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
sortextable.c
/*
* sortextable.c: Sort the kernel's exception table
*
* Copyright 2011 - 2012 Cavium, Inc.
*
* Based on code taken from recortmcount.c which is:
*
* Copyright 2009 John F. Reiser <jreiser@BitWagon.com>. All rights reserved.
* Licensed under the GNU General Public License, version 2 (GPLv2).
*
* Restructured to fit Linux format, as well as other updates:
* Copyright 2010 Steven Rostedt <srostedt@redhat.com>, Red Hat Inc.
*/
/*
* Strategy: alter the vmlinux file in-place.
*/
#include <sys/types.h>
#include <sys/mman.h>
#include <sys/stat.h>
#include <getopt.h>
#include <elf.h>
#include <fcntl.h>
#include <setjmp.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <unistd.h>
#include <tools/be_byteshift.h>
#include <tools/le_byteshift.h>
#ifndef EM_ARCOMPACT
#define EM_ARCOMPACT 93
#endif
#ifndef EM_XTENSA
#define EM_XTENSA 94
#endif
#ifndef EM_AARCH64
#define EM_AARCH64 183
#endif
#ifndef EM_MICROBLAZE
#define EM_MICROBLAZE 189
#endif
#ifndef EM_ARCV2
#define EM_ARCV2 195
#endif
static int fd_map; /* File descriptor for file being modified. */
static int mmap_failed; /* Boolean flag. */
static void *ehdr_curr; /* current ElfXX_Ehdr * for resource cleanup */
static struct stat sb; /* Remember .st_size, etc. */
static jmp_buf jmpenv; /* setjmp/longjmp per-file error escape */
/* setjmp() return values */
enum {
SJ_SETJMP = 0, /* hardwired first return */
SJ_FAIL,
SJ_SUCCEED
};
/* Per-file resource cleanup when multiple files. */
static void
cleanup(void)
{
if (!mmap_failed)
munmap(ehdr_curr, sb.st_size);
close(fd_map);
}
static void __attribute__((noreturn))
fail_file(void)
{
cleanup();
longjmp(jmpenv, SJ_FAIL);
}
/*
* Get the whole file as a programming convenience in order to avoid
* malloc+lseek+read+free of many pieces. If successful, then mmap
* avoids copying unused pieces; else just read the whole file.
* Open for both read and write.
*/
static void *mmap_file(char const *fname)
{
void *addr;
fd_map = open(fname, O_RDWR);
if (fd_map < 0 || fstat(fd_map, &sb) < 0) {
perror(fname);
fail_file();
}
if (!S_ISREG(sb.st_mode)) {
fprintf(stderr, "not a regular file: %s\n", fname);
fail_file();
}
addr = mmap(0, sb.st_size, PROT_READ|PROT_WRITE, MAP_SHARED,
fd_map, 0);
if (addr == MAP_FAILED) {
mmap_failed = 1;
fprintf(stderr, "Could not mmap file: %s\n", fname);
fail_file();
}
return addr;
}
static uint64_t r8be(const uint64_t *x)
{
return get_unaligned_be64(x);
}
static uint32_t rbe(const uint32_t *x)
{
return get_unaligned_be32(x);
}
static uint16_t r2be(const uint16_t *x)
{
return get_unaligned_be16(x);
}
static uint64_t r8le(const uint64_t *x)
{
return get_unaligned_le64(x);
}
static uint32_t rle(const uint32_t *x)
{
return get_unaligned_le32(x);
}
static uint16_t r2le(const uint16_t *x)
{
return get_unaligned_le16(x);
}
static void w8be(uint64_t val, uint64_t *x)
{
put_unaligned_be64(val, x);
}
static void wbe(uint32_t val, uint32_t *x)
{
put_unaligned_be32(val, x);
}
static void w2be(uint16_t val, uint16_t *x)
{
put_unaligned_be16(val, x);
}
static void w8le(uint64_t val, uint64_t *x)
{
put_unaligned_le64(val, x);
}
static void wle(uint32_t val, uint32_t *x)
{
put_unaligned_le32(val, x);
}
static void w2le(uint16_t val, uint16_t *x)
{
put_unaligned_le16(val, x);
}
static uint64_t (*r8)(const uint64_t *);
static uint32_t (*r)(const uint32_t *);
static uint16_t (*r2)(const uint16_t *);
static void (*w8)(uint64_t, uint64_t *);
static void (*w)(uint32_t, uint32_t *);
static void (*w2)(uint16_t, uint16_t *);
typedef void (*table_sort_t)(char *, int);
/*
* Move reserved section indices SHN_LORESERVE..SHN_HIRESERVE out of
* the way to -256..-1, to avoid conflicting with real section
* indices.
