vulkan: allow graphics queue only through env var (ggml-org#20599)

* vulkan: avoid graphics queue on non-RADV AMD drivers

* avoid graphics queues on small GPUs

* change to only use graphics queue if overridden with env var GGML_VK_ALLOW_GRAPHICS_QUEUE

* reenable transfer queue if graphics queue is not used

Author: 0cc4m

ggml/src/ggml-vulkan/ggml-vulkan.cpp

@@ -4981,8 +4981,9 @@ static vk_device ggml_vk_get_device(size_t idx) {
         std::vector<vk::QueueFamilyProperties> queue_family_props = device->physical_device.getQueueFamilyProperties();
 
         // Try to find a non-graphics compute queue and transfer-focused queues
-        // On AMD, the graphics queue seems to be faster, so don't avoid it
-        const vk::QueueFlagBits graphics_flag = device->vendor_id == VK_VENDOR_ID_AMD ? (vk::QueueFlagBits)0 : vk::QueueFlagBits::eGraphics;
+        // Allow overriding avoiding the graphics queue because it can increase performance on RADV
+        const bool allow_graphics_queue = (getenv("GGML_VK_ALLOW_GRAPHICS_QUEUE") != nullptr);
+        const vk::QueueFlagBits graphics_flag = allow_graphics_queue ? (vk::QueueFlagBits)0 : vk::QueueFlagBits::eGraphics;
         const uint32_t compute_queue_family_index = ggml_vk_find_queue_family_index(queue_family_props, vk::QueueFlagBits::eCompute, graphics_flag, -1, 1);
         const uint32_t transfer_queue_family_index = ggml_vk_find_queue_family_index(queue_family_props, vk::QueueFlagBits::eTransfer, vk::QueueFlagBits::eCompute | graphics_flag, compute_queue_family_index, 1);
 
@@ -5443,11 +5444,14 @@ static vk_device ggml_vk_get_device(size_t idx) {
 
         ggml_vk_load_shaders(device);
 
+        // Only use transfer queue on AMD non-GCN, when the graphics queue is not enabled
+        const bool prefers_transfer_queue = device->vendor_id == VK_VENDOR_ID_AMD && device->architecture != AMD_GCN && !allow_graphics_queue;
+
         if (!device->single_queue) {
             const uint32_t transfer_queue_index = compute_queue_family_index == transfer_queue_family_index ? 1 : 0;
             ggml_vk_create_queue(device, device->transfer_queue, transfer_queue_family_index, transfer_queue_index, { vk::PipelineStageFlagBits::eTransfer }, true);
 
-            device->async_use_transfer_queue = (getenv("GGML_VK_ASYNC_USE_TRANSFER_QUEUE") != nullptr);
+            device->async_use_transfer_queue = prefers_transfer_queue || (getenv("GGML_VK_ASYNC_USE_TRANSFER_QUEUE") != nullptr);
         } else {
             // TODO: Use pointer or reference to avoid copy
             device->transfer_queue.copyFrom(device->compute_queue);