Re: [PATCH] Fixes compile warning with -Wimplicit-fallthrough CFLAGS
On Tue, Jan 23, 2018 at 10:33:57AM -0800, Junio C Hamano wrote:
> Jeff King writes:
>
> >> diff --git a/apply.c b/apply.c
> >> index 321a9fa68..a22fb2881 100644
> >> --- a/apply.c
> >> +++ b/apply.c
> >> @@ -1450,7 +1450,7 @@ static void recount_diff(const char *line, int size,
> >> struct fragment *fragment)
> >>switch (*line) {
> >>case ' ': case '\n':
> >>newlines++;
> >> - /* fall through */
> >> + GIT_FALLTHROUGH;
> >
> > Ugh, the semi-colon there makes it look like it's actual code. If we go
> > this route, I wonder if it's worth hiding it inside the macro.
>
> What? You mean to shout in all caps without even terminating the
> line with any punctuation? Please don't---I am sure it will break
> auto indentation people rely on from their editors.
True, that may be even worse. I just wonder if we can do something to
make it look more obviously like a non-code attribute. The actual syntax
is something like:
[[fallthrough]];
which is pretty horrid, but at least a bit easier to see. gcc also
provides "__attribute__((fallthrough))", but I don't think it works with
clang.
I vastly prefer the comment approach if we can use it. Apparently clang
doesn't support it, but I have also not managed to get clang (either
version 4, 6, or the upcoming 7) to actually report anything via
-Wimplicit-fallthrough, either. Maybe I'm holding it wrong.
-Peff
Re: [PATCH] Fixes compile warning with -Wimplicit-fallthrough CFLAGS
Jeff King writes:
>> diff --git a/apply.c b/apply.c
>> index 321a9fa68..a22fb2881 100644
>> --- a/apply.c
>> +++ b/apply.c
>> @@ -1450,7 +1450,7 @@ static void recount_diff(const char *line, int size,
>> struct fragment *fragment)
>> switch (*line) {
>> case ' ': case '\n':
>> newlines++;
>> -/* fall through */
>> +GIT_FALLTHROUGH;
>
> Ugh, the semi-colon there makes it look like it's actual code. If we go
> this route, I wonder if it's worth hiding it inside the macro.
What? You mean to shout in all caps without even terminating the
line with any punctuation? Please don't---I am sure it will break
auto indentation people rely on from their editors.
Re: [PATCH] Fixes compile warning with -Wimplicit-fallthrough CFLAGS
On Mon, Jan 22, 2018 at 4:59 PM, Jeff King wrote:
> On Mon, Jan 22, 2018 at 07:54:18PM -0500, Eric Sunshine wrote:
>
>> On Mon, Jan 22, 2018 at 6:51 PM, Elia Pinto wrote:
>> > This patch add explicit fallthrough compiler attribute
>> > when needed on switch case statement eliminating
>> > the compile warning [-Werror=implicit-fallthrough=].
>> > It does this by means of a macro that takes into account
>> > the versions of the compilers that include that attribute.
>> > [...]
>> > Signed-off-by: Elia Pinto
>> > ---
>> > diff --git a/convert.c b/convert.c
>> > @@ -1554,7 +1554,7 @@ static int ident_filter_fn(struct stream_filter
>> > *filter,
>> > switch (ident->state) {
>> > default:
>> > strbuf_add(&ident->left, head, ident->state);
>> > - /* fallthrough */
>> > + GIT_FALLTHROUGH;
>> > case IDENT_SKIPPING:
>> > /* fallthrough */
>>
>> Why doesn't this /* fallthrough */ deserve the same treatment?
>>
>> > case IDENT_DRAINING:
>
> I can't answer that philosophically, but I can tell you why the compiler
> does not complain. :)
>
> Usually case arms with no statements between them are exempt from
> fallthrough warnings. So:
>
> switch (foo)
> case 1:
> case 2:
> case 3:
> /* do one thing */
> break;
> case 4:
> /* do another thing */
> break;
> }
>
> does not need any annotations for cases 1 and 2 to fallthrough. Which
> means that the original comment was not actually necessary for the
> compiler, though the original author considered it a useful comment.
>
> So there you get into philosophy. Should it be converted to a
> compiler-visible annotation, or is it better left as a comment?
>
> -Peff
I'd personally rather stick to the comment if we can, or use something
like "fallthrough;" to make it appear like a keyword, instead of an
all caps macro, since at least to my sensibility, the all caps is a
bit too crazy.
Also, I would not put one inside an empty case statement that just
falls through to the next branch and does nothing special itself.
