From 2891378bb17dbed110b391f9aab47d241413153d Mon Sep 17 00:00:00 2001
From: akleine <alb.kleine@gmx.de>
Date: Thu, 15 Oct 2026 16:26:27 +0200
Subject: [PATCH] fix: prevent core dump in CPU backend if using FP8 (e4m3)
 tensors (#2140)

---
 src/model/common/ggml_block.hpp | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/src/model/common/ggml_block.hpp b/src/model/common/ggml_block.hpp
index 1d35b417..ddf4a199 100644
--- a/src/model/common/ggml_block.hpp
+++ b/src/model/common/ggml_block.hpp
@@ -14,6 +14,7 @@
 
 #include "core/ggml_extend.h"
 #include "core/ggml_runner.h"
+#include "core/ggml_extend_backend.h"
 #include "model.h"
 
 class GGMLBlock {
@@ -222,7 +223,7 @@ public:
                 if (force_prec_f32) {
                     ggml_prec_set_acc(fp8_matmul, GGML_PREC_F32);
                 }
-                supports_fp8_matmul = ggml_backend_supports_op(ctx->backend, fp8_matmul);
+                supports_fp8_matmul = ggml_backend_supports_op(ctx->backend, fp8_matmul) && !sd_backend_is_cpu(ctx->backend);
             }
             if (!supports_fp8_matmul) {
                 w = ggml_cast(ctx->ggml_ctx, w, GGML_TYPE_BF16);