[android-common:android14-kiwi-6.1 249/249] mm/memory.c:5148: warning: Function parameter or member 'mm' not described in 'mm_account_fault'

[android-common:android14-kiwi-6.1 249/249] mm/memory.c:5148: warning: Function parameter or member 'mm' not described in 'mm_account_fault'

[kernel test robot]

tree:   https://android.googlesource.com/kernel/common android14-kiwi-6.1
head:   d2ae202da688a9a795c238142e01da921b3e4b81
commit: a264d8efcb8f7231898dc4cb306ef8e20e6e285e [249/249] BACKPORT: mm: do not increment pgfault stats when page fault handler retries
config: x86_64-allnoconfig (https://download.01.org/0day-ci/archive/20250214/202502142006.SCbQkTrm-lkp@intel.com/config)
compiler: clang version 19.1.3 (https://github.com/llvm/llvm-project ab51eccf88f5321e7c60591c5546b254b6afab99)
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20250214/202502142006.SCbQkTrm-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/202502142006.SCbQkTrm-lkp@intel.com/

All warnings (new ones prefixed by >>):

   mm/memory.c:1714: warning: Function parameter or member 'start_t' not described in 'unmap_vmas'
   mm/memory.c:1714: warning: Function parameter or member 'end_t' not described in 'unmap_vmas'
   mm/memory.c:1714: warning: Function parameter or member 'mm_wr_locked' not described in 'unmap_vmas'
>> mm/memory.c:5148: warning: Function parameter or member 'mm' not described in 'mm_account_fault'


