The Linux Kernel Mailing List
 help / color / mirror / Atom feed
* [PATCH] net: liquidio: fix BAR resource leak on PF number failure
@ 2026-06-20  8:37 Haoxiang Li
  2026-06-22 13:48 ` Simon Horman
  0 siblings, 1 reply; 2+ messages in thread
From: Haoxiang Li @ 2026-06-20  8:37 UTC (permalink / raw)
  To: andrew+netdev, davem, kuba, pabeni, felix.manlunas,
	ricardo.farrington
  Cc: netdev, linux-kernel, Haoxiang Li, stable

If cn23xx_get_pf_num() fails, the function returns without
unmapping either BAR. Unmap both BARs before returning from
the error path.

Fixes: 0c45d7fe12c7 ("liquidio: fix use of pf in pass-through mode in a virtual machine")
Cc: stable@vger.kernel.org
Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>
---
 drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
index 75f22f74774c..a1548ca81ecd 100644
--- a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
@@ -1167,8 +1167,11 @@ int setup_cn23xx_octeon_pf_device(struct octeon_device *oct)
 		return 1;
 	}
 
-	if (cn23xx_get_pf_num(oct) != 0)
+	if (cn23xx_get_pf_num(oct) != 0) {
+		octeon_unmap_pci_barx(oct, 0);
+		octeon_unmap_pci_barx(oct, 1);
 		return 1;
+	}
 
 	if (cn23xx_sriov_config(oct)) {
 		octeon_unmap_pci_barx(oct, 0);
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH] net: liquidio: fix BAR resource leak on PF number failure
  2026-06-20  8:37 [PATCH] net: liquidio: fix BAR resource leak on PF number failure Haoxiang Li
@ 2026-06-22 13:48 ` Simon Horman
  0 siblings, 0 replies; 2+ messages in thread
From: Simon Horman @ 2026-06-22 13:48 UTC (permalink / raw)
  To: Haoxiang Li
  Cc: andrew+netdev, davem, kuba, pabeni, felix.manlunas,
	ricardo.farrington, netdev, linux-kernel, stable

On Sat, Jun 20, 2026 at 04:37:28PM +0800, Haoxiang Li wrote:
> If cn23xx_get_pf_num() fails, the function returns without
> unmapping either BAR. Unmap both BARs before returning from
> the error path.

I think it would be worth noting how this problem was found,
and if a publicly available tool was used, naming it.

> 
> Fixes: 0c45d7fe12c7 ("liquidio: fix use of pf in pass-through mode in a virtual machine")
> Cc: stable@vger.kernel.org
> Signed-off-by: Haoxiang Li <haoxiang_li2024@163.com>

There is an AI-generated review of this patch available on sashiko.dev.
I don't think the issue raised there directly affects this patch.
But you may want to consider looking into that in the context of
a separate follow-up.

> ---
>  drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c | 5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
> index 75f22f74774c..a1548ca81ecd 100644
> --- a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
> +++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
> @@ -1167,8 +1167,11 @@ int setup_cn23xx_octeon_pf_device(struct octeon_device *oct)
>  		return 1;
>  	}
>  
> -	if (cn23xx_get_pf_num(oct) != 0)
> +	if (cn23xx_get_pf_num(oct) != 0) {
> +		octeon_unmap_pci_barx(oct, 0);
> +		octeon_unmap_pci_barx(oct, 1);
>  		return 1;
> +	}
>  
>  	if (cn23xx_sriov_config(oct)) {
>  		octeon_unmap_pci_barx(oct, 0);

I think this would be best handled by introducing an idiomatic goto unwind
ladder to this function. Something like this (compile tested only!):

diff --git a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
index 75f22f74774c..73362b92d0fd 100644
--- a/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
+++ b/drivers/net/ethernet/cavium/liquidio/cn23xx_pf_device.c
@@ -1163,18 +1163,14 @@ int setup_cn23xx_octeon_pf_device(struct octeon_device *oct)
 	if (octeon_map_pci_barx(oct, 1, MAX_BAR1_IOREMAP_SIZE)) {
 		dev_err(&oct->pci_dev->dev, "%s CN23XX BAR1 map failed\n",
 			__func__);
-		octeon_unmap_pci_barx(oct, 0);
-		return 1;
+		goto err_free_barx_0;
 	}
 
 	if (cn23xx_get_pf_num(oct) != 0)
-		return 1;
+		goto err_free_barx_1;
 
-	if (cn23xx_sriov_config(oct)) {
-		octeon_unmap_pci_barx(oct, 0);
-		octeon_unmap_pci_barx(oct, 1);
-		return 1;
-	}
+	if (cn23xx_sriov_config(oct))
+		goto err_free_barx_1;
 
 	octeon_write_csr64(oct, CN23XX_SLI_MAC_CREDIT_CNT, 0x3F802080802080ULL);
 
@@ -1205,6 +1201,12 @@ int setup_cn23xx_octeon_pf_device(struct octeon_device *oct)
 	oct->coproc_clock_rate = 1000000ULL * cn23xx_coprocessor_clock(oct);
 
 	return 0;
+
+err_free_barx_0:
+	octeon_unmap_pci_barx(oct, 0);
+err_free_barx_1:
+	octeon_unmap_pci_barx(oct, 1);
+	return 1;
 }
 EXPORT_SYMBOL_GPL(setup_cn23xx_octeon_pf_device);
 
-- 
pw-bot: changes-requested

^ permalink raw reply related	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2026-06-22 13:49 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-06-20  8:37 [PATCH] net: liquidio: fix BAR resource leak on PF number failure Haoxiang Li
2026-06-22 13:48 ` Simon Horman

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox