From patchwork Tue Jul 18 03:45:30 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: sunqiuyang X-Patchwork-Id: 9846741 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id A984760393 for ; Tue, 18 Jul 2017 03:42:22 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id A4DEC26BE9 for ; Tue, 18 Jul 2017 03:42:22 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 97FE626CFF; Tue, 18 Jul 2017 03:42:22 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-6.9 required=2.0 tests=BAYES_00,RCVD_IN_DNSWL_HI autolearn=unavailable version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 1163526BE9 for ; Tue, 18 Jul 2017 03:42:21 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751356AbdGRDlx (ORCPT ); Mon, 17 Jul 2017 23:41:53 -0400 Received: from szxga02-in.huawei.com ([45.249.212.188]:10301 "EHLO szxga02-in.huawei.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751346AbdGRDlx (ORCPT ); Mon, 17 Jul 2017 23:41:53 -0400 Received: from 172.30.72.56 (EHLO dggeml405-hub.china.huawei.com) ([172.30.72.56]) by dggrg02-dlp.huawei.com (MOS 4.4.6-GA FastPath queued) with ESMTP id ARG71839; Tue, 18 Jul 2017 11:41:39 +0800 (CST) Received: from huawei.com (10.175.124.28) by dggeml405-hub.china.huawei.com (10.3.17.49) with Microsoft SMTP Server id 14.3.301.0; Tue, 18 Jul 2017 11:41:29 +0800 From: sunqiuyang To: , , CC: , , Subject: [PATCH v7 1/1] f2fs: dax: implement direct access Date: Tue, 18 Jul 2017 11:45:30 +0800 Message-ID: <1500349530-21530-1-git-send-email-sunqiuyang@huawei.com> X-Mailer: git-send-email 2.5.0 MIME-Version: 1.0 X-Originating-IP: [10.175.124.28] X-CFilter-Loop: Reflected X-Mirapoint-Virus-RAPID-Raw: score=unknown(0), refid=str=0001.0A090205.596D8374.003A, ss=1, re=0.000, recu=0.000, reip=0.000, cl=1, cld=1, fgs=0, ip=0.0.0.0, so=2014-11-16 11:51:01, dmn=2013-03-21 17:37:32 X-Mirapoint-Loop-Id: d0f6e3756b80d6c09ba0a6cc31303b7e Sender: linux-fsdevel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-fsdevel@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP From: Qiuyang Sun This patch implements Direct Access (DAX) in F2FS, including: - a mount option to choose whether to enable DAX or not - read/write and mmap of regular files in the DAX way - zero-out of unaligned partial blocks in the DAX way - garbage collection of DAX files, by mapping both old and new physical addresses of a data page into memory and copy data between them directly - incompatibility of DAX with inline data, atomic or volatile write, etc. Signed-off-by: Qiuyang Sun Reviewed-by: Chao Yu --- Changlog v6 -> v7: - Document the mount option "dax" for this feature in f2fs.txt - Minor cleanup in dax_move_data_page() --- Documentation/filesystems/f2fs.txt | 2 + fs/f2fs/data.c | 132 +++++++++++++++++++++++-- fs/f2fs/f2fs.h | 9 ++ fs/f2fs/file.c | 192 ++++++++++++++++++++++++++++++++++++- fs/f2fs/gc.c | 105 +++++++++++++++++++- fs/f2fs/inline.c | 4 + fs/f2fs/inode.c | 8 +- fs/f2fs/namei.c | 5 + fs/f2fs/super.c | 15 +++ 9 files changed, 459 insertions(+), 13 deletions(-) diff --git a/Documentation/filesystems/f2fs.txt b/Documentation/filesystems/f2fs.txt index 273ccb2..c86c421 100644 --- a/Documentation/filesystems/f2fs.txt +++ b/Documentation/filesystems/f2fs.txt @@ -164,6 +164,8 @@ io_bits=%u Set the bit size of write IO requests. It should be set with "mode=lfs". usrquota Enable plain user disk quota accounting. grpquota Enable plain group disk quota accounting. +dax Use direct access (no