Skip to content
  1. May 03, 2018
    • Coly Li's avatar
      bcache: set CACHE_SET_IO_DISABLE in bch_cached_dev_error() · 6147305c
      Coly Li authored 
      Commit c7b7bd07 ("bcache: add io_disable to struct cached_dev") tries
to stop bcache device by calling bcache_device_stop() when too many I/O
errors happened on backing device. But if there is internal I/O happening
on cache device (writeback scan, garbage collection, etc), a regular I/O
request triggers the internal I/Os may still holds a refcount of dc->count,
and the refcount may only be dropped after the internal I/O stopped.

By this patch, bch_cached_dev_error() will check if the backing device is
attached to a cache set, if yes that CACHE_SET_IO_DISABLE will be set to
flags of this cache set. Then internal I/Os on cache device will be
rejected and stopped immediately, and the bcache device can be stopped.

For people who are not familiar with the interesting refcount dependance,
let me explain a bit more how the fix works. Example the writeback thread
will scan cache device for dirty data writeback purpose. Before it stopps,
it holds a refcount of dc->count. When CACHE_SET_IO_DISABLE bit is set,
the internal I/O will stopped and the while-loop in bch_writeback_thread()
quits and calls cached_dev_put() to drop dc->count. If this is the last
refcount to drop, then cached_dev_detach_finish() will be called. In this
call back function, in turn closure_put(dc->disk.cl) is called to drop a
refcount of closure dc->disk.cl. If this is the last refcount of this
closure to drop, then cached_dev_flush() will be called. Then the cached
device is freed. So if CACHE_SET_IO_DISABLE is not set, the bache device
can not be stopped until all inernal cache device I/O stopped. For large
size cache device, and writeback thread competes locks with gc thread,
there might be a quite long time to wait.

Fixes: c7b7bd07

 ("bcache: add io_disable to struct cached_dev")
Signed-off-by: default avatarColy Li <colyli@suse.de>
Reviewed-by: default avatarHannes Reinecke <hare@suse.com>
Signed-off-by: default avatarJens Axboe <axboe@kernel.dk>
      6147305c
    • Coly Li's avatar
      bcache: store disk name in struct cache and struct cached_dev · 6e916a7e
      Coly Li authored 
      Current code uses bdevname() or bio_devname() to reference gendisk
disk name when bcache needs to display the disk names in kernel message.
It was safe before bcache device failure handling patch set merged in,
because when devices are failed, there was deadlock to prevent bcache
printing error messages with gendisk disk name. But after the failure
handling patch set merged, the deadlock is fixed, so it is possible
that the gendisk structure bdev->hd_disk is released when bdevname() is
called to reference bdev->bd_disk->disk_name[]. This is why I receive
bug report of NULL pointers deference panic.

This patch stores gendisk disk name in a buffer inside struct cache and
struct cached_dev, then print out the offline device name won't reference
bdev->hd_disk anymore. And this patch also avoids extra function calls
of bdevname() and bio_devnmae().

Changelog:
v3, add Reviewed-by from Hannes.
v2, call bdevname() earlier in register_bdev()
v1, first version with segguestion from Junhui Tang.

Fixes: c7b7bd07 ("bcache: add io_disable to struct cached_dev")
Fixes: 5138ac67

 ("bcache: fix misleading error message in bch_count_io_errors()")
Signed-off-by: default avatarColy Li <colyli@suse.de>
Reviewed-by: default avatarHannes Reinecke <hare@suse.com>
Signed-off-by: default avatarJens Axboe <axboe@kernel.dk>
      6e916a7e
  3. Apr 26, 2018
    • Omar Sandoval's avatar
      blk-mq: fix sysfs inflight counter · bf0ddaba
      Omar Sandoval authored 
      When the blk-mq inflight implementation was added, /proc/diskstats was
converted to use it, but /sys/block/$dev/inflight was not. Fix it by
adding another helper to count in-flight requests by data direction.

Fixes: f299b7c7

 ("blk-mq: provide internal in-flight variant")
Signed-off-by: default avatarOmar Sandoval <osandov@fb.com>
Signed-off-by: default avatarJens Axboe <axboe@kernel.dk>
      bf0ddaba
    • Omar Sandoval's avatar
      blk-mq: count allocated but not started requests in iostats inflight · 6131837b
      Omar Sandoval authored 
      In the legacy block case, we increment the counter right after we
allocate the request, not when the driver handles it. In both the legacy
and blk-mq cases, part_inc_in_flight() is called from
blk_account_io_start() right after we've allocated the request. blk-mq
only considers requests started requests as inflight, but this is
inconsistent with the legacy definition and the intention in the code.
This removes the started condition and instead counts all allocated
requests.

