diff options
| -rw-r--r-- | Documentation/filesystems/exofs.txt | 176 | ||||
| -rw-r--r-- | fs/Kconfig | 2 | ||||
| -rw-r--r-- | fs/Makefile | 1 | ||||
| -rw-r--r-- | fs/exofs/BUGS | 3 | ||||
| -rw-r--r-- | fs/exofs/Kbuild | 16 | ||||
| -rw-r--r-- | fs/exofs/Kconfig | 13 | ||||
| -rw-r--r-- | fs/exofs/common.h | 184 | ||||
| -rw-r--r-- | fs/exofs/dir.c | 672 | ||||
| -rw-r--r-- | fs/exofs/exofs.h | 180 | ||||
| -rw-r--r-- | fs/exofs/file.c | 87 | ||||
| -rw-r--r-- | fs/exofs/inode.c | 1303 | ||||
| -rw-r--r-- | fs/exofs/namei.c | 342 | ||||
| -rw-r--r-- | fs/exofs/osd.c | 153 | ||||
| -rw-r--r-- | fs/exofs/super.c | 584 | ||||
| -rw-r--r-- | fs/exofs/symlink.c | 57 |
15 files changed, 3773 insertions, 0 deletions
diff --git a/Documentation/filesystems/exofs.txt b/Documentation/filesystems/exofs.txt new file mode 100644 index 000000000000..0ced74c2f73c --- /dev/null +++ b/Documentation/filesystems/exofs.txt @@ -0,0 +1,176 @@ +=============================================================================== +WHAT IS EXOFS? +=============================================================================== + +exofs is a file system that uses an OSD and exports the API of a normal Linux +file system. Users access exofs like any other local file system, and exofs +will in turn issue commands to the local OSD initiator. + +OSD is a new T10 command set that views storage devices not as a large/flat +array of sectors but as a container of objects, each having a length, quota, +time attributes and more. Each object is addressed by a 64bit ID, and is +contained in a 64bit ID partition. Each object has associated attributes +attached to it, which are integral part of the object and provide metadata about +the object. The standard defines some common obligatory attributes, but user +attributes can be added as needed. + +=============================================================================== +ENVIRONMENT +=============================================================================== + +To use this file system, you need to have an object store to run it on. You +may download a target from: +http://open-osd.org + +See Documentation/scsi/osd.txt for how to setup a working osd environment. + +=============================================================================== +USAGE +=============================================================================== + +1. Download and compile exofs and open-osd initiator: + You need an external Kernel source tree or kernel headers from your + distribution. (anything based on 2.6.26 or later). + + a. download open-osd including exofs source using: + [parent-directory]$ git clone git://git.open-osd.org/open-osd.git + + b. Build the library module like this: + [parent-directory]$ make -C KSRC=$(KER_DIR) open-osd + + This will build both the open-osd initiator as well as the exofs kernel + module. Use whatever parameters you compiled your Kernel with and + $(KER_DIR) above pointing to the Kernel you compile against. See the file + open-osd/top-level-Makefile for an example. + +2. Get the OSD initiator and target set up properly, and login to the target. + See Documentation/scsi/osd.txt for farther instructions. Also see ./do-osd + for example script that does all these steps. + +3. Insmod the exofs.ko module: + [exofs]$ insmod exofs.ko + +4. Make sure the directory where you want to mount exists. If not, create it. + (For example, mkdir /mnt/exofs) + +5. At first run you will need to invoke the mkfs.exofs application + + As an example, this will create the file system on: + /dev/osd0 partition ID 65536 + + mkfs.exofs --pid=65536 --format /dev/osd0 + + The --format is optional if not specified no OSD_FORMAT will be + preformed and a clean file system will be created in the specified pid, + in the available space of the target. (Use --format=size_in_meg to limit + the total LUN space available) + + If pid already exist it will be deleted and a new one will be created in it's + place. Be careful. + + An exofs lives inside a single OSD partition. You can create multiple exofs + filesystems on the same device using multiple pids. + + (run mkfs.exofs without any parameters for usage help message) + +6. Mount the file system. + + For example, to mount /dev/osd0, partition ID 0x10000 on /mnt/exofs: + + mount -t exofs -o pid=65536 /dev/osd0 /mnt/exofs/ + +7. For reference (See do-exofs example script): + do-exofs start - an example of how to perform the above steps. + do-exofs stop - an example of how to unmount the file system. + do-exofs format - an example of how to format and mkfs a new exofs. + +8. Extra compilation flags (uncomment in fs/exofs/Kbuild): + CONFIG_EXOFS_DEBUG - for debug messages and extra checks. + +=============================================================================== +exofs mount options +=============================================================================== +Similar to any mount command: + mount -t exofs -o exofs_options /dev/osdX mount_exofs_directory + +Where: + -t exofs: specifies the exofs file system + + /dev/osdX: X is a decimal number. /dev/osdX was created after a successful + login into an OSD target. + + mount_exofs_directory: The directory to mount the file system on + + exofs specific options: Options are separated by commas (,) + pid=<integer> - The partition number to mount/create as + container of the filesystem. + This option is mandatory + to=<integer> - Timeout in ticks for a single command + default is (60 * HZ) [for debugging only] + +=============================================================================== +DESIGN +=============================================================================== + +* The file system control block (AKA on-disk superblock) resides in an object + with a special ID (defined in common.h). + Information included in the file system control block is used to fill the + in-memory superblock structure at mount time. This object is created before + the file system is used by mkexofs.c It contains information such as: + - The file system's magic number + - The next inode number to be allocated + +* Each file resides in its own object and contains the data (and it will be + possible to extend the file over multiple objects, though this has not been + implemented yet). + +* A directory is treated as a file, and essentially contains a list of <file + name, inode #> pairs for files that are found in that directory. The object + IDs correspond to the files' inode numbers and will be allocated according to + a bitmap (stored in a separate object). Now they are allocated using a + counter. + +* Each file's control block (AKA on-disk inode) is stored in its object's + attributes. This applies to both regular files and other types (directories, + device files, symlinks, etc.). + +* Credentials are generated per object (inode and superblock) when they is + created in memory (read off disk or created). The credential works for all + operations and is used as long as the object remains in memory. + +* Async OSD operations are used whenever possible, but the target may execute + them out of order. The operations that concern us are create, delete, + readpage, writepage, update_inode, and truncate. The following pairs of + operations should execute in the order written, and we need to prevent them + from executing in reverse order: + - The following are handled with the OBJ_CREATED and OBJ_2BCREATED + flags. OBJ_CREATED is set when we know the object exists on the OSD - + in create's callback function, and when we successfully do a read_inode. + OBJ_2BCREATED is set in the beginning of the create function, so we + know that we should wait. + - create/delete: delete should wait until the object is created + on the OSD. + - create/readpage: readpage should be able to return a page + full of zeroes in this case. If there was a write already + en-route (i.e. create, writepage, readpage) then the page + would be locked, and so it would really be the same as + create/writepage. + - create/writepage: if writepage is called for a sync write, it + should wait until the object is created on the OSD. + Otherwise, it should just return. + - create/truncate: truncate should wait until the object is + created on the OSD. + - create/update_inode: update_inode should wait until the + object is created on the OSD. + - Handled by VFS locks: + - readpage/delete: shouldn't happen because of page lock. + - writepage/delete: shouldn't happen because of page lock. + - readpage/writepage: shouldn't happen because of page lock. + +=============================================================================== +LICENSE/COPYRIGHT +=============================================================================== +The exofs file system is based on ext2 v0.5b (distributed with the Linux kernel +version 2.6.10). All files include the original copyrights, and the license +is GPL version 2 (only version 2, as is true for the Linux kernel). The +Linux kernel can be downloaded from www.kernel.org. diff --git a/fs/Kconfig b/fs/Kconfig index cef8b18ceaa3..ae3b34a2ea69 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -169,6 +169,8 @@ source "fs/romfs/Kconfig" source "fs/sysv/Kconfig" source "fs/ufs/Kconfig" +source "fs/exofs/Kconfig" + endif # MISC_FILESYSTEMS menuconfig NETWORK_FILESYSTEMS diff --git a/fs/Makefile b/fs/Makefile index b5cd8e18dd9f..15f73014a208 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -120,3 +120,4 @@ obj-$(CONFIG_DEBUG_FS) += debugfs/ obj-$(CONFIG_OCFS2_FS) += ocfs2/ obj-$(CONFIG_BTRFS_FS) += btrfs/ obj-$(CONFIG_GFS2_FS) += gfs2/ +obj-$(CONFIG_EXOFS_FS) += exofs/ diff --git a/fs/exofs/BUGS b/fs/exofs/BUGS new file mode 100644 index 000000000000..1b2d4c63a579 --- /dev/null +++ b/fs/exofs/BUGS @@ -0,0 +1,3 @@ +- Out-of-space may cause a severe problem if the object (and directory entry) + were written, but the inode attributes failed. Then if the filesystem was + unmounted and mounted the kernel can get into an endless loop doing a readdir. diff --git a/fs/exofs/Kbuild b/fs/exofs/Kbuild new file mode 100644 index 000000000000..cc2d22db119c --- /dev/null +++ b/fs/exofs/Kbuild @@ -0,0 +1,16 @@ +# +# Kbuild for the EXOFS module +# +# Copyright (C) 2008 Panasas Inc. All rights reserved. +# +# Authors: +# Boaz Harrosh <bharrosh@panasas.com> +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License version 2 +# +# Kbuild - Gets included from the Kernels Makefile and build system +# + +exofs-y := osd.o inode.o file.o symlink.o namei.o dir.o super.o +obj-$(CONFIG_EXOFS_FS) += exofs.o diff --git a/fs/exofs/Kconfig b/fs/exofs/Kconfig new file mode 100644 index 000000000000..86194b2f799d --- /dev/null +++ b/fs/exofs/Kconfig @@ -0,0 +1,13 @@ +config EXOFS_FS + tristate "exofs: OSD based file system support" + depends on SCSI_OSD_ULD + help + EXOFS is a file system that uses an OSD storage device, + as its backing storage. + +# Debugging-related stuff +config EXOFS_DEBUG + bool "Enable debugging" + depends on EXOFS_FS + help + This option enables EXOFS debug prints. diff --git a/fs/exofs/common.h b/fs/exofs/common.h new file mode 100644 index 000000000000..b1512c4bb8c7 --- /dev/null +++ b/fs/exofs/common.h @@ -0,0 +1,184 @@ +/* + * common.h - Common definitions for both Kernel and user-mode utilities + * + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#ifndef __EXOFS_COM_H__ +#define __EXOFS_COM_H__ + +#include <linux/types.h> + +#include <scsi/osd_attributes.h> +#include <scsi/osd_initiator.h> +#include <scsi/osd_sec.h> + +/**************************************************************************** + * Object ID related defines + * NOTE: inode# = object ID - EXOFS_OBJ_OFF + ****************************************************************************/ +#define EXOFS_MIN_PID 0x10000 /* Smallest partition ID */ +#define EXOFS_OBJ_OFF 0x10000 /* offset for objects */ +#define EXOFS_SUPER_ID 0x10000 /* object ID for on-disk superblock */ +#define