*/
#define SPECIAL(i) ((i) - (SHN_HIRESERVE + 1))
static inline int is_shndx_special(unsigned int i)
{
return i != SHN_XINDEX && i >= SHN_LORESERVE && i <= SHN_HIRESERVE;
}
/* Accessor for sym->st_shndx, hides ugliness of "64k sections" */
static inline unsigned int get_secindex(unsigned int shndx,
unsigned int sym_offs,
const Elf32_Word *symtab_shndx_start)
{
if (is_shndx_special(shndx))
return SPECIAL(shndx);
if (shndx != SHN_XINDEX)
return shndx;
return r(&symtab_shndx_start[sym_offs]);
}
/* 32 bit and 64 bit are very similar */
#include "sortextable.h"
#define SORTEXTABLE_64
#include "sortextable.h"
static int compare_relative_table(const void *a, const void *b)
{
int32_t av = (int32_t)r(a);
int32_t bv = (int32_t)r(b);
if (av < bv)
return -1;
if (av > bv)
return 1;
return 0;
}
static void x86_sort_relative_table(char *extab_image, int image_size)
{
int i;
i = 0;
while (i < image_size) {
uint32_t *loc = (uint32_t *)(extab_image + i);
w(r(loc) + i, loc);
w(r(loc + 1) + i + 4, loc + 1);
w(r(loc + 2) + i + 8, loc + 2);
i += sizeof(uint32_t) * 3;
}
qsort(extab_image, image_size / 12, 12, compare_relative_table);
i = 0;
while (i < image_size) {
uint32_t *loc = (uint32_t *)(extab_image + i);
w(r(loc) - i, loc);
w(r(loc + 1) - (i + 4), loc + 1);
w(r(loc + 2) - (i + 8), loc + 2);
i += sizeof(uint32_t) * 3;
}
}
static void sort_relative_table(char *extab_image, int image_size)
{
int i;
/*
* Do the same thing the runtime sort does, first normalize to
* being relative to the start of the section.
*/
i = 0;
while (i < image_size) {
uint32_t *loc = (uint32_t *)(extab_image + i);
w(r(loc) + i, loc);
i += 4;
}
qsort(extab_image, image_size / 8, 8, compare_relative_table);
/* Now denormalize. */
i = 0;
while (i < image_size) {
uint32_t *loc = (uint32_t *)(extab_image + i);
w(r(loc) - i, loc);
i += 4;
}
}
static void
do_file(char const *const fname)
{
table_sort_t custom_sort;
Elf32_Ehdr *ehdr = mmap_file(fname);
ehdr_curr = ehdr;
switch (ehdr->e_ident[EI_DATA]) {
default:
fprintf(stderr, "unrecognized ELF data encoding %d: %s\n",
ehdr->e_ident[EI_DATA], fname);
fail_file();
break;
case ELFDATA2LSB:
r = rle;
r2 = r2le;
r8 = r8le;
w = wle;
w2 = w2le;
w8 = w8le;
break;
case ELFDATA2MSB:
r = rbe;
r2 = r2be;
r8 = r8be;
w = wbe;
w2 = w2be;
w8 = w8be;
break;
} /* end switch */
if (memcmp(ELFMAG, ehdr->e_ident, SELFMAG) != 0
|| (r2(&ehdr->e_type) != ET_EXEC && r2(&ehdr->e_type) != ET_DYN)
|| ehdr->e_ident[EI_VERSION] != EV_CURRENT) {
fprintf(stderr, "unrecognized ET_EXEC/ET_DYN file %s\n", fname);
fail_file();
}
custom_sort = NULL;
switch (r2(&ehdr->e_machine)) {
default:
fprintf(stderr, "unrecognized e_machine %d %s\n",
r2(&ehdr->e_machine), fname);
fail_file();
break;
case EM_386:
case EM_X86_64:
custom_sort = x86_sort_relative_table;
break;
case EM_S390:
case EM_AARCH64:
case EM_PARISC:
case EM_PPC:
case EM_PPC64:
custom_sort = sort_relative_table;
break;
case EM_ARCOMPACT:
case EM_ARCV2:
case EM_ARM:
case EM_MICROBLAZE:
case EM_MIPS:
case EM_XTENSA:
break;
} /* end switch */
switch (ehdr->e_ident[EI_CLASS]) {
default:
fprintf(stderr, "unrecognized ELF class %d %s\n",
ehdr->e_ident[EI_CLASS], fname);
fail_file();
break;
case ELFCLASS32:
if (r2(&ehdr->e_ehsize) != sizeof(Elf32_Ehdr)
|| r2(&ehdr->e_shentsize) != sizeof(Elf32_Shdr)) {
fprintf(stderr,
"unrecognized ET_EXEC/ET_DYN file: %s\n", fname);
fail_file();
}
do32(ehdr, fname, custom_sort);
break;
case ELFCLASS64: {
Elf64_Ehdr *const ghdr = (Elf64_Ehdr *)ehdr;
if (r2(&ghdr->e_ehsize) != sizeof(Elf64_Ehdr)
|| r2(&ghdr->e_shentsize) != sizeof(Elf64_Shdr)) {
fprintf(stderr,
"unrecognized ET_EXEC/ET_DYN file: %s\n", fname);
fail_file();
}
do64(ghdr, fname, custom_sort);
break;
}
} /* end switch */
cleanup();
}
int
main(int argc, char *argv[])
{
int n_error = 0; /* gcc-4.3.0 false positive complaint */
int i;
if (argc < 2) {
fprintf(stderr, "usage: sortextable vmlinux...\n");
return 0;
}
/* Process each file in turn, allowing deep failure. */
for (i = 1; i < argc; i++) {
char *file = argv[i];
int const sjval = setjmp(jmpenv);
switch (sjval) {
default:
fprintf(stderr, "internal error: %s\n", file);
exit(1);
break;
case SJ_SETJMP: /* normal sequence */
/* Avoid problems if early cleanup() */
fd_map = -1;
ehdr_curr = NULL;
mmap_failed = 1;
do_file(file);
break;
case SJ_FAIL: /* error in do_file or below */
++n_error;
break;
case SJ_SUCCEED: /* premature success */
/* do nothing */
break;
} /* end switch */
}
return !!n_error;
}
![swh spinner](/static/img/swh-spinner.gif)
Computing file changes ...