tree: https://gitee.com/openeuler/kernel.git OLK-6.6 head: c02455c4cce8052f74fc11c09e9073c01c7f33a4 commit: 3bfa35ccde9e52ca74fed047eb4d89be32dd9e5e [2195/2195] mm/numa: no task_numa_fault() call if PTE is changed config: x86_64-randconfig-2002-20250501 (https://download.01.org/0day-ci/archive/20250626/202506261637.qYfRPb91-lkp@i...) compiler: clang version 20.1.2 (https://github.com/llvm/llvm-project 58df0ef89dd64126512e4ee27b4ac3fd8ddf6247) reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250626/202506261637.qYfRPb91-lkp@i...) 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> | Closes: https://lore.kernel.org/oe-kbuild-all/202506261637.qYfRPb91-lkp@intel.com/ All warnings (new ones prefixed by >>): In file included from mm/memory.c:43: In file included from include/linux/mm.h:2247: include/linux/vmstat.h:522:36: warning: arithmetic between different enumeration types ('enum node_stat_item' and 'enum lru_list') [-Wenum-enum-conversion] 522 | return node_stat_name(NR_LRU_BASE + lru) + 3; // skip "nr_" | ~~~~~~~~~~~ ^ ~~~ In file included from mm/memory.c:44: include/linux/mm_inline.h:47:41: warning: arithmetic between different enumeration types ('enum node_stat_item' and 'enum lru_list') [-Wenum-enum-conversion] 47 | __mod_lruvec_state(lruvec, NR_LRU_BASE + lru, nr_pages); | ~~~~~~~~~~~ ^ ~~~ include/linux/mm_inline.h:49:22: warning: arithmetic between different enumeration types ('enum zone_stat_item' and 'enum lru_list') [-Wenum-enum-conversion] 49 | NR_ZONE_LRU_BASE + lru, nr_pages); | ~~~~~~~~~~~~~~~~ ^ ~~~
mm/memory.c:5277:17: warning: variable 'nr_pages' set but not used [-Wunused-but-set-variable] 5277 | int flags = 0, nr_pages; | ^ 4 warnings generated.
vim +/nr_pages +5277 mm/memory.c f569661eacc258 Baolin Wang 2024-05-11 5266 2b7403035459c7 Souptick Joarder 2018-08-23 5267 static vm_fault_t do_numa_page(struct vm_fault *vmf) d10e63f29488b0 Mel Gorman 2012-10-25 5268 { 82b0f8c39a3869 Jan Kara 2016-12-14 5269 struct vm_area_struct *vma = vmf->vma; d5e6c8d5790088 Kefeng Wang 2023-11-27 5270 struct folio *folio = NULL; d5e6c8d5790088 Kefeng Wang 2023-11-27 5271 int nid = NUMA_NO_NODE; 2cb37064a2b82c Baolin Wang 2024-06-22 5272 bool writable = false, ignore_writable = false; 2cb37064a2b82c Baolin Wang 2024-06-22 5273 bool pte_write_upgrade = vma_wants_manual_pte_write_upgrade(vma); 90572890d20252 Peter Zijlstra 2013-10-07 5274 int last_cpupid; cbee9f88ec1b8d Peter Zijlstra 2012-10-25 5275 int target_nid; 04a8645304500b Aneesh Kumar K.V 2019-03-05 5276 pte_t pte, old_pte; 2cb37064a2b82c Baolin Wang 2024-06-22 @5277 int flags = 0, nr_pages; d10e63f29488b0 Mel Gorman 2012-10-25 5278 d10e63f29488b0 Mel Gorman 2012-10-25 5279 /* e2491c31fb10ae John Hubbard 2024-05-25 5280 * The pte cannot be used safely until we verify, while holding the page e2491c31fb10ae John Hubbard 2024-05-25 5281 * table lock, that its contents have not changed during fault handling. d10e63f29488b0 Mel Gorman 2012-10-25 5282 */ 82b0f8c39a3869 Jan Kara 2016-12-14 5283 spin_lock(vmf->ptl); e2491c31fb10ae John Hubbard 2024-05-25 5284 /* Read the live PTE from the page tables: */ e2491c31fb10ae John Hubbard 2024-05-25 5285 old_pte = ptep_get(vmf->pte); e2491c31fb10ae John Hubbard 2024-05-25 5286 e2491c31fb10ae John Hubbard 2024-05-25 5287 if (unlikely(!pte_same(old_pte, vmf->orig_pte))) { 82b0f8c39a3869 Jan Kara 2016-12-14 5288 pte_unmap_unlock(vmf->pte, vmf->ptl); 3bfa35ccde9e52 Zi Yan 