Re: [PATCH-net-next] ptp: fix corrupted list in ptp_open

Re: [PATCH-net-next] ptp: fix corrupted list in ptp_open

[Richard Cochran]

On Sun, Oct 29, 2023 at 10:09:42AM +0800, Edward Adam Davis wrote:
> There is no lock protection when writing ptp->tsevqs in ptp_open(), ptp_read(),
> ptp_release(), which can cause data corruption and increase mutual exclusion
> to avoid this issue.

-ENOPARSE

How can lack of lock protection increase mutual exclusion?

> Moreover, the queue should not be released in ptp_read() and should be deleted
> together.

The queue should be deleted togther?  Huh?

> @@ -543,6 +552,8 @@ ssize_t ptp_read(struct posix_clock_context *pccontext, uint rdflags,
>  		cnt = EXTTS_BUFSIZE;
>  
>  	cnt = cnt / sizeof(struct ptp_extts_event);
> +	if (mutex_lock_interruptible(&ptp->tsevq_mux)) 
> +		return -ERESTARTSYS;

This is not needed because the spin lock (timestamp_event_queue::lock)
already protects the event queue.

Thanks,
Richard

Re: [PATCH-net-next] ptp: fix corrupted list in ptp_open

[Richard Cochran]

On Sun, Oct 29, 2023 at 10:09:42AM +0800, Edward Adam Davis wrote:

> @@ -585,7 +596,6 @@ ssize_t ptp_read(struct posix_clock_context *pccontext, uint rdflags,
>  free_event:
>  	kfree(event);
>  exit:
> -	if (result < 0)
> -		ptp_release(pccontext);
> +	mutex_unlock(&ptp->tsevq_mux);
>  	return result;
>  }

This is the only hunk that makes sense.  Keep this, but remove the
rest, just like in your previous patches.

Thanks,
Richard

Re: [PATCH net-next V2] ptp: fix corrupted list in ptp_open

[Martin Habets]

Please use a separate mail thread for a new patch revision.
See the section "Resending after review" in
Documentation/process/maintainer-netdev.rst.

Martin

On Tue, Oct 31, 2023 at 05:07:08AM +0800, Edward Adam Davis wrote:
> There is no lock protection when writing ptp->tsevqs in ptp_open(),
> ptp_release(), which can cause data corruption, use mutex lock to avoid this 
> issue.
> 
> Moreover, ptp_release() should not be used to release the queue in ptp_read(),
> and it should be deleted together.
> 
> Reported-and-tested-by: syzbot+df3f3ef31f60781fa911@syzkaller.appspotmail.com
> Fixes: 8f5de6fb2453 ("ptp: support multiple timestamp event readers")
> Signed-off-by: Edward Adam Davis <eadavis@qq.com>
> ---
>  drivers/ptp/ptp_chardev.c | 11 +++++++++--
>  drivers/ptp/ptp_clock.c   |  3 +++
>  drivers/ptp/ptp_private.h |  1 +
>  3 files changed, 13 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/ptp/ptp_chardev.c b/drivers/ptp/ptp_chardev.c
> index 282cd7d24077..e31551d2697d 100644
> --- a/drivers/ptp/ptp_chardev.c
> +++ b/drivers/ptp/ptp_chardev.c
> @@ -109,6 +109,9 @@ int ptp_open(struct posix_clock_context *pccontext, fmode_t fmode)
>  	struct timestamp_event_queue *queue;
>  	char debugfsname[32];
>  
> +	if (mutex_lock_interruptible(&ptp->tsevq_mux)) 
> +		return -ERESTARTSYS;
> +
>  	queue = kzalloc(sizeof(*queue), GFP_KERNEL);
>  	if (!queue)
>  		return -EINVAL;
> @@ -132,15 +135,20 @@ int ptp_open(struct posix_clock_context *pccontext, fmode_t fmode)
>  	debugfs_create_u32_array("mask", 0444, queue->debugfs_instance,
>  				 &queue->dfs_bitmap);
>  
> +	mutex_unlock(&ptp->tsevq_mux);
>  	return 0;
>  }
>  
>  int ptp_release(struct posix_clock_context *pccontext)
>  {
>  	struct timestamp_event_queue *queue = pccontext->private_clkdata;
> +	struct ptp_clock *ptp =
> +		container_of(pccontext->clk, struct ptp_clock, clock);
>  	unsigned long flags;
>  
>  	if (queue) {
> +		if (mutex_lock_interruptible(&ptp->tsevq_mux)) 
> +			return -ERESTARTSYS;
>  		debugfs_remove(queue->debugfs_instance);
>  		pccontext->private_clkdata = NULL;
>  		spin_lock_irqsave(&queue->lock, flags);
> @@ -148,6 +156,7 @@ int ptp_release(struct posix_clock_context *pccontext)
>  		spin_unlock_irqrestore(&queue->lock, flags);
>  		bitmap_free(queue->mask);
>  		kfree(queue);
> +		mutex_unlock(&ptp->tsevq_mux);
>  	}
>  	return 0;
>  }
> @@ -585,7 +594,5 @@ ssize_t ptp_read(struct posix_clock_context *pccontext, uint rdflags,
>  free_event:
>  	kfree(event);
>  exit:
> -	if (result < 0)
> -		ptp_release(pccontext);
>  	return result;
>  }
> diff --git a/drivers/ptp/ptp_clock.c b/drivers/ptp/ptp_clock.c
> index 3d1b0a97301c..7930db6ec18d 100644
> --- a/drivers/ptp/ptp_clock.c
> +++ b/drivers/ptp/ptp_clock.c
> @@ -176,6 +176,7 @@ static void ptp_clock_release(struct device *dev)
>  
>  	ptp_cleanup_pin_groups(ptp);
>  	kfree(ptp->vclock_index);
> +	mutex_destroy(&ptp->tsevq_mux);
>  	mutex_destroy(&ptp->pincfg_mux);
>  	mutex_destroy(&ptp->n_vclocks_mux);
>  	/* Delete first entry */
> @@ -247,6 +248,7 @@ struct ptp_clock *ptp_clock_register(struct ptp_clock_info *info,
>  	if (!queue)
>  		goto no_memory_queue;
>  	list_add_tail(&queue->qlist, &ptp->tsevqs);
> +	mutex_init(&ptp->tsevq_mux);
>  	queue->mask = bitmap_alloc(PTP_MAX_CHANNELS, GFP_KERNEL);
>  	if (!queue->mask)
>  		goto no_memory_bitmap;
> @@ -356,6 +358,7 @@ struct ptp_clock *ptp_clock_register(struct ptp_clock_info *info,
>  	if (ptp->kworker)
>  		kthread_destroy_worker(ptp->kworker);
>  kworker_err:
> +	mutex_destroy(&ptp->tsevq_mux);
>  	mutex_destroy(&ptp->pincfg_mux);
>  	mutex_destroy(&ptp->n_vclocks_mux);
>  	bitmap_free(queue->mask);
> diff --git a/drivers/ptp/ptp_private.h b/drivers/ptp/ptp_private.h
> index 52f87e394aa6..1525bd2059ba 100644
> --- a/drivers/ptp/ptp_private.h
> +++ b/drivers/ptp/ptp_private.h
> @@ -44,6 +44,7 @@ struct ptp_clock {
>  	struct pps_device *pps_source;
>  	long dialed_frequency; /* remembers the frequency adjustment */
>  	struct list_head tsevqs; /* timestamp fifo list */
> +	struct mutex tsevq_mux; /* one process at a time reading the fifo */
>  	struct mutex pincfg_mux; /* protect concurrent info->pin_config access */
>  	wait_queue_head_t tsev_wq;
>  	int defunct; /* tells readers to go away when clock is being removed */
> -- 
> 2.25.1
> 

