summaryrefslogtreecommitdiffstats
path: root/fs/gfs2/super.c
diff options
context:
space:
mode:
authorSteven Whitehouse2006-06-14 21:32:57 +0200
committerSteven Whitehouse2006-06-14 21:32:57 +0200
commitfeaa7bba026c181ce071d5a4884f7f9dd26207a1 (patch)
treec858deb225917265cb07820730e9764674d133e8 /fs/gfs2/super.c
parent[DLM] compat patch (diff)
downloadkernel-qcow2-linux-feaa7bba026c181ce071d5a4884f7f9dd26207a1.tar.gz
kernel-qcow2-linux-feaa7bba026c181ce071d5a4884f7f9dd26207a1.tar.xz
kernel-qcow2-linux-feaa7bba026c181ce071d5a4884f7f9dd26207a1.zip
[GFS2] Fix unlinked file handling
This patch fixes the way we have been dealing with unlinked, but still open files. It removes all limits (other than memory for inodes, as per every other filesystem) on numbers of these which we can support on GFS2. It also means that (like other fs) its the responsibility of the last process to close the file to deallocate the storage, rather than the person who did the unlinking. Note that with GFS2, those two events might take place on different nodes. Also there are a number of other changes: o We use the Linux inode subsystem as it was intended to be used, wrt allocating GFS2 inodes o The Linux inode cache is now the point which we use for local enforcement of only holding one copy of the inode in core at once (previous to this we used the glock layer). o We no longer use the unlinked "special" file. We just ignore it completely. This makes unlinking more efficient. o We now use the 4th block allocation state. The previously unused state is used to track unlinked but still open inodes. o gfs2_inoded is no longer needed o Several fields are now no longer needed (and removed) from the in core struct gfs2_inode o Several fields are no longer needed (and removed) from the in core superblock There are a number of future possible optimisations and clean ups which have been made possible by this patch. Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Diffstat (limited to 'fs/gfs2/super.c')
-rw-r--r--fs/gfs2/super.c45
1 files changed, 14 insertions, 31 deletions
diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c
index a943a505bc5a..f2d287660cc9 100644
--- a/fs/gfs2/super.c
+++ b/fs/gfs2/super.c
@@ -31,7 +31,6 @@
#include "rgrp.h"
#include "super.h"
#include "trans.h"
-#include "unlinked.h"
#include "util.h"
/**
@@ -55,7 +54,6 @@ void gfs2_tune_init(struct gfs2_tune *gt)
gt->gt_recoverd_secs = 60;
gt->gt_logd_secs = 1;
gt->gt_quotad_secs = 5;
- gt->gt_inoded_secs = 15;
gt->gt_quota_simul_sync = 64;
gt->gt_quota_warn_period = 10;
gt->gt_quota_scale_num = 1;
@@ -202,9 +200,6 @@ int gfs2_read_sb(struct gfs2_sbd *sdp, struct gfs2_glock *gl, int silent)
sdp->sd_hash_bsize = sdp->sd_sb.sb_bsize / 2;
sdp->sd_hash_bsize_shift = sdp->sd_sb.sb_bsize_shift - 1;
sdp->sd_hash_ptrs = sdp->sd_hash_bsize / sizeof(uint64_t);
- sdp->sd_ut_per_block = (sdp->sd_sb.sb_bsize -
- sizeof(struct gfs2_meta_header)) /
- sizeof(struct gfs2_unlinked_tag);
sdp->sd_qc_per_block = (sdp->sd_sb.sb_bsize -
sizeof(struct gfs2_meta_header)) /
sizeof(struct gfs2_quota_change);
@@ -277,7 +272,7 @@ int gfs2_read_sb(struct gfs2_sbd *sdp, struct gfs2_glock *gl, int silent)
int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
{
- struct gfs2_inode *dip = sdp->sd_jindex->u.generic_ip;
+ struct gfs2_inode *dip = GFS2_I(sdp->sd_jindex);
struct qstr name;
char buf[20];
struct gfs2_jdesc *jd;
@@ -296,8 +291,7 @@ int gfs2_jindex_hold(struct gfs2_sbd *sdp, struct gfs2_holder *ji_gh)
name.len = sprintf(buf, "journal%u", sdp->sd_journals);
name.hash = gfs2_disk_hash(name.name, name.len);
- error = gfs2_dir_search(sdp->sd_jindex,
- &name, NULL, NULL);
+ error = gfs2_dir_search(sdp->sd_jindex, &name, NULL, NULL);
if (error == -ENOENT) {
error = 0;
break;
