From 29c8bbbd6e21daa0997d1c3ee886b897ee7ad652 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:43 +0000 Subject: afs: Fix missing put_page() In afs_writepages_region(), inside the loop where we find dirty pages to deal with, one of the if-statements is missing a put_page(). Signed-off-by: David Howells --- fs/afs/write.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/afs/write.c') diff --git a/fs/afs/write.c b/fs/afs/write.c index c83c1a0e851f..e919e64cd4e0 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -513,6 +513,7 @@ static int afs_writepages_region(struct address_space *mapping, if (PageWriteback(page) || !PageDirty(page)) { unlock_page(page); + put_page(page); continue; } -- cgit v1.2.3 From 5611ef280d814042825ee17688f5751266fc538b Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:43 +0000 Subject: afs: Fix page overput in afs_fill_page() afs_fill_page() loads the page it wants to fill into the afs_read request without incrementing its refcount - but then calls afs_put_read() to clean up afterwards, which then releases a ref on the page. Fix this by getting a ref on the page before calling afs_vnode_fetch_data(). This causes sync after a write to hang in afs_writepages_region() because find_get_pages_tag() gets confused and doesn't return. Fixes: 196ee9cd2d04 ("afs: Make afs_fs_fetch_data() take a list of pages") Reported-by: Marc Dionne Signed-off-by: David Howells Tested-by: Marc Dionne --- fs/afs/write.c | 1 + 1 file changed, 1 insertion(+) (limited to 'fs/afs/write.c') diff --git a/fs/afs/write.c b/fs/afs/write.c index e919e64cd4e0..3ac52f6a96ff 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -101,6 +101,7 @@ static int afs_fill_page(struct afs_vnode *vnode, struct key *key, req->pos = pos; req->nr_pages = 1; req->pages[0] = page; + get_page(page); i_size = i_size_read(&vnode->vfs_inode); if (pos + PAGE_SIZE > i_size) -- cgit v1.2.3 From e8e581a88c5f5fc7cf1f636d122b77fbcfc8c2f6 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:44 +0000 Subject: afs: Handle a short write to an AFS page Handle the situation where afs_write_begin() is told to expect that a full-page write will be made, but this doesn't happen (EFAULT, CTRL-C, etc.), and so afs_write_end() sees a partial write took place. Currently, no attempt is to deal with the discrepency. Fix this by loading the gap from the server. Reported-by: Al Viro Signed-off-by: David Howells --- fs/afs/fsclient.c | 4 +++- fs/afs/internal.h | 2 +- fs/afs/write.c | 28 +++++++++++++++++++--------- 3 files changed, 23 insertions(+), 11 deletions(-) (limited to 'fs/afs/write.c') diff --git a/fs/afs/fsclient.c b/fs/afs/fsclient.c index bf8904a1a58f..6f917dd1238c 100644 --- a/fs/afs/fsclient.c +++ b/fs/afs/fsclient.c @@ -393,8 +393,10 @@ static int afs_deliver_fs_fetch_data(struct afs_call *call) if (req->remain > 0) { call->offset = 0; req->index++; - if (req->index >= req->nr_pages) + if (req->index >= req->nr_pages) { + call->unmarshall = 4; goto begin_discard; + } goto begin_page; } } diff --git a/fs/afs/internal.h b/fs/afs/internal.h index 7784a8bc375c..dc2cb486e127 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -130,7 +130,7 @@ struct afs_call_type { */ struct afs_read { loff_t pos; /* Where to start reading */ - loff_t len; /* How much to read */ + loff_t len; /* How much we're asking for */ loff_t actual_len; /* How much we're actually getting */ atomic_t usage; unsigned int remain; /* Amount remaining */ diff --git a/fs/afs/write.c b/fs/afs/write.c index 3ac52f6a96ff..ea66890fc188 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -84,10 +84,9 @@ void afs_put_writeback(struct afs_writeback *wb) * partly or wholly fill a page that's under preparation for writing */ static int afs_fill_page(struct afs_vnode *vnode, struct key *key, - loff_t pos, struct page *page) + loff_t pos, unsigned int len, struct page *page) { struct afs_read *req; - loff_t i_size; int ret; _enter(",,%llu", (unsigned long long)pos); @@ -99,16 +98,11 @@ static int afs_fill_page(struct afs_vnode *vnode, struct key *key, atomic_set(&req->usage, 1); req->pos = pos; + req->len = len; req->nr_pages = 1; req->pages[0] = page; get_page(page); - i_size = i_size_read(&vnode->vfs_inode); - if (pos + PAGE_SIZE > i_size) - req->len = i_size - pos; - else - req->len = PAGE_SIZE; - ret = afs_vnode_fetch_data(vnode, key, req); afs_put_read(req); if (ret < 0) { @@ -164,7 +158,7 @@ int afs_write_begin(struct file *file, struct address_space *mapping, /* page won't leak in error case: it eventually gets cleaned off LRU */ if (!PageUptodate(page) && len != PAGE_SIZE) { - ret = afs_fill_page(vnode, key, index << PAGE_SHIFT, page); + ret = afs_fill_page(vnode, key, pos & PAGE_MASK, PAGE_SIZE, page); if (ret < 0) { kfree(candidate); _leave(" = %d [prep]", ret); @@ -258,7 +252,9 @@ int afs_write_end(struct file *file, struct address_space *mapping, struct page *page, void *fsdata) { struct afs_vnode *vnode = AFS_FS_I(file_inode(file)); + struct key *key = file->private_data; loff_t i_size, maybe_i_size; + int ret; _enter("{%x:%u},{%lx}", vnode->fid.vid, vnode->fid.vnode, page->index); @@ -274,6 +270,20 @@ int afs_write_end(struct file *file, struct address_space *mapping, spin_unlock(&vnode->writeback_lock); } + if (!PageUptodate(page)) { + if (copied < len) { + /* Try and load any missing data from the server. The + * unmarshalling routine will take care of clearing any + * bits that are beyond the EOF. + */ + ret = afs_fill_page(vnode, key, pos + copied, + len - copied, page); + if (ret < 0) + return ret; + } + SetPageUptodate(page); + } + set_page_dirty(page); if (PageDirty(page)) _debug("dirtied"); -- cgit v1.2.3 From 58fed94dfb17e89556b5705f20f90e5b2971b6a1 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:45 +0000 Subject: afs: Flush outstanding writes when an fd is closed Flush outstanding writes in afs when an fd is closed. This is what NFS and CIFS do. Reported-by: Marc Dionne Signed-off-by: David Howells --- fs/afs/file.c | 1 + fs/afs/internal.h | 1 + fs/afs/write.c | 14 ++++++++++++++ 3 files changed, 16 insertions(+) (limited to 'fs/afs/write.c') diff --git a/fs/afs/file.c b/fs/afs/file.c index a38e1c30d110..b5829443ff69 100644 --- a/fs/afs/file.c +++ b/fs/afs/file.c @@ -30,6 +30,7 @@ static int afs_readpages(struct file *filp, struct address_space *mapping, const struct file_operations afs_file_operations = { .open = afs_open, + .flush = afs_flush, .release = afs_release, .llseek = generic_file_llseek, .read_iter = generic_file_read_iter, diff --git a/fs/afs/internal.h b/fs/afs/internal.h index dc2cb486e127..af1d91ec7f2c 100644 --- a/fs/afs/internal.h +++ b/fs/afs/internal.h @@ -720,6 +720,7 @@ extern int afs_writepages(struct address_space *, struct writeback_control *); extern void afs_pages_written_back(struct afs_vnode *, struct afs_call *); extern ssize_t afs_file_write(struct kiocb *, struct iov_iter *); extern int afs_writeback_all(struct afs_vnode *); +extern int afs_flush(struct file *, fl_owner_t); extern int afs_fsync(struct file *, loff_t, loff_t, int); diff --git a/fs/afs/write.c b/fs/afs/write.c index ea66890fc188..f1450ea09406 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -757,6 +757,20 @@ out: return ret; } +/* + * Flush out all outstanding writes on a file opened for writing when it is + * closed. + */ +int afs_flush(struct file *file, fl_owner_t id) +{ + _enter(""); + + if ((file->f_mode & FMODE_WRITE) == 0) + return 0; + + return vfs_fsync(file, 0); +} + /* * notification that a previously read-only page is about to become writable * - if it returns an error, the caller will deliver a bus error signal -- cgit v1.2.3 From 6d06b0d25209c80e99c1e89700f1e09694a3766b Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:48 +0000 Subject: afs: Fix page leak in afs_write_begin() afs_write_begin() leaks a ref and a lock on a page if afs_fill_page() fails. Fix the leak by unlocking and releasing the page in the error path. Signed-off-by: David Howells --- fs/afs/write.