page cache). See + Documentation/filesystems/dax.txt. ================================================================================ DEBUGFS ENTRIES diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 87c1f41..26b908a 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -910,6 +910,15 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, err = -EIO; goto sync_out; } + /* + * If newly allocated blocks are to be zeroed out later, + * a single f2fs_map_blocks must not contain both old + * and new blocks at the same time. + */ + if (flag == F2FS_GET_BLOCK_ZERO + && (map->m_flags & F2FS_MAP_MAPPED) + && !(map->m_flags & F2FS_MAP_NEW)) + goto sync_out; if (flag == F2FS_GET_BLOCK_PRE_AIO) { if (blkaddr == NULL_ADDR) { prealloc++; @@ -938,7 +947,8 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, blkaddr != NEW_ADDR) goto sync_out; } - } + } else if (flag == F2FS_GET_BLOCK_ZERO && map->m_flags & F2FS_MAP_NEW) + goto sync_out; if (flag == F2FS_GET_BLOCK_PRE_AIO) goto skip; @@ -996,6 +1006,12 @@ int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, goto next_dnode; sync_out: + if (flag == F2FS_GET_BLOCK_ZERO && map->m_flags & F2FS_MAP_NEW) { + clean_bdev_aliases(inode->i_sb->s_bdev, + map->m_pblk, map->m_len); + err = sb_issue_zeroout(inode->i_sb, map->m_pblk, + map->m_len, GFP_NOFS); + } f2fs_put_dnode(&dn); unlock_out: if (create) { @@ -1808,16 +1824,19 @@ static int f2fs_write_data_pages(struct address_space *mapping, return 0; } -static void f2fs_write_failed(struct address_space *mapping, loff_t to) +static void f2fs_write_failed(struct address_space *mapping, loff_t to, + bool lock) { struct inode *inode = mapping->host; loff_t i_size = i_size_read(inode); if (to > i_size) { - down_write(&F2FS_I(inode)->i_mmap_sem); + if (lock) + down_write(&F2FS_I(inode)->i_mmap_sem); truncate_pagecache(inode, i_size); truncate_blocks(inode, i_size, true); - up_write(&F2FS_I(inode)->i_mmap_sem); + if (lock) + up_write(&F2FS_I(inode)->i_mmap_sem); } } @@ -2000,7 +2019,7 @@ static int f2fs_write_begin(struct file *file, struct address_space *mapping, fail: f2fs_put_page(page, 1); - f2fs_write_failed(mapping, pos + len); + f2fs_write_failed(mapping, pos + len, true); return err; } @@ -2077,7 +2096,7 @@ static ssize_t f2fs_direct_IO(struct kiocb *iocb, struct iov_iter *iter) if (err > 0) set_inode_flag(inode, FI_UPDATE_WRITE); else if (err < 0) - f2fs_write_failed(mapping, offset + count); + f2fs_write_failed(mapping, offset + count, true); } trace_f2fs_direct_IO_exit(inode, offset, count, rw, err); @@ -2274,3 +2293,104 @@ int f2fs_migrate_page(struct address_space *mapping, .migratepage = f2fs_migrate_page, #endif }; + +#ifdef CONFIG_FS_DAX +#include +#include + +static int f2fs_iomap_begin(struct inode *inode, loff_t offset, + loff_t length, unsigned int flags, struct iomap *iomap) +{ + struct block_device *bdev; + unsigned long first_block = F2FS_BYTES_TO_BLK(offset); + unsigned long last_block = F2FS_BYTES_TO_BLK(offset + length - 1); + struct f2fs_map_blocks map; + int ret; + + if (WARN_ON_ONCE(f2fs_has_inline_data(inode))) + return -ERANGE; + + map.m_lblk = first_block; + map.m_len = last_block - first_block + 1; + map.m_next_pgofs = NULL; + + if (!(flags & IOMAP_WRITE)) + ret = f2fs_map_blocks(inode, &map, 0, F2FS_GET_BLOCK_FIEMAP); + else { + /* i_size should be kept here and changed later in f2fs_iomap_end */ + loff_t original_i_size = i_size_read(inode); + + ret = f2fs_map_blocks(inode, &map, 1, F2FS_GET_BLOCK_ZERO); + if (i_size_read(inode) > original_i_size) { + f2fs_i_size_write(inode, original_i_size); + if (ret) + f2fs_write_failed(inode->i_mapping, + offset + length, + !