Re: Warning on mremapped uffd-wp memory

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 

On 06.08.24 19:15, David Hildenbrand wrote:

On 06.08.24 18:58, Ryan Roberts wrote:

On 06/08/2024 17:37, David Hildenbrand wrote:

On 06.08.24 17:15, Ryan Roberts wrote:

Hi Peter, David,

syzkaller has found an issue (at least on arm64, but I suspect it will be
visible on x86_64 too) that triggers the following warning:

[ 2291.836518] ------------[ cut here ]------------
[ 2291.836528] WARNING: CPU: 3 PID: 9056 at mm/page_table_check.c:207
__page_table_check_ptes_set+0x22c/0x248
[ 2291.836541] Modules linked in:
[ 2291.836549] CPU: 3 UID: 1000 PID: 9056 Comm: bug Tainted: G
W          6.11.0-rc2-dirty #2
[ 2291.836554] Tainted: [W]=WARN
[ 2291.836557] Hardware name: linux,dummy-virt (DT)
[ 2291.836559] pstate: 80400005 (Nzcv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--)
[ 2291.836564] pc : __page_table_check_ptes_set+0x22c/0x248
[ 2291.836568] lr : ptep_modify_prot_commit+0x24c/0x2b0
[ 2291.836573] sp : ffff80008ca6ba20
[ 2291.836575] x29: ffff80008ca6ba20 x28: ffff186392d1eb00 x27: 0000000020ffd000
[ 2291.836598] x26: 0010000000000001 x25: 0000000000000001 x24: 0000000000000000
[ 2291.836605] x23: 04e800018c738f43 x22: 0000000000000001 x21: ffff1863824163c0
[ 2291.836612] x20: 04e800018c738f43 x19: 04e800018c738f43 x18: 0000fffff7f87fff
[ 2291.836619] x17: 0000000000000000 x16: 1fffe30c748d22a1 x15: 0060000000000fc3
[ 2291.836625] x14: 0000000000000000 x13: 0000000020ffd000 x12: 0000fffff7f87fff
[ 2291.836631] x11: 0000000020ffd000 x10: 0000000000000000 x9 : ffffbcab99e3ab84
[ 2291.836638] x8 : ffff186382b8f000 x7 : 0000000020ffe000 x6 : 0000000020ffd000
[ 2291.836644] x5 : ffff186392d1eb00 x4 : 04e800018c738f43 x3 : 0000000000000001
[ 2291.836650] x2 : 04e800018c738f43 x1 : ffff18639fe01fe8 x0 : ffffbcab9ce56780
[ 2291.836657] Call trace:
[ 2291.836659]  __page_table_check_ptes_set+0x22c/0x248
[ 2291.836664]  ptep_modify_prot_commit+0x24c/0x2b0
[ 2291.836667]  change_protection+0x8a0/0x1100
[ 2291.836672]  mprotect_fixup+0x124/0x2d0
[ 2291.836675]  do_mprotect_pkey.constprop.0+0x29c/0x460
[ 2291.836679]  __arm64_sys_mprotect+0x24/0xf8
[ 2291.836682]  invoke_syscall+0x50/0x120
[ 2291.836690]  el0_svc_common.constprop.0+0x48/0xf0
[ 2291.836694]  do_el0_svc+0x24/0x38
[ 2291.836699]  el0_svc+0x34/0xe0
[ 2291.836705]  el0t_64_sync_handler+0x100/0x130
[ 2291.836709]  el0t_64_sync+0x190/0x198
[ 2291.836713] ---[ end trace 0000000000000000 ]---

The generated program (see below) mmaps a 16M region (RWX). It then mlocks all
current and future memory.

Next, it registers 12K (3 pages) for use with UFFD-WP, and marks 4 pages
UFFD-WP'ed. This returns ENOENT because we only registered 3 pages, but those 3
pages are still UFFD-WP'ed in their PTE, so this error is not relavent to the
bug. At this point, there is a single VMA covering the 12K, with VM_UFFD_WP set,
amongst other flags:

     20ffb000-20ffe000 rwxp 00000000 00:00 0
     Size:                 12 kB
     KernelPageSize:        4 kB
     MMUPageSize:           4 kB
     Rss:                  12 kB
     Pss:                  12 kB
     Pss_Dirty:            12 kB
     Shared_Clean:          0 kB
     Shared_Dirty:          0 kB
     Private_Clean:         0 kB
     Private_Dirty:        12 kB
     Referenced:           12 kB
     Anonymous:            12 kB
     KSM:                   0 kB
     LazyFree:              0 kB
     AnonHugePages:         0 kB
     ShmemPmdMapped:        0 kB
     FilePmdMapped:         0 kB
     Shared_Hugetlb:        0 kB
     Private_Hugetlb:       0 kB
     Swap:                  0 kB
     SwapPss:               0 kB
     Locked:               12 kB
     THPeligible:           0
     VmFlags: rd wr ex mr mw me uw lo ac

