Message ID | 20180404010931.GA30291@embeddedor.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Quoting Gustavo A. R. Silva (2018-04-03 18:09:31) > In preparation to enabling -Wimplicit-fallthrough, mark switch cases > where we are expecting to fall through. > > Notice that in this particular case I replaced the comma wiht a dash, s/wiht/with/ > which is what GCC is expecting to find when there are other comments > in the same line in which "fall through" appears. > > Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> > --- > This code was compiled with GCC 7.3.0 What's the GCC warning? Please include in the commit text. Also what's the implicit-fallthrough level you're using? 3? Can you file a bug so GCC can update their regex to accept a comma in addition to whitespace characters? And maybe we can change the level to 2 or 1 in the kernel's makefile so we don't need to patch comments for this.
On 04/05/2018 03:24 PM, Stephen Boyd wrote: > Quoting Gustavo A. R. Silva (2018-04-03 18:09:31) >> In preparation to enabling -Wimplicit-fallthrough, mark switch cases >> where we are expecting to fall through. >> >> Notice that in this particular case I replaced the comma wiht a dash, > > s/wiht/with/ > Thanks for this. >> which is what GCC is expecting to find when there are other comments >> in the same line in which "fall through" appears. >> >> Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> >> --- >> This code was compiled with GCC 7.3.0 > > What's the GCC warning? Please include in the commit text. > drivers/clk/samsung/clk-s3c2443.c:430:3: warning: this statement may fall through [-Wimplicit-fallthrough=] samsung_clk_register_alias(ctx, s3c2450_aliases, ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ARRAY_SIZE(s3c2450_aliases)); ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ drivers/clk/samsung/clk-s3c2443.c:433:2: note: here case S3C2416: ^~~~ > Also what's the implicit-fallthrough level you're using? 3? Can you file > a bug so GCC can update their regex to accept a comma in addition to > whitespace characters? > Yeah, I can talk to the GCC guys about that. > And maybe we can change the level to 2 or 1 in the kernel's makefile so > we don't need to patch comments for this. > Yeah, with level 2 I don't get that warning. Thanks -- Gustavo
diff --git a/drivers/clk/samsung/clk-s3c2443.c b/drivers/clk/samsung/clk-s3c2443.c index c46e6d5..7ebb87f 100644 --- a/drivers/clk/samsung/clk-s3c2443.c +++ b/drivers/clk/samsung/clk-s3c2443.c @@ -429,7 +429,7 @@ void __init s3c2443_common_clk_init(struct device_node *np, unsigned long xti_f, ARRAY_SIZE(s3c2450_gates)); samsung_clk_register_alias(ctx, s3c2450_aliases, ARRAY_SIZE(s3c2450_aliases)); - /* fall through, as s3c2450 extends the s3c2416 clocks */ + /* fall through - as s3c2450 extends the s3c2416 clocks */ case S3C2416: samsung_clk_register_div(ctx, s3c2416_dividers, ARRAY_SIZE(s3c2416_dividers));
In preparation to enabling -Wimplicit-fallthrough, mark switch cases where we are expecting to fall through. Notice that in this particular case I replaced the comma wiht a dash, which is what GCC is expecting to find when there are other comments in the same line in which "fall through" appears. Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> --- This code was compiled with GCC 7.3.0 drivers/clk/samsung/clk-s3c2443.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)