From: Axel Rasmussen <axelrasmussen@google.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>,
Alexey Dobriyan <adobriyan@gmail.com>,
Andrea Arcangeli <aarcange@redhat.com>,
Andrew Morton <akpm@linux-foundation.org>,
Anshuman Khandual <anshuman.khandual@arm.com>,
Catalin Marinas <catalin.marinas@arm.com>,
Chinwen Chang <chinwen.chang@mediatek.com>,
Huang Ying <ying.huang@intel.com>,
Ingo Molnar <mingo@redhat.com>,
kbuild-all@lists.01.org, clang-built-linux@googlegroups.com,
Linux Memory Management List <linux-mm@kvack.org>
Subject: Re: [PATCH v2 7/9] userfaultfd: add UFFDIO_CONTINUE ioctl
Date: Mon, 25 Jan 2021 09:50:46 -0800 [thread overview]
Message-ID: <CAJHvVcjHmSNXVNxtR3gJSxUnB0yifr+CSi9GnZDXrHeQwi2EpA@mail.gmail.com> (raw)
In-Reply-To: <202101252146.VjG3nQXC-lkp@intel.com>
This build error seems to be caused by a missing #ifdef
CONFIG_USERFAULTFD. I'll send a v3 with this fix, after waiting for
other feedback on the v2 version.
On Mon, Jan 25, 2021 at 5:37 AM kernel test robot <lkp@intel.com> wrote:
>
> Hi Axel,
>
> Thank you for the patch! Yet something to improve:
>
> [auto build test ERROR on arm64/for-next/core]
> [also build test ERROR on powerpc/next s390/features tip/perf/core linus/master v5.11-rc5 next-20210122]
> [cannot apply to hp-parisc/for-next hnaz-linux-mm/master ia64/next sparc-next/master sparc/master]
> [If your patch is applied to the wrong git tree, kindly drop us a note.
> And when submitting patch, we suggest to use '--base' as documented in
> https://git-scm.com/docs/git-format-patch]
>
> url: https://github.com/0day-ci/linux/commits/Axel-Rasmussen/userfaultfd-add-minor-fault-handling/20210125-104035
> base: https://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git for-next/core
> config: x86_64-randconfig-a013-20210125 (attached as .config)
> compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 12d0753aca22896fda2cf76781b0ee0524d55065)
> reproduce (this is a W=1 build):
> wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
> chmod +x ~/bin/make.cross
> # install x86_64 cross compiling tool for clang build
> # apt-get install binutils-x86-64-linux-gnu
> # https://github.com/0day-ci/linux/commit/b8fb53c3a341b9b853aa3286286c807088311dbd
> git remote add linux-review https://github.com/0day-ci/linux
> git fetch --no-tags linux-review Axel-Rasmussen/userfaultfd-add-minor-fault-handling/20210125-104035
> git checkout b8fb53c3a341b9b853aa3286286c807088311dbd
> # save the attached .config to linux build tree
> COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=x86_64
>
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kernel test robot <lkp@intel.com>
>
> All errors (new ones prefixed by >>):
>
> In file included from mm/hugetlb.c:39:
> include/linux/hugetlb.h:142:10: warning: declaration of 'enum mcopy_atomic_mode' will not be visible outside of this function [-Wvisibility]
> enum mcopy_atomic_mode mode,
> ^
> mm/hugetlb.c:4659:13: warning: declaration of 'enum mcopy_atomic_mode' will not be visible outside of this function [-Wvisibility]
> enum mcopy_atomic_mode mode,
> ^
> >> mm/hugetlb.c:4654:5: error: conflicting types for 'hugetlb_mcopy_atomic_pte'
> int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm,
> ^
> include/linux/hugetlb.h:138:5: note: previous declaration is here
> int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm, pte_t *dst_pte,
> ^
> >> mm/hugetlb.c:4659:31: error: variable has incomplete type 'enum mcopy_atomic_mode'
> enum mcopy_atomic_mode mode,
> ^
> mm/hugetlb.c:4659:13: note: forward declaration of 'enum mcopy_atomic_mode'
> enum mcopy_atomic_mode mode,
> ^
> >> mm/hugetlb.c:4675:25: error: use of undeclared identifier 'MCOPY_ATOMIC_CONTINUE'
> if (!*pagep && mode != MCOPY_ATOMIC_CONTINUE) {
> ^
> mm/hugetlb.c:4692:21: error: use of undeclared identifier 'MCOPY_ATOMIC_CONTINUE'
> } else if (mode == MCOPY_ATOMIC_CONTINUE) {
> ^
> mm/hugetlb.c:4711:27: error: use of undeclared identifier 'MCOPY_ATOMIC_CONTINUE'
> if (vm_shared && mode != MCOPY_ATOMIC_CONTINUE) {
> ^
> mm/hugetlb.c:4771:14: error: use of undeclared identifier 'MCOPY_ATOMIC_CONTINUE'
> if (mode != MCOPY_ATOMIC_CONTINUE)
> ^
> 2 warnings and 6 errors generated.