(flags & IOMAP_FAULT)); + } + } + + if (ret) + return ret; + + iomap->flags = 0; + bdev = inode->i_sb->s_bdev; + iomap->bdev = bdev; + if (blk_queue_dax(bdev->bd_queue)) + iomap->dax_dev = dax_get_by_host(bdev->bd_disk->disk_name); + else + iomap->dax_dev = NULL; + iomap->offset = F2FS_BLK_TO_BYTES((u64)first_block); + + if (map.m_len == 0) { + iomap->type = IOMAP_HOLE; + iomap->blkno = IOMAP_NULL_BLOCK; + iomap->length = F2FS_BLKSIZE; + } else { + if (map.m_flags & F2FS_MAP_UNWRITTEN) + iomap->type = IOMAP_UNWRITTEN; + else if (map.m_flags & F2FS_MAP_MAPPED) + iomap->type = IOMAP_MAPPED; + else { + WARN_ON_ONCE(1); + return -EIO; + } + iomap->blkno = + (sector_t)map.m_pblk << F2FS_LOG_SECTORS_PER_BLOCK; + iomap->length = F2FS_BLK_TO_BYTES((u64)map.m_len); + } + + if (map.m_flags & F2FS_MAP_NEW) + iomap->flags |= IOMAP_F_NEW; + return 0; +} + +static int f2fs_iomap_end(struct inode *inode, loff_t offset, loff_t length, + ssize_t written, unsigned int flags, struct iomap *iomap) +{ + put_dax(iomap->dax_dev); + if (!(flags & IOMAP_WRITE) || (flags & IOMAP_FAULT)) + return 0; + + if (offset + written > i_size_read(inode)) + f2fs_i_size_write(inode, offset + written); + + if (iomap->offset + iomap->length > + ALIGN(i_size_read(inode), F2FS_BLKSIZE)) { + block_t written_blk = F2FS_BYTES_TO_BLK(offset + written); + block_t end_blk = F2FS_BYTES_TO_BLK(offset + length); + + if (written_blk < end_blk) + f2fs_write_failed(inode->i_mapping, offset + length, + true); + } + + f2fs_update_time(F2FS_I_SB(inode), REQ_TIME); + return 0; +} + +struct iomap_ops f2fs_iomap_ops = { + .iomap_begin = f2fs_iomap_begin, + .iomap_end = f2fs_iomap_end, +}; +#endif diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 94a88b2..0bbe011 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -91,6 +91,11 @@ struct f2fs_fault_info { #define F2FS_MOUNT_LFS 0x00040000 #define F2FS_MOUNT_USRQUOTA 0x00080000 #define F2FS_MOUNT_GRPQUOTA 0x00100000 +#ifdef CONFIG_FS_DAX +#define F2FS_MOUNT_DAX 0x00200000 /* Direct Access */ +#else +#define F2FS_MOUNT_DAX 0 +#endif #define clear_opt(sbi, option) ((sbi)->mount_opt.opt &= ~F2FS_MOUNT_##option) #define set_opt(sbi, option) ((sbi)->mount_opt.opt |= F2FS_MOUNT_##option) @@ -479,6 +484,7 @@ struct f2fs_map_blocks { #define F2FS_GET_BLOCK_BMAP 3 #define F2FS_GET_BLOCK_PRE_DIO 4 #define F2FS_GET_BLOCK_PRE_AIO 5 +#define F2FS_GET_BLOCK_ZERO 6 /* * i_advise uses FADVISE_XXX_BIT. We can add additional hints later. @@ -2437,6 +2443,9 @@ void f2fs_invalidate_page(struct page *page, unsigned int offset, int f2fs_migrate_page(struct address_space *mapping, struct page *newpage, struct page *page, enum migrate_mode mode); #endif +#ifdef CONFIG_FS_DAX +extern struct iomap_ops f2fs_iomap_ops; +#endif /* * gc.c diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index a0e6d2c..0e62e97 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -23,6 +23,10 @@ #include #include #include +#ifdef CONFIG_FS_DAX +#include +#include +#endif #include "f2fs.h" #include "node.h" @@ -121,6 +125,64 @@ static int f2fs_vm_page_mkwrite(struct vm_fault *vmf) .page_mkwrite = f2fs_vm_page_mkwrite, }; +#ifdef CONFIG_FS_DAX +static int f2fs_dax_huge_fault(struct vm_fault *vmf, + enum page_entry_size pe_size) +{ + int