCK_BITLEN)) == 0); r->l0 = cpu_to_be64( ((xfs_bmbt_rec_base_t)extent_flag << 63) | @@ -319,17 +319,17 @@ xfs_bmbt_disk_set_allf( r->l1 = cpu_to_be64( ((xfs_bmbt_rec_base_t)startblock << 21) | ((xfs_bmbt_rec_base_t)blockcount & - (xfs_bmbt_rec_base_t)XFS_MASK64LO(21))); + (xfs_bmbt_rec_base_t)xfs_mask64lo(21))); #else /* !XFS_BIG_BLKNOS */ if (ISNULLSTARTBLOCK(startblock)) { r->l0 = cpu_to_be64( ((xfs_bmbt_rec_base_t)extent_flag << 63) | ((xfs_bmbt_rec_base_t)startoff << 9) | - (xfs_bmbt_rec_base_t)XFS_MASK64LO(9)); - r->l1 = cpu_to_be64(XFS_MASK64HI(11) | + (xfs_bmbt_rec_base_t)xfs_mask64lo(9)); + r->l1 = cpu_to_be64(xfs_mask64hi(11) | ((xfs_bmbt_rec_base_t)startblock << 21) | ((xfs_bmbt_rec_base_t)blockcount & - (xfs_bmbt_rec_base_t)XFS_MASK64LO(21))); + (xfs_bmbt_rec_base_t)xfs_mask64lo(21))); } else { r->l0 = cpu_to_be64( ((xfs_bmbt_rec_base_t)extent_flag << 63) | @@ -337,7 +337,7 @@ xfs_bmbt_disk_set_allf( r->l1 = cpu_to_be64( ((xfs_bmbt_rec_base_t)startblock << 21) | ((xfs_bmbt_rec_base_t)blockcount & - (xfs_bmbt_rec_base_t)XFS_MASK64LO(21))); + (xfs_bmbt_rec_base_t)xfs_mask64lo(21))); } #endif /* XFS_BIG_BLKNOS */ } @@ -362,9 +362,9 @@ xfs_bmbt_set_blockcount( xfs_bmbt_rec_host_t *r, xfs_filblks_t v) { - ASSERT((v & XFS_MASK64HI(43)) == 0); - r->l1 = (r->l1 & (xfs_bmbt_rec_base_t)XFS_MASK64HI(43)) | - (xfs_bmbt_rec_base_t)(v & XFS_MASK64LO(21)); + ASSERT((v & xfs_mask64hi(43)) == 0); + r->l1 = (r->l1 & (xfs_bmbt_rec_base_t)xfs_mask64hi(43)) | + (xfs_bmbt_rec_base_t)(v & xfs_mask64lo(21)); } /* @@ -376,21 +376,21 @@ xfs_bmbt_set_startblock( xfs_fsblock_t v) { #if XFS_BIG_BLKNOS - ASSERT((v & XFS_MASK64HI(12)) == 0); - r->l0 = (r->l0 & (xfs_bmbt_rec_base_t)XFS_MASK64HI(55)) | + ASSERT((v & xfs_mask64hi(12)) == 0); + r->l0 = (r->l0 & (xfs_bmbt_rec_base_t)xfs_mask64hi(55)) | (xfs_bmbt_rec_base_t)(v >> 43); - r->l1 = (r->l1 & (xfs_bmbt_rec_base_t)XFS_MASK64LO(21)) | + r->l1 = (r->l1 & (xfs_bmbt_rec_base_t)xfs_mask64lo(21)) | (xfs_bmbt_rec_base_t)(v << 21); #else /* !XFS_BIG_BLKNOS */ if (ISNULLSTARTBLOCK(v)) { - r->l0 |= (xfs_bmbt_rec_base_t)XFS_MASK64LO(9); - r->l1 = (xfs_bmbt_rec_base_t)XFS_MASK64HI(11) | + r->l0 |= (xfs_bmbt_rec_base_t)xfs_mask64lo(9); + r->l1 = (xfs_bmbt_rec_base_t)xfs_mask64hi(11) | ((xfs_bmbt_rec_base_t)v << 21) | - (r->l1 & (xfs_bmbt_rec_base_t)XFS_MASK64LO(21)); + (r->l1 & (xfs_bmbt_rec_base_t)xfs_mask64lo(21)); } else { - r->l0 &= ~(xfs_bmbt_rec_base_t)XFS_MASK64LO(9); + r->l0 &= ~(xfs_bmbt_rec_base_t)xfs_mask64lo(9); r->l1 = ((xfs_bmbt_rec_base_t)v << 21) | - (r->l1 & (xfs_bmbt_rec_base_t)XFS_MASK64LO(21)); + (r->l1 & (xfs_bmbt_rec_base_t)xfs_mask64lo(21)); } #endif /* XFS_BIG_BLKNOS */ } @@ -403,10 +403,10 @@ xfs_bmbt_set_startoff( xfs_bmbt_rec_host_t *r, xfs_fileoff_t v) { - ASSERT((v & XFS_MASK64HI(9)) == 0); - r->l0 = (r->l0 & (xfs_bmbt_rec_base_t) XFS_MASK64HI(1)) | + ASSERT((v & xfs_mask64hi(9)) == 0); + r->l0 = (r->l0 & (xfs_bmbt_rec_base_t) xfs_mask64hi(1)) | ((xfs_bmbt_rec_base_t)v << 9) | - (r->l0 & (xfs_bmbt_rec_base_t)XFS_MASK64LO(9)); + (r->l0 & (xfs_bmbt_rec_base_t)xfs_mask64lo(9)); } /* @@ -419,9 +419,9 @@ xfs_bmbt_set_state( { ASSERT(v == XFS_EXT_NORM || v == XFS_EXT_UNWRITTEN); if (v == XFS_EXT_NORM) - r->l0 &= XFS_MASK64LO(64 - BMBT_EXNTFLAG_BITLEN); + r->l0 &= xfs_mask64lo(64 - BMBT_EXNTFLAG_BITLEN); else - r->l0 |= XFS_MASK64HI(BMBT_EXNTFLAG_BITLEN); + r->l0 |= xfs_mask64hi(BMBT_EXNTFLAG_BITLEN); } /* From e6edbd1c1cbef278d58cdd8b046599ba8ac90cfc Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 8 Jan 2009 13:42:23 -0500 Subject: [PATCH 05/10] [XFS] fix compile of xfs_btree_readahead_lblock on m68k Change the left/right variables to the proper always 64bit xfs_dfsbo_t type because otherwise compilation fails for Geert on m68k without CONFIG_LBD: | fs/xfs/xfs_btree.c: In function 'xfs_btree_readahead_lblock': | fs/xfs/xfs_btree.c:736: warning: comparison is always true due to limited range of data type | fs/xfs/xfs_btree.c:741: warning: comparison is always true due to limited range of data type Reported-by: Geert Uytterhoeven Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Lachlan McIlroy --- fs/xfs/xfs_btree.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index 7ed59267420d..2c3ef20f8842 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -730,8 +730,8 @@ xfs_btree_readahead_lblock( struct xfs_btree_block *block) { int rval = 0; - xfs_fsblock_t left = be64_to_cpu(block->bb_u.l.bb_leftsib); - xfs_fsblock_t right = be64_to_cpu(block->bb_u.l.bb_rightsib); + xfs_dfsbno_t left = be64_to_cpu(block->bb_u.l.bb_leftsib); + xfs_dfsbno_t right = be64_to_cpu(block->bb_u.l.bb_rightsib); if ((lr & XFS_BTCUR_LEFTRA) && left != NULLDFSBNO) { xfs_btree_reada_bufl(cur->bc_mp, left, 1); From 15440319767942a363f282d6585303d3d75088ba Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 8 Jan 2009 14:00:00 -0500 Subject: [PATCH 06/10] [XFS] truncate readdir offsets to signed 32 bit values John Stanley reported EOVERFLOW errors in readdir from his self-build glibc. I traced this down to glibc enabling d_off overflow checks in one of the about five million different getdents implementations. In 2.6.28 Dave Woodhouse moved our readdir double buffering required for NFS4 readdirplus into nfsd and at that point we lost the capping of the directory offsets to 32 bit signed values. Johns glibc used getdents64 to even implement readdir for normal 32 bit offset dirents, and failed with EOVERFLOW only if this happens on the first dirent in a getdents call. I managed to come up with a testcase that uses raw getdents and does the EOVERFLOW check manually. We always hit it with our last entry due to the special end of directory marker. The patch below is a dumb version of just putting back the masking, to make sure we have the same behavior as in 2.6.27 and earlier. I will work on a better and cleaner fix for 2.6.30. Reported-by: John Stanley Tested-by: John Stanley Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Lachlan McIlroy --- fs/xfs/xfs_dir2_block.c | 7 ++++--- fs/xfs/xfs_dir2_leaf.c | 6 +++--- fs/xfs/xfs_dir2_sf.c | 15 ++++++++------- 3 files changed, 15 insertions(+), 13 deletions(-) diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c index e2fa0a1d8e96..e1f0a06aaf04 100644 --- a/fs/xfs/xfs_dir2_block.c +++ b/fs/xfs/xfs_dir2_block.c @@ -517,9 +517,9 @@ xfs_dir2_block_getdents( /* * If it didn't fit, set the final offset to here & return. */ - if (filldir(dirent, dep->name, dep->namelen, cook, + if (filldir(dirent, dep->name, dep->namelen, cook & 0x7fffffff, ino, DT_UNKNOWN)) { - *offset = cook; + *offset = cook & 0x7fffffff; xfs_da_brelse(NULL, bp); return 0; } @@ -529,7 +529,8 @@ xfs_dir2_block_getdents( * Reached the end of the block. * Set the offset to a non-existent block 1 and return. */ - *offset = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk + 1, 0); + *offset = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk + 1, 0) & + 0x7fffffff; xfs_da_brelse(NULL, bp); return 0; } diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c index 93535992cb60..ef805a374eec 100644 --- a/fs/xfs/xfs_dir2_leaf.c +++ b/fs/xfs/xfs_dir2_leaf.c @@ -1092,7 +1092,7 @@ xfs_dir2_leaf_getdents( * Won't fit. Return to caller. */ if (filldir(dirent, dep->name, dep->namelen, - xfs_dir2_byte_to_dataptr(mp, curoff), + xfs_dir2_byte_to_dataptr(mp, curoff) & 0x7fffffff, ino, DT_UNKNOWN)) break; @@ -1108,9 +1108,9 @@ xfs_dir2_leaf_getdents( * All done. Set output offset value to current offset. */ if (curoff > xfs_dir2_dataptr_to_byte(mp, XFS_DIR2_MAX_DATAPTR)) - *offset = XFS_DIR2_MAX_DATAPTR; + *offset = XFS_DIR2_MAX_DATAPTR & 0x7fffffff; else - *offset = xfs_dir2_byte_to_dataptr(mp, curoff); + *offset = xfs_dir2_byte_to_dataptr(mp, curoff) & 0x7fffffff; kmem_free(map); if (bp) xfs_da_brelse(NULL, bp); diff --git a/fs/xfs/xfs_dir2_sf.c b/fs/xfs/xfs_dir2_sf.c index b46af0013ec9..a8a8a6efad5b 100644 --- a/fs/xfs/xfs_dir2_sf.c +++ b/fs/xfs/xfs_dir2_sf.c @@ -752,8 +752,8 @@ xfs_dir2_sf_getdents( #if XFS_BIG_INUMS ino += mp->m_inoadd; #endif - if (filldir(dirent, ".", 1, dot_offset, ino, DT_DIR)) { - *offset = dot_offset; + if (filldir(dirent, ".", 1, dot_offset & 0x7fffffff, ino, DT_DIR)) { + *offset = dot_offset & 0x7fffffff; return 0; } } @@ -766,8 +766,8 @@ xfs_dir2_sf_getdents( #if XFS_BIG_INUMS ino += mp->m_inoadd; #endif - if (filldir(dirent, "..", 2, dotdot_offset, ino, DT_DIR)) { - *offset = dotdot_offset; + if (filldir(dirent, "..", 2, dotdot_offset & 0x7fffffff, ino, DT_DIR)) { + *offset = dotdot_offset & 0x7fffffff; return 0; } } @@ -791,14 +791,15 @@ xfs_dir2_sf_getdents( #endif if (filldir(dirent, sfep->name, sfep->namelen, - off, ino, DT_UNKNOWN)) { - *offset = off; + off & 0x7fffffff, ino, DT_UNKNOWN)) { + *offset = off & 0x7fffffff; return 0; } sfep = xfs_dir2_sf_nextentry(sfp, sfep); } - *offset = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk + 1, 0); + *offset = xfs_dir2_db_off_to_dataptr(mp, mp->m_dirdatablk + 1, 0) & + 0x7fffffff; return 0; } From 058652a37dd9eac18d6b8c1a311137c679de9dae Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 8 Jan 2009 13:42:25 -0500 Subject: [PATCH 07/10] [XFS] make xfs_ino_t an unsigned long long Currently xfs_ino_t is defined as a u64 which can either be an unsigned long long or on some 64 bit platforms and unsigned long. Just making it and unsigned long long mean's it's still always 64 bits wide, but we don't need to resort to cases to print it. Fixes a warning regression on 64 bit powerpc in current git. Signed-off-by: Christoph Hellwig Reviewed-by: Dave Chinner Signed-off-by: Lachlan McIlroy --- fs/xfs/xfs_types.