Next we mremap the first page to the address where the last page was previously
mapped, with MREMAP_DONTUNMAP. This leads to 2 VMAs, but the new one doesn't
have VM_UFFD_WP set (Note also that the original VMA no longer has VM_LOCKED
which seems wrong to me, but I'll ignore that for now):

     20ffb000-20ffd000 rwxp 00000000 00:00 0
     Size:                  8 kB
     KernelPageSize:        4 kB
     MMUPageSize:           4 kB
     Rss:                   4 kB
     Pss:                   4 kB
     Pss_Dirty:             4 kB
     Shared_Clean:          0 kB
     Shared_Dirty:          0 kB
     Private_Clean:         0 kB
     Private_Dirty:         4 kB
     Referenced:            4 kB
     Anonymous:             4 kB
     KSM:                   0 kB
     LazyFree:              0 kB
     AnonHugePages:         0 kB
     ShmemPmdMapped:        0 kB
     FilePmdMapped:         0 kB
     Shared_Hugetlb:        0 kB
     Private_Hugetlb:       0 kB
     Swap:                  0 kB
     SwapPss:               0 kB
     Locked:                0 kB
     THPeligible:           0
     VmFlags: rd wr ex mr mw me uw ac
     20ffd000-20ffe000 rwxp 00000000 00:00 0
     Size:                  4 kB
     KernelPageSize:        4 kB
     MMUPageSize:           4 kB
     Rss:                   4 kB
     Pss:                   4 kB
     Pss_Dirty:             4 kB
     Shared_Clean:          0 kB
     Shared_Dirty:          0 kB
     Private_Clean:         0 kB
     Private_Dirty:         4 kB
     Referenced:            4 kB
     Anonymous:             4 kB
     KSM:                   0 kB
     LazyFree:              0 kB
     AnonHugePages:         0 kB
     ShmemPmdMapped:        0 kB
     FilePmdMapped:         0 kB
     Shared_Hugetlb:        0 kB
     Private_Hugetlb:       0 kB
     Swap:                  0 kB
     SwapPss:               0 kB
     Locked:                4 kB
     THPeligible:           0
     VmFlags: rd wr ex mr mw me lo ac

Finally we try to mprotect that last 4K region to remove X, and we get the
warning saying the PTE has both the UFFD-WP and WRITE bits set.

I'm guessing this is because the VM_UFFD_WP flag got spuriously dropped when
creating the final 4K VMA and so mprotect's can_change_pte_writable() check
incorrectly allowed the pte to be marked writable. But the mremap man page is
not very clear on the semantics when interacting with uffd regions; perhaps
uffd-wp bit should have been cleared when mremapping the ptes?

I'm hoping you can advice on the expected semantics and we can figure out how to
solve this?


The reproducer is as follows (with a few annotations added by me):

"""
// autogenerated by syzkaller (https://github.com/google/syzkaller)

#define _GNU_SOURCE

#include <endian.h>
#include <stdint.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/syscall.h>
#include <sys/types.h>
#include <unistd.h>

#ifndef __NR_ioctl
#define __NR_ioctl 29
#endif
#ifndef __NR_mlockall
#define __NR_mlockall 230
#endif
#ifndef __NR_mmap
#define __NR_mmap 222
#endif
#ifndef __NR_mprotect
#define __NR_mprotect 226
#endif
#ifndef __NR_mremap
#define __NR_mremap 216
#endif
#ifndef __NR_userfaultfd
#define __NR_userfaultfd 282
#endif

uint64_t r[1] = {0xffffffffffffffff};

int main(void)
{
      intptr_t res = 0;

      syscall(__NR_mmap, /*addr=*/0x1ffff000ul, /*len=*/0x1000ul, /*prot=*/0ul,
/*flags=MAP_FIXED|MAP_ANONYMOUS|MAP_PRIVATE*/0x32ul, /*fd=*/-1, /*offset=*/0ul);
      syscall(__NR_mmap, /*addr=*/0x20000000ul, /*len=*/0x1000000ul,
/*prot=PROT_WRITE|PROT_READ|PROT_EXEC*/7ul,
/*flags=MAP_FIXED|MAP_ANONYMOUS|MAP_PRIVATE*/0x32ul, /*fd=*/-1, /*offset=*/0ul);
      syscall(__NR_mmap, /*addr=*/0x21000000ul, /*len=*/0x1000ul, /*prot=*/0ul,
/*flags=MAP_FIXED|MAP_ANONYMOUS|MAP_PRIVATE*/0x32ul, /*fd=*/-1, /*offset=*/0ul);

      write(1, "executing program\n", sizeof("executing program\n") - 1);

