2018-07-30 06:11:41

by Xiao, Jin

[permalink] [raw]
Subject: [PATCH] block: blk_init_allocated_queue() set q->fq as NULL in the fail case

We find the memory use-after-free issue in __blk_drain_queue()
on the kernel 4.14. After read the latest kernel 4.18-rc6 we
think it has the same problem.

Memory is allocated for q->fq in the blk_init_allocated_queue().
If the elevator init function called with error return, it will
run into the fail case to free the q->fq.

Then the __blk_drain_queue() uses the same memory after the free
of the q->fq, it will lead to the unpredictable event.

The patch is to set q->fq as NULL in the fail case of
blk_init_allocated_queue().

Fixes: commit 7c94e1c157a2 ("block: introduce blk_flush_queue to drive flush machinery")
Signed-off-by: xiao jin <[email protected]>
Cc: Ming Lei <[email protected]>
Cc: Bart Van Assche <[email protected]>
Cc: <[email protected]>
---
block/blk-core.c | 1 +
1 file changed, 1 insertion(+)

diff --git a/block/blk-core.c b/block/blk-core.c
index b888175..52635e2 100644
--- a/block/blk-core.c
+++ b/block/blk-core.c
@@ -1072,6 +1072,7 @@ int blk_init_allocated_queue(struct request_queue *q)
q->exit_rq_fn(q, q->fq->flush_rq);
out_free_flush_queue:
blk_free_flush_queue(q->fq);
+ q->fq = NULL;
return -ENOMEM;
}
EXPORT_SYMBOL(blk_init_allocated_queue);
--
1.7.9.5



2018-07-30 12:07:49

by Ming Lei

[permalink] [raw]
Subject: Re: [PATCH] block: blk_init_allocated_queue() set q->fq as NULL in the fail case

On Mon, Jul 30, 2018 at 2:11 PM, xiao jin <[email protected]> wrote:
> We find the memory use-after-free issue in __blk_drain_queue()
> on the kernel 4.14. After read the latest kernel 4.18-rc6 we
> think it has the same problem.
>
> Memory is allocated for q->fq in the blk_init_allocated_queue().
> If the elevator init function called with error return, it will
> run into the fail case to free the q->fq.
>
> Then the __blk_drain_queue() uses the same memory after the free
> of the q->fq, it will lead to the unpredictable event.
>
> The patch is to set q->fq as NULL in the fail case of
> blk_init_allocated_queue().
>
> Fixes: commit 7c94e1c157a2 ("block: introduce blk_flush_queue to drive flush machinery")
> Signed-off-by: xiao jin <[email protected]>
> Cc: Ming Lei <[email protected]>
> Cc: Bart Van Assche <[email protected]>
> Cc: <[email protected]>
> ---
> block/blk-core.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/block/blk-core.c b/block/blk-core.c
> index b888175..52635e2 100644
> --- a/block/blk-core.c
> +++ b/block/blk-core.c
> @@ -1072,6 +1072,7 @@ int blk_init_allocated_queue(struct request_queue *q)
> q->exit_rq_fn(q, q->fq->flush_rq);
> out_free_flush_queue:
> blk_free_flush_queue(q->fq);
> + q->fq = NULL;
> return -ENOMEM;
> }
> EXPORT_SYMBOL(blk_init_allocated_queue);
> --
> 1.7.9.5
>

Reviewed-by: Ming Lei <[email protected]>

Thanks,
Ming Lei

2018-07-30 13:43:18

by Bart Van Assche

[permalink] [raw]
Subject: Re: [PATCH] block: blk_init_allocated_queue() set q->fq as NULL in the fail case

On Mon, 2018-07-30 at 14:11 +-0800, xiao jin wrote:
+AD4- We find the memory use-after-free issue in +AF8AXw-blk+AF8-drain+AF8-queue()
+AD4- on the kernel 4.14. After read the latest kernel 4.18-rc6 we
+AD4- think it has the same problem.
+AD4-
+AD4- Memory is allocated for q-+AD4-fq in the blk+AF8-init+AF8-allocated+AF8-queue().
+AD4- If the elevator init function called with error return, it will
+AD4- run into the fail case to free the q-+AD4-fq.
+AD4-
+AD4- Then the +AF8AXw-blk+AF8-drain+AF8-queue() uses the same memory after the free
+AD4- of the q-+AD4-fq, it will lead to the unpredictable event.
+AD4-
+AD4- The patch is to set q-+AD4-fq as NULL in the fail case of
+AD4- blk+AF8-init+AF8-allocated+AF8-queue().

Reviewed-by: Bart Van Assche +ADw-bart.vanassche+AEA-wdc.com+AD4-



2018-07-30 14:30:59

by Jens Axboe

[permalink] [raw]
Subject: Re: [PATCH] block: blk_init_allocated_queue() set q->fq as NULL in the fail case

On 7/30/18 12:11 AM, xiao jin wrote:
> We find the memory use-after-free issue in __blk_drain_queue()
> on the kernel 4.14. After read the latest kernel 4.18-rc6 we
> think it has the same problem.
>
> Memory is allocated for q->fq in the blk_init_allocated_queue().
> If the elevator init function called with error return, it will
> run into the fail case to free the q->fq.
>
> Then the __blk_drain_queue() uses the same memory after the free
> of the q->fq, it will lead to the unpredictable event.
>
> The patch is to set q->fq as NULL in the fail case of
> blk_init_allocated_queue().

Applied, thanks.

--
Jens Axboe