vim +5148 mm/memory.c

^1da177e4c3f41 Linus Torvalds     2005-04-16  5129  
bce617edecada0 Peter Xu           2020-08-11  5130  /**
f0953a1bbaca71 Ingo Molnar        2021-05-06  5131   * mm_account_fault - Do page fault accounting
bce617edecada0 Peter Xu           2020-08-11  5132   *
bce617edecada0 Peter Xu           2020-08-11  5133   * @regs: the pt_regs struct pointer.  When set to NULL, will skip accounting
bce617edecada0 Peter Xu           2020-08-11  5134   *        of perf event counters, but we'll still do the per-task accounting to
bce617edecada0 Peter Xu           2020-08-11  5135   *        the task who triggered this page fault.
bce617edecada0 Peter Xu           2020-08-11  5136   * @address: the faulted address.
bce617edecada0 Peter Xu           2020-08-11  5137   * @flags: the fault flags.
bce617edecada0 Peter Xu           2020-08-11  5138   * @ret: the fault retcode.
bce617edecada0 Peter Xu           2020-08-11  5139   *
f0953a1bbaca71 Ingo Molnar        2021-05-06  5140   * This will take care of most of the page fault accounting.  Meanwhile, it
bce617edecada0 Peter Xu           2020-08-11  5141   * will also include the PERF_COUNT_SW_PAGE_FAULTS_[MAJ|MIN] perf counter
f0953a1bbaca71 Ingo Molnar        2021-05-06  5142   * updates.  However, note that the handling of PERF_COUNT_SW_PAGE_FAULTS should
bce617edecada0 Peter Xu           2020-08-11  5143   * still be in per-arch page fault handlers at the entry of page fault.
bce617edecada0 Peter Xu           2020-08-11  5144   */
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5145  static inline void mm_account_fault(struct mm_struct *mm, struct pt_regs *regs,
bce617edecada0 Peter Xu           2020-08-11  5146  				    unsigned long address, unsigned int flags,
bce617edecada0 Peter Xu           2020-08-11  5147  				    vm_fault_t ret)
bce617edecada0 Peter Xu           2020-08-11 @5148  {
bce617edecada0 Peter Xu           2020-08-11  5149  	bool major;
bce617edecada0 Peter Xu           2020-08-11  5150  
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5151  	/* Incomplete faults will be accounted upon completion. */
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5152  	if (ret & VM_FAULT_RETRY)
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5153  		return;
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5154  
bce617edecada0 Peter Xu           2020-08-11  5155  	/*
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5156  	 * To preserve the behavior of older kernels, PGFAULT counters record
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5157  	 * both successful and failed faults, as opposed to perf counters,
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5158  	 * which ignore failed cases.
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5159  	 */
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5160  	count_vm_event(PGFAULT);
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5161  	count_memcg_event_mm(mm, PGFAULT);
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5162  
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5163  	/*
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5164  	 * Do not account for unsuccessful faults (e.g. when the address wasn't
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5165  	 * valid).  That includes arch_vma_access_permitted() failing before
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5166  	 * reaching here. So this is not a "this many hardware page faults"
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5167  	 * counter.  We should use the hw profiling for that.
bce617edecada0 Peter Xu           2020-08-11  5168  	 */
a264d8efcb8f72 Suren Baghdasaryan 2023-04-19  5169  	if (ret & VM_FAULT_ERROR)
bce617edecada0 Peter Xu           2020-08-11  5170  		return;
bce617edecada0 Peter Xu           2020-08-11  5171  
bce617edecada0 Peter Xu           2020-08-11  5172  	/*
bce617edecada0 Peter Xu           2020-08-11  5173  	 * We define the fault as a major fault when the final successful fault
bce617edecada0 Peter Xu           2020-08-11  5174  	 * is VM_FAULT_MAJOR, or if it retried (which implies that we couldn't
bce617edecada0 Peter Xu           2020-08-11  5175  	 * handle it immediately previously).
bce617edecada0 Peter Xu           2020-08-11  5176  	 */
bce617edecada0 Peter Xu           2020-08-11  5177  	major = (ret & VM_FAULT_MAJOR) || (flags & FAULT_FLAG_TRIED);
bce617edecada0 Peter Xu           2020-08-11  5178  
a2beb5f1efede6 Peter Xu           2020-08-11  5179  	if (major)
a2beb5f1efede6 Peter Xu           2020-08-11  5180  		current->maj_flt++;
a2beb5f1efede6 Peter Xu           2020-08-11  5181  	else
a2beb5f1efede6 Peter Xu           2020-08-11  5182  		current->min_flt++;
a2beb5f1efede6 Peter Xu           2020-08-11  5183  
bce617edecada0 Peter Xu           2020-08-11  5184  	/*
a2beb5f1efede6 Peter Xu           2020-08-11  5185  	 * If the fault is done for GUP, regs will be NULL.  We only do the
a2beb5f1efede6 Peter Xu           2020-08-11  5186  	 * accounting for the per thread fault counters who triggered the
a2beb5f1efede6 Peter Xu           2020-08-11  5187  	 * fault, and we skip the perf event updates.
bce617edecada0 Peter Xu           2020-08-11  5188  	 */
bce617edecada0 Peter Xu           2020-08-11  5189  	if (!regs)
bce617edecada0 Peter Xu           2020-08-11  5190  		return;
bce617edecada0 Peter Xu           2020-08-11  5191  
a2beb5f1efede6 Peter Xu           2020-08-11  5192  	if (major)
bce617edecada0 Peter Xu           2020-08-11  5193  		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MAJ, 1, regs, address);
a2beb5f1efede6 Peter Xu           2020-08-11  5194  	else
bce617edecada0 Peter Xu           2020-08-11  5195  		perf_sw_event(PERF_COUNT_SW_PAGE_FAULTS_MIN, 1, regs, address);
bce617edecada0 Peter Xu           2020-08-11  5196  }
bce617edecada0 Peter Xu           2020-08-11  5197  

:::::: The code at line 5148 was first introduced by commit
:::::: bce617edecada007aee8610fbe2c14d10b8de2f6 mm: do page fault accounting in handle_mm_fault

:::::: TO: Peter Xu <peterx@redhat.com>
:::::: CC: Linus Torvalds <torvalds@linux-foundation.org>

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