2014-06-27 04:17:02

by Minfei Huang

[permalink] [raw]
Subject: [PATCH] dm-io: Prevent the danging point of the sync io callback function

BUG: unable to handle kernel NULL pointer dereference at 0000000000000046
IP: [<ffffffffa0009cef>] dec_count+0x5f/0x80 [dm_mod]
PGD 0
Oops: 0000 [#1] SMP
last sysfs file: /sys/devices/pci0000:00/0000:00:02.2/0000:02:00.0/host0/scsi_host/host0/proc_name

Pid: 2708, comm: kcopyd Tainted: G W --------------- H 2.6.32-279.19.5.el6.ucloud.x86_64 #1 Dell Inc. PowerEdge R720xd/0DCWD1
RIP: 0010:[<ffffffffa0009cef>] [<ffffffffa0009cef>] dec_count+0x5f/0x80 [dm_mod]
RSP: 0018:ffff880100603c30 EFLAGS: 00010246
RAX: 0000000000000046 RBX: ffff8817968a5c30 RCX: 0000000000000000
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8817968a5c00
RBP: ffff880100603c50 R08: 0000000000000000 R09: 0000000000000000
R10: ffff880caa594cc0 R11: 0000000000000000 R12: ffff8817968a5c80
R13: ffffffff81013963 R14: 0000000000001000 R15: 0000000000000000
FS: 0000000000000000(0000) GS:ffff880100600000(0000) knlGS:0000000000000000
CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b
CR2: 0000000000000046 CR3: 000000020c309000 CR4: 00000000001426e0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Process kcopyd (pid: 2708, threadinfo ffff88180cd26000, task ffff881841c9aa80)
Stack:
ffff880100603c40 ffff880aa8b32300 0000000000000000 ffff8817968a5c00
<d> ffff880100603c80 ffffffffa000a12a 0000000000000000 ffff880aa8b32300
<d> 0000000000000000 ffff880caa594cc0 ffff880100603c90 ffffffff811bcf6d
Call Trace:
<IRQ>
[<ffffffffa000a12a>] endio+0x4a/0x70 [dm_mod]
[<ffffffff811bcf6d>] bio_endio+0x1d/0x40
[<ffffffff81260beb>] req_bio_endio+0x9b/0xe0
[<ffffffff81263114>] blk_update_request+0x104/0x500
[<ffffffff81263331>] ? blk_update_request+0x321/0x500
[<ffffffff81263537>] blk_update_bidi_request+0x27/0xa0
[<ffffffff8126419f>] blk_end_bidi_request+0x2f/0x80
[<ffffffff81264240>] blk_end_request+0x10/0x20
[<ffffffff81375c6f>] scsi_io_completion+0xaf/0x6c0
[<ffffffff8136cb92>] scsi_finish_command+0xc2/0x130
[<ffffffff813763e5>] scsi_softirq_done+0x145/0x170
[<ffffffff812698ed>] blk_done_softirq+0x8d/0xa0
[<ffffffff81074c5f>] __do_softirq+0xdf/0x210
[<ffffffff8100c2cc>] call_softirq+0x1c/0x30
[<ffffffff8100df9d>] do_softirq+0xad/0xe0
[<ffffffff81074995>] irq_exit+0x95/0xa0
[<ffffffff81510515>] do_IRQ+0x75/0xf0
[<ffffffff8100ba53>] ret_from_intr+0x0/0x16

The value of rdi register(0xffff8817968a5c00) is the io pointer,
If the sync io, the address of io point must be alloc from stack.
SO
crash> struct thread_info ffff8817968a4000
struct thread_info {
task = 0xffff88180cd9a580,
exec_domain = 0xffffffff81a98ac0,
...
}

crash> struct task_struct 0xffff88180cd9a580
struct task_struct {
state = 2,
stack = 0xffff8817968a4000,
...
}

It shows value exactly when use the value of io address.

The io address in callback function will become the danging point,
cause by the thread of sync io wakes up by other threads
and return to relieve the io address,

Signed-off-by: Minfei Huang <[email protected]>
---
drivers/md/dm-io.c | 19 +++++++++++++++----
1 files changed, 15 insertions(+), 4 deletions(-)

diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index 3842ac7..f992913 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -38,6 +38,7 @@ struct io {
void *context;
void *vma_invalidate_address;
unsigned long vma_invalidate_size;
+ atomic_t wakeup;
} __attribute__((aligned(DM_IO_MAX_REGIONS)));

static struct kmem_cache *_dm_io_cache;
@@ -121,10 +122,16 @@ static void dec_count(struct io *io, unsigned int region, int error)
invalidate_kernel_vmap_range(io->vma_invalidate_address,
io->vma_invalidate_size);

- if (io->sleeper)
- wake_up_process(io->sleeper);
+ if (io->sleeper) {
+ struct task_struct *sleeper = io->sleeper;

- else {
+ atomic_set(&io->wakeup, 1);
+/*
+ * The thread may be waked up by other threads,
+ * if then the sync io point will become the dangling pointer
+ */
+ wake_up_process(sleeper);
+ } else {
unsigned long r = io->error_bits;
io_notify_fn fn = io->callback;
void *context = io->context;
@@ -401,12 +408,14 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
io->vma_invalidate_address = dp->vma_invalidate_address;
io->vma_invalidate_size = dp->vma_invalidate_size;

