In preparation to enabling -Wimplicit-fallthrough, mark switch
cases where we are expecting to fall through.
This patch fixes the following warnings:
drivers/platform/x86/acer-wmi.c: In function ‘set_u32’:
drivers/platform/x86/acer-wmi.c:1378:33: warning: this statement may fall through [-Wimplicit-fallthrough=]
if (cap == ACER_CAP_WIRELESS ||
^
drivers/platform/x86/acer-wmi.c:1386:3: note: here
case ACER_WMID:
^~~~
drivers/platform/x86/acer-wmi.c:1393:12: warning: this statement may fall through [-Wimplicit-fallthrough=]
else if (wmi_has_guid(WMID_GUID2))
^
drivers/platform/x86/acer-wmi.c:1395:3: note: here
default:
^~~~~~~
drivers/platform/x86/acer-wmi.c: In function ‘get_u32’:
drivers/platform/x86/acer-wmi.c:1340:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
if (cap == ACER_CAP_MAILLED) {
^
drivers/platform/x86/acer-wmi.c:1344:2: note: here
case ACER_WMID:
^~~~
drivers/platform/x86/acer-wmi.c: In function ‘WMID_get_u32’:
drivers/platform/x86/acer-wmi.c:1013:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
if (quirks->mailled == 1) {
^
drivers/platform/x86/acer-wmi.c:1018:2: note: here
default:
^~~~~~~
Warning level 3 was used: -Wimplicit-fallthrough=3
This patch is part of the ongoing efforts to enable
-Wimplicit-fallthrough.
Signed-off-by: Gustavo A. R. Silva <[email protected]>
---
drivers/platform/x86/acer-wmi.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
index fcfeadd1301f..bd87f9037f95 100644
--- a/drivers/platform/x86/acer-wmi.c
+++ b/drivers/platform/x86/acer-wmi.c
@@ -1015,6 +1015,7 @@ static acpi_status WMID_get_u32(u32 *value, u32 cap)
*value = tmp & 0x1;
return 0;
}
+ /* fall through */
default:
return AE_ERROR;
}
@@ -1341,6 +1342,7 @@ static acpi_status get_u32(u32 *value, u32 cap)
status = AMW0_get_u32(value, cap);
break;
}
+ /* fall through */
case ACER_WMID:
status = WMID_get_u32(value, cap);
break;
@@ -1383,6 +1385,7 @@ static acpi_status set_u32(u32 value, u32 cap)
return AMW0_set_u32(value, cap);
}
+ /* fall through */
case ACER_WMID:
return WMID_set_u32(value, cap);
case ACER_WMID_v2:
@@ -1392,6 +1395,7 @@ static acpi_status set_u32(u32 value, u32 cap)
return wmid_v2_set_u32(value, cap);
else if (wmi_has_guid(WMID_GUID2))
return WMID_set_u32(value, cap);
+ /* fall through */
default:
return AE_BAD_PARAMETER;
}
--
2.21.0
On Wed, May 08, 2019 at 11:49:34AM -0500, Gustavo A. R. Silva wrote:
> In preparation to enabling -Wimplicit-fallthrough, mark switch
> cases where we are expecting to fall through.