EXOFS_ROOT_ID 0x10002 /* object ID for root directory */ + +/* exofs Application specific page/attribute */ +# define EXOFS_APAGE_FS_DATA (OSD_APAGE_APP_DEFINED_FIRST + 3) +# define EXOFS_ATTR_INODE_DATA 1 + +/* + * The maximum number of files we can have is limited by the size of the + * inode number. This is the largest object ID that the file system supports. + * Object IDs 0, 1, and 2 are always in use (see above defines). + */ +enum { + EXOFS_MAX_INO_ID = (sizeof(ino_t) * 8 == 64) ? ULLONG_MAX : + (1ULL << (sizeof(ino_t) * 8ULL - 1ULL)), + EXOFS_MAX_ID = (EXOFS_MAX_INO_ID - 1 - EXOFS_OBJ_OFF), +}; + +/**************************************************************************** + * Misc. + ****************************************************************************/ +#define EXOFS_BLKSHIFT 12 +#define EXOFS_BLKSIZE (1UL << EXOFS_BLKSHIFT) + +/**************************************************************************** + * superblock-related things + ****************************************************************************/ +#define EXOFS_SUPER_MAGIC 0x5DF5 + +/* + * The file system control block - stored in an object's data (mainly, the one + * with ID EXOFS_SUPER_ID). This is where the in-memory superblock is stored + * on disk. Right now it just has a magic value, which is basically a sanity + * check on our ability to communicate with the object store. + */ +struct exofs_fscb { + __le64 s_nextid; /* Highest object ID used */ + __le32 s_numfiles; /* Number of files on fs */ + __le16 s_magic; /* Magic signature */ + __le16 s_newfs; /* Non-zero if this is a new fs */ +}; + +/**************************************************************************** + * inode-related things + ****************************************************************************/ +#define EXOFS_IDATA 5 + +/* + * The file control block - stored in an object's attributes. This is where + * the in-memory inode is stored on disk. + */ +struct exofs_fcb { + __le64 i_size; /* Size of the file */ + __le16 i_mode; /* File mode */ + __le16 i_links_count; /* Links count */ + __le32 i_uid; /* Owner Uid */ + __le32 i_gid; /* Group Id */ + __le32 i_atime; /* Access time */ + __le32 i_ctime; /* Creation time */ + __le32 i_mtime; /* Modification time */ + __le32 i_flags; /* File flags (unused for now)*/ + __le32 i_generation; /* File version (for NFS) */ + __le32 i_data[EXOFS_IDATA]; /* Short symlink names and device #s */ +}; + +#define EXOFS_INO_ATTR_SIZE sizeof(struct exofs_fcb) + +/* This is the Attribute the fcb is stored in */ +static const struct __weak osd_attr g_attr_inode_data = ATTR_DEF( + EXOFS_APAGE_FS_DATA, + EXOFS_ATTR_INODE_DATA, + EXOFS_INO_ATTR_SIZE); + +/**************************************************************************** + * dentry-related things + ****************************************************************************/ +#define EXOFS_NAME_LEN 255 + +/* + * The on-disk directory entry + */ +struct exofs_dir_entry { + __le64 inode_no; /* inode number */ + __le16 rec_len; /* directory entry length */ + u8 name_len; /* name length */ + u8 file_type; /* umm...file type */ + char name[EXOFS_NAME_LEN]; /* file name */ +}; + +enum { + EXOFS_FT_UNKNOWN, + EXOFS_FT_REG_FILE, + EXOFS_FT_DIR, + EXOFS_FT_CHRDEV, + EXOFS_FT_BLKDEV, + EXOFS_FT_FIFO, + EXOFS_FT_SOCK, + EXOFS_FT_SYMLINK, + EXOFS_FT_MAX +}; + +#define EXOFS_DIR_PAD 4 +#define EXOFS_DIR_ROUND (EXOFS_DIR_PAD - 1) +#define EXOFS_DIR_REC_LEN(name_len) \ + (((name_len) + offsetof(struct exofs_dir_entry, name) + \ + EXOFS_DIR_ROUND) & ~EXOFS_DIR_ROUND) + +/************************* + * function declarations * + *************************/ +/* osd.c */ +void exofs_make_credential(u8 cred_a[OSD_CAP_LEN], + const struct osd_obj_id *obj); + +int exofs_check_ok_resid(struct osd_request *or, u64 *in_resid, u64 *out_resid); +static inline int exofs_check_ok(struct osd_request *or) +{ + return exofs_check_ok_resid(or, NULL, NULL); +} +int exofs_sync_op(struct osd_request *or, int timeout, u8 *cred); +int exofs_async_op(struct osd_request *or, + osd_req_done_fn *async_done, void *caller_context, u8 *cred); + +int extract_attr_from_req(struct osd_request *or, struct osd_attr *attr); + +int osd_req_read_kern(struct osd_request *or, + const struct osd_obj_id *obj, u64 offset, void *buff, u64 len); + +int osd_req_write_kern(struct osd_request *or, + const struct osd_obj_id *obj, u64 offset, void *buff, u64 len); + +#endif /*ifndef __EXOFS_COM_H__*/ diff --git a/fs/exofs/dir.c b/fs/exofs/dir.c new file mode 100644 index 000000000000..65b0c8c776a1 --- /dev/null +++ b/fs/exofs/dir.c @@ -0,0 +1,672 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "exofs.h" + +static inline unsigned exofs_chunk_size(struct inode *inode) +{ + return inode->i_sb->s_blocksize; +} + +static inline void exofs_put_page(struct page *page) +{ + kunmap(page); + page_cache_release(page); +} + +/* Accesses dir's inode->i_size must be called under inode lock */ +static inline unsigned long dir_pages(struct inode *inode) +{ + return (inode->i_size + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; +} + +static unsigned exofs_last_byte(struct inode *inode, unsigned long page_nr) +{ + loff_t last_byte = inode->i_size; + + last_byte -= page_nr << PAGE_CACHE_SHIFT; + if (last_byte > PAGE_CACHE_SIZE) + last_byte = PAGE_CACHE_SIZE; + return last_byte; +} + +static int exofs_commit_chunk(struct page *page, loff_t pos, unsigned len) +{ + struct address_space *mapping = page->mapping; + struct inode *dir = mapping->host; + int err = 0; + + dir->i_version++; + + if (!PageUptodate(page)) + SetPageUptodate(page); + + if (pos+len > dir->i_size) { + i_size_write(dir, pos+len); + mark_inode_dirty(dir); + } + set_page_dirty(page); + + if (IS_DIRSYNC(dir)) + err = write_one_page(page, 1); + else + unlock_page(page); + + return err; +} + +static void exofs_check_page(struct page *page) +{ + struct inode *dir = page->mapping->host; + unsigned chunk_size = exofs_chunk_size(dir); + char *kaddr = page_address(page); + unsigned offs, rec_len; + unsigned limit = PAGE_CACHE_SIZE; + struct exofs_dir_entry *p; + char *error; + + /* if the page is the last one in the directory */ + if ((dir->i_size >> PAGE_CACHE_SHIFT) == page->index) { + limit = dir->i_size & ~PAGE_CACHE_MASK; + if (limit & (chunk_size - 1)) + goto Ebadsize; + if (!limit) + goto out; + } + for (offs = 0; offs <= limit - EXOFS_DIR_REC_LEN(1); offs += rec_len) { + p = (struct exofs_dir_entry *)(kaddr + offs); + rec_len = le16_to_cpu(p->rec_len); + + if (rec_len < EXOFS_DIR_REC_LEN(1)) + goto Eshort; + if (rec_len & 3) + goto Ealign; + if (rec_len < EXOFS_DIR_REC_LEN(p->name_len)) + goto Enamelen; + if (((offs + rec_len - 1) ^ offs) & ~(chunk_size-1)) + goto Espan; + } + if (offs != limit) + goto Eend; +out: + SetPageChecked(page); + return; + +Ebadsize: + EXOFS_ERR("ERROR [exofs_check_page]: " + "size of directory #%lu is not a multiple of chunk size", + dir->i_ino + ); + goto fail; +Eshort: + error = "rec_len is smaller than minimal"; + goto bad_entry; +Ealign: + error = "unaligned directory entry"; + goto bad_entry; +Enamelen: + error = "rec_len is too small for name_len"; + goto bad_entry; +Espan: + error = "directory entry across blocks"; + goto bad_entry; +bad_entry: + EXOFS_ERR( + "ERROR [exofs_check_page]: bad entry in directory #%lu: %s - " + "offset=%lu, inode=%llu, rec_len=%d, name_len=%d", + dir->i_ino, error, (page->index<<PAGE_CACHE_SHIFT)+offs, + _LLU(le64_to_cpu(p->inode_no)), + rec_len, p->name_len); + goto fail; +Eend: + p = (struct exofs_dir_entry *)(kaddr + offs); + EXOFS_ERR("ERROR [exofs_check_page]: " + "entry in directory #%lu spans the page boundary" + "offset=%lu, inode=%llu", + dir->i_ino, (page->index<<PAGE_CACHE_SHIFT)+offs, + _LLU(le64_to_cpu(p->inode_no))); +fail: + SetPageChecked(page); + SetPageError(page); +} + +static struct page *exofs_get_page(struct inode *dir, unsigned long n) +{ + struct address_space *mapping = dir->i_mapping; + struct page *page = read_mapping_page(mapping, n, NULL); + + if (!IS_ERR(page)) { + kmap(page); + if (!PageChecked(page)) + exofs_check_page(page); + if (PageError(page)) + goto fail; + } + return page; + +fail: + exofs_put_page(page); + return ERR_PTR(-EIO); +} + +static inline int exofs_match(int len, const unsigned char *name, + struct exofs_dir_entry *de) +{ + if (len != de->name_len) + return 0; + if (!de->inode_no) + return 0; + return !memcmp(name, de->name, len); +} + +static inline +struct exofs_dir_entry *exofs_next_entry(struct exofs_dir_entry *p) +{ + return (struct exofs_dir_entry *)((char *)p + le16_to_cpu(p->rec_len)); +} + +static inline unsigned +exofs_validate_entry(char *base, unsigned offset, unsigned mask) +{ + struct exofs_dir_entry *de = (struct exofs_dir_entry *)(base + offset); + struct exofs_dir_entry *p = + (struct exofs_dir_entry *)(base + (offset&mask)); + while ((char *)p < (char *)de) { + if (p->rec_len == 0) + break; + p = exofs_next_entry(p); + } + return (char *)p - base; +} + +static unsigned char exofs_filetype_table[EXOFS_FT_MAX] = { + [EXOFS_FT_UNKNOWN] = DT_UNKNOWN, + [EXOFS_FT_REG_FILE] = DT_REG, + [EXOFS_FT_DIR] = DT_DIR, + [EXOFS_FT_CHRDEV] = DT_CHR, + [EXOFS_FT_BLKDEV] = DT_BLK, + [EXOFS_FT_FIFO] = DT_FIFO, + [EXOFS_FT_SOCK] = DT_SOCK, + [EXOFS_FT_SYMLINK] = DT_LNK, +}; + +#define S_SHIFT 12 +static unsigned char exofs_type_by_mode[S_IFMT >> S_SHIFT] = { + [S_IFREG >> S_SHIFT] = EXOFS_FT_REG_FILE, + [S_IFDIR >> S_SHIFT] = EXOFS_FT_DIR, + [S_IFCHR >> S_SHIFT] = EXOFS_FT_CHRDEV, + [S_IFBLK >> S_SHIFT] = EXOFS_FT_BLKDEV, + [S_IFIFO >> S_SHIFT] = EXOFS_FT_FIFO, + [S_IFSOCK >> S_SHIFT] = EXOFS_FT_SOCK, + [S_IFLNK >> S_SHIFT] = EXOFS_FT_SYMLINK, +}; + +static inline +void exofs_set_de_type(struct exofs_dir_entry *de, struct inode *inode) +{ + mode_t mode = inode->i_mode; + de->file_type = exofs_type_by_mode[(mode & S_IFMT) >> S_SHIFT]; +} + +static int +exofs_readdir(struct file *filp, void *dirent, filldir_t filldir) +{ + loff_t pos = filp->f_pos; + struct inode *inode = filp->f_path.dentry->d_inode; + unsigned int offset = pos & ~PAGE_CACHE_MASK; + unsigned long n = pos >> PAGE_CACHE_SHIFT; + unsigned long npages = dir_pages(inode); + unsigned chunk_mask = ~(exofs_chunk_size(inode)-1); + unsigned char *types = NULL; + int need_revalidate = (filp->f_version != inode->i_version); + + if (pos > inode->i_size - EXOFS_DIR_REC_LEN(1)) + return 0; + + types = exofs_filetype_table; + + for ( ; n < npages; n++, offset = 0) { + char *kaddr, *limit; + struct exofs_dir_entry *de; + struct page *page = exofs_get_page(inode, n); + + if (IS_ERR(page)) { + EXOFS_ERR("ERROR: " + "bad page in #%lu", + inode->i_ino); + filp->f_pos += PAGE_CACHE_SIZE - offset; + return PTR_ERR(page); + } + kaddr = page_address(page); + if (unlikely(need_revalidate)) { + if (offset) { + offset = exofs_validate_entry(kaddr, offset, + chunk_mask); + filp->f_pos = (n<<PAGE_CACHE_SHIFT) + offset; + } + filp->f_version = inode->i_version; + need_revalidate = 0; + } + de = (struct exofs_dir_entry *)(kaddr + offset); + limit = kaddr + exofs_last_byte(inode, n) - + EXOFS_DIR_REC_LEN(1); + for (; (char *)de <= limit; de = exofs_next_entry(de)) { + if (de->rec_len == 0) { + EXOFS_ERR("ERROR: " + "zero-length directory entry"); + exofs_put_page(page); + return -EIO; + } + if (de->inode_no) { + int over; + unsigned char d_type = DT_UNKNOWN; + + if (types && de->file_type < EXOFS_FT_MAX) + d_type = types[de->file_type]; + + offset = (char *)de - kaddr; + over = filldir(dirent, de->name, de->name_len, + (n<<PAGE_CACHE_SHIFT) | offset, + le64_to_cpu(de->inode_no), + d_type); + if (over) { + exofs_put_page(page); + return 0; + } + } + filp->f_pos += le16_to_cpu(de->rec_len); + } + exofs_put_page(page); + } + + return 0; +} + +struct exofs_dir_entry *exofs_find_entry(struct inode *dir, + struct dentry *dentry, struct page **res_page) +{ + const unsigned char *name = dentry->d_name.name; + int namelen = dentry->d_name.len; + unsigned reclen = EXOFS_DIR_REC_LEN(namelen); + unsigned long start, n; + unsigned long npages = dir_pages(dir); + struct page *page = NULL; + struct exofs_i_info *oi = exofs_i(dir); + struct exofs_dir_entry *de; + + if (npages == 0) + goto out; + + *res_page = NULL; + + start = oi->i_dir_start_lookup; + if (start >= npages) + start = 0; + n = start; + do { + char *kaddr; + page = exofs_get_page(dir, n); + if (!IS_ERR(page)) { + kaddr = page_address(page); + de = (struct exofs_dir_entry *) kaddr; + kaddr += exofs_last_byte(dir, n) - reclen; + while ((char *) de <= kaddr) { + if (de->rec_len == 0) { + EXOFS_ERR( + "ERROR: exofs_find_entry: " + "zero-length directory entry"); + exofs_put_page(page); + goto out; + } + if (exofs_match(namelen, name, de)) + goto found; + de = exofs_next_entry(de); + } + exofs_put_page(page); + } + if (++n >= npages) + n = 0; + } while (n != start); +out: + return NULL; + +found: + *res_page = page; + oi->i_dir_start_lookup = n; + return de; +} + +struct exofs_dir_entry *exofs_dotdot(struct inode *dir, struct page **p) +{ + struct page *page = exofs_get_page(dir, 0); + struct exofs_dir_entry *de = NULL; + + if (!IS_ERR(page)) { + de = exofs_next_entry( + (struct exofs_dir_entry *)page_address(page)); + *p = page; + } + return de; +} + +ino_t exofs_parent_ino(struct dentry *child) +{ + struct page *page; + struct exofs_dir_entry *de; + ino_t ino; + + de = exofs_dotdot(child->d_inode, &page); + if (!de) + return 0; + + ino = le64_to_cpu(de->inode_no); + exofs_put_page(page); + return ino; +} + +ino_t exofs_inode_by_name(struct inode *dir, struct dentry *dentry) +{ + ino_t res = 0; + struct exofs_dir_entry *de; + struct page *page; + + de = exofs_find_entry(dir, dentry, &page); + if (de) { + res = le64_to_cpu(de->inode_no); + exofs_put_page(page); + } + return res; +} + +int exofs_set_link(struct inode *dir, struct exofs_dir_entry *de, + struct page *page, struct inode *inode) +{ + loff_t pos = page_offset(page) + + (char *) de - (char *) page_address(page); + unsigned len = le16_to_cpu(de->rec_len); + int err; + + lock_page(page); + err = exofs_write_begin(NULL, page->mapping, pos, len, + AOP_FLAG_UNINTERRUPTIBLE, &page, NULL); + if (err) + EXOFS_ERR("exofs_set_link: exofs_write_begin FAILD => %d\n", + err); + + de->inode_no = cpu_to_le64(inode->i_ino); + exofs_set_de_type(de, inode); + if (likely(!err)) + err = exofs_commit_chunk(page, pos, len); + exofs_put_page(page); + dir->i_mtime = dir->i_ctime = CURRENT_TIME; + mark_inode_dirty(dir); + return err; +} + +int exofs_add_link(struct dentry *dentry, struct inode *inode) +{ + struct inode *dir = dentry->d_parent->d_inode; + const unsigned char *name = dentry->d_name.name; + int namelen = dentry->d_name.len; + unsigned chunk_size = exofs_chunk_size(dir); + unsigned reclen = EXOFS_DIR_REC_LEN(namelen); + unsigned short rec_len, name_len; + struct page *page = NULL; + struct exofs_sb_info *sbi = inode->i_sb->s_fs_info; + struct exofs_dir_entry *de; + unsigned long npages = dir_pages(dir); + unsigned long n; + char *kaddr; + loff_t pos; + int err; + + for (n = 0; n <= npages; n++) { + char *dir_end; + + page = exofs_get_page(dir, n); + err = PTR_ERR(page); + if (IS_ERR(page)) + goto out; + lock_page(page); + kaddr = page_address(page); + dir_end = kaddr + exofs_last_byte(dir, n); + de = (struct exofs_dir_entry *)kaddr; + kaddr += PAGE_CACHE_SIZE - reclen; + while ((char *)de <= kaddr) { + if ((char *)de == dir_end) { + name_len = 0; + rec_len = chunk_size; + de->rec_len = cpu_to_le16(chunk_size); + de->inode_no = 0; + goto got_it; + } + if (de->rec_len == 0) { + EXOFS_ERR("ERROR: exofs_add_link: " + "zero-length directory entry"); + err = -EIO; + goto out_unlock; + } + err = -EEXIST; + if (exofs_match(namelen, name, de)) + goto out_unlock; + name_len = EXOFS_DIR_REC_LEN(de->name_len); + rec_len = le16_to_cpu(de->rec_len); + if (!de->inode_no && rec_len >= reclen) + goto got_it; + if (rec_len >= name_len + reclen) + goto got_it; + de = (struct exofs_dir_entry *) ((char *) de + rec_len); + } + unlock_page(page); + exofs_put_page(page); + } + + EXOFS_ERR("exofs_add_link: BAD dentry=%p or inode=%p", dentry, inode); + return -EINVAL; + +got_it: + pos = page_offset(page) + + (char *)de - (char *)page_address(page); + err = exofs_write_begin(NULL, page->mapping, pos, rec_len, 0, + &page, NULL); + if (err) + goto out_unlock; + if (de->inode_no) { + struct exofs_dir_entry *de1 = + (struct exofs_dir_entry *)((char *)de + name_len); + de1->rec_len = cpu_to_le16(rec_len - name_len); + de->rec_len = cpu_to_le16(name_len); + de = de1; + } + de->name_len = namelen; + memcpy(de->name, name, namelen); + de->inode_no = cpu_to_le64(inode->i_ino); + exofs_set_de_type(de, inode); + err = exofs_commit_chunk(page, pos, rec_len); + dir->i_mtime = dir->i_ctime = CURRENT_TIME; + mark_inode_dirty(dir); + sbi->s_numfiles++; + +out_put: + exofs_put_page(page); +out: + return err; +out_unlock: + unlock_page(page); + goto out_put; +} + +int exofs_delete_entry(struct exofs_dir_entry *dir, struct page *page) +{ + struct address_space *mapping = page->mapping; + struct inode *inode = mapping->host; + struct exofs_sb_info *sbi = inode->i_sb->s_fs_info; + char *kaddr = page_address(page); + unsigned from = ((char *)dir - kaddr) & ~(exofs_chunk_size(inode)-1); + unsigned to = ((char *)dir - kaddr) + le16_to_cpu(dir->rec_len); + loff_t pos; + struct exofs_dir_entry *pde = NULL; + struct exofs_dir_entry *de = (struct exofs_dir_entry *) (kaddr + from); + int err; + + while (de < dir) { + if (de->rec_len == 0) { + EXOFS_ERR("ERROR: exofs_delete_entry:" + "zero-length directory entry"); + err = -EIO; + goto out; + } + pde = de; + de = exofs_next_entry(de); + } + if (pde) + from = (char *)pde - (char *)page_address(page); + pos = page_offset(page) + from; + lock_page(page); + err = exofs_write_begin(NULL, page->mapping, pos, to - from, 0, + &page, NULL); + if (err) + EXOFS_ERR("exofs_delete_entry: exofs_write_begin FAILD => %d\n", + err); + if (pde) + pde->rec_len = cpu_to_le16(to - from); + dir->inode_no = 0; + if (likely(!err)) + err = exofs_commit_chunk(page, pos, to - from); + inode->i_ctime = inode->i_mtime = CURRENT_TIME; + mark_inode_dirty(inode); + sbi->s_numfiles--; +out: + exofs_put_page(page); + return err; +} + +/* kept aligned on 4 bytes */ +#define THIS_DIR ".\0\0" +#define PARENT_DIR "..\0" + +int exofs_make_empty(struct inode *inode, struct inode *parent) +{ + struct address_space *mapping = inode->i_mapping; + struct page *page = grab_cache_page(mapping, 0); + unsigned chunk_size = exofs_chunk_size(inode); + struct exofs_dir_entry *de; + int err; + void *kaddr; + + if (!page) + return -ENOMEM; + + err = exofs_write_begin(NULL, page->mapping, 0, chunk_size, 0, + &page, NULL); + if (err) { + unlock_page(page); + goto fail; + } + + kaddr = kmap_atomic(page, KM_USER0); + de = (struct exofs_dir_entry *)kaddr; + de->name_len = 1; + de->rec_len = cpu_to_le16(EXOFS_DIR_REC_LEN(1)); + memcpy(de->name, THIS_DIR, sizeof(THIS_DIR)); + de->inode_no = cpu_to_le64(inode->i_ino); + exofs_set_de_type(de, inode); + + de = (struct exofs_dir_entry *)(kaddr + EXOFS_DIR_REC_LEN(1)); + de->name_len = 2; + de->rec_len = cpu_to_le16(chunk_size - EXOFS_DIR_REC_LEN(1)); + de->inode_no = cpu_to_le64(parent->i_ino); + memcpy(de->name, PARENT_DIR, sizeof(PARENT_DIR)); + exofs_set_de_type(de, inode); + kunmap_atomic(page, KM_USER0); + err = exofs_commit_chunk(page, 0, chunk_size); +fail: + page_cache_release(page); + return err; +} + +int exofs_empty_dir(struct inode *inode) +{ + struct page *page = NULL; + unsigned long i, npages = dir_pages(inode); + + for (i = 0; i < npages; i++) { + char *kaddr; + struct exofs_dir_entry *de; + page = exofs_get_page(inode, i); + + if (IS_ERR(page)) + continue; + + kaddr = page_address(page); + de = (struct exofs_dir_entry *)kaddr; + kaddr += exofs_last_byte(inode, i) - EXOFS_DIR_REC_LEN(1); + + while ((char *)de <= kaddr) { + if (de->rec_len == 0) { + EXOFS_ERR("ERROR: exofs_empty_dir: " + "zero-length directory entry" + "kaddr=%p, de=%p\n", kaddr, de); + goto not_empty; + } + if (de->inode_no != 0) { + /* check for . and .. */ + if (de->name[0] != '.') + goto not_empty; + if (de->name_len > 2) + goto not_empty; + if (de->name_len < 2) { + if (le64_to_cpu(de->inode_no) != + inode->i_ino) + goto not_empty; + } else if (de->name[1] != '.') + goto not_empty; + } + de = exofs_next_entry(de); + } + exofs_put_page(page); + } + return 1; + +not_empty: + exofs_put_page(page); + return 0; +} + +const struct file_operations exofs_dir_operations = { + .llseek = generic_file_llseek, + .read = generic_read_dir, + .readdir = exofs_readdir, +}; diff --git a/fs/exofs/exofs.h b/fs/exofs/exofs.h new file mode 100644 index 000000000000..0fd4c7859679 --- /dev/null +++ b/fs/exofs/exofs.h @@ -0,0 +1,180 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <linux/fs.h> +#include <linux/time.h> +#include "common.h" + +#ifndef __EXOFS_H__ +#define __EXOFS_H__ + +#define EXOFS_ERR(fmt, a...) printk(KERN_ERR "exofs: " fmt, ##a) + +#ifdef CONFIG_EXOFS_DEBUG +#define EXOFS_DBGMSG(fmt, a...) \ + printk(KERN_NOTICE "exofs @%s:%d: " fmt, __func__, __LINE__, ##a) +#else +#define EXOFS_DBGMSG(fmt, a...) \ + do { if (0) printk(fmt, ##a); } while (0) +#endif + +/* u64 has problems with printk this will cast it to unsigned long long */ +#define _LLU(x) (unsigned long long)(x) + +/* + * our extension to the in-memory superblock + */ +struct exofs_sb_info { + struct osd_dev *s_dev; /* returned by get_osd_dev */ + osd_id s_pid; /* partition ID of file system*/ + int s_timeout; /* timeout for OSD operations */ + uint64_t s_nextid; /* highest object ID used */ + uint32_t s_numfiles; /* number of files on fs */ + spinlock_t s_next_gen_lock; /* spinlock for gen # update */ + u32 s_next_generation; /* next gen # to use */ + atomic_t s_curr_pending; /* number of pending commands */ + uint8_t s_cred[OSD_CAP_LEN]; /* all-powerful credential */ +}; + +/* + * our extension to the in-memory inode + */ +struct exofs_i_info { + unsigned long i_flags; /* various atomic flags */ + uint32_t i_data[EXOFS_IDATA];/*short symlink names and device #s*/ + uint32_t i_dir_start_lookup; /* which page to start lookup */ + wait_queue_head_t i_wq; /* wait queue for inode */ + uint64_t i_commit_size; /* the object's written length */ + uint8_t i_cred[OSD_CAP_LEN];/* all-powerful credential */ + struct inode vfs_inode; /* normal in-memory inode */ +}; + +/* + * our inode flags + */ +#define OBJ_2BCREATED 0 /* object will be created soon*/ +#define OBJ_CREATED 1 /* object has been created on the osd*/ + +static inline int obj_2bcreated(struct exofs_i_info *oi) +{ + return test_bit(OBJ_2BCREATED, &oi->i_flags); +} + +static inline void set_obj_2bcreated(struct exofs_i_info *oi) +{ + set_bit(OBJ_2BCREATED, &oi->i_flags); +} + +static inline int obj_created(struct exofs_i_info *oi) +{ + return test_bit(OBJ_CREATED, &oi->i_flags); +} + +static inline void set_obj_created(struct exofs_i_info *oi) +{ + set_bit(OBJ_CREATED, &oi->i_flags); +} + +int __exofs_wait_obj_created(struct exofs_i_info *oi); +static inline int wait_obj_created(struct exofs_i_info *oi) +{ + if (likely(obj_created(oi))) + return 0; + + return __exofs_wait_obj_created(oi); +} + +/* + * get to our inode from the vfs inode + */ +static inline struct exofs_i_info *exofs_i(struct inode *inode) +{ + return container_of(inode, struct exofs_i_info, vfs_inode); +} + +/* + * Maximum count of links to a file + */ +#define EXOFS_LINK_MAX 32000 + +/************************* + * function declarations * + *************************/ +/* inode.c */ +void exofs_truncate(struct inode *inode); +int exofs_setattr(struct dentry *, struct iattr *); +int exofs_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata); +extern struct inode *exofs_iget(struct super_block *, unsigned long); +struct inode *exofs_new_inode(struct inode *, int); +extern int exofs_write_inode(struct inode *, int); +extern void