+ atomic_set(&io->wakeup, 0);
+
dispatch_io(rw, num_regions, where, dp, io, 1);

while (1) {
set_current_state(TASK_UNINTERRUPTIBLE);

- if (!atomic_read(&io->count))
+ if (atomic_read(&io->wakeup))
break;

io_schedule();
@@ -442,6 +451,8 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
io->vma_invalidate_address = dp->vma_invalidate_address;
io->vma_invalidate_size = dp->vma_invalidate_size;

+ atomic_set(&io->wakeup, 0);
+
dispatch_io(rw, num_regions, where, dp, io, 0);
return 0;
}
--
1.7.1


2014-06-27 15:11:22

by Joe Thornber

[permalink] [raw]
Subject: Re: [dm-devel] [PATCH] dm-io: Prevent the danging point of the sync io callback function

On Fri, Jun 27, 2014 at 12:01:30PM +0800, Minfei Huang wrote:
> The io address in callback function will become the danging point,
> cause by the thread of sync io wakes up by other threads
> and return to relieve the io address,

Yes, well found. I prefer the following fix however.

- Joe



Author: Joe Thornber <[email protected]>
Date: Fri Jun 27 15:49:29 2014 +0100

[dm-io] Fix a race condition in the wake up code for sync_io

There's a race condition between the atomic_dec_and_test(&io->count)
in dec_count() and the waking of the sync_io() thread. If the thread
is spuriously woken immediately after the decrement it may exit,
making the on the stack io struct invalid, yet the dec_count could
still be using it.

There are smaller fixes than the one here (eg, just take the io object
off the stack). But I feel this code could use a clean up.

- simplify dec_count().

- It always calls a callback fn now.
- It always frees the io object back to the pool.

- sync_io()

- Take the io object off the stack and allocate it from the pool the
same as async_io.
- Use a completion object rather than an explicit io_schedule()
loop. The callback triggers the completion.

Reported by: Minfei Huang

diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index 3842ac7..a0982e81 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -10,6 +10,7 @@
#include <linux/device-mapper.h>

#include <linux/bio.h>
+#include <linux/completion.h>
#include <linux/mempool.h>
#include <linux/module.h>
#include <linux/sched.h>
@@ -32,7 +33,6 @@ struct dm_io_client {
struct io {
unsigned long error_bits;
atomic_t count;
- struct task_struct *sleeper;
struct dm_io_client *client;
io_notify_fn callback;
void *context;
@@ -111,28 +111,27 @@ static void retrieve_io_and_region_from_bio(struct bio *bio, struct io **io,
* We need an io object to keep track of the number of bios that
* have been dispatched for a particular io.
*---------------------------------------------------------------*/
-static void dec_count(struct io *io, unsigned int region, int error)
+static void complete_io(struct io *io)
{
- if (error)
- set_bit(region, &io->error_bits);
+ unsigned long error_bits = io->error_bits;
+ io_notify_fn fn = io->callback;
+ void *context = io->context;

- if (atomic_dec_and_test(&io->count)) {
- if (io->vma_invalidate_size)
- invalidate_kernel_vmap_range(io->vma_invalidate_address,
- io->vma_invalidate_size);
+ if (io->vma_invalidate_size)
+ invalidate_kernel_vmap_range(io->vma_invalidate_address,
+ io->vma_invalidate_size);

- if (io->sleeper)
- wake_up_process(io->sleeper);
+ mempool_free(io, io->client->pool);
+ fn(error_bits, context);
+}

- else {
- unsigned long r = io->error_bits;
- io_notify_fn fn = io->callback;
- void *context = io->context;
+static void dec_count(struct io *io, unsigned int region, int error)
+{
+ if (error)
+ set_bit(region, &io->error_bits);

- mempool_free(io, io->client->pool);
- fn(r, context);
- }
- }
+ if (atomic_dec_and_test(&io->count))
+ complete_io(io);
}

static void endio(struct bio *bio, int error)
@@ -375,48 +374,49 @@ static void dispatch_io(int rw, unsigned int num_regions,
dec_count(io, 0, 0);
}

+struct sync_io {
+ unsigned long error_bits;
+ struct completion complete;
+};
+
+static void sync_complete(unsigned long error, void *context)
+{
+ struct sync_io *sio = context;
+ sio->error_bits = error;
+ complete(&sio->complete);
+}
+
static int sync_io(struct dm_io_client *client, unsigned int num_regions,
struct dm_io_region *where, int rw, struct dpages *dp,
unsigned long *error_bits)
{
- /*
- * gcc <= 4.3 can't do the alignment for stack variables, so we must
- * align it on our own.
- * volatile prevents the optimizer from removing or reusing
- * "io_" field from the stack frame (allowed in ANSI C).
- */
- volatile char io_[sizeof(struct io) + __alignof__(struct io) - 1];
- struct io *io = (struct io *)PTR_ALIGN(&io_, __alignof__(struct io));
+ struct io *io;
+ struct sync_io sio;

if (num_regions > 1 && (rw & RW_MASK) != WRITE) {
WARN_ON(1);
return -EIO;
}

+ init_completion(&sio.complete);
+
+ io = mempool_alloc(client->pool, GFP_NOIO);
io->error_bits = 0;
atomic_set(&io->count, 1); /* see dispatch_io() */
- io->sleeper = current;
+ io->callback = sync_complete;
+ io->context = &sio;
io->client = client;

io->vma_invalidate_address = dp->vma_invalidate_address;
io->vma_invalidate_size = dp->vma_invalidate_size;

dispatch_io(rw, num_regions, where, dp, io, 1);
-
- while (1) {
- set_current_state(TASK_UNINTERRUPTIBLE);
-
- if (!atomic_read(&io->count))
- break;
-
- io_schedule();
- }
- set_current_state(TASK_RUNNING);
+ wait_for_completion_io(&sio.complete);

if (error_bits)
- *error_bits = io->error_bits;
+ *error_bits = sio.error_bits;

