I got frustrated trying to figure out the c-for-go side of things, so I wrote my own lightweight wrapper over the vulkan API. Instead of trying to reverse engineer the c header files, I started with their published XML specification. https://github.com/ibd1279/vks
Jmw On Saturday, November 13, 2021 at 10:42:24 PM UTC+1 Jason Watson wrote: > Does anyone know how to configure c-for-go to work with output function > parameters - specifically slices? > > I was trying to play around with adding some vulkan 1.1 and vulkan 1.2 > bindings to the vulkan-go bindings (github.com/vulkan-go/vulkan), but in > order to regenerate the bindings, I needed to use c-for-go ( > github.com/xlab/c-for-go). After struggling to get the local instance to > work, I realized that c-for-go is generating different helper functions > than when vulkan was previously checked-in. > > After digging a bit further, it looks like c-for-go now copies slices that > are "arr" hinted before passing them to C. As a result, a nil slice is > copied into an empty slice when crossing the membrane and not NULL. > Additionally, if the C function is expecting to populate the slice ( e.g. > https://www.khronos.org/registry/vulkan/specs/1.2-extensions/man/html/vkEnumeratePhysicalDevices.html > > ), the results are lost as they are populated into the copied slice, and > not copied back into the original slice. > > The only way I've been able to get it to work is by configuring it as a > pointer and removing the slice header before hand, but I was curious if > anyone else was using c-for-go and had other solutions or configuration > tricks that I don't know as a c-for-go newbie. > > Jmw > -- You received this message because you are subscribed to the Google Groups "golang-nuts" group. To unsubscribe from this group and stop receiving emails from it, send an email to golang-nuts+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/golang-nuts/dfd47bf1-4235-443c-8a85-46c77d88eba6n%40googlegroups.com.