Thanks,
Jake
Re: [PATCH] Fixes compile warning with -Wimplicit-fallthrough CFLAGS
On Mon, Jan 22, 2018 at 07:54:18PM -0500, Eric Sunshine wrote:
> On Mon, Jan 22, 2018 at 6:51 PM, Elia Pinto wrote:
> > This patch add explicit fallthrough compiler attribute
> > when needed on switch case statement eliminating
> > the compile warning [-Werror=implicit-fallthrough=].
> > It does this by means of a macro that takes into account
> > the versions of the compilers that include that attribute.
> > [...]
> > Signed-off-by: Elia Pinto
> > ---
> > diff --git a/convert.c b/convert.c
> > @@ -1554,7 +1554,7 @@ static int ident_filter_fn(struct stream_filter
> > *filter,
> > switch (ident->state) {
> > default:
> > strbuf_add(&ident->left, head, ident->state);
> > - /* fallthrough */
> > + GIT_FALLTHROUGH;
> > case IDENT_SKIPPING:
> > /* fallthrough */
>
> Why doesn't this /* fallthrough */ deserve the same treatment?
>
> > case IDENT_DRAINING:
I can't answer that philosophically, but I can tell you why the compiler
does not complain. :)
Usually case arms with no statements between them are exempt from
fallthrough warnings. So:
switch (foo)
case 1:
case 2:
case 3:
/* do one thing */
break;
case 4:
/* do another thing */
break;
}
does not need any annotations for cases 1 and 2 to fallthrough. Which
means that the original comment was not actually necessary for the
compiler, though the original author considered it a useful comment.
So there you get into philosophy. Should it be converted to a
compiler-visible annotation, or is it better left as a comment?
-Peff
Re: [PATCH] Fixes compile warning with -Wimplicit-fallthrough CFLAGS
On Mon, Jan 22, 2018 at 6:51 PM, Elia Pinto wrote:
> This patch add explicit fallthrough compiler attribute
> when needed on switch case statement eliminating
> the compile warning [-Werror=implicit-fallthrough=].
> It does this by means of a macro that takes into account
> the versions of the compilers that include that attribute.
> [...]
> Signed-off-by: Elia Pinto
> ---
> diff --git a/convert.c b/convert.c
> @@ -1554,7 +1554,7 @@ static int ident_filter_fn(struct stream_filter *filter,
> switch (ident->state) {
> default:
> strbuf_add(&ident->left, head, ident->state);
> - /* fallthrough */
> + GIT_FALLTHROUGH;
> case IDENT_SKIPPING:
> /* fallthrough */
Why doesn't this /* fallthrough */ deserve the same treatment?
> case IDENT_DRAINING:
Re: [PATCH] Fixes compile warning with -Wimplicit-fallthrough CFLAGS
On Mon, Jan 22, 2018 at 11:51:18PM +, Elia Pinto wrote:
> This patch add explicit fallthrough compiler attribute
> when needed on switch case statement eliminating
> the compile warning [-Werror=implicit-fallthrough=].
> It does this by means of a macro that takes into account
> the versions of the compilers that include that attribute.
>
> The fallthrough (or clang::fallthrough) attribute is used to annotate
> intentional fall-through between switch labels.
> Traditionally these are marked with a specific comment, but
> this attribute is meant to replace comments with a more strict
> annotation, which can be checked by the compiler (gcc-7 or clang).
> The flags in question were introduced in gcc 7 and are also enabled
> with -Wextra.
Hrm. Your subject says "fixes compile warnings", but don't we already
compile cleanly with -Wimplicit-fallthrough after my 1cf01a34ea
(consistently use "fallthrough" comments in switches, 2017-09-21)?
Certainly the tip of "master" seems to pass for me on both gcc 7 and
clang 4. You can pump the warning up to level 5 on gcc to insist on the
attribute, but I think the comments are more readable (and it is not
like we have a problem with false positive comments).
> It would also have been possible to introduce a specific comment
> accepted by gcc 7 instead of the fallthrough attribute for this warning,
> but clang does not have a similar implementation. The macro replaces
> the previous, not uniform, comments and can acts as a comment itself.
Interestingly clang seems to accept -Wimplicit-fallthrough, but I could
not get it to actually trigger a warning, even after removing some of
the existing comments.
What version of clang are you using? I'm certainly puzzled by the
behavior I'm seeing.
> diff --git a/apply.c b/apply.c
> index 321a9fa68..a22fb2881 100644
> --- a/apply.c
> +++ b/apply.c
> @@ -1450,7 +1450,7 @@ static void recount_diff(const char *line, int size,
> struct fragment *fragment)
> switch (*line) {
> case ' ': case '\n':
> newlines++;
> - /* fall through */
> + GIT_FALLTHROUGH;
Ugh, the semi-colon there makes it look like it's actual code. If we go
this route, I wonder if it's worth hiding it inside the macro.
-Peff
