Revision 85bd839983778fcd0c1c043327b14a046e979b39 authored by Gu Zheng on 10 June 2015, 18:14:43 UTC, committed by Linus Torvalds on 10 June 2015, 23:43:43 UTC
Izumi found the following oops when hot re-adding a node:

    BUG: unable to handle kernel paging request at ffffc90008963690
    IP: __wake_up_bit+0x20/0x70
    Oops: 0000 [#1] SMP
    CPU: 68 PID: 1237 Comm: rs:main Q:Reg Not tainted 4.1.0-rc5 #80
    Hardware name: FUJITSU PRIMEQUEST2800E/SB, BIOS PRIMEQUEST 2000 Series BIOS Version 1.87 04/28/2015
    task: ffff880838df8000 ti: ffff880017b94000 task.ti: ffff880017b94000
    RIP: 0010:[<ffffffff810dff80>]  [<ffffffff810dff80>] __wake_up_bit+0x20/0x70
    RSP: 0018:ffff880017b97be8  EFLAGS: 00010246
    RAX: ffffc90008963690 RBX: 00000000003c0000 RCX: 000000000000a4c9
    RDX: 0000000000000000 RSI: ffffea101bffd500 RDI: ffffc90008963648
    RBP: ffff880017b97c08 R08: 0000000002000020 R09: 0000000000000000
    R10: 0000000000000000 R11: 0000000000000000 R12: ffff8a0797c73800
    R13: ffffea101bffd500 R14: 0000000000000001 R15: 00000000003c0000
    FS:  00007fcc7ffff700(0000) GS:ffff880874800000(0000) knlGS:0000000000000000
    CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
    CR2: ffffc90008963690 CR3: 0000000836761000 CR4: 00000000001407e0
    Call Trace:
      unlock_page+0x6d/0x70
      generic_write_end+0x53/0xb0
      xfs_vm_write_end+0x29/0x80 [xfs]
      generic_perform_write+0x10a/0x1e0
      xfs_file_buffered_aio_write+0x14d/0x3e0 [xfs]
      xfs_file_write_iter+0x79/0x120 [xfs]
      __vfs_write+0xd4/0x110
      vfs_write+0xac/0x1c0
      SyS_write+0x58/0xd0
      system_call_fastpath+0x12/0x76
    Code: 5d c3 66 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 55 48 89 e5 48 83 ec 20 65 48 8b 04 25 28 00 00 00 48 89 45 f8 31 c0 48 8d 47 48 <48> 39 47 48 48 c7 45 e8 00 00 00 00 48 c7 45 f0 00 00 00 00 48
    RIP  [<ffffffff810dff80>] __wake_up_bit+0x20/0x70
     RSP <ffff880017b97be8>
    CR2: ffffc90008963690

Reproduce method (re-add a node)::
  Hot-add nodeA --> remove nodeA --> hot-add nodeA (panic)

This seems an use-after-free problem, and the root cause is
zone->wait_table was not set to *NULL* after free it in
try_offline_node.

When hot re-add a node, we will reuse the pgdat of it, so does the zone
struct, and when add pages to the target zone, it will init the zone
first (including the wait_table) if the zone is not initialized.  The
judgement of zone initialized is based on zone->wait_table:

	static inline bool zone_is_initialized(struct zone *zone)
	{
		return !!zone->wait_table;
	}

so if we do not set the zone->wait_table to *NULL* after free it, the
memory hotplug routine will skip the init of new zone when hot re-add
the node, and the wait_table still points to the freed memory, then we
will access the invalid address when trying to wake up the waiting
people after the i/o operation with the page is done, such as mentioned
above.

Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
Reported-by: Taku Izumi <izumi.taku@jp.fujitsu.com>
Reviewed by: Yasuaki Ishimatsu <isimatu.yasuaki@jp.fujitsu.com>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Tang Chen <tangchen@cn.fujitsu.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 5879ae5
Raw File
cs5535.c
/*
 * Copyright (C) 2004-2005 Advanced Micro Devices, Inc.
 * Copyright (C)      2007 Bartlomiej Zolnierkiewicz
 *
 * History:
 * 09/20/2005 - Jaya Kumar <jayakumar.ide@gmail.com>
 * - Reworked tuneproc, set_drive, misc mods to prep for mainline
 * - Work was sponsored by CIS (M) Sdn Bhd.
 * Ported to Kernel 2.6.11 on June 26, 2005 by
 *   Wolfgang Zuleger <wolfgang.zuleger@gmx.de>
 *   Alexander Kiausch <alex.kiausch@t-online.de>
 * Originally developed by AMD for 2.4/2.6
 *
 * Development of this chipset driver was funded
 * by the nice folks at National Semiconductor/AMD.
 *
 * This program is free software; you can redistribute it and/or modify it
 * under the terms of the GNU General Public License version 2 as published by
 * the Free Software Foundation.
 *
 * Documentation:
 *  CS5535 documentation available from AMD
 */

