Re: [mm] f9ce0be71d: BUG:KCSAN:data-race_in_next_uptodate_page/next_uptodate_page

[Kirill A. Shutemov]

On Thu, Aug 26, 2021 at 10:41:57PM +0800, kernel test robot wrote:
> 
> 
> Greeting,
> 
> FYI, we noticed the following commit (built with clang-14):
> 
> commit: f9ce0be71d1fbb038ada15ced83474b0e63f264d ("mm: Cleanup faultaround and finish_fault() codepaths")
> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git master
> 
> 
> in testcase: trinity
> version: 
> with following parameters:
> 
> 	number: 99999
> 	group: group-04
> 
> test-description: Trinity is a linux system call fuzz tester.
> test-url: http://codemonkey.org.uk/projects/trinity/
> 
> 
> on test machine: qemu-system-x86_64 -enable-kvm -cpu SandyBridge -smp 2 -m 16G
> 
> caused below changes (please refer to attached dmesg/kmsg for entire log/backtrace):
> 
> 
> +------------------------------------------------------------------------+-----------+------------+
> |                                                                        | v5.11-rc4 | f9ce0be71d |
> +------------------------------------------------------------------------+-----------+------------+
> | BUG:KCSAN:data-race_in_next_uptodate_page/next_uptodate_page           | 0         | 19         |
> | BUG:KCSAN:data-race_in_mark_page_accessed/next_uptodate_page           | 0         | 17         |
> | BUG:KCSAN:data-race_in_next_uptodate_page/page_memcg                   | 0         | 13         |
> | BUG:KCSAN:data-race_in_next_uptodate_page/unlock_page                  | 0         | 13         |
> +------------------------------------------------------------------------+-----------+------------+
> 
> 
> If you fix the issue, kindly add following tag
> Reported-by: kernel test robot <oliver.sang@xxxxxxxxx>
> 
> 
> [  184.717904][ T1873] ==================================================================
> [  184.718938][ T1873] BUG: KCSAN: data-race in next_uptodate_page / unlock_page
> [  184.719828][ T1873]
> [  184.720103][ T1873] write (marked) to 0xffffea00050f37c0 of 8 bytes by task 1872 on cpu 1:
> [  184.721024][ T1873]  unlock_page+0x102/0x1b0
> [  184.721533][ T1873]  filemap_map_pages+0x6c6/0x890
> [  184.722102][ T1873]  handle_mm_fault+0x179c/0x27f0
> [  184.722672][ T1873]  do_user_addr_fault+0x3fb/0x830
> [  184.723263][ T1873]  exc_page_fault+0xc3/0x1a0
> [  184.723845][ T1873]  asm_exc_page_fault+0x1e/0x30
> [  184.724427][ T1873]
> [  184.724720][ T1873] read to 0xffffea00050f37c0 of 8 bytes by task 1873 on cpu 0:
> [  184.725575][ T1873]  next_uptodate_page+0x456/0x830
> [  184.726161][ T1873]  filemap_map_pages+0x728/0x890
> [  184.726747][ T1873]  handle_mm_fault+0x179c/0x27f0
> [  184.727332][ T1873]  do_user_addr_fault+0x3fb/0x830
> [  184.727905][ T1873]  exc_page_fault+0xc3/0x1a0
> [  184.728440][ T1873]  asm_exc_page_fault+0x1e/0x30
> [  184.729027][ T1873]
> [  184.729313][ T1873] Reported by Kernel Concurrency Sanitizer on:
> [  184.730019][ T1873] CPU: 0 PID: 1873 Comm: systemd-udevd Not tainted 5.11.0-rc4-00001-gf9ce0be71d1f #1
> [  184.731103][ T1873] ==================================================================

Line annotation would be helpful.

And I'm not very familiar with KCSAN. My guess it reports PageLock() vs.
clearing PG_locked. In this context it looks safe, unless I miss
something.

Do we need some annotation to help KCSAN?

-- 
 Kirill A. Shutemov