Fixes: f299b7c7

 ("blk-mq: provide internal in-flight variant")
Signed-off-by: default avatarOmar Sandoval <osandov@fb.com>
Signed-off-by: default avatarJens Axboe <axboe@kernel.dk>
      6131837b
    • Linus Torvalds's avatar
      Merge tag 'for_v4.17-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs · 69bfd470
      Linus Torvalds authored 
      Pull fsnotify fix from Jan Kara:
 "A fix of a fsnotify race causing panics / softlockups"

* tag 'for_v4.17-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs:
  fsnotify: Fix fsnotify_mark_connector race
      69bfd470
    • Linus Torvalds's avatar
      Merge tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi · 3442097b
      Linus Torvalds authored 
      Pull SCSI fixes from James Bottomley:
 "Eight bug fixes, one spelling update and one tracepoint addition.

  The most serious is probably the mptsas write same fix because it
  means anyone using these controllers sees errors when modern
  filesystems try to issue discards"

* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
  scsi: target: fix crash with iscsi target and dvd
  scsi: sd_zbc: Avoid that resetting a zone fails sporadically
  scsi: sd: Defer spinning up drive while SANITIZE is in progress
  scsi: megaraid_sas: Do not log an error if FW successfully initializes.
  scsi: ufs: add trace event for ufs upiu
  scsi: core: remove reference to scsi_show_extd_sense()
  scsi: mptsas: Disable WRITE SAME
  scsi: fnic: fix spelling mistake in fnic stats "Abord" -> "Abort"
  scsi: scsi_debug: IMMED related delay adjustments
  scsi: iscsi: respond to netlink with unicast when appropriate
      3442097b
    • Linus Torvalds's avatar
      Merge tag 'for-linus-20180425' of git://git.kernel.dk/linux-block · 8fba70b0
      Linus Torvalds authored 
      Pull block updates from Jens Axboe:
 "I ended up sitting on this about a week longer than I wanted to, since
  we were hashing out details with a timeout change. I've now killed
  that patch, so we can flush the existing queue in due time.

  This contains:

   - Fix for an old regression, where entering the queue can be
     disturbed by a signal to the process. This can cause spurious EIO.
     Fix from Alan Jenkins.

   - cdrom information leak fix from Dan.

   - Trivial helper for testing queue FUA from Dave Chinner, part of his
     O_DIRECT FUA series.

   - Series of swim fixes from Finn that actually makes it work again.

   - Loop O_DIRECT corruption fix, which caused data corruption in
     production for us. From me.

   - BFQ crash fix from me.

   - bcache maintainer update. Michael no longer has the time to do it,
     Coly has stepped up to serve as the new maintainer.

   - blkcg locking fixes from Jiang Biao.

   - Revert of a change from this merge window from Ming, that causes an
     issue on some hardware.

   - Minor clarification doc addition from Linus Walleij"

* tag 'for-linus-20180425' of git://git.kernel.dk/linux-block: (22 commits)
  Revert "blk-mq: remove code for dealing with remapping queue"
  block: mq: Add some minor doc for core structs
  bcache: mark Coly Li as bcache maintainer
  MAINTAINERS: Remove me as maintainer of bcache
  blkcg: init root blkcg_gq under lock
  blkcg: small fix on comment in blkcg_init_queue
  blkcg: don't hold blkcg lock when deactivating policy
  block: add blk_queue_fua() helper function
  cdrom: information leak in cdrom_ioctl_media_changed()
  bfq-iosched: ensure to clear bic/bfqq pointers when preparing request
  blk-mq: start request gstate with gen 1
  block/swim: Select appropriate drive on device open
  block/swim: Fix IO error at end of medium
  block/swim: Check drive type
  block/swim: Rename macros to avoid inconsistent inverted logic
  block/swim: Don't log an error message for an invalid ioctl
  block/swim: Remove extra put_disk() call from error path
  block/swim: Fix array bounds check
  m68k/mac: Don't remap SWIM MMIO region
  loop: handle short DIO reads
  ...
      8fba70b0
    • Linus Torvalds's avatar
      Merge tag 'riscv-for-linus-4.17-rc3' of... · c6dc3e71
      Linus Torvalds authored 
      Merge tag 'riscv-for-linus-4.17-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/palmer/riscv-linux

