2023-10-05 14:22:07

by Chengfeng Ye

Subject: [PATCH v2 2/2] atm: solos-pci: Fix potential deadlock on &tx_queue_lock

As &card->tx_queue_lock is acquired under softirq context along the
following call chain from solos_bh(), other acquisition of the same
lock inside process context should disable at least bh to avoid double
lock.

<deadlock #2>
pclose()
--> spin_lock(&card->tx_queue_lock)
<interrupt>
--> solos_bh()
--> fpga_tx()
--> spin_lock(&card->tx_queue_lock)

This flaw was found by an experimental static analysis tool I am
developing for irq-related deadlock.

To prevent the potential deadlock, the patch uses spin_lock_irqsave()
on &card->tx_queue_lock under process context code consistently to
prevent the possible deadlock scenario.

Fixes: 213e85d38912 ("solos-pci: clean up pclose() function")
Signed-off-by: Chengfeng Ye <[email protected]>
---
V2: add fix tag, and split into two patches

drivers/atm/solos-pci.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/atm/solos-pci.c b/drivers/atm/solos-pci.c
index 48cf9b36b61a..247e9200e312 100644
--- a/drivers/atm/solos-pci.c
+++ b/drivers/atm/solos-pci.c
@@ -955,16 +955,17 @@ static void pclose(struct atm_vcc *vcc)
unsigned char port = SOLOS_CHAN(vcc->dev);
struct sk_buff *skb, *tmpskb;
struct pkt_hdr *header;
+ unsigned long flags;

/* Remove any yet-to-be-transmitted packets from the pending queue */
- spin_lock(&card->tx_queue_lock);
+ spin_lock_irqsave(&card->tx_queue_lock, flags);
skb_queue_walk_safe(&card->tx_queue[port], skb, tmpskb) {
if (SKB_CB(skb)->vcc == vcc) {
skb_unlink(skb, &card->tx_queue[port]);
solos_pop(vcc, skb);
}
}
- spin_unlock(&card->tx_queue_lock);
+ spin_unlock_irqrestore(&card->tx_queue_lock, flags);

skb = alloc_skb(sizeof(*header), GFP_KERNEL);
if (!skb) {
--
2.17.1

2023-10-07 15:13:53

by Simon Horman

Subject: Re: [PATCH v2 2/2] atm: solos-pci: Fix potential deadlock on &tx_queue_lock

On Thu, Oct 05, 2023 at 07:49:17AM +0000, Chengfeng Ye wrote:
> As &card->tx_queue_lock is acquired under softirq context along the
> following call chain from solos_bh(), other acquisition of the same
> lock inside process context should disable at least bh to avoid double
> lock.
>
> <deadlock #2>
> pclose()
> --> spin_lock(&card->tx_queue_lock)
> <interrupt>
> --> solos_bh()
> --> fpga_tx()
> --> spin_lock(&card->tx_queue_lock)
>
> This flaw was found by an experimental static analysis tool I am
> developing for irq-related deadlock.
>
> To prevent the potential deadlock, the patch uses spin_lock_irqsave()
> on &card->tx_queue_lock under process context code consistently to
> prevent the possible deadlock scenario.
>
> Fixes: 213e85d38912 ("solos-pci: clean up pclose() function")
> Signed-off-by: Chengfeng Ye <[email protected]>
> ---
> V2: add fix tag, and split into two patches

Reviewed-by: Simon Horman <[email protected]>