- return io->error_bits ? -EIO : 0;
+ return sio.error_bits ? -EIO : 0;
}

static int async_io(struct dm_io_client *client, unsigned int num_regions,
@@ -434,7 +434,6 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
io = mempool_alloc(client->pool, GFP_NOIO);
io->error_bits = 0;
atomic_set(&io->count, 1); /* see dispatch_io() */
- io->sleeper = NULL;
io->client = client;
io->callback = fn;
io->context = context;

2014-06-27 18:30:00

by Mikulas Patocka

[permalink] [raw]
Subject: Re: [dm-devel] [PATCH] dm-io: Prevent the danging point of the sync io callback function



On Fri, 27 Jun 2014, Minfei Huang wrote:

> BUG: unable to handle kernel NULL pointer dereference at 0000000000000046
> IP: [<ffffffffa0009cef>] dec_count+0x5f/0x80 [dm_mod]
> PGD 0
> Oops: 0000 [#1] SMP
> last sysfs file: /sys/devices/pci0000:00/0000:00:02.2/0000:02:00.0/host0/scsi_host/host0/proc_name
>
> Pid: 2708, comm: kcopyd Tainted: G W --------------- H 2.6.32-279.19.5.el6.ucloud.x86_64 #1 Dell Inc. PowerEdge R720xd/0DCWD1
> RIP: 0010:[<ffffffffa0009cef>] [<ffffffffa0009cef>] dec_count+0x5f/0x80 [dm_mod]
> RSP: 0018:ffff880100603c30 EFLAGS: 00010246
> RAX: 0000000000000046 RBX: ffff8817968a5c30 RCX: 0000000000000000
> RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8817968a5c00
> RBP: ffff880100603c50 R08: 0000000000000000 R09: 0000000000000000
> R10: ffff880caa594cc0 R11: 0000000000000000 R12: ffff8817968a5c80
> R13: ffffffff81013963 R14: 0000000000001000 R15: 0000000000000000
> FS: 0000000000000000(0000) GS:ffff880100600000(0000) knlGS:0000000000000000
> CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b
> CR2: 0000000000000046 CR3: 000000020c309000 CR4: 00000000001426e0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> Process kcopyd (pid: 2708, threadinfo ffff88180cd26000, task ffff881841c9aa80)
> Stack:
> ffff880100603c40 ffff880aa8b32300 0000000000000000 ffff8817968a5c00
> <d> ffff880100603c80 ffffffffa000a12a 0000000000000000 ffff880aa8b32300
> <d> 0000000000000000 ffff880caa594cc0 ffff880100603c90 ffffffff811bcf6d
> Call Trace:
> <IRQ>
> [<ffffffffa000a12a>] endio+0x4a/0x70 [dm_mod]
> [<ffffffff811bcf6d>] bio_endio+0x1d/0x40
> [<ffffffff81260beb>] req_bio_endio+0x9b/0xe0
> [<ffffffff81263114>] blk_update_request+0x104/0x500
> [<ffffffff81263331>] ? blk_update_request+0x321/0x500
> [<ffffffff81263537>] blk_update_bidi_request+0x27/0xa0
> [<ffffffff8126419f>] blk_end_bidi_request+0x2f/0x80
> [<ffffffff81264240>] blk_end_request+0x10/0x20
> [<ffffffff81375c6f>] scsi_io_completion+0xaf/0x6c0
> [<ffffffff8136cb92>] scsi_finish_command+0xc2/0x130
> [<ffffffff813763e5>] scsi_softirq_done+0x145/0x170
> [<ffffffff812698ed>] blk_done_softirq+0x8d/0xa0
> [<ffffffff81074c5f>] __do_softirq+0xdf/0x210
> [<ffffffff8100c2cc>] call_softirq+0x1c/0x30
> [<ffffffff8100df9d>] do_softirq+0xad/0xe0
> [<ffffffff81074995>] irq_exit+0x95/0xa0
> [<ffffffff81510515>] do_IRQ+0x75/0xf0
> [<ffffffff8100ba53>] ret_from_intr+0x0/0x16
>
> The value of rdi register(0xffff8817968a5c00) is the io pointer,
> If the sync io, the address of io point must be alloc from stack.
> SO
> crash> struct thread_info ffff8817968a4000
> struct thread_info {
> task = 0xffff88180cd9a580,
> exec_domain = 0xffffffff81a98ac0,
> ...
> }
>
> crash> struct task_struct 0xffff88180cd9a580
> struct task_struct {
> state = 2,
> stack = 0xffff8817968a4000,
> ...
> }
>
> It shows value exactly when use the value of io address.
>
> The io address in callback function will become the danging point,
> cause by the thread of sync io wakes up by other threads
> and return to relieve the io address,
>
> Signed-off-by: Minfei Huang <[email protected]>
> ---
> drivers/md/dm-io.c | 19 +++++++++++++++----
> 1 files changed, 15 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
> index 3842ac7..f992913 100644
> --- a/drivers/md/dm-io.c
> +++ b/drivers/md/dm-io.c
> @@ -38,6 +38,7 @@ struct io {
> void *context;
> void *vma_invalidate_address;
> unsigned long vma_invalidate_size;
> + atomic_t wakeup;
> } __attribute__((aligned(DM_IO_MAX_REGIONS)));
>
> static struct kmem_cache *_dm_io_cache;
> @@ -121,10 +122,16 @@ static void dec_count(struct io *io, unsigned int region, int error)
> invalidate_kernel_vmap_range(io->vma_invalidate_address,
> io->vma_invalidate_size);
>
> - if (io->sleeper)
> - wake_up_process(io->sleeper);
> + if (io->sleeper) {
> + struct task_struct *sleeper = io->sleeper;
>
> - else {
> + atomic_set(&io->wakeup, 1);

