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/ext4_jbd2.c | 433 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 433 insertions(+) create mode 100644 fs/ext4/ext4_jbd2.c (limited to 'fs/ext4/ext4_jbd2.c') diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c new file mode 100644 index 000000000..53ddedb52 --- /dev/null +++ b/fs/ext4/ext4_jbd2.c @@ -0,0 +1,433 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Interface between ext4 and JBD + */ + +#include "ext4_jbd2.h" + +#include + +int ext4_inode_journal_mode(struct inode *inode) +{ + if (EXT4_JOURNAL(inode) == NULL) + return EXT4_INODE_WRITEBACK_DATA_MODE; /* writeback */ + /* We do not support data journalling with delayed allocation */ + if (!S_ISREG(inode->i_mode) || + ext4_test_inode_flag(inode, EXT4_INODE_EA_INODE) || + test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA || + (ext4_test_inode_flag(inode, EXT4_INODE_JOURNAL_DATA) && + !test_opt(inode->i_sb, DELALLOC))) { + /* We do not support data journalling for encrypted data */ + if (S_ISREG(inode->i_mode) && IS_ENCRYPTED(inode)) + return EXT4_INODE_ORDERED_DATA_MODE; /* ordered */ + return EXT4_INODE_JOURNAL_DATA_MODE; /* journal data */ + } + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_ORDERED_DATA) + return EXT4_INODE_ORDERED_DATA_MODE; /* ordered */ + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_WRITEBACK_DATA) + return EXT4_INODE_WRITEBACK_DATA_MODE; /* writeback */ + BUG(); +} + +/* Just increment the non-pointer handle value */ +static handle_t *ext4_get_nojournal(void) +{ + handle_t *handle = current->journal_info; + + BUG_ON(handle && !handle->h_invalid); + + if (!handle) { + handle = jbd2_alloc_handle(GFP_NOFS); + if (!handle) + return ERR_PTR(-ENOMEM); + handle->h_invalid = 1; + /* + * This is done by start_this_handle() if journalling + * is enabled. + */ + handle->saved_alloc_context = memalloc_nofs_save(); + current->journal_info = handle; + } + handle->h_ref++; + return handle; +} + + +/* Decrement the non-pointer handle value */ +static void ext4_put_nojournal(handle_t *handle) +{ + BUG_ON(handle->h_ref == 0); + + handle->h_ref--; + if (handle->h_ref == 0) { + memalloc_nofs_restore(handle->saved_alloc_context); + jbd2_free_handle(handle); + current->journal_info = NULL; + } +} + +/* + * Wrappers for jbd2_journal_start/end. + */ +static int ext4_journal_check_start(struct super_block *sb) +{ + int ret; + journal_t *journal; + + might_sleep(); + + ret = ext4_emergency_state(sb); + if (unlikely(ret)) + return ret; + + if (WARN_ON_ONCE(sb_rdonly(sb))) + return -EROFS; + + WARN_ON(sb->s_writers.frozen == SB_FREEZE_COMPLETE); + journal = EXT4_SB(sb)->s_journal; + /* + * Special case here: if the journal has aborted behind our + * backs (eg. EIO in the commit thread), then we still need to + * take the FS itself readonly cleanly. + */ + if (journal && is_journal_aborted(journal)) { + ext4_abort(sb, -journal->j_errno, "Detected aborted journal"); + return -EROFS; + } + return 0; +} + +handle_t *__ext4_journal_start_sb(struct inode *inode, + struct super_block *sb, unsigned int line, + int type, int blocks, int rsv_blocks, + int revoke_creds) +{ + journal_t *journal; + int err; + if (inode) + trace_ext4_journal_start_inode(inode, blocks, rsv_blocks, + revoke_creds, type, + _RET_IP_); + else + trace_ext4_journal_start_sb(sb, blocks, rsv_blocks, + revoke_creds, type, + _RET_IP_); + err = ext4_journal_check_start(sb); + if (err < 0) + return ERR_PTR(err); + + journal = EXT4_SB(sb)->s_journal; + if (!journal || (EXT4_SB(sb)->s_mount_state & EXT4_FC_REPLAY)) + return ext4_get_nojournal(); + return jbd2__journal_start(journal, blocks, rsv_blocks, revoke_creds, + GFP_NOFS, type, line); +} + +int __ext4_journal_stop(const char *where, unsigned int line, handle_t *handle) +{ + struct super_block *sb; + int err; + int rc; + + if (!ext4_handle_valid(handle)) { + ext4_put_nojournal(handle); + return 0; + } + + err = handle->h_err; + if (!handle->h_transaction) { + rc = jbd2_journal_stop(handle); + return err ? err : rc; + } + + sb = handle->h_transaction->t_journal->j_private; + rc = jbd2_journal_stop(handle); + + if (!err) + err = rc; + if (err) + __ext4_std_error(sb, where, line, err); + return err; +} + +handle_t *__ext4_journal_start_reserved(handle_t *handle, unsigned int line, + int type) +{ + struct super_block *sb; + int err; + + if (!ext4_handle_valid(handle)) + return ext4_get_nojournal(); + + sb = handle->h_journal->j_private; + trace_ext4_journal_start_reserved(sb, + jbd2_handle_buffer_credits(handle), _RET_IP_); + err = ext4_journal_check_start(sb); + if (err < 0) { + jbd2_journal_free_reserved(handle); + return ERR_PTR(err); + } + + err = jbd2_journal_start_reserved(handle, type, line); + if (err < 0) + return ERR_PTR(err); + return handle; +} + +int __ext4_journal_ensure_credits(handle_t *handle, int check_cred, + int extend_cred, int revoke_cred) +{ + if (!ext4_handle_valid(handle)) + return 0; + if (is_handle_aborted(handle)) + return -EROFS; + if (jbd2_handle_buffer_credits(handle) >= check_cred && + handle->h_revoke_credits >= revoke_cred) + return 0; + extend_cred = max(0, extend_cred - jbd2_handle_buffer_credits(handle)); + revoke_cred = max(0, revoke_cred - handle->h_revoke_credits); + return ext4_journal_extend(handle, extend_cred, revoke_cred); +} + +static void ext4_journal_abort_handle(const char *caller, unsigned int line, + const char *err_fn, + struct buffer_head *bh, + handle_t *handle, int err) +{ + char nbuf[16]; + const char *errstr = ext4_decode_error(NULL, err, nbuf); + + BUG_ON(!ext4_handle_valid(handle)); + + if (bh) + BUFFER_TRACE(bh, "abort"); + + if (!handle->h_err) + handle->h_err = err; + + if (is_handle_aborted(handle)) + return; + + printk(KERN_ERR "EXT4-fs: %s:%d: aborting transaction: %s in %s\n", + caller, line, errstr, err_fn); + + jbd2_journal_abort_handle(handle); +} + +static void ext4_check_bdev_write_error(struct super_block *sb) +{ + struct address_space *mapping = sb->s_bdev->bd_mapping; + struct ext4_sb_info *sbi = EXT4_SB(sb); + int err; + + /* + * If the block device has write error flag, it may have failed to + * async write out metadata buffers in the background. In this case, + * we could read old data from disk and write it out again, which + * may lead to on-disk filesystem inconsistency. + */ + if (errseq_check(&mapping->wb_err, READ_ONCE(sbi->s_bdev_wb_err))) { + spin_lock(&sbi->s_bdev_wb_lock); + err = errseq_check_and_advance(&mapping->wb_err, &sbi->s_bdev_wb_err); + spin_unlock(&sbi->s_bdev_wb_lock); + if (err) + ext4_error_err(sb, -err, + "Error while async write back metadata"); + } +} + +int __ext4_journal_get_write_access(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb, + struct buffer_head *bh, + enum ext4_journal_trigger_type trigger_type) +{ + int err; + + might_sleep(); + + if (ext4_handle_valid(handle)) { + err = jbd2_journal_get_write_access(handle, bh); + if (err) { + ext4_journal_abort_handle(where, line, __func__, bh, + handle, err); + return err; + } + } else + ext4_check_bdev_write_error(sb); + if (trigger_type == EXT4_JTR_NONE || + !ext4_has_feature_metadata_csum(sb)) + return 0; + BUG_ON(trigger_type >= EXT4_JOURNAL_TRIGGER_COUNT); + jbd2_journal_set_triggers(bh, + &EXT4_SB(sb)->s_journal_triggers[trigger_type].tr_triggers); + return 0; +} + +/* + * The ext4 forget function must perform a revoke if we are freeing data + * which has been journaled. Metadata (eg. indirect blocks) must be + * revoked in all cases. + * + * "bh" may be NULL: a metadata block may have been freed from memory + * but there may still be a record of it in the journal, and that record + * still needs to be revoked. + */ +int __ext4_forget(const char *where, unsigned int line, handle_t *handle, + int is_metadata, struct inode *inode, + struct buffer_head *bh, ext4_fsblk_t blocknr) +{ + int err; + + might_sleep(); + + trace_ext4_forget(inode, is_metadata, blocknr); + BUFFER_TRACE(bh, "enter"); + + ext4_debug("forgetting