From 134b2bb756c984d4fd2cabc26335f8de56b12ee1 Mon Sep 17 00:00:00 2001
From: Yash Raj Pandey <55940078+devYRPauli@users.noreply.github.com>
Date: Fri, 2 Oct 2026 10:47:30 -0400
Subject: [PATCH] ggml-cuda : fix cpy transposed path corrupting non-contiguous
 dst (#27663)

---
 ggml/src/ggml-cuda/cpy.cu  | 3 ++-
 tests/test-backend-ops.cpp | 2 ++
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/ggml/src/ggml-cuda/cpy.cu b/ggml/src/ggml-cuda/cpy.cu
index 7a99845854..90a06b97bb 100644
--- a/ggml/src/ggml-cuda/cpy.cu
+++ b/ggml/src/ggml-cuda/cpy.cu
@@ -459,7 +459,8 @@ void ggml_cuda_cpy(ggml_backend_cuda_context & ctx, const ggml_tensor * src0, gg
 
     const bool contiguous_srcs = ggml_is_contiguous(src0) && ggml_is_contiguous(src1);
     const bool can_be_transposed = nb01 == (int64_t)ggml_element_size(src0) &&
-        src0->ne[3] == 1 && nb02 == ne00 * ne01 * (int64_t)ggml_element_size(src0);
+        src0->ne[3] == 1 && nb02 == ne00 * ne01 * (int64_t)ggml_element_size(src0) &&
+        ggml_is_contiguous(src1);
 
     size_t mc_width = 0, mc_height = 0, mc_spitch = 0, mc_dpitch = 0;
 
diff --git a/tests/test-backend-ops.cpp b/tests/test-backend-ops.cpp
index 2682d9c064..d725191b7c 100644
--- a/tests/test-backend-ops.cpp
+++ b/tests/test-backend-ops.cpp
@@ -9833,6 +9833,8 @@ static std::vector<std::unique_ptr<test_case>> make_test_cases_eval() {
     test_cases.emplace_back(new test_cpy(GGML_TYPE_F32, GGML_TYPE_F32, {2, 2, 524281, 1}, {-1,-1,-1,-1}, {1, 0, 2, 3}));
     test_cases.emplace_back(new test_cpy(GGML_TYPE_F32, GGML_TYPE_F32, {128, 2, 3, 1}, {128, 2, 3, 1}, {0, 0, 0, 0}, {0, 0, 0, 0}, false, {128, 4, 3, 1})); // strided dst
     test_cases.emplace_back(new test_cpy(GGML_TYPE_F16, GGML_TYPE_F16, {128, 2, 3, 1}, {128, 2, 3, 1}, {0, 0, 0, 0}, {0, 0, 0, 0}, false, {128, 4, 3, 1})); // strided dst
+    test_cases.emplace_back(new test_cpy(GGML_TYPE_F32, GGML_TYPE_F32, {64, 16, 1, 1}, {1, 64, 16, 1}, {1, 2, 0, 3}, {0, 0, 0, 0}, false, {32, 64, 16, 1})); // permuted src, strided dst
+    test_cases.emplace_back(new test_cpy(GGML_TYPE_F16, GGML_TYPE_F16, {64, 16, 1, 1}, {1, 64, 16, 1}, {1, 2, 0, 3}, {0, 0, 0, 0}, false, {32, 64, 16, 1})); // permuted src, strided dst
 
     // CPY - different src/dst shapes (reshaping via CPY)
     // Use permutations of {3, 5, 7, 32}. Total elements: 3*5*7*32 = 3360.