Dan Albert
Dan Albert
Okay, so the default configuration of ffmpeg (that is, no `--with-feature` flags when you run `configure`) _does_ require the updated header, even if vulkan won't be used at runtime? I'm...
Yes, up to date Vulkan headers are of course something we want :) No argument there. It does sound like ffmpeg's autodetection is wrong though. Presence of those headers has...
It looks like it's trying to do that, actually: ``` if enabled vulkan; then check_pkg_config_header_only vulkan "vulkan >= 1.3.277" "vulkan/vulkan.h" "defined VK_VERSION_1_3" || check_cpp_condition vulkan "vulkan/vulkan.h" "defined(VK_VERSION_1_4) || (defined(VK_VERSION_1_3) &&...
I'm still not totally sure what this should look like. Currently I think the best path is to ship an Android.cmake module that includes helper functions for the NDK. One...
Thanks for looking at this. I'm *hoping* that we can make the interface for these more or less identical to how sanitizers are used on any other platform. For example:...
Probably something we'll tackle post (or as part of) the CMake implementation cleanup. Don't know exactly which release that will be, so moving into the unplanned bucket (it's planned, just...
No update. The CMake cleanup I mention above hasn't happened yet. It _might_ be something we can start on this year, but there's other work that needs to be finished...
> I hope it will be resolved before NDK R18 released. If we find out that we're just building something wrong then that's possible, but if it's something that will...
As far as I know this has not been resolved upstream, so moving to r20.
Still no upstream progress.