linux/fs/nilfs2/dat.c

541 lines
14 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0+
/*
* NILFS disk address translation.
*
* Copyright (C) 2006-2008 Nippon Telegraph and Telephone Corporation.
*
* Written by Koji Sato.
*/
#include <linux/types.h>
#include <linux/buffer_head.h>
#include <linux/string.h>
#include <linux/errno.h>
#include "nilfs.h"
#include "mdt.h"
#include "alloc.h"
#include "dat.h"
#define NILFS_CNO_MIN ((__u64)1)
#define NILFS_CNO_MAX (~(__u64)0)
/**
* struct nilfs_dat_info - on-memory private data of DAT file
* @mi: on-memory private data of metadata file
* @palloc_cache: persistent object allocator cache of DAT file
* @shadow: shadow map of DAT file
*/
struct nilfs_dat_info {
struct nilfs_mdt_info mi;
struct nilfs_palloc_cache palloc_cache;
struct nilfs_shadow_map shadow;
};
static inline struct nilfs_dat_info *NILFS_DAT_I(struct inode *dat)
{
return (struct nilfs_dat_info *)NILFS_MDT(dat);
}
static int nilfs_dat_prepare_entry(struct inode *dat,
struct nilfs_palloc_req *req, int create)
{
int ret;
ret = nilfs_palloc_get_entry_block(dat, req->pr_entry_nr,
create, &req->pr_entry_bh);
if (unlikely(ret == -ENOENT)) {
nilfs_err(dat->i_sb,
"DAT doesn't have a block to manage vblocknr = %llu",
(unsigned long long)req->pr_entry_nr);
/*
* Return internal code -EINVAL to notify bmap layer of
* metadata corruption.
*/
ret = -EINVAL;
}
return ret;
}
static void nilfs_dat_commit_entry(struct inode *dat,
struct nilfs_palloc_req *req)
{
mark_buffer_dirty(req->pr_entry_bh);
nilfs_mdt_mark_dirty(dat);
brelse(req->pr_entry_bh);
}
static void nilfs_dat_abort_entry(struct inode *dat,
struct nilfs_palloc_req *req)
{
brelse(req->pr_entry_bh);
}
int nilfs_dat_prepare_alloc(struct inode *dat, struct nilfs_palloc_req *req)
{
int ret;
ret = nilfs_palloc_prepare_alloc_entry(dat, req);
if (ret < 0)
return ret;
ret = nilfs_dat_prepare_entry(dat, req, 1);
if (ret < 0)
nilfs_palloc_abort_alloc_entry(dat, req);
return ret;
}
void nilfs_dat_commit_alloc(struct inode *dat, struct nilfs_palloc_req *req)
{
struct nilfs_dat_entry *entry;
void *kaddr;
kaddr = kmap_atomic(req->pr_entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr,
req->pr_entry_bh, kaddr);
entry->de_start = cpu_to_le64(NILFS_CNO_MIN);
entry->de_end = cpu_to_le64(NILFS_CNO_MAX);
entry->de_blocknr = cpu_to_le64(0);
kunmap_atomic(kaddr);
nilfs_palloc_commit_alloc_entry(dat, req);
nilfs_dat_commit_entry(dat, req);
}
void nilfs_dat_abort_alloc(struct inode *dat, struct nilfs_palloc_req *req)
{
nilfs_dat_abort_entry(dat, req);
nilfs_palloc_abort_alloc_entry(dat, req);
}
static void nilfs_dat_commit_free(struct inode *dat,
struct nilfs_palloc_req *req)
{
struct nilfs_dat_entry *entry;
void *kaddr;
kaddr = kmap_atomic(req->pr_entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr,
req->pr_entry_bh, kaddr);
entry->de_start = cpu_to_le64(NILFS_CNO_MIN);
entry->de_end = cpu_to_le64(NILFS_CNO_MIN);
entry->de_blocknr = cpu_to_le64(0);
kunmap_atomic(kaddr);
nilfs_dat_commit_entry(dat, req);
nilfs2: fix NULL pointer dereference in nilfs_palloc_commit_free_entry() Syzbot reported a null-ptr-deref bug: NILFS (loop0): segctord starting. Construction interval = 5 seconds, CP frequency < 30 seconds general protection fault, probably for non-canonical address 0xdffffc0000000002: 0000 [#1] PREEMPT SMP KASAN KASAN: null-ptr-deref in range [0x0000000000000010-0x0000000000000017] CPU: 1 PID: 3603 Comm: segctord Not tainted 6.1.0-rc2-syzkaller-00105-gb229b6ca5abb #0 Hardware name: Google Compute Engine/Google Compute Engine, BIOS