[PATCH 5/6] winevulkan: Support using __wine_unix_call interface for Vulkan functions.
Signed-off-by: Jacek Caban <jacek(a)codeweavers.com> --- vkEndCommandBuffer is used just as an example. dlls/winevulkan/make_vulkan | 152 +++++++++++++++++++++++++++++++----- 1 file changed, 131 insertions(+), 21 deletions(-)
On 09.12.21 03:16, Jacek Caban wrote:
Signed-off-by: Jacek Caban <jacek(a)codeweavers.com> --- vkEndCommandBuffer is used just as an example.
dlls/winevulkan/make_vulkan | 152 +++++++++++++++++++++++++++++++----- 1 file changed, 131 insertions(+), 21 deletions(-)
Hi Jacek, why did you decide to introduce the wrap_vkX functions on the unix side instead of modifying the functions in vulkan.c directly? Also, I'm not particularly a fan of copy-pasting needs_alignment into VkParam from VkMember, but I'm not sure if there is a better solution.
On 12/9/21 12:52 PM, Georg Lehmann wrote:
On 09.12.21 03:16, Jacek Caban wrote:
Signed-off-by: Jacek Caban <jacek(a)codeweavers.com> --- vkEndCommandBuffer is used just as an example.
dlls/winevulkan/make_vulkan | 152 +++++++++++++++++++++++++++++++----- 1 file changed, 131 insertions(+), 21 deletions(-)
Hi Jacek,
why did you decide to introduce the wrap_vkX functions on the unix side instead of modifying the functions in vulkan.c directly?
That abstracts things from vulkan.c a bit more. During mu experimentations with those patches, I found it useful to be able to change how particular functions (or sets of functions) are handled by simply changing one make_vulkan line. Now that I think of it, I could probably structure the series better and use those structs even for direct calls. I will revisit it.
Also, I'm not particularly a fan of copy-pasting needs_alignment into VkParam from VkMember, but I'm not sure if there is a better solution.
Yeah, I'm not a fan neither. I was looking at leveraging VkMember for that, but it doesn't seem to be meant to be used that way. Another solution would be to introduce a separate class, which would be shared between VkMember and VkParam, but I don't feel like needs_alignment is enough to justify such a large change. Thanks, Jacek
participants (2)
-
Georg Lehmann -
Jacek Caban