pnfs block layout updates
Subsystems (2)
cover pnfs block layout updates
This patch set applies to the pnfs-block branch of your git repo. Together with the previous 34 patch set this adds the pnfs block layout client to pnfs. I am soliciting comments on this set. The final patch set will differ in three small ways:
1. Removal of some debug printks
2. Removal of the DEVONLY patch
3. Fix up some signed-offs
I believe this set addresses the comments since the previous patch set, including rewriting the I/O path to eliminate write_begin/end and defer layoutget to flush time.
This set is also available on the for-benny branch ofgit://citi.umich.edu/projects/linux-pnfs-blk.git .
Jim Rees (2):
get rid of deprecated xdr macros
reindent
Peng Tao (4):
SQUASHME: pnfs-block: Remove write_begin/end hooks
SQUASHME: pnfs-block: skip sectors already initialized
SQUASHME: pnfs: teach layoutcommit handle multiple segments
pnfs-block: mark IO error with NFS_LAYOUT_{RW|RO}_FAILED
fs/nfs/blocklayout/blocklayout.c | 716 ++++++++++++++++------------------- fs/nfs/blocklayout/blocklayout.h | 51 +--- fs/nfs/blocklayout/blocklayoutdev.c | 30 ++- fs/nfs/blocklayout/extents.c | 115 +++--- fs/nfs/file.c | 26 +-- fs/nfs/nfs4filelayout.c | 2 +- fs/nfs/nfs4proc.c | 8 +- fs/nfs/pnfs.c | 101 ++---- fs/nfs/pnfs.h | 119 +------ fs/nfs/write.c | 12 +- include/linux/nfs_fs.h | 3 +- include/linux/nfs_xdr.h | 2 +- 12 files changed, 459 insertions(+), 726 deletions(-)
Signature · 1 line
1.7.4.1
PATCH 1/6 SQUASHME: pnfs-block: Remove write_begin/end hooks
From: Peng Tao <[email protected]>
Do not call pnfs_update_layout at write_begin, and rewrite bl_write_pagelist to handle EOF there.
Signed-off-by Peng Tao <[email protected]>
--- fs/nfs/blocklayout/blocklayout.c | 652 ++++++++++++++++++-------------------- fs/nfs/file.c | 26 +-- fs/nfs/pnfs.c | 41 --- fs/nfs/pnfs.h | 115 ------- fs/nfs/write.c | 12 +- include/linux/nfs_fs.h | 3 +- 6 files changed, 321 insertions(+), 528 deletions(-)
fs/nfs/blocklayout/blocklayout.c +313 -339
@@ -32,8 +32,8 @@#include <linux/module.h> #include <linux/init.h> -#include <linux/buffer_head.h> /* various write calls */ -#include <linux/bio.h> /* struct bio */ +#include <linux/buffer_head.h> /* various write calls */ +#include <linux/bio.h> /* struct bio */ #include <linux/vmalloc.h> #include "blocklayout.h"
@@ -75,16 +75,11 @@ static int is_hole(struct pnfs_block_extent *be, sector_t isect)*/ static int is_writable (struct pnfs_block_extent *be,sector_t isect) { -if (be->be_state ==PNFS_BLOCK_READWRITE_DATA ) -return 1 ; -else if (be->be_state !=PNFS_BLOCK_INVALID_DATA ) -return 0 ; -else -return is_sector_initialized (be->be_inval, isect); +return (be->be_state ==PNFS_BLOCK_READWRITE_DATA || + be->be_state ==PNFS_BLOCK_INVALID_DATA ); } -static int -dont_like_caller (struct nfs_page *req) +static int dont_like_caller (struct nfs_page *req) {if (atomic_read (&req->wb_complete)) {/* Called by _multi */
@@ -109,7 +104,7 @@ static inline struct parallel_io *alloc_parallel(void *data) {struct parallel_io *rv; - rv =kmalloc (sizeof (*rv),GFP_KERNEL ); + rv =kmalloc (sizeof (*rv),GFP_KERNEL );if (rv) { rv->data = data;kref_init (&rv->refcnt);
@@ -136,21 +131,19 @@ static inline void put_parallel(struct parallel_io *p)kref_put (&p->refcnt, destroy_parallel); } -static struct bio * -bl_submit_bio (int rw,struct bio *bio) +static struct bio *bl_submit_bio (int rw,struct bio *bio) {if (bio) {get_parallel (bio->bi_private);dprintk ("%s submitting %s bio %u@%llu\n" , __func__, rw ==READ ?"read" :"write" , - bio->bi_size, (u64 )bio->bi_sector); + bio->bi_size, (u64 ) bio->bi_sector);submit_bio (rw, bio); }return NULL ; } -static inline void -bl_done_with_rpage (struct page *page,const int ok) +static inline void bl_done_with_rpage (struct page *page,const int ok) {if (ok) {ClearPagePnfsErr (page);
@@ -191,8 +184,7 @@ static void bl_read_cleanup(struct work_struct *work)pnfs_ld_read_done (rdata); } -static void -bl_end_par_io_read (void *data) +static void bl_end_par_io_read (void *data) {struct nfs_read_data *rdata = data;
@@ -208,8 +200,7 @@ static void bl_rpc_do_nothing(struct rpc_task *task, void *calldata)return ; } -static enum pnfs_try_status -bl_read_pagelist (struct nfs_read_data *rdata) +static enum pnfs_try_status bl_read_pagelist (struct nfs_read_data *rdata) {int i, hole;struct bio *bio =NULL ;
@@ -222,7 +213,7 @@ bl_read_pagelist(struct nfs_read_data *rdata)int pg_index = rdata->args.pgbase >>PAGE_CACHE_SHIFT ;dprintk ("%s enter nr_pages %u offset %lld count %Zd\n" , __func__, - rdata->npages, f_offset, count); + rdata->npages, f_offset, count);if (dont_like_caller (rdata->req)) {dprintk ("%s dont_like_caller failed\n" , __func__);
@@ -260,10 +251,10 @@ bl_read_pagelist(struct nfs_read_data *rdata)break ; } extent_length = be->be_length - - (isect - be->be_f_offset); + (isect - be->be_f_offset);if (cow_read) {sector_t cow_length = cow_read->be_length - - (isect - cow_read->be_f_offset); + (isect - cow_read->be_f_offset); extent_length =min (extent_length, cow_length); } }
@@ -282,15 +273,17 @@ bl_read_pagelist(struct nfs_read_data *rdata) be_read = (hole && cow_read) ? cow_read : be;for (;;) {if (!bio) { - bio =bio_alloc (GFP_NOIO , rdata->npages - i); + bio = +bio_alloc (GFP_NOIO , + rdata->npages - i);if (!bio) {/* Error out this page */ bl_done_with_rpage (pages[i],0 );break ; } bio->bi_sector = isect - - be_read->be_f_offset + - be_read->be_v_offset; + be_read->be_f_offset + + be_read->be_v_offset; bio->bi_bdev = be_read->be_mdev; bio->bi_end_io = bl_end_io_read; bio->bi_private = par;
@@ -315,7 +308,7 @@ bl_read_pagelist(struct nfs_read_data *rdata)put_parallel (par);return PNFS_ATTEMPTED ; - use_mds: +use_mds:dprintk ("Giving up and using normal NFS\n" );return PNFS_NOT_ATTEMPTED ; }
@@ -335,18 +328,16 @@ static void mark_extents_written(struct pnfs_block_layout *bl,while (isect < end) {sector_t len; be =find_get_extent (bl, isect,NULL ); -BUG_ON (!be);/* FIXME */ +BUG_ON (!be);/* FIXME */ len =min (end, be->be_f_offset + be->be_length) - isect;if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) -mark_for_commit (be, isect, len);/* What if fails? */ +mark_for_commit (be, isect, len);/* What if fails? */ isect += len;put_extent (be); } } -/* STUB - this needs thought */ -static inline void -bl_done_with_wpage (struct page *page,const int ok) +static inline void bl_done_with_wpage (struct page *page,const int ok) {if (!ok) {SetPageError (page);
@@ -360,15 +351,37 @@ bl_done_with_wpage(struct page *page, const int ok)spin_unlock (&inode->i_lock); } } -/* end_page_writeback called in rpc_release. Should be done here. */ } -/* This is basically copied from mpage_end_io_read */ +static void bl_end_io_write_zero (struct bio *bio,int err) +{ +struct parallel_io *par = bio->bi_private; +const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags); +struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; + +do { +struct page *page = bvec->bv_page; + +if (--bvec >= bio->bi_io_vec) +prefetchw (&bvec->bv_page->flags); +bl_done_with_wpage (page, uptodate); +/* This is the zeroing page we added */ +end_page_writeback (page); +page_cache_release (page); + }while (bvec >= bio->bi_io_vec); +if (!uptodate && !wdata->pnfs_error) + wdata->pnfs_error = -EIO ; +bio_put (bio); +put_parallel (par); +} +static void bl_end_io_write (struct bio *bio,int err) { -void *data = bio->bi_private; +struct parallel_io *par = bio->bi_private;const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags);struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; do { struct page *page = bvec->bv_page;
@@ -377,8 +390,10 @@ static void bl_end_io_write(struct bio *bio, int err)prefetchw (&bvec->bv_page->flags);bl_done_with_wpage (page, uptodate); }while (bvec >= bio->bi_io_vec); +if (!uptodate && !wdata->pnfs_error) + wdata->pnfs_error = -EIO ;bio_put (bio); -put_parallel (data); +put_parallel (par); }/* Function scheduled for call during bl_end_par_io_write,
@@ -391,7 +406,7 @@ static void bl_write_cleanup(struct work_struct *work)dprintk ("%s enter\n" , __func__); task = container_of(work, struct rpc_task, u.tk_work); wdata = container_of(task, struct nfs_write_data, task); -if (!wdata->task.tk_status) { +if (!wdata->pnfs_error) {/* Marks for LAYOUTCOMMIT */ /* BUG - this should be called after each bio, not after * all finish, unless have some way of storing success/failure
@@ -403,8 +418,7 @@ static void bl_write_cleanup(struct work_struct *work) }/* Called when last of bios associated with a bl_write_pagelist call finishes */ -static void -bl_end_par_io_write (void *data) +static void bl_end_par_io_write (void *data) {struct nfs_write_data *wdata = data;
@@ -415,19 +429,118 @@ bl_end_par_io_write(void *data)schedule_work (&wdata->task.u.tk_work); } +/* STUB - mark intersection of layout and page as bad, so is not +* used again. +*/ +static void mark_bad_read (void ) +{ +return ; +} + +/* Copied from buffer.c */ +static void __end_buffer_read_notouch (struct buffer_head *bh,int uptodate) +{ +if (uptodate) { +set_buffer_uptodate (bh); + }else { +/* This happens, due to failed READA attempts. */ +clear_buffer_uptodate (bh); + } +unlock_buffer (bh); +} + +/* Copied from buffer.c */ +static void end_buffer_read_nobh (struct buffer_head *bh,int uptodate) +{ +__end_buffer_read_notouch (bh, uptodate); +} + +/* +* map_block: map a requested I/0 block (isect) into an offset in the LVM +* meta block_device +*/ +static void +map_block (sector_t isect,struct pnfs_block_extent *be,struct buffer_head *bh) +{ +dprintk ("%s enter be=%p\n" , __func__, be); + +set_buffer_mapped (bh); + bh->b_bdev = be->be_mdev; + bh->b_blocknr = (isect - be->be_f_offset + be->be_v_offset) >> + (be->be_mdev->bd_inode->i_blkbits -9 ); + +dprintk ("%s isect %ld, bh->b_blocknr %ld, using bsize %Zd\n" , + __func__, (long )isect, (long )bh->b_blocknr, bh->b_size); +return ; +} + +/* Given an unmapped page, zero it (or read in page for COW), +*/ +static int +init_page_for_write (struct page *page,struct pnfs_block_extent *cow_read) +{ +struct buffer_head *bh =NULL ; +int ret =0 ; +sector_t isect; + +dprintk ("%s enter, %p\n" , __func__, page); +BUG_ON (PageUptodate (page)); +/* not COW, zero and return */ +if (!cow_read) { +zero_user_segment (page,0 ,PAGE_SIZE ); +SetPageUptodate (page); + goto cleanup; + } + +/* COW, readin page */ + bh =alloc_page_buffers (page,PAGE_CACHE_SIZE ,0 ); +if (!bh) { + ret = -ENOMEM ; + goto cleanup; + } + + isect = (sector_t ) page->index << (PAGE_CACHE_SHIFT -9 ); +map_block (isect, cow_read, bh); +lock_buffer (bh); + bh->b_end_io = end_buffer_read_nobh; +submit_bh (READ , bh); +dprintk ("%s: Waiting for buffer read\n" , __func__); +wait_on_buffer (bh); +if (!buffer_uptodate (bh)) { + ret = -EIO ; + goto cleanup; + } +SetPageUptodate (page); +SetPageMappedToDisk (page); + +cleanup: +put_extent (cow_read); +if (bh) +free_buffer_head (bh); +if (ret) { +/* Need to mark layout with bad read...should now +* just use nfs4 for reads and writes. +*/ +mark_bad_read (); + } +return ret; +} +static enum pnfs_try_status -bl_write_pagelist (struct nfs_write_data *wdata, -int sync) +bl_write_pagelist (struct nfs_write_data *wdata,int sync) { -int i; +int i, ret, npg_zero, pg_index, last =0 ;struct bio *bio =NULL ; -struct pnfs_block_extent *be =NULL ; -sector_t isect, extent_length =0 ; +struct pnfs_block_extent *be =NULL , *cow_read =NULL ; +sector_t isect, last_isect =0 , extent_length =0 ;struct parallel_io *par;loff_t offset = wdata->args.offset;size_t count = wdata->args.count;struct page **pages = wdata->args.pages; -int pg_index = wdata->args.pgbase >>PAGE_CACHE_SHIFT ; +struct page *page; +pgoff_t index; +int npg_per_block = +NFS_SERVER (wdata->inode)->pnfs_blksize >>PAGE_CACHE_SHIFT ;dprintk ("%s enter, %Zu@%lld\n" , __func__, count, offset);if (!wdata->lseg) {
@@ -439,11 +552,8 @@ bl_write_pagelist(struct nfs_write_data *wdata,return PNFS_NOT_ATTEMPTED ; }/* At this point, wdata->pages is a (sequential) list of nfs_pages. -* We want to write each, and if there is an error remove it from -* list and call -* nfs_retry_request(req) to have it redone using nfs. -* QUEST? Do as block or per req? Think have to do per block -* as part of end_bio +* We want to write each, and if there is an error set pnfs_error +* to have it redone using nfs. */ par =alloc_parallel (wdata);if (!par)
@@ -454,7 +564,106 @@ bl_write_pagelist(struct nfs_write_data *wdata,/* At this point, have to be more careful with error handling */ isect = (sector_t) ((offset & (long )PAGE_CACHE_MASK ) >>9 ); -for (i = pg_index; i < wdata->npages ; i++) { + be =find_get_extent (BLK_LSEG2EXT (wdata->lseg), isect, &cow_read); +if (!be || !is_writable (be, isect)) { +dprintk ("%s no matching extents!\n" , __func__); + wdata->pnfs_error = -EINVAL ; + goto out; + } + +/* First page inside INVALID extent */ +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + npg_zero = (offset >>PAGE_CACHE_SHIFT ) % npg_per_block; + isect = (sector_t) (((offset - npg_zero *PAGE_CACHE_SIZE ) & + (long )PAGE_CACHE_MASK ) >>9 ); + extent_length = be->be_length - (isect - be->be_f_offset); + +fill_invalid_ext: +dprintk ("%s need to zero %d pages\n" , __func__, npg_zero); +while (npg_zero) { +/* page ref released in bl_end_io_write_zero */ + index = isect >> (PAGE_CACHE_SHIFT -9 ); +dprintk ("%s zero %dth page: index %lu isect %lu\n" , + __func__, npg_zero, index, isect); + page = +find_or_create_page (wdata->inode->i_mapping, index, +GFP_NOFS ); +if (!page) { +dprintk ("%s oom\n" , __func__); + wdata->pnfs_error = -ENOMEM ; + goto out; + } + +/* PageDirty: Other will write this out +* PageWriteback: Other is writing this out +* PageUptodate && sector_initialized: already written out +*/ +if (PageDirty (page) ||PageWriteback (page) || + (PageUptodate (page) && +is_sector_initialized (be->be_inval, isect))) { +dprintk + ("page %lu is uptodate %d dirty %d writeback %d\n" , + page->index,PageUptodate (page), +PageDirty (page),PageWriteback (page)); +unlock_page (page); +page_cache_release (page); + npg_zero--; + isect +=PAGE_CACHE_SIZE >>9 ; + extent_length -=PAGE_CACHE_SIZE >>9 ; +continue ; + } +if (!PageUptodate (page) && cow_read) { +/* New page, readin or zero it */ +init_page_for_write (page, cow_read); + } +set_page_writeback (page); +unlock_page (page); + + ret =mark_initialized_sectors (be->be_inval, isect, +PAGE_CACHE_SECTORS , +NULL ); +if (unlikely (ret)) { +dprintk ("%s mark_initialized_sectors fail %d\n" , + __func__, ret); +page_cache_release (page); + wdata->pnfs_error = ret; + goto out; + } +for (;;) { +if (!bio) { + bio =bio_alloc (GFP_NOIO , npg_zero--); +if (unlikely (!bio)) { +end_page_writeback (page); +page_cache_release (page); + wdata->pnfs_error = -ENOMEM ; + goto out; + } + bio->bi_sector = + isect - be->be_f_offset + + be->be_v_offset; + bio->bi_bdev = be->be_mdev; + bio->bi_end_io = bl_end_io_write_zero; + bio->bi_private = par; + } +if (bio_add_page (bio, page,PAGE_CACHE_SIZE ,0 )) +break ; + bio =bl_submit_bio (WRITE , bio); + } +/* FIXME: This should be done in bi_end_io */ +mark_extents_written (BLK_LSEG2EXT (wdata->lseg), + page->index <<PAGE_CACHE_SHIFT , +PAGE_CACHE_SIZE ); + isect +=PAGE_CACHE_SIZE >>9 ; + extent_length -=PAGE_CACHE_SIZE >>9 ; + } +if (last) + goto write_done; + } + bio =bl_submit_bio (WRITE , bio); + +/* Middle pages */ + pg_index = wdata->args.pgbase >>PAGE_CACHE_SHIFT ; +for (i = pg_index; i < wdata->npages; i++) {if (!extent_length) {/* We've used up the previous extent */ put_extent (be);
@@ -463,24 +672,32 @@ bl_write_pagelist(struct nfs_write_data *wdata, be =find_get_extent (BLK_LSEG2EXT (wdata->lseg), isect,NULL );if (!be || !is_writable (be, isect)) { -/* FIXME */ -bl_done_with_wpage (pages[i],0 ); -break ; + wdata->pnfs_error = -EINVAL ; + goto out; } extent_length = be->be_length - - (isect - be->be_f_offset); + (isect - be->be_f_offset); + } +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + ret =mark_initialized_sectors (be->be_inval, isect, +PAGE_CACHE_SECTORS , +NULL ); +if (unlikely (ret)) { +dprintk ("%s mark_initialized_sectors fail %d\n" , + __func__, ret); + wdata->pnfs_error = ret; + goto out; + } }for (;;) {if (!bio) { bio =bio_alloc (GFP_NOIO , wdata->npages - i);if (!bio) { -/* Error out this page */ -/* FIXME */ -bl_done_with_wpage (pages[i],0 ); -break ; + wdata->pnfs_error = -ENOMEM ; + goto out; } bio->bi_sector = isect - be->be_f_offset + - be->be_v_offset; + be->be_v_offset; bio->bi_bdev = be->be_mdev; bio->bi_end_io = bl_end_io_write; bio->bi_private = par;
@@ -490,11 +707,27 @@ bl_write_pagelist(struct nfs_write_data *wdata, bio =bl_submit_bio (WRITE , bio); } isect +=PAGE_CACHE_SIZE >>9 ; + last_isect = isect; extent_length -=PAGE_CACHE_SIZE >>9 ; } - wdata->res.count = (isect <<9 ) - (offset); -if (count < wdata->res.count) + +/* Last page inside INVALID extent */ +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + bio =bl_submit_bio (WRITE , bio); + npg_zero = npg_per_block - + (last_isect >> (PAGE_CACHE_SHIFT -9 )) % npg_per_block; +if (npg_zero < npg_per_block) { + last =1 ; + goto fill_invalid_ext; + } + } + +write_done: + wdata->res.count = (last_isect <<9 ) - (offset); +if (count < wdata->res.count) { wdata->res.count = count; + } +out:put_extent (be);bl_submit_bio (WRITE , bio);put_parallel (par);
@@ -521,8 +754,7 @@ release_extents(struct pnfs_block_layout *bl, struct pnfs_layout_range *range)spin_unlock (&bl->bl_ext_lock); } -static void -release_inval_marks (struct pnfs_inval_markings *marks) +static void release_inval_marks (struct pnfs_inval_markings *marks) {struct pnfs_inval_tracking *pos, *temp;