Pull RISC-V fixes from Palmer Dabbelt:
 "This contains three small fixes related to the RISC-V port that I'd
  like to target for 4.17-rc3:

   - a Kconfig cleanup to select DMA_DIRECT_OPS instead of redefining it
     in arch/riscv

   - the removal of asm/handle_irq.h, which doesn't exist, from our arch
     header list

   - the addition of "-no-pie" the link rules for our VDSO-related
     files, which fixes the build on systems where PIE is enabled by
     default"

* tag 'riscv-for-linus-4.17-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/palmer/riscv-linux:
  RISC-V: build vdso-dummy.o with -no-pie
  riscv: there is no <asm/handle_irq.h>
  riscv: select DMA_DIRECT_OPS instead of redefining it
      c6dc3e71
    • Linus Torvalds's avatar
      Merge tag 'dma-mapping-4.17-3' of git://git.infradead.org/users/hch/dma-mapping · 26ed24e4
      Linus Torvalds authored 
      Pull dma-mapping fixes from Christoph Hellwig:
 "A few small dma-mapping fixes for Linux 4.17-rc3:

   - don't loop to try GFP_DMA allocations if ZONE_DMA is not actually
     enabled (regression in 4.16)

   - don't try to do virt_to_page before we know we actuall have a valid
     page in dma_common_mmap

   - a comment fixup related to the above fix"

* tag 'dma-mapping-4.17-3' of git://git.infradead.org/users/hch/dma-mapping:
  dma-mapping: postpone cpu addr translation on mmap
  dma-coherent: clarify dma_mmap_from_dev_coherent documentation
  dma-direct: don't retry allocation for no-op GFP_DMA
      26ed24e4
  5. Apr 25, 2018
  7. Apr 24, 2018
  9. Apr 23, 2018
    • Xin Long's avatar
      bonding: do not set slave_dev npinfo before slave_enable_netpoll in bond_enslave · ddea788c
      Xin Long authored 
      After Commit 8a8efa22 ("bonding: sync netpoll code with bridge"), it
would set slave_dev npinfo in slave_enable_netpoll when enslaving a dev
if bond->dev->npinfo was set.

However now slave_dev npinfo is set with bond->dev->npinfo before calling
slave_enable_netpoll. With slave_dev npinfo set, __netpoll_setup called
in slave_enable_netpoll will not call slave dev's .ndo_netpoll_setup().
It causes that the lower dev of this slave dev can't set its npinfo.

One way to reproduce it:

  # modprobe bonding
  # brctl addbr br0
  # brctl addif br0 eth1
  # ifconfig bond0 192.168.122.1/24 up
  # ifenslave bond0 eth2
  # systemctl restart netconsole
  # ifenslave bond0 br0
  # ifconfig eth2 down
  # systemctl restart netconsole

The netpoll won't really work.

This patch is to remove that slave_dev npinfo setting in bond_enslave().

Fixes: 8a8efa22

 ("bonding: sync netpoll code with bridge")
Signed-off-by: default avatarXin Long <lucien.xin@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      ddea788c
    • Jann Horn's avatar
      tcp: don't read out-of-bounds opsize · 7e5a206a
      Jann Horn authored 
      The old code reads the "opsize" variable from out-of-bounds memory (first
byte behind the segment) if a broken TCP segment ends directly after an
opcode that is neither EOL nor NOP.

The result of the read isn't used for anything, so the worst thing that
could theoretically happen is a pagefault; and since the physmap is usually
mostly contiguous, even that seems pretty unlikely.

The following C reproducer triggers the uninitialized read - however, you
can't actually see anything happen unless you put something like a
pr_warn() in tcp_parse_md5sig_option() to print the opsize.

