[PATCH] net: macb: fix ignored return value of clk_prepare_enable() in runtime resume

[PATCH] net: macb: fix ignored return value of clk_prepare_enable() in runtime resume

[Gustavo Kenji Mendonça Kaneko]

macb_runtime_resume() calls clk_prepare_enable() five times without
checking the return value. If any clock fails to enable, the driver
silently continues with potentially unclocked hardware, which can
cause undefined behavior or system instability on resume from runtime
suspend.

The existing macb_clks_disable() already uses clk_bulk_disable_unprepare()
for the symmetric disable path. Align the enable path by using
clk_bulk_prepare_enable(), which atomically enables all clocks and
automatically rolls back any successfully enabled clocks on failure.

Signed-off-by: Gustavo Kenji Mendonça Kaneko <kaneko.dev@pm.me>
---
 drivers/net/ethernet/cadence/macb_main.c | 21 ++++++++++++---------
 1 file changed, 12 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/cadence/macb_main.c b/drivers/net/ethernet/cadence/macb_main.c
index a12aa21244e8..85bd4ff0e4e6 100644
--- a/drivers/net/ethernet/cadence/macb_main.c
+++ b/drivers/net/ethernet/cadence/macb_main.c
@@ -6185,16 +6185,19 @@ static int __maybe_unused macb_runtime_resume(struct device *dev)
 {
 	struct net_device *netdev = dev_get_drvdata(dev);
 	struct macb *bp = netdev_priv(netdev);
+	struct clk_bulk_data clks[] = {
+		{ .clk = bp->pclk },
+		{ .clk = bp->hclk },
+		{ .clk = bp->tx_clk },
+		{ .clk = bp->rx_clk },
+		{ .clk = bp->tsu_clk },
+	};
 
-	if (!(device_may_wakeup(dev))) {
-		clk_prepare_enable(bp->pclk);
-		clk_prepare_enable(bp->hclk);
-		clk_prepare_enable(bp->tx_clk);
-		clk_prepare_enable(bp->rx_clk);
-		clk_prepare_enable(bp->tsu_clk);
-	} else if (!(bp->caps & MACB_CAPS_NEED_TSUCLK)) {
-		clk_prepare_enable(bp->tsu_clk);
-	}
+	if (!(device_may_wakeup(dev)))
+		return clk_bulk_prepare_enable(ARRAY_SIZE(clks), clks);
+
+	if (!(bp->caps & MACB_CAPS_NEED_TSUCLK))
+		return clk_prepare_enable(bp->tsu_clk);
 
 	return 0;
 }
-- 
2.54.0

Re: [PATCH] net: macb: fix ignored return value of clk_prepare_enable() in runtime resume

[Théo Lebrun]

Hello Gustavo,

Your CC list is laking some net maintainers, how did you get who to send
this email to?

   ./scripts/get_maintainer.pl -f drivers/net/ethernet/cadence/macb_main.c

On Sun May 31, 2026 at 7:18 AM CEST, Gustavo Kenji Mendonça Kaneko wrote:
> macb_runtime_resume() calls clk_prepare_enable() five times without
> checking the return value. If any clock fails to enable, the driver
> silently continues with potentially unclocked hardware, which can
> cause undefined behavior or system instability on resume from runtime
> suspend.

Could you point out if this was figured out to address a bug or if it is
only a theoretical defense?

> The existing macb_clks_disable() already uses clk_bulk_disable_unprepare()
> for the symmetric disable path. Align the enable path by using
> clk_bulk_prepare_enable(), which atomically enables all clocks and
> automatically rolls back any successfully enabled clocks on failure.

"atomically" usually refers to some locking mechanism, is that the
signification you had in mind here? Because I see none.

> Signed-off-by: Gustavo Kenji Mendonça Kaneko <kaneko.dev@pm.me>

Thanks!

--
Théo Lebrun, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com