@@ -615,7 +847,8 @@ bl_cleanup_layoutcommit(struct pnfs_layout_hdr *lo, struct nfs4_layoutcommit_data *lcdata) {dprintk ("%s enter\n" , __func__); -clean_pnfs_block_layoutupdate (BLK_LO2EXT (lo), &lcdata->args, lcdata->res.status); +clean_pnfs_block_layoutupdate (BLK_LO2EXT (lo), &lcdata->args, + lcdata->res.status); }static void free_blk_mountid (struct block_mount_id *mid)
@@ -625,8 +858,7 @@ static void free_blk_mountid(struct block_mount_id *mid)spin_lock (&mid->bm_lock);while (!list_empty (&mid->bm_devlist)) { dev =list_first_entry (&mid->bm_devlist, - struct pnfs_block_dev, - bm_node); + struct pnfs_block_dev, bm_node);list_del (&dev->bm_node);free_block_dev (dev); }
@@ -638,10 +870,11 @@ static void free_blk_mountid(struct block_mount_id *mid)/* This is mostly copied from the filelayout's get_device_info function. * It seems much of this should be at the generic pnfs level. */ -static struct pnfs_block_dev * -nfs4_blk_get_deviceinfo (struct nfs_server *server,const struct nfs_fh *fh, -struct nfs4_deviceid *d_id, -struct list_head *sdlist) +static struct pnfs_block_dev *nfs4_blk_get_deviceinfo (struct nfs_server *server, +const struct nfs_fh *fh, +struct nfs4_deviceid + *d_id, +struct list_head *sdlist) {struct pnfs_device *dev;struct pnfs_block_dev *rv =NULL ;
@@ -695,7 +928,7 @@ nfs4_blk_get_deviceinfo(struct nfs_server *server, const struct nfs_fh *fh, goto out_free; rv =nfs4_blk_decode_device (server, dev, sdlist); - out_free: +out_free:if (dev->area !=NULL )vunmap (dev->area);for (i =0 ; i < max_pages; i++)
@@ -748,8 +981,8 @@ bl_set_layoutdriver(struct nfs_server *server, const struct nfs_fh *fh)*/ for (i =0 ; i < dlist->num_devs; i++) { bdev =nfs4_blk_get_deviceinfo (server, fh, - &dlist->dev_id[i], - &block_disklist); + &dlist->dev_id[i], + &block_disklist);if (!bdev) { status = -ENODEV ; goto out_error;
@@ -762,18 +995,17 @@ bl_set_layoutdriver(struct nfs_server *server, const struct nfs_fh *fh)dprintk ("%s SUCCESS\n" , __func__); server->pnfs_ld_data = b_mt_id; - out_return: +out_return:kfree (dlist);return status; - out_error: +out_error:free_blk_mountid (b_mt_id);kfree (mtype); goto out_return; } -static int -bl_clear_layoutdriver (struct nfs_server *server) +static int bl_clear_layoutdriver (struct nfs_server *server) {struct block_mount_id *b_mt_id = server->pnfs_ld_data;
@@ -783,268 +1015,14 @@ bl_clear_layoutdriver(struct nfs_server *server)return 0 ; } -/* STUB - mark intersection of layout and page as bad, so is not -* used again. -*/ -static void mark_bad_read (void ) -{ -return ; -} - -/* Copied from buffer.c */ -static void __end_buffer_read_notouch (struct buffer_head *bh,int uptodate) -{ -if (uptodate) { -set_buffer_uptodate (bh); - }else { -/* This happens, due to failed READA attempts. */ -clear_buffer_uptodate (bh); - } -unlock_buffer (bh); -} - -/* Copied from buffer.c */ -static void end_buffer_read_nobh (struct buffer_head *bh,int uptodate) -{ -__end_buffer_read_notouch (bh, uptodate); -} - -/* -* map_block: map a requested I/0 block (isect) into an offset in the LVM -* meta block_device -*/ -static void -map_block (sector_t isect,struct pnfs_block_extent *be,struct buffer_head *bh) -{ -dprintk ("%s enter be=%p\n" , __func__, be); - -set_buffer_mapped (bh); - bh->b_bdev = be->be_mdev; - bh->b_blocknr = (isect - be->be_f_offset + be->be_v_offset) >> - (be->be_mdev->bd_inode->i_blkbits -9 ); - -dprintk ("%s isect %ld, bh->b_blocknr %ld, using bsize %Zd\n" , - __func__, (long )isect, - (long )bh->b_blocknr, - bh->b_size); -return ; -} - -/* Given an unmapped page, zero it (or read in page for COW), -* and set appropriate flags/markings, but it is safe to not initialize -* the range given in [from, to). -*/ -/* This is loosely based on nobh_write_begin */ -static int -init_page_for_write (struct pnfs_block_layout *bl,struct page *page, -unsigned from,unsigned to,sector_t **pages_to_mark) -{ -struct buffer_head *bh; -int inval, ret = -EIO ; -struct pnfs_block_extent *be =NULL , *cow_read =NULL ; -sector_t isect; - -dprintk ("%s enter, %p\n" , __func__, page); - bh =alloc_page_buffers (page,PAGE_CACHE_SIZE ,0 ); -if (!bh) { - ret = -ENOMEM ; - goto cleanup; - } - - isect = (sector_t )page->index << (PAGE_CACHE_SHIFT -9 ); - be =find_get_extent (bl, isect, &cow_read); -if (!be) - goto cleanup; - inval =is_hole (be, isect); -dprintk ("%s inval=%i, from=%u, to=%u\n" , __func__, inval, from, to); -if (inval) { -if (be->be_state ==PNFS_BLOCK_NONE_DATA ) { -dprintk ("%s PANIC - got NONE_DATA extent %p\n" , - __func__, be); - goto cleanup; - } -map_block (isect, be, bh); -unmap_underlying_metadata (bh->b_bdev, bh->b_blocknr); - } -if (PageUptodate (page)) { -/* Do nothing */ - }else if (inval & !cow_read) { -zero_user_segments (page,0 , from, to,PAGE_CACHE_SIZE ); - }else if (0 < from ||PAGE_CACHE_SIZE > to) { -struct pnfs_block_extent *read_extent; - - read_extent = (inval && cow_read) ? cow_read : be; -map_block (isect, read_extent, bh); -lock_buffer (bh); - bh->b_end_io = end_buffer_read_nobh; -submit_bh (READ , bh); -dprintk ("%s: Waiting for buffer read\n" , __func__); -/* XXX Don't really want to hold layout lock here */ -wait_on_buffer (bh); -if (!buffer_uptodate (bh)) - goto cleanup; - } -if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { -/* There is a BUG here if is a short copy after write_begin, -* but I think this is a generic fs bug. The problem is that -* we have marked the page as initialized, but it is possible -* that the section not copied may never get copied. -*/ - ret =mark_initialized_sectors (be->be_inval, isect, -PAGE_CACHE_SECTORS , - pages_to_mark); -/* Want to preallocate mem so above can't fail */ -if (ret) - goto cleanup; - } -SetPageMappedToDisk (page); - ret =0 ; - -cleanup: -free_buffer_head (bh); -put_extent (be); -put_extent (cow_read); -if (ret) { -/* Need to mark layout with bad read...should now -* just use nfs4 for reads and writes. -*/ -mark_bad_read (); - } -return ret; -} - -static int -bl_write_begin (struct pnfs_layout_segment *lseg,struct page *page,loff_t pos, -unsigned count,struct pnfs_fsdata *fsdata) -{ -unsigned from, to; -int ret; -sector_t *pages_to_mark =NULL ; -struct pnfs_block_layout *bl =BLK_LSEG2EXT (lseg); - -dprintk ("%s enter, %u@%lld\n" , __func__, count, pos); -print_page (page); -/* The following code assumes blocksize >= PAGE_CACHE_SIZE */ -if (bl->bl_blocksize < (PAGE_CACHE_SIZE >>9 )) { -dprintk ("%s Can't handle blocksize %llu\n" , __func__, - (u64 )bl->bl_blocksize); -put_lseg (fsdata->lseg); - fsdata->lseg =NULL ; -return 0 ; - } -if (PageMappedToDisk (page)) { -/* Basically, this is a flag that says we have -* successfully called write_begin already on this page. -*/ -/* NOTE - there are cache consistency issues here. -* For example, what if the layout is recalled, then regained? -* If the file is closed and reopened, will the page flags -* be reset? If not, we'll have to use layout info instead of -* the page flag. -*/ -return 0 ; - } - from = pos & (PAGE_CACHE_SIZE -1 ); - to = from + count; - ret =init_page_for_write (bl, page, from, to, &pages_to_mark); -if (ret) { -dprintk ("%s init page failed with %i" , __func__, ret); -/* Revert back to plain NFS and just continue on with -* write. This assumes there is no request attached, which -* should be true if we get here. -*/ -BUG_ON (PagePrivate (page)); -put_lseg (fsdata->lseg); - fsdata->lseg =NULL ; -kfree (pages_to_mark); - ret =0 ; - }else { - fsdata->private = pages_to_mark; - } -return ret; -} - -/* CAREFUL - what happens if copied < count??? */ -static int -bl_write_end (struct inode *inode,struct page *page,loff_t pos, -unsigned count,unsigned copied,struct pnfs_layout_segment *lseg) -{ -dprintk ("%s enter, %u@%lld, lseg=%p\n" , __func__, count, pos, lseg); -print_page (page); -if (lseg) -SetPageUptodate (page); -return 0 ; -} - -/* Return any memory allocated to fsdata->private, and take advantage -* of no page locks to mark pages noted in write_begin as needing -* initialization. -*/ -static void -bl_write_end_cleanup (struct file *filp,struct pnfs_fsdata *fsdata) -{ -struct page *page; -pgoff_t index; -sector_t *pos; -struct address_space *mapping = filp->f_mapping; -struct pnfs_fsdata *fake_data; -struct pnfs_layout_segment *lseg; - -if (!fsdata) -return ; - lseg = fsdata->lseg; -if (!lseg) -return ; - pos = fsdata->private; -if (!pos) -return ; -dprintk ("%s enter with pos=%llu\n" , __func__, (u64)(*pos)); -for (; *pos != ~0 ; pos++) { - index = *pos >> (PAGE_CACHE_SHIFT -9 ); -/* XXX How do we properly deal with failures here??? */ - page =grab_cache_page_write_begin (mapping, index,0 ); -if (!page) { -printk (KERN_ERR "%s BUG BUG BUG NoMem\n" , __func__); -continue ; - } -dprintk ("%s: Examining block page\n" , __func__); -print_page (page); -if (!PageMappedToDisk (page)) { -/* XXX How do we properly deal with failures here??? */ -dprintk ("%s Marking block page\n" , __func__); -init_page_for_write (BLK_LSEG2EXT (fsdata->lseg), page, -PAGE_CACHE_SIZE ,PAGE_CACHE_SIZE , -NULL ); -print_page (page); - fake_data =kzalloc (sizeof (*fake_data),GFP_KERNEL ); -if (!fake_data) { -printk (KERN_ERR "%s BUG BUG BUG NoMem\n" , - __func__); -unlock_page (page); -continue ; - } -get_lseg (lseg); - fake_data->lseg = lseg; - fake_data->bypass_eof =1 ; - mapping->a_ops->write_end (filp, mapping, - index <<PAGE_CACHE_SHIFT , -PAGE_CACHE_SIZE , -PAGE_CACHE_SIZE , - page, fake_data); -/* Note fake_data is freed by nfs_write_end */ - }else -unlock_page (page); - } -kfree (fsdata->private); - fsdata->private =NULL ; -} -static const struct nfs_pageio_ops bl_pg_read_ops = { + .pg_init = pnfs_generic_pg_init_read, .pg_test = pnfs_generic_pg_test, .pg_doio = nfs_generic_pg_readpages, };static const struct nfs_pageio_ops bl_pg_write_ops = { + .pg_init = pnfs_generic_pg_init_write, .pg_test = pnfs_generic_pg_test, .pg_doio = nfs_generic_pg_writepages, };
@@ -1054,9 +1032,6 @@ static struct pnfs_layoutdriver_type blocklayout_type = { .name = "LAYOUT_BLOCK_VOLUME", .read_pagelist = bl_read_pagelist, .write_pagelist = bl_write_pagelist, - .write_begin = bl_write_begin, - .write_end = bl_write_end, - .write_end_cleanup = bl_write_end_cleanup, .alloc_layout_hdr = bl_alloc_layout_hdr, .free_layout_hdr = bl_free_layout_hdr, .alloc_lseg = bl_alloc_lseg,
@@ -1083,8 +1058,7 @@ static int __init nfs4blocklayout_init(void) static void __exit nfs4blocklayout_exit(void) { -dprintk ("%s: NFSv4 Block Layout Driver Unregistering...\n" , - __func__); +dprintk ("%s: NFSv4 Block Layout Driver Unregistering...\n" , __func__);pnfs_unregister_layoutdriver (&blocklayout_type);bl_pipe_exit ();
fs/nfs/file.c +3 -23
@@ -384,15 +384,12 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,pgoff_t index = pos >>PAGE_CACHE_SHIFT ;struct page *page;int once_thru =0 ; -struct pnfs_layout_segment *lseg;dfprintk (PAGECACHE ,"NFS: write_begin(%s/%s(%ld), %u@%lld)\n" , file->f_path.dentry->d_parent->d_name.name, file->f_path.dentry->d_name.name, mapping->host->i_ino, len, (long long ) pos); - lseg =pnfs_update_layout (mapping->host, -nfs_file_open_context (file), - pos, len,IOMODE_RW ,GFP_NOFS ); + start:/* * Prevent starvation issues if someone is doing a consistency
@@ -412,9 +409,6 @@ start:if (ret) {unlock_page (page);page_cache_release (page); - *pagep =NULL ; - *fsdata =NULL ; - goto out; }else if (!once_thru &&nfs_want_read_modify_write (file, page, pos, len)) { once_thru =1 ;
@@ -423,12 +417,6 @@ start:if (!ret) goto start; } - ret =pnfs_write_begin (file, page, pos, len, lseg, fsdata); - out: -if (ret) { -put_lseg (lseg); - *fsdata =NULL ; - }return ret; }
@@ -438,7 +426,6 @@ static int nfs_write_end(struct file *file, struct address_space *mapping, {unsigned offset = pos & (PAGE_CACHE_SIZE -1 );int status; -struct pnfs_layout_segment *lseg; dfprintk(PAGECACHE, "NFS: write_end(%s/%s(%ld), %u@%lld)\n", file->f_path.dentry->d_parent->d_name.name,
@@ -465,17 +452,10 @@ static int nfs_write_end(struct file *file, struct address_space *mapping,zero_user_segment (page, pglen,PAGE_CACHE_SIZE ); } - lseg =nfs4_pull_lseg_from_fsdata (file, fsdata); - status =pnfs_write_end (file, page, pos, len, copied, lseg); -if (status) - goto out; - status =nfs_updatepage (file, page, offset, copied, lseg, fsdata); + status =nfs_updatepage (file, page, offset, copied); -out:unlock_page (page);page_cache_release (page); -pnfs_write_end_cleanup (file, fsdata); -put_lseg (lseg);if (status <0 )return status;
@@ -597,7 +577,7 @@ static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) ret =VM_FAULT_LOCKED ;if (nfs_flush_incompatible (filp, page) ==0 && -nfs_updatepage (filp, page,0 , pagelen,NULL ,NULL ) ==0 ) +nfs_updatepage (filp, page,0 , pagelen) ==0 ) goto out; ret =VM_FAULT_SIGBUS ;
fs/nfs/pnfs.c +0 -41
53 lines not on this page · raw message
fs/nfs/pnfs.h +0 -115
175 lines not on this page · raw message
fs/nfs/write.c +4 -8
@@ -673,9 +673,7 @@ out: }static int nfs_writepage_setup (struct nfs_open_context *ctx,struct page *page, -unsigned int offset,unsigned int count, -struct pnfs_layout_segment *lseg,void *fsdata) - +unsigned int offset,unsigned int count) {struct nfs_page *req;
@@ -683,8 +681,7 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page,if (IS_ERR (req))return PTR_ERR (req);/* Update file length */ -if (pnfs_grow_ok (lseg, fsdata)) -nfs_grow_file (page, offset, count); +nfs_grow_file (page, offset, count);nfs_mark_uptodate (page, req->wb_pgbase, req->wb_bytes);nfs_mark_request_dirty (req);nfs_clear_page_tag_locked (req);
@@ -737,8 +734,7 @@ static int nfs_write_pageuptodate(struct page *page, struct inode *inode)* things with a page scheduled for an RPC call (e.g. invalidate it). */ int nfs_updatepage (struct file *file,struct page *page, -unsigned int offset,unsigned int count, -struct pnfs_layout_segment *lseg,void *fsdata) +unsigned int offset,unsigned int count) {struct nfs_open_context *ctx =nfs_file_open_context (file);struct inode *inode = page->mapping->host;
@@ -763,7 +759,7 @@ int nfs_updatepage(struct file *file, struct page *page, offset =0 ; } - status =nfs_writepage_setup (ctx, page, offset, count, lseg, fsdata); + status =nfs_writepage_setup (ctx, page, offset, count);if (status <0 )nfs_set_pageerror (page);
include/linux/nfs_fs.h +1 -2
@@ -510,8 +510,7 @@ extern int nfs_congestion_kb;extern int nfs_writepage (struct page *page,struct writeback_control *wbc);extern int nfs_writepages (struct address_space *,struct writeback_control *);extern int nfs_flush_incompatible (struct file *file,struct page *page); -extern int nfs_updatepage (struct file *,struct page *,unsigned int , -unsigned int ,struct pnfs_layout_segment *,void *); +extern int nfs_updatepage (struct file *,struct page *,unsigned int ,unsigned int );extern void nfs_writeback_done (struct rpc_task *,struct nfs_write_data *);/*
Signature · 1 line
1.7.4.1
Re: [PATCH 1/6] SQUASHME: pnfs-block: Remove write_begin/end hooks
Hi, sorry for my late response.
I like the direction this is going!
However I have a few comments inline below.
In general, combining cosmetic changes or cleanups with the real functionality makes the patch really hard to review.
Please separate the patches so that each one is boiled down to the minimum to make it self-sufficient, even if that ends up with more patches to review, (see Documentation/SubmittingPatches)
more inline...
Quoted 63 lines · Jim Rees
From: Peng Tao <[email protected]>
Do not call
pnfs_update_layoutatwrite_begin, and rewritebl_write_pagelistto handle EOF there.Signed-off-by Peng Tao <[email protected]>--- fs/nfs/blocklayout/blocklayout.c | 652 ++++++++++++++++++-------------------- fs/nfs/file.c | 26 +-- fs/nfs/pnfs.c | 41 --- fs/nfs/pnfs.h | 115 ------- fs/nfs/write.c | 12 +- include/linux/nfs_fs.h | 3 +- 6 files changed, 321 insertions(+), 528 deletions(-)fs/nfs/blocklayout/blocklayout.c +6
-11@@ -32,8 +32,8 @@#include <linux/module.h> #include <linux/init.h> -#include <linux/buffer_head.h> /* various write calls */ -#include <linux/bio.h> /* struct bio */ +#include <linux/buffer_head.h> /* various write calls */ +#include <linux/bio.h> /* struct bio */ #include <linux/vmalloc.h> #include "blocklayout.h" @@ -75,16 +75,11 @@ static int is_hole(struct pnfs_block_extent *be, sector_t isect)*/ static int is_writable (struct pnfs_block_extent *be,sector_t isect) { -if (be->be_state ==PNFS_BLOCK_READWRITE_DATA ) -return 1 ; -else if (be->be_state !=PNFS_BLOCK_INVALID_DATA ) -return 0 ; -else -return is_sector_initialized (be->be_inval, isect); +return (be->be_state ==PNFS_BLOCK_READWRITE_DATA || + be->be_state ==PNFS_BLOCK_INVALID_DATA ); } -static int -dont_like_caller (struct nfs_page *req) +static int dont_like_caller (struct nfs_page *req) {if (atomic_read (&req->wb_complete)) {/* Called by _multi */ @@ -109,7 +104,7 @@ static inline struct parallel_io *alloc_parallel(void *data) {struct parallel_io *rv; - rv =kmalloc (sizeof (*rv),GFP_KERNEL ); + rv =kmalloc (sizeof (*rv),GFP_KERNEL );if (rv) { rv->data = data;kref_init (&rv->refcnt);@@ -136,21 +131,19 @@ static inline void put_parallel(struct parallel_io *p)kref_put (&p->refcnt, destroy_parallel); }
fs/nfs/blocklayout/blocklayout.c static inline void put_parallel(struct parallel_io *p) · line 140
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +2
-3-static struct bio * -bl_submit_bio (int rw,struct bio *bio) +static struct bio *bl_submit_bio (int rw,struct bio *bio) {if (bio) {get_parallel (bio->bi_private);dprintk ("%s submitting %s bio %u@%llu\n" , __func__, rw ==READ ?"read" :"write" , - bio->bi_size, (u64 )bio->bi_sector); + bio->bi_size, (u64 ) bio->bi_sector);
Do you even support or want to support 32 bit bi_sectors? If not, you could just add BUILD_BUG_ON(sizeof(sector_t) != 8); Otherwise, typecasting to (unsigned long long) would be more portable.
fs/nfs/blocklayout/blocklayout.c static inline void put_parallel(struct parallel_io *p) · line 143
Quoted 3 lines
fs/nfs/blocklayout/blocklayout.c +0
-0submit_bio (rw, bio); }return NULL ;
What's the point of always returning NULL?