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'fs/afs/write.c') diff --git a/fs/afs/write.c b/fs/afs/write.c index f1450ea09406..6e13e96c3db0 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -154,12 +154,12 @@ int afs_write_begin(struct file *file, struct address_space *mapping, kfree(candidate); return -ENOMEM; } - *pagep = page; - /* page won't leak in error case: it eventually gets cleaned off LRU */ if (!PageUptodate(page) && len != PAGE_SIZE) { ret = afs_fill_page(vnode, key, pos & PAGE_MASK, PAGE_SIZE, page); if (ret < 0) { + unlock_page(page); + put_page(page); kfree(candidate); _leave(" = %d [prep]", ret); return ret; @@ -167,6 +167,9 @@ int afs_write_begin(struct file *file, struct address_space *mapping, SetPageUptodate(page); } + /* page won't leak in error case: it eventually gets cleaned off LRU */ + *pagep = page; + try_again: spin_lock(&vnode->writeback_lock); -- cgit v1.2.3 From 7286a35e893176169b09715096a4aca557e2ccd2 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:48 +0000 Subject: afs: Fix afs_kill_pages() Fix afs_kill_pages() in two ways: (1) If a writeback has been partially flushed, then if we try and kill the pages it contains, some of them may no longer be undergoing writeback and end_page_writeback() will assert. Fix this by checking to see whether the page in question is actually undergoing writeback before ending that writeback. (2) The loop that scans for pages to kill doesn't increase the first page index, and so the loop may not terminate, but it will try to process the same pages over and over again. Fix this by increasing the first page index to one after the last page we processed. Signed-off-by: David Howells --- fs/afs/write.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'fs/afs/write.c') diff --git a/fs/afs/write.c b/fs/afs/write.c index 6e13e96c3db0..134de0667898 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -321,10 +321,14 @@ static void afs_kill_pages(struct afs_vnode *vnode, bool error, ASSERTCMP(pv.nr, ==, count); for (loop = 0; loop < count; loop++) { - ClearPageUptodate(pv.pages[loop]); + struct page *page = pv.pages[loop]; + ClearPageUptodate(page); if (error) - SetPageError(pv.pages[loop]); - end_page_writeback(pv.pages[loop]); + SetPageError(page); + if (PageWriteback(page)) + end_page_writeback(page); + if (page->index >= first) + first = page->index + 1; } __pagevec_release(&pv); -- cgit v1.2.3 From 65a151094edeb04e8f5f6f1502028e2383e81bb8 Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:49 +0000 Subject: afs: ->writepage() shouldn't call clear_page_dirty_for_io() The ->writepage() op shouldn't call clear_page_dirty_for_io() as that has already been called by the caller. Fix afs_writepage() by moving the call out of afs_write_back_from_locked_page() to afs_writepages_region() where it is needed. Signed-off-by: David Howells --- fs/afs/write.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'fs/afs/write.c') diff --git a/fs/afs/write.c b/fs/afs/write.c index 134de0667898..e5f150bccfb5 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -231,7 +231,7 @@ flush_conflicting_wb: if (wb->state == AFS_WBACK_PENDING) wb->state = AFS_WBACK_CONFLICTING; spin_unlock(&vnode->writeback_lock); - if (PageDirty(page)) { + if (clear_page_dirty_for_io(page)) { ret = afs_write_back_from_locked_page(wb, page); if (ret < 0) { afs_put_writeback(candidate); @@ -353,8 +353,6 @@ static int afs_write_back_from_locked_page(struct afs_writeback *wb, _enter(",%lx", primary_page->index); count = 1; - if (!clear_page_dirty_for_io(primary_page)) - BUG(); if (test_set_page_writeback(primary_page)) BUG(); @@ -542,6 +540,8 @@ static int afs_writepages_region(struct address_space *mapping, wb->state = AFS_WBACK_WRITING; spin_unlock(&wb->vnode->writeback_lock); + if (!clear_page_dirty_for_io(page)) + BUG(); ret = afs_write_back_from_locked_page(wb, page); unlock_page(page); put_page(page); -- cgit v1.2.3 From c5051c7bc777dffa5661569dec5997f432b9a34a Mon Sep 17 00:00:00 2001 From: David Howells Date: Thu, 16 Mar 2017 16:27:49 +0000 Subject: afs: Don't wait for page writeback with the page lock held Drop the page lock before waiting for page writeback. Signed-off-by: David Howells --- fs/afs/write.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) (limited to 'fs/afs/write.c') diff --git a/fs/afs/write.c b/fs/afs/write.c index e5f150bccfb5..2d2fccd5044b 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -518,17 +518,16 @@ static int afs_writepages_region(struct address_space *mapping, */ lock_page(page); - if (page->mapping != mapping) { + if (page->mapping != mapping || !PageDirty(page)) { unlock_page(page); put_page(page); continue; } - if (wbc->sync_mode != WB_SYNC_NONE) - wait_on_page_writeback(page); - - if (PageWriteback(page) || !PageDirty(page)) { + if (PageWriteback(page)) { unlock_page(page); + if (wbc->sync_mode != WB_SYNC_NONE) + wait_on_page_writeback(page); put_page(page); continue; } -- cgit v1.2.3