The problem here is that the processor may reorder the read of io->sleeper
with atomic_set(&io->wakeup, 1); (performing atomic_set first and "sleeper
= io->sleeper" afterwards) exposing the same race condition.

You need to use memory barriers to avoid reordering, but I think the
solution with the completion is better (the completion takes care of
barriers automatically).

Mikulas


> +/*
> + * The thread may be waked up by other threads,
> + * if then the sync io point will become the dangling pointer
> + */
> + wake_up_process(sleeper);
> + } else {
> unsigned long r = io->error_bits;
> io_notify_fn fn = io->callback;
> void *context = io->context;
> @@ -401,12 +408,14 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
> io->vma_invalidate_address = dp->vma_invalidate_address;
> io->vma_invalidate_size = dp->vma_invalidate_size;
>
> + atomic_set(&io->wakeup, 0);
> +
> dispatch_io(rw, num_regions, where, dp, io, 1);
>
> while (1) {
> set_current_state(TASK_UNINTERRUPTIBLE);
>
> - if (!atomic_read(&io->count))
> + if (atomic_read(&io->wakeup))
> break;
>
> io_schedule();
> @@ -442,6 +451,8 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
> io->vma_invalidate_address = dp->vma_invalidate_address;
> io->vma_invalidate_size = dp->vma_invalidate_size;
>
> + atomic_set(&io->wakeup, 0);
> +
> dispatch_io(rw, num_regions, where, dp, io, 0);
> return 0;
> }
> --
> 1.7.1
>
>
> --
> dm-devel mailing list
> [email protected]
> https://www.redhat.com/mailman/listinfo/dm-devel
>

2014-06-27 18:44:52

by Mikulas Patocka

[permalink] [raw]
Subject: Re: [dm-devel] [PATCH] dm-io: Prevent the danging point of the sync io callback function



On Fri, 27 Jun 2014, Joe Thornber wrote:

> On Fri, Jun 27, 2014 at 12:01:30PM +0800, Minfei Huang wrote:
> > The io address in callback function will become the danging point,
> > cause by the thread of sync io wakes up by other threads
> > and return to relieve the io address,
>
> Yes, well found. I prefer the following fix however.
>
> - Joe

It seems ok.

The patch is too big, I think the only change that needs to be done to fix
the bug is to replace "struct task_struct *sleeper;" with "struct
completion *completion;", replace "if (io->sleeper)
wake_up_process(io->sleeper);" with "if (io->completion)
complete(io->completion);" and declare the completion in sync_io() and
wait on it instead of "while (1)" loop there.

I suggest - split it to two patches, a minimal patch that fixes the bug by
changing sleeper to completion and the second patch with remaining changes
- so that only the first patch can be backported to stable kernels.

The smaller patch is less likely to produce conflicts (or bugs introduced
by incorrectly resolved conflicts) when being backported.

