Commit 711f60bee for llama.cpp

commit 711f60beeb9e983f6d8e01bda25302dee9226c8d
Author: Mostafa <mostafaaafaheem@gmail.com>
Date:   Mon Sep 21 13:38:48 2026 +0300

    tests : remove stale comment (#29140)

diff --git a/tests/test-backend-ops.cpp b/tests/test-backend-ops.cpp
index 5c7de196d..887add245 100644
--- a/tests/test-backend-ops.cpp
+++ b/tests/test-backend-ops.cpp
@@ -6219,13 +6219,6 @@ struct test_conv_2d : public test_case {
     // Whether the inputs are contiguous in the channel dim or the width dim
     const bool                   cwhn;

-    // If true, the direct CONV_2D will be used in the graph, otherwise it
-    // uses ggml_conv_2d:
-    // * if the program is called with -o CONV_2D_DIRECT_IMPL, the
-    // CONV_2D graph will be built, while
-    // * if the program is called with -o CONV_2D_INDIRECT_IMPL, the
-    // IM2COL -> MUL_MM graph will be built.
-
     std::string vars() override {
         return VARS_TO_STR10(ne_input, ne_kernel, type_kernel, stride0, stride1, padding0, padding1, dilation0, dilation1, cwhn);
     }