On Tue, 25 Mar 2025, Zhao Zhili wrote:
On Mar 25, 2025, at 18:46, Martin Storsjö <mar...@martin.st> wrote:
On Tue, 25 Mar 2025, Zhao Zhili wrote:
在 2025年3月25日，下午6:01，Martin Storsjö <mar...@martin.st> 写道：
This error constant was only added in the macOS 12 and iOS 15
SDKs.
If building with an older SDK, assume that the error doesn't
match this constant.
This fixes build errors with older SDKs after
c6214b0d691566c7cb0f2ff5be08a24c3534e5bb.
As this constant is declared unconditionally (without any target
version guards) in the headers, one can't reproduce this issue
while building with a newer SDK and targeting an older version
with -mmacosx-version-min=.
Technically, checking just one of MAC_OS_VERSION_12_0 and
__IPHONE_15_0 is enough; the macOS SDKs and iOS SDKs share these
files, so files from a similar point in time do have the same
defines. Therefore, we don't really need to expand this with
checks for tvOS, watchOS and other OS variants.
---
libavcodec/videotoolbox.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/libavcodec/videotoolbox.c b/libavcodec/videotoolbox.c
index aaa6129576..7f7b910b26 100644
--- a/libavcodec/videotoolbox.c
+++ b/libavcodec/videotoolbox.c
@@ -728,8 +728,13 @@ static void videotoolbox_decoder_callback(void *opaque,
}
if (!image_buffer) {
+#if defined(MAC_OS_VERSION_12_0) || defined(__IPHONE_15_0)
+ // kVTVideoDecoderReferenceMissingErr was defined in the macOS 12/iOS
15 SDK
if (status != kVTVideoDecoderReferenceMissingErr)
vtctx->reconfig_needed = true;
+#else
+ vtctx->reconfig_needed = true;
+#endif
I used its numeric value initially, but after checking the doc,
kVTVideoDecoderReferenceMissingErr
iOS 6.0+
iPadOS 6.0+
Mac Catalyst 13.0+
macOS 10.8+
tvOS 10.2+
visionOS 1.0+
I replaced it with the enum.
Now I wondering what does this macOS 10.8 mean.
Anyway, can we use the numeric value here?
Can you link to which docs declare this to be available since macOS 10.8?
https://developer.apple.com/documentation/videotoolbox/kvtvideodecoderreferencemissingerr?language=objc
Thanks, that's interesting. Apparently they don't really track the
availability of the individual enum entries very thoroughly here...
// Martin
_______________________________________________
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".