Google 10/11/2022 RIP: 0010:nilfs_palloc_commit_free_entry+0xe5/0x6b0 fs/nilfs2/alloc.c:608 Code: 00 00 00 00 fc ff df 80 3c 02 00 0f 85 cd 05 00 00 48 b8 00 00 00 00 00 fc ff df 4c 8b 73 08 49 8d 7e 10 48 89 fa 48 c1 ea 03 <80> 3c 02 00 0f 85 26 05 00 00 49 8b 46 10 be a6 00 00 00 48 c7 c7 RSP: 0018:ffffc90003dff830 EFLAGS: 00010212 RAX: dffffc0000000000 RBX: ffff88802594e218 RCX: 000000000000000d RDX: 0000000000000002 RSI: 0000000000002000 RDI: 0000000000000010 RBP: ffff888071880222 R08: 0000000000000005 R09: 000000000000003f R10: 000000000000000d R11: 0000000000000000 R12: ffff888071880158 R13: ffff88802594e220 R14: 0000000000000000 R15: 0000000000000004 FS: 0000000000000000(0000) GS:ffff8880b9b00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00007fb1c08316a8 CR3: 0000000018560000 CR4: 0000000000350ee0 Call Trace: <TASK> nilfs_dat_commit_free fs/nilfs2/dat.c:114 [inline] nilfs_dat_commit_end+0x464/0x5f0 fs/nilfs2/dat.c:193 nilfs_dat_commit_update+0x26/0x40 fs/nilfs2/dat.c:236 nilfs_btree_commit_update_v+0x87/0x4a0 fs/nilfs2/btree.c:1940 nilfs_btree_commit_propagate_v fs/nilfs2/btree.c:2016 [inline] nilfs_btree_propagate_v fs/nilfs2/btree.c:2046 [inline] nilfs_btree_propagate+0xa00/0xd60 fs/nilfs2/btree.c:2088 nilfs_bmap_propagate+0x73/0x170 fs/nilfs2/bmap.c:337 nilfs_collect_file_data+0x45/0xd0 fs/nilfs2/segment.c:568 nilfs_segctor_apply_buffers+0x14a/0x470 fs/nilfs2/segment.c:1018 nilfs_segctor_scan_file+0x3f4/0x6f0 fs/nilfs2/segment.c:1067 nilfs_segctor_collect_blocks fs/nilfs2/segment.c:1197 [inline] nilfs_segctor_collect fs/nilfs2/segment.c:1503 [inline] nilfs_segctor_do_construct+0x12fc/0x6af0 fs/nilfs2/segment.c:2045 nilfs_segctor_construct+0x8e3/0xb30 fs/nilfs2/segment.c:2379 nilfs_segctor_thread_construct fs/nilfs2/segment.c:2487 [inline] nilfs_segctor_thread+0x3c3/0xf30 fs/nilfs2/segment.c:2570 kthread+0x2e4/0x3a0 kernel/kthread.c:376 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:306 </TASK> ... If DAT metadata file is corrupted on disk, there is a case where req->pr_desc_bh is NULL and blocknr is 0 at nilfs_dat_commit_end() during a b-tree operation that cascadingly updates ancestor nodes of the b-tree, because nilfs_dat_commit_alloc() for a lower level block can initialize the blocknr on the same DAT entry between nilfs_dat_prepare_end() and nilfs_dat_commit_end(). If this happens, nilfs_dat_commit_end() calls nilfs_dat_commit_free() without valid buffer heads in req->pr_desc_bh and req->pr_bitmap_bh, and causes the NULL pointer dereference above in nilfs_palloc_commit_free_entry() function, which leads to a crash. Fix this by adding a NULL check on req->pr_desc_bh and req->pr_bitmap_bh before nilfs_palloc_commit_free_entry() in nilfs_dat_commit_free(). This also calls nilfs_error() in that case to notify that there is a fatal flaw in the filesystem metadata and prevent further operations. Link: https://lkml.kernel.org/r/00000000000097c20205ebaea3d6@google.com Link: https://lkml.kernel.org/r/20221114040441.1649940-1-zhangpeng362@huawei.com Link: https://lkml.kernel.org/r/20221119120542.17204-1-konishi.ryusuke@gmail.com Signed-off-by: ZhangPeng <zhangpeng362@huawei.com> Signed-off-by: Ryusuke Konishi <konishi.ryusuke@gmail.com> Reported-by: syzbot+ebe05ee8e98f755f61d0@syzkaller.appspotmail.com Tested-by: Ryusuke Konishi <konishi.ryusuke@gmail.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2022-11-19 20:05:42 +08:00
if (unlikely(req->pr_desc_bh == NULL || req->pr_bitmap_bh == NULL)) {
nilfs_error(dat->i_sb,