exofs_delete_inode(struct inode *); + +/* dir.c: */ +int exofs_add_link(struct dentry *, struct inode *); +ino_t exofs_inode_by_name(struct inode *, struct dentry *); +int exofs_delete_entry(struct exofs_dir_entry *, struct page *); +int exofs_make_empty(struct inode *, struct inode *); +struct exofs_dir_entry *exofs_find_entry(struct inode *, struct dentry *, + struct page **); +int exofs_empty_dir(struct inode *); +struct exofs_dir_entry *exofs_dotdot(struct inode *, struct page **); +ino_t exofs_parent_ino(struct dentry *child); +int exofs_set_link(struct inode *, struct exofs_dir_entry *, struct page *, + struct inode *); + +/********************* + * operation vectors * + *********************/ +/* dir.c: */ +extern const struct file_operations exofs_dir_operations; + +/* file.c */ +extern const struct inode_operations exofs_file_inode_operations; +extern const struct file_operations exofs_file_operations; + +/* inode.c */ +extern const struct address_space_operations exofs_aops; + +/* namei.c */ +extern const struct inode_operations exofs_dir_inode_operations; +extern const struct inode_operations exofs_special_inode_operations; + +/* symlink.c */ +extern const struct inode_operations exofs_symlink_inode_operations; +extern const struct inode_operations exofs_fast_symlink_inode_operations; + +#endif diff --git a/fs/exofs/file.c b/fs/exofs/file.c new file mode 100644 index 000000000000..6ed7fe484752 --- /dev/null +++ b/fs/exofs/file.c @@ -0,0 +1,87 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <linux/buffer_head.h> + +#include "exofs.h" + +static int exofs_release_file(struct inode *inode, struct file *filp) +{ + return 0; +} + +static int exofs_file_fsync(struct file *filp, struct dentry *dentry, + int datasync) +{ + int ret; + struct address_space *mapping = filp->f_mapping; + + ret = filemap_write_and_wait(mapping); + if (ret) + return ret; + + /*Note: file_fsync below also calles sync_blockdev, which is a no-op + * for exofs, but other then that it does sync_inode and + * sync_superblock which is what we need here. + */ + return file_fsync(filp, dentry, datasync); +} + +static int exofs_flush(struct file *file, fl_owner_t id) +{ + exofs_file_fsync(file, file->f_path.dentry, 1); + /* TODO: Flush the OSD target */ + return 0; +} + +const struct file_operations exofs_file_operations = { + .llseek = generic_file_llseek, + .read = do_sync_read, + .write = do_sync_write, + .aio_read = generic_file_aio_read, + .aio_write = generic_file_aio_write, + .mmap = generic_file_mmap, + .open = generic_file_open, + .release = exofs_release_file, + .fsync = exofs_file_fsync, + .flush = exofs_flush, + .splice_read = generic_file_splice_read, + .splice_write = generic_file_splice_write, +}; + +const struct inode_operations exofs_file_inode_operations = { + .truncate = exofs_truncate, + .setattr = exofs_setattr, +}; diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c new file mode 100644 index 000000000000..ba8d9fab4693 --- /dev/null +++ b/fs/exofs/inode.c @@ -0,0 +1,1303 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <linux/writeback.h> +#include <linux/buffer_head.h> +#include <scsi/scsi_device.h> + +#include "exofs.h" + +#ifdef CONFIG_EXOFS_DEBUG +# define EXOFS_DEBUG_OBJ_ISIZE 1 +#endif + +struct page_collect { + struct exofs_sb_info *sbi; + struct request_queue *req_q; + struct inode *inode; + unsigned expected_pages; + + struct bio *bio; + unsigned nr_pages; + unsigned long length; + loff_t pg_first; /* keep 64bit also in 32-arches */ +}; + +static void _pcol_init(struct page_collect *pcol, unsigned expected_pages, + struct inode *inode) +{ + struct exofs_sb_info *sbi = inode->i_sb->s_fs_info; + struct request_queue *req_q = sbi->s_dev->scsi_device->request_queue; + + pcol->sbi = sbi; + pcol->req_q = req_q; + pcol->inode = inode; + pcol->expected_pages = expected_pages; + + pcol->bio = NULL; + pcol->nr_pages = 0; + pcol->length = 0; + pcol->pg_first = -1; + + EXOFS_DBGMSG("_pcol_init ino=0x%lx expected_pages=%u\n", inode->i_ino, + expected_pages); +} + +static void _pcol_reset(struct page_collect *pcol) +{ + pcol->expected_pages -= min(pcol->nr_pages, pcol->expected_pages); + + pcol->bio = NULL; + pcol->nr_pages = 0; + pcol->length = 0; + pcol->pg_first = -1; + EXOFS_DBGMSG("_pcol_reset ino=0x%lx expected_pages=%u\n", + pcol->inode->i_ino, pcol->expected_pages); + + /* this is probably the end of the loop but in writes + * it might not end here. don't be left with nothing + */ + if (!pcol->expected_pages) + pcol->expected_pages = 128; +} + +static int pcol_try_alloc(struct page_collect *pcol) +{ + int pages = min_t(unsigned, pcol->expected_pages, BIO_MAX_PAGES); + + for (; pages; pages >>= 1) { + pcol->bio = bio_alloc(GFP_KERNEL, pages); + if (likely(pcol->bio)) + return 0; + } + + EXOFS_ERR("Failed to kcalloc expected_pages=%u\n", + pcol->expected_pages); + return -ENOMEM; +} + +static void pcol_free(struct page_collect *pcol) +{ + bio_put(pcol->bio); + pcol->bio = NULL; +} + +static int pcol_add_page(struct page_collect *pcol, struct page *page, + unsigned len) +{ + int added_len = bio_add_pc_page(pcol->req_q, pcol->bio, page, len, 0); + if (unlikely(len != added_len)) + return -ENOMEM; + + ++pcol->nr_pages; + pcol->length += len; + return 0; +} + +static int update_read_page(struct page *page, int ret) +{ + if (ret == 0) { + /* Everything is OK */ + SetPageUptodate(page); + if (PageError(page)) + ClearPageError(page); + } else if (ret == -EFAULT) { + /* In this case we were trying to read something that wasn't on + * disk yet - return a page full of zeroes. This should be OK, + * because the object should be empty (if there was a write + * before this read, the read would be waiting with the page + * locked */ + clear_highpage(page); + + SetPageUptodate(page); + if (PageError(page)) + ClearPageError(page); + ret = 0; /* recovered error */ + EXOFS_DBGMSG("recovered read error\n"); + } else /* Error */ + SetPageError(page); + + return ret; +} + +static void update_write_page(struct page *page, int ret) +{ + if (ret) { + mapping_set_error(page->mapping, ret); + SetPageError(page); + } + end_page_writeback(page); +} + +/* Called at the end of reads, to optionally unlock pages and update their + * status. + */ +static int __readpages_done(struct osd_request *or, struct page_collect *pcol, + bool do_unlock) +{ + struct bio_vec *bvec; + int i; + u64 resid; + u64 good_bytes; + u64 length = 0; + int ret = exofs_check_ok_resid(or, &resid, NULL); + + osd_end_request(or); + + if (likely(!ret)) + good_bytes = pcol->length; + else if (!resid) + good_bytes = 0; + else + good_bytes = pcol->length - resid; + + EXOFS_DBGMSG("readpages_done(0x%lx) good_bytes=0x%llx" + " length=0x%lx nr_pages=%u\n", + pcol->inode->i_ino, _LLU(good_bytes), pcol->length, + pcol->nr_pages); + + __bio_for_each_segment(bvec, pcol->bio, i, 0) { + struct page *page = bvec->bv_page; + struct inode *inode = page->mapping->host; + int page_stat; + + if (inode != pcol->inode) + continue; /* osd might add more pages at end */ + + if (likely(length < good_bytes)) + page_stat = 0; + else + page_stat = ret; + + EXOFS_DBGMSG(" readpages_done(0x%lx, 0x%lx) %s\n", + inode->i_ino, page->index, + page_stat ? "bad_bytes" : "good_bytes"); + + ret = update_read_page(page, page_stat); + if (do_unlock) + unlock_page(page); + length += bvec->bv_len; + } + + pcol_free(pcol); + EXOFS_DBGMSG("readpages_done END\n"); + return ret; +} + +/* callback of async reads */ +static void readpages_done(struct osd_request *or, void *p) +{ + struct page_collect *pcol = p; + + __readpages_done(or, pcol, true); + atomic_dec(&pcol->sbi->s_curr_pending); + kfree(p); +} + +static void _unlock_pcol_pages(struct page_collect *pcol, int ret, int rw) +{ + struct bio_vec *bvec; + int i; + + __bio_for_each_segment(bvec, pcol->bio, i, 0) { + struct page *page = bvec->bv_page; + + if (rw == READ) + update_read_page(page, ret); + else + update_write_page(page, ret); + + unlock_page(page); + } + pcol_free(pcol); +} + +static int read_exec(struct page_collect *pcol, bool is_sync) +{ + struct exofs_i_info *oi = exofs_i(pcol->inode); + struct osd_obj_id obj = {pcol->sbi->s_pid, + pcol->inode->i_ino + EXOFS_OBJ_OFF}; + struct osd_request *or = NULL; + struct page_collect *pcol_copy = NULL; + loff_t i_start = pcol->pg_first << PAGE_CACHE_SHIFT; + int ret; + + if (!pcol->bio) + return 0; + + /* see comment in _readpage() about sync reads */ + WARN_ON(is_sync && (pcol->nr_pages != 1)); + + or = osd_start_request(pcol->sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + ret = -ENOMEM; + goto err; + } + + osd_req_read(or, &obj, pcol->bio, i_start); + + if (is_sync) { + exofs_sync_op(or, pcol->sbi->s_timeout, oi->i_cred); + return __readpages_done(or, pcol, false); + } + + pcol_copy = kmalloc(sizeof(*pcol_copy), GFP_KERNEL); + if (!pcol_copy) { + ret = -ENOMEM; + goto err; + } + + *pcol_copy = *pcol; + ret = exofs_async_op(or, readpages_done, pcol_copy, oi->i_cred); + if (unlikely(ret)) + goto err; + + atomic_inc(&pcol->sbi->s_curr_pending); + + EXOFS_DBGMSG("read_exec obj=0x%llx start=0x%llx length=0x%lx\n", + obj.id, _LLU(i_start), pcol->length); + + /* pages ownership was passed to pcol_copy */ + _pcol_reset(pcol); + return 0; + +err: + if (!is_sync) + _unlock_pcol_pages(pcol, ret, READ); + kfree(pcol_copy); + if (or) + osd_end_request(or); + return ret; +} + +/* readpage_strip is called either directly from readpage() or by the VFS from + * within read_cache_pages(), to add one more page to be read. It will try to + * collect as many contiguous pages as posible. If a discontinuity is + * encountered, or it runs out of resources, it will submit the previous segment + * and will start a new collection. Eventually caller must submit the last + * segment if present. + */ +static int readpage_strip(void *data, struct page *page) +{ + struct page_collect *pcol = data; + struct inode *inode = pcol->inode; + struct exofs_i_info *oi = exofs_i(inode); + loff_t i_size = i_size_read(inode); + pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT; + size_t len; + int ret; + + /* FIXME: Just for debugging, will be removed */ + if (PageUptodate(page)) + EXOFS_ERR("PageUptodate(0x%lx, 0x%lx)\n", pcol->inode->i_ino, + page->index); + + if (page->index < end_index) + len = PAGE_CACHE_SIZE; + else if (page->index == end_index) + len = i_size & ~PAGE_CACHE_MASK; + else + len = 0; + + if (!len || !obj_created(oi)) { + /* this will be out of bounds, or doesn't exist yet. + * Current page is cleared and the request is split + */ + clear_highpage(page); + + SetPageUptodate(page); + if (PageError(page)) + ClearPageError(page); + + unlock_page(page); + EXOFS_DBGMSG("readpage_strip(0x%lx, 0x%lx) empty page," + " splitting\n", inode->i_ino, page->index); + + return read_exec(pcol, false); + } + +try_again: + + if (unlikely(pcol->pg_first == -1)) { + pcol->pg_first = page->index; + } else if (unlikely((pcol->pg_first + pcol->nr_pages) != + page->index)) { + /* Discontinuity detected, split the request */ + ret = read_exec(pcol, false); + if (unlikely(ret)) + goto fail; + goto try_again; + } + + if (!pcol->bio) { + ret = pcol_try_alloc(pcol); + if (unlikely(ret)) + goto fail; + } + + if (len != PAGE_CACHE_SIZE) + zero_user(page, len, PAGE_CACHE_SIZE - len); + + EXOFS_DBGMSG(" readpage_strip(0x%lx, 0x%lx) len=0x%zx\n", + inode->i_ino, page->index, len); + + ret = pcol_add_page(pcol, page, len); + if (ret) { + EXOFS_DBGMSG("Failed pcol_add_page pages[i]=%p " + "this_len=0x%zx nr_pages=%u length=0x%lx\n", + page, len, pcol->nr_pages, pcol->length); + + /* split the request, and start again with current page */ + ret = read_exec(pcol, false); + if (unlikely(ret)) + goto fail; + + goto try_again; + } + + return 0; + +fail: + /* SetPageError(page); ??? */ + unlock_page(page); + return ret; +} + +static int exofs_readpages(struct file *file, struct address_space *mapping, + struct list_head *pages, unsigned nr_pages) +{ + struct page_collect