[davidhildenbrand:pfn_cow 3/5] arch/x86/mm/pat/memtype.c:199:36: warning: right shift count >= width of type

[kernel test robot <lkp@intel.com>]

tree:   https://github.com/davidhildenbrand/linux pfn_cow
head:   e522fd14b455b4012d7c9388026dc849daf6f8b5
commit: 17725734e525b4cc4ae1f26107e45c0ca19c17a6 [3/5] x86/mm/pat: store PAT information for VM_PAT in VMA
config: i386-buildonly-randconfig-003-20240308 (https://download.01.org/0day-ci/archive/20240308/202403081901.TbUIkbal-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20240308/202403081901.TbUIkbal-lkp@intel.com/reproduce)

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/202403081901.TbUIkbal-lkp@intel.com/

All warnings (new ones prefixed by >>):

   In file included from arch/x86/include/asm/bug.h:87,
                    from include/linux/bug.h:5,
                    from include/linux/fortify-string.h:5,
                    from include/linux/string.h:292,
                    from include/linux/seq_file.h:6,
                    from arch/x86/mm/pat/memtype.c:34:
   arch/x86/mm/pat/memtype.c: In function 'vma_set_pat_info':
>> arch/x86/mm/pat/memtype.c:199:36: warning: right shift count >= width of type [-Wshift-count-overflow]
     199 |         WARN_ON_ONCE(vma->vm_pgoff >> 61);
         |                                    ^~
   include/asm-generic/bug.h:111:32: note: in definition of macro 'WARN_ON_ONCE'
     111 |         int __ret_warn_on = !!(condition);                      \
         |                                ^~~~~~~~~


vim +199 arch/x86/mm/pat/memtype.c

   178	
   179	static inline void vma_set_pat_info(struct vm_area_struct *vma,
   180					    resource_size_t paddr, pgprot_t pgprot)
   181	{
   182		const int64_t diff = (paddr >> PAGE_SHIFT) - vma->vm_pgoff;
   183	
   184		/*
   185		 * With VM_PAT, we will indicate that the whole VMA maps a contiguous
   186		 * PFN range. To avoid walking page tables, we store the start PFN
   187		 * and the requested cachemode in the VMA.
   188		 *
   189		 * In order to obtain the start PFN in a VMA even after the VMA was
   190		 * split or mremapped without additional hooks, store the diff
   191		 * of the PFN to the vm_pgoff. Merging o extending VMAs does not apply
   192		 * to "special" mappings.
   193		 *
   194		 * If our vm_pgoff would exceed 61 bit, we would not be able to reuse
   195		 * 3 bit for our cachemode. We really don't expect that big vm_pgoffs
   196		 * in VM_PFNMAP mappings, so let's keep it simple and avoid storing
   197		 * the cachemode separately.
   198		 */
 > 199		WARN_ON_ONCE(vma->vm_pgoff >> 61);
   200		BUILD_BUG_ON((_PAGE_CACHE_MODE_NUM - 1) & ~0x7);
   201	
   202		vm_flags_set(vma, VM_PAT);
   203		vma->pat_info = (diff << 3) | pgprot2cachemode(pgprot);
   204	}
   205	

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