>
> This patch fixes the following warnings:
>
> drivers/platform/x86/acer-wmi.c: In function ‘set_u32’:
> drivers/platform/x86/acer-wmi.c:1378:33: warning: this statement may fall through [-Wimplicit-fallthrough=]
> if (cap == ACER_CAP_WIRELESS ||
> ^
> drivers/platform/x86/acer-wmi.c:1386:3: note: here
> case ACER_WMID:
> ^~~~
> drivers/platform/x86/acer-wmi.c:1393:12: warning: this statement may fall through [-Wimplicit-fallthrough=]
> else if (wmi_has_guid(WMID_GUID2))
> ^
> drivers/platform/x86/acer-wmi.c:1395:3: note: here
> default:
> ^~~~~~~
> drivers/platform/x86/acer-wmi.c: In function ‘get_u32’:
> drivers/platform/x86/acer-wmi.c:1340:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
> if (cap == ACER_CAP_MAILLED) {
> ^
> drivers/platform/x86/acer-wmi.c:1344:2: note: here
> case ACER_WMID:
> ^~~~
> drivers/platform/x86/acer-wmi.c: In function ‘WMID_get_u32’:
> drivers/platform/x86/acer-wmi.c:1013:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
> if (quirks->mailled == 1) {
> ^
> drivers/platform/x86/acer-wmi.c:1018:2: note: here
> default:
> ^~~~~~~
>
> Warning level 3 was used: -Wimplicit-fallthrough=3
>
> This patch is part of the ongoing efforts to enable
> -Wimplicit-fallthrough.
>
> Signed-off-by: Gustavo A. R. Silva <[email protected]>
> ---
> drivers/platform/x86/acer-wmi.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
> index fcfeadd1301f..bd87f9037f95 100644
> --- a/drivers/platform/x86/acer-wmi.c
> +++ b/drivers/platform/x86/acer-wmi.c
> @@ -1015,6 +1015,7 @@ static acpi_status WMID_get_u32(u32 *value, u32 cap)
> *value = tmp & 0x1;
> return 0;
> }
> + /* fall through */
> default:
> return AE_ERROR;
> }
> @@ -1341,6 +1342,7 @@ static acpi_status get_u32(u32 *value, u32 cap)
> status = AMW0_get_u32(value, cap);
> break;
> }
> + /* fall through */
This doesn't strike me as obviously the right thing to do here. If the interface
type is AMW0_V2, why is it the right thing to do to use WMID_get_u32 if the cap
isn't ACER_CAP_MAILLED?
> case ACER_WMID:
> status = WMID_get_u32(value, cap);
> break;
> @@ -1383,6 +1385,7 @@ static acpi_status set_u32(u32 value, u32 cap)
>
> return AMW0_set_u32(value, cap);
> }
> + /* fall through */
Similarly here.
Are we documenting intended behavior, or covering up a bug.
> case ACER_WMID:
> return WMID_set_u32(value, cap);
--
Darren Hart
VMware Open Source Technology Center
Darren,
Please, see my comments below...
On 5/8/19 6:06 PM, Darren Hart wrote:
> On Wed, May 08, 2019 at 11:49:34AM -0500, Gustavo A. R. Silva wrote:
>> In preparation to enabling -Wimplicit-fallthrough, mark switch
>> cases where we are expecting to fall through.
>>
>> This patch fixes the following warnings:
>>
>> drivers/platform/x86/acer-wmi.c: In function ‘set_u32’:
>> drivers/platform/x86/acer-wmi.c:1378:33: warning: this statement may fall through [-Wimplicit-fallthrough=]
>> if (cap == ACER_CAP_WIRELESS ||
>> ^
>> drivers/platform/x86/acer-wmi.c:1386:3: note: here
>> case ACER_WMID:
>> ^~~~
>> drivers/platform/x86/acer-wmi.c:1393:12: warning: this statement may fall through [-Wimplicit-fallthrough=]
>> else if (wmi_has_guid(WMID_GUID2))
>> ^
>> drivers/platform/x86/acer-wmi.c:1395:3: note: here
>> default:
>> ^~~~~~~
>> drivers/platform/x86/acer-wmi.c: In function ‘get_u32’:
>> drivers/platform/x86/acer-wmi.c:1340:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
>> if (cap == ACER_CAP_MAILLED) {
>> ^
>> drivers/platform/x86/acer-wmi.c:1344:2: note: here
>> case ACER_WMID:
>> ^~~~
>> drivers/platform/x86/acer-wmi.c: In function ‘WMID_get_u32’:
>> drivers/platform/x86/acer-wmi.c:1013:6: warning: this statement may fall through [-Wimplicit-fallthrough=]
>> if (quirks->mailled == 1) {
>> ^
>> drivers/platform/x86/acer-wmi.c:1018:2: note: here
>> default:
>> ^~~~~~~
>>
>> Warning level 3 was used: -Wimplicit-fallthrough=3
>>
>> This patch is part of the ongoing efforts to enable
>> -Wimplicit-fallthrough.