2024-08-09 5289 return 0; 4daae3b4b9e49b Mel Gorman 2012-11-02 5290 } 4daae3b4b9e49b Mel Gorman 2012-11-02 5291 04a8645304500b Aneesh Kumar K.V 2019-03-05 5292 pte = pte_modify(old_pte, vma->vm_page_prot); d10e63f29488b0 Mel Gorman 2012-10-25 5293 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5294 /* 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5295 * Detect now whether the PTE could be writable; this information 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5296 * is only valid while holding the PT lock. 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5297 */ 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5298 writable = pte_write(pte); 2cb37064a2b82c Baolin Wang 2024-06-22 5299 if (!writable && pte_write_upgrade && 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5300 can_change_pte_writable(vma, vmf->address, pte)) 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5301 writable = true; 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5302 d5e6c8d5790088 Kefeng Wang 2023-11-27 5303 folio = vm_normal_folio(vma, vmf->address, pte); d5e6c8d5790088 Kefeng Wang 2023-11-27 5304 if (!folio || folio_is_zone_device(folio)) b99a342d4f11a5 Huang Ying 2021-04-29 5305 goto out_map; d10e63f29488b0 Mel Gorman 2012-10-25 5306 6688cc05473b36 Peter Zijlstra 2013-10-07 5307 /* bea66fbd11af1c Mel Gorman 2015-03-25 5308 * Avoid grouping on RO pages in general. RO pages shouldn't hurt as bea66fbd11af1c Mel Gorman 2015-03-25 5309 * much anyway since they can be in shared cache state. This misses bea66fbd11af1c Mel Gorman 2015-03-25 5310 * the case where a mapping is writable but the process never writes bea66fbd11af1c Mel Gorman 2015-03-25 5311 * to it but pte_write gets cleared during protection updates and bea66fbd11af1c Mel Gorman 2015-03-25 5312 * pte_dirty has unpredictable behaviour between PTE scan updates, bea66fbd11af1c Mel Gorman 2015-03-25 5313 * background writeback, dirty balancing and application behaviour. bea66fbd11af1c Mel Gorman 2015-03-25 5314 */ 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5315 if (!writable) 6688cc05473b36 Peter Zijlstra 2013-10-07 5316 flags |= TNF_NO_GROUP; 6688cc05473b36 Peter Zijlstra 2013-10-07 5317 dabe1d992414a6 Rik van Riel 2013-10-07 5318 /* d5e6c8d5790088 Kefeng Wang 2023-11-27 5319 * Flag if the folio is shared between multiple address spaces. This dabe1d992414a6 Rik van Riel 2013-10-07 5320 * is later used when determining whether to group tasks together dabe1d992414a6 Rik van Riel 2013-10-07 5321 */ b8eba2d7e4a309 David Hildenbrand 2024-05-11 5322 if (folio_likely_mapped_shared(folio) && (vma->vm_flags & VM_SHARED)) dabe1d992414a6 Rik van Riel 2013-10-07 5323 flags |= TNF_SHARED; dabe1d992414a6 Rik van Riel 2013-10-07 5324 d5e6c8d5790088 Kefeng Wang 2023-11-27 5325 nid = folio_nid(folio); 2cb37064a2b82c Baolin Wang 2024-06-22 5326 nr_pages = folio_nr_pages(folio); 33024536bafd91 Huang Ying 2022-07-13 5327 /* 33024536bafd91 Huang Ying 2022-07-13 5328 * For memory tiering mode, cpupid of slow memory page is used 33024536bafd91 Huang Ying 2022-07-13 5329 * to record page access time. So use default value. 