mtd: nand: lpc32xx: fix invalid error handling of a requested irq

PATCH New Vladimir Zapolskiy 4 participants 7 messages Dec 5, 2016 – Jan 5, 2017

Subsystems (3)

Memory technology devices (MTD) NAND flash Arm/LPC32XX SoC support

lore

Share & tools

lore

Vladimir Zapolskiy

author maintainer

2016-12-05 01:47 UTC #0 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

Semantics of NR_IRQS is different on machines with SPARSE_IRQ option disabled or enabled, in the latter case IRQs are allocated starting at least from the value specified by NR_IRQS and going upwards, so the check of (irq >= NR_IRQ) to decide about an error code returned by platform_get_irq() is completely invalid, don't attempt to overrule irq subsystem in the driver.

The change fixes LPC32xx NAND MLC driver initialization on boot.

Signed-off-by Vladimir Zapolskiy <[email protected]>

---
 drivers/mtd/nand/lpc32xx_mlc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

drivers/mtd/nand/lpc32xx_mlc.c +1 -1

index 8523881..bc6e49a 100644
@@ -776,7 +776,7 @@ static int lpc32xx_nand_probe(struct platform_device *pdev)
 	init_completion(&host->comp_controller);
 
 	host->irq = platform_get_irq(pdev, 0);
-	if ((host->irq < 0) || (host->irq >= NR_IRQS)) {
+	if (host->irq < 0) {
 		dev_err(&pdev->dev, "failed to get platform irq\n");
 		res = -EINVAL;
 		goto err_exit3;

Signature · 1 line

2.10.2

Sylvain Lemieux

2016-12-07 18:09 UTC #1 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

Quoted 7 lines · Vladimir Zapolskiy

Semantics of NR_IRQS is different on machines with SPARSE_IRQ option disabled or enabled, in the latter case IRQs are allocated starting at least from the value specified by NR_IRQS and going upwards, so the check of (irq >= NR_IRQ) to decide about an error code returned by platform_get_irq() is completely invalid, don't attempt to overrule irq subsystem in the driver.

Quoted 7 lines

The change fixes LPC32xx NAND MLC driver initialization on boot.
Signed-off-by Vladimir Zapolskiy <[email protected]>
---
 drivers/mtd/nand/lpc32xx_mlc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Acked-by Sylvain Lemieux <[email protected]>

Boris Brezillon

2017-01-03 09:12 UTC #2 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

In reply to Vladimir Zapolskiy

Hi Vladimir

Quoted 8 lines · On Mon, 5 Dec 2016 03:47:10 +0200 Vladimir Zapolskiy <[email protected]> wrote:

Semantics of NR_IRQS is different on machines with SPARSE_IRQ option disabled or enabled, in the latter case IRQs are allocated starting at least from the value specified by NR_IRQS and going upwards, so the check of (irq >= NR_IRQ) to decide about an error code returned by platform_get_irq() is completely invalid, don't attempt to overrule irq subsystem in the driver.
The change fixes LPC32xx NAND MLC driver initialization on boot.

Do you need to backport this fix to stable releases? If that's the case, I'll add the Cc: stable tag when applying.

Thanks,

Boris

Quoted 19 lines

Signed-off-by Vladimir Zapolskiy <[email protected]>

---
 drivers/mtd/nand/lpc32xx_mlc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

drivers/mtd/nand/lpc32xx_mlc.c +1 -1

index 8523881..bc6e49a 100644
@@ -776,7 +776,7 @@ static int lpc32xx_nand_probe(struct platform_device *pdev)
 	init_completion(&host->comp_controller);
 
 	host->irq = platform_get_irq(pdev, 0);
-	if ((host->irq < 0) || (host->irq >= NR_IRQS)) {
+	if (host->irq < 0) {
 		dev_err(&pdev->dev, "failed to get platform irq\n");
 		res = -EINVAL;
 		goto err_exit3;

Vladimir Zapolskiy

author maintainer

2017-01-03 10:16 UTC #3 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

Hi Boris,

Quoted 12 lines · Boris Brezillon

Hi Vladimir
Quoted 7 lines · On Mon, 5 Dec 2016 03:47:10 +0200 Vladimir Zapolskiy <[email protected]> wrote:
Semantics of NR_IRQS is different on machines with SPARSE_IRQ option disabled or enabled, in the latter case IRQs are allocated starting at least from the value specified by NR_IRQS and going upwards, so the check of (irq >= NR_IRQ) to decide about an error code returned by platform_get_irq() is completely invalid, don't attempt to overrule irq subsystem in the driver.