Quoted 177 lines
fs/nfs/blocklayout/blocklayout.c +45
-25} -static inline void -bl_done_with_rpage (struct page *page,const int ok) +static inline void bl_done_with_rpage (struct page *page,const int ok) {if (ok) {ClearPagePnfsErr (page);@@ -191,8 +184,7 @@ static void bl_read_cleanup(struct work_struct *work)pnfs_ld_read_done (rdata); } -static void -bl_end_par_io_read (void *data) +static void bl_end_par_io_read (void *data) {struct nfs_read_data *rdata = data;@@ -208,8 +200,7 @@ static void bl_rpc_do_nothing(struct rpc_task *task, void *calldata)return ; } -static enum pnfs_try_status -bl_read_pagelist (struct nfs_read_data *rdata) +static enum pnfs_try_status bl_read_pagelist (struct nfs_read_data *rdata) {int i, hole;struct bio *bio =NULL ;@@ -222,7 +213,7 @@ bl_read_pagelist(struct nfs_read_data *rdata)int pg_index = rdata->args.pgbase >>PAGE_CACHE_SHIFT ;dprintk ("%s enter nr_pages %u offset %lld count %Zd\n" , __func__, - rdata->npages, f_offset, count); + rdata->npages, f_offset, count);if (dont_like_caller (rdata->req)) {dprintk ("%s dont_like_caller failed\n" , __func__);@@ -260,10 +251,10 @@ bl_read_pagelist(struct nfs_read_data *rdata)break ; } extent_length = be->be_length - - (isect - be->be_f_offset); + (isect - be->be_f_offset);if (cow_read) {sector_t cow_length = cow_read->be_length - - (isect - cow_read->be_f_offset); + (isect - cow_read->be_f_offset); extent_length =min (extent_length, cow_length); } }@@ -282,15 +273,17 @@ bl_read_pagelist(struct nfs_read_data *rdata) be_read = (hole && cow_read) ? cow_read : be;for (;;) {if (!bio) { - bio =bio_alloc (GFP_NOIO , rdata->npages - i); + bio = +bio_alloc (GFP_NOIO , + rdata->npages - i);if (!bio) {/* Error out this page */ bl_done_with_rpage (pages[i],0 );break ; } bio->bi_sector = isect - - be_read->be_f_offset + - be_read->be_v_offset; + be_read->be_f_offset + + be_read->be_v_offset; bio->bi_bdev = be_read->be_mdev; bio->bi_end_io = bl_end_io_read; bio->bi_private = par;@@ -315,7 +308,7 @@ bl_read_pagelist(struct nfs_read_data *rdata)put_parallel (par);return PNFS_ATTEMPTED ; - use_mds: +use_mds:dprintk ("Giving up and using normal NFS\n" );return PNFS_NOT_ATTEMPTED ; }@@ -335,18 +328,16 @@ static void mark_extents_written(struct pnfs_block_layout *bl,while (isect < end) {sector_t len; be =find_get_extent (bl, isect,NULL ); -BUG_ON (!be);/* FIXME */ +BUG_ON (!be);/* FIXME */ len =min (end, be->be_f_offset + be->be_length) - isect;if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) -mark_for_commit (be, isect, len);/* What if fails? */ +mark_for_commit (be, isect, len);/* What if fails? */ isect += len;put_extent (be); } } -/* STUB - this needs thought */ -static inline void -bl_done_with_wpage (struct page *page,const int ok) +static inline void bl_done_with_wpage (struct page *page,const int ok) {if (!ok) {SetPageError (page);@@ -360,15 +351,37 @@ bl_done_with_wpage(struct page *page, const int ok)spin_unlock (&inode->i_lock); } } -/* end_page_writeback called in rpc_release. Should be done here. */ } -/* This is basically copied from mpage_end_io_read */ +static void bl_end_io_write_zero (struct bio *bio,int err) +{ +struct parallel_io *par = bio->bi_private; +const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags); +struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; + +do { +struct page *page = bvec->bv_page; + +if (--bvec >= bio->bi_io_vec) +prefetchw (&bvec->bv_page->flags); +bl_done_with_wpage (page, uptodate); +/* This is the zeroing page we added */ +end_page_writeback (page); +page_cache_release (page); + }while (bvec >= bio->bi_io_vec); +if (!uptodate && !wdata->pnfs_error) + wdata->pnfs_error = -EIO ; +bio_put (bio); +put_parallel (par); +} +static void bl_end_io_write (struct bio *bio,int err) { -void *data = bio->bi_private; +struct parallel_io *par = bio->bi_private;const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags);struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; do { struct page *page = bvec->bv_page;@@ -377,8 +390,10 @@ static void bl_end_io_write(struct bio *bio, int err)prefetchw (&bvec->bv_page->flags);bl_done_with_wpage (page, uptodate); }while (bvec >= bio->bi_io_vec); +if (!uptodate && !wdata->pnfs_error) + wdata->pnfs_error = -EIO ;bio_put (bio); -put_parallel (data); +put_parallel (par); }/* Function scheduled for call during bl_end_par_io_write, @@ -391,7 +406,7 @@ static void bl_write_cleanup(struct work_struct *work)dprintk ("%s enter\n" , __func__); task = container_of(work, struct rpc_task, u.tk_work); wdata = container_of(task, struct nfs_write_data, task); -if (!wdata->task.tk_status) { +if (!wdata->pnfs_error) {/* Marks for LAYOUTCOMMIT */ /* BUG - this should be called after each bio, not after * all finish, unless have some way of storing success/failure @@ -403,8 +418,7 @@ static void bl_write_cleanup(struct work_struct *work) }/* Called when last of bios associated with a bl_write_pagelist call finishes */ -static void -bl_end_par_io_write (void *data) +static void bl_end_par_io_write (void *data) {struct nfs_write_data *wdata = data;@@ -415,19 +429,118 @@ bl_end_par_io_write(void *data)schedule_work (&wdata->task.u.tk_work); }
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 440
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +9
-0+/* STUB - mark intersection of layout and page as bad, so is not +* used again. +*/ +static void mark_bad_read (void ) +{ +return ; +} + +/* Copied from buffer.c */
NAK. If this code needs to be shared internally you should make it public, declare it properly in linux/buffer_head.h, and EXPORT_SYMBOL_GPL it
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 452
Quoted 12 lines
fs/nfs/blocklayout/blocklayout.c +12
-0+static void __end_buffer_read_notouch (struct buffer_head *bh,int uptodate) +{ +if (uptodate) { +set_buffer_uptodate (bh); + }else { +/* This happens, due to failed READA attempts. */ +clear_buffer_uptodate (bh); + } +unlock_buffer (bh); +} + +/* Copied from buffer.c */
This should be turned into a static inline helper in buffer_head.h since it just an alias for __end_buffer_read_notouch, but anyway, why do _you_ need it here?
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 460
Quoted 8 lines
fs/nfs/blocklayout/blocklayout.c +8
-0+static void end_buffer_read_nobh (struct buffer_head *bh,int uptodate) +{ +__end_buffer_read_notouch (bh, uptodate); +} + +/* +* map_block: map a requested I/0 block (isect) into an offset in the LVM +* meta block_device
What's a "meta block_device"?
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 463
Quoted 3 lines
fs/nfs/blocklayout/blocklayout.c +3
-0+*/ +static void +map_block (sector_t isect,struct pnfs_block_extent *be,struct buffer_head *bh)
I'd put the output parameter, bh, first, followed by the input parameters.
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 470
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+{ +dprintk ("%s enter be=%p\n" , __func__, be); +bl_end_io_write_zero +set_buffer_mapped (bh ); + bh->b_bdev = be->be_mdev; + bh->b_blocknr = (isect - be->be_f_offset + be->be_v_offset) >> + (be->be_mdev->bd_inode->i_blkbits -9 );
Please use SECTOR_SHIFT from <linux/device-mapper.h> rather than '9'.
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 478
Quoted 8 lines
fs/nfs/blocklayout/blocklayout.c +8
-0+ +dprintk ("%s isect %ld, bh->b_blocknr %ld, using bsize %Zd\n" , + __func__, (long )isect, (long )bh->b_blocknr, bh->b_size); +return ; +} + +/* Given an unmapped page, zero it (or read in page for COW), +*/
Why is the (or read ...) in parenthesis?
It's of exactly the same importance as the zeroing part of this function. Also, single line comments can have the closing "*/" on the same line.
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 488
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+static int +init_page_for_write (struct page *page,struct pnfs_block_extent *cow_read) +{ +struct buffer_head *bh =NULL ; +int ret =0 ; +sector_t isect; + +dprintk ("%s enter, %p\n" , __func__, page); +BUG_ON (PageUptodate (page)); +/* not COW, zero and return */
This comment is not needed. The code is straight forward as it is.
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 495
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+if (!cow_read) { +zero_user_segment (page,0 ,PAGE_SIZE ); +SetPageUptodate (page); + goto cleanup; + } + +/* COW, readin page */
This one can be removed too, unless you have something non-trivial to document.
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 502
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+ bh =alloc_page_buffers (page,PAGE_CACHE_SIZE ,0 ); +if (!bh) { + ret = -ENOMEM ; + goto cleanup; + } + + isect = (sector_t ) page->index << (PAGE_CACHE_SHIFT -9 );
ditto SECTOR_SHIFT
I see you use (PAGE_CACHE_SHIFT - 9) in other places too so even better,
#define PAGE_CACHE_SECTOR_SHIFT (PAGE_CACHE_SHIFT - SECTOR_SHIFT)
Also, no space after the type cast (not sure why this is not in Documentation/CodingStyle)
Quoted 71 lines
fs/nfs/blocklayout/blocklayout.c +38
-12+map_block (isect, cow_read, bh); +lock_buffer (bh); + bh->b_end_io = end_buffer_read_nobh; +submit_bh (READ , bh); +dprintk ("%s: Waiting for buffer read\n" , __func__); +wait_on_buffer (bh); +if (!buffer_uptodate (bh)) { + ret = -EIO ; + goto cleanup; + } +SetPageUptodate (page); +SetPageMappedToDisk (page); + +cleanup: +put_extent (cow_read); +if (bh) +free_buffer_head (bh); +if (ret) { +/* Need to mark layout with bad read...should now +* just use nfs4 for reads and writes. +*/ +mark_bad_read (); + } +return ret; +} +static enum pnfs_try_status -bl_write_pagelist (struct nfs_write_data *wdata, -int sync) +bl_write_pagelist (struct nfs_write_data *wdata,int sync) { -int i; +int i, ret, npg_zero, pg_index, last =0 ;struct bio *bio =NULL ; -struct pnfs_block_extent *be =NULL ; -sector_t isect, extent_length =0 ; +struct pnfs_block_extent *be =NULL , *cow_read =NULL ; +sector_t isect, last_isect =0 , extent_length =0 ;struct parallel_io *par;loff_t offset = wdata->args.offset;size_t count = wdata->args.count;struct page **pages = wdata->args.pages; -int pg_index = wdata->args.pgbase >>PAGE_CACHE_SHIFT ; +struct page *page; +pgoff_t index; +int npg_per_block = +NFS_SERVER (wdata->inode)->pnfs_blksize >>PAGE_CACHE_SHIFT ;dprintk ("%s enter, %Zu@%lld\n" , __func__, count, offset);if (!wdata->lseg) {@@ -439,11 +552,8 @@ bl_write_pagelist(struct nfs_write_data *wdata,return PNFS_NOT_ATTEMPTED ; }/* At this point, wdata->pages is a (sequential) list of nfs_pages. -* We want to write each, and if there is an error remove it from -* list and call -* nfs_retry_request(req) to have it redone using nfs. -* QUEST? Do as block or per req? Think have to do per block -* as part of end_bio +* We want to write each, and if there is an error set pnfs_error +* to have it redone using nfs. */ par =alloc_parallel (wdata);if (!par)@@ -454,7 +564,106 @@ bl_write_pagelist(struct nfs_write_data *wdata,/* At this point, have to be more careful with error handling */ isect = (sector_t) ((offset & (long )PAGE_CACHE_MASK ) >>9 ); -for (i = pg_index; i < wdata->npages ; i++) { + be =find_get_extent (BLK_LSEG2EXT (wdata->lseg), isect, &cow_read); +if (!be || !is_writable (be, isect)) {
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 578
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+dprintk ("%s no matching extents!\n" , __func__); + wdata->pnfs_error = -EINVAL ; + goto out; + } + +/* First page inside INVALID extent */ +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + npg_zero = (offset >>PAGE_CACHE_SHIFT ) % npg_per_block; + isect = (sector_t) (((offset - npg_zero *PAGE_CACHE_SIZE ) & + (long )PAGE_CACHE_MASK ) >>9 );
ditto SECTOR_SHIFT.
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 583
Quoted 5 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+ extent_length = be->be_length - (isect - be->be_f_offset); + +fill_invalid_ext: +dprintk ("%s need to zero %d pages\n" , __func__, npg_zero); +while (npg_zero) {
a for loop seems more appropriate here
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 585
Quoted 2 lines
fs/nfs/blocklayout/blocklayout.c +2
-0+/* page ref released in bl_end_io_write_zero */ + index = isect >> (PAGE_CACHE_SHIFT -9 );
ditto PAGE_CACHE_SECTOR_SHIFT
Quoted 17 lines
fs/nfs/blocklayout/blocklayout.c +17
-0+dprintk ("%s zero %dth page: index %lu isect %lu\n" , + __func__, npg_zero, index, isect); + page = +find_or_create_page (wdata->inode->i_mapping, index, +GFP_NOFS ); +if (!page) { +dprintk ("%s oom\n" , __func__); + wdata->pnfs_error = -ENOMEM ; + goto out; + } + +/* PageDirty: Other will write this out +* PageWriteback: Other is writing this out +* PageUptodate && sector_initialized: already written out +*/ + if (PageDirty(page) || PageWriteback(page) || + (PageUptodate(page) &&
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 612
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+ is_sector_initialized(be->be_inval, isect))) { +dprintk + ("page %lu is uptodate %d dirty %d writeback %d\n" , + page->index,PageUptodate (page), +PageDirty (page),PageWriteback (page)); +unlock_page (page); +page_cache_release (page); + npg_zero--; + isect +=PAGE_CACHE_SIZE >>9 ; + extent_length -=PAGE_CACHE_SIZE >>9 ;
#define SECTORS_PER_PAGE (PAGE_CACHE_SIZE >> SECTOR_SHIFT)
(should this be done in the iteration leg of the for statement or at a label just before the end of the loop?)
Quoted 5 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+continue ; + } +if (!PageUptodate (page) && cow_read) { +/* New page, readin or zero it */ +init_page_for_write (page, cow_read);
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 627
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+ } +set_page_writeback (page); +unlock_page (page); + + ret =mark_initialized_sectors (be->be_inval, isect, +PAGE_CACHE_SECTORS , +NULL ); +if (unlikely (ret)) { +dprintk ("%s mark_initialized_sectors fail %d\n" , + __func__, ret);
what about end_page_writeback(page)?
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 632
Quoted 5 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+page_cache_release (page); + wdata->pnfs_error = ret; + goto out; + } +for (;;) {
Doing an infinite loop here seems like a bad idea.
My understanding is that eventually you want to submit bios for all npg_zero pages.
Each time you go through this section you add one page, possibly allocating a bio if this is the first time, or if you previously submitted the bio (and set bio to NULL, which shouldn't be a side effect of bl_submit_bio, but rather it should be clearly spelled out by its callers)
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 634
Quoted 2 lines
fs/nfs/blocklayout/blocklayout.c +2
-0+if (!bio) { + bio =bio_alloc (GFP_NOIO , npg_zero--);
The post-decrement of npg_zero is non intuitive since bio_alloc allocates npg_zero iovecs and the decrement logically belongs to the bio_add_page below.
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 639
Quoted 5 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+if (unlikely (!bio)) { +end_page_writeback (page); +page_cache_release (page); + wdata->pnfs_error = -ENOMEM ; + goto out;
this calls for a common error path handling path.
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 650
Quoted 11 lines
fs/nfs/blocklayout/blocklayout.c +11
-0+ } + bio->bi_sector = + isect - be->be_f_offset + + be->be_v_offset; + bio->bi_bdev = be->be_mdev; + bio->bi_end_io = bl_end_io_write_zero; + bio->bi_private = par; + } +if (bio_add_page (bio, page,PAGE_CACHE_SIZE ,0 )) +break ; + bio =bl_submit_bio (WRITE , bio);
so if bio_add_page fails, you want to submit the current bio, allocate a new one and bio_add_page to the new one. but if it fails the second time you absolutely don't want to submit the one you've just allocated but rather you want to pass the error on.
I think it'd be better to move the contents of this loop into a couple functions that will do the additive iteration. something along these lines:
static struct bio *bl_alloc_init_bio(int npg, struct pnfs_block_extent *be, void (*end_io)(struct bio *, int err), struct parallel_io *par)
{
struct bio *bio;
bio = bio_alloc(GFP_NOIO, npg);
if (!bio)
return NULL;
bio->bi_sector = isect - be->be_f_offset + be->be_v_offset;
bio->bi_bdev = be->be_mdev;
bio->bi_end_io = end_io;
bio->bi_private = par;
return bio;
}static struct bio *bl_add_page_to_bio(struct bio *bio, int npg, struct page *page, struct pnfs_block_extent *be, void (*end_io)(struct bio *, int err), struct parallel_io *par)
{
int retried = 0;retry:
if (!bio) {
bio = bl_alloc_init_bio(npg, be, end_io, par);
if (!bio)
return PTR_ERR(-ENOMEM);
}
if (!bl_add_page(bio, page, PAGE_CACHE_SIZE, 0)) {
if (!retried++) {
bl_submit_bio(bio);
bio = NULL;
goto retry;
}
bio_free(bio);
return ERR_PTR(-EIO);
}
return bio;
}and the caller should decrement npg_zero or handle the error with respect to the page and wdata->pnfs_error.
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 657
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+ } +/* FIXME: This should be done in bi_end_io */ +mark_extents_written (BLK_LSEG2EXT (wdata->lseg), + page->index <<PAGE_CACHE_SHIFT , +PAGE_CACHE_SIZE ); + isect +=PAGE_CACHE_SIZE >>9 ; + extent_length -=PAGE_CACHE_SIZE >>9 ;
common iteration path, see comment above
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 664
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+ } +if (last) + goto write_done; + } + bio =bl_submit_bio (WRITE , bio); + +/* Middle pages */
this whole function is way too long.
Is it possible to refactor out the different parts of it into their own static functions?
Quoted 21 lines
fs/nfs/blocklayout/blocklayout.c +8
-4+ pg_index = wdata->args.pgbase >>PAGE_CACHE_SHIFT ; +for (i = pg_index; i < wdata->npages; i++) {if (!extent_length) {/* We've used up the previous extent */ put_extent (be);@@ -463,24 +672,32 @@ bl_write_pagelist(struct nfs_write_data *wdata, be =find_get_extent (BLK_LSEG2EXT (wdata->lseg), isect,NULL );if (!be || !is_writable (be, isect)) { -/* FIXME */ -bl_done_with_wpage (pages[i],0 ); -break ; + wdata->pnfs_error = -EINVAL ; + goto out; } extent_length = be->be_length - - (isect - be->be_f_offset); + (isect - be->be_f_offset); + } +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + ret = mark_initialized_sectors(be->be_inval, isect,
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 692
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +8
-0+PAGE_CACHE_SECTORS , + NULL); +if (unlikely (ret)) { +dprintk ("%s mark_initialized_sectors fail %d\n" , + __func__, ret); + wdata->pnfs_error = ret; + goto out; + } }for (;;) {
ditto.
can we use the same code structure as I suggested above also here and in bl_read_pagelist?
