Skip to contentSkip to conversation

net: stmmac: platform: Fix misleading interrupt error msg

PATCH Applied Markus Fuchs 2 participants 5 messages Mar 6, 2020 – Mar 15, 2020

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net: stmmac: platform: Fix misleading interrupt error msg

Not every stmmac based platform makes use of the eth_wake_irq or eth_lpi interrupts. Use the platform_get_irq_byname_optional variant for these interrupts, so no error message is displayed, if they can't be found. Rather print an information to hint something might be wrong to assist debugging on platforms which use these interrupts.

---
On my cyclone V socfpga platform I get error messages after updating to Linux Kernel 5.4.24

Starting kernel ...

Deasserting all peripheral resets

[    1.206363] socfpga-dwmac ff700000.ethernet: IRQ eth_wake_irq not found
[    1.213023] socfpga-dwmac ff700000.ethernet: IRQ eth_lpi not found

These interrupts don't matter for my platform and many other stmmac based ones, as we can see by grepping for 'macirq'.

socfpga.dtsi:                   interrupt-names = "macirq";
socfpga.dtsi:                   interrupt-names = "macirq";
sun7i-a20.dtsi:                 interrupt-names = "macirq";
spear600.dtsi:                  interrupt-names = "macirq", "eth_wake_irq";
artpec6.dtsi:                   interrupt-names = "macirq", "eth_lpi";
rk322x.dtsi:                    interrupt-names = "macirq";
sun9i-a80.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
stih407-family.dtsi:            interrupt-names = "macirq", "eth_wake_irq";
stm32f429.dtsi:                 interrupt-names = "macirq";
sun6i-a31.dtsi:                 interrupt-names = "macirq";
meson.dtsi:                     interrupt-names = "macirq";
rk3288.dtsi:                    interrupt-names = "macirq", "eth_wake_irq";
sun8i-r40.dtsi:                 interrupt-names = "macirq";
sunxi-h3-h5.dtsi:               interrupt-names = "macirq";
spear3xx.dtsi:                  interrupt-names = "macirq", "eth_wake_irq";
lpc18xx.dtsi:                   interrupt-names = "macirq";
stm32h743.dtsi:                 interrupt-names = "macirq";
socfpga_arria10.dtsi:           interrupt-names = "macirq";
socfpga_arria10.dtsi:           interrupt-names = "macirq";
socfpga_arria10.dtsi:           interrupt-names = "macirq";
rv1108.dtsi:                    interrupt-names = "macirq", "eth_wake_irq";
spear13xx.dtsi:                 interrupt-names = "macirq", "eth_wake_irq";
stm32mp151.dtsi:                interrupt-names = "macirq";
ox820.dtsi:                     interrupt-names = "macirq", "eth_wake_irq";
sun8i-a83t.dtsi:                interrupt-names = "macirq";

So, in my opinion, the error messages are missleading. I believe the right way to handle this would require more changes though. Some kind of configuration information, telling which interrupts are required by the platform and than conditionally call platform_get_irq_byname(). This would print an error message, if something is wrong, on the right platforms and nothing at all on the others.

.../net/ethernet/stmicro/stmmac/stmmac_platform.c | 14 ++++++++++----

 1 file changed, 10 insertions(+), 4 deletions(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_platform.c +10 -4
index d10ac54bf385..13fafd905db8 100644
@@ -663,16 +663,22 @@ int stmmac_get_platform_resources(struct platform_device *pdev,
 	 * In case the wake up interrupt is not passed from the platform
 	 * so the driver will continue to use the mac irq (ndev->irq)
 	 */
-	stmmac_res->wol_irq = platform_get_irq_byname(pdev, "eth_wake_irq");
+	stmmac_res->wol_irq =
+		platform_get_irq_byname_optional(pdev, "eth_wake_irq");
 	if (stmmac_res->wol_irq < 0) {
 		if (stmmac_res->wol_irq == -EPROBE_DEFER)
 			return -EPROBE_DEFER;
+		dev_info(&pdev->dev, "IRQ eth_wake_irq not found\n");
 		stmmac_res->wol_irq = stmmac_res->irq;
 	}
 
-	stmmac_res->lpi_irq = platform_get_irq_byname(pdev, "eth_lpi");
-	if (stmmac_res->lpi_irq == -EPROBE_DEFER)
-		return -EPROBE_DEFER;
+	stmmac_res->lpi_irq =
+		platform_get_irq_byname_optional(pdev, "eth_lpi");
+	if (stmmac_res->lpi_irq < 0) {
+		if (stmmac_res->lpi_irq == -EPROBE_DEFER)
+			return -EPROBE_DEFER;
+		dev_info(&pdev->dev, "IRQ eth_lpi not found\n");
+	}
 
 	res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	stmmac_res->addr = devm_ioremap_resource(&pdev->dev, res);
Signature · 1 line
2.25.1
maintainer
#1 lore raw

From: Markus Fuchs <[email protected]>
Date: Fri, 6 Mar 2020 17:38:48 +0100

Quoted 7 lines

Not every stmmac based platform makes use of the eth_wake_irq or eth_lpi interrupts. Use the platform_get_irq_byname_optional variant for these interrupts, so no error message is displayed, if they can't be found. Rather print an information to hint something might be wrong to assist debugging on platforms which use these interrupts.

What do you mean the error message is misleading right now?

It isn't printing anything out at the moment in this situation.