33024536bafd91 Huang Ying 2022-07-13 5330 */ 33024536bafd91 Huang Ying 2022-07-13 5331 if ((sysctl_numa_balancing_mode & NUMA_BALANCING_MEMORY_TIERING) && d5e6c8d5790088 Kefeng Wang 2023-11-27 5332 !node_is_toptier(nid)) 33024536bafd91 Huang Ying 2022-07-13 5333 last_cpupid = (-1 & LAST_CPUPID_MASK); 33024536bafd91 Huang Ying 2022-07-13 5334 else 28e1ec746768bb Kefeng Wang 2023-11-27 5335 last_cpupid = folio_last_cpupid(folio); bf7c2ab764599b Kefeng Wang 2023-11-27 5336 target_nid = numa_migrate_prep(folio, vma, vmf->address, nid, &flags); 98fa15f34cb379 Anshuman Khandual 2019-03-05 5337 if (target_nid == NUMA_NO_NODE) { d5e6c8d5790088 Kefeng Wang 2023-11-27 5338 folio_put(folio); b99a342d4f11a5 Huang Ying 2021-04-29 5339 goto out_map; 4daae3b4b9e49b Mel Gorman 2012-11-02 5340 } b99a342d4f11a5 Huang Ying 2021-04-29 5341 pte_unmap_unlock(vmf->pte, vmf->ptl); 6a56ccbcf6c695 David Hildenbrand 2022-11-08 5342 writable = false; 2cb37064a2b82c Baolin Wang 2024-06-22 5343 ignore_writable = true; 4daae3b4b9e49b Mel Gorman 2012-11-02 5344 4daae3b4b9e49b Mel Gorman 2012-11-02 5345 /* Migrate to the requested node */ d5e6c8d5790088 Kefeng Wang 2023-11-27 5346 if (migrate_misplaced_folio(folio, vma, target_nid)) { d5e6c8d5790088 Kefeng Wang 2023-11-27 5347 nid = target_nid; 6688cc05473b36 Peter Zijlstra 2013-10-07 5348 flags |= TNF_MIGRATED; 3bfa35ccde9e52 Zi Yan 2024-08-09 5349 task_numa_fault(last_cpupid, nid, 1, flags); 3bfa35ccde9e52 Zi Yan 2024-08-09 5350 return 0; 3bfa35ccde9e52 Zi Yan 2024-08-09 5351 } 3bfa35ccde9e52 Zi Yan 2024-08-09 5352 074c238177a75f Mel Gorman 2015-03-25 5353 flags |= TNF_MIGRATE_FAIL; c7ad08804fae5b Hugh Dickins 2023-06-08 5354 vmf->pte = pte_offset_map_lock(vma->vm_mm, vmf->pmd, c7ad08804fae5b Hugh Dickins 2023-06-08 5355 vmf->address, &vmf->ptl); c7ad08804fae5b Hugh Dickins 2023-06-08 5356 if (unlikely(!vmf->pte)) 3bfa35ccde9e52 Zi Yan 2024-08-09 5357 return 0; c33c794828f212 Ryan Roberts 2023-06-12 5358 if (unlikely(!pte_same(ptep_get(vmf->pte), vmf->orig_pte))) { b99a342d4f11a5 Huang Ying 2021-04-29 5359 pte_unmap_unlock(vmf->pte, vmf->ptl); d10e63f29488b0 Mel Gorman 2012-10-25 5360 return 0; 3bfa35ccde9e52 Zi Yan 2024-08-09 5361 } b99a342d4f11a5 Huang Ying 2021-04-29 5362 out_map: b99a342d4f11a5 Huang Ying 2021-04-29 5363 /* b99a342d4f11a5 Huang Ying 2021-04-29 5364 * Make it present again, depending on how arch implements b99a342d4f11a5 Huang Ying 2021-04-29 5365 * non-accessible ptes, some can allow access by kernel mode. b99a342d4f11a5 Huang Ying 2021-04-29 5366 */ 2cb37064a2b82c Baolin Wang 2024-06-22 5367 if (folio && folio_test_large(folio)) 2cb37064a2b82c Baolin Wang 2024-06-22 5368 numa_rebuild_large_mapping(vmf, vma, folio, pte, ignore_writable, 2cb37064a2b82c Baolin Wang 2024-06-22 5369 pte_write_upgrade); 2cb37064a2b82c Baolin Wang 2024-06-22 5370 else 2cb37064a2b82c Baolin Wang 2024-06-22 5371 numa_rebuild_single_mapping(vmf, vma, vmf->address, vmf->pte, 2cb37064a2b82c Baolin Wang 2024-06-22 5372 writable); b99a342d4f11a5 Huang Ying 2021-04-29 5373 pte_unmap_unlock(vmf->pte, vmf->ptl); 3bfa35ccde9e52 Zi Yan 2024-08-09 5374 3bfa35ccde9e52 Zi Yan 2024-08-09 5375 if (nid != NUMA_NO_NODE) 3bfa35ccde9e52 Zi Yan 2024-08-09 5376 task_numa_fault(last_cpupid, nid, 1, flags); 3bfa35ccde9e52 Zi Yan 2024-08-09 5377 return 0; d10e63f29488b0 Mel Gorman 2012-10-25 5378 } d10e63f29488b0 Mel Gorman 2012-10-25 5379 :::::: The code at line 5277 was first introduced by commit :::::: 2cb37064a2b82cd8ebeeaaed21259b5912dff51d mm: support multi-size THP numa balancing :::::: TO: Baolin Wang <baolin.wang@linux.alibaba.com> :::::: CC: Liu Shixin <liushixin2@huawei.com> -- 0-DAY CI Kernel Test Service https://github.com/intel/lkp-tests/wiki