On 04/05/2018 03:00 PM, Jakub Jelinek wrote:
> On Thu, Apr 05, 2018 at 02:51:22PM +0200, Martin Liška wrote:
>> On 04/04/2018 09:31 PM, Marek Polacek wrote:
>>> On Tue, Apr 03, 2018 at 02:29:37PM +0200, Martin Liška wrote:
>>>> Hi.
>>>>
>>>> This helps the warning with --save-temps. Doing that one needs to preserve
>>>> comments
>>>> in preprocessed source file.
>>>
>>> Do we really want to only use -C when -Wimplicit-fallthrough is in effect?
>>> I
>>> mean, shouldn't we always use -C when -save-temps?
>>
>> Why not, Jakub what do you think? Note that it was originally Jakub's idea
>> to do that.
>
> I'd prefer to do that only when we actually care about the comments, it is a
> behavior change in any case, and might be undesirable to some people.
>
> Note that we do not care about the comments if -Wimplicit-fallthrough=0
> or -Wimplicit-fallthrough=5, but do care for:
> -Wimplicit-fallthrough
> -Wimplicit-fallthrough=1
> -Wimplicit-fallthrough=2
> -Wimplicit-fallthrough=3
> -Wimplicit-fallthrough=4
> -Wextra
> -W
Or you can trigger the warning via
-Werror=implicit-fallthrough*
which would complicate the option matching.
> (unless -Wno-implicit-fallthrough). So, it would be desirable to:
> 1) swap the order, put save-temps to the outer level
> 2) use
> {Wimplicit-fallthrough*:{!Wimplicit-fallthrough=0:{!Wimplicit-fallthrough=5:...}}}
> 3) verify (including adding testcases) that it doesn't emit comments for the
> =0, =5 or -W -Wno-implicit-fallthrough cases, but does for -W etc.
That would also complicate the exclusion of negative forms.
Sigh. I'm thinking if really worth it..
Martin
>
> Jakub
>