"state inconsistency probably due to duplicate use of vblocknr = %llu",
(unsigned long long)req->pr_entry_nr);
return;
}
nilfs_palloc_commit_free_entry(dat, req);
}
int nilfs_dat_prepare_start(struct inode *dat, struct nilfs_palloc_req *req)
{
return nilfs_dat_prepare_entry(dat, req, 0);
}
void nilfs_dat_commit_start(struct inode *dat, struct nilfs_palloc_req *req,
sector_t blocknr)
{
struct nilfs_dat_entry *entry;
void *kaddr;
kaddr = kmap_atomic(req->pr_entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr,
req->pr_entry_bh, kaddr);
entry->de_start = cpu_to_le64(nilfs_mdt_cno(dat));
entry->de_blocknr = cpu_to_le64(blocknr);
kunmap_atomic(kaddr);
nilfs_dat_commit_entry(dat, req);
}
int nilfs_dat_prepare_end(struct inode *dat, struct nilfs_palloc_req *req)
{
struct nilfs_dat_entry *entry;
__u64 start;
sector_t blocknr;
void *kaddr;
int ret;
ret = nilfs_dat_prepare_entry(dat, req, 0);
if (ret < 0)
return ret;
kaddr = kmap_atomic(req->pr_entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr,
req->pr_entry_bh, kaddr);
start = le64_to_cpu(entry->de_start);
blocknr = le64_to_cpu(entry->de_blocknr);
kunmap_atomic(kaddr);
if (blocknr == 0) {
ret = nilfs_palloc_prepare_free_entry(dat, req);
if (ret < 0) {
nilfs_dat_abort_entry(dat, req);
return ret;
}
}
if (unlikely(start > nilfs_mdt_cno(dat))) {
nilfs_err(dat->i_sb,
"vblocknr = %llu has abnormal lifetime: start cno (= %llu) > current cno (= %llu)",
(unsigned long long)req->pr_entry_nr,
(unsigned long long)start,
(unsigned long long)nilfs_mdt_cno(dat));
nilfs_dat_abort_entry(dat, req);
return -EINVAL;
}
return 0;
}
void nilfs_dat_commit_end(struct inode *dat, struct nilfs_palloc_req *req,
int dead)
{
struct nilfs_dat_entry *entry;
__u64 start, end;
sector_t blocknr;
void *kaddr;
kaddr = kmap_atomic(req->pr_entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr,
req->pr_entry_bh, kaddr);
end = start = le64_to_cpu(entry->de_start);
if (!dead) {
end = nilfs_mdt_cno(dat);
WARN_ON(start > end);
}
entry->de_end = cpu_to_le64(end);
blocknr = le64_to_cpu(entry->de_blocknr);
kunmap_atomic(kaddr);
if (blocknr == 0)
nilfs_dat_commit_free(dat, req);
else
nilfs_dat_commit_entry(dat, req);
}
void nilfs_dat_abort_end(struct inode *dat, struct nilfs_palloc_req *req)
{
struct nilfs_dat_entry *entry;
__u64 start;
sector_t blocknr;
void *kaddr;
kaddr = kmap_atomic(req->pr_entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, req->pr_entry_nr,
req->pr_entry_bh, kaddr);
start = le64_to_cpu(entry->de_start);
blocknr = le64_to_cpu(entry->de_blocknr);
kunmap_atomic(kaddr);
if (start == nilfs_mdt_cno(dat) && blocknr == 0)
nilfs_palloc_abort_free_entry(dat, req);
nilfs_dat_abort_entry(dat, req);
}
int nilfs_dat_prepare_update(struct inode *dat,
struct nilfs_palloc_req *oldreq,
struct nilfs_palloc_req *newreq)
{
int ret;
ret = nilfs_dat_prepare_end(dat, oldreq);
if (!ret) {
ret = nilfs_dat_prepare_alloc(dat, newreq);
if (ret < 0)
nilfs_dat_abort_end(dat, oldreq);
}
return ret;
}
void nilfs_dat_commit_update(struct inode *dat,
struct nilfs_palloc_req *oldreq,
struct nilfs_palloc_req *newreq, int dead)
{
nilfs_dat_commit_end(dat, oldreq, dead);
nilfs_dat_commit_alloc(dat, newreq);
}
void nilfs_dat_abort_update(struct inode *dat,
struct nilfs_palloc_req *oldreq,
struct nilfs_palloc_req *newreq)
{
nilfs_dat_abort_end(dat, oldreq);
nilfs_dat_abort_alloc(dat, newreq);
}
/**
* nilfs_dat_mark_dirty -
* @dat: DAT file inode
* @vblocknr: virtual block number
*
* Description:
*
* Return Value: On success, 0 is returned. On error, one of the following
* negative error codes is returned.