pcol; + int ret; + + _pcol_init(&pcol, nr_pages, mapping->host); + + ret = read_cache_pages(mapping, pages, readpage_strip, &pcol); + if (ret) { + EXOFS_ERR("read_cache_pages => %d\n", ret); + return ret; + } + + return read_exec(&pcol, false); +} + +static int _readpage(struct page *page, bool is_sync) +{ + struct page_collect pcol; + int ret; + + _pcol_init(&pcol, 1, page->mapping->host); + + /* readpage_strip might call read_exec(,async) inside at several places + * but this is safe for is_async=0 since read_exec will not do anything + * when we have a single page. + */ + ret = readpage_strip(&pcol, page); + if (ret) { + EXOFS_ERR("_readpage => %d\n", ret); + return ret; + } + + return read_exec(&pcol, is_sync); +} + +/* + * We don't need the file + */ +static int exofs_readpage(struct file *file, struct page *page) +{ + return _readpage(page, false); +} + +/* Callback for osd_write. All writes are asynchronouse */ +static void writepages_done(struct osd_request *or, void *p) +{ + struct page_collect *pcol = p; + struct bio_vec *bvec; + int i; + u64 resid; + u64 good_bytes; + u64 length = 0; + + int ret = exofs_check_ok_resid(or, NULL, &resid); + + osd_end_request(or); + atomic_dec(&pcol->sbi->s_curr_pending); + + if (likely(!ret)) + good_bytes = pcol->length; + else if (!resid) + good_bytes = 0; + else + good_bytes = pcol->length - resid; + + EXOFS_DBGMSG("writepages_done(0x%lx) good_bytes=0x%llx" + " length=0x%lx nr_pages=%u\n", + pcol->inode->i_ino, _LLU(good_bytes), pcol->length, + pcol->nr_pages); + + __bio_for_each_segment(bvec, pcol->bio, i, 0) { + struct page *page = bvec->bv_page; + struct inode *inode = page->mapping->host; + int page_stat; + + if (inode != pcol->inode) + continue; /* osd might add more pages to a bio */ + + if (likely(length < good_bytes)) + page_stat = 0; + else + page_stat = ret; + + update_write_page(page, page_stat); + unlock_page(page); + EXOFS_DBGMSG(" writepages_done(0x%lx, 0x%lx) status=%d\n", + inode->i_ino, page->index, page_stat); + + length += bvec->bv_len; + } + + pcol_free(pcol); + kfree(pcol); + EXOFS_DBGMSG("writepages_done END\n"); +} + +static int write_exec(struct page_collect *pcol) +{ + struct exofs_i_info *oi = exofs_i(pcol->inode); + struct osd_obj_id obj = {pcol->sbi->s_pid, + pcol->inode->i_ino + EXOFS_OBJ_OFF}; + struct osd_request *or = NULL; + struct page_collect *pcol_copy = NULL; + loff_t i_start = pcol->pg_first << PAGE_CACHE_SHIFT; + int ret; + + if (!pcol->bio) + return 0; + + or = osd_start_request(pcol->sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("write_exec: Faild to osd_start_request()\n"); + ret = -ENOMEM; + goto err; + } + + pcol_copy = kmalloc(sizeof(*pcol_copy), GFP_KERNEL); + if (!pcol_copy) { + EXOFS_ERR("write_exec: Faild to kmalloc(pcol)\n"); + ret = -ENOMEM; + goto err; + } + + *pcol_copy = *pcol; + + osd_req_write(or, &obj, pcol_copy->bio, i_start); + ret = exofs_async_op(or, writepages_done, pcol_copy, oi->i_cred); + if (unlikely(ret)) { + EXOFS_ERR("write_exec: exofs_async_op() Faild\n"); + goto err; + } + + atomic_inc(&pcol->sbi->s_curr_pending); + EXOFS_DBGMSG("write_exec(0x%lx, 0x%llx) start=0x%llx length=0x%lx\n", + pcol->inode->i_ino, pcol->pg_first, _LLU(i_start), + pcol->length); + /* pages ownership was passed to pcol_copy */ + _pcol_reset(pcol); + return 0; + +err: + _unlock_pcol_pages(pcol, ret, WRITE); + kfree(pcol_copy); + if (or) + osd_end_request(or); + return ret; +} + +/* writepage_strip is called either directly from writepage() or by the VFS from + * within write_cache_pages(), to add one more page to be written to storage. + * It will try to collect as many contiguous pages as possible. If a + * discontinuity is encountered or it runs out of resources it will submit the + * previous segment and will start a new collection. + * Eventually caller must submit the last segment if present. + */ +static int writepage_strip(struct page *page, + struct writeback_control *wbc_unused, void *data) +{ + struct page_collect *pcol = data; + struct inode *inode = pcol->inode; + struct exofs_i_info *oi = exofs_i(inode); + loff_t i_size = i_size_read(inode); + pgoff_t end_index = i_size >> PAGE_CACHE_SHIFT; + size_t len; + int ret; + + BUG_ON(!PageLocked(page)); + + ret = wait_obj_created(oi); + if (unlikely(ret)) + goto fail; + + if (page->index < end_index) + /* in this case, the page is within the limits of the file */ + len = PAGE_CACHE_SIZE; + else { + len = i_size & ~PAGE_CACHE_MASK; + + if (page->index > end_index || !len) { + /* in this case, the page is outside the limits + * (truncate in progress) + */ + ret = write_exec(pcol); + if (unlikely(ret)) + goto fail; + if (PageError(page)) + ClearPageError(page); + unlock_page(page); + return 0; + } + } + +try_again: + + if (unlikely(pcol->pg_first == -1)) { + pcol->pg_first = page->index; + } else if (unlikely((pcol->pg_first + pcol->nr_pages) != + page->index)) { + /* Discontinuity detected, split the request */ + ret = write_exec(pcol); + if (unlikely(ret)) + goto fail; + goto try_again; + } + + if (!pcol->bio) { + ret = pcol_try_alloc(pcol); + if (unlikely(ret)) + goto fail; + } + + EXOFS_DBGMSG(" writepage_strip(0x%lx, 0x%lx) len=0x%zx\n", + inode->i_ino, page->index, len); + + ret = pcol_add_page(pcol, page, len); + if (unlikely(ret)) { + EXOFS_DBGMSG("Failed pcol_add_page " + "nr_pages=%u total_length=0x%lx\n", + pcol->nr_pages, pcol->length); + + /* split the request, next loop will start again */ + ret = write_exec(pcol); + if (unlikely(ret)) { + EXOFS_DBGMSG("write_exec faild => %d", ret); + goto fail; + } + + goto try_again; + } + + BUG_ON(PageWriteback(page)); + set_page_writeback(page); + + return 0; + +fail: + set_bit(AS_EIO, &page->mapping->flags); + unlock_page(page); + return ret; +} + +static int exofs_writepages(struct address_space *mapping, + struct writeback_control *wbc) +{ + struct page_collect pcol; + long start, end, expected_pages; + int ret; + + start = wbc->range_start >> PAGE_CACHE_SHIFT; + end = (wbc->range_end == LLONG_MAX) ? + start + mapping->nrpages : + wbc->range_end >> PAGE_CACHE_SHIFT; + + if (start || end) + expected_pages = min(end - start + 1, 32L); + else + expected_pages = mapping->nrpages; + + EXOFS_DBGMSG("inode(0x%lx) wbc->start=0x%llx wbc->end=0x%llx" + " m->nrpages=%lu start=0x%lx end=0x%lx\n", + mapping->host->i_ino, wbc->range_start, wbc->range_end, + mapping->nrpages, start, end); + + _pcol_init(&pcol, expected_pages, mapping->host); + + ret = write_cache_pages(mapping, wbc, writepage_strip, &pcol); + if (ret) { + EXOFS_ERR("write_cache_pages => %d\n", ret); + return ret; + } + + return write_exec(&pcol); +} + +static int exofs_writepage(struct page *page, struct writeback_control *wbc) +{ + struct page_collect pcol; + int ret; + + _pcol_init(&pcol, 1, page->mapping->host); + + ret = writepage_strip(page, NULL, &pcol); + if (ret) { + EXOFS_ERR("exofs_writepage => %d\n", ret); + return ret; + } + + return write_exec(&pcol); +} + +int exofs_write_begin(struct file *file, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) +{ + int ret = 0; + struct page *page; + + page = *pagep; + if (page == NULL) { + ret = simple_write_begin(file, mapping, pos, len, flags, pagep, + fsdata); + if (ret) { + EXOFS_DBGMSG("simple_write_begin faild\n"); + return ret; + } + + page = *pagep; + } + + /* read modify write */ + if (!PageUptodate(page) && (len != PAGE_CACHE_SIZE)) { + ret = _readpage(page, true); + if (ret) { + /*SetPageError was done by _readpage. Is it ok?*/ + unlock_page(page); + EXOFS_DBGMSG("__readpage_filler faild\n"); + } + } + + return ret; +} + +static int exofs_write_begin_export(struct file *file, + struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) +{ + *pagep = NULL; + + return exofs_write_begin(file, mapping, pos, len, flags, pagep, + fsdata); +} + +const struct address_space_operations exofs_aops = { + .readpage = exofs_readpage, + .readpages = exofs_readpages, + .writepage = exofs_writepage, + .writepages = exofs_writepages, + .write_begin = exofs_write_begin_export, + .write_end = simple_write_end, +}; + +/****************************************************************************** + * INODE OPERATIONS + *****************************************************************************/ + +/* + * Test whether an inode is a fast symlink. + */ +static inline int exofs_inode_is_fast_symlink(struct inode *inode) +{ + struct exofs_i_info *oi = exofs_i(inode); + + return S_ISLNK(inode->i_mode) && (oi->i_data[0] != 0); +} + +/* + * get_block_t - Fill in a buffer_head + * An OSD takes care of block allocation so we just fake an allocation by + * putting in the inode's sector_t in the buffer_head. + * TODO: What about the case of create==0 and @iblock does not exist in the + * object? + */ +static int exofs_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + map_bh(bh_result, inode->i_sb, iblock); + return 0; +} + +const struct osd_attr g_attr_logical_length = ATTR_DEF( + OSD_APAGE_OBJECT_INFORMATION, OSD_ATTR_OI_LOGICAL_LENGTH, 8); + +/* + * Truncate a file to the specified size - all we have to do is set the size + * attribute. We make sure the object exists first. + */ +void exofs_truncate(struct inode *inode) +{ + struct exofs_sb_info *sbi = inode->i_sb->s_fs_info; + struct exofs_i_info *oi = exofs_i(inode); + struct osd_obj_id obj = {sbi->s_pid, inode->i_ino + EXOFS_OBJ_OFF}; + struct osd_request *or; + struct osd_attr attr; + loff_t isize = i_size_read(inode); + __be64 newsize; + int ret; + + if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) + || S_ISLNK(inode->i_mode))) + return; + if (exofs_inode_is_fast_symlink(inode)) + return; + if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) + return; + inode->i_mtime = inode->i_ctime = CURRENT_TIME; + + nobh_truncate_page(inode->i_mapping, isize, exofs_get_block); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("ERROR: exofs_truncate: osd_start_request failed\n"); + goto fail; + } + + osd_req_set_attributes(or, &obj); + + newsize = cpu_to_be64((u64)isize); + attr = g_attr_logical_length; + attr.val_ptr = &newsize; + osd_req_add_set_attr_list(or, &attr, 1); + + /* if we are about to truncate an object, and it hasn't been + * created yet, wait + */ + if (unlikely(wait_obj_created(oi))) + goto fail; + + ret = exofs_sync_op(or, sbi->s_timeout, oi->i_cred); + osd_end_request(or); + if (ret) + goto fail; + +out: + mark_inode_dirty(inode); + return; +fail: + make_bad_inode(inode); + goto out; +} + +/* + * Set inode attributes - just call generic functions. + */ +int exofs_setattr(struct dentry *dentry, struct iattr *iattr) +{ + struct inode *inode = dentry->d_inode; + int error; + + error = inode_change_ok(inode, iattr); + if (error) + return error; + + error = inode_setattr(inode, iattr); + return error; +} + +/* + * Read an inode from the OSD, and return it as is. We also return the size + * attribute in the 'sanity' argument if we got compiled with debugging turned + * on. + */ +static int exofs_get_inode(struct super_block *sb, struct exofs_i_info *oi, + struct exofs_fcb *inode, uint64_t *sanity) +{ + struct exofs_sb_info *sbi = sb->s_fs_info; + struct osd_request *or; + struct osd_attr attr; + struct osd_obj_id obj = {sbi->s_pid, + oi->vfs_inode.i_ino + EXOFS_OBJ_OFF}; + int ret; + + exofs_make_credential(oi->i_cred, &obj); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("exofs_get_inode: osd_start_request failed.