From ff8f33c8b30d7b7efdcf2548c7f6e64db6a89b29 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Wed, 11 Aug 2010 09:37:53 +0100 Subject: GFS2: New truncate sequence This updates GFS2's truncate code to use the new truncate sequence correctly. This is a stepping stone to being able to remove ip->i_disksize in favour of using i_size everywhere now that the two sizes are always identical. Signed-off-by: Steven Whitehouse Cc: Nick Piggin Cc: Christoph Hellwig --- fs/gfs2/aops.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) (limited to 'fs/gfs2/aops.c') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 194fe16d8418..f687f25fb7ff 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -696,13 +696,11 @@ out: page_cache_release(page); - /* - * XXX(truncate): the call below should probably be replaced with - * a call to the gfs2-specific truncate blocks helper to actually - * release disk blocks.. - */ + gfs2_trans_end(sdp); if (pos + len > ip->i_inode.i_size) - truncate_setsize(&ip->i_inode, ip->i_inode.i_size); + gfs2_trim_blocks(&ip->i_inode); + goto out_trans_fail; + out_endtrans: gfs2_trans_end(sdp); out_trans_fail: -- cgit v1.2.3 From a2e0f79939e09e74698564b88dee709db208e1e2 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Wed, 11 Aug 2010 09:53:11 +0100 Subject: GFS2: Remove i_disksize With the update of the truncate code, ip->i_disksize and inode->i_size are merely copies of each other. This means we can remove ip->i_disksize and use inode->i_size exclusively reducing the size of a GFS2 inode by 8 bytes. Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 6 +----- fs/gfs2/bmap.c | 12 +++++------- fs/gfs2/dir.c | 28 +++++++++++++++------------- fs/gfs2/file.c | 2 +- fs/gfs2/glops.c | 5 ++--- fs/gfs2/incore.h | 1 - fs/gfs2/inode.c | 9 ++++----- fs/gfs2/inode.h | 13 +++++++++++++ fs/gfs2/ops_fstype.c | 2 +- fs/gfs2/ops_inode.c | 10 +++++----- fs/gfs2/quota.c | 13 +++++-------- fs/gfs2/rgrp.c | 6 +++--- fs/gfs2/super.c | 11 +++++------ 13 files changed, 60 insertions(+), 58 deletions(-) (limited to 'fs/gfs2/aops.c') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index f687f25fb7ff..c92f36ba3fc9 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -800,10 +800,8 @@ static int gfs2_stuffed_write_end(struct inode *inode, struct buffer_head *dibh, page_cache_release(page); if (copied) { - if (inode->i_size < to) { + if (inode->i_size < to) i_size_write(inode, to); - ip->i_disksize = inode->i_size; - } gfs2_dinode_out(ip, di); mark_inode_dirty(inode); } @@ -874,8 +872,6 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, ret = generic_write_end(file, mapping, pos, len, copied, page, fsdata); if (ret > 0) { - if (inode->i_size > ip->i_disksize) - ip->i_disksize = inode->i_size; gfs2_dinode_out(ip, dibh->b_data); mark_inode_dirty(inode); } diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 20b971ad4973..04513e997df6 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -131,7 +131,7 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) if (error) goto out; - if (ip->i_disksize) { + if (i_size_read(&ip->i_inode)) { /* Get a free block, fill it with the stuffed data, and write it out to disk */ @@ -160,7 +160,7 @@ int gfs2_unstuff_dinode(struct gfs2_inode *ip, struct page *page) di = (struct gfs2_dinode *)dibh->b_data; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); - if (ip->i_disksize) { + if (i_size_read(&ip->i_inode)) { *(__be64 *)(di + 1) = cpu_to_be64(block); gfs2_add_inode_blocks(&ip->i_inode, 1); di->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); @@ -985,7 +985,6 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) } i_size_write(inode, newsize); - ip->i_disksize = newsize; ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_dinode_out(ip, dibh->b_data); @@ -1051,7 +1050,7 @@ static int trunc_end(struct gfs2_inode *ip) if (error) goto out; - if (!ip->i_disksize) { + if (!i_size_read(&ip->i_inode)) { ip->i_height = 0; ip->i_goal = ip->i_no_addr; gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); @@ -1167,7 +1166,6 @@ static int do_grow(struct inode *inode, u64 size) goto do_end_trans; i_size_write(inode, size); - ip->i_disksize = size; ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_trans_add_bh(ip->i_gl, dibh, 1); gfs2_dinode_out(ip, dibh->b_data); @@ -1219,7 +1217,7 @@ int gfs2_setattr_size(struct inode *inode, u64 newsize) int gfs2_truncatei_resume(struct gfs2_inode *ip) { int error; - error = trunc_dealloc(ip, ip->i_disksize); + error = trunc_dealloc(ip, i_size_read(&ip->i_inode)); if (!error) error = trunc_end(ip); return error; @@ -1260,7 +1258,7 @@ int gfs2_write_alloc_required(struct gfs2_inode *ip, u64 offset, shift = sdp->sd_sb.sb_bsize_shift; BUG_ON(gfs2_is_dir(ip)); - end_of_file = (ip->i_disksize + sdp->sd_sb.sb_bsize - 1) >> shift; + end_of_file = (i_size_read(&ip->i_inode) + sdp->sd_sb.sb_bsize - 1) >> shift; lblock = offset >> shift; lblock_stop = (offset + len + sdp->sd_sb.sb_bsize - 1) >> shift; if (lblock_stop > end_of_file) diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index b9dd88a78dd4..c1042ae438cc 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -127,8 +127,8 @@ static int gfs2_dir_write_stuffed(struct gfs2_inode *ip, const char *buf, gfs2_trans_add_bh(ip->i_gl, dibh, 1); memcpy(dibh->b_data + offset + sizeof(struct gfs2_dinode), buf, size); - if (ip->i_disksize < offset + size) - ip->i_disksize = offset + size; + if (ip->i_inode.i_size < offset + size) + i_size_write(&ip->i_inode, offset + size); ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_dinode_out(ip, dibh->b_data); @@ -225,8 +225,8 @@ out: if (error) return error; - if (ip->i_disksize < offset + copied) - ip->i_disksize = offset + copied; + if (ip->i_inode.i_size < offset + copied) + i_size_write(&ip->i_inode, offset + copied); ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_trans_add_bh(ip->i_gl, dibh, 1); @@ -275,12 +275,13 @@ static int gfs2_dir_read_data(struct gfs2_inode *ip, char *buf, u64 offset, unsigned int o; int copied = 0; int error = 0; + u64 disksize = i_size_read(&ip->i_inode); - if (offset >= ip->i_disksize) + if (offset >= disksize) return 0; - if (offset + size > ip->i_disksize) - size = ip->i_disksize - offset; + if (offset + size > disksize) + size = disksize - offset; if (!size) return 0; @@ -727,7 +728,7 @@ static struct gfs2_dirent *gfs2_dirent_search(struct inode *inode, unsigned hsize = 1 << ip->i_depth; unsigned index; u64 ln; - if (hsize * sizeof(u64) != ip->i_disksize) { + if (hsize * sizeof(u64) != i_size_read(inode)) { gfs2_consist_inode(ip); return ERR_PTR(-EIO); } @@ -879,7 +880,7 @@ static int dir_make_exhash(struct inode *inode) for (x = sdp->sd_hash_ptrs; x--; lp++) *lp = cpu_to_be64(bn); - dip->i_disksize = sdp->sd_sb.sb_bsize / 2; + i_size_write(inode, sdp->sd_sb.sb_bsize / 2); gfs2_add_inode_blocks(&dip->i_inode, 1); dip->i_diskflags |= GFS2_DIF_EXHASH; @@ -1057,11 +1058,12 @@ static int dir_double_exhash(struct gfs2_inode *dip) u64 *buf; u64 *from, *to; u64 block; + u64 disksize = i_size_read(&dip->i_inode); int x; int error = 0; hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != dip->i_disksize) { + if (hsize * sizeof(u64) != disksize) { gfs2_consist_inode(dip); return -EIO; } @@ -1072,7 +1074,7 @@ static int dir_double_exhash(struct gfs2_inode *dip) if (!buf) return -ENOMEM; - for (block = dip->i_disksize >> sdp->sd_hash_bsize_shift; block--;) { + for (block = disksize >> sdp->sd_hash_bsize_shift; block--;) { error = gfs2_dir_read_data(dip, (char *)buf, block * sdp->sd_hash_bsize, sdp->sd_hash_bsize, 1); @@ -1370,7 +1372,7 @@ static int dir_e_read(struct inode *inode, u64 *offset, void *opaque, unsigned depth = 0; hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != dip->i_disksize) { + if (hsize * sizeof(u64) != i_size_read(inode)) { gfs2_consist_inode(dip); return -EIO; } @@ -1784,7 +1786,7 @@ static int foreach_leaf(struct gfs2_inode *dip, leaf_call_t lc, void *data) int error = 0; hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != dip->i_disksize) { + if (hsize * sizeof(u64) != i_size_read(&dip->i_inode)) { gfs2_consist_inode(dip); return -EIO; } diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 4edd662c8232..daadcd2e755f 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -491,7 +491,7 @@ static int gfs2_open(struct inode *inode, struct file *file) goto fail; if (!