Refine handling of shim presence

This allows the CPU only builds to work on systems with Radeon cards

llm/llm.go

@@ -22,6 +22,9 @@ type LLM interface {
 	Close()
 }
 
+// Set to false on linux/windows if we are able to load the shim
+var ShimPresent = false
+
 func New(workDir, model string, adapters, projectors []string, opts api.Options) (LLM, error) {
 	if _, err := os.Stat(model); err != nil {
 		return nil, err
@@ -79,11 +82,10 @@ func New(workDir, model string, adapters, projectors []string, opts api.Options)
 	opts.RopeFrequencyBase = 0.0
 	opts.RopeFrequencyScale = 0.0
 	gpuInfo := gpu.GetGPUInfo()
-	switch gpuInfo.Driver {
-	case "ROCM":
+	if gpuInfo.Driver == "ROCM" && ShimPresent {
 		return newRocmShimExtServer(model, adapters, projectors, ggml.NumLayers(), opts)
-	default:
-		// Rely on the built-in CUDA based server which will fall back to CPU
+	} else {
+		// Rely on the built-in CUDA/Metal based server which will fall back to CPU
 		return newLlamaExtServer(model, adapters, projectors, ggml.NumLayers(), opts)
 	}
 }

llm/shim_ext_server.go

@@ -30,7 +30,6 @@ import (
 var libEmbed embed.FS
 
 var RocmShimMissing = fmt.Errorf("ROCm shim library not included in this build of ollama. Radeon GPUs are not supported")
-var NoShim = true
 
 type shimExtServer struct {
 	s       C.struct_rocm_llama_server
@@ -78,7 +77,7 @@ func (llm *shimExtServer) llama_server_release_json_resp(json_resp **C.char) {
 }
 
 func newRocmShimExtServer(model string, adapters, projectors []string, numLayers int64, opts api.Options) (extServer, error) {
-	if NoShim {
+	if !ShimPresent {
 		return nil, RocmShimMissing
 	}
 	log.Printf("Loading ROCM llm server")
@@ -207,6 +206,6 @@ func extractLib(workDir string) error {
 	case err != nil:
 		return fmt.Errorf("stat ROCm shim %s: %v", files[0], err)
 	}
-	NoShim = false
+	ShimPresent = true
 	return nil
 }