result; + struct inode *inode = file_inode(vmf->vma->vm_file); + struct super_block *sb = inode->i_sb; + bool write = vmf->flags & FAULT_FLAG_WRITE; + + if (write) { + sb_start_pagefault(sb); + file_update_time(vmf->vma->vm_file); + } + down_read(&F2FS_I(inode)->i_mmap_sem); + result = dax_iomap_fault(vmf, pe_size, &f2fs_iomap_ops); + up_read(&F2FS_I(inode)->i_mmap_sem); + if (write) + sb_end_pagefault(sb); + + return result; +} + +static int f2fs_dax_fault(struct vm_fault *vmf) +{ + return f2fs_dax_huge_fault(vmf, PE_SIZE_PTE); +} + +static int f2fs_dax_pfn_mkwrite(struct vm_fault *vmf) +{ + struct inode *inode = file_inode(vmf->vma->vm_file); + struct super_block *sb = inode->i_sb; + loff_t size; + int ret; + + sb_start_pagefault(sb); + file_update_time(vmf->vma->vm_file); + down_read(&F2FS_I(inode)->i_mmap_sem); + size = (i_size_read(inode) + PAGE_SIZE - 1) >> PAGE_SHIFT; + if (vmf->pgoff >= size) + ret = VM_FAULT_SIGBUS; + else + ret = dax_pfn_mkwrite(vmf); + up_read(&F2FS_I(inode)->i_mmap_sem); + sb_end_pagefault(sb); + + return ret; +} + +static const struct vm_operations_struct f2fs_dax_vm_ops = { + .fault = f2fs_dax_fault, + .huge_fault = f2fs_dax_huge_fault, + .page_mkwrite = f2fs_dax_fault, + .pfn_mkwrite = f2fs_dax_pfn_mkwrite, +}; +#else +#define f2fs_dax_vm_ops f2fs_file_vm_ops +#endif + static int get_parent_ino(struct inode *inode, nid_t *pino) { struct dentry *dentry; @@ -436,7 +498,13 @@ static int f2fs_file_mmap(struct file *file, struct vm_area_struct *vma) return err; file_accessed(file); - vma->vm_ops = &f2fs_file_vm_ops; + + if (IS_DAX(inode)) { + vma->vm_ops = &f2fs_dax_vm_ops; + vma->vm_flags |= VM_MIXEDMAP | VM_HUGEPAGE; + } else + vma->vm_ops = &f2fs_file_vm_ops; + return 0; } @@ -519,6 +587,17 @@ static int truncate_partial_data_page(struct inode *inode, u64 from, if (!offset && !cache_only) return 0; +#ifdef CONFIG_FS_DAX + if (IS_DAX(inode)) { + int ret; + + down_read(&F2FS_I(inode)->dio_rwsem[WRITE]); + ret = iomap_zero_range(inode, from, PAGE_SIZE - offset, + NULL, &f2fs_iomap_ops); + up_read(&F2FS_I(inode)->dio_rwsem[WRITE]); + return ret; + } +#endif if (cache_only) { page = find_lock_page(mapping, index); if (page && PageUptodate(page)) @@ -799,6 +878,18 @@ static int fill_zero(struct inode *inode, pgoff_t index, if (!len) return 0; +#ifdef CONFIG_FS_DAX + if (IS_DAX(inode)) { + int ret; + + down_read(&F2FS_I(inode)->dio_rwsem[WRITE]); + ret = iomap_zero_range(inode, + F2FS_BLK_TO_BYTES((loff_t)index) + start, + len, NULL, &f2fs_iomap_ops); + up_read(&F2FS_I(inode)->dio_rwsem[WRITE]); + return ret; + } +#endif f2fs_balance_fs(sbi, true); f2fs_lock_op(sbi); @@ -1121,6 +1212,11 @@ static int f2fs_collapse_range(struct inode *inode, loff_t offset, loff_t len) loff_t new_size; int ret; +#ifdef CONFIG_FS_DAX + /* The current implementation does not apply to DAX files. */ + if (IS_DAX(inode)) + return -EINVAL; +#endif if (offset + len >= i_size_read(inode)) return -EINVAL; @@ -1311,6 +1407,11 @@ static int f2fs_insert_range(struct inode *inode, loff_t offset, loff_t len) loff_t new_size; int ret = 0; +#ifdef CONFIG_FS_DAX + /* The current implementation does not apply to DAX files. */ + if (IS_DAX(inode)) + return -EINVAL; +#endif new_size = i_size_read(inode) + len; ret = inode_newsize_ok(inode, new_size); if (ret) @@ -1581,6 +1682,10 @@ static int f2fs_ioc_start_atomic_write(struct file *filp) struct inode *inode = file_inode(filp); int ret; +#ifdef