On 3/14/22 13:52, Francisco Casas wrote:
By the way, maybe we should also make hlsl_new_array_type() free the memory of the basic_type argument on failure. Also we are assuming that hlsl_new_array_type() can't fail in some parts of the code that call it. However, I suggest fixing that in the future, to avoid drowning in a glass of water now.
There's no need to do that; new types are stored in the "ctx->types" array, and will be freed when the context is.