summaryrefslogtreecommitdiffstats
path: root/drivers/atm
diff options
context:
space:
mode:
authorChengfeng Ye <dg573847474@gmail.com>2023-12-07 13:34:37 +0100
committerDavid S. Miller <davem@davemloft.net>2023-12-09 22:17:45 +0100
commitd5dba32b8f6cb39be708b726044ba30dbc088b30 (patch)
tree4fdc081a10e22cf42a9ae95a7e55fbfac576e988 /drivers/atm
parentMerge branch 'bnxt_en-misc-fixes' (diff)
downloadlinux-d5dba32b8f6cb39be708b726044ba30dbc088b30.tar.xz
linux-d5dba32b8f6cb39be708b726044ba30dbc088b30.zip
atm: solos-pci: Fix potential deadlock on &cli_queue_lock
As &card->cli_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 #1> console_show() --> spin_lock(&card->cli_queue_lock) <interrupt> --> solos_bh() --> spin_lock(&card->cli_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_bh() on the card->cli_queue_lock under process context code consistently to prevent the possible deadlock scenario. Fixes: 9c54004ea717 ("atm: Driver for Solos PCI ADSL2+ card.") Signed-off-by: Chengfeng Ye <dg573847474@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'drivers/atm')
-rw-r--r--drivers/atm/solos-pci.c4
1 files changed, 2 insertions, 2 deletions
diff --git a/drivers/atm/solos-pci.c b/drivers/atm/solos-pci.c
index 94fbc3abe60e..95f768b28a5e 100644
--- a/drivers/atm/solos-pci.c
+++ b/drivers/atm/solos-pci.c
@@ -449,9 +449,9 @@ static ssize_t console_show(struct device *dev, struct device_attribute *attr,
struct sk_buff *skb;
unsigned int len;
- spin_lock(&card->cli_queue_lock);
+ spin_lock_bh(&card->cli_queue_lock);
skb = skb_dequeue(&card->cli_queue[SOLOS_CHAN(atmdev)]);
- spin_unlock(&card->cli_queue_lock);
+ spin_unlock_bh(&card->cli_queue_lock);
if(skb == NULL)
return sprintf(buf, "No data.\n");