Re: [PATCH v3 4/6] mm: shmem: add mTHP support for anonymous shmem

[Dan Carpenter <dan.carpenter@linaro.org>]

Hi Baolin,

kernel test robot noticed the following build warnings:

https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Baolin-Wang/mm-memory-extend-finish_fault-to-support-large-folio/20240530-100805
base:   https://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm.git mm-everything
patch link:    https://lore.kernel.org/r/ec35a23026dd016705d211e85163cabe07681516.1717033868.git.baolin.wang%40linux.alibaba.com
patch subject: [PATCH v3 4/6] mm: shmem: add mTHP support for anonymous shmem
config: powerpc64-randconfig-r071-20240531 (https://download.01.org/0day-ci/archive/20240602/202406020203.14sT311e-lkp@intel.com/config)
compiler: clang version 19.0.0git (https://github.com/llvm/llvm-project bafda89a0944d947fc4b3b5663185e07a397ac30)

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 <dan.carpenter@linaro.org>
| Closes: https://lore.kernel.org/r/202406020203.14sT311e-lkp@intel.com/

smatch warnings:
mm/shmem.c:1766 shmem_alloc_and_add_folio() error: uninitialized symbol 'suitable_orders'.

vim +/suitable_orders +1766 mm/shmem.c

ededbc2c2f28a1 Baolin Wang             2024-05-30  1729  static struct folio *shmem_alloc_and_add_folio(struct vm_fault *vmf,
ededbc2c2f28a1 Baolin Wang             2024-05-30  1730  		gfp_t gfp, struct inode *inode, pgoff_t index,
ededbc2c2f28a1 Baolin Wang             2024-05-30  1731  		struct mm_struct *fault_mm, unsigned long orders)
800d8c63b2e989 Kirill A. Shutemov      2016-07-26  1732  {
3022fd7af9604d Hugh Dickins            2023-09-29  1733  	struct address_space *mapping = inode->i_mapping;
0f0796945614b7 Mike Rapoport           2017-09-06  1734  	struct shmem_inode_info *info = SHMEM_I(inode);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1735  	struct vm_area_struct *vma = vmf ? vmf->vma : NULL;
ededbc2c2f28a1 Baolin Wang             2024-05-30  1736  	unsigned long suitable_orders;
ededbc2c2f28a1 Baolin Wang             2024-05-30  1737  	struct folio *folio = NULL;
3022fd7af9604d Hugh Dickins            2023-09-29  1738  	long pages;
ededbc2c2f28a1 Baolin Wang             2024-05-30  1739  	int error, order;
800d8c63b2e989 Kirill A. Shutemov      2016-07-26  1740  
396bcc5299c281 Matthew Wilcox (Oracle  2020-04-06  1741) 	if (!IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE))
ededbc2c2f28a1 Baolin Wang             2024-05-30  1742  		orders = 0;
800d8c63b2e989 Kirill A. Shutemov      2016-07-26  1743  
ededbc2c2f28a1 Baolin Wang             2024-05-30  1744  	if (orders > 0) {
ededbc2c2f28a1 Baolin Wang             2024-05-30  1745  		if (vma && vma_is_anon_shmem(vma)) {
ededbc2c2f28a1 Baolin Wang             2024-05-30  1746  			suitable_orders = anon_shmem_suitable_orders(inode, vmf,
ededbc2c2f28a1 Baolin Wang             2024-05-30  1747  							mapping, index, orders);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1748  		} else if (orders & BIT(HPAGE_PMD_ORDER)) {
3022fd7af9604d Hugh Dickins            2023-09-29  1749  			pages = HPAGE_PMD_NR;
ededbc2c2f28a1 Baolin Wang             2024-05-30  1750  			suitable_orders = BIT(HPAGE_PMD_ORDER);
3022fd7af9604d Hugh Dickins            2023-09-29  1751  			index = round_down(index, HPAGE_PMD_NR);
3022fd7af9604d Hugh Dickins            2023-09-29  1752  
3022fd7af9604d Hugh Dickins            2023-09-29  1753  			/*
3022fd7af9604d Hugh Dickins            2023-09-29  1754  			 * Check for conflict before waiting on a huge allocation.
3022fd7af9604d Hugh Dickins            2023-09-29  1755  			 * Conflict might be that a huge page has just been allocated
3022fd7af9604d Hugh Dickins            2023-09-29  1756  			 * and added to page cache by a racing thread, or that there
3022fd7af9604d Hugh Dickins            2023-09-29  1757  			 * is already at least one small page in the huge extent.
3022fd7af9604d Hugh Dickins            2023-09-29  1758  			 * Be careful to retry when appropriate, but not forever!
3022fd7af9604d Hugh Dickins            2023-09-29  1759  			 * Elsewhere -EEXIST would be the right code, but not here.
3022fd7af9604d Hugh Dickins            2023-09-29  1760  			 */
3022fd7af9604d Hugh Dickins            2023-09-29  1761  			if (xa_find(&mapping->i_pages, &index,
3022fd7af9604d Hugh Dickins            2023-09-29  1762  				    index + HPAGE_PMD_NR - 1, XA_PRESENT))
3022fd7af9604d Hugh Dickins            2023-09-29  1763  				return ERR_PTR(-E2BIG);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1764  		}

suitable_orders uninitialized on else path.

52cd3b074050dd Lee Schermerhorn        2008-04-28  1765  
ededbc2c2f28a1 Baolin Wang             2024-05-30 @1766  		order = highest_order(suitable_orders);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1767  		while (suitable_orders) {
ededbc2c2f28a1 Baolin Wang             2024-05-30  1768  			pages = 1UL << order;
ededbc2c2f28a1 Baolin Wang             2024-05-30  1769  			index = round_down(index, pages);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1770  			folio = shmem_alloc_folio(gfp, order, info, index);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1771  			if (folio)
ededbc2c2f28a1 Baolin Wang             2024-05-30  1772  				goto allocated;
ededbc2c2f28a1 Baolin Wang             2024-05-30  1773  
ededbc2c2f28a1 Baolin Wang             2024-05-30  1774  			if (pages == HPAGE_PMD_NR)
3022fd7af9604d Hugh Dickins            2023-09-29  1775  				count_vm_event(THP_FILE_FALLBACK);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1776  			order = next_order(&suitable_orders, order);
ededbc2c2f28a1 Baolin Wang             2024-05-30  1777  		}
3022fd7af9604d Hugh Dickins            2023-09-29  1778  	} else {
3022fd7af9604d Hugh Dickins            2023-09-29  1779  		pages = 1;

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki