From: kernel test robot <lkp@intel.com>
To: oe-kbuild@lists.linux.dev
Cc: lkp@intel.com, Dan Carpenter <error27@gmail.com>
Subject: [kas:rcu-free/clean 11/13] mm/filemap.c:2869 filemap_read_rcu() warn: bitwise AND condition is false here
Date: Thu, 15 Oct 2026 00:18:33 +0200 [thread overview]
Message-ID: <202610150009.YSyjG3rF-lkp@intel.com> (raw)
Unfiltered warning: this smatch report is sent as-is and has not been
reviewed for false positives, so please treat it with due skepticism.
BCC: lkp@intel.com
CC: oe-kbuild-all@lists.linux.dev
TO: "Kiryl Shutsemau (Meta)" <kas@kernel.org>
tree: https://git.kernel.org/pub/scm/linux/kernel/git/kas/linux.git rcu-free/clean
head: 25a123685f783fe2d511f44330e977304997fecc
commit: 7945beac7bf45f6e6fe50b519f2b55c3fed1cc2f [11/13] mm/filemap: read from the page cache without folio references
:::::: branch date: 34 hours ago
:::::: commit date: 34 hours ago
config: nios2-randconfig-r2200-20261014 (https://download.01.org/0day-ci/archive/20261015/202610150009.YSyjG3rF-lkp@intel.com/config)
compiler: nios2-linux-gcc (GCC) 11.5.0
smatch: v0.5.0-9187-g5189e3fb
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Reported-by: Dan Carpenter <error27@gmail.com>
| Closes: https://lore.kernel.org/r/202610150009.YSyjG3rF-lkp@intel.com/
smatch warnings:
mm/filemap.c:2869 filemap_read_rcu() warn: bitwise AND condition is false here
vim +2869 mm/filemap.c
5ccc944dce3df5f Matthew Wilcox (Oracle 2022-06-10 2830)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2831) /*
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2832) * Copy from folios with PG_rcu_free without taking references, and return
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2833) * how much was copied; the caller reads the rest. The memory of a folio
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2834) * is not reused before rcu_read_unlock() if the flag was seen and then
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2835) * the folio was still in its slot. A split can change a folio's size
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2836) * meanwhile, so copy page by page, each page looked up in i_pages.
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2837) */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2838) static size_t filemap_read_rcu(struct kiocb *iocb, struct iov_iter *iter)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2839) {
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2840) struct address_space *mapping = iocb->ki_filp->f_mapping;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2841) struct inode *inode = mapping->host;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2842) XA_STATE(xas, &mapping->i_pages, 0);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2843) size_t copied = 0;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2844)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2845) if (!IS_ENABLED(CONFIG_PAGE_RCU_FREE) || !mapping_rcu_read(mapping))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2846) return 0;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2847) /* An aliasing D-cache needs flush_dcache_folio(), under a reference */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2848) if (cpu_dcache_is_aliasing() && mapping_writably_mapped(mapping))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2849) return 0;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2850)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2851) rcu_read_lock();
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2852) /* A fault ends the copy: it cannot be handled here */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2853) pagefault_disable();
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2854) while (iov_iter_count(iter) && copied < SZ_128K) {
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2855) loff_t pos = iocb->ki_pos;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2856) pgoff_t index = pos >> PAGE_SHIFT;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2857) size_t offset = offset_in_page(pos);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2858) struct folio *folio;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2859) struct page *page;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2860) size_t bytes, n;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2861) loff_t isize;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2862)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2863) xas_set(&xas, index);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2864) folio = xas_load(&xas);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2865) if (xas_retry(&xas, folio) || !folio || xa_is_value(folio))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2866) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2867)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2868) /* Not folio_test_rcu_free(): may be a tail page by now */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 @2869) if (!(READ_ONCE(folio->flags.f) & __PG_RCU_FREE))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2870) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2871) /* Pairs with folio_test_set_rcu_free() */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2872) smp_rmb();
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2873) if (xas_reload(&xas) != folio)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2874) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2875) /* Pairs with the store into i_pages */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2876) smp_rmb();
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2877)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2878) if (!folio_test_uptodate(folio) || folio_test_readahead(folio))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2879) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2880) /* folio_mark_accessed() needs a reference */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2881) if (!folio_mark_accessed_noop(folio))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2882) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2883) page = folio_page(folio, index - folio->index);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2884) /* memory_failure() waits only for readers that got past this */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2885) if (PageHWPoison(page))
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2886) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2887)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2888) isize = i_size_read(inode);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2889) if (pos >= isize)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2890) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2891) /* A write that moved i_size used the folio in the slot */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2892) if (xas_reload(&xas) != folio)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2893) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2894) bytes = min_t(loff_t, isize - pos,
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2895) min(iov_iter_count(iter), PAGE_SIZE - offset));
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2896)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2897) /* Not _nofault(): its usercopy check reads the folio's order */
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2898) n = copy_page_to_iter(page, offset, bytes, iter);
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2899) copied += n;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2900) iocb->ki_pos += n;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2901) if (n < bytes)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2902) break;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2903) }
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2904) pagefault_enable();
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2905) rcu_read_unlock();
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2906)
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2907) return copied;
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2908) }
7945beac7bf45f6 Kiryl Shutsemau (Meta 2026-10-13 2909)
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
reply other threads:[~2026-10-14 22:19 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=202610150009.YSyjG3rF-lkp@intel.com \
--to=lkp@intel.com \
--cc=error27@gmail.com \
--cc=oe-kbuild@lists.linux.dev \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.