#2 lore raw
Quoted 10 lines · David Miller

From: Markus Fuchs <[email protected]>
Date: Fri, 6 Mar 2020 17:38:48 +0100

Quoted 6 lines

Not every stmmac based platform makes use of the eth_wake_irq or eth_lpi interrupts. Use the platform_get_irq_byname_optional variant for these interrupts, so no error message is displayed, if they can't be found. Rather print an information to hint something might be wrong to assist debugging on platforms which use these interrupts.

Quoted 5 lines
Quoted 1 line

What do you mean the error message is misleading right now?

It isn't printing anything out at the moment in this situation.

Hello,

the error messages are

[    1.206363] socfpga-dwmac ff700000.ethernet: IRQ eth_wake_irq not found
[    1.213023] socfpga-dwmac ff700000.ethernet: IRQ eth_lpi not found

I tried to explain this in my original post between the --- lines of the patch. Maybe this was wrong, so I repost it.

On my cyclone V socfpga platform I get error messages after updating to Linux Kernel 5.4.24

Starting kernel ...

Deasserting all peripheral resets

[    1.206363] socfpga-dwmac ff700000.ethernet: IRQ eth_wake_irq not found
[    1.213023] socfpga-dwmac ff700000.ethernet: IRQ eth_lpi not found

These interrupts don't matter for my platform and many other stmmac based ones, as we can see by grepping for 'macirq'.

socfpga.dtsi:                   interrupt-names = "macirq";
socfpga.dtsi:                   interrupt-names = "macirq";
sun7i-a20.dtsi:                 interrupt-names = "macirq";
spear600.dtsi:                  interrupt-names = "macirq", "eth_wake_irq";
artpec6.dtsi:                   interrupt-names = "macirq", "eth_lpi";
rk322x.dtsi:                    interrupt-names = "macirq";
sun9i-a80.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
spear1310.dtsi:                 interrupt-names = "macirq";
stih407-family.dtsi:            interrupt-names = "macirq", "eth_wake_irq";
stm32f429.dtsi:                 interrupt-names = "macirq";
sun6i-a31.dtsi:                 interrupt-names = "macirq";
meson.dtsi:                     interrupt-names = "macirq";
rk3288.dtsi:                    interrupt-names = "macirq", "eth_wake_irq";
sun8i-r40.dtsi:                 interrupt-names = "macirq";
sunxi-h3-h5.dtsi:               interrupt-names = "macirq";
spear3xx.dtsi:                  interrupt-names = "macirq", "eth_wake_irq";
lpc18xx.dtsi:                   interrupt-names = "macirq";
stm32h743.dtsi:                 interrupt-names = "macirq";
socfpga_arria10.dtsi:           interrupt-names = "macirq";
socfpga_arria10.dtsi:           interrupt-names = "macirq";
socfpga_arria10.dtsi:           interrupt-names = "macirq";
rv1108.dtsi:                    interrupt-names = "macirq", "eth_wake_irq";
spear13xx.dtsi:                 interrupt-names = "macirq", "eth_wake_irq";
stm32mp151.dtsi:                interrupt-names = "macirq";
ox820.dtsi:                     interrupt-names = "macirq", "eth_wake_irq";
sun8i-a83t.dtsi:                interrupt-names = "macirq";

So, in my opinion, the error messages are missleading. I believe the right way to handle this would require more changes though. Some kind of configuration information, telling which interrupts are required by the platform and than conditionally call platform_get_irq_byname(). This would print an error message, if something is wrong, on the right platforms and nothing at all on the others.

#3 lore raw
In reply to David Miller
Quoted 9 lines · David Miller

From: Markus Fuchs <[email protected]>
Date: Fri, 6 Mar 2020 17:38:48 +0100

Quoted 6 lines

Not every stmmac based platform makes use of the eth_wake_irq or eth_lpi interrupts. Use the platform_get_irq_byname_optional variant for these interrupts, so no error message is displayed, if they can't be found. Rather print an information to hint something might be wrong to assist debugging on platforms which use these interrupts.

Quoted 5 lines
Quoted 1 line

What do you mean the error message is misleading right now?

It isn't printing anything out at the moment in this situation.

Commit 7723f4c5ecdb driver core: platform: Add an error message to

    platform_get_irq*()

The above commit added a generic dev_err() output to the platform_get_irq_byname function.

My patch uses the platform_get_irq_byname_optional function, which doesn't print anything and adds the original dev_err output as dev_info output to the driver.
Otherwise there would be no output at all even for platforms in need of these irqs.

maintainer Applied
#4 lore raw

From: Markus Fuchs <[email protected]>
Date: Thu, 12 Mar 2020 20:46:25 +0100

Quoted 19 lines · David Miller
Quoted 9 lines

From: Markus Fuchs <[email protected]>
Date: Fri, 6 Mar 2020 17:38:48 +0100

Quoted 6 lines

Not every stmmac based platform makes use of the eth_wake_irq or eth_lpi interrupts. Use the platform_get_irq_byname_optional variant for these interrupts, so no error message is displayed, if they can't be found. Rather print an information to hint something might be wrong to assist debugging on platforms which use these interrupts.

Quoted 5 lines
Quoted 1 line

What do you mean the error message is misleading right now?

It isn't printing anything out at the moment in this situation.

Commit 7723f4c5ecdb driver core: platform: Add an error message to

    platform_get_irq*()
Quoted 8 lines

The above commit added a generic dev_err() output to the platform_get_irq_byname function.

My patch uses the platform_get_irq_byname_optional function, which doesn't print anything and adds the original dev_err output as dev_info output to the driver.
Otherwise there would be no output at all even for platforms in need of these irqs.

Aha, now I get it, thanks for explaining.

Applied, thank you.