Mikulas


> Author: Joe Thornber <[email protected]>
> Date: Fri Jun 27 15:49:29 2014 +0100
>
> [dm-io] Fix a race condition in the wake up code for sync_io
>
> There's a race condition between the atomic_dec_and_test(&io->count)
> in dec_count() and the waking of the sync_io() thread. If the thread
> is spuriously woken immediately after the decrement it may exit,
> making the on the stack io struct invalid, yet the dec_count could
> still be using it.
>
> There are smaller fixes than the one here (eg, just take the io object
> off the stack). But I feel this code could use a clean up.
>
> - simplify dec_count().
>
> - It always calls a callback fn now.
> - It always frees the io object back to the pool.
>
> - sync_io()
>
> - Take the io object off the stack and allocate it from the pool the
> same as async_io.
> - Use a completion object rather than an explicit io_schedule()
> loop. The callback triggers the completion.
>
> Reported by: Minfei Huang
>
> diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
> index 3842ac7..a0982e81 100644
> --- a/drivers/md/dm-io.c
> +++ b/drivers/md/dm-io.c
> @@ -10,6 +10,7 @@
> #include <linux/device-mapper.h>
>
> #include <linux/bio.h>
> +#include <linux/completion.h>
> #include <linux/mempool.h>
> #include <linux/module.h>
> #include <linux/sched.h>
> @@ -32,7 +33,6 @@ struct dm_io_client {
> struct io {
> unsigned long error_bits;
> atomic_t count;
> - struct task_struct *sleeper;
> struct dm_io_client *client;
> io_notify_fn callback;
> void *context;
> @@ -111,28 +111,27 @@ static void retrieve_io_and_region_from_bio(struct bio *bio, struct io **io,
> * We need an io object to keep track of the number of bios that
> * have been dispatched for a particular io.
> *---------------------------------------------------------------*/
> -static void dec_count(struct io *io, unsigned int region, int error)
> +static void complete_io(struct io *io)
> {
> - if (error)
> - set_bit(region, &io->error_bits);
> + unsigned long error_bits = io->error_bits;
> + io_notify_fn fn = io->callback;
> + void *context = io->context;
>
> - if (atomic_dec_and_test(&io->count)) {
> - if (io->vma_invalidate_size)
> - invalidate_kernel_vmap_range(io->vma_invalidate_address,
> - io->vma_invalidate_size);
> + if (io->vma_invalidate_size)
> + invalidate_kernel_vmap_range(io->vma_invalidate_address,
> + io->vma_invalidate_size);
>
> - if (io->sleeper)
> - wake_up_process(io->sleeper);
> + mempool_free(io, io->client->pool);
> + fn(error_bits, context);
> +}
>
> - else {
> - unsigned long r = io->error_bits;
> - io_notify_fn fn = io->callback;
> - void *context = io->context;
> +static void dec_count(struct io *io, unsigned int region, int error)
> +{
> + if (error)
> + set_bit(region, &io->error_bits);
>
> - mempool_free(io, io->client->pool);
> - fn(r, context);
> - }
> - }
> + if (atomic_dec_and_test(&io->count))
> + complete_io(io);
> }
>
> static void endio(struct bio *bio, int error)
> @@ -375,48 +374,49 @@ static void dispatch_io(int rw, unsigned int num_regions,
> dec_count(io, 0, 0);
> }
>
> +struct sync_io {
> + unsigned long error_bits;
> + struct completion complete;
> +};
> +
> +static void sync_complete(unsigned long error, void *context)
> +{
> + struct sync_io *sio = context;
> + sio->error_bits = error;
> + complete(&sio->complete);
> +}
> +
> static int sync_io(struct dm_io_client *client, unsigned int num_regions,
> struct dm_io_region *where, int rw, struct dpages *dp,
> unsigned long *error_bits)
> {
> - /*
> - * gcc <= 4.3 can't do the alignment for stack variables, so we must
> - * align it on our own.
> - * volatile prevents the optimizer from removing or reusing
> - * "io_" field from the stack frame (allowed in ANSI C).
> - */
> - volatile char io_[sizeof(struct io) + __alignof__(struct io) - 1];
> - struct io *io = (struct io *)PTR_ALIGN(&io_, __alignof__(struct io));
> + struct io *io;
> + struct sync_io sio;
>
> if (num_regions > 1 && (rw & RW_MASK) != WRITE) {
> WARN_ON(1);
> return -EIO;
> }
>
> + init_completion(&sio.complete);
> +
> + io = mempool_alloc(client->pool, GFP_NOIO);
> io->error_bits = 0;
> atomic_set(&io->count, 1); /* see dispatch_io() */
> - io->sleeper = current;
> + io->callback = sync_complete;
> + io->context = &sio;
> io->client = client;
>
> io->vma_invalidate_address = dp->vma_invalidate_address;
> io->vma_invalidate_size = dp->vma_invalidate_size;
>
> dispatch_io(rw, num_regions, where, dp, io, 1);
> -
> - while (1) {
> - set_current_state(TASK_UNINTERRUPTIBLE);
> -
> - if (!atomic_read(&io->count))
> - break;
> -
> - io_schedule();
> - }
> - set_current_state(TASK_RUNNING);
> + wait_for_completion_io(&sio.complete);
>
> if (error_bits)
> - *error_bits = io->error_bits;
> + *error_bits = sio.error_bits;
>
> - return io->error_bits ? -EIO : 0;
> + return sio.error_bits ? -EIO : 0;
> }
>
> static int async_io(struct dm_io_client *client, unsigned int num_regions,
> @@ -434,7 +434,6 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
> io = mempool_alloc(client->pool, GFP_NOIO);
> io->error_bits = 0;
> atomic_set(&io->count, 1); /* see dispatch_io() */
> - io->sleeper = NULL;
> io->client = client;
> io->callback = fn;
> io->context = context;
>
> --
> dm-devel mailing list
> [email protected]
> https://www.redhat.com/mailman/listinfo/dm-devel
>

2014-06-27 19:29:16

by Mike Snitzer

[permalink] [raw]
Subject: Re: dm-io: Prevent the danging point of the sync io callback function

On Fri, Jun 27 2014 at 2:44pm -0400,
Mikulas Patocka <[email protected]> wrote:

>
>
> On Fri, 27 Jun 2014, Joe Thornber wrote:
>
> > On Fri, Jun 27, 2014 at 12:01:30PM +0800, Minfei Huang wrote:
> > > The io address in callback function will become the danging point,
> > > cause by the thread of sync io wakes up by other threads
> > > and return to relieve the io address,
> >
> > Yes, well found. I prefer the following fix however.
> >
> > - Joe
>
> It seems ok.
>
> The patch is too big, I think the only change that needs to be done to fix
> the bug is to replace "struct task_struct *sleeper;" with "struct
> completion *completion;", replace "if (io->sleeper)
> wake_up_process(io->sleeper);" with "if (io->completion)
> complete(io->completion);" and declare the completion in sync_io() and
> wait on it instead of "while (1)" loop there.

