From 51ff87bdd9f21a5d3672517b75d25ab5842d94a8 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Mon, 15 Oct 2007 14:42:35 +0100 Subject: [GFS2] Clean up internal read function As requested by Christoph, this patch cleans up GFS2's internal read function so that it no longer uses the do_generic_mapping_read function. This function is obsolete and GFS2 is the last user of it. As a side effect the internal read code gets smaller and easier to read and gfs2_readpage is split into two. One function has the locking and the other function has the rest of the logic. Signed-off-by: Steven Whitehouse Cc: Christoph Hellwig --- fs/gfs2/ops_file.c | 45 --------------------------------------------- 1 file changed, 45 deletions(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index bb11fd6752d3..a729c86b8be1 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -33,7 +33,6 @@ #include "lm.h" #include "log.h" #include "meta_io.h" -#include "ops_file.h" #include "ops_vm.h" #include "quota.h" #include "rgrp.h" @@ -41,50 +40,6 @@ #include "util.h" #include "eaops.h" -/* - * Most fields left uninitialised to catch anybody who tries to - * use them. f_flags set to prevent file_accessed() from touching - * any other part of this. Its use is purely as a flag so that we - * know (in readpage()) whether or not do to locking. - */ -struct file gfs2_internal_file_sentinel = { - .f_flags = O_NOATIME|O_RDONLY, -}; - -static int gfs2_read_actor(read_descriptor_t *desc, struct page *page, - unsigned long offset, unsigned long size) -{ - char *kaddr; - unsigned long count = desc->count; - - if (size > count) - size = count; - - kaddr = kmap(page); - memcpy(desc->arg.data, kaddr + offset, size); - kunmap(page); - - desc->count = count - size; - desc->written += size; - desc->arg.buf += size; - return size; -} - -int gfs2_internal_read(struct gfs2_inode *ip, struct file_ra_state *ra_state, - char *buf, loff_t *pos, unsigned size) -{ - struct inode *inode = &ip->i_inode; - read_descriptor_t desc; - desc.written = 0; - desc.arg.data = buf; - desc.count = size; - desc.error = 0; - do_generic_mapping_read(inode->i_mapping, ra_state, - &gfs2_internal_file_sentinel, pos, &desc, - gfs2_read_actor); - return desc.written ? desc.written : desc.error; -} - /** * gfs2_llseek - seek to a location in a file * @file: the file -- cgit v1.2.2 From 3cc3f710ce0effe397b830826a1a081fa81f11c7 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Mon, 15 Oct 2007 15:40:33 +0100 Subject: [GFS2] Use ->page_mkwrite() for mmap() This cleans up the mmap() code path for GFS2 by implementing the page_mkwrite function for GFS2. We are thus able to use the generic filemap_fault function for our ->fault() implementation. This now means that shared writable mappings will be much more efficiently shared across the cluster if there is a reasonable proportion of read activity (the greater proportion, the better). As a side effect, it also reduces the size of the code, removes special cases from readpage and readpages, and makes the code path easier to follow. Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 131 +++++++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 121 insertions(+), 10 deletions(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index a729c86b8be1..6f3aeb059c61 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -33,7 +33,6 @@ #include "lm.h" #include "log.h" #include "meta_io.h" -#include "ops_vm.h" #include "quota.h" #include "rgrp.h" #include "trans.h" @@ -169,7 +168,7 @@ static int gfs2_get_flags(struct file *filp, u32 __user *ptr) if (put_user(fsflags, ptr)) error = -EFAULT; - gfs2_glock_dq_m(1, &gh); + gfs2_glock_dq(&gh); gfs2_holder_uninit(&gh); return error; } @@ -293,6 +292,125 @@ static long gfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return -ENOTTY; } +/** + * gfs2_allocate_page_backing - Use bmap to allocate blocks + * @page: The (locked) page to allocate