Quoted 811 lines
fs/nfs/blocklayout/blocklayout.c +39
-284if (!bio) { bio =bio_alloc (GFP_NOIO , wdata->npages - i);if (!bio) { -/* Error out this page */ -/* FIXME */ -bl_done_with_wpage (pages[i],0 ); -break ; + wdata->pnfs_error = -ENOMEM ; + goto out; } bio->bi_sector = isect - be->be_f_offset + - be->be_v_offset; + be->be_v_offset; bio->bi_bdev = be->be_mdev; bio->bi_end_io = bl_end_io_write; bio->bi_private = par;@@ -490,11 +707,27 @@ bl_write_pagelist(struct nfs_write_data *wdata, bio =bl_submit_bio (WRITE , bio); } isect +=PAGE_CACHE_SIZE >>9 ; + last_isect = isect; extent_length -=PAGE_CACHE_SIZE >>9 ; } - wdata->res.count = (isect <<9 ) - (offset); -if (count < wdata->res.count) + +/* Last page inside INVALID extent */ +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + bio =bl_submit_bio (WRITE , bio); + npg_zero = npg_per_block - + (last_isect >> (PAGE_CACHE_SHIFT -9 )) % npg_per_block; +if (npg_zero < npg_per_block) { + last =1 ; + goto fill_invalid_ext; + } + } + +write_done: + wdata->res.count = (last_isect <<9 ) - (offset); +if (count < wdata->res.count) { wdata->res.count = count; + } +out:put_extent (be);bl_submit_bio (WRITE , bio);put_parallel (par);@@ -521,8 +754,7 @@ release_extents(struct pnfs_block_layout *bl, struct pnfs_layout_range *range)spin_unlock (&bl->bl_ext_lock); } -static void -release_inval_marks (struct pnfs_inval_markings *marks) +static void release_inval_marks (struct pnfs_inval_markings *marks) {struct pnfs_inval_tracking *pos, *temp;@@ -615,7 +847,8 @@ bl_cleanup_layoutcommit(struct pnfs_layout_hdr *lo, struct nfs4_layoutcommit_data *lcdata) {dprintk ("%s enter\n" , __func__); -clean_pnfs_block_layoutupdate (BLK_LO2EXT (lo), &lcdata->args, lcdata->res.status); +clean_pnfs_block_layoutupdate (BLK_LO2EXT (lo), &lcdata->args, + lcdata->res.status); }static void free_blk_mountid (struct block_mount_id *mid)@@ -625,8 +858,7 @@ static void free_blk_mountid(struct block_mount_id *mid)spin_lock (&mid->bm_lock);while (!list_empty (&mid->bm_devlist)) { dev =list_first_entry (&mid->bm_devlist, - struct pnfs_block_dev, - bm_node); + struct pnfs_block_dev, bm_node);list_del (&dev->bm_node);free_block_dev (dev); }@@ -638,10 +870,11 @@ static void free_blk_mountid(struct block_mount_id *mid)/* This is mostly copied from the filelayout's get_device_info function. * It seems much of this should be at the generic pnfs level. */ -static struct pnfs_block_dev * -nfs4_blk_get_deviceinfo (struct nfs_server *server,const struct nfs_fh *fh, -struct nfs4_deviceid *d_id, -struct list_head *sdlist) +static struct pnfs_block_dev *nfs4_blk_get_deviceinfo (struct nfs_server *server, +const struct nfs_fh *fh, +struct nfs4_deviceid + *d_id, +struct list_head *sdlist) {struct pnfs_device *dev;struct pnfs_block_dev *rv =NULL ;@@ -695,7 +928,7 @@ nfs4_blk_get_deviceinfo(struct nfs_server *server, const struct nfs_fh *fh, goto out_free; rv =nfs4_blk_decode_device (server, dev, sdlist); - out_free: +out_free:if (dev->area !=NULL )vunmap (dev->area);for (i =0 ; i < max_pages; i++)@@ -748,8 +981,8 @@ bl_set_layoutdriver(struct nfs_server *server, const struct nfs_fh *fh)*/ for (i =0 ; i < dlist->num_devs; i++) { bdev =nfs4_blk_get_deviceinfo (server, fh, - &dlist->dev_id[i], - &block_disklist); + &dlist->dev_id[i], + &block_disklist);if (!bdev) { status = -ENODEV ; goto out_error;@@ -762,18 +995,17 @@ bl_set_layoutdriver(struct nfs_server *server, const struct nfs_fh *fh)dprintk ("%s SUCCESS\n" , __func__); server->pnfs_ld_data = b_mt_id; - out_return: +out_return:kfree (dlist);return status; - out_error: +out_error:free_blk_mountid (b_mt_id);kfree (mtype); goto out_return; } -static int -bl_clear_layoutdriver (struct nfs_server *server) +static int bl_clear_layoutdriver (struct nfs_server *server) {struct block_mount_id *b_mt_id = server->pnfs_ld_data;@@ -783,268 +1015,14 @@ bl_clear_layoutdriver(struct nfs_server *server)return 0 ; } -/* STUB - mark intersection of layout and page as bad, so is not -* used again. -*/ -static void mark_bad_read (void ) -{ -return ; -} - -/* Copied from buffer.c */ -static void __end_buffer_read_notouch (struct buffer_head *bh,int uptodate) -{ -if (uptodate) { -set_buffer_uptodate (bh); - }else { -/* This happens, due to failed READA attempts. */ -clear_buffer_uptodate (bh); - } -unlock_buffer (bh); -} - -/* Copied from buffer.c */ -static void end_buffer_read_nobh (struct buffer_head *bh,int uptodate) -{ -__end_buffer_read_notouch (bh, uptodate); -} - -/* -* map_block: map a requested I/0 block (isect) into an offset in the LVM -* meta block_device -*/ -static void -map_block (sector_t isect,struct pnfs_block_extent *be,struct buffer_head *bh) -{ -dprintk ("%s enter be=%p\n" , __func__, be); - -set_buffer_mapped (bh); - bh->b_bdev = be->be_mdev; - bh->b_blocknr = (isect - be->be_f_offset + be->be_v_offset) >> - (be->be_mdev->bd_inode->i_blkbits -9 ); - -dprintk ("%s isect %ld, bh->b_blocknr %ld, using bsize %Zd\n" , - __func__, (long )isect, - (long )bh->b_blocknr, - bh->b_size); -return ; -} - -/* Given an unmapped page, zero it (or read in page for COW), -* and set appropriate flags/markings, but it is safe to not initialize -* the range given in [from, to). -*/ -/* This is loosely based on nobh_write_begin */ -static int -init_page_for_write (struct pnfs_block_layout *bl,struct page *page, -unsigned from,unsigned to,sector_t **pages_to_mark) -{ -struct buffer_head *bh; -int inval, ret = -EIO ; -struct pnfs_block_extent *be =NULL , *cow_read =NULL ; -sector_t isect; - -dprintk ("%s enter, %p\n" , __func__, page); - bh =alloc_page_buffers (page,PAGE_CACHE_SIZE ,0 ); -if (!bh) { - ret = -ENOMEM ; - goto cleanup; - } - - isect = (sector_t )page->index << (PAGE_CACHE_SHIFT -9 ); - be =find_get_extent (bl, isect, &cow_read); -if (!be) - goto cleanup; - inval =is_hole (be, isect); -dprintk ("%s inval=%i, from=%u, to=%u\n" , __func__, inval, from, to); -if (inval) { -if (be->be_state ==PNFS_BLOCK_NONE_DATA ) { -dprintk ("%s PANIC - got NONE_DATA extent %p\n" , - __func__, be); - goto cleanup; - } -map_block (isect, be, bh); -unmap_underlying_metadata (bh->b_bdev, bh->b_blocknr); - } -if (PageUptodate (page)) { -/* Do nothing */ - }else if (inval & !cow_read) { -zero_user_segments (page,0 , from, to,PAGE_CACHE_SIZE ); - }else if (0 < from ||PAGE_CACHE_SIZE > to) { -struct pnfs_block_extent *read_extent; - - read_extent = (inval && cow_read) ? cow_read : be; -map_block (isect, read_extent, bh); -lock_buffer (bh); - bh->b_end_io = end_buffer_read_nobh; -submit_bh (READ , bh); -dprintk ("%s: Waiting for buffer read\n" , __func__); -/* XXX Don't really want to hold layout lock here */ -wait_on_buffer (bh); -if (!buffer_uptodate (bh)) - goto cleanup; - } -if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { -/* There is a BUG here if is a short copy after write_begin, -* but I think this is a generic fs bug. The problem is that -* we have marked the page as initialized, but it is possible -* that the section not copied may never get copied. -*/ - ret =mark_initialized_sectors (be->be_inval, isect, -PAGE_CACHE_SECTORS , - pages_to_mark); -/* Want to preallocate mem so above can't fail */ -if (ret) - goto cleanup; - } -SetPageMappedToDisk (page); - ret =0 ; - -cleanup: -free_buffer_head (bh); -put_extent (be); -put_extent (cow_read); -if (ret) { -/* Need to mark layout with bad read...should now -* just use nfs4 for reads and writes. -*/ -mark_bad_read (); - } -return ret; -} - -static int -bl_write_begin (struct pnfs_layout_segment *lseg,struct page *page,loff_t pos, -unsigned count,struct pnfs_fsdata *fsdata) -{ -unsigned from, to; -int ret; -sector_t *pages_to_mark =NULL ; -struct pnfs_block_layout *bl =BLK_LSEG2EXT (lseg); - -dprintk ("%s enter, %u@%lld\n" , __func__, count, pos); -print_page (page); -/* The following code assumes blocksize >= PAGE_CACHE_SIZE */ -if (bl->bl_blocksize < (PAGE_CACHE_SIZE >>9 )) { -dprintk ("%s Can't handle blocksize %llu\n" , __func__, - (u64 )bl->bl_blocksize); -put_lseg (fsdata->lseg); - fsdata->lseg =NULL ; -return 0 ; - } -if (PageMappedToDisk (page)) { -/* Basically, this is a flag that says we have -* successfully called write_begin already on this page. -*/ -/* NOTE - there are cache consistency issues here. -* For example, what if the layout is recalled, then regained? -* If the file is closed and reopened, will the page flags -* be reset? If not, we'll have to use layout info instead of -* the page flag. -*/ -return 0 ; - } - from = pos & (PAGE_CACHE_SIZE -1 ); - to = from + count; - ret =init_page_for_write (bl, page, from, to, &pages_to_mark); -if (ret) { -dprintk ("%s init page failed with %i" , __func__, ret); -/* Revert back to plain NFS and just continue on with -* write. This assumes there is no request attached, which -* should be true if we get here. -*/ -BUG_ON (PagePrivate (page)); -put_lseg (fsdata->lseg); - fsdata->lseg =NULL ; -kfree (pages_to_mark); - ret =0 ; - }else { - fsdata->private = pages_to_mark; - } -return ret; -} - -/* CAREFUL - what happens if copied < count??? */ -static int -bl_write_end (struct inode *inode,struct page *page,loff_t pos, -unsigned count,unsigned copied,struct pnfs_layout_segment *lseg) -{ -dprintk ("%s enter, %u@%lld, lseg=%p\n" , __func__, count, pos, lseg); -print_page (page); -if (lseg) -SetPageUptodate (page); -return 0 ; -} - -/* Return any memory allocated to fsdata->private, and take advantage -* of no page locks to mark pages noted in write_begin as needing -* initialization. -*/ -static void -bl_write_end_cleanup (struct file *filp,struct pnfs_fsdata *fsdata) -{ -struct page *page; -pgoff_t index; -sector_t *pos; -struct address_space *mapping = filp->f_mapping; -struct pnfs_fsdata *fake_data; -struct pnfs_layout_segment *lseg; - -if (!fsdata) -return ; - lseg = fsdata->lseg; -if (!lseg) -return ; - pos = fsdata->private; -if (!pos) -return ; -dprintk ("%s enter with pos=%llu\n" , __func__, (u64)(*pos)); -for (; *pos != ~0 ; pos++) { - index = *pos >> (PAGE_CACHE_SHIFT -9 ); -/* XXX How do we properly deal with failures here??? */ - page =grab_cache_page_write_begin (mapping, index,0 ); -if (!page) { -printk (KERN_ERR "%s BUG BUG BUG NoMem\n" , __func__); -continue ; - } -dprintk ("%s: Examining block page\n" , __func__); -print_page (page); -if (!PageMappedToDisk (page)) { -/* XXX How do we properly deal with failures here??? */ -dprintk ("%s Marking block page\n" , __func__); -init_page_for_write (BLK_LSEG2EXT (fsdata->lseg), page, -PAGE_CACHE_SIZE ,PAGE_CACHE_SIZE , -NULL ); -print_page (page); - fake_data =kzalloc (sizeof (*fake_data),GFP_KERNEL ); -if (!fake_data) { -printk (KERN_ERR "%s BUG BUG BUG NoMem\n" , - __func__); -unlock_page (page); -continue ; - } -get_lseg (lseg); - fake_data->lseg = lseg; - fake_data->bypass_eof =1 ; - mapping->a_ops->write_end (filp, mapping, - index <<PAGE_CACHE_SHIFT , -PAGE_CACHE_SIZE , -PAGE_CACHE_SIZE , - page, fake_data); -/* Note fake_data is freed by nfs_write_end */ - }else -unlock_page (page); - } -kfree (fsdata->private); - fsdata->private =NULL ; -} -static const struct nfs_pageio_ops bl_pg_read_ops = { + .pg_init = pnfs_generic_pg_init_read, .pg_test = pnfs_generic_pg_test, .pg_doio = nfs_generic_pg_readpages, };static const struct nfs_pageio_ops bl_pg_write_ops = { + .pg_init = pnfs_generic_pg_init_write, .pg_test = pnfs_generic_pg_test, .pg_doio = nfs_generic_pg_writepages, };@@ -1054,9 +1032,6 @@ static struct pnfs_layoutdriver_type blocklayout_type = { .name = "LAYOUT_BLOCK_VOLUME", .read_pagelist = bl_read_pagelist, .write_pagelist = bl_write_pagelist, - .write_begin = bl_write_begin, - .write_end = bl_write_end, - .write_end_cleanup = bl_write_end_cleanup, .alloc_layout_hdr = bl_alloc_layout_hdr, .free_layout_hdr = bl_free_layout_hdr, .alloc_lseg = bl_alloc_lseg,@@ -1083,8 +1058,7 @@ static int __init nfs4blocklayout_init(void) static void __exit nfs4blocklayout_exit(void) { -dprintk ("%s: NFSv4 Block Layout Driver Unregistering...\n" , - __func__); +dprintk ("%s: NFSv4 Block Layout Driver Unregistering...\n" , __func__);pnfs_unregister_layoutdriver (&blocklayout_type);bl_pipe_exit ();fs/nfs/file.c +3
-23@@ -384,15 +384,12 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping,pgoff_t index = pos >>PAGE_CACHE_SHIFT ;struct page *page;int once_thru =0 ; -struct pnfs_layout_segment *lseg;dfprintk (PAGECACHE ,"NFS: write_begin(%s/%s(%ld), %u@%lld)\n" , file->f_path.dentry->d_parent->d_name.name, file->f_path.dentry->d_name.name, mapping->host->i_ino, len, (long long ) pos); - lseg =pnfs_update_layout (mapping->host, -nfs_file_open_context (file), - pos, len,IOMODE_RW ,GFP_NOFS ); + start:/* * Prevent starvation issues if someone is doing a consistency @@ -412,9 +409,6 @@ start:if (ret) {unlock_page (page);page_cache_release (page); - *pagep =NULL ; - *fsdata =NULL ; - goto out; }else if (!once_thru &&nfs_want_read_modify_write (file, page, pos, len)) { once_thru =1 ;@@ -423,12 +417,6 @@ start:if (!ret) goto start; } - ret =pnfs_write_begin (file, page, pos, len, lseg, fsdata); - out: -if (ret) { -put_lseg (lseg); - *fsdata =NULL ; - }return ret; }@@ -438,7 +426,6 @@ static int nfs_write_end(struct file *file, struct address_space *mapping, {unsigned offset = pos & (PAGE_CACHE_SIZE -1 );int status; -struct pnfs_layout_segment *lseg; dfprintk(PAGECACHE, "NFS: write_end(%s/%s(%ld), %u@%lld)\n", file->f_path.dentry->d_parent->d_name.name,@@ -465,17 +452,10 @@ static int nfs_write_end(struct file *file, struct address_space *mapping,zero_user_segment (page, pglen,PAGE_CACHE_SIZE ); } - lseg =nfs4_pull_lseg_from_fsdata (file, fsdata); - status =pnfs_write_end (file, page, pos, len, copied, lseg); -if (status) - goto out; - status =nfs_updatepage (file, page, offset, copied, lseg, fsdata); + status =nfs_updatepage (file, page, offset, copied); -out:unlock_page (page);page_cache_release (page); -pnfs_write_end_cleanup (file, fsdata); -put_lseg (lseg);if (status <0 )return status;@@ -597,7 +577,7 @@ static int nfs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) ret =VM_FAULT_LOCKED ;if (nfs_flush_incompatible (filp, page) ==0 && -nfs_updatepage (filp, page,0 , pagelen,NULL ,NULL ) ==0 ) +nfs_updatepage (filp, page,0 , pagelen) ==0 ) goto out; ret =VM_FAULT_SIGBUS ;fs/nfs/pnfs.c +0
-41@@ -1223,41 +1223,6 @@ pnfs_try_to_write_data(struct nfs_write_data *wdata, }/* -* This gives the layout driver an opportunity to read in page "around" -* the data to be written. It returns 0 on success, otherwise an error code -* which will either be passed up to user, or ignored if -* some previous part of write succeeded. -* Note the range [pos, pos+len-1] is entirely within the page. -*/ -int _pnfs_write_begin (struct inode *inode,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -struct pnfs_fsdata **fsdata) -{ -struct pnfs_fsdata *data; -int status =0 ; - -dprintk ("--> %s: pos=%llu len=%u\n" , - __func__, (unsigned long long )pos, len); - data =kzalloc (sizeof (struct pnfs_fsdata ),GFP_KERNEL ); -if (!data) { - status = -ENOMEM ; - goto out; - } - data->lseg = lseg;/* refcount passed into data to be managed there */ - status =NFS_SERVER (inode)->pnfs_curr_ld->write_begin ( - lseg, page, pos, len, data); -if (status) { -kfree (data); - data =NULL ; - } -out: - *fsdata = data; -dprintk ("<-- %s: status=%d\n" , __func__, status); -return status; -} - -/* * Called by non rpc-based layout drivers */ int @@ -1373,12 +1338,6 @@ void pnfs_cleanup_layoutcommit(struct inode *inode, data); } -void pnfs_free_fsdata (struct pnfs_fsdata *fsdata) -{ -/* lseg refcounting handled directly in nfs_write_end */ -kfree (fsdata); -} -/* * For the LAYOUT4_NFSV4_1_FILES layout type, NFS_DATA_SYNC WRITEs and * NFS_UNSTABLE WRITEs with a COMMIT to data servers must store enough fs/nfs/pnfs.h +0
-115@@ -54,12 +54,6 @@ enum pnfs_try_status { PNFS_NOT_ATTEMPTED = 1, }; -struct pnfs_fsdata { -struct pnfs_layout_segment *lseg; -int bypass_eof; -void *private; -}; -#ifdef CONFIG_NFS_V4_1 #define LAYOUT_NFSV4_1_MODULE_PREFIX "nfs-layouttype4"@@ -113,14 +107,6 @@ struct pnfs_layoutdriver_type {*/ enum pnfs_try_status (*read_pagelist) (struct nfs_read_data *nfs_data);enum pnfs_try_status (*write_pagelist) (struct nfs_write_data *nfs_data,int how); -int (*write_begin) (struct pnfs_layout_segment *lseg,struct page *page, -loff_t pos,unsigned count, -struct pnfs_fsdata *fsdata); -int (*write_end)(struct inode *inode,struct page *page,loff_t pos, -unsigned count,unsigned copied, -struct pnfs_layout_segment *lseg); -void (*write_end_cleanup)(struct file *filp, -struct pnfs_fsdata *fsdata);void (*free_deviceid_node) (struct nfs4_deviceid_node *);@@ -196,7 +182,6 @@ enum pnfs_try_status pnfs_try_to_read_data(struct nfs_read_data *,void pnfs_generic_pg_init_read (struct nfs_pageio_descriptor *,struct nfs_page *);void pnfs_generic_pg_init_write (struct nfs_pageio_descriptor *,struct nfs_page *);bool pnfs_generic_pg_test (struct nfs_pageio_descriptor *pgio,struct nfs_page *prev,struct nfs_page *req); -void pnfs_free_fsdata (struct pnfs_fsdata *fsdata);int pnfs_layout_process (struct nfs4_layoutget *lgp);void pnfs_free_lseg_list (struct list_head *tmp_list);void pnfs_destroy_layout (struct nfs_inode *);@@ -208,10 +193,6 @@ void pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo,int pnfs_choose_layoutget_stateid (nfs4_stateid *dst,struct pnfs_layout_hdr *lo,struct nfs4_state *open_state); -int _pnfs_write_begin (struct inode *inode,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -struct pnfs_fsdata **fsdata);int mark_matching_lsegs_invalid (struct pnfs_layout_hdr *lo,struct list_head *tmp_list,struct pnfs_layout_range *recall_range);@@ -329,13 +310,6 @@ static inline void pnfs_clear_request_commit(struct nfs_page *req)put_lseg (req->wb_commit_lseg); } -static inline int pnfs_grow_ok (struct pnfs_layout_segment *lseg, -struct pnfs_fsdata *fsdata) -{ -return !fsdata || ((struct pnfs_layout_segment *)fsdata == lseg) || - !fsdata->bypass_eof; -} -/* Should the pNFS client commit and return the layout upon a setattr */ static inline bool pnfs_ld_layoutret_on_setattr (struct inode *inode)@@ -346,49 +320,6 @@ pnfs_ld_layoutret_on_setattr(struct inode *inode)PNFS_LAYOUTRET_ON_SETATTR ; } -static inline int pnfs_write_begin (struct file *filp,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -void **fsdata) -{ -struct inode *inode = filp->f_dentry->d_inode; -struct nfs_server *nfss =NFS_SERVER (inode); -int status =0 ; - - *fsdata = lseg; -if (lseg && nfss->pnfs_curr_ld->write_begin) - status =_pnfs_write_begin (inode, page, pos, len, lseg, - (struct pnfs_fsdata **) fsdata); -return status; -} - -/* CAREFUL - what happens if copied < len??? */ -static inline int pnfs_write_end (struct file *filp,struct page *page, -loff_t pos,unsigned len,unsigned copied, -struct pnfs_layout_segment *lseg) -{ -struct inode *inode = filp->f_dentry->d_inode; -struct nfs_server *nfss =NFS_SERVER (inode); - -if (nfss->pnfs_curr_ld && nfss->pnfs_curr_ld->write_end) -return nfss->pnfs_curr_ld->write_end (inode, page, pos, len, - copied, lseg); -else -return 0 ; -} - -static inline void pnfs_write_end_cleanup (struct file *filp,void *fsdata) -{ -struct nfs_server *nfss =NFS_SERVER (filp->f_dentry->d_inode); - -if (fsdata && nfss->pnfs_curr_ld) { -if (nfss->pnfs_curr_ld->write_end_cleanup) - nfss->pnfs_curr_ld->write_end_cleanup (filp, fsdata); -if (nfss->pnfs_curr_ld->write_begin) -pnfs_free_fsdata (fsdata); - } -} -static inline int pnfs_return_layout (struct inode *ino) {struct nfs_inode *nfsi =NFS_I (ino);@@ -400,19 +331,6 @@ static inline int pnfs_return_layout(struct inode *ino) return 0; } -static inline struct pnfs_layout_segment * -nfs4_pull_lseg_from_fsdata (struct file *filp,void *fsdata) -{ -if (fsdata) { -struct nfs_server *nfss =NFS_SERVER (filp->f_dentry->d_inode); - -if (nfss->pnfs_curr_ld && nfss->pnfs_curr_ld->write_begin) -return ((struct pnfs_fsdata *) fsdata)->lseg; -return (struct pnfs_layout_segment *)fsdata; - } -return NULL ; -} -#else /* CONFIG_NFS_V4_1 */ static inline void pnfs_destroy_all_layouts (struct nfs_client *clp)@@ -433,12 +351,6 @@ static inline void put_lseg(struct pnfs_layout_segment *lseg) { } -static inline int pnfs_grow_ok (struct pnfs_layout_segment *lseg, -struct pnfs_fsdata *fsdata) -{ -return 1 ; -} -static inline enum pnfs_try_status pnfs_try_to_read_data (struct nfs_read_data *data,const struct rpc_call_ops *call_ops)@@ -458,26 +370,6 @@ static inline int pnfs_return_layout(struct inode *ino)return 0 ; } -static inline int pnfs_write_begin (struct file *filp,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -void **fsdata) -{ - *fsdata =NULL ; -return 0 ; -} - -static inline int pnfs_write_end (struct file *filp,struct page *page, -loff_t pos,unsigned len,unsigned copied, -struct pnfs_layout_segment *lseg) -{ -return 0 ; -} - -static inline void pnfs_write_end_cleanup (struct file *filp,void *fsdata) -{ -} -static inline bool pnfs_ld_layoutret_on_setattr (struct inode *inode) {@@ -554,13 +446,6 @@ static inline int pnfs_layoutcommit_inode(struct inode *inode, bool sync)static inline void nfs4_deviceid_purge_client (struct nfs_client *ncl) { } - -static inline struct pnfs_layout_segment * -nfs4_pull_lseg_from_fsdata (struct file *filp,void *fsdata) -{ -return NULL ; -} -#endif /* CONFIG_NFS_V4_1 */ #endif /* FS_NFS_PNFS_H */ fs/nfs/write.c +4
-8@@ -673,9 +673,7 @@ out: }static int nfs_writepage_setup (struct nfs_open_context *ctx,struct page *page, -unsigned int offset,unsigned int count, -struct pnfs_layout_segment *lseg,void *fsdata) - +unsigned int offset,unsigned int count) {struct nfs_page *req;@@ -683,8 +681,7 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page,if (IS_ERR (req))return PTR_ERR (req);/* Update file length */ -if (pnfs_grow_ok (lseg, fsdata)) -nfs_grow_file (page, offset, count); +nfs_grow_file (page, offset, count);nfs_mark_uptodate (page, req->wb_pgbase, req->wb_bytes);nfs_mark_request_dirty (req);nfs_clear_page_tag_locked (req);@@ -737,8 +734,7 @@ static int nfs_write_pageuptodate(struct page *page, struct inode *inode)* things with a page scheduled for an RPC call (e.g. invalidate it). */ int nfs_updatepage (struct file *file,struct page *page, -unsigned int offset,unsigned int count, -struct pnfs_layout_segment *lseg,void *fsdata) +unsigned int offset,unsigned int count) {struct nfs_open_context *ctx =nfs_file_open_context (file);struct inode *inode = page->mapping->host;@@ -763,7 +759,7 @@ int nfs_updatepage(struct file *file, struct page *page, offset =0 ; } - status =nfs_writepage_setup (ctx, page, offset, count, lseg, fsdata); + status =nfs_writepage_setup (ctx, page, offset, count);if (status <0 )nfs_set_pageerror (page);include/linux/nfs_fs.h +1
-2@@ -510,8 +510,7 @@ extern int nfs_congestion_kb;extern int nfs_writepage (struct page *page,struct writeback_control *wbc);extern int nfs_writepages (struct address_space *,struct writeback_control *);extern int nfs_flush_incompatible (struct file *file,struct page *page); -extern int nfs_updatepage (struct file *,struct page *,unsigned int , -unsigned int ,struct pnfs_layout_segment *,void *); +extern int nfs_updatepage (struct file *,struct page *,unsigned int ,unsigned int );extern void nfs_writeback_done (struct rpc_task *,struct nfs_write_data *);/*
Re: [PATCH 1/6] SQUASHME: pnfs-block: Remove write_begin/end hooks
Benny Halevy wrote:
Hi, sorry for my late response.
I like the direction this is going!
However I have a few comments inline below.
Thanks very much for the comments. We actually had a review with Trond and many of these concerns are already being addressed. Others are new and I will certainly look at them. I will have a new patch set later this week, based on Trond's nfs-for-next branch. I will send that to the list and would appreciate any comments you might have.
RE: [PATCH 1/6] SQUASHME: pnfs-block: Remove write_begin/end hooks
In reply to Benny HalevyHi, Benny,
Thanks a lot for your comments.
Cheers,
-Bergwolf
Quoted 11 lines
-----Original Message-----
From: Benny Halevy [mailto:[email protected]]
Sent: Wednesday, July 13, 2011 8:52 PM
To: Jim Rees; Peng, Tao
Cc: [email protected]; peter honeyman
Subject: Re: [PATCH 1/6] SQUASHME: pnfs-block: Remove write_begin/end hooksHi, sorry for my late response.
I like the direction this is going!
However I have a few comments inline below.
