Re: [FFmpeg-devel] [PATCH v8 06/18] swscale: add SWS_UNSTABLE flag

From: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
To: ffmpeg-devel@ffmpeg.org
Subject: Re: [FFmpeg-devel] [PATCH v8 06/18] swscale: add SWS_UNSTABLE flag
Date: Sun, 13 Jul 2025 22:05:21 +0200
Message-ID: <GV1P250MB07370EAE6808B1166FB7839E8F55A@GV1P250MB0737.EURP250.PROD.OUTLOOK.COM> (raw)
In-Reply-To: <20250712104459.52153-6-ffmpeg@haasn.xyz>

Niklas Haas:
> From: Niklas Haas <git@haasn.dev>
> 
> Give users and developers a way to opt in to the new format conversion code,
> and more code from the swscale rewrite in general, even while development is
> still ongoing.
> ---
>  doc/APIchanges       | 3 +++
>  doc/scaler.texi      | 4 ++++
>  libswscale/options.c | 1 +
>  libswscale/swscale.h | 7 +++++++
>  libswscale/version.h | 2 +-
>  5 files changed, 16 insertions(+), 1 deletion(-)
> 
> diff --git a/doc/APIchanges b/doc/APIchanges
> index 6018da37e1..978af2bd50 100644
> --- a/doc/APIchanges
> +++ b/doc/APIchanges
> @@ -2,6 +2,9 @@ The last version increases of all libraries were on 2025-03-28
>  
>  API changes, most recent first:
>  
> +2025-07-xx - xxxxxxxxxx - lsws 9.1.100 - swscale.h
> +  Add SWS_UNSTABLE flag.
> +
>  2025-07-07 - eca477da52 - lavc 62.6.100 - packet.h
>    Add AV_PKT_DATA_RTCP_SR.
>  
> diff --git a/doc/scaler.texi b/doc/scaler.texi
> index a9ed10e34e..0c2b952395 100644
> --- a/doc/scaler.texi
> +++ b/doc/scaler.texi
> @@ -68,6 +68,10 @@ Select full chroma input.
>  
>  @item bitexact
>  Enable bitexact output.
> +
> +@item unstable
> +Allow the use of experimental new code. May subtly affect the output or even
> +produce wrong results. For testing only.
>  @end table
>  
>  @item srcw @var{(API only)}
> diff --git a/libswscale/options.c b/libswscale/options.c
> index feecae8c89..06e51dcfe9 100644
> --- a/libswscale/options.c
> +++ b/libswscale/options.c
> @@ -50,6 +50,7 @@ static const AVOption swscale_options[] = {
>          { "full_chroma_inp", "full chroma input",             0,  AV_OPT_TYPE_CONST, { .i64 = SWS_FULL_CHR_H_INP }, .flags = VE, .unit = "sws_flags" },
>          { "bitexact",        "bit-exact mode",                0,  AV_OPT_TYPE_CONST, { .i64 = SWS_BITEXACT       }, .flags = VE, .unit = "sws_flags" },
>          { "error_diffusion", "error diffusion dither",        0,  AV_OPT_TYPE_CONST, { .i64 = SWS_ERROR_DIFFUSION}, .flags = VE, .unit = "sws_flags" },
> +        { "unstable",        "allow experimental new code",   0,  AV_OPT_TYPE_CONST, { .i64 = SWS_UNSTABLE       }, .flags = VE, .unit = "sws_flags" },
>  
>      { "param0",          "scaler param 0", OFFSET(scaler_params[0]), AV_OPT_TYPE_DOUBLE, { .dbl = SWS_PARAM_DEFAULT  }, INT_MIN, INT_MAX, VE },
>      { "param1",          "scaler param 1", OFFSET(scaler_params[1]), AV_OPT_TYPE_DOUBLE, { .dbl = SWS_PARAM_DEFAULT  }, INT_MIN, INT_MAX, VE },
> diff --git a/libswscale/swscale.h b/libswscale/swscale.h
> index b04aa182d2..4aa072009c 100644
> --- a/libswscale/swscale.h
> +++ b/libswscale/swscale.h
> @@ -155,6 +155,13 @@ typedef enum SwsFlags {
>      SWS_ACCURATE_RND   = 1 << 18,
>      SWS_BITEXACT       = 1 << 19,
>  
> +    /**
> +     * Allow using experimental new code paths. This may be faster, slower,
> +     * or produce different output, with semantics subject to change at any
> +     * point in time. For testing and debugging purposes only.
> +     */
> +    SWS_UNSTABLE = 1 << 20,
> +
>      /**
>       * Deprecated flags.
>       */
> diff --git a/libswscale/version.h b/libswscale/version.h
> index 148efd83eb..4e54701aba 100644
> --- a/libswscale/version.h
> +++ b/libswscale/version.h
> @@ -28,7 +28,7 @@
>  
>  #include "version_major.h"
>  
> -#define LIBSWSCALE_VERSION_MINOR   0
> +#define LIBSWSCALE_VERSION_MINOR   1
>  #define LIBSWSCALE_VERSION_MICRO 100
>  
>  #define LIBSWSCALE_VERSION_INT  AV_VERSION_INT(LIBSWSCALE_VERSION_MAJOR, \

This patchset adds 214992B of .text, 6304B of .rodata, 42176B of
.data.rel.ro for an opt-in feature. There should be a configure option
to disable the new additions (i.e. the rest of the patchset) from being
compiled in for users that know that they will not use it.

- Andreas

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

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