backing for + * + * We try to allocate all the blocks required for the page in + * one go. This might fail for various reasons, so we keep + * trying until all the blocks to back this page are allocated. + * If some of the blocks are already allocated, thats ok too. + */ + +static int gfs2_allocate_page_backing(struct page *page) +{ + struct inode *inode = page->mapping->host; + struct buffer_head bh; + unsigned long size = PAGE_CACHE_SIZE; + u64 lblock = page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); + + do { + bh.b_state = 0; + bh.b_size = size; + gfs2_block_map(inode, lblock, 1, &bh); + if (!buffer_mapped(&bh)) + return -EIO; + size -= bh.b_size; + lblock += (bh.b_size >> inode->i_blkbits); + } while(size > 0); + return 0; +} + +/** + * gfs2_page_mkwrite - Make a shared, mmap()ed, page writable + * @vma: The virtual memory area + * @page: The page which is about to become writable + * + * When the page becomes writable, we need to ensure that we have + * blocks allocated on disk to back that page. + */ + +static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct page *page) +{ + struct inode *inode = vma->vm_file->f_path.dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_sbd *sdp = GFS2_SB(inode); + unsigned long last_index; + u64 pos = page->index << (PAGE_CACHE_SIZE - inode->i_blkbits); + unsigned int data_blocks, ind_blocks, rblocks; + int alloc_required = 0; + struct gfs2_holder gh; + struct gfs2_alloc *al; + int ret; + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_ATIME, &gh); + ret = gfs2_glock_nq_atime(&gh); + if (ret) + goto out; + + set_bit(GIF_SW_PAGED, &ip->i_flags); + gfs2_write_calc_reserv(ip, PAGE_CACHE_SIZE, &data_blocks, &ind_blocks); + ret = gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE, &alloc_required); + if (ret || !alloc_required) + goto out_unlock; + + ip->i_alloc.al_requested = 0; + al = gfs2_alloc_get(ip); + ret = gfs2_quota_lock(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + if (ret) + goto out_alloc_put; + ret = gfs2_quota_check(ip, ip->i_inode.i_uid, ip->i_inode.i_gid); + if (ret) + goto out_quota_unlock; + al->al_requested = data_blocks + ind_blocks; + ret = gfs2_inplace_reserve(ip); + if (ret) + goto out_quota_unlock; + + rblocks = RES_DINODE + ind_blocks; + if (gfs2_is_jdata(ip)) + rblocks += data_blocks ? data_blocks : 1; + if (ind_blocks || data_blocks) + rblocks += RES_STATFS + RES_QUOTA; + ret = gfs2_trans_begin(sdp, rblocks, 0); + if (ret) + goto out_trans_fail; + + lock_page(page); + ret = -EINVAL; + last_index = ip->i_inode.i_size >> PAGE_CACHE_SHIFT; + if (page->index > last_index) + goto out_unlock_page; + if (!PageUptodate(page) || page->mapping != ip->i_inode.i_mapping) + goto out_unlock_page; + if (gfs2_is_stuffed(ip)) { + ret = gfs2_unstuff_dinode(ip, page); + if (ret) + goto out_unlock_page; + } + ret = gfs2_allocate_page_backing(page); + +out_unlock_page: + unlock_page(page); + gfs2_trans_end(sdp); +out_trans_fail: + gfs2_inplace_release(ip); +out_quota_unlock: + gfs2_quota_unlock(ip); +out_alloc_put: + gfs2_alloc_put(ip); +out_unlock: + gfs2_glock_dq(&gh); +out: + gfs2_holder_uninit(&gh); + return ret; +} + +static struct vm_operations_struct gfs2_vm_ops = { + .fault = filemap_fault, + .page_mkwrite = gfs2_page_mkwrite, +}; + /** * gfs2_mmap - @@ -315,14 +433,7 @@ static int gfs2_mmap(struct file *file, struct vm_area_struct *vma) return error; } - /* This is VM_MAYWRITE instead of VM_WRITE because a call - to mprotect() can turn on VM_WRITE later. */ - - if ((vma->vm_flags & (VM_MAYSHARE | VM_MAYWRITE)) == - (VM_MAYSHARE | VM_MAYWRITE)) - vma->vm_ops = &gfs2_vm_ops_sharewrite; - else - vma->vm_ops = &gfs2_vm_ops_private; + vma->vm_ops = &gfs2_vm_ops; gfs2_glock_dq_uninit(&i_gh); -- cgit v1.2.2 From 5561093e2cac9f7d2a77e39cc689b8d2b7f9b2bc Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Wed, 17 Oct 2007 08:47:38 +0100 Subject: [GFS2] Introduce gfs2_set_aops() Just like ext3 we now have three sets of address space operations to cover the cases of writeback, ordered and journalled data writes. This means that the individual operations can now become less complicated as we are able to remove some of the tests for file data mode from the code. Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index 6f3aeb059c61..ad5daaa6babc 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -38,6 +38,7 @@ #include "trans.h" #include "util.h" #include "eaops.h" +#include "ops_address.h" /** * gfs2_llseek - seek to a location in a file @@ -245,7 +246,16 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) if (error) goto out; } - + if ((flags ^ new_flags) & GFS2_DIF_JDATA) { + if (flags & GFS2_DIF_JDATA) + gfs2_log_flush(sdp, ip->i_gl); + error = filemap_fdatawrite(inode->i_mapping); + if (error) + goto out; + error = filemap_fdatawait(inode->i_mapping); + if (error) + goto out; + } error = gfs2_trans_begin(sdp, RES_DINODE, 0); if (error) goto out; @@ -257,6 +267,7 @@ static int do_gfs2_set_flags(struct file *filp, u32 reqflags, u32 mask) gfs2_dinode_out(ip, bh->b_data); brelse(bh); gfs2_set_inode_flags(inode); + gfs2_set_aops(inode); out_trans_end: gfs2_trans_end(sdp); out: -- cgit v1.2.2 From c97bfe4351771675963e02f34d31e206fd2d7150 Mon Sep 17 00:00:00 2001 From: Wendy Cheng Date: Thu, 29 Nov 2007 17:56:51 -0500 Subject: [GFS2] Remove lock methods for lock_nolock protocol GFS2 supports two modes of locking - lock_nolock for single node filesystem and lock_dlm for cluster mode locking. The gfs2 lock methods are removed from file operation table for lock_nolock protocol. This would allow VFS to handle posix lock and flock logics just like other in-tree filesystems without duplication. Signed-off-by: S. Wendy Cheng Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 36 ++++++++++++++++++++++++------------ 1 file changed, 24 insertions(+), 12 deletions(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index ad5daaa6babc..db76ac1947e7 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -615,15 +615,6 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl) if (__mandatory_lock(&ip->i_inode)) return -ENOLCK; - if (sdp->sd_args.ar_localflocks) { - if (IS_GETLK(cmd)) { - posix_test_lock(file, fl); - return 0; - } else { - return posix_lock_file_wait(file, fl); - } - } - if (cmd == F_CANCELLK) { /* Hack: */ cmd = F_SETLK; @@ -716,9 +707,6 @@ static int gfs2_flock(struct file *file, int cmd, struct file_lock *fl) if (__mandatory_lock(&ip->i_inode)) return -ENOLCK; - if (sdp->sd_args.ar_localflocks) - return flock_lock_file_wait(file, fl); - if (fl->fl_type == F_UNLCK) { do_unflock(file, fl); return 0; @@ -755,3 +743,27 @@ const struct file_operations gfs2_dir_fops = { .flock = gfs2_flock, }; +const struct file_operations gfs2_file_fops_nolock = { + .llseek = gfs2_llseek, + .read = do_sync_read, + .aio_read = generic_file_aio_read, + .write = do_sync_write, + .aio_write = generic_file_aio_write, + .unlocked_ioctl = gfs2_ioctl, + .mmap = gfs2_mmap, + .open = gfs2_open, + .release = gfs2_close, + .fsync = gfs2_fsync, + .splice_read = generic_file_splice_read, + .splice_write = generic_file_splice_write, + .setlease = gfs2_setlease, +}; + +const struct file_operations gfs2_dir_fops_nolock = { + .readdir = gfs2_readdir, + .unlocked_ioctl = gfs2_ioctl, + .open = gfs2_open, + .release = gfs2_close, + .fsync = gfs2_fsync, +}; + -- cgit v1.2.2 From dbee2199c37336e89060fbe9abdfd1ca8454372a Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Fri, 30 Nov 2007 08:17:15 +0000 Subject: [GFS2] Remove unused variable Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 1 - 1 file changed, 1 deletion(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index db76ac1947e7..2569c13eb108 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -700,7 +700,6 @@ static void do_unflock(struct file *file, struct file_lock *fl) static int gfs2_flock(struct file *file, int cmd, struct file_lock *fl) { struct gfs2_inode *ip = GFS2_I(file->f_mapping->host); - struct gfs2_sbd *sdp = GFS2_SB(file->f_mapping->host); if (!