Commit 19d337dff95cbf76ed ("rfkill: rewrite") incorrectly reversed
the meaning of 'state' in acer_rfkill_update() when it changed
rfkill_force_state() to rfkill_set_sw_state(). Fix it.
Signed-off-by: Troy Moure <[email protected]>
---
This fixes the rfkill switch on my Acer laptop, which was behaving
backwards (the state reported in the logs was the opposite of the
state indicated by the LED).
drivers/platform/x86/acer-wmi.c | 4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
index 09a503e..be2fd6f 100644
--- a/drivers/platform/x86/acer-wmi.c
+++ b/drivers/platform/x86/acer-wmi.c
@@ -958,12 +958,12 @@ static void acer_rfkill_update(struct work_struct *ignored)
status = get_u32(&state, ACER_CAP_WIRELESS);
if (ACPI_SUCCESS(status))
- rfkill_set_sw_state(wireless_rfkill, !!state);
+ rfkill_set_sw_state(wireless_rfkill, !state);
if (has_cap(ACER_CAP_BLUETOOTH)) {
status = get_u32(&state, ACER_CAP_BLUETOOTH);
if (ACPI_SUCCESS(status))
- rfkill_set_sw_state(bluetooth_rfkill, !!state);
+ rfkill_set_sw_state(bluetooth_rfkill, !state);
}
schedule_delayed_work(&acer_rfkill_work, round_jiffies_relative(HZ));
--
1.6.2.2
On Wed, 2009-06-17 at 11:51 +0100, Troy Moure wrote:
> Commit 19d337dff95cbf76ed ("rfkill: rewrite") incorrectly reversed
> the meaning of 'state' in acer_rfkill_update() when it changed
> rfkill_force_state() to rfkill_set_sw_state(). Fix it.
>
> Signed-off-by: Troy Moure <[email protected]>
> ---
> This fixes the rfkill switch on my Acer laptop, which was behaving
> backwards (the state reported in the logs was the opposite of the
> state indicated by the LED).
Thanks. Looks like I made this mistake a few times.
johannes
> drivers/platform/x86/acer-wmi.c | 4 ++--
> 1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
> index 09a503e..be2fd6f 100644
> --- a/drivers/platform/x86/acer-wmi.c
> +++ b/drivers/platform/x86/acer-wmi.c
> @@ -958,12 +958,12 @@ static void acer_rfkill_update(struct work_struct *ignored)
>
> status = get_u32(&state, ACER_CAP_WIRELESS);
> if (ACPI_SUCCESS(status))
> - rfkill_set_sw_state(wireless_rfkill, !!state);
> + rfkill_set_sw_state(wireless_rfkill, !state);
>
> if (has_cap(ACER_CAP_BLUETOOTH)) {
> status = get_u32(&state, ACER_CAP_BLUETOOTH);
> if (ACPI_SUCCESS(status))
> - rfkill_set_sw_state(bluetooth_rfkill, !!state);
> + rfkill_set_sw_state(bluetooth_rfkill, !state);
> }
>
> schedule_delayed_work(&acer_rfkill_work, round_jiffies_relative(HZ));
On 6/17/09, Troy Moure <[email protected]> wrote:
> Commit 19d337dff95cbf76ed ("rfkill: rewrite") incorrectly reversed
> the meaning of 'state' in acer_rfkill_update() when it changed
> rfkill_force_state() to rfkill_set_sw_state(). Fix it.
>
> Signed-off-by: Troy Moure <[email protected]>
> ---
> This fixes the rfkill switch on my Acer laptop, which was behaving
> backwards (the state reported in the logs was the opposite of the
> state indicated by the LED).
I think there's another reversal error that you didn't notice :-). I found that acer-wmi disables the wireless when it is loaded.
The core tries to initialize the wireless to enabled, but the polarity is wrong in acer_rfkill_set(). It ends up disabling the wireless instead.
Here's a patch for it.
--->
>From 75953366b4958a77019729662a997f5d361a7529 Mon Sep 17 00:00:00 2001
From: Alan Jenkins <[email protected]>
Date: Sun, 19 Jul 2009 09:29:21 +0100
Subject: [PATCH] acer-wmi: fix rfkill conversion
Fix another polarity error introduced by the rfkill rewrite,
this time in acer_rfkill_set().
Signed-off-by: Alan Jenkins <[email protected]>
---
drivers/platform/x86/acer-wmi.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
index be2fd6f..fb45f5e 100644
--- a/drivers/platform/x86/acer-wmi.c
+++ b/drivers/platform/x86/acer-wmi.c
@@ -973,7 +973,7 @@ static int acer_rfkill_set(void *data, bool blocked)
{
acpi_status status;
u32 cap = (unsigned long)data;
- status = set_u32(!!blocked, cap);
+ status = set_u32(!blocked, cap);
if (ACPI_FAILURE(status))
return -ENODEV;
return 0;
--
1.6.3.2
On Sun, 2009-07-19 at 09:48 +0100, Alan Jenkins wrote:
> On 6/17/09, Troy Moure <[email protected]> wrote:
> > Commit 19d337dff95cbf76ed ("rfkill: rewrite") incorrectly reversed
> > the meaning of 'state' in acer_rfkill_update() when it changed
> > rfkill_force_state() to rfkill_set_sw_state(). Fix it.
> >
> > Signed-off-by: Troy Moure <[email protected]>
> > ---
> > This fixes the rfkill switch on my Acer laptop, which was behaving
> > backwards (the state reported in the logs was the opposite of the
> > state indicated by the LED).
>
> I think there's another reversal error that you didn't notice :-). I
> found that acer-wmi disables the wireless when it is loaded.
Ahrg.
> The core tries to initialize the wireless to enabled, but the polarity
> is wrong in acer_rfkill_set(). It ends up disabling the wireless
> instead.
>
> Here's a patch for it.
I really should go back and audit it again, but every instance takes me
about 5 minutes to do the (not state != unblocked) thing in my
head ... :/
Thanks!
johannes
> --->
> From 75953366b4958a77019729662a997f5d361a7529 Mon Sep 17 00:00:00 2001
> From: Alan Jenkins <[email protected]>
> Date: Sun, 19 Jul 2009 09:29:21 +0100
> Subject: [PATCH] acer-wmi: fix rfkill conversion
>
> Fix another polarity error introduced by the rfkill rewrite,
> this time in acer_rfkill_set().
>
> Signed-off-by: Alan Jenkins <[email protected]>
> ---
> drivers/platform/x86/acer-wmi.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
> index be2fd6f..fb45f5e 100644
> --- a/drivers/platform/x86/acer-wmi.c
> +++ b/drivers/platform/x86/acer-wmi.c
> @@ -973,7 +973,7 @@ static int acer_rfkill_set(void *data, bool blocked)
> {
> acpi_status status;
> u32 cap = (unsigned long)data;
> - status = set_u32(!!blocked, cap);
> + status = set_u32(!blocked, cap);
> if (ACPI_FAILURE(status))
> return -ENODEV;
> return 0;