Messages in this thread | | | Date | Fri, 22 May 2015 09:50:32 -0700 | From | Mike Kravetz <> | Subject | Re: [RFC v3 PATCH 04/10] mm/hugetlb: expose hugetlb fault mutex for use by fallocate |
| |
On 05/21/2015 11:23 PM, Naoya Horiguchi wrote: > On Thu, May 21, 2015 at 08:47:38AM -0700, Mike Kravetz wrote: >> hugetlb page faults are currently synchronized by the table of >> mutexes (htlb_fault_mutex_table). fallocate code will need to >> synchronize with the page fault code when it allocates or >> deletes pages. Expose interfaces so that fallocate operations >> can be synchronized with page faults. >> >> Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com> >> --- >> include/linux/hugetlb.h | 3 +++ >> mm/hugetlb.c | 23 ++++++++++++++++++++++- >> 2 files changed, 25 insertions(+), 1 deletion(-) >> >> diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h >> index fd337f2..d0d033e 100644 >> --- a/include/linux/hugetlb.h >> +++ b/include/linux/hugetlb.h >> @@ -82,6 +82,9 @@ void putback_active_hugepage(struct page *page); >> bool is_hugepage_active(struct page *page); >> void free_huge_page(struct page *page); >> void hugetlb_fix_reserve_counts(struct inode *inode, bool restore_reserve); >> +u32 hugetlb_fault_mutex_shared_hash(struct address_space *mapping, pgoff_t idx); >> +void hugetlb_fault_mutex_lock(u32 hash); >> +void hugetlb_fault_mutex_unlock(u32 hash); >> >> #ifdef CONFIG_ARCH_WANT_HUGE_PMD_SHARE >> pte_t *huge_pmd_share(struct mm_struct *mm, unsigned long addr, pud_t *pud); >> diff --git a/mm/hugetlb.c b/mm/hugetlb.c >> index 620cc9e..df0d32a 100644 >> --- a/mm/hugetlb.c >> +++ b/mm/hugetlb.c >> @@ -3183,7 +3183,8 @@ static u32 fault_mutex_hash(struct hstate *h, struct mm_struct *mm, >> unsigned long key[2]; >> u32 hash; >> >> - if (vma->vm_flags & VM_SHARED) { >> + /* !vma implies this was called from hugetlbfs fallocate code */ >> + if (!vma || vma->vm_flags & VM_SHARED) { >> key[0] = (unsigned long) mapping; >> key[1] = idx; >> } else { >> @@ -3209,6 +3210,26 @@ static u32 fault_mutex_hash(struct hstate *h, struct mm_struct *mm, >> } >> #endif >> >> +/* >> + * Interfaces to the fault mutex routines for use by hugetlbfs >> + * fallocate code. Faults must be synchronized with page adds or >> + * deletes by fallocate. fallocate only deals with shared mappings. >> + */ >> +u32 hugetlb_fault_mutex_shared_hash(struct address_space *mapping, pgoff_t idx) >> +{ >> + return fault_mutex_hash(NULL, NULL, NULL, mapping, idx, 0); >> +} >> + >> +void hugetlb_fault_mutex_lock(u32 hash) >> +{ >> + mutex_lock(&htlb_fault_mutex_table[hash]); >> +} >> + >> +void hugetlb_fault_mutex_unlock(u32 hash) >> +{ >> + mutex_unlock(&htlb_fault_mutex_table[hash]); >> +} >> + >> int hugetlb_fault(struct mm_struct *mm, struct vm_area_struct *vma, >> unsigned long address, unsigned int flags) >> { > > You introduce new lock/unlock interfaces, so how about making the existing > user of this lock (i.e. hugetlb_fault()) use them? >
I will make that change in the next version of the patch set.
Thanks, -- Mike Kravetz
> Thanks, > Naoya Horiguchi >
| |