(fl->fl_flags & FL_FLOCK)) return -ENOLCK; -- cgit v1.2.2 From e9e1ef2b6ee401d7c1e1eb38052857b4b206d172 Mon Sep 17 00:00:00 2001 From: Bob Peterson Date: Mon, 10 Dec 2007 14:13:27 -0600 Subject: [GFS2] Remove function gfs2_get_block This patch is just a cleanup. Function gfs2_get_block() just calls function gfs2_block_map reversing the last two parameters. By reversing the parameters, gfs2_block_map() may be called directly and function gfs2_get_block may be eliminated altogether. Since this function is done for every block operation, this streamlines the code and makes it a little bit more efficient. Signed-off-by: Bob Peterson Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index 2569c13eb108..597f7ff2bc11 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -323,7 +323,7 @@ static int gfs2_allocate_page_backing(struct page *page) do { bh.b_state = 0; bh.b_size = size; - gfs2_block_map(inode, lblock, 1, &bh); + gfs2_block_map(inode, lblock, &bh, 1); if (!buffer_mapped(&bh)) return -EIO; size -= bh.b_size; -- cgit v1.2.2 From 6dbd822487d0a9f14432cb4680415b80656b63a2 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Thu, 10 Jan 2008 15:18:55 +0000 Subject: [GFS2] Reduce inode size by moving i_alloc out of line It is possible to reduce the size of GFS2 inodes by taking the i_alloc structure out of the gfs2_inode. This patch allocates the i_alloc structure whenever its needed, and frees it afterward. This decreases the amount of low memory we use at the expense of requiring a memory allocation for each page or partial page that we write. A quick test with postmark shows that the overhead is not measurable and I also note that OCFS2 use the same approach. In the future I'd like to solve the problem by shrinking down the size of the members of the i_alloc structure, but for now, this reduces the immediate problem of using too much low-memory on x86 and doesn't add too much overhead. Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index 597f7ff2bc11..d7f4726ae0ce 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -364,9 +364,11 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct page *page) ret = gfs2_write_alloc_required(ip, pos, PAGE_CACHE_SIZE, &alloc_required); if (ret || !alloc_required) goto out_unlock; - - ip->i_alloc.al_requested = 0; + ret = -ENOMEM; al = gfs2_alloc_get(ip); + if (al == NULL) + goto out_unlock; + ret = gfs2_quota_lock(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); if (ret) goto out_alloc_put; -- cgit v1.2.2 From b7fe2e391ee7b711d6dfd6a694d60c4f21113cbb Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Thu, 17 Jan 2008 15:12:03 +0000 Subject: [GFS2] Fix page_mkwrite truncation race path There was a bug in the truncation/invalidation race path for ->page_mkwrite for gfs2. It ought to return 0 so that the effect is the same as if the page was truncated at any of the other points at which the page_lock is dropped. This will result in the restart of the whole page fault path. If it was due to a real truncation (as opposed to an invalidate because we let a glock go) then the ->fault path will pick that up when it gets called again. Signed-off-by: Steven Whitehouse --- fs/gfs2/ops_file.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/gfs2/ops_file.c') diff --git a/fs/gfs2/ops_file.c b/fs/gfs2/ops_file.c index d7f4726ae0ce..f4842f2548cd 100644 --- a/fs/gfs2/ops_file.c +++ b/fs/gfs2/ops_file.c @@ -394,6 +394,7 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct page *page) last_index = ip->i_inode.i_size >> PAGE_CACHE_SHIFT; if (page->index > last_index) goto out_unlock_page; + ret = 0; if (!PageUptodate(page) || page->mapping != ip->i_inode.i_mapping) goto out_unlock_page; if (gfs2_is_stuffed(ip)) { -- cgit v1.2.2