Revision e2093926a098a8ccf0f1d10f6df8dad452cb28d3 authored by Ross Zwisler on 02 June 2017, 21:46:37 UTC, committed by Linus Torvalds on 02 June 2017, 22:07:37 UTC
We currently have two related PMD vs PTE races in the DAX code.  These
can both be easily triggered by having two threads reading and writing
simultaneously to the same private mapping, with the key being that
private mapping reads can be handled with PMDs but private mapping
writes are always handled with PTEs so that we can COW.

Here is the first race:

  CPU 0					CPU 1

  (private mapping write)
  __handle_mm_fault()
    create_huge_pmd() - FALLBACK
    handle_pte_fault()
      passes check for pmd_devmap()

					(private mapping read)
					__handle_mm_fault()
					  create_huge_pmd()
					    dax_iomap_pmd_fault() inserts PMD

      dax_iomap_pte_fault() does a PTE fault, but we already have a DAX PMD
      			  installed in our page tables at this spot.

Here's the second race:

  CPU 0					CPU 1

  (private mapping read)
  __handle_mm_fault()
    passes check for pmd_none()
    create_huge_pmd()
      dax_iomap_pmd_fault() inserts PMD

  (private mapping write)
  __handle_mm_fault()
    create_huge_pmd() - FALLBACK
					(private mapping read)
					__handle_mm_fault()
					  passes check for pmd_none()
					  create_huge_pmd()

    handle_pte_fault()
      dax_iomap_pte_fault() inserts PTE
					    dax_iomap_pmd_fault() inserts PMD,
					       but we already have a PTE at
					       this spot.

The core of the issue is that while there is isolation between faults to
the same range in the DAX fault handlers via our DAX entry locking,
there is no isolation between faults in the code in mm/memory.c.  This
means for instance that this code in __handle_mm_fault() can run:

	if (pmd_none(*vmf.pmd) && transparent_hugepage_enabled(vma)) {
		ret = create_huge_pmd(&vmf);

But by the time we actually get to run the fault handler called by
create_huge_pmd(), the PMD is no longer pmd_none() because a racing PTE
fault has installed a normal PMD here as a parent.  This is the cause of
the 2nd race.  The first race is similar - there is the following check
in handle_pte_fault():

	} else {
		/* See comment in pte_alloc_one_map() */
		if (pmd_devmap(*vmf->pmd) || pmd_trans_unstable(vmf->pmd))
			return 0;

So if a pmd_devmap() PMD (a DAX PMD) has been installed at vmf->pmd, we
will bail and retry the fault.  This is correct, but there is nothing
preventing the PMD from being installed after this check but before we
actually get to the DAX PTE fault handlers.

In my testing these races result in the following types of errors:

  BUG: Bad rss-counter state mm:ffff8800a817d280 idx:1 val:1
  BUG: non-zero nr_ptes on freeing mm: 15

Fix this issue by having the DAX fault handlers verify that it is safe
to continue their fault after they have taken an entry lock to block
other racing faults.

[ross.zwisler@linux.intel.com: improve fix for colliding PMD & PTE entries]
  Link: http://lkml.kernel.org/r/20170526195932.32178-1-ross.zwisler@linux.intel.com
Link: http://lkml.kernel.org/r/20170522215749.23516-2-ross.zwisler@linux.intel.com
Signed-off-by: Ross Zwisler <ross.zwisler@linux.intel.com>
Reported-by: Pawel Lebioda <pawel.lebioda@intel.com>
Reviewed-by: Jan Kara <jack@suse.cz>
Cc: "Darrick J. Wong" <darrick.wong@oracle.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Matthew Wilcox <mawilcox@microsoft.com>
Cc: "Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Pawel Lebioda <pawel.lebioda@intel.com>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Xiong Zhou <xzhou@redhat.com>
Cc: Eryu Guan <eguan@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent d0f0931
Raw File
headers_check.pl
#!/usr/bin/perl -w
#
# headers_check.pl execute a number of trivial consistency checks
#
# Usage: headers_check.pl dir arch [files...]
# dir:   dir to look for included files
# arch:  architecture
# files: list of files to check
#
# The script reads the supplied files line by line and:
#
# 1) for each include statement it checks if the
#    included file actually exists.
#    Only include files located in asm* and linux* are checked.
#    The rest are assumed to be system include files.
#
# 2) It is checked that prototypes does not use "extern"
#
# 3) Check for leaked CONFIG_ symbols