Quoted 6 lines
In general, combining cosmetic changes or cleanups with the real functionality makes the patch really hard to review.
Please separate the patches so that each one is boiled down to the minimum to make it self-sufficient, even if that ends up with more patches to review, (see
Documentation/SubmittingPatches)
After squashing, this patch will only be rewriting bl_write_pagelist.
Quoted 68 lines
more inline...
fs/nfs/blocklayout/blocklayout.c static int is_hole(struct pnfs_block_extent *be, sector_t
Quoted 31 lines · Jim Rees
From: Peng Tao <[email protected]>
Do not call
pnfs_update_layoutatwrite_begin, and rewritebl_write_pagelistto handle EOF there.Signed-off-by Peng Tao <[email protected]>--- fs/nfs/blocklayout/blocklayout.c | 652 ++++++++++++++++++-------------------- fs/nfs/file.c | 26 +-- fs/nfs/pnfs.c | 41 --- fs/nfs/pnfs.h | 115 ------- fs/nfs/write.c | 12 +- include/linux/nfs_fs.h | 3 +- 6 files changed, 321 insertions(+), 528 deletions(-)fs/nfs/blocklayout/blocklayout.c +2
-2@@ -32,8 +32,8 @@#include <linux/module.h> #include <linux/init.h> -#include <linux/buffer_head.h> /* various write calls */ -#include <linux/bio.h> /* struct bio */ +#include <linux/buffer_head.h> /* various write calls */ +#include <linux/bio.h> /* struct bio */ #include <linux/vmalloc.h> #include "blocklayout.h" @@ -75,16 +75,11 @@ static int is_hole(struct pnfs_block_extent *be, sector_tisect)
Quoted 32 lines
fs/nfs/blocklayout/blocklayout.c +4
-9*/ static int is_writable (struct pnfs_block_extent *be,sector_t isect) { -if (be->be_state ==PNFS_BLOCK_READWRITE_DATA ) -return 1 ; -else if (be->be_state !=PNFS_BLOCK_INVALID_DATA ) -return 0 ; -else -return is_sector_initialized (be->be_inval, isect); +return (be->be_state ==PNFS_BLOCK_READWRITE_DATA || + be->be_state ==PNFS_BLOCK_INVALID_DATA ); } -static int -dont_like_caller (struct nfs_page *req) +static int dont_like_caller (struct nfs_page *req) {if (atomic_read (&req->wb_complete)) {/* Called by _multi */ @@ -109,7 +104,7 @@ static inline struct parallel_io *alloc_parallel(void *data) {struct parallel_io *rv; - rv =kmalloc (sizeof (*rv),GFP_KERNEL ); + rv =kmalloc (sizeof (*rv),GFP_KERNEL );if (rv) { rv->data = data;kref_init (&rv->refcnt);@@ -136,21 +131,19 @@ static inline void put_parallel(struct parallel_io *p)kref_put (&p->refcnt, destroy_parallel); }
Quoted 14 lines
fs/nfs/blocklayout/blocklayout.c static inline void put_parallel(struct parallel_io *p) · line 140
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +2
-3-static struct bio * -bl_submit_bio (int rw,struct bio *bio) +static struct bio *bl_submit_bio (int rw,struct bio *bio) {if (bio) {get_parallel (bio->bi_private);dprintk ("%s submitting %s bio %u@%llu\n" , __func__, rw ==READ ?"read" :"write" , - bio->bi_size, (u64 )bio->bi_sector); + bio->bi_size, (u64 ) bio->bi_sector);Do you even support or want to support 32 bit
bi_sectors? If not, you could just addBUILD_BUG_ON(sizeof(sector_t) != 8); Otherwise, typecasting to (unsigned long long) would be more portable.
Will change it to (unsigned long long). Thanks.
Quoted 6 lines
fs/nfs/blocklayout/blocklayout.c static inline void put_parallel(struct parallel_io *p) · line 143
Quoted 4 lines
fs/nfs/blocklayout/blocklayout.c +0
-0submit_bio (rw, bio); }return NULL ;What's the point of always returning NULL?
It intends to submit bio and let caller re-allocate it. The logic is similar to mpage_bio_submit().
Quoted 181 lines
fs/nfs/blocklayout/blocklayout.c static void bl_rpc_do_nothing(struct rpc_task *task, void
Quoted 20 lines
fs/nfs/blocklayout/blocklayout.c +2
-4} -static inline void -bl_done_with_rpage (struct page *page,const int ok) +static inline void bl_done_with_rpage (struct page *page,const int ok) {if (ok) {ClearPagePnfsErr (page);@@ -191,8 +184,7 @@ static void bl_read_cleanup(struct work_struct *work)pnfs_ld_read_done (rdata); } -static void -bl_end_par_io_read (void *data) +static void bl_end_par_io_read (void *data) {struct nfs_read_data *rdata = data;@@ -208,8 +200,7 @@ static void bl_rpc_do_nothing(struct rpc_task *task, void*calldata)
fs/nfs/blocklayout/blocklayout.c static void mark_extents_written(struct
Quoted 62 lines
fs/nfs/blocklayout/blocklayout.c +10
-9return ; } -static enum pnfs_try_status -bl_read_pagelist (struct nfs_read_data *rdata) +static enum pnfs_try_status bl_read_pagelist (struct nfs_read_data *rdata) {int i, hole;struct bio *bio =NULL ;@@ -222,7 +213,7 @@ bl_read_pagelist(struct nfs_read_data *rdata)int pg_index = rdata->args.pgbase >>PAGE_CACHE_SHIFT ;dprintk ("%s enter nr_pages %u offset %lld count %Zd\n" , __func__, - rdata->npages, f_offset, count); + rdata->npages, f_offset, count);if (dont_like_caller (rdata->req)) {dprintk ("%s dont_like_caller failed\n" , __func__);@@ -260,10 +251,10 @@ bl_read_pagelist(struct nfs_read_data *rdata)break ; } extent_length = be->be_length - - (isect - be->be_f_offset); + (isect - be->be_f_offset);if (cow_read) {sector_t cow_length = cow_read->be_length - - (isect - cow_read->be_f_offset); + (isect - cow_read->be_f_offset); extent_length =min (extent_length, cow_length); } }@@ -282,15 +273,17 @@ bl_read_pagelist(struct nfs_read_data *rdata) be_read = (hole && cow_read) ? cow_read : be;for (;;) {if (!bio) { - bio =bio_alloc (GFP_NOIO , rdata->npages - i); + bio = +bio_alloc (GFP_NOIO , + rdata->npages - i);if (!bio) {/* Error out this page */ bl_done_with_rpage (pages[i],0 );break ; } bio->bi_sector = isect - - be_read->be_f_offset + - be_read->be_v_offset; + be_read->be_f_offset + + be_read->be_v_offset; bio->bi_bdev = be_read->be_mdev; bio->bi_end_io = bl_end_io_read; bio->bi_private = par;@@ -315,7 +308,7 @@ bl_read_pagelist(struct nfs_read_data *rdata)put_parallel (par);return PNFS_ATTEMPTED ; - use_mds: +use_mds:dprintk ("Giving up and using normal NFS\n" );return PNFS_NOT_ATTEMPTED ; }@@ -335,18 +328,16 @@ static void mark_extents_written(struct
pnfs_block_layout*bl,Quoted 97 lines
fs/nfs/blocklayout/blocklayout.c +33
-12while (isect < end) {sector_t len; be =find_get_extent (bl, isect,NULL ); -BUG_ON (!be);/* FIXME */ +BUG_ON (!be);/* FIXME */ len =min (end, be->be_f_offset + be->be_length) - isect;if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) -mark_for_commit (be, isect, len);/* What if fails? */ +mark_for_commit (be, isect, len);/* What if fails? */ isect += len;put_extent (be); } } -/* STUB - this needs thought */ -static inline void -bl_done_with_wpage (struct page *page,const int ok) +static inline void bl_done_with_wpage (struct page *page,const int ok) {if (!ok) {SetPageError (page);@@ -360,15 +351,37 @@ bl_done_with_wpage(struct page *page, const int ok)spin_unlock (&inode->i_lock); } } -/* end_page_writeback called in rpc_release. Should be done here. */ } -/* This is basically copied from mpage_end_io_read */ +static void bl_end_io_write_zero (struct bio *bio,int err) +{ +struct parallel_io *par = bio->bi_private; +const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags); +struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; + +do { +struct page *page = bvec->bv_page; + +if (--bvec >= bio->bi_io_vec) +prefetchw (&bvec->bv_page->flags); +bl_done_with_wpage (page, uptodate); +/* This is the zeroing page we added */ +end_page_writeback (page); +page_cache_release (page); + }while (bvec >= bio->bi_io_vec); +if (!uptodate && !wdata->pnfs_error) + wdata->pnfs_error = -EIO ; +bio_put (bio); +put_parallel (par); +} +static void bl_end_io_write (struct bio *bio,int err) { -void *data = bio->bi_private; +struct parallel_io *par = bio->bi_private;const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags);struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; do { struct page *page = bvec->bv_page;@@ -377,8 +390,10 @@ static void bl_end_io_write(struct bio *bio, int err)prefetchw (&bvec->bv_page->flags);bl_done_with_wpage (page, uptodate); }while (bvec >= bio->bi_io_vec); +if (!uptodate && !wdata->pnfs_error) + wdata->pnfs_error = -EIO ;bio_put (bio); -put_parallel (data); +put_parallel (par); }/* Function scheduled for call during bl_end_par_io_write, @@ -391,7 +406,7 @@ static void bl_write_cleanup(struct work_struct *work)dprintk ("%s enter\n" , __func__); task = container_of(work, struct rpc_task, u.tk_work); wdata = container_of(task, struct nfs_write_data, task); -if (!wdata->task.tk_status) { +if (!wdata->pnfs_error) {/* Marks for LAYOUTCOMMIT */ /* BUG - this should be called after each bio, not after * all finish, unless have some way of storing success/failure @@ -403,8 +418,7 @@ static void bl_write_cleanup(struct work_struct *work) }/* Called when last of bios associated with a bl_write_pagelist call finishes */ -static void -bl_end_par_io_write (void *data) +static void bl_end_par_io_write (void *data) {struct nfs_write_data *wdata = data;@@ -415,19 +429,118 @@ bl_end_par_io_write(void *data)schedule_work (&wdata->task.u.tk_work); }
Quoted 12 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 440
Quoted 9 lines
fs/nfs/blocklayout/blocklayout.c +9
-0+/* STUB - mark intersection of layout and page as bad, so is not +* used again. +*/ +static void mark_bad_read (void ) +{ +return ; +} + +/* Copied from buffer.c */ NAK. If this code needs to be shared internally you should make it public, declare it properly in
linux/buffer_head.h, andEXPORT_SYMBOL_GPLit
This will be removed as expained bellow.
Quoted 17 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 452
Quoted 13 lines
fs/nfs/blocklayout/blocklayout.c +12
-0+static void __end_buffer_read_notouch (struct buffer_head *bh,int uptodate) +{ +if (uptodate) { +set_buffer_uptodate (bh); + }else { +/* This happens, due to failed READA attempts. */ +clear_buffer_uptodate (bh); + } +unlock_buffer (bh); +} + +/* Copied from buffer.c */ This should be turned into a static inline helper in
buffer_head.hsince it just an alias for__end_buffer_read_notouch, but anyway, why do _you_ need it here?
After looking into it more closely, I think we can just remove all the end_buffer_read_xxx functions here and call bh_submit_read() in init_page_for_write(). I will make the change. Thanks.
Quoted 11 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 460
Quoted 9 lines
fs/nfs/blocklayout/blocklayout.c +8
-0+static void end_buffer_read_nobh (struct buffer_head *bh,int uptodate) +{ +__end_buffer_read_notouch (bh, uptodate); +} + +/* +* map_block: map a requested I/0 block (isect) into an offset in the LVM +* meta block_device What's a "meta
block_device"?
We can remove the *meta* if you don't like the name.
Quoted 6 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 463
Quoted 4 lines
fs/nfs/blocklayout/blocklayout.c +3
-0+*/ +static void +map_block (sector_t isect,struct pnfs_block_extent *be,struct buffer_head *bh)I'd put the output parameter, bh, first, followed by the input parameters.
Will do that. Thanks.
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 470
Quoted 8 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+{ +dprintk ("%s enter be=%p\n" , __func__, be); +bl_end_io_write_zero +set_buffer_mapped (bh ); + bh->b_bdev = be->be_mdev; + bh->b_blocknr = (isect - be->be_f_offset + be->be_v_offset) >> + (be->be_mdev->bd_inode->i_blkbits -9 );Please use
SECTOR_SHIFTfrom <linux/device-mapper.h> rather than '9'.
Will do that. Thanks.
Quoted 13 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 478
Quoted 9 lines
fs/nfs/blocklayout/blocklayout.c +8
-0+ +dprintk ("%s isect %ld, bh->b_blocknr %ld, using bsize %Zd\n" , + __func__, (long )isect, (long )bh->b_blocknr, bh->b_size); +return ; +} + +/* Given an unmapped page, zero it (or read in page for COW), +*/ Why is the (or read ...) in parenthesis?
It's of exactly the same importance as the zeroing part of this function. Also, single line comments can have the closing "*/" on the same line.
Will change it. Thanks.
Quoted 13 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 488
Quoted 11 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+static int +init_page_for_write (struct page *page,struct pnfs_block_extent *cow_read) +{ +struct buffer_head *bh =NULL ; +int ret =0 ; +sector_t isect; + +dprintk ("%s enter, %p\n" , __func__, page); +BUG_ON (PageUptodate (page)); +/* not COW, zero and return */ This comment is not needed. The code is straight forward as it is.
Will remove it.
Quoted 11 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 495
Quoted 8 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+if (!cow_read) { +zero_user_segment (page,0 ,PAGE_SIZE ); +SetPageUptodate (page); + goto cleanup; + } + +/* COW, readin page */ This one can be removed too, unless you have something non-trivial to document.
Will remove it.
Quoted 9 lines
fs/nfs/blocklayout/blocklayout.c bl_end_par_io_write(void *data) · line 502
Quoted 8 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+ bh =alloc_page_buffers (page,PAGE_CACHE_SIZE ,0 ); +if (!bh) { + ret = -ENOMEM ; + goto cleanup; + } + + isect = (sector_t ) page->index << (PAGE_CACHE_SHIFT -9 );
Quoted 4 lines
ditto
SECTOR_SHIFT
I see you use (PAGE_CACHE_SHIFT- 9) in other places too so even better,
#definePAGE_CACHE_SECTOR_SHIFT(PAGE_CACHE_SHIFT-SECTOR_SHIFT)
Agreed, will do it.
Quoted 88 lines
Also, no space after the type cast (not sure why this is not in
Documentation/CodingStyle)Quoted 71 lines
fs/nfs/blocklayout/blocklayout.c +38
-12+map_block (isect, cow_read, bh); +lock_buffer (bh); + bh->b_end_io = end_buffer_read_nobh; +submit_bh (READ , bh); +dprintk ("%s: Waiting for buffer read\n" , __func__); +wait_on_buffer (bh); +if (!buffer_uptodate (bh)) { + ret = -EIO ; + goto cleanup; + } +SetPageUptodate (page); +SetPageMappedToDisk (page); + +cleanup: +put_extent (cow_read); +if (bh) +free_buffer_head (bh); +if (ret) { +/* Need to mark layout with bad read...should now +* just use nfs4 for reads and writes. +*/ +mark_bad_read (); + } +return ret; +} +static enum pnfs_try_status -bl_write_pagelist (struct nfs_write_data *wdata, -int sync) +bl_write_pagelist (struct nfs_write_data *wdata,int sync) { -int i; +int i, ret, npg_zero, pg_index, last =0 ;struct bio *bio =NULL ; -struct pnfs_block_extent *be =NULL ; -sector_t isect, extent_length =0 ; +struct pnfs_block_extent *be =NULL , *cow_read =NULL ; +sector_t isect, last_isect =0 , extent_length =0 ;struct parallel_io *par;loff_t offset = wdata->args.offset;size_t count = wdata->args.count;struct page **pages = wdata->args.pages; -int pg_index = wdata->args.pgbase >>PAGE_CACHE_SHIFT ; +struct page *page; +pgoff_t index; +int npg_per_block = +NFS_SERVER (wdata->inode)->pnfs_blksize >>PAGE_CACHE_SHIFT ;dprintk ("%s enter, %Zu@%lld\n" , __func__, count, offset);if (!wdata->lseg) {@@ -439,11 +552,8 @@ bl_write_pagelist(struct nfs_write_data *wdata,return PNFS_NOT_ATTEMPTED ; }/* At this point, wdata->pages is a (sequential) list of nfs_pages. -* We want to write each, and if there is an error remove it from -* list and call -* nfs_retry_request(req) to have it redone using nfs. -* QUEST? Do as block or per req? Think have to do per block -* as part of end_bio +* We want to write each, and if there is an error set pnfs_error +* to have it redone using nfs. */ par =alloc_parallel (wdata);if (!par)@@ -454,7 +564,106 @@ bl_write_pagelist(struct nfs_write_data *wdata,/* At this point, have to be more careful with error handling */ isect = (sector_t) ((offset & (long )PAGE_CACHE_MASK ) >>9 ); -for (i = pg_index; i < wdata->npages ; i++) { + be =find_get_extent (BLK_LSEG2EXT (wdata->lseg), isect, &cow_read); +if (!be || !is_writable (be, isect)) {fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 578
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+dprintk ("%s no matching extents!\n" , __func__); + wdata->pnfs_error = -EINVAL ; + goto out; + } + +/* First page inside INVALID extent */ +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + npg_zero = (offset >>PAGE_CACHE_SHIFT ) % npg_per_block; + isect = (sector_t) (((offset - npg_zero *PAGE_CACHE_SIZE ) & + (long )PAGE_CACHE_MASK ) >>9 );ditto
SECTOR_SHIFT.
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 583
Quoted 5 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+ extent_length = be->be_length - (isect - be->be_f_offset); + +fill_invalid_ext: +dprintk ("%s need to zero %d pages\n" , __func__, npg_zero); +while (npg_zero) {a for loop seems more appropriate here
OK. Will change it.
Quoted 34 lines
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 585
Quoted 3 lines
fs/nfs/blocklayout/blocklayout.c +2
-0+/* page ref released in bl_end_io_write_zero */ + index = isect >> (PAGE_CACHE_SHIFT -9 );ditto
PAGE_CACHE_SECTOR_SHIFTQuoted 17 lines
fs/nfs/blocklayout/blocklayout.c +17
-0+dprintk ("%s zero %dth page: index %lu isect %lu\n" , + __func__, npg_zero, index, isect); + page = +find_or_create_page (wdata->inode->i_mapping, index, +GFP_NOFS ); +if (!page) { +dprintk ("%s oom\n" , __func__); + wdata->pnfs_error = -ENOMEM ; + goto out; + } + +/* PageDirty: Other will write this out +* PageWriteback: Other is writing this out +* PageUptodate && sector_initialized: already written out +*/ + if (PageDirty(page) || PageWriteback(page) || + (PageUptodate(page) &&fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 612
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+ is_sector_initialized(be->be_inval, isect))) { +dprintk + ("page %lu is uptodate %d dirty %d writeback %d\n" , + page->index,PageUptodate (page), +PageDirty (page),PageWriteback (page)); +unlock_page (page); +page_cache_release (page); + npg_zero--; + isect +=PAGE_CACHE_SIZE >>9 ; + extent_length -=PAGE_CACHE_SIZE >>9 ;
Quoted 4 lines
#define
SECTORS_PER_PAGE(PAGE_CACHE_SIZE>>SECTOR_SHIFT)(should this be done in the iteration leg of the for statement or at a label just before the end of the loop?)
Will add goto a lable to the end of the loop. Thanks.
Quoted 17 lines
Quoted 6 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+continue ; + } +if (!PageUptodate (page) && cow_read) { +/* New page, readin or zero it */ +init_page_for_write (page, cow_read);fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 627
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +10
-0+ } +set_page_writeback (page); +unlock_page (page); + + ret =mark_initialized_sectors (be->be_inval, isect, +PAGE_CACHE_SECTORS , +NULL ); +if (unlikely (ret)) { +dprintk ("%s mark_initialized_sectors fail %d\n" , + __func__, ret);
Quoted 1 line
what about
end_page_writeback(page)?
Nice catch. Will add it here. Thanks.
Quoted 9 lines
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 632
Quoted 6 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+page_cache_release (page); + wdata->pnfs_error = ret; + goto out; + } +for (;;) {Doing an infinite loop here seems like a bad idea.
Quoted 8 lines
My understanding is that eventually you want to submit bios for all
npg_zeropages.Each time you go through this section you add one page, possibly allocating a bio if this is the first time, or if you previously submitted the bio (and set bio to NULL, which shouldn't be a side effect of
bl_submit_bio, but rather it should be clearly spelled out by its callers)
The idea is that when current bio is full, we submit and re-allocate it. And then add the page to it.
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 634
Quoted 3 lines
fs/nfs/blocklayout/blocklayout.c +2
-0+if (!bio) { + bio =bio_alloc (GFP_NOIO , npg_zero--);The post-decrement of
npg_zerois non intuitive sincebio_allocallocatesnpg_zeroiovecs and the decrement logically belongs to thebio_add_pagebelow.
I'll move npg_zero-- to the end of the loop so that is where we've done with current page.
Quoted 8 lines
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 639
Quoted 6 lines
fs/nfs/blocklayout/blocklayout.c +5
-0+if (unlikely (!bio)) { +end_page_writeback (page); +page_cache_release (page); + wdata->pnfs_error = -ENOMEM ; + goto out;this calls for a common error path handling path.
The out path is the error handling path. My concern is that if we take care of page_writeback and release in error handling path, we will have two more goto lable. But the error only happens in two places as well. So it may not worth the code jump.
Quoted 13 lines
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 650
Quoted 12 lines
fs/nfs/blocklayout/blocklayout.c +11
-0+ } + bio->bi_sector = + isect - be->be_f_offset + + be->be_v_offset; + bio->bi_bdev = be->be_mdev; + bio->bi_end_io = bl_end_io_write_zero; + bio->bi_private = par; + } +if (bio_add_page (bio, page,PAGE_CACHE_SIZE ,0 )) +break ; + bio =bl_submit_bio (WRITE , bio);
Quoted 4 lines
so if
bio_add_pagefails, you want to submit the current bio, allocate a new one andbio_add_pageto the new one. but if it fails the second time you absolutely don't want to submit the one you've just allocated but rather you want to pass the error on.
According to the comments of bio_add_page:
The target block device must allow bio's up to PAGE_SIZE,
so it is always possible to add a single page to an empty bio.
I think if we allocate a new bio and call bio_add_page, it will always succeed. Or am I misunderstanding it?
And the same allocate/submit looping forever exists in generic code as well. Please see do_mpage_readpage and __mpage_writepage for example.