====================================
#define _GNU_SOURCE
#include <arpa/inet.h>
#include <stdlib.h>
#include <errno.h>
#include <stdarg.h>
#include <net/if.h>
#include <linux/if.h>
#include <linux/ip.h>
#include <linux/tcp.h>
#include <linux/in.h>
#include <linux/if_tun.h>
#include <err.h>
#include <sys/types.h>
#include <sys/stat.h>
#include <fcntl.h>
#include <string.h>
#include <stdio.h>
#include <unistd.h>
#include <sys/ioctl.h>
#include <assert.h>

void systemf(const char *command, ...) {
  char *full_command;
  va_list ap;
  va_start(ap, command);
  if (vasprintf(&full_command, command, ap) == -1)
    err(1, "vasprintf");
  va_end(ap);
  printf("systemf: <<<%s>>>\n", full_command);
  system(full_command);
}

char *devname;

int tun_alloc(char *name) {
  int fd = open("/dev/net/tun", O_RDWR);
  if (fd == -1)
    err(1, "open tun dev");
  static struct ifreq req = { .ifr_flags = IFF_TUN|IFF_NO_PI };
  strcpy(req.ifr_name, name);
  if (ioctl(fd, TUNSETIFF, &req))
    err(1, "TUNSETIFF");
  devname = req.ifr_name;
  printf("device name: %s\n", devname);
  return fd;
}

#define IPADDR(a,b,c,d) (((a)<<0)+((b)<<8)+((c)<<16)+((d)<<24))

void sum_accumulate(unsigned int *sum, void *data, int len) {
  assert((len&2)==0);
  for (int i=0; i<len/2; i++) {
    *sum += ntohs(((unsigned short *)data)[i]);
  }
}

unsigned short sum_final(unsigned int sum) {
  sum = (sum >> 16) + (sum & 0xffff);
  sum = (sum >> 16) + (sum & 0xffff);
  return htons(~sum);
}

void fix_ip_sum(struct iphdr *ip) {
  unsigned int sum = 0;
  sum_accumulate(&sum, ip, sizeof(*ip));
  ip->check = sum_final(sum);
}

void fix_tcp_sum(struct iphdr *ip, struct tcphdr *tcp) {
  unsigned int sum = 0;
  struct {
    unsigned int saddr;
    unsigned int daddr;
    unsigned char pad;
    unsigned char proto_num;
    unsigned short tcp_len;
  } fakehdr = {
    .saddr = ip->saddr,
    .daddr = ip->daddr,
    .proto_num = ip->protocol,
    .tcp_len = htons(ntohs(ip->tot_len) - ip->ihl*4)
  };
  sum_accumulate(&sum, &fakehdr, sizeof(fakehdr));
  sum_accumulate(&sum, tcp, tcp->doff*4);
  tcp->check = sum_final(sum);
}

int main(void) {
  int tun_fd = tun_alloc("inject_dev%d");
  systemf("ip link set %s up", devname);
  systemf("ip addr add 192.168.42.1/24 dev %s", devname);

  struct {
    struct iphdr ip;
    struct tcphdr tcp;
    unsigned char tcp_opts[20];
  } __attribute__((packed)) syn_packet = {
    .ip = {
      .ihl = sizeof(struct iphdr)/4,
      .version = 4,
      .tot_len = htons(sizeof(syn_packet)),
      .ttl = 30,
      .protocol = IPPROTO_TCP,
      /* FIXUP check */
      .saddr = IPADDR(192,168,42,2),
      .daddr = IPADDR(192,168,42,1)
    },
    .tcp = {
      .source = htons(1),
      .dest = htons(1337),
      .seq = 0x12345678,
      .doff = (sizeof(syn_packet.tcp)+sizeof(syn_packet.tcp_opts))/4,
      .syn = 1,
      .window = htons(64),
      .check = 0 /*FIXUP*/
    },
    .tcp_opts = {
      /* INVALID: trailing MD5SIG opcode after NOPs */
      1, 1, 1, 1, 1,
      1, 1, 1, 1, 1,
      1, 1, 1, 1, 1,
      1, 1, 1, 1, 19
    }
  };
  fix_ip_sum(&syn_packet.ip);
  fix_tcp_sum(&syn_packet.ip, &syn_packet.tcp);
  while (1) {
    int write_res = write(tun_fd, &syn_packet, sizeof(syn_packet));
    if (write_res != sizeof(syn_packet))
      err(1, "packet write failed");
  }
}
====================================

Fixes: cfb6eeb4

 ("[TCP]: MD5 Signature Option (RFC2385) support.")
Signed-off-by: default avatarJann Horn <jannh@google.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      7e5a206a