Revision 6bc9b56433b76e40d11099338d27fbc5cd2935ca authored by Naoya Horiguchi on 24 August 2018, 00:00:38 UTC, committed by Linus Torvalds on 24 August 2018, 01:48:43 UTC
Patch series "mm: soft-offline: fix race against page allocation".

Xishi recently reported the issue about race on reusing the target pages
of soft offlining.  Discussion and analysis showed that we need make
sure that setting PG_hwpoison should be done in the right place under
zone->lock for soft offline.  1/2 handles free hugepage's case, and 2/2
hanldes free buddy page's case.

This patch (of 2):

There's a race condition between soft offline and hugetlb_fault which
causes unexpected process killing and/or hugetlb allocation failure.

The process killing is caused by the following flow:

  CPU 0               CPU 1              CPU 2

  soft offline
    get_any_page
    // find the hugetlb is free
                      mmap a hugetlb file
                      page fault
                        ...
                          hugetlb_fault
                            hugetlb_no_page
                              alloc_huge_page
                              // succeed
      soft_offline_free_page
      // set hwpoison flag
                                         mmap the hugetlb file
                                         page fault
                                           ...
                                             hugetlb_fault
                                               hugetlb_no_page
                                                 find_lock_page
                                                   return VM_FAULT_HWPOISON
                                           mm_fault_error
                                             do_sigbus
                                             // kill the process

The hugetlb allocation failure comes from the following flow:

  CPU 0                          CPU 1

                                 mmap a hugetlb file
                                 // reserve all free page but don't fault-in
  soft offline
    get_any_page
    // find the hugetlb is free
      soft_offline_free_page
      // set hwpoison flag
        dissolve_free_huge_page
        // fail because all free hugepages are reserved
                                 page fault
                                   ...
                                     hugetlb_fault
                                       hugetlb_no_page
                                         alloc_huge_page
                                           ...
                                             dequeue_huge_page_node_exact
                                             // ignore hwpoisoned hugepage
                                             // and finally fail due to no-mem

The root cause of this is that current soft-offline code is written based
on an assumption that PageHWPoison flag should be set at first to avoid
accessing the corrupted data.  This makes sense for memory_failure() or
hard offline, but does not for soft offline because soft offline is about
corrected (not uncorrected) error and is safe from data lost.  This patch
changes soft offline semantics where it sets PageHWPoison flag only after
containment of the error page completes successfully.

Link: http://lkml.kernel.org/r/1531452366-11661-2-git-send-email-n-horiguchi@ah.jp.nec.com
Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Reported-by: Xishi Qiu <xishi.qiuxishi@alibaba-inc.com>
Suggested-by: Xishi Qiu <xishi.qiuxishi@alibaba-inc.com>
Tested-by: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: <zy.zhengyi@alibaba-inc.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
1 parent 30aba66
Raw File
sun.c
// SPDX-License-Identifier: GPL-2.0
/*
 *  fs/partitions/sun.c
 *
 *  Code extracted from drivers/block/genhd.c
 *
 *  Copyright (C) 1991-1998  Linus Torvalds
 *  Re-organised Feb 1998 Russell King
 */

#include "check.h"
#include "sun.h"