*
* %-EIO - I/O error.
*
* %-ENOMEM - Insufficient amount of memory available.
*/
int nilfs_dat_mark_dirty(struct inode *dat, __u64 vblocknr)
{
struct nilfs_palloc_req req;
int ret;
req.pr_entry_nr = vblocknr;
ret = nilfs_dat_prepare_entry(dat, &req, 0);
if (ret == 0)
nilfs_dat_commit_entry(dat, &req);
return ret;
}
/**
* nilfs_dat_freev - free virtual block numbers
* @dat: DAT file inode
* @vblocknrs: array of virtual block numbers
* @nitems: number of virtual block numbers
*
* Description: nilfs_dat_freev() frees the virtual block numbers specified by
* @vblocknrs and @nitems.
*
* Return Value: On success, 0 is returned. On error, one of the following
* negative error codes is returned.
*
* %-EIO - I/O error.
*
* %-ENOMEM - Insufficient amount of memory available.
*
* %-ENOENT - The virtual block number have not been allocated.
*/
int nilfs_dat_freev(struct inode *dat, __u64 *vblocknrs, size_t nitems)
{
return nilfs_palloc_freev(dat, vblocknrs, nitems);
}
/**
* nilfs_dat_move - change a block number
* @dat: DAT file inode
* @vblocknr: virtual block number
* @blocknr: block number
*
* Description: nilfs_dat_move() changes the block number associated with
* @vblocknr to @blocknr.
*
* Return Value: On success, 0 is returned. On error, one of the following
* negative error codes is returned.
*
* %-EIO - I/O error.
*
* %-ENOMEM - Insufficient amount of memory available.
*/
int nilfs_dat_move(struct inode *dat, __u64 vblocknr, sector_t blocknr)
{
struct buffer_head *entry_bh;
struct nilfs_dat_entry *entry;
void *kaddr;
int ret;
ret = nilfs_palloc_get_entry_block(dat, vblocknr, 0, &entry_bh);
if (ret < 0)
return ret;
/*
* The given disk block number (blocknr) is not yet written to
* the device at this point.
*
* To prevent nilfs_dat_translate() from returning the
* uncommitted block number, this makes a copy of the entry
* buffer and redirects nilfs_dat_translate() to the copy.
*/
if (!buffer_nilfs_redirected(entry_bh)) {
ret = nilfs_mdt_freeze_buffer(dat, entry_bh);
if (ret) {
brelse(entry_bh);
return ret;
}
}
kaddr = kmap_atomic(entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, vblocknr, entry_bh, kaddr);
if (unlikely(entry->de_blocknr == cpu_to_le64(0))) {
nilfs_crit(dat->i_sb,
"%s: invalid vblocknr = %llu, [%llu, %llu)",
__func__, (unsigned long long)vblocknr,
(unsigned long long)le64_to_cpu(entry->de_start),
(unsigned long long)le64_to_cpu(entry->de_end));
kunmap_atomic(kaddr);
brelse(entry_bh);
return -EINVAL;
}
WARN_ON(blocknr == 0);
entry->de_blocknr = cpu_to_le64(blocknr);
kunmap_atomic(kaddr);
mark_buffer_dirty(entry_bh);
nilfs_mdt_mark_dirty(dat);
brelse(entry_bh);
return 0;
}
/**
* nilfs_dat_translate - translate a virtual block number to a block number
* @dat: DAT file inode
* @vblocknr: virtual block number
* @blocknrp: pointer to a block number
*
* Description: nilfs_dat_translate() maps the virtual block number @vblocknr
* to the corresponding block number.
*
* Return Value: On success, 0 is returned and the block number associated
* with @vblocknr is stored in the place pointed by @blocknrp. On error, one
* of the following negative error codes is returned.
*
* %-EIO - I/O error.
