[PATCH 5.12 061/242] pinctrl: mcp23s08: fix race condition in irq handler

From: Greg Kroah-Hartman
Date: Thu Jul 15 2021 - 15:08:08 EST


From: Radim Pavlik <radim.pavlik@xxxxxxxxxxxxxxxxxx>

[ Upstream commit 897120d41e7afd9da435cb00041a142aeeb53c07 ]

Checking value of MCP_INTF in mcp23s08_irq suggests that the handler may be
called even when there is no interrupt pending.

But the actual interrupt could happened between reading MCP_INTF and MCP_GPIO.
In this situation we got nothing from MCP_INTF, but the event gets acknowledged
on the expander by reading MCP_GPIO. This leads to losing events.

Fix the problem by not reading any register until we see something in MCP_INTF.

The error was reproduced and fix tested on MCP23017.

Signed-off-by: Radim Pavlik <radim.pavlik@xxxxxxxxxxxxxxxxxx>
Link: https://lore.kernel.org/r/AM7PR06MB6769E1183F68DEBB252F665ABA3E9@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Signed-off-by: Linus Walleij <linus.walleij@xxxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
---
drivers/pinctrl/pinctrl-mcp23s08.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/pinctrl/pinctrl-mcp23s08.c b/drivers/pinctrl/pinctrl-mcp23s08.c
index ce2d8014b7e0..799d596a1a4b 100644
--- a/drivers/pinctrl/pinctrl-mcp23s08.c
+++ b/drivers/pinctrl/pinctrl-mcp23s08.c
@@ -351,6 +351,11 @@ static irqreturn_t mcp23s08_irq(int irq, void *data)
if (mcp_read(mcp, MCP_INTF, &intf))
goto unlock;

+ if (intf == 0) {
+ /* There is no interrupt pending */
+ return IRQ_HANDLED;
+ }
+
if (mcp_read(mcp, MCP_INTCAP, &intcap))
goto unlock;

@@ -368,11 +373,6 @@ static irqreturn_t mcp23s08_irq(int irq, void *data)
mcp->cached_gpio = gpio;
mutex_unlock(&mcp->lock);

- if (intf == 0) {
- /* There is no interrupt pending */
- return IRQ_HANDLED;
- }
-
dev_dbg(mcp->chip.parent,
"intcap 0x%04X intf 0x%04X gpio_orig 0x%04X gpio 0x%04X\n",
intcap, intf, gpio_orig, gpio);
--
2.30.2