>>
>> Signed-off-by: Gustavo A. R. Silva <[email protected]>
>> ---
>> drivers/platform/x86/acer-wmi.c | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/platform/x86/acer-wmi.c b/drivers/platform/x86/acer-wmi.c
>> index fcfeadd1301f..bd87f9037f95 100644
>> --- a/drivers/platform/x86/acer-wmi.c
>> +++ b/drivers/platform/x86/acer-wmi.c
>> @@ -1015,6 +1015,7 @@ static acpi_status WMID_get_u32(u32 *value, u32 cap)
>> *value = tmp & 0x1;
>> return 0;
>> }
>> + /* fall through */
>> default:
>> return AE_ERROR;
>> }
>> @@ -1341,6 +1342,7 @@ static acpi_status get_u32(u32 *value, u32 cap)
>> status = AMW0_get_u32(value, cap);
>> break;
>> }
>> + /* fall through */
>
> This doesn't strike me as obviously the right thing to do here. If the interface
> type is AMW0_V2, why is it the right thing to do to use WMID_get_u32 if the cap
> isn't ACER_CAP_MAILLED?
>
In commit commit 745a5d2126926808295742932d0e36d485efa485 case ACER_AMW0_V2 falls
through to case ACER_WMID deliberately in function set_u32(), without reporting
any error or warning. So, I thought it was fair to assume that the fall-through
is intentional in both functions get_u32() and set_u32(). Otherwise I would
expect to see a message indicating that interface ACER_AMW0_V2 is unavailable
in function set_u32().
This is also complemented by the following...
>> case ACER_WMID:
>> status = WMID_get_u32(value, cap);
>> break;
>> @@ -1383,6 +1385,7 @@ static acpi_status set_u32(u32 value, u32 cap)
>>
>> return AMW0_set_u32(value, cap);
>> }
>> + /* fall through */
>
> Similarly here.
>
> Are we documenting intended behavior, or covering up a bug.
>
Commit 5c742b45dd5fbbb6cf74d3378341704f4b23c5e8 mentions that "This was fixed
in acer_acpi some time ago, but I forgot to port the patch over to acer-wmi
when it was merged." Notice that this driver (acer-wmi) is based on the
no-longer existing acer_acpi driver. But after googling for a while I could
found the fix the original author talks about:
https://repo.or.cz/acer_acpi.git/commitdiff/74c08a38875ffa9989c3100947650ac8a388c189
So, the fix is indeed similar and contains the same fall-throughs from case
ACER_AMW0_V2 to case ACER_WMID in both functions get_u32() and set_u32().
Thanks
--
Gustavo
On Thu, May 9, 2019 at 4:48 AM Gustavo A. R. Silva
<[email protected]> wrote:
> Commit 5c742b45dd5fbbb6cf74d3378341704f4b23c5e8 mentions that "This was fixed
> in acer_acpi some time ago, but I forgot to port the patch over to acer-wmi
> when it was merged." Notice that this driver (acer-wmi) is based on the
> no-longer existing acer_acpi driver. But after googling for a while I could
> found the fix the original author talks about:
>
> https://repo.or.cz/acer_acpi.git/commitdiff/74c08a38875ffa9989c3100947650ac8a388c189
>
> So, the fix is indeed similar and contains the same fall-throughs from case
> ACER_AMW0_V2 to case ACER_WMID in both functions get_u32() and set_u32().
Pushed to my review and testing queue, thanks!
--
With Best Regards,
Andy Shevchenko