      // userfaultfd(UFFD_USER_MODE_ONLY)        = 3
      res = syscall(__NR_userfaultfd, /*flags=UFFD_USER_MODE_ONLY*/1ul);
      if (res != -1)
          r[0] = res;

      // ioctl(3, UFFDIO_API, {api=0xaa, features=0 =>
features=UFFD_FEATURE_PAGEFAULT_FLAG_WP|UFFD_FEATURE_EVENT_FORK|UFFD_FEATURE_EVENT_REMAP|UFFD_FEATURE_EVENT_REMOVE|UFFD_FEATURE_MISSING_HUGETLBFS|UFFD_FEATURE_MISSING_SHMEM|UFFD_FEATURE_EVENT_UNMAP|UFFD_FEATURE_SIGBUS|UFFD_FEATURE_THREAD_ID|UFFD_FEATURE_MINOR_HUGETLBFS|UFFD_FEATURE_MINOR_SHMEM|0x1f800, ioctls=1<<_UFFDIO_REGISTER|1<<_UFFDIO_UNREGISTER|1<<_UFFDIO_API}) = 0
      *(uint64_t*)0x20000000 = 0xaa;
      *(uint64_t*)0x20000008 = 0;
      *(uint64_t*)0x20000010 = 0;
      syscall(__NR_ioctl, /*fd=*/r[0], /*cmd=*/0xc018aa3f, /*arg=*/0x20000000ul);

      syscall(__NR_mlockall, /*flags=MCL_FUTURE|MCL_CURRENT*/3ul);

      // ioctl(3, UFFDIO_REGISTER, {range={start=0x20ffb000, len=0x3000},
mode=UFFDIO_REGISTER_MODE_WP,
ioctls=1<<_UFFDIO_WAKE|1<<_UFFDIO_COPY|1<<_UFFDIO_ZEROPAGE|1<<_UFFDIO_WRITEPROTECT|0x120}) = 0
      *(uint64_t*)0x20000180 = 0x20ffb000;
      *(uint64_t*)0x20000188 = 0x3000;
      *(uint64_t*)0x20000190 = 2;
      *(uint64_t*)0x20000198 = 0;
      syscall(__NR_ioctl, /*fd=*/r[0], /*cmd=*/0xc020aa00, /*arg=*/0x20000180ul);

      // ioctl(3, UFFDIO_WRITEPROTECT, 0x20000080) = -1 ENOENT (No such file or
directory)
      *(uint64_t*)0x20000080 = 0x20ffb000;
      *(uint64_t*)0x20000088 = 0x4000;
      *(uint64_t*)0x20000090 = 1;
      syscall(__NR_ioctl, /*fd=*/r[0], /*cmd=*/0xc018aa06, /*arg=*/0x20000080ul);

      syscall(__NR_mremap, /*addr=*/0x20ffb000ul, /*len=*/0x1000ul,
/*newlen=*/0x1000ul,
/*flags=MREMAP_DONTUNMAP|MREMAP_FIXED|MREMAP_MAYMOVE*/7ul,
/*newaddr=*/0x20ffd000ul);
      syscall(__NR_mprotect, /*addr=*/0x20ffd000ul, /*len=*/0x1000ul,
/*prot=PROT_WRITE|PROT_READ*/3ul);

      return 0;
}
"""

I'd appreciate any thoughts you may have!


Interesting. Either the vma flag shouldn't get dropped or we should un-mark the
PTEs.


Yes, agreed. But which? I guess Peter is the expert here?


Staring at mremap_userfaultfd_prep(), it looks like we unconditionally
want to drop the vma flag: userfaultfd_reset_ctx().

While this might have made sense in the MISSING mode (although I don't
really think so ...) it might not make sense in the WP mode, espcially
when used for softdirty-like monitoring. But maybe simply clearing the
WP flag everywhere would keep that use case also happy ("everything was
dirtied").

Hopefully Peter knows what's right or wrong :)


Staring at

commit 3cfd22be0ad663248fadfc8f6ffa3e255c394552
Author: Peter Xu <peterx@xxxxxxxxxx>
Date:   Fri Dec 28 00:38:47 2018 -0800

    userfaultfd: clear flag if remap event not enabled
When the process being tracked does mremap() without 
    UFFD_FEATURE_EVENT_REMAP on the corresponding tracking uffd file handle,
    we should not generate the remap event, and at the same time we should
    clear all the uffd flags on the new VMA.  Without this patch, we can still
    have the VM_UFFD_MISSING|VM_UFFD_WP flags on the new VMA even the fault
    handling process does not even know the existance of the VMA.

I guess the reason about "existence of the VMA" was that if virtual addresses
change, how should an external monitor know how to handle the new addresses.

But if it's not an external monitor (i.e., the app that uses uffd+mremap), that
might not be the case -- the monitor knows about VMAs.

I'm not sure if clearing uffd flags was the right decision, but I'm no expert.

--
Cheers,

David / dhildenb
[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux