reiser4-batch-write.patch
PATCH 2/2 reiser4-batch-write.patch
This changes reiser4' write to go via generic_file_write and batch_write address space operation implementation.
fs/reiser4/plugin/file/file.c renamed from fs/reiser4/plugin/file/file.c~reiser4-batch-write +195 -217
@@ -20,6 +20,7 @@#include <linux/writeback.h> #include <linux/pagevec.h> +#include <linux/uio.h> #include <linux/syscalls.h>
@@ -905,6 +906,199 @@ commit_write_unix_file(struct file *filereturn result; } +static void drop_access (unix_file_info_t *uf_info) +{ +if (uf_info->exclusive_use) +drop_exclusive_access (uf_info); +else +drop_nonexclusive_access (uf_info); +} + +#define NEITHER_OBTAINED 0 +#define EA_OBTAINED 1 +#define NEA_OBTAINED 2 + +long batch_write_unix_file (struct file *file,const write_descriptor_t *desc, +struct pagevec *lru_pvec,struct page **cached_page, +size_t *written) +{ +int result; +reiser4_context *ctx; +struct inode *inode; +unix_file_info_t *uf_info; +ssize_t written1; +int try_free_space; +int to_write =PAGE_CACHE_SIZE *WRITE_GRANULARITY ; +size_t left; +ssize_t (*write_op)(struct file *,const char __user *,size_t , +loff_t *pos); +int ea; +loff_t new_size; +loff_t pos; +char __user *buf; + + inode = file->f_dentry->d_inode; +assert ("vs-947" , !inode_get_flag (inode,REISER4_NO_SD )); +assert ("vs-9471" , (!inode_get_flag (inode,REISER4_PART_MIXED ))); + + ctx =init_context (inode->i_sb); +if (IS_ERR (ctx)) +return PTR_ERR (ctx); + + uf_info =unix_file_inode_data (inode); + + *written =0 ; + result =0 ; + try_free_space =0 ; + left = desc->count; + pos = desc->pos; + buf = desc->buf; +/* make sure that we are to write within one i/o vector */ +assert ("" , desc->cur_iov->iov_base + desc->iov_off == buf); +assert ("" , desc->iov_off + desc->count <= desc->cur_iov->iov_len); + + ea =NEITHER_OBTAINED ; + + new_size =i_size_read (inode); +if (desc->pos + desc->count > new_size) + new_size = desc->pos + desc->count; + +while (left) { +if (left < to_write) + to_write = left; + +if (uf_info->container ==UF_CONTAINER_EMPTY ) { +get_exclusive_access (uf_info); + ea =EA_OBTAINED ; +if (uf_info->container !=UF_CONTAINER_EMPTY ) { +/* file is made not empty by another process */ +drop_exclusive_access (uf_info); + ea =NEITHER_OBTAINED ; +continue ; + } + }else if (uf_info->container ==UF_CONTAINER_UNKNOWN ) { +/* +* get exclusive access directly just to not have to +* re-obtain it if file will appear empty +*/ +get_exclusive_access (uf_info); + ea =EA_OBTAINED ; + result =find_file_state (inode, uf_info); +if (result) { +drop_exclusive_access (uf_info); + ea =NEITHER_OBTAINED ; +break ; + } + }else { +get_nonexclusive_access (uf_info); + ea =NEA_OBTAINED ; + } + +/* either EA or NEA is obtained. Choose item write method */ +if (uf_info->container ==UF_CONTAINER_EXTENTS ) { +/* file is built of extent items */ + write_op = write_extent; + }else if (uf_info->container ==UF_CONTAINER_EMPTY ) { +/* file is empty */ +if (should_have_notail (uf_info, new_size)) + write_op = write_extent; +else + write_op = write_tail; + }else { +/* file is built of tail items */ +if (should_have_notail (uf_info, new_size)) { +if (ea ==NEA_OBTAINED ) { +drop_nonexclusive_access (uf_info); +get_exclusive_access (uf_info); + ea =EA_OBTAINED ; + } +if (uf_info->container ==UF_CONTAINER_TAILS ) { +/* +* if file is being convered by another +* process - wait until it completes +*/ +while (1 ) { +if (inode_get_flag (inode,REISER4_PART_IN_CONV )) { +drop_exclusive_access (uf_info); +schedule (); +get_exclusive_access (uf_info); +continue ; + } +break ; + } +if (uf_info->container ==UF_CONTAINER_TAILS ) { + result =tail2extent (uf_info); +if (result) +break ; + } + } +drop_exclusive_access (uf_info); + ea =NEITHER_OBTAINED ; +continue ; + } + write_op = write_tail; + } + + written1 =write_op (file, desc->buf, to_write, &pos); +if (written1 == -ENOSPC && try_free_space) { +drop_access (uf_info); +txnmgr_force_commit_all (inode->i_sb,0 ); + try_free_space =0 ; +continue ; + } +if (written1 <0 ) { +drop_access (uf_info); + result = written1; +break ; + } +/* something is written. */ +if (uf_info->container ==UF_CONTAINER_EMPTY ) { +assert ("" , ea ==EA_OBTAINED ); + uf_info->container = (write_op == write_extent) ? +UF_CONTAINER_EXTENTS :UF_CONTAINER_TAILS ; + }else { +assert ("" ,ergo (uf_info->container ==UF_CONTAINER_EXTENTS , + write_op == write_extent)); +assert ("" ,ergo (uf_info->container ==UF_CONTAINER_TAILS , + write_op == write_tail)); + } +if (desc->pos + written1 > inode->i_size) +INODE_SET_FIELD (inode, i_size, desc->pos + written1); +file_update_time (file); + result =reiser4_update_sd (inode); +if (result) { +drop_access (uf_info); +context_set_commit_async (ctx); +reiser4_exit_context (ctx); +return result; + } +drop_access (uf_info); + ea =NEITHER_OBTAINED ; +txn_restart (ctx); + current->journal_info =NULL ; +/* +* tell VM how many pages were dirtied. Maybe number of pages +* which were dirty already should not be counted +*/ +balance_dirty_pages_ratelimited_nr (inode->i_mapping, + (written1 +PAGE_CACHE_SIZE -1 ) /PAGE_CACHE_SIZE ); + current->journal_info = ctx; + + left -= written1; + buf += written1; + pos += written1; + *written += written1; + } +/* +* return number of written bytes or error code if nothing is +* written. Note, that it does not work correctly in case when +* sync_unix_file returns error +*/ +context_set_commit_async (ctx); +reiser4_exit_context (ctx); +return (*written - left) ? (*written - left) : result; +} +/* * Support for "anonymous" pages and jnodes. *
@@ -1984,21 +2178,6 @@ int open_unix_file(struct inode *inode,return result; } -#define NEITHER_OBTAINED 0 -#define EA_OBTAINED 1 -#define NEA_OBTAINED 2 - -static void drop_access (unix_file_info_t *uf_info) -{ -if (uf_info->exclusive_use) -drop_exclusive_access (uf_info); -else -drop_nonexclusive_access (uf_info); -} - -#define debug_wuf (format, ...) printk("%s: %d: %s: " format "\n", \ - __FILE__, __LINE__, __FUNCTION__, ## __VA_ARGS__) -/** * write_unix_file - write of struct file_operations * @file: file to write to
@@ -2012,208 +2191,7 @@ static void drop_access(unix_file_info_tssize_t write_unix_file (struct file *file,const char __user *buf,size_t count,loff_t *pos) { -int result; -reiser4_context *ctx; -struct inode *inode; -unix_file_info_t *uf_info; -ssize_t written; -int try_free_space; -int to_write =PAGE_CACHE_SIZE *WRITE_GRANULARITY ; -size_t left; -ssize_t (*write_op)(struct file *,const char __user *,size_t , -loff_t *pos); -int ea; -loff_t new_size; - - inode = file->f_dentry->d_inode; - ctx =init_context (inode->i_sb); -if (IS_ERR (ctx)) -return PTR_ERR (ctx); - -mutex_lock (&inode->i_mutex); - -assert ("vs-947" , !inode_get_flag (inode,REISER4_NO_SD )); -assert ("vs-9471" , (!inode_get_flag (inode,REISER4_PART_MIXED ))); - -/* check amount of bytes to write and writing position */ - result =generic_write_checks (file, pos, &count,0 ); -if (result) { -mutex_unlock (&inode->i_mutex); -context_set_commit_async (ctx); -reiser4_exit_context (ctx); -return result; - } - - result =remove_suid (file->f_dentry); -if (result) { -mutex_unlock (&inode->i_mutex); -context_set_commit_async (ctx); -reiser4_exit_context (ctx); -return result; - } - - uf_info =unix_file_inode_data (inode); - - current->backing_dev_info = inode->i_mapping->backing_dev_info; - written =0 ; - try_free_space =0 ; - left = count; - ea =NEITHER_OBTAINED ; - - new_size =i_size_read (inode); -if (*pos + count > new_size) - new_size = *pos + count; - -while (left) { -if (left < to_write) - to_write = left; - -if (uf_info->container ==UF_CONTAINER_EMPTY ) { -get_exclusive_access (uf_info); - ea =EA_OBTAINED ; -if (uf_info->container !