Quoted 3 lines
I think it'd be better to move the contents of this loop into a couple functions that will do the additive iteration. something along these lines:
Nice example. I will add these. Thank you very much!
Quoted 62 lines
static struct bio *bl_alloc_init_bio(int npg, struct
pnfs_block_extent*be, void (*end_io)(struct bio *, int err), structparallel_io*par){ struct bio *bio; bio = bio_alloc(GFP_NOIO, npg); if (!bio) return NULL; bio->bi_sector = isect - be->be_f_offset + be->be_v_offset; bio->bi_bdev = be->be_mdev; bio->bi_end_io = end_io; bio->bi_private = par; return bio; }static struct bio *bl_add_page_to_bio(struct bio *bio, int npg, struct page *page, struct
pnfs_block_extent*be, void (*end_io)(struct bio *, int err), structparallel_io*par){ int retried = 0;retry:
if (!bio) { bio = bl_alloc_init_bio(npg, be, end_io, par); if (!bio) return PTR_ERR(-ENOMEM); } if (!bl_add_page(bio, page, PAGE_CACHE_SIZE, 0)) { if (!retried++) { bl_submit_bio(bio); bio = NULL; goto retry; } bio_free(bio); return ERR_PTR(-EIO); } return bio; }and the caller should decrement
npg_zeroor handle the error with respect to the page andwdata->pnfs_error.fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 657
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+ } +/* FIXME: This should be done in bi_end_io */ +mark_extents_written (BLK_LSEG2EXT (wdata->lseg), + page->index <<PAGE_CACHE_SHIFT , +PAGE_CACHE_SIZE ); + isect +=PAGE_CACHE_SIZE >>9 ; + extent_length -=PAGE_CACHE_SIZE >>9 ;common iteration path, see comment above
fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 664
Quoted 7 lines
fs/nfs/blocklayout/blocklayout.c +7
-0+ } +if (last) + goto write_done; + } + bio =bl_submit_bio (WRITE , bio); + +/* Middle pages */
Quoted 3 lines
this whole function is way too long.
Is it possible to refactor out the different parts of it into their own static functions?
Will do it.
Quoted 35 lines
Quoted 22 lines
fs/nfs/blocklayout/blocklayout.c +8
-4+ pg_index = wdata->args.pgbase >>PAGE_CACHE_SHIFT ; +for (i = pg_index; i < wdata->npages; i++) {if (!extent_length) {/* We've used up the previous extent */ put_extent (be);@@ -463,24 +672,32 @@ bl_write_pagelist(struct nfs_write_data *wdata, be =find_get_extent (BLK_LSEG2EXT (wdata->lseg), isect,NULL );if (!be || !is_writable (be, isect)) { -/* FIXME */ -bl_done_with_wpage (pages[i],0 ); -break ; + wdata->pnfs_error = -EINVAL ; + goto out; } extent_length = be->be_length - - (isect - be->be_f_offset); + (isect - be->be_f_offset); + } +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + ret = mark_initialized_sectors(be->be_inval, isect,fs/nfs/blocklayout/blocklayout.c bl_write_pagelist(struct nfs_write_data *wdata, · line 692
Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +8
-0+PAGE_CACHE_SECTORS , + NULL); +if (unlikely (ret)) { +dprintk ("%s mark_initialized_sectors fail %d\n" , + __func__, ret); + wdata->pnfs_error = ret; + goto out; + } }for (;;) {ditto.
Quoted 2 lines
can we use the same code structure as I suggested above also here and in
bl_read_pagelist?
Yes, I will change it. Thanks.
Quoted 834 lines
fs/nfs/blocklayout/blocklayout.c release_extents(struct pnfs_block_layout *bl, struct
Quoted 48 lines
fs/nfs/blocklayout/blocklayout.c +21
-7if (!bio) { bio =bio_alloc (GFP_NOIO , wdata->npages - i);if (!bio) { -/* Error out this page */ -/* FIXME */ -bl_done_with_wpage (pages[i],0 ); -break ; + wdata->pnfs_error = -ENOMEM ; + goto out; } bio->bi_sector = isect - be->be_f_offset + - be->be_v_offset; + be->be_v_offset; bio->bi_bdev = be->be_mdev; bio->bi_end_io = bl_end_io_write; bio->bi_private = par;@@ -490,11 +707,27 @@ bl_write_pagelist(struct nfs_write_data *wdata, bio =bl_submit_bio (WRITE , bio); } isect +=PAGE_CACHE_SIZE >>9 ; + last_isect = isect; extent_length -=PAGE_CACHE_SIZE >>9 ; } - wdata->res.count = (isect <<9 ) - (offset); -if (count < wdata->res.count) + +/* Last page inside INVALID extent */ +if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { + bio =bl_submit_bio (WRITE , bio); + npg_zero = npg_per_block - + (last_isect >> (PAGE_CACHE_SHIFT -9 )) % npg_per_block; +if (npg_zero < npg_per_block) { + last =1 ; + goto fill_invalid_ext; + } + } + +write_done: + wdata->res.count = (last_isect <<9 ) - (offset); +if (count < wdata->res.count) { wdata->res.count = count; + } +out:put_extent (be);bl_submit_bio (WRITE , bio);put_parallel (par);@@ -521,8 +754,7 @@ release_extents(struct pnfs_block_layout *bl, struct
pnfs_layout_range*range)Quoted 10 lines
fs/nfs/blocklayout/blocklayout.c +1
-2spin_unlock (&bl->bl_ext_lock); } -static void -release_inval_marks (struct pnfs_inval_markings *marks) +static void release_inval_marks (struct pnfs_inval_markings *marks) {struct pnfs_inval_tracking *pos, *temp;fs/nfs/blocklayout/blocklayout.c bl_cleanup_layoutcommit(struct pnfs_layout_hdr *lo, · line 618
Quoted 4 lines
fs/nfs/blocklayout/blocklayout.c +0
-1@@ -615,7 +847,8 @@ bl_cleanup_layoutcommit(struct pnfs_layout_hdr *lo, struct nfs4_layoutcommit_data *lcdata) {dprintk ("%s enter\n" , __func__); - clean_pnfs_block_layoutupdate(BLK_LO2EXT(lo), &lcdata->args,
lcdata->res.status);fs/nfs/blocklayout/blocklayout.c static void free_blk_mountid(struct block_mount_id
Quoted 16 lines
fs/nfs/blocklayout/blocklayout.c +3
-2+clean_pnfs_block_layoutupdate (BLK_LO2EXT (lo), &lcdata->args, + lcdata->res.status); }static void free_blk_mountid (struct block_mount_id *mid)@@ -625,8 +858,7 @@ static void free_blk_mountid(struct block_mount_id *mid)spin_lock (&mid->bm_lock);while (!list_empty (&mid->bm_devlist)) { dev =list_first_entry (&mid->bm_devlist, - struct pnfs_block_dev, - bm_node); + struct pnfs_block_dev, bm_node);list_del (&dev->bm_node);free_block_dev (dev); }@@ -638,10 +870,11 @@ static void free_blk_mountid(struct block_mount_id*mid)
fs/nfs/blocklayout/blocklayout.c nfs4_blk_get_deviceinfo(struct nfs_server *server, const
Quoted 16 lines
fs/nfs/blocklayout/blocklayout.c +5
-4/* This is mostly copied from the filelayout's get_device_info function. * It seems much of this should be at the generic pnfs level. */ -static struct pnfs_block_dev * -nfs4_blk_get_deviceinfo (struct nfs_server *server,const struct nfs_fh *fh, -struct nfs4_deviceid *d_id, -struct list_head *sdlist) +static struct pnfs_block_dev *nfs4_blk_get_deviceinfo (struct nfs_server *server, +const struct nfs_fh *fh, +struct nfs4_deviceid + *d_id, +struct list_head *sdlist) {struct pnfs_device *dev;struct pnfs_block_dev *rv =NULL ;@@ -695,7 +928,7 @@ nfs4_blk_get_deviceinfo(struct nfs_server *server, conststruct
nfs_fh*fh,fs/nfs/blocklayout/blocklayout.c bl_set_layoutdriver(struct nfs_server *server, const struct
Quoted 9 lines
fs/nfs/blocklayout/blocklayout.c +1
-1goto out_free; rv =nfs4_blk_decode_device (server, dev, sdlist); - out_free: +out_free:if (dev->area !=NULL )vunmap (dev->area);for (i =0 ; i < max_pages; i++)@@ -748,8 +981,8 @@ bl_set_layoutdriver(struct nfs_server *server, const struct
nfs_fh*fh)fs/nfs/blocklayout/blocklayout.c bl_set_layoutdriver(struct nfs_server *server, const
Quoted 11 lines
fs/nfs/blocklayout/blocklayout.c +2
-2*/ for (i =0 ; i < dlist->num_devs; i++) { bdev =nfs4_blk_get_deviceinfo (server, fh, - &dlist->dev_id[i], - &block_disklist); + &dlist->dev_id[i], + &block_disklist);if (!bdev) { status = -ENODEV ; goto out_error;@@ -762,18 +995,17 @@ bl_set_layoutdriver(struct nfs_server *server, conststruct
nfs_fh*fh)fs/nfs/blocklayout/blocklayout.c static struct pnfs_layoutdriver_type blocklayout_type =
Quoted 293 lines
fs/nfs/blocklayout/blocklayout.c +5
-260dprintk ("%s SUCCESS\n" , __func__); server->pnfs_ld_data = b_mt_id; - out_return: +out_return:kfree (dlist);return status; - out_error: +out_error:free_blk_mountid (b_mt_id);kfree (mtype); goto out_return; } -static int -bl_clear_layoutdriver (struct nfs_server *server) +static int bl_clear_layoutdriver (struct nfs_server *server) {struct block_mount_id *b_mt_id = server->pnfs_ld_data;@@ -783,268 +1015,14 @@ bl_clear_layoutdriver(struct nfs_server *server)return 0 ; } -/* STUB - mark intersection of layout and page as bad, so is not -* used again. -*/ -static void mark_bad_read (void ) -{ -return ; -} - -/* Copied from buffer.c */ -static void __end_buffer_read_notouch (struct buffer_head *bh,int uptodate) -{ -if (uptodate) { -set_buffer_uptodate (bh); - }else { -/* This happens, due to failed READA attempts. */ -clear_buffer_uptodate (bh); - } -unlock_buffer (bh); -} - -/* Copied from buffer.c */ -static void end_buffer_read_nobh (struct buffer_head *bh,int uptodate) -{ -__end_buffer_read_notouch (bh, uptodate); -} - -/* -* map_block: map a requested I/0 block (isect) into an offset in the LVM -* meta block_device -*/ -static void -map_block (sector_t isect,struct pnfs_block_extent *be,struct buffer_head *bh) -{ -dprintk ("%s enter be=%p\n" , __func__, be); - -set_buffer_mapped (bh); - bh->b_bdev = be->be_mdev; - bh->b_blocknr = (isect - be->be_f_offset + be->be_v_offset) >> - (be->be_mdev->bd_inode->i_blkbits -9 ); - -dprintk ("%s isect %ld, bh->b_blocknr %ld, using bsize %Zd\n" , - __func__, (long )isect, - (long )bh->b_blocknr, - bh->b_size); -return ; -} - -/* Given an unmapped page, zero it (or read in page for COW), -* and set appropriate flags/markings, but it is safe to not initialize -* the range given in [from, to). -*/ -/* This is loosely based on nobh_write_begin */ -static int -init_page_for_write (struct pnfs_block_layout *bl,struct page *page, -unsigned from,unsigned to,sector_t **pages_to_mark) -{ -struct buffer_head *bh; -int inval, ret = -EIO ; -struct pnfs_block_extent *be =NULL , *cow_read =NULL ; -sector_t isect; - -dprintk ("%s enter, %p\n" , __func__, page); - bh =alloc_page_buffers (page,PAGE_CACHE_SIZE ,0 ); -if (!bh) { - ret = -ENOMEM ; - goto cleanup; - } - - isect = (sector_t )page->index << (PAGE_CACHE_SHIFT -9 ); - be =find_get_extent (bl, isect, &cow_read); -if (!be) - goto cleanup; - inval =is_hole (be, isect); -dprintk ("%s inval=%i, from=%u, to=%u\n" , __func__, inval, from, to); -if (inval) { -if (be->be_state ==PNFS_BLOCK_NONE_DATA ) { -dprintk ("%s PANIC - got NONE_DATA extent %p\n" , - __func__, be); - goto cleanup; - } -map_block (isect, be, bh); -unmap_underlying_metadata (bh->b_bdev, bh->b_blocknr); - } -if (PageUptodate (page)) { -/* Do nothing */ - }else if (inval & !cow_read) { -zero_user_segments (page,0 , from, to,PAGE_CACHE_SIZE ); - }else if (0 < from ||PAGE_CACHE_SIZE > to) { -struct pnfs_block_extent *read_extent; - - read_extent = (inval && cow_read) ? cow_read : be; -map_block (isect, read_extent, bh); -lock_buffer (bh); - bh->b_end_io = end_buffer_read_nobh; -submit_bh (READ , bh); -dprintk ("%s: Waiting for buffer read\n" , __func__); -/* XXX Don't really want to hold layout lock here */ -wait_on_buffer (bh); -if (!buffer_uptodate (bh)) - goto cleanup; - } -if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) { -/* There is a BUG here if is a short copy after write_begin, -* but I think this is a generic fs bug. The problem is that -* we have marked the page as initialized, but it is possible -* that the section not copied may never get copied. -*/ - ret =mark_initialized_sectors (be->be_inval, isect, -PAGE_CACHE_SECTORS , - pages_to_mark); -/* Want to preallocate mem so above can't fail */ -if (ret) - goto cleanup; - } -SetPageMappedToDisk (page); - ret =0 ; - -cleanup: -free_buffer_head (bh); -put_extent (be); -put_extent (cow_read); -if (ret) { -/* Need to mark layout with bad read...should now -* just use nfs4 for reads and writes. -*/ -mark_bad_read (); - } -return ret; -} - -static int -bl_write_begin (struct pnfs_layout_segment *lseg,struct page *page,loff_t pos, -unsigned count,struct pnfs_fsdata *fsdata) -{ -unsigned from, to; -int ret; -sector_t *pages_to_mark =NULL ; -struct pnfs_block_layout *bl =BLK_LSEG2EXT (lseg); - -dprintk ("%s enter, %u@%lld\n" , __func__, count, pos); -print_page (page); -/* The following code assumes blocksize >= PAGE_CACHE_SIZE */ -if (bl->bl_blocksize < (PAGE_CACHE_SIZE >>9 )) { -dprintk ("%s Can't handle blocksize %llu\n" , __func__, - (u64 )bl->bl_blocksize); -put_lseg (fsdata->lseg); - fsdata->lseg =NULL ; -return 0 ; - } -if (PageMappedToDisk (page)) { -/* Basically, this is a flag that says we have -* successfully called write_begin already on this page. -*/ -/* NOTE - there are cache consistency issues here. -* For example, what if the layout is recalled, then regained? -* If the file is closed and reopened, will the page flags -* be reset? If not, we'll have to use layout info instead of -* the page flag. -*/ -return 0 ; - } - from = pos & (PAGE_CACHE_SIZE -1 ); - to = from + count; - ret =init_page_for_write (bl, page, from, to, &pages_to_mark); -if (ret) { -dprintk ("%s init page failed with %i" , __func__, ret); -/* Revert back to plain NFS and just continue on with -* write. This assumes there is no request attached, which -* should be true if we get here. -*/ -BUG_ON (PagePrivate (page)); -put_lseg (fsdata->lseg); - fsdata->lseg =NULL ; -kfree (pages_to_mark); - ret =0 ; - }else { - fsdata->private = pages_to_mark; - } -return ret; -} - -/* CAREFUL - what happens if copied < count??? */ -static int -bl_write_end (struct inode *inode,struct page *page,loff_t pos, -unsigned count,unsigned copied,struct pnfs_layout_segment *lseg) -{ -dprintk ("%s enter, %u@%lld, lseg=%p\n" , __func__, count, pos, lseg); -print_page (page); -if (lseg) -SetPageUptodate (page); -return 0 ; -} - -/* Return any memory allocated to fsdata->private, and take advantage -* of no page locks to mark pages noted in write_begin as needing -* initialization. -*/ -static void -bl_write_end_cleanup (struct file *filp,struct pnfs_fsdata *fsdata) -{ -struct page *page; -pgoff_t index; -sector_t *pos; -struct address_space *mapping = filp->f_mapping; -struct pnfs_fsdata *fake_data; -struct pnfs_layout_segment *lseg; - -if (!fsdata) -return ; - lseg = fsdata->lseg; -if (!lseg) -return ; - pos = fsdata->private; -if (!pos) -return ; -dprintk ("%s enter with pos=%llu\n" , __func__, (u64)(*pos)); -for (; *pos != ~0 ; pos++) { - index = *pos >> (PAGE_CACHE_SHIFT -9 ); -/* XXX How do we properly deal with failures here??? */ - page =grab_cache_page_write_begin (mapping, index,0 ); -if (!page) { -printk (KERN_ERR "%s BUG BUG BUG NoMem\n" , __func__); -continue ; - } -dprintk ("%s: Examining block page\n" , __func__); -print_page (page); -if (!PageMappedToDisk (page)) { -/* XXX How do we properly deal with failures here??? */ -dprintk ("%s Marking block page\n" , __func__); -init_page_for_write (BLK_LSEG2EXT (fsdata->lseg), page, -PAGE_CACHE_SIZE ,PAGE_CACHE_SIZE , -NULL ); -print_page (page); - fake_data =kzalloc (sizeof (*fake_data),GFP_KERNEL ); -if (!fake_data) { -printk (KERN_ERR "%s BUG BUG BUG NoMem\n" , - __func__); -unlock_page (page); -continue ; - } -get_lseg (lseg); - fake_data->lseg = lseg; - fake_data->bypass_eof =1 ; - mapping->a_ops->write_end (filp, mapping, - index <<PAGE_CACHE_SHIFT , -PAGE_CACHE_SIZE , -PAGE_CACHE_SIZE , - page, fake_data); -/* Note fake_data is freed by nfs_write_end */ - }else -unlock_page (page); - } -kfree (fsdata->private); - fsdata->private =NULL ; -} -static const struct nfs_pageio_ops bl_pg_read_ops = { + .pg_init = pnfs_generic_pg_init_read, .pg_test = pnfs_generic_pg_test, .pg_doio = nfs_generic_pg_readpages, };static const struct nfs_pageio_ops bl_pg_write_ops = { + .pg_init = pnfs_generic_pg_init_write, .pg_test = pnfs_generic_pg_test, .pg_doio = nfs_generic_pg_writepages, };@@ -1054,9 +1032,6 @@ static struct pnfs_layoutdriver_type blocklayout_type ={
fs/nfs/file.c static int nfs_write_begin(struct file *file, struct
Quoted 24 lines
fs/nfs/blocklayout/blocklayout.c +1
-5.name = "LAYOUT_BLOCK_VOLUME", .read_pagelist = bl_read_pagelist, .write_pagelist = bl_write_pagelist, - .write_begin = bl_write_begin, - .write_end = bl_write_end, - .write_end_cleanup = bl_write_end_cleanup, .alloc_layout_hdr = bl_alloc_layout_hdr, .free_layout_hdr = bl_free_layout_hdr, .alloc_lseg = bl_alloc_lseg,@@ -1083,8 +1058,7 @@ static int __init nfs4blocklayout_init(void) static void __exit nfs4blocklayout_exit(void) { -dprintk ("%s: NFSv4 Block Layout Driver Unregistering...\n" , - __func__); +dprintk ("%s: NFSv4 Block Layout Driver Unregistering...\n" , __func__);pnfs_unregister_layoutdriver (&blocklayout_type);bl_pipe_exit ();fs/nfs/file.c +0
-0@@ -384,15 +384,12 @@ static int nfs_write_begin(struct file *file, struct
address_space*mapping,fs/nfs/file.c static int nfs_write_end(struct file *file, struct
Quoted 40 lines
fs/nfs/file.c +1
-13pgoff_t index = pos >>PAGE_CACHE_SHIFT ;struct page *page;int once_thru =0 ; -struct pnfs_layout_segment *lseg;dfprintk (PAGECACHE ,"NFS: write_begin(%s/%s(%ld), %u@%lld)\n" , file->f_path.dentry->d_parent->d_name.name, file->f_path.dentry->d_name.name, mapping->host->i_ino, len, (long long ) pos); - lseg =pnfs_update_layout (mapping->host, -nfs_file_open_context (file), - pos, len,IOMODE_RW ,GFP_NOFS ); + start:/* * Prevent starvation issues if someone is doing a consistency @@ -412,9 +409,6 @@ start:if (ret) {unlock_page (page);page_cache_release (page); - *pagep =NULL ; - *fsdata =NULL ; - goto out; }else if (!once_thru &&nfs_want_read_modify_write (file, page, pos, len)) { once_thru =1 ;@@ -423,12 +417,6 @@ start:if (!ret) goto start; } - ret =pnfs_write_begin (file, page, pos, len, lseg, fsdata); - out: -if (ret) { -put_lseg (lseg); - *fsdata =NULL ; - }return ret; }@@ -438,7 +426,6 @@ static int nfs_write_end(struct file *file, struct
address_space*mapping,fs/nfs/file.c static int nfs_write_end(struct file *file, struct
Quoted 8 lines
fs/nfs/file.c +0
-1{unsigned offset = pos & (PAGE_CACHE_SIZE -1 );int status; -struct pnfs_layout_segment *lseg; dfprintk(PAGECACHE, "NFS: write_end(%s/%s(%ld), %u@%lld)\n", file->f_path.dentry->d_parent->d_name.name,@@ -465,17 +452,10 @@ static int nfs_write_end(struct file *file, struct
address_space*mapping,fs/nfs/file.c static int nfs_vm_page_mkwrite(struct vm_area_struct
Quoted 19 lines
fs/nfs/file.c +1
-8zero_user_segment (page, pglen,PAGE_CACHE_SIZE ); } - lseg =nfs4_pull_lseg_from_fsdata (file, fsdata); - status =pnfs_write_end (file, page, pos, len, copied, lseg); -if (status) - goto out; - status =nfs_updatepage (file, page, offset, copied, lseg, fsdata); + status =nfs_updatepage (file, page, offset, copied); -out:unlock_page (page);page_cache_release (page); -pnfs_write_end_cleanup (file, fsdata); -put_lseg (lseg);if (status <0 )return status;@@ -597,7 +577,7 @@ static int nfs_vm_page_mkwrite(struct vm_area_struct*vma, struct
