From ab7a7f9f1c97d99d5dd13da0ef371904c5cc5e6b Mon Sep 17 00:00:00 2001
From: Daniel Han <danielhanchen@gmail.com>
Date: Mon, 1 Jun 2026 02:55:57 -0700
Subject: [PATCH] Studio: match the Linux llama.cpp prebuilt to the runtime
 cudart major (#5914)

* Studio: match the Linux llama.cpp prebuilt to the runtime cudart major

* [pre-commit.ci] auto fixes from pre-commit.com hooks

for more information, see https://pre-commit.ci

---------

Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
---
 studio/install_llama_prebuilt.py              | 11 ++-
 .../test_install_llama_prebuilt_logic.py      | 74 +++++++++++++++++++
 2 files changed, 84 insertions(+), 1 deletion(-)

diff --git a/studio/install_llama_prebuilt.py b/studio/install_llama_prebuilt.py
index 2afd54bec9..77d0905020 100644
--- a/studio/install_llama_prebuilt.py
+++ b/studio/install_llama_prebuilt.py
@@ -1402,7 +1402,16 @@ def direct_linux_release_plan(
 
     attempts: list[AssetChoice] = []
     if host.has_usable_nvidia:
-        selection = linux_cuda_choice_from_release(host, bundle)
+        # Prefer the cudart major Studio loads at runtime (torch's bundled
+        # libcudart), not the newest detected on disk. Without this a stray
+        # cuda13 runtime outranks the torch cuda12 the binary links against.
+        torch_preference = detect_torch_cuda_runtime_preference(host)
+        selection = linux_cuda_choice_from_release(
+            host,
+            bundle,
+            preferred_runtime_line = torch_preference.runtime_line,
+            selection_preamble = torch_preference.selection_log,
+        )
         if selection is not None:
             attempts.extend(selection.attempts)
     if host.has_rocm and not host.has_usable_nvidia:
diff --git a/tests/studio/install/test_install_llama_prebuilt_logic.py b/tests/studio/install/test_install_llama_prebuilt_logic.py
index 49ad6da177..4a427d6f54 100644
--- a/tests/studio/install/test_install_llama_prebuilt_logic.py
+++ b/tests/studio/install/test_install_llama_prebuilt_logic.py
@@ -438,6 +438,80 @@ def test_simple_linux_direct_release_uses_published_source_checksums_for_branch(
     assert exact_source is True
 
 
+def test_simple_linux_direct_release_honors_torch_cudart_preference(
+    monkeypatch: pytest.MonkeyPatch,
+):
+    # Regression: a Blackwell host (sm_120, driver 13.0) with BOTH cudart majors
+    # visible -- a stray cuda13 wheel plus torch's cuda12 -- must install the
+    # cuda12 build that matches the runtime torch, not the newest-major cuda13
+    # build (which loads no GPU and silently falls back to CPU).
+    release = {
+        "tag_name": "b9334",
+        "assets": [
+            {
+                "name": f"app-b9334-linux-x64-{profile}.tar.gz",
+                "browser_download_url": f"https://example.test/app-b9334-linux-x64-{profile}.tar.gz",
+            }
+            for profile in (
+                "cuda12-newer",
+                "cuda12-portable",
+                "cuda13-newer",
+                "cuda13-portable",
+            )
+        ],
+    }
+    # cuda13 detected first (newest-major order); both compatible with driver 13.0.
+    monkeypatch.setattr(
+        INSTALL_LLAMA_PREBUILT,
+        "detected_linux_runtime_lines",
+        lambda: (
+            ["cuda13", "cuda12"],
+            {
+                "cuda13": ["/usr/local/lib/python3.13/site-packages/nvidia/cu13/lib"],
+                "cuda12": [
+                    "/venv/lib/python3.13/site-packages/nvidia/cuda_runtime/lib"
+                ],
+            },
+        ),
+    )
+    host = HostInfo(
+        system = "Linux",
+        machine = "x86_64",
+        is_windows = False,
+        is_linux = True,
+        is_macos = False,
+        is_x86_64 = True,
+        is_arm64 = False,
+        nvidia_smi = "nvidia-smi",
+        driver_cuda_version = (13, 0),
+        compute_caps = ["120"],
+        visible_cuda_devices = None,
+        has_physical_nvidia = True,
+        has_usable_nvidia = True,
+    )
+
+    def first_asset_for_torch(line):
+        monkeypatch.setattr(
+            INSTALL_LLAMA_PREBUILT,
+            "detect_torch_cuda_runtime_preference",
+            lambda h: INSTALL_LLAMA_PREBUILT.CudaRuntimePreference(
+                runtime_line = line, selection_log = []
+            ),
+        )
+        plan = INSTALL_LLAMA_PREBUILT.direct_linux_release_plan(
+            release, host, "unslothai/llama.cpp", "latest"
+        )
+        return plan.attempts[0]
+
+    # torch reports cuda12 (the cu128 runtime) -> install the cuda12 build.
+    primary = first_asset_for_torch("cuda12")
+    assert primary.name == "app-b9334-linux-x64-cuda12-newer.tar.gz"
+    assert primary.runtime_line == "cuda12"
+
+    # torch unavailable -> unchanged newest-major fallback (documents the residual).
+    assert first_asset_for_torch(None).name == "app-b9334-linux-x64-cuda13-newer.tar.gz"
+
+
 @pytest.mark.parametrize(
     "mutate, expected_match",
     [