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/xfs/xfs_types.h b/fs/xfs/xfs_types.h index baedbd14dc21..b2f724502f1b 100644 --- a/fs/xfs/xfs_types.h +++ b/fs/xfs/xfs_types.h @@ -45,7 +45,7 @@ typedef __uint32_t prid_t; /* project ID */ typedef __uint32_t inst_t; /* an instruction */ typedef __s64 xfs_off_t; /* type */ -typedef __u64 xfs_ino_t; /* type */ +typedef unsigned long long xfs_ino_t; /* type */ typedef __s64 xfs_daddr_t; /* type */ typedef char * xfs_caddr_t; /* type */ typedef __u32 xfs_dev_t; From 958f8c0e4fc311e23a40635a530c01aec366a6e8 Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:40:44 +1100 Subject: [PATCH 08/10] [XFS] remove old vmap cache XFS's vmap batching simply defers a number (up to 64) of vunmaps, and keeps track of them in a list. To purge the batch, it just goes through the list and calls vunamp on each one. This is pretty poor: a global TLB flush is generally still performed on each vunmap, with the most expensive parts of the operation being the broadcast IPIs and locking involved in the SMP callouts, and the locking involved in the vmap management -- none of these are avoided by just batching up the calls. I'm actually surprised it ever made much difference. (Now that the lazy vmap allocator is upstream, this description is not quite right, but the vunmap batching still doesn't seem to do much) Rip all this logic out of XFS completely. I will improve vmap performance and scalability directly in subsequent patch. Signed-off-by: Nick Piggin Reviewed-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/xfs/linux-2.6/xfs_buf.c | 75 +------------------------------------- 1 file changed, 1 insertion(+), 74 deletions(-) diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index cb329edc925b..0b2177a9fbdc 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -165,75 +165,6 @@ test_page_region( return (mask && (page_private(page) & mask) == mask); } -/* - * Mapping of multi-page buffers into contiguous virtual space - */ - -typedef struct a_list { - void *vm_addr; - struct a_list *next; -} a_list_t; - -static a_list_t *as_free_head; -static int as_list_len; -static DEFINE_SPINLOCK(as_lock); - -/* - * Try to batch vunmaps because they are costly. - */ -STATIC void -free_address( - void *addr) -{ - a_list_t *aentry; - -#ifdef CONFIG_XEN - /* - * Xen needs to be able to make sure it can get an exclusive - * RO mapping of pages it wants to turn into a pagetable. If - * a newly allocated page is also still being vmap()ed by xfs, - * it will cause pagetable construction to fail. This is a - * quick workaround to always eagerly unmap pages so that Xen - * is happy. - */ - vunmap(addr); - return; -#endif - - aentry = kmalloc(sizeof(a_list_t), GFP_NOWAIT); - if (likely(aentry)) { - spin_lock(&as_lock); - aentry->next = as_free_head; - aentry->vm_addr = addr; - as_free_head = aentry; - as_list_len++; - spin_unlock(&as_lock); - } else { - vunmap(addr); - } -} - -STATIC