#include <linux/module.h>
#include <linux/pci.h>
#include <linux/ide.h>

#define DRV_NAME "cs5535"

#define MSR_ATAC_BASE		0x51300000
#define ATAC_GLD_MSR_CAP	(MSR_ATAC_BASE+0)
#define ATAC_GLD_MSR_CONFIG	(MSR_ATAC_BASE+0x01)
#define ATAC_GLD_MSR_SMI	(MSR_ATAC_BASE+0x02)
#define ATAC_GLD_MSR_ERROR	(MSR_ATAC_BASE+0x03)
#define ATAC_GLD_MSR_PM		(MSR_ATAC_BASE+0x04)
#define ATAC_GLD_MSR_DIAG	(MSR_ATAC_BASE+0x05)
#define ATAC_IO_BAR		(MSR_ATAC_BASE+0x08)
#define ATAC_RESET		(MSR_ATAC_BASE+0x10)
#define ATAC_CH0D0_PIO		(MSR_ATAC_BASE+0x20)
#define ATAC_CH0D0_DMA		(MSR_ATAC_BASE+0x21)
#define ATAC_CH0D1_PIO		(MSR_ATAC_BASE+0x22)
#define ATAC_CH0D1_DMA		(MSR_ATAC_BASE+0x23)
#define ATAC_PCI_ABRTERR	(MSR_ATAC_BASE+0x24)
#define ATAC_BM0_CMD_PRIM	0x00
#define ATAC_BM0_STS_PRIM	0x02
#define ATAC_BM0_PRD		0x04
#define CS5535_CABLE_DETECT	0x48

/* Format I PIO settings. We separate out cmd and data for safer timings */

static unsigned int cs5535_pio_cmd_timings[5] =
{ 0xF7F4, 0x53F3, 0x13F1, 0x5131, 0x1131 };
static unsigned int cs5535_pio_dta_timings[5] =
{ 0xF7F4, 0xF173, 0x8141, 0x5131, 0x1131 };

static unsigned int cs5535_mwdma_timings[3] =
{ 0x7F0FFFF3, 0x7F035352, 0x7f024241 };

static unsigned int cs5535_udma_timings[5] =
{ 0x7F7436A1, 0x7F733481, 0x7F723261, 0x7F713161, 0x7F703061 };

/* Macros to check if the register is the reset value -  reset value is an
   invalid timing and indicates the register has not been set previously */

#define CS5535_BAD_PIO(timings) ( (timings&~0x80000000UL) == 0x00009172 )
#define CS5535_BAD_DMA(timings) ( (timings & 0x000FFFFF) == 0x00077771 )

/****
 *	cs5535_set_speed         -     Configure the chipset to the new speed
 *	@drive: Drive to set up
 *	@speed: desired speed
 *
 *	cs5535_set_speed() configures the chipset to a new speed.
 */
static void cs5535_set_speed(ide_drive_t *drive, const u8 speed)
{
	u32 reg = 0, dummy;
	u8 unit = drive->dn & 1;

	/* Set the PIO timings */
	if (speed < XFER_SW_DMA_0) {
		ide_drive_t *pair = ide_get_pair_dev(drive);
		u8 cmd, pioa;

		cmd = pioa = speed - XFER_PIO_0;

		if (pair) {
			u8 piob = pair->pio_mode - XFER_PIO_0;

			if (piob < cmd)
				cmd = piob;
		}

		/* Write the speed of the current drive */
		reg = (cs5535_pio_cmd_timings[cmd] << 16) |
			cs5535_pio_dta_timings[pioa];
		wrmsr(unit ? ATAC_CH0D1_PIO : ATAC_CH0D0_PIO, reg, 0);

