From a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 Mon Sep 17 00:00:00 2001 From: Jinke Han Date: Tue, 8 Sep 2026 15:37:42 +0800 Subject: x86/kprobes: Fix crash when probing CS CALL instructions When using eBPF to probe CS CALL instructions within a function, a crash can be triggered. The eBPF tool probes offset 257 of the __hrtimer_run_queues() function: <__hrtimer_run_queues+249>: nopl 0x0(%rax,%rax,1) <__hrtimer_run_queues+254>: mov %r14,%rdi <__hrtimer_run_queues+257>: cs call <__x86_indirect_thunk_r12> <__hrtimer_run_queues+263>: mov %eax,%r12d <__hrtimer_run_queues+266>: xchg %ax,%ax <__hrtimer_run_queues+268>: mov %r13,%rdi Which triggers this crash: BUG: unable to handle page fault for address: 00000000000f41c9 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 0 P4D 0 Oops: 0002 [#1] SMP NOPTI CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P RIP: 0010:__hrtimer_run_queues+0x106/0x230 Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is at the 6th byte of the above CS CALL instruction. Since the CS CALL instruction occupies 6 bytes, the exception occurred in the middle of that call instruction. The root cause is that when using eBPF tools to probe in the middle of a function, a kprobe with INT3 is used as the underlying implementation. During single-step emulation of the original CALL instruction, int3_emulate_call() assumes that the probed CALL instruction is 5 bytes long. However, the actual CS-prefixed CALL instruction occupies 6 bytes, so it constructs an incorrect exception return address. When the CPU returns from the kprobe handler, the next instruction to be executed is at the address of the last byte of that CS CALL instruction. Coincidentally, starting from that address, the CPU fetches and decodes a completely different instruction, which ultimately triggers a kernel crash. Fix the issue by using the actual instruction length obtained from the instruction decoder when constructing the exception return address, rather than relying on the hardcoded CALL_INSN_SIZE macro. [ mingo: Refined the changelog ] Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step") Suggested-by: Masami Hiramatsu (Google) Signed-off-by: Jinke Han Signed-off-by: Ingo Molnar Reviewed-by: Masami Hiramatsu (Google) Acked-by: Yafang Shao Acked-by: Borislav Petkov Cc: Peter Zijlstra Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000 --- fs/ext4/block_validity.c | 370 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 370 insertions(+) create mode 100644 fs/ext4/block_validity.c (limited to 'fs/ext4/block_validity.c') diff --git a/fs/ext4/block_validity.c b/fs/ext4/block_validity.c new file mode 100644 index 000000000..3a31e3d71 --- /dev/null +++ b/fs/ext4/block_validity.c @@ -0,0 +1,370 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * linux/fs/ext4/block_validity.c + * + * Copyright (C) 2009 + * Theodore Ts'o (tytso@mit.edu) + * + * Track which blocks in the filesystem are metadata blocks that + * should never be used as data blocks by files or directories. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include "ext4.h" + +struct ext4_system_zone { + struct rb_node node; + ext4_fsblk_t start_blk; + unsigned int count; + u32 ino; +}; + +static struct kmem_cache *ext4_system_zone_cachep; + +int __init ext4_init_system_zone(void) +{ + ext4_system_zone_cachep = KMEM_CACHE(ext4_system_zone, 0); + if (ext4_system_zone_cachep == NULL) + return -ENOMEM; + return 0; +} + +void ext4_exit_system_zone(void) +{ + rcu_barrier(); + kmem_cache_destroy(ext4_system_zone_cachep); +} + +static inline int can_merge(struct ext4_system_zone *entry1, + struct ext4_system_zone *entry2) +{ + if ((entry1->start_blk + entry1->count) == entry2->start_blk && + entry1->ino == entry2->ino) + return 1; + return 0; +} + +static void release_system_zone(struct ext4_system_blocks *system_blks) +{ + struct ext4_system_zone *entry, *n; + + rbtree_postorder_for_each_entry_safe(entry, n, + &system_blks->root, node) + kmem_cache_free(ext4_system_zone_cachep, entry); +} + +/* + * Mark a range of blocks as belonging to the "system zone" --- that + * is, filesystem metadata blocks which should never be used by + * inodes. + */ +static int add_system_zone(struct ext4_system_blocks *system_blks, + ext4_fsblk_t start_blk, + unsigned int count, u32 ino) +{ + struct ext4_system_zone *new_entry, *entry; + struct rb_node **n = &system_blks->root.rb_node, *node; + struct rb_node *parent = NULL, *new_node; + + while (*n) { + parent = *n; + entry = rb_entry(parent, struct ext4_system_zone, node); + if (start_blk < entry->start_blk) + n = &(*n)->rb_left; + else if (start_blk >= (entry->start_blk + entry->count)) + n = &(*n)->rb_right; + else /* Unexpected overlap of system zones. */ + return -EFSCORRUPTED; + } + + new_entry = kmem_cache_alloc(ext4_system_zone_cachep, + GFP_KERNEL); + if (!new_entry) + return -ENOMEM; + new_entry->start_blk = start_blk; + new_entry->count = count; + new_entry->ino = ino; + new_node = &new_entry->node; + + rb_link_node(new_node, parent, n); + rb_insert_color(new_node, &system_blks->root); + + /* Can we merge to the left? */ + node = rb_prev(new_node); + if (node) { + entry = rb_entry(node, struct ext4_system_zone, node); + if (can_merge(entry, new_entry)) { + new_entry->start_blk = entry->start_blk; + new_entry->count += entry->count; + rb_erase(node, &system_blks->root); + kmem_cache_free(ext4_system_zone_cachep, entry); + } + } + + /* Can we merge to the right? */ + node = rb_next(new_node); + if (node) { + entry = rb_entry(node, struct ext4_system_zone, node); + if (can_merge(new_entry, entry)) { + new_entry->count += entry->count; + rb_erase(node, &system_blks->root); + kmem_cache_free(ext4_system_zone_cachep, entry); + } + } + return 0; +} + +static void debug_print_tree(struct ext4_sb_info *sbi) +{ + struct rb_node *node; + struct ext4_system_zone *entry; + struct ext4_system_blocks *system_blks; + int first = 1; + + printk(KERN_INFO "System zones: "); + rcu_read_lock(); + system_blks = rcu_dereference(sbi->s_system_blks); + node = rb_first(&system_blks->root); + while (node) { + entry = rb_entry(node, struct ext4_system_zone, node); + printk(KERN_CONT "%s%llu-%llu", first ? "" : ", ", + entry->start_blk, entry->start_blk + entry->count - 1); + first = 0; + node = rb_next(node); + } + rcu_read_unlock(); + printk(KERN_CONT "\n"); +} + +static int ext4_protect_reserved_inode(struct super_block *sb, + struct ext4_system_blocks *system_blks, + u32 ino) +{ + struct inode *inode; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_map_blocks map; + u32 i = 0, num; + int err = 0, n; + + if ((ino < EXT4_ROOT_INO) || + (ino > le32_to_cpu(sbi->s_es->s_inodes_count))) + return -EINVAL; + inode = ext4_iget(sb, ino, EXT4_IGET_SPECIAL); + if (IS_ERR(inode)) + return PTR_ERR(inode); + num = (inode->i_size + sb->s_blocksize - 1) >> sb->s_blocksize_bits; + while (i < num) { + cond_resched(); + map.m_lblk = i; + map.m_len = num - i; + n = ext4_map_blocks(NULL, inode, &map, 0); + if (n < 0) { + err = n; + break; + } + if (n == 0) { + i++; + } else { + err = add_system_zone(system_blks, map.m_pblk, n, ino); + if (err < 0) { + if (err == -EFSCORRUPTED) { + EXT4_ERROR_INODE_ERR(inode, -err, + "blocks %llu-%llu from inode overlap system zone", + map.m_pblk, + map.m_pblk + map.m_len - 1); + } + break; + } + i += n; + } + } + iput(inode); + return err; +} + +static void ext4_destroy_system_zone(struct rcu_head *rcu) +{ + struct ext4_system_blocks *system_blks; + + system_blks = container_of(rcu, struct ext4_system_blocks, rcu); + release_system_zone(system_blks); + kfree(system_blks); +} + +/* + * Build system zone rbtree which is used for block validity checking. + * + * The update of system_blks pointer in this function is protected by + * sb->s_umount semaphore. However we have to be careful as we can be + * racing with ext4_inode_block_valid() calls reading system_blks rbtree + * protected only by RCU. That's why we first build the rbtree and then + * swap it in place. + */ +int ext4_setup_system_zone(struct super_block *sb) +{ + ext4_group_t ngroups = ext4_get_groups_count(sb); + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_system_blocks *system_blks; + struct