linux.git/fs/proc/task_mmu.c, branch v4.12

mm: larger stack guard gap, between vmas

2017-06-19T13:50:20+00:00

Stack guard page is a useful feature to reduce a risk of stack smashing
into a different mapping. We have been using a single page gap which
is sufficient to prevent having stack adjacent to a different mapping.
But this seems to be insufficient in the light of the stack usage in
userspace. E.g. glibc uses as large as 64kB alloca() in many commonly
used functions. Others use constructs liks gid_t buffer[NGROUPS_MAX]
which is 256kB or stack strings with MAX_ARG_STRLEN.

This will become especially dangerous for suid binaries and the default
no limit for the stack size limit because those applications can be
tricked to consume a large portion of the stack and a single glibc call
could jump over the guard page. These attacks are not theoretical,
unfortunatelly.

Make those attacks less probable by increasing the stack guard gap
to 1MB (on systems with 4k pages; but make it depend on the page size
because systems with larger base pages might cap stack allocations in
the PAGE_SIZE units) which should cover larger alloca() and VLA stack
allocations. It is obviously not a full fix because the problem is
somehow inherent, but it should reduce attack space a lot.

One could argue that the gap size should be configurable from userspace,
but that can be done later when somebody finds that the new 1MB is wrong
for some special case applications.  For now, add a kernel command line
option (stack_guard_gap) to specify the stack gap size (in page units).

Implementation wise, first delete all the old code for stack guard page:
because although we could get away with accounting one extra page in a
stack vma, accounting a larger gap can break userspace - case in point,
a program run with "ulimit -S -v 20000" failed when the 1MB gap was
counted for RLIMIT_AS; similar problems could come with RLIMIT_MLOCK
and strict non-overcommit mode.

Instead of keeping gap inside the stack vma, maintain the stack guard
gap as a gap between vmas: using vm_start_gap() in place of vm_start
(or vm_end_gap() in place of vm_end if VM_GROWSUP) in just those few
places which need to respect the gap - mainly arch_get_unmapped_area(),
and and the vma tree's subtree_gap support for that.

Original-patch-by: Oleg Nesterov 
Original-patch-by: Michal Hocko 
Signed-off-by: Hugh Dickins 
Acked-by: Michal Hocko 
Tested-by: Helge Deller  # parisc
Signed-off-by: Linus Torvalds

proc: show MADV_FREE pages info in smaps

2017-05-03T22:52:08+00:00

Show MADV_FREE pages info of each vma in smaps.  The interface is for
diganose or monitoring purpose, userspace could use it to understand
what happens in the application.  Since userspace could dirty MADV_FREE
pages without notice from kernel, this interface is the only place we
can get accurate accounting info about MADV_FREE pages.

[mhocko@kernel.org: update Documentation/filesystems/proc.txt]
Link: http://lkml.kernel.org/r/89efde633559de1ec07444f2ef0f4963a97a2ce8.1487965799.git.shli@fb.com
Signed-off-by: Shaohua Li 
Acked-by: Johannes Weiner 
Acked-by: Minchan Kim 
Acked-by: Michal Hocko 
Acked-by: Hillf Danton 
Cc: Hugh Dickins 
Cc: Rik van Riel 
Cc: Mel Gorman 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

thp: fix MADV_DONTNEED vs clear soft dirty race

2017-04-14T01:24:21+00:00

Yet another instance of the same race.

Fix is identical to change_huge_pmd().

See "thp: fix MADV_DONTNEED vs.  numa balancing race" for more details.

Link: http://lkml.kernel.org/r/20170302151034.27829-5-kirill.shutemov@linux.intel.com
Signed-off-by: Kirill A. Shutemov 
Cc: Andrea Arcangeli 
Cc: Hillf Danton 
Cc: 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

sched/headers: Prepare for new header dependencies before moving code to

2017-03-02T07:42:28+00:00

We are going to split  out of , which
will have to be picked up from other headers and a couple of .c files.

Create a trivial placeholder  file that just
maps to  to make this patch obviously correct and
bisectable.

The APIs that are going to be moved first are:

   mm_alloc()
   __mmdrop()
   mmdrop()
   mmdrop_async_fn()
   mmdrop_async()
   mmget_not_zero()
   mmput()
   mmput_async()
   get_task_mm()
   mm_access()
   mm_release()

Include the new header in the files that are going to need it.

Acked-by: Linus Torvalds 
Cc: Mike Galbraith 
Cc: Peter Zijlstra 
Cc: Thomas Gleixner 
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar

mm: use mmget_not_zero() helper

2017-02-28T02:43:48+00:00

We already have the helper, we can convert the rest of the kernel
mechanically using:

  git grep -l 'atomic_inc_not_zero.*mm_users' | xargs sed -i 's/atomic_inc_not_zero(&\(.*\)->mm_users)/mmget_not_zero\(\1\)/'

This is needed for a later patch that hooks into the helper, but might
be a worthwhile cleanup on its own.