Here is the minimalist fix you suggested (I agree that splitting out a
minimalist fix is useful):

drivers/md/dm-io.c | 22 ++++++++--------------
1 file changed, 8 insertions(+), 14 deletions(-)

diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
index 2a20986..e60c2ea 100644
--- a/drivers/md/dm-io.c
+++ b/drivers/md/dm-io.c
@@ -10,6 +10,7 @@
#include <linux/device-mapper.h>

#include <linux/bio.h>
+#include <linux/completion.h>
#include <linux/mempool.h>
#include <linux/module.h>
#include <linux/sched.h>
@@ -32,7 +33,7 @@ struct dm_io_client {
struct io {
unsigned long error_bits;
atomic_t count;
- struct task_struct *sleeper;
+ struct completion *wait;
struct dm_io_client *client;
io_notify_fn callback;
void *context;
@@ -121,8 +122,8 @@ static void dec_count(struct io *io, unsigned int region, int error)
invalidate_kernel_vmap_range(io->vma_invalidate_address,
io->vma_invalidate_size);

- if (io->sleeper)
- wake_up_process(io->sleeper);
+ if (io->wait)
+ complete(io->wait);

else {
unsigned long r = io->error_bits;
@@ -385,6 +386,7 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
*/
volatile char io_[sizeof(struct io) + __alignof__(struct io) - 1];
struct io *io = (struct io *)PTR_ALIGN(&io_, __alignof__(struct io));
+ DECLARE_COMPLETION_ONSTACK(wait);

if (num_regions > 1 && (rw & RW_MASK) != WRITE) {
WARN_ON(1);
@@ -393,7 +395,7 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,

io->error_bits = 0;
atomic_set(&io->count, 1); /* see dispatch_io() */
- io->sleeper = current;
+ io->wait = &wait;
io->client = client;

io->vma_invalidate_address = dp->vma_invalidate_address;
@@ -401,15 +403,7 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,

dispatch_io(rw, num_regions, where, dp, io, 1);

- while (1) {
- set_current_state(TASK_UNINTERRUPTIBLE);
-
- if (!atomic_read(&io->count))
- break;
-
- io_schedule();
- }
- set_current_state(TASK_RUNNING);
+ wait_for_completion_io(&wait);

if (error_bits)
*error_bits = io->error_bits;
@@ -432,7 +426,7 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
io = mempool_alloc(client->pool, GFP_NOIO);
io->error_bits = 0;
atomic_set(&io->count, 1); /* see dispatch_io() */
- io->sleeper = NULL;
+ io->wait = NULL;
io->client = client;
io->callback = fn;
io->context = context;

2014-06-27 20:48:06

by Mikulas Patocka

[permalink] [raw]
Subject: Re: [dm-devel] [PATCH] dm-io: Prevent the danging point of the sync io callback function



On Fri, 27 Jun 2014, Mikulas Patocka wrote:

>
>
> On Fri, 27 Jun 2014, Minfei Huang wrote:
>
> > BUG: unable to handle kernel NULL pointer dereference at 0000000000000046
> > IP: [<ffffffffa0009cef>] dec_count+0x5f/0x80 [dm_mod]
> > PGD 0
> > Oops: 0000 [#1] SMP
> > last sysfs file: /sys/devices/pci0000:00/0000:00:02.2/0000:02:00.0/host0/scsi_host/host0/proc_name
> >
> > Pid: 2708, comm: kcopyd Tainted: G W --------------- H 2.6.32-279.19.5.el6.ucloud.x86_64 #1 Dell Inc. PowerEdge R720xd/0DCWD1
> > RIP: 0010:[<ffffffffa0009cef>] [<ffffffffa0009cef>] dec_count+0x5f/0x80 [dm_mod]
> > RSP: 0018:ffff880100603c30 EFLAGS: 00010246
> > RAX: 0000000000000046 RBX: ffff8817968a5c30 RCX: 0000000000000000
> > RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8817968a5c00
> > RBP: ffff880100603c50 R08: 0000000000000000 R09: 0000000000000000
> > R10: ffff880caa594cc0 R11: 0000000000000000 R12: ffff8817968a5c80
> > R13: ffffffff81013963 R14: 0000000000001000 R15: 0000000000000000
> > FS: 0000000000000000(0000) GS:ffff880100600000(0000) knlGS:0000000000000000
> > CS: 0010 DS: 0018 ES: 0018 CR0: 000000008005003b
> > CR2: 0000000000000046 CR3: 000000020c309000 CR4: 00000000001426e0
> > DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> > DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> > Process kcopyd (pid: 2708, threadinfo ffff88180cd26000, task ffff881841c9aa80)
> > Stack:
> > ffff880100603c40 ffff880aa8b32300 0000000000000000 ffff8817968a5c00
> > <d> ffff880100603c80 ffffffffa000a12a 0000000000000000 ffff880aa8b32300
> > <d> 0000000000000000 ffff880caa594cc0 ffff880100603c90 ffffffff811bcf6d
> > Call Trace:
> > <IRQ>
> > [<ffffffffa000a12a>] endio+0x4a/0x70 [dm_mod]
> > [<ffffffff811bcf6d>] bio_endio+0x1d/0x40
> > [<ffffffff81260beb>] req_bio_endio+0x9b/0xe0
> > [<ffffffff81263114>] blk_update_request+0x104/0x500
> > [<ffffffff81263331>] ? blk_update_request+0x321/0x500
> > [<ffffffff81263537>] blk_update_bidi_request+0x27/0xa0
> > [<ffffffff8126419f>] blk_end_bidi_request+0x2f/0x80
> > [<ffffffff81264240>] blk_end_request+0x10/0x20
> > [<ffffffff81375c6f>] scsi_io_completion+0xaf/0x6c0
> > [<ffffffff8136cb92>] scsi_finish_command+0xc2/0x130
> > [<ffffffff813763e5>] scsi_softirq_done+0x145/0x170
> > [<ffffffff812698ed>] blk_done_softirq+0x8d/0xa0
> > [<ffffffff81074c5f>] __do_softirq+0xdf/0x210
> > [<ffffffff8100c2cc>] call_softirq+0x1c/0x30
> > [<ffffffff8100df9d>] do_softirq+0xad/0xe0
> > [<ffffffff81074995>] irq_exit+0x95/0xa0
> > [<ffffffff81510515>] do_IRQ+0x75/0xf0
> > [<ffffffff8100ba53>] ret_from_intr+0x0/0x16
> >
> > The value of rdi register(0xffff8817968a5c00) is the io pointer,
> > If the sync io, the address of io point must be alloc from stack.
> > SO
> > crash> struct thread_info ffff8817968a4000
> > struct thread_info {
> > task = 0xffff88180cd9a580,
> > exec_domain = 0xffffffff81a98ac0,
> > ...
> > }
> >
> > crash> struct task_struct 0xffff88180cd9a580
> > struct task_struct {
> > state = 2,
> > stack = 0xffff8817968a4000,
> > ...
> > }
> >
> > It shows value exactly when use the value of io address.
> >
> > The io address in callback function will become the danging point,
> > cause by the thread of sync io wakes up by other threads
> > and return to relieve the io address,
> >
> > Signed-off-by: Minfei Huang <[email protected]>
> > ---
> > drivers/md/dm-io.c | 19 +++++++++++++++----
> > 1 files changed, 15 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
> > index 3842ac7..f992913 100644
> > --- a/drivers/md/dm-io.c
> > +++ b/drivers/md/dm-io.c
> > @@ -38,6 +38,7 @@ struct io {
> > void *context;
> > void *vma_invalidate_address;
> > unsigned long vma_invalidate_size;
> > + atomic_t wakeup;
> > } __attribute__((aligned(DM_IO_MAX_REGIONS)));
> >
> > static struct kmem_cache *_dm_io_cache;
> > @@ -121,10 +122,16 @@ static void dec_count(struct io *io, unsigned int region, int error)
> > invalidate_kernel_vmap_range(io->vma_invalidate_address,
> > io->vma_invalidate_size);
> >
> > - if (io->sleeper)
> > - wake_up_process(io->sleeper);
> > + if (io->sleeper) {
> > + struct task_struct *sleeper = io->sleeper;
> >
> > - else {
> > + atomic_set(&io->wakeup, 1);
>
> The problem here is that the processor may reorder the read of io->sleeper
> with atomic_set(&io->wakeup, 1); (performing atomic_set first and "sleeper
> = io->sleeper" afterwards) exposing the same race condition.
>
> You need to use memory barriers to avoid reordering, but I think the
> solution with the completion is better (the completion takes care of
> barriers automatically).
>
> Mikulas

Also - there is another race - after atomic_set(&io->wakeup, 1), the
target process may terminate, so wake_up_process(sleeper) operates on
non-existing process. You need to declare a special wait queue or use
wait_on_atomic_t+wake_up_atomic_t (that uses uses pre-initialized hash of
wait queues) to avoid that race.

But as I said - the completion approach is better. It doesn't suffer from
these problems.

Mikulas

> > +/*
> > + * The thread may be waked up by other threads,
> > + * if then the sync io point will become the dangling pointer
> > + */
> > + wake_up_process(sleeper);
> > + } else {
> > unsigned long r = io->error_bits;
> > io_notify_fn fn = io->callback;
> > void *context = io->context;
> > @@ -401,12 +408,14 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
> > io->vma_invalidate_address = dp->vma_invalidate_address;
> > io->vma_invalidate_size = dp->vma_invalidate_size;
> >
> > + atomic_set(&io->wakeup, 0);
> > +
> > dispatch_io(rw, num_regions, where, dp, io, 1);
> >
> > while (1) {
> > set_current_state(TASK_UNINTERRUPTIBLE);
> >
> > - if (!atomic_read(&io->count))
> > + if (atomic_read(&io->wakeup))
> > break;
> >
> > io_schedule();
> > @@ -442,6 +451,8 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
> > io->vma_invalidate_address = dp->vma_invalidate_address;
> > io->vma_invalidate_size = dp->vma_invalidate_size;
> >
> > + atomic_set(&io->wakeup, 0);
> > +
> > dispatch_io(rw, num_regions, where, dp, io, 0);
> > return 0;
> > }
> > --
> > 1.7.1
> >
> >
> > --
> > dm-devel mailing list
> > [email protected]
> > https://www.redhat.com/mailman/listinfo/dm-devel
> >
>

