On Tue, Apr 16, 2019 at 12:55:03AM +0200, Carl Eugen Hoyos wrote:
> 2019-04-14 1:02 GMT+02:00, Michael Niedermayer <mich...@niedermayer.cc>:
> > On Sat, Apr 13, 2019 at 07:25:54PM -0300, James Almer wrote:
> >> On 4/13/2019 7:04 PM, Carl Eugen Hoyos wrote:
> >> > Hi!
> >> >
> >> > Attached patch silences two warnings shown when compiling with clang:
> >> > libavfilter/vf_fspp.h:51:42: warning: implicit conversion from 'int'
> >> > to 'int16_t' (aka 'short') changes value from 44130 to -21406
> >> >
> >> > Please comment, Carl Eugen
> >> >
> >> >
> >> > 0001-lavfi-vf_fspp-Use-the-intended-cast-to-int16_t.patch
> >> >
> >> > From d731e523d9c5854183d20d37fe921f49fb048498 Mon Sep 17 00:00:00 2001
> >> > From: Carl Eugen Hoyos <ceffm...@gmail.com>
> >> > Date: Sat, 13 Apr 2019 23:05:44 +0200
> >> > Subject: [PATCH] lavfi/fspp: Add two casts to int16_t.
> >> >
> >> > Silences a warning with clang:
> >> > warning: implicit conversion from 'int' to 'int16_t' (aka 'short')
> >> > changes value from 44130 to -21406
> >> > ---
> >> >  libavfilter/vf_fspp.h |    2 +-
> >> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >> >
> >> > diff --git a/libavfilter/vf_fspp.h b/libavfilter/vf_fspp.h
> >> > index 74a3447..e346439 100644
> >> > --- a/libavfilter/vf_fspp.h
> >> > +++ b/libavfilter/vf_fspp.h
> >> > @@ -31,7 +31,7 @@
> >> >  #define DCTSIZE 8
> >> >  #define DCTSIZE_S "8"
> >> >
> >> > -#define FIX(x,s)  ((int) ((x) * (1 << s) + 0.5) & 0xffff)
> >> > +#define FIX(x,s)  ((int16_t)((x) * (1 << s) + 0.5) & (int16_t)0xffff)
> >>
> >> The "& 0xffff" is turning -21406 into 44130 since it zeroes all the high
> >> bits, and then that value gets converted back to -21406 when stored into
> >> an int16_t variable.
> >>
> >> Change the int cast into int16_t if you want since it doesn't hurt (all
> >> the values are safely within the int16_t range anyway), but remove the
> >> "& 0xffff" altogether instead of casting it, to get rid of this warning.
> >>
> >> >  #define C64(x)    ((uint64_t)((x) | (x) << 16)) <<32 | (uint64_t)(x) |
> >> > (uint64_t)(x) << 16
> >> >  #define FIX64(x,s)  C64(FIX(x,s))
> >
> > i think the C64 / FIX64 macros depend on the & 0xFFFF. They are maybe
> > unused
> > and just referenced in comments but having them not give correct values
> > could still be confusing. The 0xFFFF if its removed should be added to the
> > code / comments that refer to it in a way that needs it
> 
> New patch attached.
> Not sure if / where to add comments.
> 
> Please comment, Carl Eugen

>  vf_fspp.h |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 17e468ee43988ef39d3ce882774dc3e2d625c9e6  
> 0001-lavfi-fspp-Cast-float-values-to-int16_t-instead-of-i.patch
> From dfda9d8a6714649015f252e0e0bf5e3e52ce1b5c Mon Sep 17 00:00:00 2001
> From: Carl Eugen Hoyos <ceffm...@gmail.com>
> Date: Tue, 16 Apr 2019 00:50:50 +0200
> Subject: [PATCH] lavfi/fspp: Cast float values to int16_t instead of int.
> 
> Silences a warning with clang:
> warning: implicit conversion from 'int' to 'int16_t' (aka 'short') changes 
> value from 44130 to -21406
> ---
>  libavfilter/vf_fspp.h |    2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/libavfilter/vf_fspp.h b/libavfilter/vf_fspp.h
> index 802db14..bdafe8e 100644
> --- a/libavfilter/vf_fspp.h
> +++ b/libavfilter/vf_fspp.h
> @@ -31,7 +31,7 @@
>  #define DCTSIZE 8
>  #define DCTSIZE_S "8"
>  
> -#define FIX(x,s)  ((int) ((x) * (1 << s) + 0.5) & 0xffff)
> +#define FIX(x,s)  ((int16_t)((x) * (1 << s) + 0.5))

does it need the cast at all to avoid the warning ?

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Many things microsoft did are stupid, but not doing something just because
microsoft did it is even more stupid. If everything ms did were stupid they
would be bankrupt already.

Attachment: signature.asc
Description: PGP signature

_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel

To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".

Reply via email to