Re: [RFC net-next 13/15] eth: fbnic: Support AF_XDP zero-copy receive

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] net: deadlock on 32-bit during u64_stats_sync update with preemption enabled
--

commit 1bc54e591deb70184971cd19a4b1a9d18715cb4e
Author: Björn Töpel <bjorn@kernel.org>

eth: fbnic: Support AF_XDP zero-copy receive

This commit adds support for AF_XDP zero-copy receive in the fbnic driver.
It integrates the XSK buffer pool as a memory provider for page_pool and
refactors the receive ring refill loops to balance header and payload buffers.

> diff --git a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c
> --- a/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c
> +++ b/drivers/net/ethernet/meta/fbnic/fbnic_txrx.c
[ ... ]
> -static void fbnic_fill_bdq(struct fbnic_ring *bdq)
> +static bool fbnic_fill_bdq_one(struct fbnic_ring *bdq, unsigned int *tail)
>  {
> -	unsigned int count = fbnic_desc_unused(bdq);
> -	unsigned int i = bdq->tail;
> +	netmem_ref netmem;
>  
> -	if (!count)
> -		return;
> +	netmem = page_pool_dev_alloc_netmems(bdq->page_pool);
> +	if (!netmem) {
> +		u64_stats_update_begin(&bdq->stats.syncp);

[Severity: High]
Does this code risk a deadlock on 32-bit systems when called from process
context with preemption enabled?

If memory allocation fails here during interface bring-up (e.g., via
ndo_open -> fbnic_up() -> fbnic_fill() -> fbnic_nv_fill() -> fbnic_fill_bdqs()
-> fbnic_fill_bdq_one()), the sequence counter is updated without disabling
preemption.

If the thread is preempted while the sequence counter is odd, and another
thread attempts to read network statistics (like ndo_get_stats64 or
get_queue_stats_rx), the reader will spin forever preventing the writer
from finishing, leading to a hard deadlock.

> +		bdq->stats.bdq.alloc_failed++;
> +		u64_stats_update_end(&bdq->stats.syncp);
> +		return false;
> +	}

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261002190018.696925-1-bjorn@kernel.org?part=13