ext4_group_desc *gdp; + ext4_group_t i; + int ret; + + system_blks = kzalloc_obj(*system_blks); + if (!system_blks) + return -ENOMEM; + + for (i=0; i < ngroups; i++) { + unsigned int meta_blks = ext4_num_base_meta_blocks(sb, i); + + cond_resched(); + if (meta_blks != 0) { + ret = add_system_zone(system_blks, + ext4_group_first_block_no(sb, i), + meta_blks, 0); + if (ret) + goto err; + } + gdp = ext4_get_group_desc(sb, i, NULL); + ret = add_system_zone(system_blks, + ext4_block_bitmap(sb, gdp), 1, 0); + if (ret) + goto err; + ret = add_system_zone(system_blks, + ext4_inode_bitmap(sb, gdp), 1, 0); + if (ret) + goto err; + ret = add_system_zone(system_blks, + ext4_inode_table(sb, gdp), + sbi->s_itb_per_group, 0); + if (ret) + goto err; + } + if (ext4_has_feature_journal(sb) && sbi->s_es->s_journal_inum) { + ret = ext4_protect_reserved_inode(sb, system_blks, + le32_to_cpu(sbi->s_es->s_journal_inum)); + if (ret) + goto err; + } + + /* + * System blks rbtree complete, announce it once to prevent racing + * with ext4_inode_block_valid() accessing the rbtree at the same + * time. + */ + rcu_assign_pointer(sbi->s_system_blks, system_blks); + + if (test_opt(sb, DEBUG)) + debug_print_tree(sbi); + return 0; +err: + release_system_zone(system_blks); + kfree(system_blks); + return ret; +} + +/* + * Called when the filesystem is unmounted or when remounting it with + * noblock_validity specified. + * + * The update of system_blks pointer in this function is protected by + * sb->s_umount semaphore. However we have to be careful as we can be + * racing with ext4_inode_block_valid() calls reading system_blks rbtree + * protected only by RCU. So we first clear the system_blks pointer and + * then free the rbtree only after RCU grace period expires. + */ +void ext4_release_system_zone(struct super_block *sb) +{ + struct ext4_system_blocks *system_blks; + + system_blks = rcu_dereference_protected(EXT4_SB(sb)->s_system_blks, + lockdep_is_held(&sb->s_umount)); + rcu_assign_pointer(EXT4_SB(sb)->s_system_blks, NULL); + + if (system_blks) + call_rcu(&system_blks->rcu, ext4_destroy_system_zone); +} + +int ext4_sb_block_valid(struct super_block *sb, struct inode *inode, + ext4_fsblk_t start_blk, unsigned int count) +{ + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_system_blocks *system_blks; + struct ext4_system_zone *entry; + struct rb_node *n; + int ret = 1; + + if ((start_blk <= le32_to_cpu(sbi->s_es->s_first_data_block)) || + (start_blk + count < start_blk) || + (start_blk + count > ext4_blocks_count(sbi->s_es))) + return 0; + + /* + * Lock the system zone to prevent it being released concurrently + * when doing a remount which inverse current "[no]block_validity" + * mount option. + */ + rcu_read_lock(); + system_blks = rcu_dereference(sbi->s_system_blks); + if (system_blks == NULL) + goto out_rcu; + + n = system_blks->root.rb_node; + while (n) { + entry = rb_entry(n, struct ext4_system_zone, node); + if (start_blk + count - 1 < entry->start_blk) + n = n->rb_left; + else if (start_blk >= (entry->start_blk + entry->count)) + n = n->rb_right; + else { + ret = 0; + if (inode) + ret = (entry->ino == inode->i_ino); + break; + } + } +out_rcu: + rcu_read_unlock(); + return ret; +} + +/* + * Returns 1 if the passed-in block region (start_blk, + * start_blk+count) is valid; 0 if some part of the block region + * overlaps with some other filesystem metadata blocks. + */ +int ext4_inode_block_valid(struct inode *inode, ext4_fsblk_t start_blk, + unsigned int count) +{ + return ext4_sb_block_valid(inode->i_sb, inode, start_blk, count); +} + +int ext4_check_blockref(const char *function, unsigned int line, + struct inode *inode, __le32 *p, unsigned int max) +{ + __le32 *bref = p; + unsigned int blk; + journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; + + if (journal && inode == journal->j_inode) + return 0; + + while (bref < p+max) { + blk = le32_to_cpu(*bref++); + if (blk && + unlikely(!ext4_inode_block_valid(inode, blk, 1))) { + ext4_error_inode(inode, function, line, blk, + "invalid block"); + return -EFSCORRUPTED; + } + } + return 0; +} + -- cgit v1.3.1