Link: http://lkml.kernel.org/r/20161218123229.22952-3-vegard.nossum@oracle.com
Signed-off-by: Vegard Nossum 
Acked-by: Michal Hocko 
Acked-by: Peter Zijlstra (Intel) 
Acked-by: David Rientjes 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

Replace with globally

2016-12-24T19:46:01+00:00

This was entirely automated, using the script by Al:

  PATT='^[[:blank:]]*#[[:blank:]]*include[[:blank:]]*'
  sed -i -e "s!$PATT!#include !" \
        $(git grep -l "$PATT"|grep -v ^include/linux/uaccess.h)

to do the replacement at the end of the merge window.

Requested-by: Al Viro 
Signed-off-by: Linus Torvalds

mm: add cond_resched() in gather_pte_stats()

2016-12-13T02:55:09+00:00

The other pagetable walks in task_mmu.c have a cond_resched() after
walking their ptes: add a cond_resched() in gather_pte_stats() too, for
reading /proc//numa_maps.  Only pagemap_pmd_range() has a
cond_resched() in its (unusually expensive) pmd_trans_huge case: more
should probably be added, but leave them unchanged for now.

Link: http://lkml.kernel.org/r/alpine.LSU.2.11.1612052157400.13021@eggly.anvils
Signed-off-by: Hugh Dickins 
Acked-by: Michal Hocko 
Cc: David Rientjes 
Cc: Gerald Schaefer 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fs/proc: Stop trying to report thread stacks

2016-10-20T07:21:41+00:00

This reverts more of:

  b76437579d13 ("procfs: mark thread stack correctly in proc//maps")

... which was partially reverted by:

  65376df58217 ("proc: revert /proc//maps [stack:TID] annotation")

Originally, /proc/PID/task/TID/maps was the same as /proc/TID/maps.

In current kernels, /proc/PID/maps (or /proc/TID/maps even for
threads) shows "[stack]" for VMAs in the mm's stack address range.

In contrast, /proc/PID/task/TID/maps uses KSTK_ESP to guess the
target thread's stack's VMA.  This is racy, probably returns garbage
and, on arches with CONFIG_TASK_INFO_IN_THREAD=y, is also crash-prone:
KSTK_ESP is not safe to use on tasks that aren't known to be running
ordinary process-context kernel code.

This patch removes the difference and just shows "[stack]" for VMAs
in the mm's stack range.  This is IMO much more sensible -- the
actual "stack" address really is treated specially by the VM code,
and the current thread stack isn't even well-defined for programs
that frequently switch stacks on their own.

Reported-by: Jann Horn 
Signed-off-by: Andy Lutomirski 
Acked-by: Thomas Gleixner 
Cc: Al Viro 
Cc: Andrew Morton 
Cc: Borislav Petkov 
Cc: Brian Gerst 
Cc: Johannes Weiner 
Cc: Kees Cook 
Cc: Linus Torvalds 
Cc: Linux API 
Cc: Peter Zijlstra 
Cc: Tycho Andersen 
Link: http://lkml.kernel.org/r/3e678474ec14e0a0ec34c611016753eea2e1b8ba.1475257877.git.luto@kernel.org
Signed-off-by: Ingo Molnar

mm, proc: fix region lost in /proc/self/smaps

2016-10-08T01:46:30+00:00

Recently, Redhat reported that nvml test suite failed on QEMU/KVM,
more detailed info please refer to:

   https://bugzilla.redhat.com/show_bug.cgi?id=1365721

Actually, this bug is not only for NVDIMM/DAX but also for any other
file systems.  This simple test case abstracted from nvml can easily
reproduce this bug in common environment:

-------------------------- testcase.c -----------------------------

int
is_pmem_proc(const void *addr, size_t len)
{
        const char *caddr = addr;

        FILE *fp;
        if ((fp = fopen("/proc/self/smaps", "r")) == NULL) {
                printf("!/proc/self/smaps");
                return 0;
        }

        int retval = 0;         /* assume false until proven otherwise */
        char line[PROCMAXLEN];  /* for fgets() */
        char *lo = NULL;        /* beginning of current range in smaps file */
        char *hi = NULL;        /* end of current range in smaps file */
        int needmm = 0;         /* looking for mm flag for current range */
        while (fgets(line, PROCMAXLEN, fp) != NULL) {
                static const char vmflags[] = "VmFlags:";
                static const char mm[] = " wr";

                /* check for range line */
                if (sscanf(line, "%p-%p", &lo, &hi) == 2) {
                        if (needmm) {
                                /* last range matched, but no mm flag found */
                                printf("never found mm flag.\n");
                                break;
                        } else if (caddr < lo) {
                                /* never found the range for caddr */
                                printf("#######no match for addr %p.\n", caddr);
                                break;
                        } else if (caddr < hi) {
                                /* start address is in this range */
                                size_t rangelen = (size_t)(hi - caddr);

                                /* remember that matching has started */
                                needmm = 1;

