==================================================================
BUG: KASAN: use-after-free in ext4_ext_binsearch fs/ext4/extents.c:826 [inline]
BUG: KASAN: use-after-free in ext4_find_extent+0xbeb/0xe20 fs/ext4/extents.c:945
Read of size 4 at addr ffff888121a8bec4 by task kworker/u4:0/7

CPU: 1 PID: 7 Comm: kworker/u4:0 Not tainted 5.10.240-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 08/14/2025
Workqueue: writeback wb_workfn (flush-7:0)
Call Trace:
 __dump_stack+0x21/0x24 lib/dump_stack.c:77
 dump_stack_lvl+0x169/0x1d8 lib/dump_stack.c:118
 print_address_description+0x7f/0x2c0 mm/kasan/report.c:248
 __kasan_report mm/kasan/report.c:435 [inline]
 kasan_report+0xe2/0x130 mm/kasan/report.c:452
 __asan_report_load4_noabort+0x14/0x20 mm/kasan/report_generic.c:308
 ext4_ext_binsearch fs/ext4/extents.c:826 [inline]
 ext4_find_extent+0xbeb/0xe20 fs/ext4/extents.c:945
 ext4_ext_map_blocks+0x1de/0x5d40 fs/ext4/extents.c:4160
 ext4_map_blocks+0x978/0x1bc0 fs/ext4/inode.c:674
 mpage_map_one_extent fs/ext4/inode.c:2452 [inline]
 mpage_map_and_submit_extent fs/ext4/inode.c:2505 [inline]
 ext4_writepages+0x11d5/0x2e00 fs/ext4/inode.c:2873
 do_writepages+0x12a/0x270 mm/page-writeback.c:2380
 __writeback_single_inode+0xd5/0xa20 fs/fs-writeback.c:1470
 writeback_sb_inodes+0x860/0x1400 fs/fs-writeback.c:1733
 wb_writeback+0x3e3/0xb90 fs/fs-writeback.c:1908
 wb_do_writeback fs/fs-writeback.c:2053 [inline]
 wb_workfn+0x38f/0xe20 fs/fs-writeback.c:2094
 process_one_work+0x6e1/0xba0 kernel/workqueue.c:2301
 worker_thread+0xa6a/0x13b0 kernel/workqueue.c:2447
 kthread+0x346/0x3d0 kernel/kthread.c:313
 ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:298

The buggy address belongs to the page:
page:ffffea000486a2c0 refcount:0 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x121a8b
flags: 0x4000000000000000()
raw: 4000000000000000 ffffea000486a2c8 ffffea000486a2c8 0000000000000000
raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
page dumped because: kasan: bad access detected
page_owner info is not present (never set?)

Memory state around the buggy address:
 ffff888121a8bd80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 ffff888121a8be00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
>ffff888121a8be80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
                                           ^
 ffff888121a8bf00: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
 ffff888121a8bf80: ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff
==================================================================
EXT4-fs error (device loop0): ext4_map_blocks:740: inode #15: block 1: comm kworker/u4:0: lblock 1 mapped to illegal pblock 1 (length 3)
EXT4-fs (loop0): Delayed block allocation failed for inode 15 at logical offset 1 with max blocks 3 with error 117
EXT4-fs (loop0): This should not happen!! Data will be lost

EXT4-fs error (device loop0): __ext4_get_inode_loc:4438: comm kworker/u4:0: Invalid inode table block 8391460049216894068 in block_group 0
EXT4-fs error (device loop0): ext4_map_blocks:740: inode #15: block 16: comm kworker/u4:0: lblock 16 mapped to illegal pblock 16 (length 1)
EXT4-fs (loop0): Delayed block allocation failed for inode 15 at logical offset 16 with max blocks 1 with error 117
EXT4-fs (loop0): This should not happen!! Data will be lost