On Wed, 10 Sep 2014 16:24:46 -0700 Peter Feiner <pfeiner@xxxxxxxxxx> wrote: > If a /proc/pid/pagemap read spans a [VMA, an unmapped region, then a > VM_SOFTDIRTY VMA], the virtual pages in the unmapped region are reported > as softdirty. Here's a program to demonstrate the bug: > > int main() { > const uint64_t PAGEMAP_SOFTDIRTY = 1ul << 55; > uint64_t pme[3]; > int fd = open("/proc/self/pagemap", O_RDONLY);; > char *m = mmap(NULL, 3 * getpagesize(), PROT_READ, > MAP_ANONYMOUS | MAP_SHARED, -1, 0); > munmap(m + getpagesize(), getpagesize()); > pread(fd, pme, 24, (unsigned long) m / getpagesize() * 8); > assert(pme[0] & PAGEMAP_SOFTDIRTY); /* passes */ > assert(!(pme[1] & PAGEMAP_SOFTDIRTY)); /* fails */ > assert(pme[2] & PAGEMAP_SOFTDIRTY); /* passes */ > return 0; > } > > (Note that all pages in new VMAs are softdirty until cleared). > > Tested: > Used the program given above. I'm going to include this code in > a selftest in the future. > > ... > > --- a/fs/proc/task_mmu.c > +++ b/fs/proc/task_mmu.c > > ... > > @@ -1048,32 +1048,51 @@ static int pagemap_pte_range(pmd_t *pmd, unsigned long addr, unsigned long end, > > if (pmd_trans_unstable(pmd)) > return 0; > - for (; addr != end; addr += PAGE_SIZE) { > - int flags2; > - > - /* check to see if we've left 'vma' behind > - * and need a new, higher one */ > - if (vma && (addr >= vma->vm_end)) { > - vma = find_vma(walk->mm, addr); > - if (vma && (vma->vm_flags & VM_SOFTDIRTY)) > - flags2 = __PM_SOFT_DIRTY; > - else > - flags2 = 0; > - pme = make_pme(PM_NOT_PRESENT(pm->v2) | PM_STATUS2(pm->v2, flags2)); > + > + while (1) { > + unsigned long vm_start = end; Did you really mean to do that? If so, perhaps a little comment to explain how it works? > + unsigned long vm_end = end; > + unsigned long vm_flags = 0; > + > + if (vma) { > + /* > + * We can't possibly be in a hugetlb VMA. In general, > + * for a mm_walk with a pmd_entry and a hugetlb_entry, > + * the pmd_entry can only be called on addresses in a > + * hugetlb if the walk starts in a non-hugetlb VMA and > + * spans a hugepage VMA. Since pagemap_read walks are > + * PMD-sized and PMD-aligned, this will never be true. > + */ > + BUG_ON(is_vm_hugetlb_page(vma)); > + vm_start = vma->vm_start; > + vm_end = min(end, vma->vm_end); > + vm_flags = vma->vm_flags; > + } > + > + /* Addresses before the VMA. */ > + for (; addr < vm_start; addr += PAGE_SIZE) { > + pagemap_entry_t pme = make_pme(PM_NOT_PRESENT(pm->v2)); > + > + err = add_to_pagemap(addr, &pme, pm); > + if (err) > + return err; > > ... > -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx"> email@xxxxxxxxx </a>