\n"); + return -ENOMEM; + } + osd_req_get_attributes(or, &obj); + + /* we need the inode attribute */ + osd_req_add_get_attr_list(or, &g_attr_inode_data, 1); + +#ifdef EXOFS_DEBUG_OBJ_ISIZE + /* we get the size attributes to do a sanity check */ + osd_req_add_get_attr_list(or, &g_attr_logical_length, 1); +#endif + + ret = exofs_sync_op(or, sbi->s_timeout, oi->i_cred); + if (ret) + goto out; + + attr = g_attr_inode_data; + ret = extract_attr_from_req(or, &attr); + if (ret) { + EXOFS_ERR("exofs_get_inode: extract_attr_from_req failed\n"); + goto out; + } + + WARN_ON(attr.len != EXOFS_INO_ATTR_SIZE); + memcpy(inode, attr.val_ptr, EXOFS_INO_ATTR_SIZE); + +#ifdef EXOFS_DEBUG_OBJ_ISIZE + attr = g_attr_logical_length; + ret = extract_attr_from_req(or, &attr); + if (ret) { + EXOFS_ERR("ERROR: extract attr from or failed\n"); + goto out; + } + *sanity = get_unaligned_be64(attr.val_ptr); +#endif + +out: + osd_end_request(or); + return ret; +} + +/* + * Fill in an inode read from the OSD and set it up for use + */ +struct inode *exofs_iget(struct super_block *sb, unsigned long ino) +{ + struct exofs_i_info *oi; + struct exofs_fcb fcb; + struct inode *inode; + uint64_t uninitialized_var(sanity); + int ret; + + inode = iget_locked(sb, ino); + if (!inode) + return ERR_PTR(-ENOMEM); + if (!(inode->i_state & I_NEW)) + return inode; + oi = exofs_i(inode); + + /* read the inode from the osd */ + ret = exofs_get_inode(sb, oi, &fcb, &sanity); + if (ret) + goto bad_inode; + + init_waitqueue_head(&oi->i_wq); + set_obj_created(oi); + + /* copy stuff from on-disk struct to in-memory struct */ + inode->i_mode = le16_to_cpu(fcb.i_mode); + inode->i_uid = le32_to_cpu(fcb.i_uid); + inode->i_gid = le32_to_cpu(fcb.i_gid); + inode->i_nlink = le16_to_cpu(fcb.i_links_count); + inode->i_ctime.tv_sec = (signed)le32_to_cpu(fcb.i_ctime); + inode->i_atime.tv_sec = (signed)le32_to_cpu(fcb.i_atime); + inode->i_mtime.tv_sec = (signed)le32_to_cpu(fcb.i_mtime); + inode->i_ctime.tv_nsec = + inode->i_atime.tv_nsec = inode->i_mtime.tv_nsec = 0; + oi->i_commit_size = le64_to_cpu(fcb.i_size); + i_size_write(inode, oi->i_commit_size); + inode->i_blkbits = EXOFS_BLKSHIFT; + inode->i_generation = le32_to_cpu(fcb.i_generation); + +#ifdef EXOFS_DEBUG_OBJ_ISIZE + if ((inode->i_size != sanity) && + (!exofs_inode_is_fast_symlink(inode))) { + EXOFS_ERR("WARNING: Size of object from inode and " + "attributes differ (%lld != %llu)\n", + inode->i_size, _LLU(sanity)); + } +#endif + + oi->i_dir_start_lookup = 0; + + if ((inode->i_nlink == 0) && (inode->i_mode == 0)) { + ret = -ESTALE; + goto bad_inode; + } + + if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) { + if (fcb.i_data[0]) + inode->i_rdev = + old_decode_dev(le32_to_cpu(fcb.i_data[0])); + else + inode->i_rdev = + new_decode_dev(le32_to_cpu(fcb.i_data[1])); + } else { + memcpy(oi->i_data, fcb.i_data, sizeof(fcb.i_data)); + } + + if (S_ISREG(inode->i_mode)) { + inode->i_op = &exofs_file_inode_operations; + inode->i_fop = &exofs_file_operations; + inode->i_mapping->a_ops = &exofs_aops; + } else if (S_ISDIR(inode->i_mode)) { + inode->i_op = &exofs_dir_inode_operations; + inode->i_fop = &exofs_dir_operations; + inode->i_mapping->a_ops = &exofs_aops; + } else if (S_ISLNK(inode->i_mode)) { + if (exofs_inode_is_fast_symlink(inode)) + inode->i_op = &exofs_fast_symlink_inode_operations; + else { + inode->i_op = &exofs_symlink_inode_operations; + inode->i_mapping->a_ops = &exofs_aops; + } + } else { + inode->i_op = &exofs_special_inode_operations; + if (fcb.i_data[0]) + init_special_inode(inode, inode->i_mode, + old_decode_dev(le32_to_cpu(fcb.i_data[0]))); + else + init_special_inode(inode, inode->i_mode, + new_decode_dev(le32_to_cpu(fcb.i_data[1]))); + } + + unlock_new_inode(inode); + return inode; + +bad_inode: + iget_failed(inode); + return ERR_PTR(ret); +} + +int __exofs_wait_obj_created(struct exofs_i_info *oi) +{ + if (!obj_created(oi)) { + BUG_ON(!obj_2bcreated(oi)); + wait_event(oi->i_wq, obj_created(oi)); + } + return unlikely(is_bad_inode(&oi->vfs_inode)) ? -EIO : 0; +} +/* + * Callback function from exofs_new_inode(). The important thing is that we + * set the obj_created flag so that other methods know that the object exists on + * the OSD. + */ +static void create_done(struct osd_request *or, void *p) +{ + struct inode *inode = p; + struct exofs_i_info *oi = exofs_i(inode); + struct exofs_sb_info *sbi = inode->i_sb->s_fs_info; + int ret; + + ret = exofs_check_ok(or); + osd_end_request(or); + atomic_dec(&sbi->s_curr_pending); + + if (unlikely(ret)) { + EXOFS_ERR("object=0x%llx creation faild in pid=0x%llx", + _LLU(sbi->s_pid), _LLU(inode->i_ino + EXOFS_OBJ_OFF)); + make_bad_inode(inode); + } else + set_obj_created(oi); + + atomic_dec(&inode->i_count); + wake_up(&oi->i_wq); +} + +/* + * Set up a new inode and create an object for it on the OSD + */ +struct inode *exofs_new_inode(struct inode *dir, int mode) +{ + struct super_block *sb; + struct inode *inode; + struct exofs_i_info *oi; + struct exofs_sb_info *sbi; + struct osd_request *or; + struct osd_obj_id obj; + int ret; + + sb = dir->i_sb; + inode = new_inode(sb); + if (!inode) + return ERR_PTR(-ENOMEM); + + oi = exofs_i(inode); + + init_waitqueue_head(&oi->i_wq); + set_obj_2bcreated(oi); + + sbi = sb->s_fs_info; + + sb->s_dirt = 1; + inode->i_uid = current->cred->fsuid; + if (dir->i_mode & S_ISGID) { + inode->i_gid = dir->i_gid; + if (S_ISDIR(mode)) + mode |= S_ISGID; + } else { + inode->i_gid = current->cred->fsgid; + } + inode->i_mode = mode; + + inode->i_ino = sbi->s_nextid++; + inode->i_blkbits = EXOFS_BLKSHIFT; + inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; + oi->i_commit_size = inode->i_size = 0; + spin_lock(&sbi->s_next_gen_lock); + inode->i_generation = sbi->s_next_generation++; + spin_unlock(&sbi->s_next_gen_lock); + insert_inode_hash(inode); + + mark_inode_dirty(inode); + + obj.partition = sbi->s_pid; + obj.id = inode->i_ino + EXOFS_OBJ_OFF; + exofs_make_credential(oi->i_cred, &obj); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("exofs_new_inode: osd_start_request failed\n"); + return ERR_PTR(-ENOMEM); + } + + osd_req_create_object(or, &obj); + + /* increment the refcount so that the inode will still be around when we + * reach the callback + */ + atomic_inc(&inode->i_count); + + ret = exofs_async_op(or, create_done, inode, oi->i_cred); + if (ret) { + atomic_dec(&inode->i_count); + osd_end_request(or); + return ERR_PTR(-EIO); + } + atomic_inc(&sbi->s_curr_pending); + + return inode; +} + +/* + * struct to pass two arguments to update_inode's callback + */ +struct updatei_args { + struct exofs_sb_info *sbi; + struct exofs_fcb fcb; +}; + +/* + * Callback function from exofs_update_inode(). + */ +static void updatei_done(struct osd_request *or, void *p) +{ + struct updatei_args *args = p; + + osd_end_request(or); + + atomic_dec(&args->sbi->s_curr_pending); + + kfree(args); +} + +/* + * Write the inode to the OSD. Just fill up the struct, and set the attribute + * synchronously or asynchronously depending on the do_sync flag. + */ +static int exofs_update_inode(struct inode *inode, int do_sync) +{ + struct exofs_i_info *oi = exofs_i(inode); + struct super_block *sb = inode->i_sb; + struct exofs_sb_info *sbi = sb->s_fs_info; + struct osd_obj_id obj = {sbi->s_pid, inode->i_ino + EXOFS_OBJ_OFF}; + struct osd_request *or; + struct osd_attr attr; + struct exofs_fcb *fcb; + struct updatei_args *args; + int ret; + + args = kzalloc(sizeof(*args), GFP_KERNEL); + if (!args) + return -ENOMEM; + + fcb = &args->fcb; + + fcb->i_mode = cpu_to_le16(inode->i_mode); + fcb->i_uid = cpu_to_le32(inode->i_uid); + fcb->i_gid = cpu_to_le32(inode->i_gid); + fcb->i_links_count = cpu_to_le16(inode->i_nlink); + fcb->i_ctime = cpu_to_le32(inode->i_ctime.tv_sec); + fcb->i_atime = cpu_to_le32(inode->i_atime.tv_sec); + fcb->i_mtime = cpu_to_le32(inode->i_mtime.tv_sec); + oi->i_commit_size = i_size_read(inode); + fcb->i_size = cpu_to_le64(oi->i_commit_size); + fcb->i_generation = cpu_to_le32(inode->i_generation); + + if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) { + if (old_valid_dev(inode->i_rdev)) { + fcb->i_data[0] = + cpu_to_le32(old_encode_dev(inode->i_rdev)); + fcb->i_data[1] = 0; + } else { + fcb->i_data[0] = 0; + fcb->i_data[1] = + cpu_to_le32(new_encode_dev(inode->i_rdev)); + fcb->i_data[2] = 0; + } + } else + memcpy(fcb->i_data, oi->i_data, sizeof(fcb->i_data)); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("exofs_update_inode: osd_start_request failed.\n"); + ret = -ENOMEM; + goto free_args; + } + + osd_req_set_attributes(or, &obj); + + attr = g_attr_inode_data; + attr.val_ptr = fcb; + osd_req_add_set_attr_list(or, &attr, 1); + + if (!obj_created(oi)) { + EXOFS_DBGMSG("!obj_created\n"); + BUG_ON(!obj_2bcreated(oi)); + wait_event(oi->i_wq, obj_created(oi)); + EXOFS_DBGMSG("wait_event done\n"); + } + + if (do_sync) { + ret = exofs_sync_op(or, sbi->s_timeout, oi->i_cred); + osd_end_request(or); + goto free_args; + } else { + args->sbi = sbi; + + ret = exofs_async_op(or, updatei_done, args, oi->i_cred); + if (ret) { + osd_end_request(or); + goto free_args; + } + atomic_inc(&sbi->s_curr_pending); + goto out; /* deallocation in updatei_done */ + } + +free_args: + kfree(args); +out: + EXOFS_DBGMSG("ret=>%d\n", ret); + return ret; +} + +int exofs_write_inode(struct inode *inode, int wait) +{ + return exofs_update_inode(inode, wait); +} + +/* + * Callback function from exofs_delete_inode() - don't have much cleaning up to + * do. + */ +static void delete_done(struct osd_request *or, void *p) +{ + struct exofs_sb_info *sbi; + osd_end_request(or); + sbi = p; + atomic_dec(&sbi->s_curr_pending); +} + +/* + * Called when the refcount of an inode reaches zero. We remove the object + * from the OSD here. We make sure the object was created before we try and + * delete it. + */ +void exofs_delete_inode(struct inode *inode) +{ + struct exofs_i_info *oi = exofs_i(inode); + struct super_block *sb = inode->i_sb; + struct exofs_sb_info *sbi = sb->s_fs_info; + struct osd_obj_id obj = {sbi->s_pid, inode->i_ino + EXOFS_OBJ_OFF}; + struct osd_request *or; + int ret; + + truncate_inode_pages(&inode->i_data, 0); + + if (is_bad_inode(inode)) + goto no_delete; + + mark_inode_dirty(inode); + exofs_update_inode(inode, inode_needs_sync(inode)); + + inode->i_size = 0; + if (inode->i_blocks) + exofs_truncate(inode); + + clear_inode(inode); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("exofs_delete_inode: osd_start_request failed\n"); + return; + } + + osd_req_remove_object(or, &obj); + + /* if we are deleting an obj that hasn't been created yet, wait */ + if (!obj_created(oi)) { + BUG_ON(!obj_2bcreated(oi)); + wait_event(oi->i_wq, obj_created(oi)); + } + + ret = exofs_async_op(or, delete_done, sbi, oi->i_cred); + if (ret) { + EXOFS_ERR( + "ERROR: @exofs_delete_inode exofs_async_op failed\n"); + osd_end_request(or); + return; + } + atomic_inc(&sbi->s_curr_pending); + + return; + +no_delete: + clear_inode(inode); +} diff --git a/fs/exofs/namei.c b/fs/exofs/namei.c new file mode 100644 index 000000000000..77fdd765e76d --- /dev/null +++ b/fs/exofs/namei.c @@ -0,0 +1,342 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "exofs.h" + +static inline int exofs_add_nondir(struct dentry *dentry, struct inode *inode) +{ + int err = exofs_add_link(dentry, inode); + if (!err) { + d_instantiate(dentry, inode); + return 0; + } + inode_dec_link_count(inode); + iput(inode); + return err; +} + +static struct dentry *exofs_lookup(struct inode *dir, struct dentry *dentry, + struct nameidata *nd) +{ + struct inode *inode; + ino_t ino; + + if (dentry->d_name.len > EXOFS_NAME_LEN) + return ERR_PTR(-ENAMETOOLONG); + + ino = exofs_inode_by_name(dir, dentry); + inode = NULL; + if (ino) { + inode = exofs_iget(dir->i_sb, ino); + if (IS_ERR(inode)) + return ERR_CAST(inode); + } + return d_splice_alias(inode, dentry); +} + +static int exofs_create(struct inode *dir, struct dentry *dentry, int mode, + struct nameidata *nd) +{ + struct inode *inode = exofs_new_inode(dir, mode); + int err = PTR_ERR(inode); + if (!IS_ERR(inode)) { + inode->i_op = &exofs_file_inode_operations; + inode->i_fop = &exofs_file_operations; + inode->i_mapping->a_ops = &exofs_aops; + mark_inode_dirty(inode); + err = exofs_add_nondir(dentry, inode); + } + return err; +} + +static int exofs_mknod(struct inode *dir, struct dentry *dentry, int mode, + dev_t rdev) +{ + struct inode *inode; + int err; + + if (!new_valid_dev(rdev)) + return -EINVAL; + + inode = exofs_new_inode(dir, mode); + err = PTR_ERR(inode); + if (!IS_ERR(inode)) { + init_special_inode(inode, inode->i_mode, rdev); + mark_inode_dirty(inode); + err = exofs_add_nondir(dentry, inode); + } + return err; +} + +static int exofs_symlink(struct inode *dir, struct dentry *dentry, + const char *symname) +{ + struct super_block *sb = dir->i_sb; + int err = -ENAMETOOLONG; + unsigned l = strlen(symname)+1; + struct inode *inode; + struct exofs_i_info *oi; + + if (l > sb->s_blocksize) + goto out; + + inode = exofs_new_inode(dir, S_IFLNK | S_IRWXUGO); + err = PTR_ERR(inode); + if (IS_ERR(inode)) + goto out; + + oi = exofs_i(inode); + if (l > sizeof(oi->i_data)) { + /* slow symlink */ + inode->i_op = &exofs_symlink_inode_operations; + inode->i_mapping->a_ops = &exofs_aops; + memset(oi->i_data, 0, sizeof(oi->i_data)); + + err = page_symlink(inode, symname, l); + if (err) + goto out_fail; + } else { + /* fast symlink */ + inode->i_op = &exofs_fast_symlink_inode_operations; + memcpy(oi->i_data, symname, l); + inode->i_size = l-1; + } + mark_inode_dirty(inode); + + err = exofs_add_nondir(dentry, inode); +out: + return err; + +out_fail: + inode_dec_link_count(inode); + iput(inode); + goto out; +} + +static int exofs_link(struct dentry *old_dentry, struct inode *dir, + struct dentry *dentry) +{ + struct inode *inode = old_dentry->d_inode; + + if (inode->i_nlink >= EXOFS_LINK_MAX) + return -EMLINK; + + inode->i_ctime = CURRENT_TIME; + inode_inc_link_count(inode); + atomic_inc(&inode->i_count); + + return exofs_add_nondir(dentry, inode); +} + +static int exofs_mkdir(struct inode *dir, struct dentry *dentry, int mode) +{ + struct inode *inode; + int err = -EMLINK; + + if (dir->i_nlink >= EXOFS_LINK_MAX) + goto out; + + inode_inc_link_count(dir); + + inode = exofs_new_inode(dir, S_IFDIR | mode); + err = PTR_ERR(inode); + if (IS_ERR(inode)) + goto out_dir; + + inode->i_op = &exofs_dir_inode_operations; + inode->i_fop = &exofs_dir_operations; + inode->i_mapping->a_ops = &exofs_aops; + + inode_inc_link_count(inode); + + err = exofs_make_empty(inode, dir); + if (err) + goto out_fail; + + err = exofs_add_link(dentry, inode); + if (err) + goto out_fail; + + d_instantiate(dentry, inode); +out: + return err; + +out_fail: + inode_dec_link_count(inode); + inode_dec_link_count(inode); + iput(inode); +out_dir: + inode_dec_link_count(dir); + goto out; +} + +static int exofs_unlink(struct inode *dir, struct dentry *dentry) +{ + struct inode *inode = dentry->d_inode; + struct exofs_dir_entry *de; + struct page *page; + int err = -ENOENT; + + de = exofs_find_entry(dir, dentry, &page); + if (!de) + goto out; + + err = exofs_delete_entry(de, page); + if (err) + goto out; + + inode->i_ctime = dir->i_ctime; + inode_dec_link_count(inode); + err = 0; +out: + return err; +} + +static int exofs_rmdir(struct inode *dir, struct dentry *dentry) +{ + struct inode *inode = dentry->d_inode; + int err = -ENOTEMPTY; + + if (exofs_empty_dir(inode)) { + err = exofs_unlink(dir, dentry); + if (!err) { + inode->i_size = 0; + inode_dec_link_count(inode); + inode_dec_link_count(dir); + } + } + return err; +} + +static int exofs_rename(struct inode *old_dir, struct dentry *old_dentry, + struct inode *new_dir, struct dentry *new_dentry) +{ + struct inode *old_inode = old_dentry->d_inode; + struct inode *new_inode = new_dentry->d_inode; + struct page *dir_page = NULL; + struct exofs_dir_entry *dir_de = NULL; + struct page *old_page; + struct exofs_dir_entry *old_de; + int err = -ENOENT; + + old_de = exofs_find_entry(old_dir, old_dentry, &old_page); + if (!old_de) + goto out; + + if (S_ISDIR(old_inode->i_mode)) { + err = -EIO; + dir_de = exofs_dotdot(old_inode, &dir_page); + if (!dir_de) + goto out_old; + } + + if (new_inode) { + struct page *new_page; + struct exofs_dir_entry *new_de; + + err = -ENOTEMPTY; + if (dir_de && !exofs_empty_dir(new_inode)) + goto out_dir; + + err = -ENOENT; + new_de = exofs_find_entry(new_dir, new_dentry, &new_page); + if (!new_de) + goto out_dir; + inode_inc_link_count(old_inode); + err = exofs_set_link(new_dir, new_de, new_page, old_inode); + new_inode->i_ctime = CURRENT_TIME; + if (dir_de) + drop_nlink(new_inode); + inode_dec_link_count(new_inode); + if (err) + goto out_dir; + } else { + if (dir_de) { + err = -EMLINK; + if (new_dir->i_nlink >= EXOFS_LINK_MAX) + goto out_dir; + } + inode_inc_link_count(old_inode); + err = exofs_add_link(new_dentry, old_inode); + if (err) { + inode_dec_link_count(old_inode); + goto out_dir; + } + if (dir_de) + inode_inc_link_count(new_dir); + } + + old_inode->i_ctime = CURRENT_TIME; + + exofs_delete_entry(old_de, old_page); + inode_dec_link_count(old_inode); + + if (dir_de) { + err = exofs_set_link(old_inode, dir_de, dir_page, new_dir); + inode_dec_link_count(old_dir); + if (err) + goto out_dir; + } + return 0; + + +out_dir: + if (dir_de) { + kunmap(dir_page); + page_cache_release(dir_page); + } +out_old: + kunmap(old_page); + page_cache_release(old_page); +out: + return err; +} + +const struct inode_operations exofs_dir_inode_operations = { + .create = exofs_create, + .lookup = exofs_lookup, + .link = exofs_link, + .unlink = exofs_unlink, + .symlink = exofs_symlink, + .mkdir = exofs_mkdir, + .rmdir = exofs_rmdir, + .mknod = exofs_mknod, + .rename = exofs_rename, + .setattr = exofs_setattr, +}; + +const struct inode_operations exofs_special_inode_operations = { + .setattr = exofs_setattr, +}; diff --git a/fs/exofs/osd.c b/fs/exofs/osd.c new file mode 100644 index 000000000000..b249ae97fb15 --- /dev/null +++ b/fs/exofs/osd.c @@ -0,0 +1,153 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <scsi/scsi_device.h> +#include <scsi/osd_sense.h> + +#include "exofs.h" + +int exofs_check_ok_resid(struct osd_request *or, u64 *in_resid, u64 *out_resid) +{ + struct osd_sense_info osi; + int ret = osd_req_decode_sense(or, &osi); + + if (ret) { /* translate to Linux codes */ + if (osi.additional_code == scsi_invalid_field_in_cdb) { + if (osi.cdb_field_offset == OSD_CFO_STARTING_BYTE) + ret = -EFAULT; + if (osi.cdb_field_offset == OSD_CFO_OBJECT_ID) + ret = -ENOENT; + else + ret = -EINVAL; + } else if (osi.additional_code == osd_quota_error) + ret = -ENOSPC; + else + ret = -EIO; + } + + /* FIXME: should be include in osd_sense_info */ + if (in_resid) + *in_resid = or->in.req ? or->in.req->data_len : 0; + + if (out_resid) + *out_resid = or->out.req ? or->out.req->data_len : 0; + + return ret; +} + +void exofs_make_credential(u8 cred_a[OSD_CAP_LEN], const struct osd_obj_id *obj) +{ + osd_sec_init_nosec_doall_caps(cred_a, obj, false, true); +} + +/* + * Perform a synchronous OSD operation. + */ +int exofs_sync_op(struct osd_request *or, int timeout, uint8_t *credential) +{ + int ret; + + or->timeout = timeout; + ret = osd_finalize_request(or, 0, credential, NULL); + if (ret) { + EXOFS_DBGMSG("Faild to osd_finalize_request() => %d\n", ret); + return ret; + } + + ret = osd_execute_request(or); + + if (ret) + EXOFS_DBGMSG("osd_execute_request() => %d\n", ret); + /* osd_req_decode_sense(or, ret); */ + return ret; +} + +/* + * Perform an asynchronous OSD operation. + */ +int exofs_async_op(struct osd_request *or, osd_req_done_fn *async_done, + void *caller_context, u8 *cred) +{ + int ret; + + ret = osd_finalize_request(or, 0, cred, NULL); + if (ret) { + EXOFS_DBGMSG("Faild to osd_finalize_request() => %d\n", ret); + return ret; + } + + ret = osd_execute_request_async(or, async_done, caller_context); + + if (ret) + EXOFS_DBGMSG("osd_execute_request_async() => %d\n", ret); + return ret; +} + +int extract_attr_from_req(struct osd_request *or, struct osd_attr *attr) +{ + struct osd_attr cur_attr = {.attr_page = 0}; /* start with zeros */ + void *iter = NULL; + int nelem; + + do { + nelem = 1; + osd_req_decode_get_attr_list(or, &cur_attr, &nelem, &iter); + if ((cur_attr.attr_page == attr->attr_page) && + (cur_attr.attr_id == attr->attr_id)) { + attr->len = cur_attr.len; + attr->val_ptr = cur_attr.val_ptr; + return 0; + } + } while (iter); + + return -EIO; +} + +int osd_req_read_kern(struct osd_request *or, + const struct osd_obj_id *obj, u64 offset, void* buff, u64 len) +{ + struct request_queue *req_q = or->osd_dev->scsi_device->request_queue; + struct bio *bio = bio_map_kern(req_q, buff, len, GFP_KERNEL); + + if (!bio) + return -ENOMEM; + + osd_req_read(or, obj, bio, offset); + return 0; +} + +int osd_req_write_kern(struct osd_request *or, + const struct osd_obj_id *obj, u64 offset, void* buff, u64 len) +{ + struct request_queue *req_q = or->osd_dev->scsi_device->request_queue; + struct bio *bio = bio_map_kern(req_q, buff, len, GFP_KERNEL); + + if (!bio) + return -ENOMEM; + + osd_req_write(or, obj, bio, offset); + return 0; +} diff --git a/fs/exofs/super.c b/fs/exofs/super.c new file mode 100644 index 000000000000..9f1985e857e2 --- /dev/null +++ b/fs/exofs/super.c @@ -0,0 +1,584 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <linux/string.h> +#include <linux/parser.h> +#include <linux/vfs.h> +#include <linux/random.h> +#include <linux/exportfs.h> + +#include "exofs.h" + +/****************************************************************************** + * MOUNT OPTIONS + *****************************************************************************/ + +/* + * struct to hold what we get from mount options + */ +struct exofs_mountopt { + const char *dev_name; + uint64_t pid; + int timeout; +}; + +/* + * exofs-specific mount-time options. + */ +enum { Opt_pid, Opt_to, Opt_mkfs, Opt_format, Opt_err }; + +/* + * Our mount-time options. These should ideally be 64-bit unsigned, but the + * kernel's parsing functions do not currently support that. 32-bit should be + * sufficient for most applications now. + */ +static match_table_t tokens = { + {Opt_pid, "pid=%u"}, + {Opt_to, "to=%u"}, + {Opt_err, NULL} +}; + +/* + * The main option parsing method. Also makes sure that all of the mandatory + * mount options were set. + */ +static int parse_options(char *options, struct exofs_mountopt *opts) +{ + char *p; + substring_t args[MAX_OPT_ARGS]; + int option; + bool s_pid = false; + + EXOFS_DBGMSG("parse_options %s\n", options); + /* defaults */ + memset(opts, 0, sizeof(*opts)); + opts->timeout = BLK_DEFAULT_SG_TIMEOUT; + + while ((p = strsep(&options, ",")) != NULL) { + int token; + char str[32]; + + if (!*p) + continue; + + token = match_token(p, tokens, args); + switch (token) { + case Opt_pid: + if (0 == match_strlcpy(str, &args[0], sizeof(str))) + return -EINVAL; + opts->pid = simple_strtoull(str, NULL, 0); + if (opts->pid < EXOFS_MIN_PID) { + EXOFS_ERR("Partition ID must be >= %u", + EXOFS_MIN_PID); + return -EINVAL; + } + s_pid = 1; + break; + case Opt_to: + if (match_int(&args[0], &option)) + return -EINVAL; + if (option <= 0) { + EXOFS_ERR("Timout must be > 0"); + return -EINVAL; + } + opts->timeout = option * HZ; + break; + } + } + + if (!s_pid) { + EXOFS_ERR("Need to specify the following options:\n"); + EXOFS_ERR(" -o pid=pid_no_to_use\n"); + return -EINVAL; + } + + return 0; +} + +/****************************************************************************** + * INODE CACHE + *****************************************************************************/ + +/* + * Our inode cache. Isn't it pretty? + */ +static struct kmem_cache *exofs_inode_cachep; + +/* + * Allocate an inode in the cache + */ +static struct inode *exofs_alloc_inode(struct super_block *sb) +{ + struct exofs_i_info *oi; + + oi = kmem_cache_alloc(exofs_inode_cachep, GFP_KERNEL); + if (!oi) + return NULL; + + oi->vfs_inode.i_version = 1; + return &oi->vfs_inode; +} + +/* + * Remove an inode from the cache + */ +static void exofs_destroy_inode(struct inode *inode) +{ + kmem_cache_free(exofs_inode_cachep, exofs_i(inode)); +} + +/* + * Initialize the inode + */ +static void exofs_init_once(void *foo) +{ + struct exofs_i_info *oi = foo; + + inode_init_once(&oi->vfs_inode); +} + +/* + * Create and initialize the inode cache + */ +static int init_inodecache(void) +{ + exofs_inode_cachep = kmem_cache_create("exofs_inode_cache", + sizeof(struct exofs_i_info), 0, + SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, + exofs_init_once); + if (exofs_inode_cachep == NULL) + return -ENOMEM; + return 0; +} + +/* + * Destroy the inode cache + */ +static void destroy_inodecache(void) +{ + kmem_cache_destroy(exofs_inode_cachep); +} + +/****************************************************************************** + * SUPERBLOCK FUNCTIONS + *****************************************************************************/ +static const struct super_operations exofs_sops; +static const struct export_operations exofs_export_ops; + +/* + * Write the superblock to the OSD + */ +static void exofs_write_super(struct super_block *sb) +{ + struct exofs_sb_info *sbi; + struct exofs_fscb *fscb; + struct osd_request *or; + struct osd_obj_id obj; + int ret; + + fscb = kzalloc(sizeof(struct exofs_fscb), GFP_KERNEL); + if (!fscb) { + EXOFS_ERR("exofs_write_super: memory allocation failed.