                                /* calculate remaining range to search for */
                                if (len > rangelen) {
                                        len -= rangelen;
                                        caddr += rangelen;
                                        printf("matched %zu bytes in range "
                                                "%p-%p, %zu left over.\n",
                                                        rangelen, lo, hi, len);
                                } else {
                                        len = 0;
                                        printf("matched all bytes in range "
                                                        "%p-%p.\n", lo, hi);
                                }
                        }
                } else if (needmm && strncmp(line, vmflags,
                                        sizeof(vmflags) - 1) == 0) {
                        if (strstr(&line[sizeof(vmflags) - 1], mm) != NULL) {
                                printf("mm flag found.\n");
                                if (len == 0) {
                                        /* entire range matched */
                                        retval = 1;
                                        break;
                                }
                                needmm = 0;     /* saw what was needed */
                        } else {
                                /* mm flag not set for some or all of range */
                                printf("range has no mm flag.\n");
                                break;
                        }
                }
        }

        fclose(fp);

        printf("returning %d.\n", retval);
        return retval;
}

void *Addr;
size_t Size;

/*
 * worker -- the work each thread performs
 */
static void *
worker(void *arg)
{
        int *ret = (int *)arg;
        *ret =  is_pmem_proc(Addr, Size);
        return NULL;
}

int main(int argc, char *argv[])
{
        if (argc <  2 || argc > 3) {
                printf("usage: %s file [env].\n", argv[0]);
                return -1;
        }

        int fd = open(argv[1], O_RDWR);

        struct stat stbuf;
        fstat(fd, &stbuf);

        Size = stbuf.st_size;
        Addr = mmap(0, stbuf.st_size, PROT_READ|PROT_WRITE, MAP_PRIVATE, fd, 0);

        close(fd);

        pthread_t threads[NTHREAD];
        int ret[NTHREAD];

        /* kick off NTHREAD threads */
        for (int i = 0; i < NTHREAD; i++)
                pthread_create(&threads[i], NULL, worker, &ret[i]);

        /* wait for all the threads to complete */
        for (int i = 0; i < NTHREAD; i++)
                pthread_join(threads[i], NULL);

        /* verify that all the threads return the same value */
        for (int i = 1; i < NTHREAD; i++) {
                if (ret[0] != ret[i]) {
                        printf("Error i %d ret[0] = %d ret[i] = %d.\n", i,
                                ret[0], ret[i]);
                }
        }

        printf("%d", ret[0]);
        return 0;
}

It failed as some threads can not find the memory region in
"/proc/self/smaps" which is allocated in the main process

It is caused by proc fs which uses 'file->version' to indicate the VMA that
is the last one has already been handled by read() system call. When the
next read() issues, it uses the 'version' to find the VMA, then the next
VMA is what we want to handle, the related code is as follows:

        if (last_addr) {
                vma = find_vma(mm, last_addr);
                if (vma && (vma = m_next_vma(priv, vma)))
                        return vma;
        }

However, VMA will be lost if the last VMA is gone, e.g:

The process VMA list is A->B->C->D

CPU 0                                  CPU 1
read() system call
   handle VMA B
   version = B
return to userspace

                                   unmap VMA B

issue read() again to continue to get
the region info
   find_vma(version) will get VMA C
   m_next_vma(C) will get VMA D
   handle D
   !!! VMA C is lost !!!

In order to fix this bug, we make 'file->version' indicate the end address
of the current VMA.  m_start will then look up a vma which with vma_start
< last_vm_end and moves on to the next vma if we found the same or an
overlapping vma.  This will guarantee that we will not miss an exclusive
vma but we can still miss one if the previous vma was shrunk.  This is
acceptable because guaranteeing "never miss a vma" is simply not feasible.
User has to cope with some inconsistencies if the file is not read in one
go.

[mhocko@suse.com: changelog fixes]
Link: http://lkml.kernel.org/r/1475296958-27652-1-git-send-email-robert.hu@intel.com
Acked-by: Dave Hansen 
Signed-off-by: Xiao Guangrong 
Signed-off-by: Robert Hu 
Acked-by: Michal Hocko 
Acked-by: Oleg Nesterov 
Cc: Paolo Bonzini 
Cc: Dan Williams 
Cc: Gleb Natapov 
Cc: Marcelo Tosatti 
Cc: Stefan Hajnoczi 
Cc: Ross Zwisler 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds

fs/proc/task_mmu.c: make the task_mmu walk_page_range() limit in clear_refs_write() obvious

2016-10-08T01:46:28+00:00

Trying to walk all of virtual memory requires architecture specific
knowledge.  On x86_64, addresses must be sign extended from bit 48,
whereas on arm64 the top VA_BITS of address space have their own set of
page tables.

clear_refs_write() calls walk_page_range() on the range 0 to ~0UL, it
provides a test_walk() callback that only expects to be walking over
VMAs.  Currently walk_pmd_range() will skip memory regions that don't
have a VMA, reporting them as a hole.

As this call only expects to walk user address space, make it walk 0 to
'highest_vm_end'.

Link: http://lkml.kernel.org/r/1472655792-22439-1-git-send-email-james.morse@arm.com
Signed-off-by: James Morse 
Acked-by: Naoya Horiguchi 
Signed-off-by: Andrew Morton 
Signed-off-by: Linus Torvalds