*
* %-ENOMEM - Insufficient amount of memory available.
*
* %-ENOENT - A block number associated with @vblocknr does not exist.
*/
int nilfs_dat_translate(struct inode *dat, __u64 vblocknr, sector_t *blocknrp)
{
struct buffer_head *entry_bh, *bh;
struct nilfs_dat_entry *entry;
sector_t blocknr;
void *kaddr;
int ret;
ret = nilfs_palloc_get_entry_block(dat, vblocknr, 0, &entry_bh);
if (ret < 0)
return ret;
if (!nilfs_doing_gc() && buffer_nilfs_redirected(entry_bh)) {
bh = nilfs_mdt_get_frozen_buffer(dat, entry_bh);
if (bh) {
WARN_ON(!buffer_uptodate(bh));
brelse(entry_bh);
entry_bh = bh;
}
}
kaddr = kmap_atomic(entry_bh->b_page);
entry = nilfs_palloc_block_get_entry(dat, vblocknr, entry_bh, kaddr);
blocknr = le64_to_cpu(entry->de_blocknr);
if (blocknr == 0) {
ret = -ENOENT;
goto out;
}
*blocknrp = blocknr;
out:
kunmap_atomic(kaddr);
brelse(entry_bh);
return ret;
}
ssize_t nilfs_dat_get_vinfo(struct inode *dat, void *buf, unsigned int visz,
size_t nvi)
{
struct buffer_head *entry_bh;
struct nilfs_dat_entry *entry;
struct nilfs_vinfo *vinfo = buf;
__u64 first, last;
void *kaddr;
unsigned long entries_per_block = NILFS_MDT(dat)->mi_entries_per_block;
int i, j, n, ret;
for (i = 0; i < nvi; i += n) {
ret = nilfs_palloc_get_entry_block(dat, vinfo->vi_vblocknr,
0, &entry_bh);
if (ret < 0)
return ret;
kaddr = kmap_atomic(entry_bh->b_page);
/* last virtual block number in this block */
first = vinfo->vi_vblocknr;
do_div(first, entries_per_block);
first *= entries_per_block;
last = first + entries_per_block - 1;
for (j = i, n = 0;
j < nvi && vinfo->vi_vblocknr >= first &&
vinfo->vi_vblocknr <= last;
j++, n++, vinfo = (void *)vinfo + visz) {
entry = nilfs_palloc_block_get_entry(
dat, vinfo->vi_vblocknr, entry_bh, kaddr);
vinfo->vi_start = le64_to_cpu(entry->de_start);
vinfo->vi_end = le64_to_cpu(entry->de_end);
vinfo->vi_blocknr = le64_to_cpu(entry->de_blocknr);
}
kunmap_atomic(kaddr);
brelse(entry_bh);
}
return nvi;
}
/**
* nilfs_dat_read - read or get dat inode
* @sb: super block instance
* @entry_size: size of a dat entry
* @raw_inode: on-disk dat inode
* @inodep: buffer to store the inode
*/
int nilfs_dat_read(struct super_block *sb, size_t entry_size,
struct nilfs_inode *raw_inode, struct inode **inodep)
{
static struct lock_class_key dat_lock_key;
struct inode *dat;
struct nilfs_dat_info *di;
int err;
if (entry_size > sb->s_blocksize) {
nilfs_err(sb, "too large DAT entry size: %zu bytes",
entry_size);
return -EINVAL;
} else if (entry_size < NILFS_MIN_DAT_ENTRY_SIZE) {
nilfs_err(sb, "too small DAT entry size: %zu bytes",
entry_size);
return -EINVAL;
}
dat = nilfs_iget_locked(sb, NULL, NILFS_DAT_INO);
if (unlikely(!dat))
return -ENOMEM;
if (!(dat->i_state & I_NEW))
goto out;
err = nilfs_mdt_init(dat, NILFS_MDT_GFP, sizeof(*di));
if (err)
goto failed;
err = nilfs_palloc_init_blockgroup(dat, entry_size);
if (err)
goto failed;
di = NILFS_DAT_I(dat);
lockdep_set_class(&di->mi.mi_sem, &dat_lock_key);
nilfs_palloc_setup_cache(dat, &di->palloc_cache);
err = nilfs_mdt_setup_shadow_map(dat, &di->shadow);
if (err)
goto failed;
err = nilfs_read_inode_common(dat, raw_inode);
if (err)
goto failed;
unlock_new_inode(dat);
out:
*inodep = dat;
return 0;
failed:
iget_failed(dat);
return err;
}