\n"); + return; + } + + lock_kernel(); + sbi = sb->s_fs_info; + fscb->s_nextid = cpu_to_le64(sbi->s_nextid); + fscb->s_numfiles = cpu_to_le32(sbi->s_numfiles); + fscb->s_magic = cpu_to_le16(sb->s_magic); + fscb->s_newfs = 0; + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_ERR("exofs_write_super: osd_start_request failed.\n"); + goto out; + } + + obj.partition = sbi->s_pid; + obj.id = EXOFS_SUPER_ID; + ret = osd_req_write_kern(or, &obj, 0, fscb, sizeof(*fscb)); + if (unlikely(ret)) { + EXOFS_ERR("exofs_write_super: osd_req_write_kern failed.\n"); + goto out; + } + + ret = exofs_sync_op(or, sbi->s_timeout, sbi->s_cred); + if (unlikely(ret)) { + EXOFS_ERR("exofs_write_super: exofs_sync_op failed.\n"); + goto out; + } + sb->s_dirt = 0; + +out: + if (or) + osd_end_request(or); + unlock_kernel(); + kfree(fscb); +} + +/* + * This function is called when the vfs is freeing the superblock. We just + * need to free our own part. + */ +static void exofs_put_super(struct super_block *sb) +{ + int num_pend; + struct exofs_sb_info *sbi = sb->s_fs_info; + + /* make sure there are no pending commands */ + for (num_pend = atomic_read(&sbi->s_curr_pending); num_pend > 0; + num_pend = atomic_read(&sbi->s_curr_pending)) { + wait_queue_head_t wq; + init_waitqueue_head(&wq); + wait_event_timeout(wq, + (atomic_read(&sbi->s_curr_pending) == 0), + msecs_to_jiffies(100)); + } + + osduld_put_device(sbi->s_dev); + kfree(sb->s_fs_info); + sb->s_fs_info = NULL; +} + +/* + * Read the superblock from the OSD and fill in the fields + */ +static int exofs_fill_super(struct super_block *sb, void *data, int silent) +{ + struct inode *root; + struct exofs_mountopt *opts = data; + struct exofs_sb_info *sbi; /*extended info */ + struct exofs_fscb fscb; /*on-disk superblock info */ + struct osd_request *or = NULL; + struct osd_obj_id obj; + int ret; + + sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); + if (!sbi) + return -ENOMEM; + sb->s_fs_info = sbi; + + /* use mount options to fill superblock */ + sbi->s_dev = osduld_path_lookup(opts->dev_name); + if (IS_ERR(sbi->s_dev)) { + ret = PTR_ERR(sbi->s_dev); + sbi->s_dev = NULL; + goto free_sbi; + } + + sbi->s_pid = opts->pid; + sbi->s_timeout = opts->timeout; + + /* fill in some other data by hand */ + memset(sb->s_id, 0, sizeof(sb->s_id)); + strcpy(sb->s_id, "exofs"); + sb->s_blocksize = EXOFS_BLKSIZE; + sb->s_blocksize_bits = EXOFS_BLKSHIFT; + sb->s_maxbytes = MAX_LFS_FILESIZE; + atomic_set(&sbi->s_curr_pending, 0); + sb->s_bdev = NULL; + sb->s_dev = 0; + + /* read data from on-disk superblock object */ + obj.partition = sbi->s_pid; + obj.id = EXOFS_SUPER_ID; + exofs_make_credential(sbi->s_cred, &obj); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + if (!silent) + EXOFS_ERR( + "exofs_fill_super: osd_start_request failed.\n"); + ret = -ENOMEM; + goto free_sbi; + } + ret = osd_req_read_kern(or, &obj, 0, &fscb, sizeof(fscb)); + if (unlikely(ret)) { + if (!silent) + EXOFS_ERR( + "exofs_fill_super: osd_req_read_kern failed.\n"); + ret = -ENOMEM; + goto free_sbi; + } + + ret = exofs_sync_op(or, sbi->s_timeout, sbi->s_cred); + if (unlikely(ret)) { + if (!silent) + EXOFS_ERR("exofs_fill_super: exofs_sync_op failed.\n"); + ret = -EIO; + goto free_sbi; + } + + sb->s_magic = le16_to_cpu(fscb.s_magic); + sbi->s_nextid = le64_to_cpu(fscb.s_nextid); + sbi->s_numfiles = le32_to_cpu(fscb.s_numfiles); + + /* make sure what we read from the object store is correct */ + if (sb->s_magic != EXOFS_SUPER_MAGIC) { + if (!silent) + EXOFS_ERR("ERROR: Bad magic value\n"); + ret = -EINVAL; + goto free_sbi; + } + + /* start generation numbers from a random point */ + get_random_bytes(&sbi->s_next_generation, sizeof(u32)); + spin_lock_init(&sbi->s_next_gen_lock); + + /* set up operation vectors */ + sb->s_op = &exofs_sops; + sb->s_export_op = &exofs_export_ops; + root = exofs_iget(sb, EXOFS_ROOT_ID - EXOFS_OBJ_OFF); + if (IS_ERR(root)) { + EXOFS_ERR("ERROR: exofs_iget failed\n"); + ret = PTR_ERR(root); + goto free_sbi; + } + sb->s_root = d_alloc_root(root); + if (!sb->s_root) { + iput(root); + EXOFS_ERR("ERROR: get root inode failed\n"); + ret = -ENOMEM; + goto free_sbi; + } + + if (!S_ISDIR(root->i_mode)) { + dput(sb->s_root); + sb->s_root = NULL; + EXOFS_ERR("ERROR: corrupt root inode (mode = %hd)\n", + root->i_mode); + ret = -EINVAL; + goto free_sbi; + } + + ret = 0; +out: + if (or) + osd_end_request(or); + return ret; + +free_sbi: + osduld_put_device(sbi->s_dev); /* NULL safe */ + kfree(sbi); + goto out; +} + +/* + * Set up the superblock (calls exofs_fill_super eventually) + */ +static int exofs_get_sb(struct file_system_type *type, + int flags, const char *dev_name, + void *data, struct vfsmount *mnt) +{ + struct exofs_mountopt opts; + int ret; + + ret = parse_options(data, &opts); + if (ret) + return ret; + + opts.dev_name = dev_name; + return get_sb_nodev(type, flags, &opts, exofs_fill_super, mnt); +} + +/* + * Return information about the file system state in the buffer. This is used + * by the 'df' command, for example. + */ +static int exofs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + struct super_block *sb = dentry->d_sb; + struct exofs_sb_info *sbi = sb->s_fs_info; + struct osd_obj_id obj = {sbi->s_pid, 0}; + struct osd_attr attrs[] = { + ATTR_DEF(OSD_APAGE_PARTITION_QUOTAS, + OSD_ATTR_PQ_CAPACITY_QUOTA, sizeof(__be64)), + ATTR_DEF(OSD_APAGE_PARTITION_INFORMATION, + OSD_ATTR_PI_USED_CAPACITY, sizeof(__be64)), + }; + uint64_t capacity = ULLONG_MAX; + uint64_t used = ULLONG_MAX; + struct osd_request *or; + uint8_t cred_a[OSD_CAP_LEN]; + int ret; + + /* get used/capacity attributes */ + exofs_make_credential(cred_a, &obj); + + or = osd_start_request(sbi->s_dev, GFP_KERNEL); + if (unlikely(!or)) { + EXOFS_DBGMSG("exofs_statfs: osd_start_request failed.\n"); + return -ENOMEM; + } + + osd_req_get_attributes(or, &obj); + osd_req_add_get_attr_list(or, attrs, ARRAY_SIZE(attrs)); + ret = exofs_sync_op(or, sbi->s_timeout, cred_a); + if (unlikely(ret)) + goto out; + + ret = extract_attr_from_req(or, &attrs[0]); + if (likely(!ret)) + capacity = get_unaligned_be64(attrs[0].val_ptr); + else + EXOFS_DBGMSG("exofs_statfs: get capacity failed.\n"); + + ret = extract_attr_from_req(or, &attrs[1]); + if (likely(!ret)) + used = get_unaligned_be64(attrs[1].val_ptr); + else + EXOFS_DBGMSG("exofs_statfs: get used-space failed.\n"); + + /* fill in the stats buffer */ + buf->f_type = EXOFS_SUPER_MAGIC; + buf->f_bsize = EXOFS_BLKSIZE; + buf->f_blocks = (capacity >> EXOFS_BLKSHIFT); + buf->f_bfree = ((capacity - used) >> EXOFS_BLKSHIFT); + buf->f_bavail = buf->f_bfree; + buf->f_files = sbi->s_numfiles; + buf->f_ffree = EXOFS_MAX_ID - sbi->s_numfiles; + buf->f_namelen = EXOFS_NAME_LEN; + +out: + osd_end_request(or); + return ret; +} + +static const struct super_operations exofs_sops = { + .alloc_inode = exofs_alloc_inode, + .destroy_inode = exofs_destroy_inode, + .write_inode = exofs_write_inode, + .delete_inode = exofs_delete_inode, + .put_super = exofs_put_super, + .write_super = exofs_write_super, + .statfs = exofs_statfs, +}; + +/****************************************************************************** + * EXPORT OPERATIONS + *****************************************************************************/ + +struct dentry *exofs_get_parent(struct dentry *child) +{ + unsigned long ino = exofs_parent_ino(child); + + if (!ino) + return NULL; + + return d_obtain_alias(exofs_iget(child->d_inode->i_sb, ino)); +} + +static struct inode *exofs_nfs_get_inode(struct super_block *sb, + u64 ino, u32 generation) +{ + struct inode *inode; + + inode = exofs_iget(sb, ino); + if (IS_ERR(inode)) + return ERR_CAST(inode); + if (generation && inode->i_generation != generation) { + /* we didn't find the right inode.. */ + iput(inode); + return ERR_PTR(-ESTALE); + } + return inode; +} + +static struct dentry *exofs_fh_to_dentry(struct super_block *sb, + struct fid *fid, int fh_len, int fh_type) +{ + return generic_fh_to_dentry(sb, fid, fh_len, fh_type, + exofs_nfs_get_inode); +} + +static struct dentry *exofs_fh_to_parent(struct super_block *sb, + struct fid *fid, int fh_len, int fh_type) +{ + return generic_fh_to_parent(sb, fid, fh_len, fh_type, + exofs_nfs_get_inode); +} + +static const struct export_operations exofs_export_ops = { + .fh_to_dentry = exofs_fh_to_dentry, + .fh_to_parent = exofs_fh_to_parent, + .get_parent = exofs_get_parent, +}; + +/****************************************************************************** + * INSMOD/RMMOD + *****************************************************************************/ + +/* + * struct that describes this file system + */ +static struct file_system_type exofs_type = { + .owner = THIS_MODULE, + .name = "exofs", + .get_sb = exofs_get_sb, + .kill_sb = generic_shutdown_super, +}; + +static int __init init_exofs(void) +{ + int err; + + err = init_inodecache(); + if (err) + goto out; + + err = register_filesystem(&exofs_type); + if (err) + goto out_d; + + return 0; +out_d: + destroy_inodecache(); +out: + return err; +} + +static void __exit exit_exofs(void) +{ + unregister_filesystem(&exofs_type); + destroy_inodecache(); +} + +MODULE_AUTHOR("Avishay Traeger <avishay@gmail.com>"); +MODULE_DESCRIPTION("exofs"); +MODULE_LICENSE("GPL"); + +module_init(init_exofs) +module_exit(exit_exofs) diff --git a/fs/exofs/symlink.c b/fs/exofs/symlink.c new file mode 100644 index 000000000000..36e2d7bc7f7b --- /dev/null +++ b/fs/exofs/symlink.c @@ -0,0 +1,57 @@ +/* + * Copyright (C) 2005, 2006 + * Avishay Traeger (avishay@gmail.com) (avishay@il.ibm.com) + * Copyright (C) 2005, 2006 + * International Business Machines + * Copyright (C) 2008, 2009 + * Boaz Harrosh <bharrosh@panasas.com> + * + * Copyrights for code taken from ext2: + * Copyright (C) 1992, 1993, 1994, 1995 + * Remy Card (card@masi.ibp.fr) + * Laboratoire MASI - Institut Blaise Pascal + * Universite Pierre et Marie Curie (Paris VI) + * from + * linux/fs/minix/inode.c + * Copyright (C) 1991, 1992 Linus Torvalds + * + * This file is part of exofs. + * + * exofs is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation. Since it is based on ext2, and the only + * valid version of GPL for the Linux kernel is version 2, the only valid + * version of GPL for exofs is version 2. + * + * exofs is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with exofs; if not, write to the Free Software + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include <linux/namei.h> + +#include "exofs.h" + +static void *exofs_follow_link(struct dentry *dentry, struct nameidata *nd) +{ + struct exofs_i_info *oi = exofs_i(dentry->d_inode); + + nd_set_link(nd, (char *)oi->i_data); + return NULL; +} + +const struct inode_operations exofs_symlink_inode_operations = { + .readlink = generic_readlink, + .follow_link = page_follow_link_light, + .put_link = page_put_link, +}; + +const struct inode_operations exofs_fast_symlink_inode_operations = { + .readlink = generic_readlink, + .follow_link = exofs_follow_link, +}; |