int sun_partition(struct parsed_partitions *state)
{
	int i;
	__be16 csum;
	int slot = 1;
	__be16 *ush;
	Sector sect;
	struct sun_disklabel {
		unsigned char info[128];   /* Informative text string */
		struct sun_vtoc {
		    __be32 version;     /* Layout version */
		    char   volume[8];   /* Volume name */
		    __be16 nparts;      /* Number of partitions */
		    struct sun_info {           /* Partition hdrs, sec 2 */
			__be16 id;
			__be16 flags;
		    } infos[8];
		    __be16 padding;     /* Alignment padding */
		    __be32 bootinfo[3];  /* Info needed by mboot */
		    __be32 sanity;       /* To verify vtoc sanity */
		    __be32 reserved[10]; /* Free space */
		    __be32 timestamp[8]; /* Partition timestamp */
		} vtoc;
		__be32 write_reinstruct; /* sectors to skip, writes */
		__be32 read_reinstruct;  /* sectors to skip, reads */
		unsigned char spare[148]; /* Padding */
		__be16 rspeed;     /* Disk rotational speed */
		__be16 pcylcount;  /* Physical cylinder count */
		__be16 sparecyl;   /* extra sects per cylinder */
		__be16 obs1;       /* gap1 */
		__be16 obs2;       /* gap2 */
		__be16 ilfact;     /* Interleave factor */
		__be16 ncyl;       /* Data cylinder count */
		__be16 nacyl;      /* Alt. cylinder count */
		__be16 ntrks;      /* Tracks per cylinder */
		__be16 nsect;      /* Sectors per track */
		__be16 obs3;       /* bhead - Label head offset */
		__be16 obs4;       /* ppart - Physical Partition */
		struct sun_partition {
			__be32 start_cylinder;
			__be32 num_sectors;
		} partitions[8];
		__be16 magic;      /* Magic number */
		__be16 csum;       /* Label xor'd checksum */
	} * label;
	struct sun_partition *p;
	unsigned long spc;
	char b[BDEVNAME_SIZE];
	int use_vtoc;
	int nparts;

	label = read_part_sector(state, 0, &sect);
	if (!label)
		return -1;

	p = label->partitions;
	if (be16_to_cpu(label->magic) != SUN_LABEL_MAGIC) {
/*		printk(KERN_INFO "Dev %s Sun disklabel: bad magic %04x\n",
		       bdevname(bdev, b), be16_to_cpu(label->magic)); */
		put_dev_sector(sect);
		return 0;
	}
	/* Look at the checksum */
	ush = ((__be16 *) (label+1)) - 1;
	for (csum = 0; ush >= ((__be16 *) label);)
		csum ^= *ush--;
	if (csum) {
		printk("Dev %s Sun disklabel: Csum bad, label corrupted\n",
		       bdevname(state->bdev, b));
		put_dev_sector(sect);
		return 0;
	}

	/* Check to see if we can use the VTOC table */
	use_vtoc = ((be32_to_cpu(label->vtoc.sanity) == SUN_VTOC_SANITY) &&
		    (be32_to_cpu(label->vtoc.version) == 1) &&
		    (be16_to_cpu(label->vtoc.nparts) <= 8));

	/* Use 8 partition entries if not specified in validated VTOC */
	nparts = (use_vtoc) ? be16_to_cpu(label->vtoc.nparts) : 8;

	/*
	 * So that old Linux-Sun partitions continue to work,
	 * alow the VTOC to be used under the additional condition ...
	 */
	use_vtoc = use_vtoc || !(label->vtoc.sanity ||
				 label->vtoc.version || label->vtoc.nparts);
	spc = be16_to_cpu(label->ntrks) * be16_to_cpu(label->nsect);
	for (i = 0; i < nparts; i++, p++) {
		unsigned long st_sector;
		unsigned int num_sectors;

		st_sector = be32_to_cpu(p->start_cylinder) * spc;
		num_sectors = be32_to_cpu(p->num_sectors);
		if (num_sectors) {
			put_partition(state, slot, st_sector, num_sectors);
			state->parts[slot].flags = 0;
			if (use_vtoc) {
				if (be16_to_cpu(label->vtoc.infos[i].id) == LINUX_RAID_PARTITION)
					state->parts[slot].flags |= ADDPART_FLAG_RAID;
				else if (be16_to_cpu(label->vtoc.infos[i].id) == SUN_WHOLE_DISK)
					state->parts[slot].flags |= ADDPART_FLAG_WHOLEDISK;
			}
		}
		slot++;
	}
	strlcat(state->pp_buf, "\n", PAGE_SIZE);
	put_dev_sector(sect);
	return 1;
}
back to top