void -purge_addresses(void) -{ - a_list_t *aentry, *old; - - if (as_free_head == NULL) - return; - - spin_lock(&as_lock); - aentry = as_free_head; - as_free_head = NULL; - as_list_len = 0; - spin_unlock(&as_lock); - - while ((old = aentry) != NULL) { - vunmap(aentry->vm_addr); - aentry = aentry->next; - kfree(old); - } -} - /* * Internal xfs_buf_t object manipulation */ @@ -333,7 +264,7 @@ xfs_buf_free( uint i; if ((bp->b_flags & XBF_MAPPED) && (bp->b_page_count > 1)) - free_address(bp->b_addr - bp->b_offset); + vunmap(bp->b_addr - bp->b_offset); for (i = 0; i < bp->b_page_count; i++) { struct page *page = bp->b_pages[i]; @@ -455,8 +386,6 @@ _xfs_buf_map_pages( bp->b_addr = page_address(bp->b_pages[0]) + bp->b_offset; bp->b_flags |= XBF_MAPPED; } else if (flags & XBF_MAPPED) { - if (as_list_len > 64) - purge_addresses(); bp->b_addr = vmap(bp->b_pages, bp->b_page_count, VM_MAP, PAGE_KERNEL); if (unlikely(bp->b_addr == NULL)) @@ -1743,8 +1672,6 @@ xfsbufd( count++; } - if (as_list_len > 0) - purge_addresses(); if (count) blk_run_address_space(target->bt_mapping); From 0087167c9d5b1273e7e6bbe39a9ab13bdb9a39bb Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:43:09 +1100 Subject: [PATCH 09/10] [XFS] use scalable vmap API Implement XFS's large buffer support with the new vmap APIs. See the vmap rewrite (db64fe02) for some numbers. The biggest improvement that comes from using the new APIs is avoiding the global KVA allocation lock on every call. Signed-off-by: Nick Piggin Reviewed-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/xfs/linux-2.6/xfs_buf.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 0b2177a9fbdc..d71dc44e21ed 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -264,7 +264,7 @@ xfs_buf_free( uint i; if ((bp->b_flags & XBF_MAPPED) && (bp->b_page_count > 1)) - vunmap(bp->b_addr - bp->b_offset); + vm_unmap_ram(bp->b_addr - bp->b_offset, bp->b_page_count); for (i = 0; i < bp->b_page_count; i++) { struct page *page = bp->b_pages[i]; @@ -386,8 +386,8 @@ _xfs_buf_map_pages( bp->b_addr = page_address(bp->b_pages[0]) + bp->b_offset; bp->b_flags |= XBF_MAPPED; } else if (flags & XBF_MAPPED) { - bp->b_addr = vmap(bp->b_pages, bp->b_page_count, - VM_MAP, PAGE_KERNEL); + bp->b_addr = vm_map_ram(bp->b_pages, bp->b_page_count, + -1, PAGE_KERNEL); if (unlikely(bp->b_addr == NULL)) return -ENOMEM; bp->b_addr += bp->b_offset; From 0335cb76aa3fa913a2164bc9b669e5aef9d56fa3 Mon Sep 17 00:00:00 2001 From: Lachlan McIlroy Date: Wed, 31 Dec 2008 12:10:12 +1100 Subject: [PATCH 10/10] [XFS] Update maintainers New maintainer contact and new tree location. Reviewed-by: Bill O`Donnell Signed-off-by: Lachlan McIlroy --- MAINTAINERS | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index a01884407fe4..5f74bce18546 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4824,11 +4824,11 @@ S: Supported XFS FILESYSTEM P: Silicon Graphics Inc -P: Tim Shimmin +P: Bill O'Donnell M: xfs-masters@oss.sgi.com L: xfs@oss.sgi.com W: http://oss.sgi.com/projects/xfs -T: git git://oss.sgi.com:8090/xfs/xfs-2.6.git +T: git://oss.sgi.com/xfs/xfs.git S: Supported XILINX SYSTEMACE DRIVER