use strict;
use File::Basename;

my ($dir, $arch, @files) = @ARGV;

my $ret = 0;
my $line;
my $lineno = 0;
my $filename;

foreach my $file (@files) {
	$filename = $file;

	open(my $fh, '<', $filename)
		or die "$filename: $!\n";
	$lineno = 0;
	while ($line = <$fh>) {
		$lineno++;
		&check_include();
		&check_asm_types();
		&check_sizetypes();
		&check_declarations();
		# Dropped for now. Too much noise &check_config();
	}
	close $fh;
}
exit $ret;

sub check_include
{
	if ($line =~ m/^\s*#\s*include\s+<((asm|linux).*)>/) {
		my $inc = $1;
		my $found;
		$found = stat($dir . "/" . $inc);
		if (!$found) {
			$inc =~ s#asm/#asm-$arch/#;
			$found = stat($dir . "/" . $inc);
		}
		if (!$found) {
			printf STDERR "$filename:$lineno: included file '$inc' is not exported\n";
			$ret = 1;
		}
	}
}

sub check_declarations
{
	# soundcard.h is what it is
	if ($line =~ m/^void seqbuf_dump\(void\);/) {
		return;
	}
	# drm headers are being C++ friendly
	if ($line =~ m/^extern "C"/) {
		return;
	}
	if ($line =~ m/^(\s*extern|unsigned|char|short|int|long|void)\b/) {
		printf STDERR "$filename:$lineno: " .
			      "userspace cannot reference function or " .
			      "variable defined in the kernel\n";
	}
}

sub check_config
{
	if ($line =~ m/[^a-zA-Z0-9_]+CONFIG_([a-zA-Z0-9_]+)[^a-zA-Z0-9_]/) {
		printf STDERR "$filename:$lineno: leaks CONFIG_$1 to userspace where it is not valid\n";
	}
}

my $linux_asm_types;
sub check_asm_types
{
	if ($filename =~ /types.h|int-l64.h|int-ll64.h/o) {
		return;
	}
	if ($lineno == 1) {
		$linux_asm_types = 0;
	} elsif ($linux_asm_types >= 1) {
		return;
	}
	if ($line =~ m/^\s*#\s*include\s+<asm\/types.h>/) {
		$linux_asm_types = 1;
		printf STDERR "$filename:$lineno: " .
		"include of <linux/types.h> is preferred over <asm/types.h>\n"
		# Warn until headers are all fixed
		#$ret = 1;
	}
}

my $linux_types;
my %import_stack = ();
sub check_include_typesh
{
	my $path = $_[0];
	my $import_path;

	my $fh;
	my @file_paths = ($path, $dir . "/" .  $path, dirname($filename) . "/" . $path);
	for my $possible ( @file_paths ) {
	    if (not $import_stack{$possible} and open($fh, '<', $possible)) {
		$import_path = $possible;
		$import_stack{$import_path} = 1;
		last;
	    }
	}
	if (eof $fh) {
	    return;
	}

	my $line;
	while ($line = <$fh>) {
		if ($line =~ m/^\s*#\s*include\s+<linux\/types.h>/) {
			$linux_types = 1;
			last;
		}
		if (my $included = ($line =~ /^\s*#\s*include\s+[<"](\S+)[>"]/)[0]) {
			check_include_typesh($included);
		}
	}
	close $fh;
	delete $import_stack{$import_path};
}

sub check_sizetypes
{
	if ($filename =~ /types.h|int-l64.h|int-ll64.h/o) {
		return;
	}
	if ($lineno == 1) {
		$linux_types = 0;
	} elsif ($linux_types >= 1) {
		return;
	}
	if ($line =~ m/^\s*#\s*include\s+<linux\/types.h>/) {
		$linux_types = 1;
		return;
	}
	if (my $included = ($line =~ /^\s*#\s*include\s+[<"](\S+)[>"]/)[0]) {
		check_include_typesh($included);
	}
	if ($line =~ m/__[us](8|16|32|64)\b/) {
		printf STDERR "$filename:$lineno: " .
		              "found __[us]{8,16,32,64} type " .
		              "without #include <linux/types.h>\n";
		$linux_types = 2;
		# Warn until headers are all fixed
		#$ret = 1;
	}
}
back to top