vm_fault*vmf)fs/nfs/pnfs.c pnfs_try_to_write_data(struct nfs_write_data
Quoted 13 lines
fs/nfs/file.c +1
-1ret =VM_FAULT_LOCKED ;if (nfs_flush_incompatible (filp, page) ==0 && -nfs_updatepage (filp, page,0 , pagelen,NULL ,NULL ) ==0 ) +nfs_updatepage (filp, page,0 , pagelen) ==0 ) goto out; ret =VM_FAULT_SIGBUS ;fs/nfs/pnfs.c +0
-0@@ -1223,41 +1223,6 @@ pnfs_try_to_write_data(struct nfs_write_data*wdata,
Quoted 72 lines
fs/nfs/pnfs.c +0
-41}/* -* This gives the layout driver an opportunity to read in page "around" -* the data to be written. It returns 0 on success, otherwise an error code -* which will either be passed up to user, or ignored if -* some previous part of write succeeded. -* Note the range [pos, pos+len-1] is entirely within the page. -*/ -int _pnfs_write_begin (struct inode *inode,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -struct pnfs_fsdata **fsdata) -{ -struct pnfs_fsdata *data; -int status =0 ; - -dprintk ("--> %s: pos=%llu len=%u\n" , - __func__, (unsigned long long )pos, len); - data =kzalloc (sizeof (struct pnfs_fsdata ),GFP_KERNEL ); -if (!data) { - status = -ENOMEM ; - goto out; - } - data->lseg = lseg;/* refcount passed into data to be managed there */ - status =NFS_SERVER (inode)->pnfs_curr_ld->write_begin ( - lseg, page, pos, len, data); -if (status) { -kfree (data); - data =NULL ; - } -out: - *fsdata = data; -dprintk ("<-- %s: status=%d\n" , __func__, status); -return status; -} - -/* * Called by non rpc-based layout drivers */ int @@ -1373,12 +1338,6 @@ void pnfs_cleanup_layoutcommit(struct inode *inode, data); } -void pnfs_free_fsdata (struct pnfs_fsdata *fsdata) -{ -/* lseg refcounting handled directly in nfs_write_end */ -kfree (fsdata); -} -/* * For the LAYOUT4_NFSV4_1_FILES layout type, NFS_DATA_SYNC WRITEs and * NFS_UNSTABLE WRITEs with a COMMIT to data servers must store enough fs/nfs/pnfs.h +0
-6@@ -54,12 +54,6 @@ enum pnfs_try_status { PNFS_NOT_ATTEMPTED = 1, }; -struct pnfs_fsdata { -struct pnfs_layout_segment *lseg; -int bypass_eof; -void *private; -}; -#ifdef CONFIG_NFS_V4_1 #define LAYOUT_NFSV4_1_MODULE_PREFIX "nfs-layouttype4"fs/nfs/pnfs.h struct pnfs_layoutdriver_type { · line 109
Quoted 3 lines
fs/nfs/pnfs.h +0
-0@@ -113,14 +107,6 @@ struct pnfs_layoutdriver_type {*/ enum pnfs_try_status (*read_pagelist) (struct nfs_read_data *nfs_data); enum pnfs_try_status (*write_pagelist) (struct nfs_write_data *nfs_data, inthow);
fs/nfs/pnfs.h enum pnfs_try_status pnfs_try_to_read_data(struct
Quoted 12 lines
fs/nfs/pnfs.h +0
-8-int (*write_begin) (struct pnfs_layout_segment *lseg,struct page *page, -loff_t pos,unsigned count, -struct pnfs_fsdata *fsdata); -int (*write_end)(struct inode *inode,struct page *page,loff_t pos, -unsigned count,unsigned copied, -struct pnfs_layout_segment *lseg); -void (*write_end_cleanup)(struct file *filp, -struct pnfs_fsdata *fsdata);void (*free_deviceid_node) (struct nfs4_deviceid_node *);@@ -196,7 +182,6 @@ enum pnfs_try_status pnfs_try_to_read_data(struct
nfs_read_data*,fs/nfs/pnfs.h enum pnfs_try_status pnfs_try_to_read_data(struct · line 182
Quoted 1 line
fs/nfs/pnfs.h +0
-0void pnfs_generic_pg_init_read(struct nfs_pageio_descriptor *, struct nfs_page*);
fs/nfs/pnfs.h enum pnfs_try_status pnfs_try_to_read_data(struct · line 183
Quoted 1 line
fs/nfs/pnfs.h +0
-0void pnfs_generic_pg_init_write(struct nfs_pageio_descriptor *, struct nfs_page*);
fs/nfs/pnfs.h enum pnfs_try_status pnfs_try_to_read_data(struct · line 184
Quoted 1 line
fs/nfs/pnfs.h +0
-0bool pnfs_generic_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page*prev, struct
nfs_page*req);fs/nfs/pnfs.h static inline void pnfs_clear_request_commit(struct
Quoted 16 lines
fs/nfs/pnfs.h +0
-5-void pnfs_free_fsdata (struct pnfs_fsdata *fsdata);int pnfs_layout_process (struct nfs4_layoutget *lgp);void pnfs_free_lseg_list (struct list_head *tmp_list);void pnfs_destroy_layout (struct nfs_inode *);@@ -208,10 +193,6 @@ void pnfs_set_layout_stateid(struct pnfs_layout_hdr *lo,int pnfs_choose_layoutget_stateid (nfs4_stateid *dst,struct pnfs_layout_hdr *lo,struct nfs4_state *open_state); -int _pnfs_write_begin (struct inode *inode,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -struct pnfs_fsdata **fsdata);int mark_matching_lsegs_invalid (struct pnfs_layout_hdr *lo,struct list_head *tmp_list,struct pnfs_layout_range *recall_range);@@ -329,13 +310,6 @@ static inline void pnfs_clear_request_commit(struct
nfs_page*req)fs/nfs/pnfs.h static inline void put_lseg(struct pnfs_layout_segment
Quoted 84 lines
fs/nfs/pnfs.h +0
-63put_lseg (req->wb_commit_lseg); } -static inline int pnfs_grow_ok (struct pnfs_layout_segment *lseg, -struct pnfs_fsdata *fsdata) -{ -return !fsdata || ((struct pnfs_layout_segment *)fsdata == lseg) || - !fsdata->bypass_eof; -} -/* Should the pNFS client commit and return the layout upon a setattr */ static inline bool pnfs_ld_layoutret_on_setattr (struct inode *inode)@@ -346,49 +320,6 @@ pnfs_ld_layoutret_on_setattr(struct inode *inode)PNFS_LAYOUTRET_ON_SETATTR ; } -static inline int pnfs_write_begin (struct file *filp,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -void **fsdata) -{ -struct inode *inode = filp->f_dentry->d_inode; -struct nfs_server *nfss =NFS_SERVER (inode); -int status =0 ; - - *fsdata = lseg; -if (lseg && nfss->pnfs_curr_ld->write_begin) - status =_pnfs_write_begin (inode, page, pos, len, lseg, - (struct pnfs_fsdata **) fsdata); -return status; -} - -/* CAREFUL - what happens if copied < len??? */ -static inline int pnfs_write_end (struct file *filp,struct page *page, -loff_t pos,unsigned len,unsigned copied, -struct pnfs_layout_segment *lseg) -{ -struct inode *inode = filp->f_dentry->d_inode; -struct nfs_server *nfss =NFS_SERVER (inode); - -if (nfss->pnfs_curr_ld && nfss->pnfs_curr_ld->write_end) -return nfss->pnfs_curr_ld->write_end (inode, page, pos, len, - copied, lseg); -else -return 0 ; -} - -static inline void pnfs_write_end_cleanup (struct file *filp,void *fsdata) -{ -struct nfs_server *nfss =NFS_SERVER (filp->f_dentry->d_inode); - -if (fsdata && nfss->pnfs_curr_ld) { -if (nfss->pnfs_curr_ld->write_end_cleanup) - nfss->pnfs_curr_ld->write_end_cleanup (filp, fsdata); -if (nfss->pnfs_curr_ld->write_begin) -pnfs_free_fsdata (fsdata); - } -} -static inline int pnfs_return_layout (struct inode *ino) {struct nfs_inode *nfsi =NFS_I (ino);@@ -400,19 +331,6 @@ static inline int pnfs_return_layout(struct inode *ino) return 0; } -static inline struct pnfs_layout_segment * -nfs4_pull_lseg_from_fsdata (struct file *filp,void *fsdata) -{ -if (fsdata) { -struct nfs_server *nfss =NFS_SERVER (filp->f_dentry->d_inode); - -if (nfss->pnfs_curr_ld && nfss->pnfs_curr_ld->write_begin) -return ((struct pnfs_fsdata *) fsdata)->lseg; -return (struct pnfs_layout_segment *)fsdata; - } -return NULL ; -} -#else /* CONFIG_NFS_V4_1 */ static inline void pnfs_destroy_all_layouts (struct nfs_client *clp)@@ -433,12 +351,6 @@ static inline void put_lseg(struct pnfs_layout_segment*lseg)
fs/nfs/pnfs.h static inline int pnfs_layoutcommit_inode(struct inode
Quoted 40 lines
fs/nfs/pnfs.h +0
-26{ } -static inline int pnfs_grow_ok (struct pnfs_layout_segment *lseg, -struct pnfs_fsdata *fsdata) -{ -return 1 ; -} -static inline enum pnfs_try_status pnfs_try_to_read_data (struct nfs_read_data *data,const struct rpc_call_ops *call_ops)@@ -458,26 +370,6 @@ static inline int pnfs_return_layout(struct inode *ino)return 0 ; } -static inline int pnfs_write_begin (struct file *filp,struct page *page, -loff_t pos,unsigned len, -struct pnfs_layout_segment *lseg, -void **fsdata) -{ - *fsdata =NULL ; -return 0 ; -} - -static inline int pnfs_write_end (struct file *filp,struct page *page, -loff_t pos,unsigned len,unsigned copied, -struct pnfs_layout_segment *lseg) -{ -return 0 ; -} - -static inline void pnfs_write_end_cleanup (struct file *filp,void *fsdata) -{ -} -static inline bool pnfs_ld_layoutret_on_setattr (struct inode *inode) {@@ -554,13 +446,6 @@ static inline int pnfs_layoutcommit_inode(struct inode*inode, bool sync)
fs/nfs/write.c static int nfs_writepage_setup(struct nfs_open_context
Quoted 29 lines
fs/nfs/pnfs.h +0
-7static inline void nfs4_deviceid_purge_client (struct nfs_client *ncl) { } - -static inline struct pnfs_layout_segment * -nfs4_pull_lseg_from_fsdata (struct file *filp,void *fsdata) -{ -return NULL ; -} -#endif /* CONFIG_NFS_V4_1 */ #endif /* FS_NFS_PNFS_H */ fs/nfs/write.c +1
-3@@ -673,9 +673,7 @@ out: }static int nfs_writepage_setup (struct nfs_open_context *ctx,struct page *page, -unsigned int offset,unsigned int count, -struct pnfs_layout_segment *lseg,void *fsdata) - +unsigned int offset,unsigned int count) {struct nfs_page *req;@@ -683,8 +681,7 @@ static int nfs_writepage_setup(struct nfs_open_context*ctx, struct page *page,
fs/nfs/write.c static int nfs_write_pageuptodate(struct page *page,
Quoted 10 lines
fs/nfs/write.c +1
-2if (IS_ERR (req))return PTR_ERR (req);/* Update file length */ -if (pnfs_grow_ok (lseg, fsdata)) -nfs_grow_file (page, offset, count); +nfs_grow_file (page, offset, count);nfs_mark_uptodate (page, req->wb_pgbase, req->wb_bytes);nfs_mark_request_dirty (req);nfs_clear_page_tag_locked (req);@@ -737,8 +734,7 @@ static int nfs_write_pageuptodate(struct page *page,struct inode *inode)
Quoted 32 lines
fs/nfs/write.c +2
-3* things with a page scheduled for an RPC call (e.g. invalidate it). */ int nfs_updatepage (struct file *file,struct page *page, -unsigned int offset,unsigned int count, -struct pnfs_layout_segment *lseg,void *fsdata) +unsigned int offset,unsigned int count) {struct nfs_open_context *ctx =nfs_file_open_context (file);struct inode *inode = page->mapping->host;@@ -763,7 +759,7 @@ int nfs_updatepage(struct file *file, struct page *page, offset =0 ; } - status =nfs_writepage_setup (ctx, page, offset, count, lseg, fsdata); + status =nfs_writepage_setup (ctx, page, offset, count);if (status <0 )nfs_set_pageerror (page);include/linux/nfs_fs.h +1
-2@@ -510,8 +510,7 @@ extern int nfs_congestion_kb;extern int nfs_writepage (struct page *page,struct writeback_control *wbc);extern int nfs_writepages (struct address_space *,struct writeback_control *);extern int nfs_flush_incompatible (struct file *file,struct page *page); -extern int nfs_updatepage (struct file *,struct page *,unsigned int , -unsigned int ,struct pnfs_layout_segment *,void *); +extern int nfs_updatepage (struct file *,struct page *,unsigned int ,unsigned int );extern void nfs_writeback_done (struct rpc_task *,struct nfs_write_data *);/*
Re: [PATCH 1/6] SQUASHME: pnfs-block: Remove write_begin/end hooks
Quoted 4 lines · [email protected] wrote:
Quoted 2 lines
+ *
map_block: map a requested I/0 block (isect) into an offset in the LVM
+ * metablock_deviceWhat's a "meta
block_device"?
We can remove the *meta* if you don't like the name.
I've been calling this the "mapped device." It's produced by blkmapd, which is short for "block map daemon." In the kernel it's consistently called the "meta device." I'm not crazy about the name either but it needs to be consistent. Should I change it?
I just noticed too that there are some outdated comments left over from when the mapping was done in the kernel:
/* For each device returned in dlist, call GETDEVICEINFO, and
* decode the opaque topology encoding to create a flat
* volume topology, matching VOLUME_SIMPLE disk signatures
* to disks in the visible block disk list.
* Construct an LVM meta device from the flat volume topology. */
What's really going on here is the devices are just being added to a list. The device is constructed later by the daemon. I will fix this comment.
PATCH 2/6 SQUASHME: pnfs-block: skip sectors already initialized
In reply to Jim ReesFrom: Peng Tao <[email protected]>
No need to test PageUptodate when checking for old pages.
Signed-off-by Peng Tao <[email protected]>
--- fs/nfs/blocklayout/blocklayout.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)
fs/nfs/blocklayout/blocklayout.c +4 -4
@@ -596,11 +596,11 @@ fill_invalid_ext:/* PageDirty: Other will write this out * PageWriteback: Other is writing this out -* PageUptodate && sector_initialized: already written out +* PageUptodate: It was read before +* sector_initialized: already written out */ if (PageDirty (page) ||PageWriteback (page) || - (PageUptodate (page) && -is_sector_initialized (be->be_inval, isect))) { +is_sector_initialized (be->be_inval, isect)) { dprintk ("page %lu is uptodate %d dirty %d writeback %d\n", page->index, PageUptodate(page),
@@ -612,7 +612,7 @@ fill_invalid_ext: extent_length -=PAGE_CACHE_SIZE >>9 ;continue ; } -if (!PageUptodate (page) && cow_read) { +if (!PageUptodate (page)) {/* New page, readin or zero it */ init_page_for_write (page, cow_read); }
Signature · 1 line
1.7.4.1
PATCH 3/6 SQUASHME: pnfs: teach layoutcommit handle multiple segments
In reply to Jim ReesFrom: Peng Tao <[email protected]>
This rewrites commit 9658e7532d35faf0c09b8b34fd5808e5e58070ac in a way more reasonable. Layoutcommit takes a list of segments and last write offset is saved at inode level.
Signed-off-by Peng Tao <[email protected]>
--- fs/nfs/nfs4filelayout.c | 2 +- fs/nfs/nfs4proc.c | 8 +++++- fs/nfs/pnfs.c | 60 ++++++++++++++++++---------------------------- fs/nfs/pnfs.h | 4 ++- include/linux/nfs_xdr.h | 2 +- 5 files changed, 36 insertions(+), 40 deletions(-)
fs/nfs/nfs4filelayout.c +1 -1
@@ -170,7 +170,7 @@ filelayout_set_layoutcommit(struct nfs_write_data *wdata)pnfs_set_layoutcommit (wdata);dprintk ("%s ionde %lu pls_end_pos %lu\n" , __func__, wdata->inode->i_ino, - (unsigned long ) wdata->lseg->pls_end_pos); + (unsigned long )NFS_I (wdata->inode)->layout->plh_lwb); }/*
fs/nfs/nfs4proc.c +7 -1
@@ -5963,10 +5963,16 @@ nfs4_layoutcommit_done(struct rpc_task *task, void *calldata)static void nfs4_layoutcommit_release (void *calldata) {struct nfs4_layoutcommit_data *data = calldata; +struct pnfs_layout_segment *lseg, *tmp;pnfs_cleanup_layoutcommit (data->args.inode, data);/* Matched by references in pnfs_set_layoutcommit */ -put_lseg (data->lseg); +list_for_each_entry_safe (lseg, tmp, &data->lseg_list, pls_lc_list) { +list_del_init (&lseg->pls_lc_list); +if (test_and_clear_bit (NFS_LSEG_LAYOUTCOMMIT , + &lseg->pls_flags)) +put_lseg (lseg); + }put_rpccred (data->cred);kfree (data); }
fs/nfs/pnfs.c +24 -36
112 lines not on this page · raw message
fs/nfs/pnfs.h +3 -1
@@ -41,12 +41,12 @@ enum {struct pnfs_layout_segment {struct list_head pls_list; +struct list_head pls_lc_list;struct pnfs_layout_range pls_range;atomic_t pls_refcount;unsigned long pls_flags;struct pnfs_layout_hdr *pls_layout;struct rpc_cred *pls_lc_cred;/* LAYOUTCOMMIT credential */ -loff_t pls_end_pos;/* LAYOUTCOMMIT write end */ };enum pnfs_try_status {
@@ -132,6 +132,8 @@ struct pnfs_layout_hdr {unsigned long plh_block_lgets;/* block LAYOUTGET if >0 */ u32 plh_barrier;/* ignore lower seqids */ unsigned long plh_flags; +loff_t plh_lwb;/* last write byte for layoutcommit */ +struct rpc_cred *plh_lc_cred;/* layoutcommit cred */ struct inode *plh_inode; };
include/linux/nfs_xdr.h +1 -1
@@ -275,7 +275,7 @@ struct nfs4_layoutcommit_res {struct nfs4_layoutcommit_data {struct rpc_task task;struct nfs_fattr fattr; -struct pnfs_layout_segment *lseg; +struct list_head lseg_list;struct rpc_cred *cred;struct nfs4_layoutcommit_args args;struct nfs4_layoutcommit_res res;
Signature · 1 line
1.7.4.1
PATCH 4/6 get rid of deprecated xdr macros
In reply to Jim ReesSigned-off-by Jim Rees <[email protected]>
--- fs/nfs/blocklayout/blocklayout.h | 46 +---------------------------------- fs/nfs/blocklayout/blocklayoutdev.c | 30 +++++++++++++++++----- fs/nfs/blocklayout/extents.c | 10 ++++---- 3 files changed, 29 insertions(+), 57 deletions(-)
fs/nfs/blocklayout/blocklayout.h +1 -45
@@ -193,52 +193,8 @@ BLK_LSEG2EXT(struct pnfs_layout_segment *lseg)return BLK_LO2EXT (lseg->pls_layout); } -uint32_t *blk_overflow (uint32_t *p,uint32_t *end,size_t nbytes); - -#define BLK_READBUF (p, e, nbytes) do { \ - p = blk_overflow(p, e, nbytes); \ - if (!p) { \ - printk(KERN_WARNING \ - "%s: reply buffer overflowed in line %d.\n", \ - __func__, __LINE__); \ - goto out_err; \ - } \ -} while (0) - -#define READ32 (x) (x) = ntohl(*p++) -#define READ64 (x) do { \ - (x) = (uint64_t)ntohl(*p++) << 32; \ - (x) |= ntohl(*p++); \ -} while (0) -#define COPYMEM (x, nbytes) do { \ - memcpy((x), p, nbytes); \ - p += XDR_QUADLEN(nbytes); \ -} while (0) -#define READ_DEVID (x) COPYMEM((x)->data, NFS4_DEVICEID4_SIZE) -#define READ_SECTOR (x) do { \ - READ64(tmp); \ - if (tmp & 0x1ff) { \ - printk(KERN_WARNING \ - "%s Value not 512-byte aligned at line %d\n", \ - __func__, __LINE__); \ - goto out_err; \ - } \ - (x) = tmp >> 9; \ -} while (0) - -#define WRITE32 (n) do { \ - *p++ = htonl(n); \ - } while (0) -#define WRITE64 (n) do { \ - *p++ = htonl((uint32_t)((n) >> 32)); \ - *p++ = htonl((uint32_t)(n)); \ -} while (0) -#define WRITEMEM (ptr, nbytes) do { \ - p = xdr_encode_opaque_fixed(p, ptr, nbytes); \ -} while (0) -#define WRITE_DEVID (x) WRITEMEM((x)->data, NFS4_DEVICEID4_SIZE) -/* blocklayoutdev.c */ +uint32_t *blk_overflow (uint32_t *p,uint32_t *end,size_t nbytes);struct block_device *nfs4_blkdev_get (dev_t dev);int nfs4_blkdev_put (struct block_device *bdev); struct pnfs_block_dev *nfs4_blk_decode_device(struct nfs_server *server,
fs/nfs/blocklayout/blocklayoutdev.c +23 -7
@@ -49,6 +49,19 @@ uint32_t *blk_overflow(uint32_t *p, uint32_t *end, size_t nbytes) }EXPORT_SYMBOL (blk_overflow); +static int decode_sector_number (__be32 **rp,sector_t *sp) +{ +uint64_t s; + + *rp =xdr_decode_hyper (*rp, &s); +if (s &0x1ff ) { +printk (KERN_WARNING "%s: sector not aligned\n" , __func__); +return -1 ; + } + *sp = s >>9 ; +return 0 ; +} +/* Open a block_device by device number. */ struct block_device *nfs4_blkdev_get (dev_t dev) {
@@ -241,7 +254,6 @@ nfs4_blk_process_layoutget(struct pnfs_layout_hdr *lo,struct xdr_buf buf;struct page *scratch;__be32 *p; -uint64_t tmp;/* Used by READSECTOR */ struct layout_verification lv = { .mode = lgr->range.iomode, .start = lgr->range.offset >>9 ,
@@ -263,7 +275,7 @@ nfs4_blk_process_layoutget(struct pnfs_layout_hdr *lo,if (unlikely (!p)) goto out_err; -READ32 (count); + count =be32_to_cpup (p++);dprintk ("%s enter, number of extents %i\n" , __func__, count); p =xdr_inline_decode (&stream, (28 +NFS4_DEVICEID4_SIZE ) * count);
