Re: [RFC PATCH v3] remoteproc: xlnx: initialize mailbox work before requesting channels

[Mathieu Poirier <mathieu.poirier@linaro.org>]

On Tue, Aug 18, 2026 at 06:12:01PM +0800, Runyu Xiao wrote:
> zynqmp_r5_setup_mbox() installs zynqmp_r5_mb_rx_cb() as the mailbox RX
> callback before requesting the mailbox channels, but initializes
> ipi->mbox_work only after both channels have been requested. Once the RX
> channel is active, a notification delivered before the late INIT_WORK()
> would make the callback queue an uninitialized work item.
> 
> Initialize the work item before requesting channels. Also drain the work
> before freeing the mailbox state, after the channels have been released so
> no new callbacks can queue it. Use the same drain before freeing ipi from
> the channel request error paths, since the callback is already published
> there as well.
> 
> This issue was found by our static analysis tool and then confirmed by
> manual review of the mailbox setup sequence. The callback is published
> before the channel requests complete, so the work item should be ready
> before the mailbox provider can invoke it.
> 
> A QEMU PoC modeled a mailbox notification delivered after the RX callback
> became reachable but before the delayed INIT_WORK(). DEBUG_OBJECTS reported
> queueing an uninitialized work item from the zynqmp_r5_setup_mbox() path.
> 
> This is sent as an RFC because the practical trigger depends on the ZynqMP
> IPI mailbox provider and firmware delivery timing. If the provider cannot
> invoke the RX callback until after setup returns, this is a defensive
> lifecycle cleanup rather than a reachable race on current systems.
> 
> Fixes: 5dfb28c257b7 ("remoteproc: xilinx: Add mailbox channels for rpmsg")
> Signed-off-by: Runyu Xiao <runyu.xiao@seu.edu.cn>
> ---
> Changes in v3:
> - Follow Mathieu's feedback and call cancel_work_sync() before freeing ipi
>   from the tx/rx channel request error paths.
> 
> Changes in v2:
> - Follow Tanmay's suggestion and keep zynqmp_r5_setup_mbox() taking the
>   child device pointer. Do not move the r5_core assignment in this patch.
> - Only move INIT_WORK() before channel requests and drain the work in
>   zynqmp_r5_free_mbox().
> 
>  drivers/remoteproc/xlnx_r5_remoteproc.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
>

I will pick this up when RC1 comes out next week.

Thanks,
Mathieu
 
> diff --git a/drivers/remoteproc/xlnx_r5_remoteproc.c b/drivers/remoteproc/xlnx_r5_remoteproc.c
> index c685bb5fa62c..d47fbf143acd 100644
> --- a/drivers/remoteproc/xlnx_r5_remoteproc.c
> +++ b/drivers/remoteproc/xlnx_r5_remoteproc.c
> @@ -318,6 +318,8 @@ static struct mbox_info *zynqmp_r5_setup_mbox(struct device *cdev)
>  	if (!ipi)
>  		return NULL;
>  
> +	INIT_WORK(&ipi->mbox_work, handle_event_notified);
> +
>  	mbox_cl = &ipi->mbox_cl;
>  	mbox_cl->rx_callback = zynqmp_r5_mb_rx_cb;
>  	mbox_cl->tx_block = false;
> @@ -329,6 +331,7 @@ static struct mbox_info *zynqmp_r5_setup_mbox(struct device *cdev)
>  	ipi->tx_chan = mbox_request_channel_byname(mbox_cl, "tx");
>  	if (IS_ERR(ipi->tx_chan)) {
>  		ipi->tx_chan = NULL;
> +		cancel_work_sync(&ipi->mbox_work);
>  		kfree(ipi);
>  		dev_warn(cdev, "mbox tx channel request failed\n");
>  		return NULL;
> @@ -339,13 +342,12 @@ static struct mbox_info *zynqmp_r5_setup_mbox(struct device *cdev)
>  		mbox_free_channel(ipi->tx_chan);
>  		ipi->rx_chan = NULL;
>  		ipi->tx_chan = NULL;
> +		cancel_work_sync(&ipi->mbox_work);
>  		kfree(ipi);
>  		dev_warn(cdev, "mbox rx channel request failed\n");
>  		return NULL;
>  	}
>  
> -	INIT_WORK(&ipi->mbox_work, handle_event_notified);
> -
>  	return ipi;
>  }
>  
> @@ -364,6 +366,8 @@ static void zynqmp_r5_free_mbox(struct mbox_info *ipi)
>  		ipi->rx_chan = NULL;
>  	}
>  
> +	cancel_work_sync(&ipi->mbox_work);
> +
>  	kfree(ipi);
>  }
>  
> -- 
> 2.34.1
>