>
>
> vim +/hugetlb_mcopy_atomic_pte +4654 mm/hugetlb.c
>
> 86e5216f8d8aa25 Adam Litke 2006-01-06 4649
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4650 /*
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4651 * Used by userfaultfd UFFDIO_COPY. Based on mcopy_atomic_pte with
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4652 * modifications for huge pages.
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4653 */
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 @4654 int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4655 pte_t *dst_pte,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4656 struct vm_area_struct *dst_vma,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4657 unsigned long dst_addr,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4658 unsigned long src_addr,
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 @4659 enum mcopy_atomic_mode mode,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4660 struct page **pagep)
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4661 {
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4662 struct address_space *mapping;
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4663 pgoff_t idx;
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4664 unsigned long size;
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4665 int vm_shared = dst_vma->vm_flags & VM_SHARED;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4666 struct hstate *h = hstate_vma(dst_vma);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4667 pte_t _dst_pte;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4668 spinlock_t *ptl;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4669 int ret;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4670 struct page *page;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4671
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4672 mapping = dst_vma->vm_file->f_mapping;
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4673 idx = vma_hugecache_offset(h, dst_vma, dst_addr);
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4674
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 @4675 if (!*pagep && mode != MCOPY_ATOMIC_CONTINUE) {
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4676 ret = -ENOMEM;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4677 page = alloc_huge_page(dst_vma, dst_addr, 0);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4678 if (IS_ERR(page))
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4679 goto out;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4680
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4681 ret = copy_huge_page_from_user(page,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4682 (const void __user *) src_addr,
> 810a56b943e265b Mike Kravetz 2017-02-22 4683 pages_per_huge_page(h), false);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4684
> c1e8d7c6a7a682e Michel Lespinasse 2020-06-08 4685 /* fallback to copy_from_user outside mmap_lock */
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4686 if (unlikely(ret)) {
> 9e368259ad98835 Andrea Arcangeli 2018-11-30 4687 ret = -ENOENT;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4688 *pagep = page;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4689 /* don't free the page */
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4690 goto out;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4691 }
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4692 } else if (mode == MCOPY_ATOMIC_CONTINUE) {
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4693 ret = -EFAULT;
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4694 page = find_lock_page(mapping, idx);
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4695 *pagep = NULL;
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4696 if (!page)
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4697 goto out;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4698 } else {
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4699 page = *pagep;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4700 *pagep = NULL;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4701 }
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4702
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4703 /*
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4704 * The memory barrier inside __SetPageUptodate makes sure that
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4705 * preceding stores to the page contents become visible before
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4706 * the set_pte_at() write.
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4707 */
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4708 __SetPageUptodate(page);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4709
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4710 /* Add shared, newly allocated pages to the page cache. */
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4711 if (vm_shared && mode != MCOPY_ATOMIC_CONTINUE) {
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4712 size = i_size_read(mapping->host) >> huge_page_shift(h);
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4713 ret = -EFAULT;
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4714 if (idx >= size)
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4715 goto out_release_nounlock;
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4716
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4717 /*
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4718 * Serialization between remove_inode_hugepages() and
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4719 * huge_add_to_page_cache() below happens through the
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4720 * hugetlb_fault_mutex_table that here must be hold by
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4721 * the caller.
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4722 */
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4723 ret = huge_add_to_page_cache(page, mapping, idx);
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4724 if (ret)
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4725 goto out_release_nounlock;
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4726 }
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4727
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4728 ptl = huge_pte_lockptr(h, dst_mm, dst_pte);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4729 spin_lock(ptl);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4730
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4731 /*
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4732 * Recheck the i_size after holding PT lock to make sure not
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4733 * to leave any page mapped (as page_mapped()) beyond the end
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4734 * of the i_size (remove_inode_hugepages() is strict about
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4735 * enforcing that). If we bail out here, we'll also leave a
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4736 * page in the radix tree in the vm_shared case beyond the end
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4737 * of the i_size, but remove_inode_hugepages() will take care
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4738 * of it as soon as we drop the hugetlb_fault_mutex_table.