@@ -280,7 +292,8 @@ nfs4_blk_process_layoutget(struct pnfs_layout_hdr *lo, status = -ENOMEM ; goto out_err; } -READ_DEVID (&be->be_devid); +memcpy (&be->be_devid, p,NFS4_DEVICEID4_SIZE ); + p +=XDR_QUADLEN (NFS4_DEVICEID4_SIZE ); be->be_mdev =translate_devid (lo, &be->be_devid);if (!be->be_mdev) goto out_err;
@@ -288,10 +301,13 @@ nfs4_blk_process_layoutget(struct pnfs_layout_hdr *lo,/* The next three values are read in as bytes, * but stored as 512-byte sector lengths */ -READ_SECTOR (be->be_f_offset); -READ_SECTOR (be->be_length); -READ_SECTOR (be->be_v_offset); -READ32 (be->be_state); +if (decode_sector_number (&p, &be->be_f_offset) <0 ) + goto out_err; +if (decode_sector_number (&p, &be->be_length) <0 ) + goto out_err; +if (decode_sector_number (&p, &be->be_v_offset) <0 ) + goto out_err; + be->be_state =be32_to_cpup (p++);if (be->be_state ==PNFS_BLOCK_INVALID_DATA ) be->be_inval = &bl->bl_inval;if (verify_extent (be, &lv)) {
fs/nfs/blocklayout/extents.c +5 -5
@@ -761,11 +761,11 @@ encode_pnfs_block_layoutupdate(struct pnfs_block_layout *bl, p =xdr_reserve_space (xdr,7 *4 +sizeof (lce->bse_devid.data));if (!p)break ; -WRITE_DEVID (&lce->bse_devid); -WRITE64 (lce->bse_f_offset <<9 ); -WRITE64 (lce->bse_length <<9 ); -WRITE64 (0LL ); -WRITE32 (PNFS_BLOCK_READWRITE_DATA ); + p =xdr_encode_opaque_fixed (p, lce->bse_devid.data,NFS4_DEVICEID4_SIZE ); + p =xdr_encode_hyper (p, lce->bse_f_offset <<9 ); + p =xdr_encode_hyper (p, lce->bse_length <<9 ); + p =xdr_encode_hyper (p,0LL ); + *p++ =cpu_to_be32 (PNFS_BLOCK_READWRITE_DATA );list_del (&lce->bse_node);list_add_tail (&lce->bse_node, ranges); bl->bl_count--;
Signature · 1 line
1.7.4.1
PATCH 5/6 reindent
In reply to Jim ReesSigned-off-by Jim Rees <[email protected]>
--- fs/nfs/blocklayout/extents.c | 107 +++++++++++++++++++++--------------------- 1 files changed, 53 insertions(+), 54 deletions(-)
fs/nfs/blocklayout/extents.c +53 -54
@@ -43,7 +43,7 @@/* Returns largest t<=s s.t. t%base==0 */ static inline sector_t normalize (sector_t s,int base) { -sector_t tmp = s;/* Since do_div modifies its argument */ +sector_t tmp = s;/* Since do_div modifies its argument */ return s -do_div (tmp, base); }
@@ -71,8 +71,7 @@ static int32_t _find_entry(struct my_tree_t *tree, u64 s)return -ENOENT ; } -static inline -int _has_tag (struct my_tree_t *tree,u64 s,int32_t tag) +static inline int _has_tag (struct my_tree_t *tree,u64 s,int32_t tag) {int32_t tags;
@@ -169,7 +168,7 @@ static int _preload_range(struct my_tree_t *tree, u64 offset, u64 length)/* Unlock - HOW??? */ status =0 ; - out_cleanup: +out_cleanup:for (i = used; i < count; i++) {if (!storage[i])break ;
@@ -179,7 +178,7 @@ static int _preload_range(struct my_tree_t *tree, u64 offset, u64 length)return status; } -static void set_needs_init (sector_t *array,sector_t offset) +static void set_needs_init (sector_t * array,sector_t offset) {sector_t *p = array;
@@ -196,7 +195,7 @@ static void set_needs_init(sector_t *array, sector_t offset)return ; }else {sector_t *save = p; -dprintk ("%s Adding %llu\n" , __func__, (u64 )offset); +dprintk ("%s Adding %llu\n" , __func__, (u64 ) offset);while (*p != ~0 ) p++; p++;
@@ -232,7 +231,8 @@ _range_has_tag(struct my_tree_t *tree, u64 start, u64 end, int32_t tag)if ((pos->it_sector == end - tree->mtt_step_size) && (pos->it_tags & (1 << tag))) { expect = pos->it_sector - tree->mtt_step_size; -if (pos->it_sector < tree->mtt_step_size || expect < start) +if (pos->it_sector < tree->mtt_step_size + || expect < start)return 1 ;continue ; }else {
@@ -270,13 +270,13 @@ int mark_initialized_sectors(struct pnfs_inval_markings *marks, sector_t **pages) {sector_t s, start, end; -sector_t *array =NULL ;/* Pages to mark */ +sector_t *array =NULL ;/* Pages to mark */ dprintk ("%s(offset=%llu,len=%llu) enter\n" , - __func__, (u64 )offset, (u64 )length); + __func__, (u64 ) offset, (u64 ) length); s =max ((sector_t )3 ,2 * (marks->im_block_size / (PAGE_CACHE_SECTORS ))); -dprintk ("%s set max=%llu\n" , __func__, (u64 )s); +dprintk ("%s set max=%llu\n" , __func__, (u64 ) s);if (pages) { array =kmalloc (s *sizeof (sector_t),GFP_KERNEL );if (!array)
@@ -318,9 +318,9 @@ int mark_initialized_sectors(struct pnfs_inval_markings *marks, }return 0 ; - out_unlock: +out_unlock:spin_unlock (&marks->im_lock); - outerr: +outerr:if (pages) {kfree (array); *pages =NULL ;
@@ -337,7 +337,7 @@ int mark_written_sectors(struct pnfs_inval_markings *marks,int status;dprintk ("%s(offset=%llu,len=%llu) enter\n" , __func__, - (u64 )offset, (u64 )length); + (u64 ) offset, (u64 ) length);spin_lock (&marks->im_lock); status =_set_range (&marks->im_tree,EXTENT_WRITTEN , offset, length);spin_unlock (&marks->im_lock);
@@ -348,8 +348,8 @@ static void print_short_extent(struct pnfs_block_short_extent *be) {dprintk ("PRINT SHORT EXTENT extent %p\n" , be);if (be) { -dprintk (" be_f_offset %llu\n" , (u64 )be->bse_f_offset); -dprintk (" be_length %llu\n" , (u64 )be->bse_length); +dprintk (" be_f_offset %llu\n" , (u64 ) be->bse_f_offset); +dprintk (" be_length %llu\n" , (u64 ) be->bse_length); } }
@@ -396,12 +396,12 @@ static void add_to_commitlist(struct pnfs_block_layout *bl,kfree (new);return ; }else if (new->bse_f_offset <= - old->bse_f_offset + old->bse_length) { + old->bse_f_offset + old->bse_length) {/* new overlaps or abuts existing be */ if (new->bse_mdev == old->bse_mdev) {/* extend new to fully replace old */ new->bse_length += new->bse_f_offset - - old->bse_f_offset; + old->bse_f_offset; new->bse_f_offset = old->bse_f_offset;list_del (&old->bse_node); bl->bl_count--;
@@ -493,15 +493,14 @@ static void print_bl_extent(struct pnfs_block_extent *be) {dprintk ("PRINT EXTENT extent %p\n" , be);if (be) { -dprintk (" be_f_offset %llu\n" , (u64 )be->be_f_offset); -dprintk (" be_length %llu\n" , (u64 )be->be_length); -dprintk (" be_v_offset %llu\n" , (u64 )be->be_v_offset); +dprintk (" be_f_offset %llu\n" , (u64 ) be->be_f_offset); +dprintk (" be_length %llu\n" , (u64 ) be->be_length); +dprintk (" be_v_offset %llu\n" , (u64 ) be->be_v_offset);dprintk (" be_state %d\n" , be->be_state); } } -static void -destroy_extent (struct kref *kref) +static void destroy_extent (struct kref *kref) {struct pnfs_block_extent *be;
@@ -510,8 +509,7 @@ destroy_extent(struct kref *kref)kfree (be); } -void -put_extent (struct pnfs_block_extent *be) +void put_extent (struct pnfs_block_extent *be) {if (be) { dprintk("%s enter %p (%i)\n", __func__, be,
@@ -533,8 +531,7 @@ struct pnfs_block_extent *alloc_extent(void)return be; } -struct pnfs_block_extent * -get_extent (struct pnfs_block_extent *be) +struct pnfs_block_extent *get_extent (struct pnfs_block_extent *be) {if (be)kref_get (&be->be_refcnt);
@@ -557,10 +554,10 @@ extents_consistent(struct pnfs_block_extent *old, struct pnfs_block_extent *new) {/* Note this assumes new->be_f_offset >= old->be_f_offset */ return (new->be_state == old->be_state) && - ((new->be_state ==PNFS_BLOCK_NONE_DATA ) || - ((new->be_v_offset - old->be_v_offset == - new->be_f_offset - old->be_f_offset) && - new->be_mdev == old->be_mdev)); + ((new->be_state ==PNFS_BLOCK_NONE_DATA ) || + ((new->be_v_offset - old->be_v_offset == + new->be_f_offset - old->be_f_offset) && + new->be_mdev == old->be_mdev)); }/* Adds new to appropriate list in bl, modifying new and removing existing
@@ -594,7 +591,7 @@ add_and_merge_extent(struct pnfs_block_layout *bl,break ;if (new->be_f_offset >= be->be_f_offset) {if (end <= be->be_f_offset + be->be_length) { -/* new is a subset of existing be*/ +/* new is a subset of existing be */ if (extents_consistent (be, new)) {dprintk ("%s: new is subset, ignoring\n" , __func__);
@@ -609,10 +606,11 @@ add_and_merge_extent(struct pnfs_block_layout *bl,if (extents_consistent (be, new)) {/* extend new to fully replace be */ new->be_length += new->be_f_offset - - be->be_f_offset; + be->be_f_offset; new->be_f_offset = be->be_f_offset; new->be_v_offset = be->be_v_offset; -dprintk ("%s: removing %p\n" , __func__, be); +dprintk ("%s: removing %p\n" , __func__, + be);list_del (&be->be_node);put_extent (be); }else {
@@ -634,8 +632,9 @@ add_and_merge_extent(struct pnfs_block_layout *bl,*|<-- new -->| */ if (extents_consistent (new, be)) {/* extend new to fully replace be */ - new->be_length += be->be_f_offset + be->be_length - - new->be_f_offset - new->be_length; + new->be_length += + be->be_f_offset + be->be_length - + new->be_f_offset - new->be_length;dprintk ("%s: removing %p\n" , __func__, be);list_del (&be->be_node);put_extent (be);
@@ -655,7 +654,7 @@ add_and_merge_extent(struct pnfs_block_layout *bl,*/ return 0 ; - out_err: +out_err:put_extent (new);return -EIO ; }
@@ -668,14 +667,14 @@ add_and_merge_extent(struct pnfs_block_layout *bl,* 1. Extents are ordered by file offset. * 2. For any given isect, there is at most one extents that matches. */ -struct pnfs_block_extent * -find_get_extent (struct pnfs_block_layout *bl,sector_t isect, -struct pnfs_block_extent **cow_read) +struct pnfs_block_extent *find_get_extent (struct pnfs_block_layout *bl, +sector_t isect, +struct pnfs_block_extent **cow_read) {struct pnfs_block_extent *be, *cow, *ret;int i; -dprintk ("%s enter with isect %llu\n" , __func__, (u64 )isect); +dprintk ("%s enter with isect %llu\n" , __func__, (u64 ) isect); cow = ret =NULL ;spin_lock (&bl->bl_ext_lock);for (i =0 ; i <EXTENT_LISTS ; i++) {
@@ -708,13 +707,13 @@ find_get_extent(struct pnfs_block_layout *bl, sector_t isect, }/* Similar to find_get_extent, but called with lock held, and ignores cow */ -static struct pnfs_block_extent * -find_get_extent_locked (struct pnfs_block_layout *bl,sector_t isect) +static struct pnfs_block_extent *find_get_extent_locked (struct pnfs_block_layout + *bl,sector_t isect) {struct pnfs_block_extent *be, *ret =NULL ;int i; -dprintk ("%s enter with isect %llu\n" , __func__, (u64 )isect); +dprintk ("%s enter with isect %llu\n" , __func__, (u64 ) isect);for (i =0 ; i <EXTENT_LISTS ; i++) {if (ret)break ;
@@ -761,7 +760,8 @@ encode_pnfs_block_layoutupdate(struct pnfs_block_layout *bl, p =xdr_reserve_space (xdr,7 *4 +sizeof (lce->bse_devid.data));if (!p)break ; - p =xdr_encode_opaque_fixed (p, lce->bse_devid.data,NFS4_DEVICEID4_SIZE ); + p =xdr_encode_opaque_fixed (p, lce->bse_devid.data, +NFS4_DEVICEID4_SIZE ); p =xdr_encode_hyper (p, lce->bse_f_offset <<9 ); p =xdr_encode_hyper (p, lce->bse_length <<9 ); p =xdr_encode_hyper (p,0LL );
@@ -799,9 +799,9 @@ _prep_new_extent(struct pnfs_block_extent *new,/* Tries to merge be with extent in front of it in list. * Frees storage if not used. */ -static struct pnfs_block_extent * -_front_merge (struct pnfs_block_extent *be,struct list_head *head, -struct pnfs_block_extent *storage) +static struct pnfs_block_extent *_front_merge (struct pnfs_block_extent *be, +struct list_head *head, +struct pnfs_block_extent *storage) {struct pnfs_block_extent *prev;
@@ -821,13 +821,12 @@ _front_merge(struct pnfs_block_extent *be, struct list_head *head,put_extent (be);return storage; - no_merge: +no_merge:kfree (storage);return be; } -static u64 -set_to_rw (struct pnfs_block_layout *bl,u64 offset,u64 length) +static u64 set_to_rw (struct pnfs_block_layout *bl,u64 offset,u64 length) {u64 rv = offset + length;struct pnfs_block_extent *be, *e1, *e2, *e3, *new, *old;
@@ -868,8 +867,7 @@ set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length)if (offset + length < be->be_f_offset + be->be_length) {_prep_new_extent (e3, be, e2->be_f_offset + e2->be_length, be->be_f_offset + be->be_length - - offset - length, -PNFS_BLOCK_INVALID_DATA ); + offset - length,PNFS_BLOCK_INVALID_DATA ); children[i++] = e3;print_bl_extent (e3); }else
@@ -905,7 +903,7 @@ set_to_rw(struct pnfs_block_layout *bl, u64 offset, u64 length)dprintk ("%s returns %llu after split\n" , __func__, rv);return rv; - out_nosplit: +out_nosplit:kfree (e1);kfree (e2);kfree (e3);
@@ -921,7 +919,8 @@ clean_pnfs_block_layoutupdate(struct pnfs_block_layout *bl,struct pnfs_block_short_extent *lce, *save;dprintk ("%s status %d\n" , __func__, status); -list_for_each_entry_safe_reverse (lce, save, &bl->bl_committing, bse_node) { +list_for_each_entry_safe_reverse (lce, save, &bl->bl_committing, + bse_node) {if (likely (!status)) {u64 offset = lce->bse_f_offset;u64 end = offset + lce->bse_length;
Signature · 1 line
1.7.4.1
PATCH 6/6 pnfs-block: mark IO error with NFS_LAYOUT_{RW|RO}_FAILED
In reply to Jim ReesFrom: Peng Tao <[email protected]>
Signed-off-by Peng Tao <[email protected]>
--- fs/nfs/blocklayout/blocklayout.c | 90 ++++++++++++++----------------------- fs/nfs/blocklayout/blocklayout.h | 5 -- 2 files changed, 34 insertions(+), 61 deletions(-)
fs/nfs/blocklayout/blocklayout.c +34 -56
@@ -143,35 +143,40 @@ static struct bio *bl_submit_bio(int rw, struct bio *bio)return NULL ; } -static inline void bl_done_with_rpage (struct page *page,const int ok) +static void bl_set_lo_fail (struct pnfs_layout_segment *lseg) { -if (ok) { -ClearPagePnfsErr (page); -SetPageUptodate (page); +if (lseg->pls_range.iomode ==IOMODE_RW ) { +dprintk ("%s Setting layout IOMODE_RW fail bit\n" , __func__); +set_bit (lo_fail_bit (IOMODE_RW ), &lseg->pls_layout->plh_flags); }else { -ClearPageUptodate (page); -SetPageError (page); -SetPagePnfsErr (page); +dprintk ("%s Setting layout IOMODE_READ fail bit\n" , __func__); +set_bit (lo_fail_bit (IOMODE_READ ), &lseg->pls_layout->plh_flags); } -/* Page is unlocked via rpc_release. Should really be done here. */ }/* This is basically copied from mpage_end_io_read */ static void bl_end_io_read (struct bio *bio,int err) { -void *data = bio->bi_private; +struct parallel_io *par = bio->bi_private;const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags);struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ; +struct nfs_read_data *rdata = (struct nfs_read_data *)par->data;do {struct page *page = bvec->bv_page;if (--bvec >= bio->bi_io_vec)prefetchw (&bvec->bv_page->flags); -bl_done_with_rpage (page, uptodate); +if (uptodate) +SetPageUptodate (page); }while (bvec >= bio->bi_io_vec); +if (!uptodate) { +if (!rdata->pnfs_error) + rdata->pnfs_error = -EIO ; +bl_set_lo_fail (rdata->lseg); + }bio_put (bio); -put_parallel (data); +put_parallel (par); }static void bl_read_cleanup (struct work_struct *work)
@@ -219,13 +224,7 @@ static enum pnfs_try_status bl_read_pagelist(struct nfs_read_data *rdata)dprintk ("%s dont_like_caller failed\n" , __func__); goto use_mds; } -if ((rdata->npages ==1 ) &&PagePnfsErr (rdata->req->wb_page)) { -/* We want to fall back to mds in case of read_page -* after error on read_pages. -*/ -dprintk ("%s PG_pnfserr set\n" , __func__); - goto use_mds; - } + par =alloc_parallel (rdata);if (!par) goto use_mds;
@@ -246,9 +245,8 @@ static enum pnfs_try_status bl_read_pagelist(struct nfs_read_data *rdata) be =find_get_extent (BLK_LSEG2EXT (rdata->lseg), isect, &cow_read);if (!be) { -/* Error out this page */ -bl_done_with_rpage (pages[i],0 ); -break ; + rdata->pnfs_error = -EIO ; + goto out; } extent_length = be->be_length - (isect - be->be_f_offset);
@@ -263,10 +261,9 @@ static enum pnfs_try_status bl_read_pagelist(struct nfs_read_data *rdata) bio =bl_submit_bio (READ , bio);/* Fill hole w/ zeroes w/o accessing device */ dprintk ("%s Zeroing page for hole\n" , __func__); -zero_user (pages[i],0 , -min_t (int,PAGE_CACHE_SIZE , count)); +zero_user_segment (pages[i],0 ,PAGE_CACHE_SIZE );print_page (pages[i]); -bl_done_with_rpage (pages[i],1 ); +SetPageUptodate (pages[i]); }else {struct pnfs_block_extent *be_read;
@@ -277,9 +274,8 @@ static enum pnfs_try_status bl_read_pagelist(struct nfs_read_data *rdata)bio_alloc (GFP_NOIO , rdata->npages - i);if (!bio) { -/* Error out this page */ -bl_done_with_rpage (pages[i],0 ); -break ; + rdata->pnfs_error = -ENOMEM ; + goto out; } bio->bi_sector = isect - be_read->be_f_offset +
@@ -302,6 +298,7 @@ static enum pnfs_try_status bl_read_pagelist(struct nfs_read_data *rdata) }else { rdata->res.count = (isect <<9 ) - f_offset; } +out:put_extent (be);put_extent (cow_read);bl_submit_bio (READ , bio);
@@ -337,22 +334,6 @@ static void mark_extents_written(struct pnfs_block_layout *bl, } } -static inline void bl_done_with_wpage (struct page *page,const int ok) -{ -if (!ok) { -SetPageError (page); -SetPagePnfsErr (page); -/* This is an inline copy of nfs_zap_mapping */ -/* This is oh so fishy, and needs deep thought */ -if (page->mapping->nrpages !=0 ) { -struct inode *inode = page->mapping->host; -spin_lock (&inode->i_lock); -NFS_I (inode)->cache_validity |=NFS_INO_INVALID_DATA ; -spin_unlock (&inode->i_lock); - } - } -} -static void bl_end_io_write_zero (struct bio *bio,int err) {struct parallel_io *par = bio->bi_private;
@@ -365,13 +346,15 @@ static void bl_end_io_write_zero(struct bio *bio, int err)if (--bvec >= bio->bi_io_vec)prefetchw (&bvec->bv_page->flags); -bl_done_with_wpage (page, uptodate);/* This is the zeroing page we added */ end_page_writeback (page);page_cache_release (page); }while (bvec >= bio->bi_io_vec); -if (!uptodate && !wdata->pnfs_error) - wdata->pnfs_error = -EIO ; +if (!uptodate) { +if (!wdata->pnfs_error) + wdata->pnfs_error = -EIO ; +bl_set_lo_fail (wdata->lseg); + }bio_put (bio);put_parallel (par); }
@@ -380,18 +363,13 @@ static void bl_end_io_write(struct bio *bio, int err) {struct parallel_io *par = bio->bi_private;const int uptodate =test_bit (BIO_UPTODATE , &bio->bi_flags); -struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt -1 ;struct nfs_write_data *wdata = (struct nfs_write_data *)par->data; -do { -struct page *page = bvec->bv_page; - -if (--bvec >= bio->bi_io_vec) -prefetchw (&bvec->bv_page->flags); -bl_done_with_wpage (page, uptodate); - }while (bvec >= bio->bi_io_vec); -if (!uptodate && !wdata->pnfs_error) - wdata->pnfs_error = -EIO ; +if (!uptodate) { +if (!wdata->pnfs_error) + wdata->pnfs_error = -EIO ; +bl_set_lo_fail (wdata->lseg); + }bio_put (bio);put_parallel (par); }
fs/nfs/blocklayout/blocklayout.h +0 -5
@@ -37,11 +37,6 @@#define PAGE_CACHE_SECTORS (PAGE_CACHE_SIZE >> 9) -#define PG_pnfserr PG_owner_priv_1 -#define PagePnfsErr (page) test_bit(PG_pnfserr, &(page)->flags) -#define SetPagePnfsErr (page) set_bit(PG_pnfserr, &(page)->flags) -#define ClearPagePnfsErr (page) clear_bit(PG_pnfserr, &(page)->flags) -struct block_mount_id {spinlock_t bm_lock;/* protects list */ struct list_head bm_devlist;/* holds pnfs_block_dev */
Signature · 1 line
1.7.4.1