Skip to contentSkip to conversation

i2c: mv64xxx: Fix reading invalid status value in atomic mode

PATCH Applied Marek Behún 3 participants 3 messages May 21, 2023 – Jun 5, 2023

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 i2c: mv64xxx: Fix reading invalid status value in atomic mode

There seems to be a bug within the mv64xxx I2C controller, wherein the status register may not necessarily contain valid value immediately after the IFLG flag is set in the control register.

My theory is that the controller:
- first sets the IFLG in control register
- then updates the status register
- then raises an interrupt

This may sometime cause weird bugs when in atomic mode, since in this mode we do not wait for an interrupt, but instead we poll the control register for IFLG and read status register immediately after.

I encountered -ENXIO from mv64xxx_i2c_fsm() due to this issue when using this driver in atomic mode.

Note that I've only seen this issue on Armada 385, I don't know whether other SOCs with this controller are also affected. Also note that this fix has been in U-Boot for over 4 years [1] without anybody complaining, so it should not cause regressions.

[1] https://source.denx.de/u-boot/u-boot/-/commit/d50e29662f78

Fixes 544a8d75f3d6 ("i2c: mv64xxx: Add atomic_xfer method to driver")
---
 drivers/i2c/busses/i2c-mv64xxx.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/i2c/busses/i2c-mv64xxx.c +11 -0
index 047dfef7a657..878c076ebdc6 100644
@@ -520,6 +520,17 @@ mv64xxx_i2c_intr(int irq, void *dev_id)
 
 	while (readl(drv_data->reg_base + drv_data->reg_offsets.control) &
 						MV64XXX_I2C_REG_CONTROL_IFLG) {
+		/*
+		 * It seems that sometime the controller updates the status
+		 * register only after it asserts IFLG in control register.
+		 * This may result in weird bugs when in atomic mode. A delay
+		 * of 100 ns before reading the status register solves this
+		 * issue. This bug does not seem to appear when using
+		 * interrupts.
+		 */
+		if (drv_data->atomic)
+			ndelay(100);
+
 		status = readl(drv_data->reg_base + drv_data->reg_offsets.status);
 		mv64xxx_i2c_fsm(drv_data, status);
 		mv64xxx_i2c_do_action(drv_data);
Signature · 1 line
2.39.3
#1 lore raw
Quoted 13 lines · Marek Behún

There seems to be a bug within the mv64xxx I2C controller, wherein the status register may not necessarily contain valid value immediately after the IFLG flag is set in the control register.

My theory is that the controller:
- first sets the IFLG in control register
- then updates the status register
- then raises an interrupt

This may sometime cause weird bugs when in atomic mode, since in this mode we do not wait for an interrupt, but instead we poll the control register for IFLG and read status register immediately after.

Quoted 7 lines

I encountered -ENXIO from mv64xxx_i2c_fsm() due to this issue when using this driver in atomic mode.

Note that I've only seen this issue on Armada 385, I don't know whether other SOCs with this controller are also affected. Also note that this fix has been in U-Boot for over 4 years [1] without anybody complaining, so it should not cause regressions.

I've never seen this bug before, but don't suspect it should cause any issues for me. Thank you for finding/fixing this.

Quoted 34 lines

[1] https://source.denx.de/u-boot/u-boot/-/commit/d50e29662f78

Fixes 544a8d75f3d6 ("i2c: mv64xxx: Add atomic_xfer method to driver")
---
 drivers/i2c/busses/i2c-mv64xxx.c | 11 +++++++++++
 1 file changed, 11 insertions(+)
drivers/i2c/busses/i2c-mv64xxx.c +11 -0
index 047dfef7a657..878c076ebdc6 100644
@@ -520,6 +520,17 @@ mv64xxx_i2c_intr(int irq, void *dev_id)
 
 	while (readl(drv_data->reg_base + drv_data->reg_offsets.control) &
 						MV64XXX_I2C_REG_CONTROL_IFLG) {
+		/*
+		 * It seems that sometime the controller updates the status
+		 * register only after it asserts IFLG in control register.
+		 * This may result in weird bugs when in atomic mode. A delay
+		 * of 100 ns before reading the status register solves this
+		 * issue. This bug does not seem to appear when using
+		 * interrupts.
+		 */
+		if (drv_data->atomic)
+			ndelay(100);
+
 		status = readl(drv_data->reg_base + drv_data->reg_offsets.status);
 		mv64xxx_i2c_fsm(drv_data, status);
 		mv64xxx_i2c_do_action(drv_data);
Signature · 1 line
2.39.3
Applied
#2 lore raw
In reply to Marek Behún
Quoted 21 lines · Marek Behún

There seems to be a bug within the mv64xxx I2C controller, wherein the status register may not necessarily contain valid value immediately after the IFLG flag is set in the control register.

My theory is that the controller:
- first sets the IFLG in control register
- then updates the status register
- then raises an interrupt

This may sometime cause weird bugs when in atomic mode, since in this mode we do not wait for an interrupt, but instead we poll the control register for IFLG and read status register immediately after.

I encountered -ENXIO from mv64xxx_i2c_fsm() due to this issue when using this driver in atomic mode.

Note that I've only seen this issue on Armada 385, I don't know whether other SOCs with this controller are also affected. Also note that this fix has been in U-Boot for over 4 years [1] without anybody complaining, so it should not cause regressions.

Quoted 4 lines

[1] https://source.denx.de/u-boot/u-boot/-/commit/d50e29662f78

Fixes 544a8d75f3d6 ("i2c: mv64xxx: Add atomic_xfer method to driver")

Applied to for-current, thanks!