From b9da3089cfd6fdec6dc425cbd5f4d4453e3a63ba Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ole=20Sch=C3=BCtt?= <ole.schuett@cp2k.org>
Date: Thu, 2 Oct 2025 13:44:02 +0200
Subject: [PATCH] Move __DATA_DIR to C code to avoid Fortran's line length
 limit

---
 cmake/CompilerConfiguration.cmake |  2 +-
 src/CMakeLists.txt                |  9 +++++++-
 src/common/cp_data_dir.c          | 26 ++++++++++++++++++++++
 src/common/cp_files.F             | 36 +++++++++++++++++++++----------
 4 files changed, 60 insertions(+), 13 deletions(-)
 create mode 100644 src/common/cp_data_dir.c

diff --git a/cmake/CompilerConfiguration.cmake b/cmake/CompilerConfiguration.cmake
index 6469558f56..1a5631d954 100644
--- a/cmake/CompilerConfiguration.cmake
+++ b/cmake/CompilerConfiguration.cmake
@@ -35,7 +35,7 @@ endif()
 
 # Baseline
 add_compile_options(
-  "$<$<COMPILE_LANG_AND_ID:Fortran,GNU>:-std=f2008;-ffree-form;-ffree-line-length-none;-fimplicit-none>"
+  "$<$<COMPILE_LANG_AND_ID:Fortran,GNU>:-std=f2008;-ffree-form;-fimplicit-none>"
   "$<$<COMPILE_LANG_AND_ID:Fortran,GNU>:-g;-fno-omit-frame-pointer;-fbacktrace>"
   "$<$<COMPILE_LANG_AND_ID:Fortran,GNU>:$<$<VERSION_GREATER_EQUAL:${CMAKE_Fortran_COMPILER_VERSION},11>:-fallow-argument-mismatch>>"
   "$<$<COMPILE_LANG_AND_ID:Fortran,GNU>:-Wno-deprecated-declarations;-Wno-maybe-uninitialized;-Wuninitialized;-Wuse-without-only>"
diff --git a/src/CMakeLists.txt b/src/CMakeLists.txt
index d2baf2a32b..227091d670 100644
--- a/src/CMakeLists.txt
+++ b/src/CMakeLists.txt
@@ -1385,7 +1385,14 @@ list(
   xc/xc_xpbe_hole_t_c_lr.F
   xc/xc_xwpbe.F)
 
-list(APPEND CP2K_SRCS_C sockets.c base/machine_cpuid.c base/openmp_trace.c)
+list(
+  APPEND
+  CP2K_SRCS_C
+  sockets.c
+  base/machine_cpuid.c
+  base/openmp_trace.c
+  base/openmp_trace.c
+  common/cp_data_dir.c)
 
 set(CP2K_OFFLOAD_SRCS_C offload/offload_buffer.c offload/offload_library.c
                         offload/offload_mempool.c)
diff --git a/src/common/cp_data_dir.c b/src/common/cp_data_dir.c
new file mode 100644
index 0000000000..955888bfba
--- /dev/null
+++ b/src/common/cp_data_dir.c
@@ -0,0 +1,26 @@
+/*----------------------------------------------------------------------------*/
+/*  CP2K: A general program to perform molecular dynamics simulations         */
+/*  Copyright 2000-2025 CP2K developers group <https://cp2k.org>              */
+/*                                                                            */
+/*  SPDX-License-Identifier: GPL-2.0-or-later                                 */
+/*----------------------------------------------------------------------------*/
+
+#include <stdlib.h>
+
+// Since __DATA_DIR can be arbitrarily long we must not let the preprocessor
+// expand that macro in Fortran code as it could exceed the line length limit.
+static const char *data_dir = __DATA_DIR;
+
+/*******************************************************************************
+ * \brief Returns path of data directory if set, otherwise an empty string.
+ * \author Ole Schuett
+ ******************************************************************************/
+const char *get_data_dir() {
+  const char *overwrite = getenv("CP2K_DATA_DIR");
+  if (overwrite != NULL) {
+    return overwrite;
+  }
+  return data_dir;
+}
+
+// EOF
diff --git a/src/common/cp_files.F b/src/common/cp_files.F
index 1de58f7994..f01746636e 100644
--- a/src/common/cp_files.F
+++ b/src/common/cp_files.F
@@ -14,7 +14,10 @@
 !> \author Matthias Krack (MK)
 ! **************************************************************************************************
 MODULE cp_files
-
+   USE ISO_C_BINDING,                   ONLY: C_CHAR,&
+                                              C_F_POINTER,&
+                                              C_NULL_CHAR,&
+                                              C_PTR
    USE kinds,                           ONLY: default_path_length
    USE machine,                         ONLY: default_input_unit,&
                                               default_output_unit,&
@@ -551,19 +554,30 @@ CONTAINS
 !> \return ...
 !> \author Ole Schuett
 ! **************************************************************************************************
-   FUNCTION get_data_dir() RESULT(data_dir_path)
-      CHARACTER(LEN=default_path_length)                 :: data_dir_path
+   FUNCTION get_data_dir() RESULT(res)
+      CHARACTER(len=default_path_length)                 :: res
 
-      INTEGER                                            :: stat
+      CHARACTER(LEN=1, KIND=C_CHAR), DIMENSION(:), &
+         POINTER                                         :: path_f
+      INTEGER                                            :: i
+      TYPE(C_PTR)                                        :: path_c
+      INTERFACE
+         FUNCTION get_data_dir_c() BIND(C, name="get_data_dir")
+            IMPORT :: C_PTR
+            TYPE(C_PTR)                               :: get_data_dir_c
+         END FUNCTION get_data_dir_c
+      END INTERFACE
 
-      CALL GET_ENVIRONMENT_VARIABLE("CP2K_DATA_DIR", data_dir_path, status=stat)
-      IF (stat == 0) RETURN
+      path_c = get_data_dir_c()
+      CALL C_F_POINTER(path_c, path_f, shape=(/default_path_length/))
 
-#if defined(__DATA_DIR)
-      data_dir_path = __DATA_DIR
-#else
-      data_dir_path = "" !data-dir not set
-#endif
+      res = ""
+      DO i = 1, default_path_length
+         IF (path_f(i) == C_NULL_CHAR) RETURN
+         res(i:i) = path_f(i)
+      END DO
+
+      CPABORT("CP2K_DATA_DIR path is too long")
 
    END FUNCTION get_data_dir