Quoted 4 lines

Quoted 1 line
The change fixes LPC32xx NAND MLC driver initialization on boot.
Do you need to backport this fix to stable releases? If that's the case, I'll add the Cc: stable tag when applying.

that will be great if you can add

Cc stable at kernel.org # v4.7+

Please feel free to add also the tag

Fixes 8cb17b5ed017 ("irqchip: Add LPC32xx interrupt controller driver")

--
With best wishes,
Vladimir

Boris Brezillon

Applied

2017-01-03 10:47 UTC #4 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

Quoted 20 lines · On Tue, 3 Jan 2017 12:16:26 +0200 Vladimir Zapolskiy <[email protected]> wrote:

Hi Boris,
Quoted 12 lines · Boris Brezillon
Hi Vladimir
Quoted 7 lines · On Mon, 5 Dec 2016 03:47:10 +0200 Vladimir Zapolskiy <[email protected]> wrote:
Semantics of NR_IRQS is different on machines with SPARSE_IRQ option disabled or enabled, in the latter case IRQs are allocated starting at least from the value specified by NR_IRQS and going upwards, so the check of (irq >= NR_IRQ) to decide about an error code returned by platform_get_irq() is completely invalid, don't attempt to overrule irq subsystem in the driver.
Quoted 4 lines
Quoted 1 line
The change fixes LPC32xx NAND MLC driver initialization on boot.
Do you need to backport this fix to stable releases? If that's the case, I'll add the Cc: stable tag when applying.

Quoted 7 lines

that will be great if you can add
Cc stable at kernel.org # v4.7+
Please feel free to add also the tag
Fixes 8cb17b5ed017 ("irqchip: Add LPC32xx interrupt controller driver")

Applied to nand/next (this patch will appear in 4.11).

Thanks,

Boris

Quoted 5 lines

--
With best wishes,
Vladimir

Brian Norris

2017-01-05 17:58 UTC #5 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

Quoted 8 lines · Boris Brezillon

Quoted 6 lines · On Tue, 3 Jan 2017 12:16:26 +0200 Vladimir Zapolskiy <[email protected]> wrote:
Quoted 2 lines · Boris Brezillon
Do you need to backport this fix to stable releases? If that's the case, I'll add the Cc: stable tag when applying.
that will be great if you can add

Quoted 7 lines

Quoted 5 lines
Cc stable at kernel.org # v4.7+
Please feel free to add also the tag
Fixes 8cb17b5ed017 ("irqchip: Add LPC32xx interrupt controller driver")
Applied to nand/next (this patch will appear in 4.11).

For next time, checkpatch noticed that you got the stable address wrong:

ERROR:STABLE_ADDRESS: The 'stable' address should be 'stable at vger.kernel.org' #17: Cc: stable at kernel.org # v4.7+

Brian

Vladimir Zapolskiy

author maintainer

2017-01-05 18:19 UTC #6 lore raw

PATCH 1/1 mtd: nand: lpc32xx: fix invalid error handling of a requested irq

Quoted 15 lines · Brian Norris

Quoted 14 lines · Boris Brezillon
Quoted 11 lines · On Tue, 3 Jan 2017 12:16:26 +0200 Vladimir Zapolskiy <[email protected]> wrote:
Quoted 2 lines · Boris Brezillon
Do you need to backport this fix to stable releases? If that's the case, I'll add the Cc: stable tag when applying.
that will be great if you can add
Cc stable at kernel.org # v4.7+
Please feel free to add also the tag
Fixes 8cb17b5ed017 ("irqchip: Add LPC32xx interrupt controller driver")

Quoted 9 lines

Quoted 1 line
Applied to nand/next (this patch will appear in 4.11).
For next time, checkpatch noticed that you got the stable address wrong:
ERROR:STABLE_ADDRESS: The 'stable' address should be 'stable at vger.kernel.org' #17: Cc: stable at kernel.org # v4.7+
Brian

sorry, I've git-logged to commit 52bce91165 and pasted the address from its commit message.

--
With best wishes,
Vladimir