=UF_CONTAINER_EMPTY ) { -/* file is made not empty by another process */ -drop_exclusive_access (uf_info); - ea =NEITHER_OBTAINED ; -continue ; - } - }else if (uf_info->container ==UF_CONTAINER_UNKNOWN ) { -/* -* get exclusive access directly just to not have to -* re-obtain it if file will appear empty -*/ -get_exclusive_access (uf_info); - ea =EA_OBTAINED ; - result =find_file_state (inode, uf_info); -if (result) { -drop_exclusive_access (uf_info); - ea =NEITHER_OBTAINED ; -break ; - } - }else { -get_nonexclusive_access (uf_info); - ea =NEA_OBTAINED ; - } - -/* either EA or NEA is obtained. Choose item write method */ -if (uf_info->container ==UF_CONTAINER_EXTENTS ) { -/* file is built of extent items */ - write_op = write_extent; - }else if (uf_info->container ==UF_CONTAINER_EMPTY ) { -/* file is empty */ -if (should_have_notail (uf_info, new_size)) - write_op = write_extent; -else - write_op = write_tail; - }else { -/* file is built of tail items */ -if (should_have_notail (uf_info, new_size)) { -if (ea ==NEA_OBTAINED ) { -drop_nonexclusive_access (uf_info); -get_exclusive_access (uf_info); - ea =EA_OBTAINED ; - } -if (uf_info->container ==UF_CONTAINER_TAILS ) { -/* -* if file is being convered by another -* process - wait until it completes -*/ -while (1 ) { -if (inode_get_flag (inode,REISER4_PART_IN_CONV )) { -drop_exclusive_access (uf_info); -schedule (); -get_exclusive_access (uf_info); -continue ; - } -break ; - } -if (uf_info->container ==UF_CONTAINER_TAILS ) { - result =tail2extent (uf_info); -if (result) -break ; - } - } -drop_exclusive_access (uf_info); - ea =NEITHER_OBTAINED ; -continue ; - } - write_op = write_tail; - } - - written =write_op (file, buf, to_write, pos); -if (written == -ENOSPC && try_free_space) { -drop_access (uf_info); -txnmgr_force_commit_all (inode->i_sb,0 ); - try_free_space =0 ; -continue ; - } -if (written <0 ) { -drop_access (uf_info); - result = written; -break ; - } -/* something is written. */ -if (uf_info->container ==UF_CONTAINER_EMPTY ) { -assert ("" , ea ==EA_OBTAINED ); - uf_info->container = (write_op == write_extent) ? -UF_CONTAINER_EXTENTS :UF_CONTAINER_TAILS ; - }else { -assert ("" ,ergo (uf_info->container ==UF_CONTAINER_EXTENTS , - write_op == write_extent)); -assert ("" ,ergo (uf_info->container ==UF_CONTAINER_TAILS , - write_op == write_tail)); - } -if (*pos + written > inode->i_size) -INODE_SET_FIELD (inode, i_size, *pos + written); -file_update_time (file); - result =reiser4_update_sd (inode); -if (result) { -mutex_unlock (&inode->i_mutex); - current->backing_dev_info =NULL ; -drop_access (uf_info); -context_set_commit_async (ctx); -reiser4_exit_context (ctx); -return result; - } -drop_access (uf_info); - ea =NEITHER_OBTAINED ; -txn_restart (ctx); - current->journal_info =NULL ; -/* -* tell VM how many pages were dirtied. Maybe number of pages -* which were dirty already should not be counted -*/ -balance_dirty_pages_ratelimited_nr (inode->i_mapping, - (written +PAGE_CACHE_SIZE -1 ) /PAGE_CACHE_SIZE ); - current->journal_info = ctx; - - left -= written; - buf += written; - *pos += written; - } - -mutex_unlock (&inode->i_mutex); - -if (result ==0 && ((file->f_flags &O_SYNC ) ||IS_SYNC (inode))) { -txn_restart_current (); -grab_space_enable (); - result =sync_unix_file (file, file->f_dentry, -0 /* data and stat data */ ); -if (result) -warning ("reiser4-7" ,"failed to sync file %llu" , - (unsigned long long )get_inode_oid (inode)); - } - - current->backing_dev_info =NULL ; - -reiser4_exit_context (ctx); - -/* -* return number of written bytes or error code if nothing is -* written. Note, that it does not work correctly in case when -* sync_unix_file returns error -*/ -return (count - left) ? (count - left) : result; +return generic_file_write (file, buf, count, pos); }/**
fs/reiser4/plugin/object.c renamed from fs/reiser4/plugin/object.c~reiser4-batch-write +1 -0
@@ -117,6 +117,7 @@ file_plugin file_plugins[LAST_FILE_PLUGI .readpages = reiser4_readpages, .prepare_write = prepare_write_unix_file, .commit_write = commit_write_unix_file, + .batch_write = batch_write_unix_file, .bmap = bmap_unix_file, .invalidatepage = reiser4_invalidatepage, .releasepage = reiser4_releasepage
fs/reiser4/plugin/file/file.h renamed from fs/reiser4/plugin/file/file.h~reiser4-batch-write +3 -0
@@ -34,6 +34,9 @@ int prepare_write_unix_file(struct file unsigned to);int commit_write_unix_file (struct file *,struct page *,unsigned from,unsigned to); +long batch_write_unix_file (struct file *,const write_descriptor_t *, +struct pagevec *,struct page **, +size_t *written);sector_t bmap_unix_file (struct address_space *,sector_t lblock);/* file plugin operations */
_