Re: [B.A.T.M.A.N.] [PATCH 2/2] batctl: suppress implicit-fallthrough compiler warning

[Simon Wunderlich <sw@simonwunderlich.de>]

[-- Attachment #1: Type: text/plain, Size: 1655 bytes --]

On Tuesday, June 13, 2017 10:26:00 AM CEST Philipp Psurek wrote:
> GCC 7.1.0 complains about an intended fallthrough.
> “__attribute__ ((fallthrough))” in this part of code would suppress this
> warning. Because older GCC compiler don’t understand this statement
> attribute and because there is already a comment in the source containing
> “falls?[ \t-]*thr(ough|u)” we can suppress the warning with the
> “-Wimplicit-fallthrough=2” warning option. Unintended fallthroughs without a
> comment will trigger this warning.
> 
> Older GCC compiler don’t understand this warning option and exits with
> error. There has to be a compiler recognition after that only GCC =>7.1.0
> receives this option.
> 
> Signed-off-by: Philipp Psurek <philipp.psurek@gmail.com>
> ---
>  Makefile | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/Makefile b/Makefile
> index 6bebb7d..29b7e2f 100755
> --- a/Makefile
> +++ b/Makefile
> @@ -101,6 +101,11 @@ MKDIR ?= mkdir -p
>  COMPILE.c = $(Q_CC)$(CC) $(CFLAGS) $(CPPFLAGS) $(TARGET_ARCH) -c
>  LINK.o = $(Q_LD)$(CC) $(CFLAGS) $(LDFLAGS) $(TARGET_ARCH)
> 
> +# Check for GCC >=7
> +ifeq ($(shell $(CC) -x c++ --std=c++17 -E -P - <<< __cplusplus),201703L)
> +       CFLAGS += -Wimplicit-fallthrough=2
> +endif
> +

Hmm, I don't like this one too much, since it looks cryptic and we will have 
to add up new versions, which I'd like to avoid.

Did you try to change the wording? I've googled a bit and it seems a comment 
containing the wording "fallthrough" will avoid it. Maybe we should just 
change "fall through" to "fallthrough"?

Thanks,
      Simon

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]