		/* And if nessesary - change the speed of the other drive */
		rdmsr(unit ?  ATAC_CH0D0_PIO : ATAC_CH0D1_PIO, reg, dummy);

		if (((reg >> 16) & cs5535_pio_cmd_timings[cmd]) !=
			cs5535_pio_cmd_timings[cmd]) {
			reg &= 0x0000FFFF;
			reg |= cs5535_pio_cmd_timings[cmd] << 16;
			wrmsr(unit ? ATAC_CH0D0_PIO : ATAC_CH0D1_PIO, reg, 0);
		}

		/* Set bit 31 of the DMA register for PIO format 1 timings */
		rdmsr(unit ?  ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, reg, dummy);
		wrmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA,
					reg | 0x80000000UL, 0);
	} else {
		rdmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, reg, dummy);

		reg &= 0x80000000UL;  /* Preserve the PIO format bit */

		if (speed >= XFER_UDMA_0 && speed <= XFER_UDMA_4)
			reg |= cs5535_udma_timings[speed - XFER_UDMA_0];
		else if (speed >= XFER_MW_DMA_0 && speed <= XFER_MW_DMA_2)
			reg |= cs5535_mwdma_timings[speed - XFER_MW_DMA_0];
		else
			return;

		wrmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, reg, 0);
	}
}

/**
 *	cs5535_set_dma_mode	-	set host controller for DMA mode
 *	@hwif: port
 *	@drive: drive
 *
 *	Programs the chipset for DMA mode.
 */

static void cs5535_set_dma_mode(ide_hwif_t *hwif, ide_drive_t *drive)
{
	cs5535_set_speed(drive, drive->dma_mode);
}

/**
 *	cs5535_set_pio_mode	-	set host controller for PIO mode
 *	@hwif: port
 *	@drive: drive
 *
 *	A callback from the upper layers for PIO-only tuning.
 */

static void cs5535_set_pio_mode(ide_hwif_t *hwif, ide_drive_t *drive)
{
	cs5535_set_speed(drive, drive->pio_mode);
}

static u8 cs5535_cable_detect(ide_hwif_t *hwif)
{
	struct pci_dev *dev = to_pci_dev(hwif->dev);
	u8 bit;

	/* if a 80 wire cable was detected */
	pci_read_config_byte(dev, CS5535_CABLE_DETECT, &bit);

	return (bit & 1) ? ATA_CBL_PATA80 : ATA_CBL_PATA40;
}

static const struct ide_port_ops cs5535_port_ops = {
	.set_pio_mode		= cs5535_set_pio_mode,
	.set_dma_mode		= cs5535_set_dma_mode,
	.cable_detect		= cs5535_cable_detect,
};

static const struct ide_port_info cs5535_chipset = {
	.name		= DRV_NAME,
	.port_ops	= &cs5535_port_ops,
	.host_flags	= IDE_HFLAG_SINGLE | IDE_HFLAG_POST_SET_MODE,
	.pio_mask	= ATA_PIO4,
	.mwdma_mask	= ATA_MWDMA2,
	.udma_mask	= ATA_UDMA4,
};

static int cs5535_init_one(struct pci_dev *dev, const struct pci_device_id *id)
{
	return ide_pci_init_one(dev, &cs5535_chipset, NULL);
}

static const struct pci_device_id cs5535_pci_tbl[] = {
	{ PCI_VDEVICE(NS, PCI_DEVICE_ID_NS_CS5535_IDE), 0 },
	{ PCI_VDEVICE(AMD, PCI_DEVICE_ID_AMD_CS5535_IDE), },
	{ 0, },
};

MODULE_DEVICE_TABLE(pci, cs5535_pci_tbl);

static struct pci_driver cs5535_pci_driver = {
	.name		= "CS5535_IDE",
	.id_table	= cs5535_pci_tbl,
	.probe		= cs5535_init_one,
	.remove		= ide_pci_remove,
	.suspend	= ide_pci_suspend,
	.resume		= ide_pci_resume,
};

static int __init cs5535_ide_init(void)
{
	return ide_pci_register_driver(&cs5535_pci_driver);
}

static void __exit cs5535_ide_exit(void)
{
	pci_unregister_driver(&cs5535_pci_driver);
}

module_init(cs5535_ide_init);
module_exit(cs5535_ide_exit);

MODULE_AUTHOR("AMD");
MODULE_DESCRIPTION("PCI driver module for AMD/NS CS5535 IDE");
MODULE_LICENSE("GPL");
back to top