Re: [PATCH net-next V2] ptp: fix corrupted list in ptp_open

[Richard Cochran]

On Tue, Oct 31, 2023 at 05:07:08AM +0800, Edward Adam Davis wrote:
> There is no lock protection when writing ptp->tsevqs in ptp_open(),
> ptp_release(), which can cause data corruption,

Really?  How?

> use mutex lock to avoid this 
> issue.
> 
> Moreover, ptp_release() should not be used to release the queue in ptp_read(),
> and it should be deleted together.
> 
> Reported-and-tested-by: syzbot+df3f3ef31f60781fa911@syzkaller.appspotmail.com
> Fixes: 8f5de6fb2453 ("ptp: support multiple timestamp event readers")
> Signed-off-by: Edward Adam Davis <eadavis@qq.com>
> ---
>  drivers/ptp/ptp_chardev.c | 11 +++++++++--
>  drivers/ptp/ptp_clock.c   |  3 +++
>  drivers/ptp/ptp_private.h |  1 +
>  3 files changed, 13 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/ptp/ptp_chardev.c b/drivers/ptp/ptp_chardev.c
> index 282cd7d24077..e31551d2697d 100644
> --- a/drivers/ptp/ptp_chardev.c
> +++ b/drivers/ptp/ptp_chardev.c
> @@ -109,6 +109,9 @@ int ptp_open(struct posix_clock_context *pccontext, fmode_t fmode)
>  	struct timestamp_event_queue *queue;
>  	char debugfsname[32];
>  
> +	if (mutex_lock_interruptible(&ptp->tsevq_mux)) 
> +		return -ERESTARTSYS;
> +

This mutex is not needed.

Please don't ignore review comments.

Thanks,
Richard

Re: [PATCH net-next V2] ptp: fix corrupted list in ptp_open

[Richard Cochran]

On Thu, Nov 02, 2023 at 07:16:17PM +0800, Edward Adam Davis wrote:
> The above two logs can clearly indicate that there is corruption when 
> executing the operation of writing ptp->tsevqs in ptp_open() and ptp_release().

So just remove the bogus call to ptp_release.

Thanks,
Richard