bh %p: is_metadata=%d, mode %o, data mode %x\n", + bh, is_metadata, inode->i_mode, + test_opt(inode->i_sb, DATA_FLAGS)); + + /* + * In the no journal case, we should wait for the ongoing buffer + * to complete and do a forget. + */ + if (!ext4_handle_valid(handle)) { + if (bh) { + clear_buffer_dirty(bh); + wait_on_buffer(bh); + __bforget(bh); + } + return 0; + } + + /* Never use the revoke function if we are doing full data + * journaling: there is no need to, and a V1 superblock won't + * support it. Otherwise, only skip the revoke on un-journaled + * data blocks. */ + + if (test_opt(inode->i_sb, DATA_FLAGS) == EXT4_MOUNT_JOURNAL_DATA || + (!is_metadata && !ext4_should_journal_data(inode))) { + if (bh) { + BUFFER_TRACE(bh, "call jbd2_journal_forget"); + err = jbd2_journal_forget(handle, bh); + if (err) + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); + return err; + } + return 0; + } + + /* + * data!=journal && (is_metadata || should_journal_data(inode)) + */ + BUFFER_TRACE(bh, "call jbd2_journal_revoke"); + err = jbd2_journal_revoke(handle, blocknr, bh); + if (err) { + ext4_journal_abort_handle(where, line, __func__, + bh, handle, err); + __ext4_error(inode->i_sb, where, line, true, -err, 0, + "error %d when attempting revoke", err); + } + BUFFER_TRACE(bh, "exit"); + return err; +} + +int __ext4_journal_get_create_access(const char *where, unsigned int line, + handle_t *handle, struct super_block *sb, + struct buffer_head *bh, + enum ext4_journal_trigger_type trigger_type) +{ + int err; + + if (!ext4_handle_valid(handle)) + return 0; + + err = jbd2_journal_get_create_access(handle, bh); + if (err) { + ext4_journal_abort_handle(where, line, __func__, bh, handle, + err); + return err; + } + if (trigger_type == EXT4_JTR_NONE || + !ext4_has_feature_metadata_csum(sb)) + return 0; + BUG_ON(trigger_type >= EXT4_JOURNAL_TRIGGER_COUNT); + jbd2_journal_set_triggers(bh, + &EXT4_SB(sb)->s_journal_triggers[trigger_type].tr_triggers); + return 0; +} + +static void ext4_inode_attach_mmb(struct inode *inode) +{ + struct mapping_metadata_bhs *mmb; + + /* + * It's difficult to handle failure when marking buffer dirty without + * leaving filesystem corrupted + */ + mmb = kmalloc_obj(*mmb, GFP_NOFS | __GFP_NOFAIL | __GFP_ACCOUNT); + mmb_init(mmb, &inode->i_data); + /* Someone swapped another mmb before us? */ + if (cmpxchg(&EXT4_I(inode)->i_metadata_bhs, NULL, mmb)) + kfree(mmb); +} + +int __ext4_handle_dirty_metadata(const char *where, unsigned int line, + handle_t *handle, struct inode *inode, + struct buffer_head *bh) +{ + int err = 0; + + might_sleep(); + + set_buffer_meta(bh); + set_buffer_prio(bh); + set_buffer_uptodate(bh); + if (ext4_handle_valid(handle)) { + err = jbd2_journal_dirty_metadata(handle, bh); + /* Errors can only happen due to aborted journal or a nasty bug */ + if (!is_handle_aborted(handle) && WARN_ON_ONCE(err)) { + ext4_journal_abort_handle(where, line, __func__, bh, + handle, err); + if (inode == NULL) { + pr_err("EXT4: jbd2_journal_dirty_metadata " + "failed: handle type %u started at " + "line %u, credits %u/%u, errcode %d", + handle->h_type, + handle->h_line_no, + handle->h_requested_credits, + jbd2_handle_buffer_credits(handle), err); + return err; + } + ext4_error_inode(inode, where, line, + bh->b_blocknr, + "journal_dirty_metadata failed: " + "handle type %u started at line %u, " + "credits %u/%u, errcode %d", + handle->h_type, + handle->h_line_no, + handle->h_requested_credits, + jbd2_handle_buffer_credits(handle), + err); + } + } else { + if (inode) { + if (!ext4_i_metadata_bhs(inode)) + ext4_inode_attach_mmb(inode); + mmb_mark_buffer_dirty(bh, ext4_i_metadata_bhs(inode)); + } else { + mark_buffer_dirty(bh); + } + if (inode && inode_needs_sync(inode)) { + sync_dirty_buffer(bh); + if (buffer_req(bh) && !buffer_uptodate(bh)) { + ext4_error_inode_err(inode, where, line, + bh->b_blocknr, EIO, + "IO error syncing itable block"); + err = -EIO; + } + } + } + return err; +} -- cgit v1.3.1