2014-06-27 20:57:12

by Mikulas Patocka

[permalink] [raw]
Subject: Re: dm-io: Prevent the danging point of the sync io callback function



On Fri, 27 Jun 2014, Mike Snitzer wrote:

> On Fri, Jun 27 2014 at 2:44pm -0400,
> Mikulas Patocka <[email protected]> wrote:
>
> >
> >
> > On Fri, 27 Jun 2014, Joe Thornber wrote:
> >
> > > On Fri, Jun 27, 2014 at 12:01:30PM +0800, Minfei Huang wrote:
> > > > The io address in callback function will become the danging point,
> > > > cause by the thread of sync io wakes up by other threads
> > > > and return to relieve the io address,
> > >
> > > Yes, well found. I prefer the following fix however.
> > >
> > > - Joe
> >
> > It seems ok.
> >
> > The patch is too big, I think the only change that needs to be done to fix
> > the bug is to replace "struct task_struct *sleeper;" with "struct
> > completion *completion;", replace "if (io->sleeper)
> > wake_up_process(io->sleeper);" with "if (io->completion)
> > complete(io->completion);" and declare the completion in sync_io() and
> > wait on it instead of "while (1)" loop there.
>
> Here is the minimalist fix you suggested (I agree that splitting out a
> minimalist fix is useful):

Acked-by: Mikulas Patocka <[email protected]>

> drivers/md/dm-io.c | 22 ++++++++--------------
> 1 file changed, 8 insertions(+), 14 deletions(-)
>
> diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c
> index 2a20986..e60c2ea 100644
> --- a/drivers/md/dm-io.c
> +++ b/drivers/md/dm-io.c
> @@ -10,6 +10,7 @@
> #include <linux/device-mapper.h>
>
> #include <linux/bio.h>
> +#include <linux/completion.h>
> #include <linux/mempool.h>
> #include <linux/module.h>
> #include <linux/sched.h>
> @@ -32,7 +33,7 @@ struct dm_io_client {
> struct io {
> unsigned long error_bits;
> atomic_t count;
> - struct task_struct *sleeper;
> + struct completion *wait;
> struct dm_io_client *client;
> io_notify_fn callback;
> void *context;
> @@ -121,8 +122,8 @@ static void dec_count(struct io *io, unsigned int region, int error)
> invalidate_kernel_vmap_range(io->vma_invalidate_address,
> io->vma_invalidate_size);
>
> - if (io->sleeper)
> - wake_up_process(io->sleeper);
> + if (io->wait)
> + complete(io->wait);
>
> else {
> unsigned long r = io->error_bits;
> @@ -385,6 +386,7 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
> */
> volatile char io_[sizeof(struct io) + __alignof__(struct io) - 1];
> struct io *io = (struct io *)PTR_ALIGN(&io_, __alignof__(struct io));
> + DECLARE_COMPLETION_ONSTACK(wait);
>
> if (num_regions > 1 && (rw & RW_MASK) != WRITE) {
> WARN_ON(1);
> @@ -393,7 +395,7 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
>
> io->error_bits = 0;
> atomic_set(&io->count, 1); /* see dispatch_io() */
> - io->sleeper = current;
> + io->wait = &wait;
> io->client = client;
>
> io->vma_invalidate_address = dp->vma_invalidate_address;
> @@ -401,15 +403,7 @@ static int sync_io(struct dm_io_client *client, unsigned int num_regions,
>
> dispatch_io(rw, num_regions, where, dp, io, 1);
>
> - while (1) {
> - set_current_state(TASK_UNINTERRUPTIBLE);
> -
> - if (!atomic_read(&io->count))
> - break;
> -
> - io_schedule();
> - }
> - set_current_state(TASK_RUNNING);
> + wait_for_completion_io(&wait);
>
> if (error_bits)
> *error_bits = io->error_bits;
> @@ -432,7 +426,7 @@ static int async_io(struct dm_io_client *client, unsigned int num_regions,
> io = mempool_alloc(client->pool, GFP_NOIO);
> io->error_bits = 0;
> atomic_set(&io->count, 1); /* see dispatch_io() */
> - io->sleeper = NULL;
> + io->wait = NULL;
> io->client = client;
> io->callback = fn;
> io->context = context;
>

2014-06-30 08:22:07

by Joe Thornber

[permalink] [raw]
Subject: Re: [dm-devel] [PATCH] dm-io: Prevent the danging point of the sync io callback function

On Fri, Jun 27, 2014 at 02:44:41PM -0400, Mikulas Patocka wrote:
> I suggest - split it to two patches, a minimal patch that fixes the bug by
> changing sleeper to completion and the second patch with remaining changes
> - so that only the first patch can be backported to stable kernels.

Agreed. Thanks for reviewing.