CONFIG_FS_DAX + if (IS_DAX(inode)) + return -EINVAL; +#endif if (!inode_owner_or_capable(inode)) return -EACCES; @@ -1630,6 +1735,9 @@ static int f2fs_ioc_commit_atomic_write(struct file *filp) struct inode *inode = file_inode(filp); int ret; +#ifdef CONFIG_FS_DAX + f2fs_bug_on(F2FS_I_SB(inode), IS_DAX(inode)); +#endif if (!inode_owner_or_capable(inode)) return -EACCES; @@ -1666,6 +1774,10 @@ static int f2fs_ioc_start_volatile_write(struct file *filp) struct inode *inode = file_inode(filp); int ret; +#ifdef CONFIG_FS_DAX + if (IS_DAX(inode)) + return -EINVAL; +#endif if (!inode_owner_or_capable(inode)) return -EACCES; @@ -1701,6 +1813,9 @@ static int f2fs_ioc_release_volatile_write(struct file *filp) struct inode *inode = file_inode(filp); int ret; +#ifdef CONFIG_FS_DAX + f2fs_bug_on(F2FS_I_SB(inode), IS_DAX(inode)); +#endif if (!inode_owner_or_capable(inode)) return -EACCES; @@ -1730,6 +1845,9 @@ static int f2fs_ioc_abort_volatile_write(struct file *filp) struct inode *inode = file_inode(filp); int ret; +#ifdef CONFIG_FS_DAX + f2fs_bug_on(F2FS_I_SB(inode), IS_DAX(inode)); +#endif if (!inode_owner_or_capable(inode)) return -EACCES; @@ -2144,6 +2262,10 @@ static int f2fs_ioc_defragment(struct file *filp, unsigned long arg) struct f2fs_defragment range; int err; +#ifdef CONFIG_FS_DAX + if (IS_DAX(inode)) + return -EINVAL; +#endif if (!capable(CAP_SYS_ADMIN)) return -EPERM; @@ -2193,6 +2315,10 @@ static int f2fs_move_file_range(struct file *file_in, loff_t pos_in, size_t dst_osize; int ret; +#ifdef CONFIG_FS_DAX + if (IS_DAX(src) || IS_DAX(dst)) + return -EINVAL; +#endif if (file_in->f_path.mnt != file_out->f_path.mnt || src->i_sb != dst->i_sb) return -EXDEV; @@ -2434,6 +2560,61 @@ long f2fs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) } } +#ifdef CONFIG_FS_DAX +static ssize_t f2fs_dax_read_iter(struct kiocb *iocb, struct iov_iter *to) +{ + struct inode *inode = file_inode(iocb->ki_filp); + ssize_t ret; + + inode_lock_shared(inode); + + if (!IS_DAX(inode)) { + inode_unlock_shared(inode); + return generic_file_read_iter(iocb, to); + } + + down_read(&F2FS_I(inode)->dio_rwsem[READ]); + ret = dax_iomap_rw(iocb, to, &f2fs_iomap_ops); + up_read(&F2FS_I(inode)->dio_rwsem[READ]); + inode_unlock_shared(inode); + + file_accessed(iocb->ki_filp); + return ret; +} + +static ssize_t f2fs_file_read_iter(struct kiocb *iocb, struct iov_iter *to) +{ + if (!iov_iter_count(to)) + return 0; /* skip atime */ + + if (IS_DAX(file_inode(iocb->ki_filp))) + return f2fs_dax_read_iter(iocb, to); + + return generic_file_read_iter(iocb, to); +} + +static ssize_t f2fs_dax_write_iter(struct kiocb *iocb, struct iov_iter *from) +{ + struct inode *inode = file_inode(iocb->ki_filp); + ssize_t ret; + + ret = file_remove_privs(iocb->ki_filp); + if (ret) + return ret; + ret = file_update_time(iocb->ki_filp); + if (ret) + return ret; + + down_read(&F2FS_I(inode)->dio_rwsem[WRITE]); + ret = dax_iomap_rw(iocb, from, &f2fs_iomap_ops); + up_read(&F2FS_I(inode)->dio_rwsem[WRITE]); + + return ret; +} +#else +#define f2fs_dax_write_iter __generic_file_write_iter +#endif + static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *file = iocb->ki_filp; @@ -2455,7 +2636,10 @@ static ssize_t f2fs_file_write_iter(struct kiocb *iocb, struct iov_iter *from) return err; } blk_start_plug(&plug); - ret = __generic_file_write_iter(iocb, from); + if (IS_DAX(inode)) + ret = f2fs_dax_write_iter(iocb, from); + else + ret = __generic_file_write_iter(iocb, from); blk_finish_plug(&plug); clear_inode_flag(inode, FI_NO_PREALLOC); } @@ -2504,7 +2688,11 @@ long f2fs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) const struct file_operations f2fs_file_operations = { .llseek = f2fs_llseek, +#ifdef CONFIG_FS_DAX + .read_iter = f2fs_file_read_iter, +#else .read_iter = generic_file_read_iter, +#endif .write_iter = f2fs_file_write_iter, .open = f2fs_file_open, .release = f2fs_release_file, diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index fa3d2e2..289d748 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -700,6 +700,101 @@ static void move_encrypted_block(struct inode *inode, block_t bidx, f2fs_put_page(page, 1); } +#ifdef CONFIG_FS_DAX +#include + +static void dax_move_data_page(struct inode *inode, block_t bidx, + unsigned int segno, int off) +{ + struct block_device *bdev = inode->i_sb->s_bdev; + struct dax_device *dax_dev; + struct dnode_of_data dn; + struct f2fs_sb_info *sbi = F2FS_I_SB(inode); + struct f2fs_summary sum; + struct node_info ni; + block_t old_blkaddr, new_blkaddr; + int err, id; + long map_len; + pgoff_t pgoff; + void *kaddr_old, *kaddr_new; + pfn_t pfn; + + f2fs_bug_on(sbi, f2fs_is_atomic_file(inode)); + + if (!check_valid_map(sbi, segno, off)) + return; + + if (!blk_queue_dax(bdev->bd_queue)) + return; + + dax_dev = dax_get_by_host(bdev->bd_disk->disk_name); + + if (!down_write_trylock(&F2FS_I(inode)->i_mmap_sem)) + goto release; + + unmap_mapping_range(inode->i_mapping, (loff_t)bidx << PAGE_SHIFT, + PAGE_SIZE, 1); + /* find the old block address */ + set_new_dnode(&dn, inode, NULL, NULL, 0); + err = get_dnode_of_data(&dn, bidx, LOOKUP_NODE); + if (err) + goto out; + old_blkaddr = dn.data_blkaddr; + /* This page is already truncated */ + if (old_blkaddr == NULL_ADDR) + goto put_dn; + + /* allocate a new block address */ + get_node_info(sbi, dn.nid, &ni); + set_summary(&sum, dn.nid, dn.ofs_in_node, ni.version); + allocate_data_block(sbi, NULL, old_blkaddr, &new_blkaddr, + &sum, CURSEG_COLD_DATA, NULL, false); + + /* copy data page from old to new address in dax_bdev */ + id = dax_read_lock(); + err = bdev_dax_pgoff(bdev, SECTOR_FROM_BLOCK(old_blkaddr), + PAGE_SIZE, &pgoff); + if (err) + goto recover; + map_len = dax_direct_access(dax_dev, pgoff, 1, &kaddr_old, &pfn); + if (map_len < 0) + goto recover; + err = bdev_dax_pgoff(bdev, SECTOR_FROM_BLOCK(new_blkaddr), + PAGE_SIZE, &pgoff); + if (err) + goto recover; + map_len = dax_direct_access(dax_dev, pgoff, 1, &kaddr_new, &pfn); + if (map_len < 0) + goto recover; + copy_page((void __force *)kaddr_new, (void __force *)kaddr_old); + + f2fs_update_data_blkaddr(&dn, new_blkaddr); + set_inode_flag(inode, FI_APPEND_WRITE); + if (bidx == 0) + set_inode_flag(inode, FI_FIRST_BLOCK_WRITTEN); + +recover: + if (err || map_len < 0) + __f2fs_replace_block(sbi, &sum, new_blkaddr, old_blkaddr, + true, true); + dax_read_unlock(id); +put_dn: + f2fs_put_dnode(&dn); +out: + unmap_mapping_range(inode->i_mapping, (loff_t)bidx << PAGE_SHIFT, + PAGE_SIZE, 1); + up_write(&F2FS_I(inode)->i_mmap_sem); +release: + put_dax(dax_dev); +} +#else +static void dax_move_data_page(struct inode *inode, block_t bidx, + unsigned int segno, int off) +{ + return; +} +#endif + static void move_data_page(struct inode *inode, block_t bidx, int gc_type, unsigned int segno, int off) { @@ -818,9 +913,9 @@ static void gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, if (IS_ERR(inode) || is_bad_inode(inode)) continue; - /* if encrypted inode, let's go phase 3 */ - if (f2fs_encrypted_inode(inode) && - S_ISREG(inode->i_mode)) { + /* if DAX or encrypted inode, let's go phase 3 */ + if (IS_DAX(inode) || (f2fs_encrypted_inode(inode) && + S_ISREG(inode->i_mode))) { add_gc_inode(gc_list, inode); continue; } @@ -858,7 +953,9 @@ static void gc_data_segment(struct f2fs_sb_info *sbi, struct f2fs_summary *sum, start_bidx = start_bidx_of_node(nofs, inode) + ofs_in_node; - if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) + if (IS_DAX(inode)) + dax_move_data_page(inode, start_bidx, segno, off); + else if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) move_encrypted_block(inode, start_bidx, segno, off); else move_data_page(inode, start_bidx, gc_type, segno, off); diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c index e0fd437..fd8b290 100644 --- a/fs/f2fs/inline.c +++ b/fs/f2fs/inline.c @@ -28,6 +28,10 @@ bool f2fs_may_inline_data(struct inode *inode) if (f2fs_encrypted_inode(inode) && S_ISREG(inode->i_mode)) return false; +#ifdef CONFIG_FS_DAX + if (IS_DAX(inode)) + return false; +#endif return true; } diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 6cd312a..7741461 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -43,8 +43,14 @@ void f2fs_set_inode_flags(struct inode *inode) new_fl |= S_NOATIME; if (flags & FS_DIRSYNC_FL) new_fl |= S_DIRSYNC; +#ifdef CONFIG_FS_DAX + if (test_opt(F2FS_I_SB(inode), DAX) && S_ISREG(inode->i_mode) && + !f2fs_has_inline_data(inode) && !f2fs_encrypted_inode(inode) && + !f2fs_is_atomic_file(inode) && !f2fs_is_volatile_file(inode)) + new_fl |= S_DAX; +#endif inode_set_flags(inode, new_fl, - S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC); + S_SYNC|S_APPEND|S_IMMUTABLE|S_NOATIME|S_DIRSYNC|S_DAX); } static void __get_inode_rdev(struct inode *inode, struct f2fs_inode *ri) diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index 760d852..afc52e0 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -70,6 +70,11 @@ static struct inode *f2fs_new_inode(struct inode *dir, umode_t mode) if (f2fs_encrypted_inode(dir) && f2fs_may_encrypt(inode)) f2fs_set_encrypted_inode(inode); +#ifdef CONFIG_FS_DAX + if (test_opt(sbi, DAX) && S_ISREG(inode->i_mode) && + !f2fs_has_inline_data(inode) && !f2fs_encrypted_inode(inode)) + inode->i_flags |= S_DAX; +#endif set_inode_flag(inode, FI_NEW_INODE); if (test_opt(sbi, INLINE_XATTR)) diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c index 32e4c02..aefe931 100644 --- a/fs/f2fs/super.c +++ b/fs/f2fs/super.c @@ -109,6 +109,7 @@ enum { Opt_nolazytime, Opt_usrquota, Opt_grpquota, + Opt_dax, Opt_err, }; @@ -146,6 +147,7 @@ enum { {Opt_nolazytime, "nolazytime"}, {Opt_usrquota, "usrquota"}, {Opt_grpquota, "grpquota"}, + {Opt_dax, "dax"}, {Opt_err, NULL}, }; @@ -399,6 +401,15 @@ static int parse_options(struct super_block *sb, char *options) "quota operations not supported"); break; #endif +#ifdef CONFIG_FS_DAX + case Opt_dax: + set_opt(sbi, DAX); + break; +#else + case Opt_dax: + f2fs_msg(sb, KERN_INFO, "dax option not supported"); + break; +#endif default: f2fs_msg(sb, KERN_ERR, "Unrecognized mount option \"%s\" or missing value", @@ -814,6 +825,10 @@ static int f2fs_show_options(struct seq_file *seq, struct dentry *root) if (test_opt(sbi, GRPQUOTA)) seq_puts(seq, ",grpquota"); #endif +#ifdef CONFIG_FS_DAX + if (test_opt(sbi, DAX)) + seq_puts(seq, ",dax"); +#endif return 0; }