(file->f_flags & O_LARGEFILE) && - ip->i_disksize > MAX_NON_LFS) { + i_size_read(inode) > MAX_NON_LFS) { error = -EOVERFLOW; goto fail_gunlock; } diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 49f97d3bb690..621d80e8fb2a 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -262,13 +262,12 @@ static int inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl) const struct gfs2_inode *ip = gl->gl_object; if (ip == NULL) return 0; - gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu/%llu\n", + gfs2_print_dbg(seq, " I: n:%llu/%llu t:%u f:0x%02lx d:0x%08x s:%llu\n", (unsigned long long)ip->i_no_formal_ino, (unsigned long long)ip->i_no_addr, IF2DT(ip->i_inode.i_mode), ip->i_flags, (unsigned int)ip->i_diskflags, - (unsigned long long)ip->i_inode.i_size, - (unsigned long long)ip->i_disksize); + (unsigned long long)i_size_read(&ip->i_inode)); return 0; } diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index fdbf4b366fa5..c11971775275 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -267,7 +267,6 @@ struct gfs2_inode { u64 i_no_formal_ino; u64 i_generation; u64 i_eattr; - loff_t i_disksize; unsigned long i_flags; /* GIF_... */ struct gfs2_glock *i_gl; /* Move into i_gh? */ struct gfs2_holder i_iopen_gh; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 08140f185a37..06370f8bd8cf 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -359,8 +359,7 @@ static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) * to do that. */ ip->i_inode.i_nlink = be32_to_cpu(str->di_nlink); - ip->i_disksize = be64_to_cpu(str->di_size); - i_size_write(&ip->i_inode, ip->i_disksize); + i_size_write(&ip->i_inode, be64_to_cpu(str->di_size)); gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); atime.tv_sec = be64_to_cpu(str->di_atime); atime.tv_nsec = be32_to_cpu(str->di_atime_nsec); @@ -1055,7 +1054,7 @@ void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) str->di_uid = cpu_to_be32(ip->i_inode.i_uid); str->di_gid = cpu_to_be32(ip->i_inode.i_gid); str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); - str->di_size = cpu_to_be64(ip->i_disksize); + str->di_size = cpu_to_be64(i_size_read(&ip->i_inode)); str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec); str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec); @@ -1085,8 +1084,8 @@ void gfs2_dinode_print(const struct gfs2_inode *ip) (unsigned long long)ip->i_no_formal_ino); printk(KERN_INFO " no_addr = %llu\n", (unsigned long long)ip->i_no_addr); - printk(KERN_INFO " i_disksize = %llu\n", - (unsigned long long)ip->i_disksize); + printk(KERN_INFO " i_size = %llu\n", + (unsigned long long)i_size_read(&ip->i_inode)); printk(KERN_INFO " blocks = %llu\n", (unsigned long long)gfs2_get_inode_blocks(&ip->i_inode)); printk(KERN_INFO " i_goal = %llu\n", diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index 300ada3f21de..15ff4df20aab 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -80,6 +80,19 @@ static inline void gfs2_inum_out(const struct gfs2_inode *ip, dent->de_inum.no_addr = cpu_to_be64(ip->i_no_addr); } +static inline int gfs2_check_internal_file_size(struct inode *inode, + u64 minsize, u64 maxsize) +{ + u64 size = i_size_read(inode); + if (size < minsize || size > maxsize) + goto err; + if (size & ((1 << inode->i_blkbits) - 1)) + goto err; + return 0; +err: + gfs2_consist_inode(GFS2_I(inode)); + return -EIO; +} extern void gfs2_set_iop(struct inode *inode); extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 4d4b1e8ac64c..5b5c87dfbfee 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -586,7 +586,7 @@ static int map_journal_extents(struct gfs2_sbd *sdp) prev_db = 0; - for (lb = 0; lb < ip->i_disksize >> sdp->sd_sb.sb_bsize_shift; lb++) { + for (lb = 0; lb < i_size_read(jd->jd_inode) >> sdp->sd_sb.sb_bsize_shift; lb++) { bh.b_state = 0; bh.b_blocknr = 0; bh.b_size = 1 << ip->i_inode.i_blkbits; diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index 1d3f2fb466bd..ee6ffd590418 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -406,7 +406,6 @@ static int gfs2_symlink(struct inode *dir, struct dentry *dentry, ip = ghs[1].gh_gl->gl_object; - ip->i_disksize = size; i_size_write(inode, size); error = gfs2_meta_inode_buffer(ip, &dibh); @@ -461,7 +460,7 @@ static int gfs2_mkdir(struct inode *dir, struct dentry *dentry, int mode) ip = ghs[1].gh_gl->gl_object; ip->i_inode.i_nlink = 2; - ip->i_disksize = sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode); + i_size_write(inode, sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)); ip->i_diskflags |= GFS2_DIF_JDATA; ip->i_entries = 2; @@ -990,7 +989,7 @@ static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) struct gfs2_inode *ip = GFS2_I(dentry->d_inode); struct gfs2_holder i_gh; struct buffer_head *dibh; - unsigned int x; + unsigned int x, size; char *buf; int error; @@ -1002,7 +1001,8 @@ static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) return NULL; } - if (!ip->i_disksize) { + size = (unsigned int)i_size_read(&ip->i_inode); + if (size == 0) { gfs2_consist_inode(ip); buf = ERR_PTR(-EIO); goto out; @@ -1014,7 +1014,7 @@ static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) goto out; } - x = ip->i_disksize + 1; + x = size + 1; buf = kmalloc(x, GFP_NOFS); if (!buf) buf = ERR_PTR(-ENOMEM); diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 1bc6b5695e6d..9bc6dd9a5443 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -735,10 +735,8 @@ get_a_page: goto out; size = loc + sizeof(struct gfs2_quota); - if (size > inode->i_size) { - ip->i_disksize = size; + if (size > inode->i_size) i_size_write(inode, size); - } inode->i_mtime = inode->i_atime = CURRENT_TIME; gfs2_trans_add_bh(ip->i_gl, dibh, 1); gfs2_dinode_out(ip, dibh->b_data); @@ -1190,18 +1188,17 @@ static void gfs2_quota_change_in(struct gfs2_quota_change_host *qc, const void * int gfs2_quota_init(struct gfs2_sbd *sdp) { struct gfs2_inode *ip = GFS2_I(sdp->sd_qc_inode); - unsigned int blocks = ip->i_disksize >> sdp->sd_sb.sb_bsize_shift; + u64 size = i_size_read(sdp->sd_qc_inode); + unsigned int blocks = size >> sdp->sd_sb.sb_bsize_shift; unsigned int x, slot = 0; unsigned int found = 0; u64 dblock; u32 extlen = 0; int error; - if (!ip->i_disksize || ip->i_disksize > (64 << 20) || - ip->i_disksize & (sdp->sd_sb.sb_bsize - 1)) { - gfs2_consist_inode(ip); + if (gfs2_check_internal_file_size(sdp->sd_qc_inode, 1, 64 << 20)) return -EIO; - } + sdp->sd_quota_slots = blocks * sdp->sd_qc_per_block; sdp->sd_quota_chunks = DIV_ROUND_UP(sdp->sd_quota_slots, 8 * PAGE_SIZE); diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 171a744f8e45..370c29b536ea 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -500,7 +500,7 @@ u64 gfs2_ri_total(struct gfs2_sbd *sdp) for (rgrps = 0;; rgrps++) { loff_t pos = rgrps * sizeof(struct gfs2_rindex); - if (pos + sizeof(struct gfs2_rindex) >= ip->i_disksize) + if (pos + sizeof(struct gfs2_rindex) >= i_size_read(inode)) break; error = gfs2_internal_read(ip, &ra_state, buf, &pos, sizeof(struct gfs2_rindex)); @@ -588,7 +588,7 @@ static int gfs2_ri_update(struct gfs2_inode *ip) struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct inode *inode = &ip->i_inode; struct file_ra_state ra_state; - u64 rgrp_count = ip->i_disksize; + u64 rgrp_count = i_size_read(inode); int error; do_div(rgrp_count, sizeof(struct gfs2_rindex)); @@ -628,7 +628,7 @@ static int gfs2_ri_update_special(struct gfs2_inode *ip) for (sdp->sd_rgrps = 0;; sdp->sd_rgrps++) { /* Ignore partials */ if ((sdp->sd_rgrps + 1) * sizeof(struct gfs2_rindex) > - ip->i_disksize) + i_size_read(inode)) break; error = read_rindex_entry(ip, &ra_state); if (error) { diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 77cb9f830ee4..e031fa4965a3 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -342,15 +342,14 @@ int gfs2_jdesc_check(struct gfs2_jdesc *jd) { struct gfs2_inode *ip = GFS2_I(jd->jd_inode); struct gfs2_sbd *sdp = GFS2_SB(jd->jd_inode); + u64 size = i_size_read(jd->jd_inode); - if (ip->i_disksize < (8 << 20) || ip->i_disksize > (1 << 30) || - (ip->i_disksize & (sdp->sd_sb.sb_bsize - 1))) { - gfs2_consist_inode(ip); + if (gfs2_check_internal_file_size(jd->jd_inode, 8 << 20, 1 << 30)) return -EIO; - } - jd->jd_blocks = ip->i_disksize >> sdp->sd_sb.sb_bsize_shift; - if (gfs2_write_alloc_required(ip, 0, ip->i_disksize)) { + jd->jd_blocks = size >> sdp->sd_sb.sb_bsize_shift; + + if (gfs2_write_alloc_required(ip, 0, size)) { gfs2_consist_inode(ip); return -EIO; } -- cgit v1.2.3 From 3921120e757f9167f3fcd3a1781239824471b14d Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Fri, 20 Aug 2010 00:21:02 -0500 Subject: GFS2: fallocate support This patch adds support for fallocate to gfs2. Since the gfs2 does not support uninitialized data blocks, it must write out zeros to all the blocks. However, since it does not need to lock any pages to read from, gfs2 can write out the zero blocks much more efficiently. On a moderately full filesystem, fallocate works around 5 times faster on average. The fallocate call also allows gfs2 to add blocks to the file without changing the filesize, which will make it possible for gfs2 to preallocate space for the rindex file, so that gfs2 can grow a completely full filesystem. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 4 +- fs/gfs2/incore.h | 1 + fs/gfs2/inode.h | 2 + fs/gfs2/ops_inode.c | 254 ++++++++++++++++++++++++++++++++++++++++++++++++++++ fs/gfs2/rgrp.c | 12 +++ fs/gfs2/trans.h | 1 + 6 files changed, 272 insertions(+), 2 deletions(-) (limited to 'fs/gfs2/aops.c') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index c92f36ba3fc9..180ef8a6de6b 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -36,8 +36,8 @@ #include "glops.h" -static void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, - unsigned int from, unsigned int to) +void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, + unsigned int from, unsigned int to) { struct buffer_head *head = page_buffers(page); unsigned int bsize = head->b_size; diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index c11971775275..578234bb03f8 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -571,6 +571,7 @@ struct gfs2_sbd { struct list_head sd_rindex_mru_list; struct gfs2_rgrpd *sd_rindex_forward; unsigned int sd_rgrps; + unsigned int sd_max_rg_data; /* Journal index stuff */ diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index 15ff4df20aab..6720d7d5fbc6 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -19,6 +19,8 @@ extern int gfs2_releasepage(struct page *page, gfp_t gfp_mask); extern int gfs2_internal_read(struct gfs2_inode *ip, struct file_ra_state *ra_state, char *buf, loff_t *pos, unsigned size); +extern void gfs2_page_add_databufs(struct gfs2_inode *ip, struct page *page, + unsigned int from, unsigned int to); extern void gfs2_set_aops(struct inode *inode); static inline int gfs2_is_stuffed(const struct gfs2_inode *ip) diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index ee6ffd590418..f6da0d7676e2 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -18,6 +18,8 @@ #include #include #include +#include +#include #include #include "gfs2.h" @@ -1277,6 +1279,257 @@ static int gfs2_removexattr(struct dentry *dentry, const char *name) return ret; } +static void empty_write_end(struct page *page, unsigned from, + unsigned to) +{ + struct gfs2_inode *ip = GFS2_I(page->mapping->host); + + page_zero_new_buffers(page, from, to); + flush_dcache_page(page); + mark_page_accessed(page); + + if (!gfs2_is_writeback(ip)) + gfs2_page_add_databufs(ip, page, from, to); + + block_commit_write(page, from, to); +} + + +static int write_empty_blocks(struct page *page, unsigned from, unsigned to) +{ + unsigned start, end, next; + struct buffer_head *bh, *head; + int error; + + if (!page_has_buffers(page)) { + error = block_prepare_write(page, from, to, gfs2_block_map); + if (unlikely(error)) + return error; + + empty_write_end(page, from, to); + return 0; + } + + bh = head = page_buffers(page); + next = end = 0; + while (next < from) { + next += bh->b_size; + bh = bh->b_this_page; + } + start = next; + do { + next += bh->b_size; + if (buffer_mapped(bh)) { + if (end) { + error = block_prepare_write(page, start, end, + gfs2_block_map); + if (unlikely(error)) + return error; + empty_write_end(page, start, end); + end = 0; + } + start = next; + } + else + end = next; + bh = bh->b_this_page; + } while (next < to); + + if (end) { + error = block_prepare_write(page, start, end, gfs2_block_map); + if (unlikely(error)) + return error; + empty_write_end(page, start, end); + } + + return 0; +} + +static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, + int mode) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct buffer_head *dibh; + int error; + u64 start = offset >> PAGE_CACHE_SHIFT; + unsigned int start_offset = offset & ~PAGE_CACHE_MASK; + u64 end = (offset + len - 1) >> PAGE_CACHE_SHIFT; + pgoff_t curr; + struct page *page; + unsigned int end_offset = (offset + len) & ~PAGE_CACHE_MASK; + unsigned int from, to; + + if (!end_offset) + end_offset = PAGE_CACHE_SIZE; + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (unlikely(error)) + goto out; + + gfs2_trans_add_bh(ip->i_gl, dibh, 1); + + if (gfs2_is_stuffed(ip)) { + error = gfs2_unstuff_dinode(ip, NULL); + if (unlikely(error)) + goto out; + } + + curr = start; + offset = start << PAGE_CACHE_SHIFT; + from = start_offset; + to = PAGE_CACHE_SIZE; + while (curr <= end) { + page = grab_cache_page_write_begin(inode->i_mapping, curr, + AOP_FLAG_NOFS); + if (unlikely(!page)) { + error = -ENOMEM; + goto out; + } + + if (curr == end) + to = end_offset; + error = write_empty_blocks(page, from, to); + if (!error && offset + to > inode->i_size && + !(mode & FALLOC_FL_KEEP_SIZE)) { + i_size_write(inode, offset + to); + } + unlock_page(page); + page_cache_release(page); + if (error) + goto out; + curr++; + offset += PAGE_CACHE_SIZE; + from = 0; + } + + gfs2_dinode_out(ip, dibh->b_data); + mark_inode_dirty(inode); + + brelse(dibh); + +out: + return error; +} + +static void calc_max_reserv(struct gfs2_inode *ip, loff_t max, loff_t *len, + unsigned int *data_blocks, unsigned int *ind_blocks) +{ + const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + unsigned int max_blocks = ip->i_alloc->al_rgd->rd_free_clone; + unsigned int tmp, max_data = max_blocks - 3 * (sdp->sd_max_height - 1); + + for (tmp = max_data; tmp > sdp->sd_diptrs;) { + tmp = DIV_ROUND_UP(tmp, sdp->sd_inptrs); + max_data -= tmp; + } + /* This calculation isn't the exact reverse of gfs2_write_calc_reserve, + so it might end up with fewer data blocks */ + if (max_data <= *data_blocks) + return; + *data_blocks = max_data; + *ind_blocks = max_blocks - max_data; + *len = ((loff_t)max_data - 3) << sdp->sd_sb.sb_bsize_shift; + if (*len > max) { + *len = max; + gfs2_write_calc_reserv(ip, max, data_blocks, ind_blocks); + } +} + +static long gfs2_fallocate(struct inode *inode, int mode, loff_t offset, + loff_t len) +{ + struct gfs2_sbd *sdp = GFS2_SB(inode); + struct gfs2_inode *ip = GFS2_I(inode); + unsigned int data_blocks = 0, ind_blocks = 0, rblocks; + loff_t bytes, max_bytes; + struct gfs2_alloc *al; + int error; + loff_t next = (offset + len - 1) >> sdp->sd_sb.sb_bsize_shift; + next = (next + 1) << sdp->sd_sb.sb_bsize_shift; + + offset = (offset >> sdp->sd_sb.sb_bsize_shift) << + sdp->sd_sb.sb_bsize_shift; + + len = next - offset; + bytes = sdp->sd_max_rg_data * sdp->sd_sb.sb_bsize / 2; + if (!bytes) + bytes = UINT_MAX; + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &ip->i_gh); + error = gfs2_glock_nq(&ip->i_gh); + if (unlikely(error)) + goto out_uninit; + + if (!gfs2_write_alloc_required(ip, offset, len)) + goto out_unlock; + + while (len > 0) { + if (len < bytes) + bytes = len; + al = gfs2_alloc_get(ip); + if (!al) { + error = -ENOMEM; + goto out_unlock; + } + + error = gfs2_quota_lock_check(ip); + if (error) + goto out_alloc_put; + +retry: + gfs2_write_calc_reserv(ip, bytes, &data_blocks, &ind_blocks); + + al->al_requested = data_blocks + ind_blocks; + error = gfs2_inplace_reserve(ip); + if (error) { + if (error == -ENOSPC && bytes > sdp->sd_sb.sb_bsize) { + bytes >>= 1; + goto retry; + } + goto out_qunlock; + } + max_bytes = bytes; + calc_max_reserv(ip, len, &max_bytes, &data_blocks, &ind_blocks); + al->al_requested = data_blocks + ind_blocks; + + rblocks = RES_DINODE + ind_blocks + RES_STATFS + RES_QUOTA + + RES_RG_HDR + ip->i_alloc->al_rgd->rd_length; + if (gfs2_is_jdata(ip)) + rblocks += data_blocks ? data_blocks : 1; + + error = gfs2_trans_begin(sdp, rblocks, + PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize); + if (error) + goto out_trans_fail; + + error = fallocate_chunk(inode, offset, max_bytes, mode); + gfs2_trans_end(sdp); + + if (error) + goto out_trans_fail; + + len -= max_bytes; + offset += max_bytes; + gfs2_inplace_release(ip); + gfs2_quota_unlock(ip); + gfs2_alloc_put(ip); + } + goto out_unlock; + +out_trans_fail: + gfs2_inplace_release(ip); +out_qunlock: + gfs2_quota_unlock(ip); +out_alloc_put: + gfs2_alloc_put(ip); +out_unlock: + gfs2_glock_dq(&ip->i_gh); +out_uninit: + gfs2_holder_uninit(&ip->i_gh); + return error; +} + + static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, u64 start, u64 len) { @@ -1327,6 +1580,7 @@ const struct inode_operations gfs2_file_iops = { .getxattr = gfs2_getxattr, .listxattr = gfs2_listxattr, .removexattr = gfs2_removexattr, + .fallocate = gfs2_fallocate, .fiemap = gfs2_fiemap, }; diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 66b6d4d8b1d2..f9ddcf401753 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -589,6 +589,8 @@ static int gfs2_ri_update(struct gfs2_inode *ip) struct inode *inode = &ip->i_inode; struct file_ra_state ra_state; u64 rgrp_count = i_size_read(inode); + struct gfs2_rgrpd *rgd; + unsigned int max_data = 0; int error; do_div(rgrp_count, sizeof(struct gfs2_rindex)); @@ -603,6 +605,10 @@ static int gfs2_ri_update(struct gfs2_inode *ip) } } + list_for_each_entry(rgd, &sdp->sd_rindex_list, rd_list) + if (rgd->rd_data > max_data) + max_data = rgd->rd_data; + sdp->sd_max_rg_data = max_data; sdp->sd_rindex_uptodate = 1; return 0; } @@ -622,6 +628,8 @@ static int gfs2_ri_update_special(struct gfs2_inode *ip) struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct inode *inode = &ip->i_inode; struct file_ra_state ra_state; + struct gfs2_rgrpd *rgd; + unsigned int max_data = 0; int error; file_ra_state_init(&ra_state, inode->i_mapping); @@ -636,6 +644,10 @@ static int gfs2_ri_update_special(struct gfs2_inode *ip) return error; } } + list_for_each_entry(rgd, &sdp->sd_rindex_list, rd_list) + if (rgd->rd_data > max_data) + max_data = rgd->rd_data; + sdp->sd_max_rg_data = max_data; sdp->sd_rindex_uptodate = 1; return 0; diff --git a/fs/gfs2/trans.h b/fs/gfs2/trans.h index edf9d4bd908e..b849eb7ad37d 100644 --- a/fs/gfs2/trans.h +++ b/fs/gfs2/trans.h @@ -20,6 +20,7 @@ struct gfs2_glock; #define RES_JDATA 1 #define RES_DATA 1 #define RES_LEAF 1 +#define RES_RG_HDR 1 #define RES_RG_BIT 2 #define RES_EATTR 1 #define RES_STATFS 1 -- cgit v1.2.3 From bf97b6734e027cc18abad420ab88f861f65d7816 Mon Sep 17 00:00:00 2001 From: Benjamin Marzinski Date: Mon, 27 Sep 2010 16:00:04 -0500 Subject: GFS2: reserve more blocks for transactions Some of the functions in GFS2 were not reserving space in the transaction for the resource group header and the resource groups bitblocks that get added when you do allocation. GFS2 now makes sure to reserve space for the resource group header and either all the bitblocks in the resource group, or one for each block that it may allocate, whichever is smaller using the new gfs2_rg_blocks() inline function. Signed-off-by: Benjamin Marzinski Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 2 ++ fs/gfs2/bmap.c | 2 +- fs/gfs2/file.c | 4 +++- fs/gfs2/ops_inode.c | 6 +++--- fs/gfs2/quota.c | 3 ++- fs/gfs2/trans.h | 8 ++++++++ fs/gfs2/xattr.c | 2 +- 7 files changed, 20 insertions(+), 7 deletions(-) (limited to 'fs/gfs2/aops.c') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 180ef8a6de6b..1bf178831ae8 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -663,6 +663,8 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, rblocks += RES_STATFS + RES_QUOTA; if (&ip->i_inode == sdp->sd_rindex) rblocks += 2 * RES_STATFS; + if (alloc_required) + rblocks += gfs2_rg_blocks(al); error = gfs2_trans_begin(sdp, rblocks, PAGE_CACHE_SIZE/sdp->sd_sb.sb_bsize); diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 04513e997df6..5476c066d4ee 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -1151,7 +1151,7 @@ static int do_grow(struct inode *inode, u64 size) goto do_grow_qunlock; } - error = gfs2_trans_begin(sdp, RES_DINODE + 1, 0); + error = gfs2_trans_begin(sdp, RES_DINODE + RES_STATFS + RES_RG_BIT, 0); if (error) goto do_grow_release; diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index daadcd2e755f..237ee6a940df 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -382,8 +382,10 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) rblocks = RES_DINODE + ind_blocks; if (gfs2_is_jdata(ip)) rblocks += data_blocks ? data_blocks : 1; - if (ind_blocks || data_blocks) + if (ind_blocks || data_blocks) { rblocks += RES_STATFS + RES_QUOTA; + rblocks += gfs2_rg_blocks(al); + } ret = gfs2_trans_begin(sdp, rblocks, 0); if (ret) goto out_trans_fail; diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index 98a94cfc2bb2..fba00171d915 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -219,7 +219,7 @@ static int gfs2_link(struct dentry *old_dentry, struct inode *dir, goto out_gunlock_q; error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - al->al_rgd->rd_length + + gfs2_rg_blocks(al) + 2 * RES_DINODE + RES_STATFS + RES_QUOTA, 0); if (error) @@ -884,7 +884,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry, goto out_gunlock_q; error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - al->al_rgd->rd_length + + gfs2_rg_blocks(al) + 4 * RES_DINODE + 4 * RES_LEAF + RES_STATFS + RES_QUOTA + 4, 0); if (error) @@ -1481,7 +1481,7 @@ retry: al->al_requested = data_blocks + ind_blocks; rblocks = RES_DINODE + ind_blocks + RES_STATFS + RES_QUOTA + - RES_RG_HDR + ip->i_alloc->al_rgd->rd_length; + RES_RG_HDR + gfs2_rg_blocks(al); if (gfs2_is_jdata(ip)) rblocks += data_blocks ? data_blocks : 1; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index 9bc6dd9a5443..58a9b9998b42 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -815,7 +815,7 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) goto out_alloc; if (nalloc) - blocks += al->al_rgd->rd_length + nalloc * ind_blocks + RES_STATFS; + blocks += gfs2_rg_blocks(al) + nalloc * ind_blocks + RES_STATFS; error = gfs2_trans_begin(sdp, blocks, 0); if (error) @@ -1586,6 +1586,7 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, error = gfs2_inplace_reserve(ip); if (error) goto out_alloc; + blocks += gfs2_rg_blocks(al); } error = gfs2_trans_begin(sdp, blocks + RES_DINODE + 1, 0); diff --git a/fs/gfs2/trans.h b/fs/gfs2/trans.h index b849eb7ad37d..fb56b783e028 100644 --- a/fs/gfs2/trans.h +++ b/fs/gfs2/trans.h @@ -26,6 +26,14 @@ struct gfs2_glock; #define RES_STATFS 1 #define RES_QUOTA 2 +/* reserve either the number of blocks to be allocated plus the rg header + * block, or all of the blocks in the rg, whichever is smaller */ +static inline unsigned int gfs2_rg_blocks(const struct gfs2_alloc *al) +{ + return (al->al_requested < al->al_rgd->rd_length)? + al->al_requested + 1 : al->al_rgd->rd_length; +} + int gfs2_trans_begin(struct gfs2_sbd *sdp, unsigned int blocks, unsigned int revokes); diff --git a/fs/gfs2/xattr.c b/fs/gfs2/xattr.c index 776af6eb4bcb..30b58f07c8a6 100644 --- a/fs/gfs2/xattr.c +++ b/fs/gfs2/xattr.c @@ -734,7 +734,7 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er, goto out_gunlock_q; error = gfs2_trans_begin(GFS2_SB(&ip->i_inode), - blks + al->al_rgd->rd_length + + blks + gfs2_rg_blocks(al) + RES_DINODE + RES_STATFS + RES_QUOTA, 0); if (error) goto out_ipres; -- cgit v1.2.3 From d594845106f34c079a6b05be01a37e4883c3bf4c Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Tue, 28 Sep 2010 10:17:47 +0100 Subject: GFS2: Fix compiler warning from previous patch This shouldn't really be required, but gcc can't tell that "al" is only accessed when initialised. Signed-off-by: Steven Whitehouse --- fs/gfs2/aops.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/gfs2/aops.c') diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 1bf178831ae8..6b24afb96aae 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -615,7 +615,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, unsigned int data_blocks = 0, ind_blocks = 0, rblocks; int alloc_required; int error = 0; - struct gfs2_alloc *al; + struct gfs2_alloc *al = NULL; pgoff_t index = pos >> PAGE_CACHE_SHIFT; unsigned from = pos & (PAGE_CACHE_SIZE - 1); unsigned to = from + len; -- cgit v1.2.3