During my regression testing I noticed the cadence GPIO driver
fails on the latest gpio for-next tree.
I think the reason is this patch:
commit 96cd559817f2 ("Merge branch 'devel' into for-next")
Here is a part of the test log:
Loopback 8 -> 24
TESTING: gpio: 488: output direction PASSED
TESTING: gpio: 504: input direction PASSED
TESTING: gpio: 488: 0 PASSED
TESTING: gpio: 488 -> 504: 0 PASSED
TESTING: gpio: 488: 1 FAILED
TESTING: gpio: 488 -> 504: 1 FAILED
TESTING: gpio: 488: 0 PASSED
TESTING: gpio: 488 -> 504: 0 PASSED
It looks like the issue is that gc->bgpio_dir has changed its meaning.
It used to be set to the register value (so it was being inverted).
Now it's always set to 1 for output and 0 for input.
However the bgpio_get_set functions were not updated.
So they invert the bit again, which means a wrong register
is being accessed.
This patch fixes that by removing the unnecessary inversion.
Signed-off-by: Jan Kotas <[email protected]>
---
drivers/gpio/gpio-mmio.c | 21 ++-------------------
1 file changed, 2 insertions(+), 19 deletions(-)
diff --git a/drivers/gpio/gpio-mmio.c b/drivers/gpio/gpio-mmio.c
index f172b4382d..04be18b954 100644
--- a/drivers/gpio/gpio-mmio.c
+++ b/drivers/gpio/gpio-mmio.c
@@ -134,17 +134,6 @@ static int bgpio_get_set(struct gpio_chip *gc, unsigned int gpio)
unsigned long pinmask = bgpio_line2mask(gc, gpio);
bool dir = !!(gc->bgpio_dir & pinmask);
- /*
- * If the direction is OUT we read the value from the SET
- * register, and if the direction is IN we read the value
- * from the DAT register.
- *
- * If the direction bits are inverted, naturally this gets
- * inverted too.
- */
- if (gc->bgpio_dir_inverted)
- dir = !dir;
-
if (dir)
return !!(gc->read_reg(gc->reg_set) & pinmask);
else
@@ -164,14 +153,8 @@ static int bgpio_get_set_multiple(struct gpio_chip *gc, unsigned long *mask,
/* Make sure we first clear any bits that are zero when we read the register */
*bits &= ~*mask;
- /* Exploit the fact that we know which directions are set */
- if (gc->bgpio_dir_inverted) {
- set_mask = *mask & ~gc->bgpio_dir;
- get_mask = *mask & gc->bgpio_dir;
- } else {
- set_mask = *mask & gc->bgpio_dir;
- get_mask = *mask & ~gc->bgpio_dir;
- }
+ set_mask = *mask & gc->bgpio_dir;
+ get_mask = *mask & ~gc->bgpio_dir;
if (set_mask)
*bits |= gc->read_reg(gc->reg_set) & set_mask;
--
2.15.0
On Mon, Apr 1, 2019 at 4:10 PM Jan Kotas <[email protected]> wrote:
> During my regression testing I noticed the cadence GPIO driver
> fails on the latest gpio for-next tree.
>
> I think the reason is this patch:
> commit 96cd559817f2 ("Merge branch 'devel' into for-next")
>
> Here is a part of the test log:
>
> Loopback 8 -> 24
> TESTING: gpio: 488: output direction PASSED
> TESTING: gpio: 504: input direction PASSED
> TESTING: gpio: 488: 0 PASSED
> TESTING: gpio: 488 -> 504: 0 PASSED
> TESTING: gpio: 488: 1 FAILED
> TESTING: gpio: 488 -> 504: 1 FAILED
> TESTING: gpio: 488: 0 PASSED
> TESTING: gpio: 488 -> 504: 0 PASSED
>
> It looks like the issue is that gc->bgpio_dir has changed its meaning.
> It used to be set to the register value (so it was being inverted).
>
> Now it's always set to 1 for output and 0 for input.
> However the bgpio_get_set functions were not updated.
> So they invert the bit again, which means a wrong register
> is being accessed.
>
> This patch fixes that by removing the unnecessary inversion.
>
> Signed-off-by: Jan Kotas <[email protected]>
Patch applied!
Yours,
Linus Walleij