@@ -423,22 +417,19 @@ struct gfs2_jdesc *gfs2_jdesc_find_dirty(struct gfs2_sbd *sdp)
int gfs2_jdesc_check(struct gfs2_jdesc *jd)
{
- struct gfs2_inode *ip = jd->jd_inode->u.generic_ip;
- struct gfs2_sbd *sdp = ip->i_sbd;
+ struct gfs2_inode *ip = GFS2_I(jd->jd_inode);
+ struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode);
int ar;
int error;
- if (ip->i_di.di_size < (8 << 20) ||
- ip->i_di.di_size > (1 << 30) ||
+ if (ip->i_di.di_size < (8 << 20) || ip->i_di.di_size > (1 << 30) ||
(ip->i_di.di_size & (sdp->sd_sb.sb_bsize - 1))) {
gfs2_consist_inode(ip);
return -EIO;
}
jd->jd_blocks = ip->i_di.di_size >> sdp->sd_sb.sb_bsize_shift;
- error = gfs2_write_alloc_required(ip,
- 0, ip->i_di.di_size,
- &ar);
+ error = gfs2_write_alloc_required(ip, 0, ip->i_di.di_size, &ar);
if (!error && ar) {
gfs2_consist_inode(ip);
error = -EIO;
@@ -456,7 +447,7 @@ int gfs2_jdesc_check(struct gfs2_jdesc *jd)
int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
{
- struct gfs2_inode *ip = sdp->sd_jdesc->jd_inode->u.generic_ip;
+ struct gfs2_inode *ip = GFS2_I(sdp->sd_jdesc->jd_inode);
struct gfs2_glock *j_gl = ip->i_gl;
struct gfs2_holder t_gh;
struct gfs2_log_header head;
@@ -484,9 +475,6 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
sdp->sd_log_sequence = head.lh_sequence + 1;
gfs2_log_pointers_init(sdp, head.lh_blkno);
- error = gfs2_unlinked_init(sdp);
- if (error)
- goto fail;
error = gfs2_quota_init(sdp);
if (error)
goto fail_unlinked;
@@ -498,7 +486,6 @@ int gfs2_make_fs_rw(struct gfs2_sbd *sdp)
return 0;
fail_unlinked:
- gfs2_unlinked_cleanup(sdp);
fail:
t_gh.gh_flags |= GL_NOCACHE;
@@ -519,7 +506,6 @@ int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
struct gfs2_holder t_gh;
int error;
- gfs2_unlinked_dealloc(sdp);
gfs2_quota_sync(sdp);
gfs2_statfs_sync(sdp);
@@ -537,7 +523,6 @@ int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
if (t_gh.gh_gl)
gfs2_glock_dq_uninit(&t_gh);
- gfs2_unlinked_cleanup(sdp);
gfs2_quota_cleanup(sdp);
return error;
@@ -545,9 +530,9 @@ int gfs2_make_fs_ro(struct gfs2_sbd *sdp)
int gfs2_statfs_init(struct gfs2_sbd *sdp)
{
- struct gfs2_inode *m_ip = sdp->sd_statfs_inode->u.generic_ip;
+ struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
struct gfs2_statfs_change *m_sc = &sdp->sd_statfs_master;
- struct gfs2_inode *l_ip = sdp->sd_sc_inode->u.generic_ip;
+ struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
struct gfs2_statfs_change *l_sc = &sdp->sd_statfs_local;
struct buffer_head *m_bh, *l_bh;
struct gfs2_holder gh;
@@ -594,7 +579,7 @@ int gfs2_statfs_init(struct gfs2_sbd *sdp)
void gfs2_statfs_change(struct gfs2_sbd *sdp, int64_t total, int64_t free,
int64_t dinodes)
{
- struct gfs2_inode *l_ip = sdp->sd_sc_inode->u.generic_ip;
+ struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
struct gfs2_statfs_change *l_sc = &sdp->sd_statfs_local;
struct buffer_head *l_bh;
int error;
@@ -620,8 +605,8 @@ void gfs2_statfs_change(struct gfs2_sbd *sdp, int64_t total, int64_t free,
int gfs2_statfs_sync(struct gfs2_sbd *sdp)
{
- struct gfs2_inode *m_ip = sdp->sd_statfs_inode->u.generic_ip;
- struct gfs2_inode *l_ip = sdp->sd_sc_inode->u.generic_ip;
+ struct gfs2_inode *m_ip = GFS2_I(sdp->sd_statfs_inode);
+ struct gfs2_inode *l_ip = GFS2_I(sdp->sd_sc_inode);
struct gfs2_statfs_change *m_sc = &sdp->sd_statfs_master;
struct gfs2_statfs_change *l_sc = &sdp->sd_statfs_local;
struct gfs2_holder gh;
@@ -852,10 +837,8 @@ static int gfs2_lock_fs_check_clean(struct gfs2_sbd *sdp,
error = -ENOMEM;
goto out;
}
- ip = jd->jd_inode->u.generic_ip;
- error = gfs2_glock_nq_init(ip->i_gl,
- LM_ST_SHARED, 0,
- &lfcc->gh);
+ ip = GFS2_I(jd->jd_inode);
+ error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &lfcc->gh);
if (error) {
kfree(lfcc);
goto out;