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4739 */
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4740 size = i_size_read(mapping->host) >> huge_page_shift(h);
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4741 ret = -EFAULT;
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4742 if (idx >= size)
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4743 goto out_release_unlock;
> 1e3921471354244 Andrea Arcangeli 2017-11-02 4744
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4745 ret = -EEXIST;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4746 if (!huge_pte_none(huge_ptep_get(dst_pte)))
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4747 goto out_release_unlock;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4748
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4749 if (vm_shared) {
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4750 page_dup_rmap(page, true);
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4751 } else {
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4752 ClearPagePrivate(page);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4753 hugepage_add_new_anon_rmap(page, dst_vma, dst_addr);
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4754 }
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4755
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4756 _dst_pte = make_huge_pte(dst_vma, page, dst_vma->vm_flags & VM_WRITE);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4757 if (dst_vma->vm_flags & VM_WRITE)
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4758 _dst_pte = huge_pte_mkdirty(_dst_pte);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4759 _dst_pte = pte_mkyoung(_dst_pte);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4760
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4761 set_huge_pte_at(dst_mm, dst_addr, dst_pte, _dst_pte);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4762
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4763 (void)huge_ptep_set_access_flags(dst_vma, dst_addr, dst_pte, _dst_pte,
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4764 dst_vma->vm_flags & VM_WRITE);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4765 hugetlb_count_add(pages_per_huge_page(h), dst_mm);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4766
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4767 /* No need to invalidate - it was non-present before */
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4768 update_mmu_cache(dst_vma, dst_addr, dst_pte);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4769
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4770 spin_unlock(ptl);
> b8fb53c3a341b9b Axel Rasmussen 2021-01-22 4771 if (mode != MCOPY_ATOMIC_CONTINUE)
> cb6acd01e2e43fd Mike Kravetz 2019-02-28 4772 set_page_huge_active(page);
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4773 if (vm_shared)
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4774 unlock_page(page);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4775 ret = 0;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4776 out:
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4777 return ret;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4778 out_release_unlock:
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4779 spin_unlock(ptl);
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4780 if (vm_shared)
> 1c9e8def43a3452 Mike Kravetz 2017-02-22 4781 unlock_page(page);
> 5af10dfd0afc559 Andrea Arcangeli 2017-08-10 4782 out_release_nounlock:
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4783 put_page(page);
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4784 goto out;
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4785 }
> 8fb5debc5fcd450 Mike Kravetz 2017-02-22 4786
>
> ---
> 0-DAY CI Kernel Test Service, Intel Corporation
> https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
next prev parent reply other threads:[~2021-01-25 17:51 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-22 21:29 [PATCH 0/9] userfaultfd: add minor fault handling Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 1/9] hugetlb: Pass vma into huge_pte_alloc() Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 2/9] hugetlb/userfaultfd: Forbid huge pmd sharing when uffd enabled Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 3/9] mm/hugetlb: Move flush_hugetlb_tlb_range() into hugetlb.h Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 4/9] hugetlb/userfaultfd: Unshare all pmds for hugetlbfs when register wp Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 5/9] userfaultfd: add minor fault registration mode Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 6/9] userfaultfd: disable huge PMD sharing for MINOR registered VMAs Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 7/9] userfaultfd: add UFFDIO_CONTINUE ioctl Axel Rasmussen
2021-01-25 7:53 ` kernel test robot
2021-01-25 8:31 ` kernel test robot
2021-01-25 13:37 ` kernel test robot
2021-01-25 17:50 ` Axel Rasmussen [this message]
2021-01-22 21:29 ` [PATCH v2 8/9] userfaultfd: update documentation to describe minor fault handling Axel Rasmussen
2021-01-22 21:29 ` [PATCH v2 9/9] userfaultfd/selftests: add test exercising " Axel Rasmussen
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=CAJHvVcjHmSNXVNxtR3gJSxUnB0yifr+CSi9GnZDXrHeQwi2EpA@mail.gmail.com \
--to=axelrasmussen@google.com \
--cc=aarcange@redhat.com \
--cc=adobriyan@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=anshuman.khandual@arm.com \
--cc=catalin.marinas@arm.com \
--cc=chinwen.chang@mediatek.com \
--cc=clang-built-linux@googlegroups.com \
--cc=kbuild-all@lists.01.org \
--cc=linux-mm@kvack.org \
--cc=mingo@redhat.com \
--cc=viro@zeniv.linux.org.uk \
--cc=ying.huang@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox