From 41785a0bf9d505a3647a10ddc330417f52fd4528 Mon Sep 17 00:00:00 2001
From: seyko <seyko2@gmail.com>
Date: Sun, 3 Apr 2016 11:42:15 +0300
Subject: [PATCH] -fnormalize-inc-dirs

    remove non-existent or duplicate directories from include paths
    if -fnormalize-inc-dirs is specified. This will help
    to compile current coreutils package
---
 libtcc.c | 93 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 tcc.h    |  3 +-
 tccpp.c  |  1 +
 3 files changed, 96 insertions(+), 1 deletion(-)

diff --git a/libtcc.c b/libtcc.c
index f51774f..8fd8183 100644
--- a/libtcc.c
+++ b/libtcc.c
@@ -1508,6 +1508,96 @@ LIBTCCAPI int tcc_add_symbol(TCCState *s, const char *name, const void *val)
     return 0;
 }
 
+
+/* Windows stat* ( https://msdn.microsoft.com/en-us/library/14h5k7ff.aspx ):
+ * - st_gid, st_ino, st_uid: only valid on "unix" file systems (not FAT, NTFS, etc)
+ * - st_atime, st_ctime: not valid on FAT, valid on NTFS.
+ * - Other fields should be reasonably compatible (and S_ISDIR should work).
+ *
+ * BY_HANDLE_FILE_INFORMATION ( https://msdn.microsoft.com/en-us/library/windows/desktop/aa363788%28v=vs.85%29.aspx ):
+ * - File index (combined nFileIndexHigh and nFileIndexLow) _may_ change when the file is opened.
+ *   - But on NTFS: it's guaranteed to be the same value until the file is deleted.
+ * - On windows server 2012 there's a 128b file id, and the 64b one via
+ *   nFileIndex* is not guaranteed to be unique.
+ *
+ * - MS Docs suggest to that volume number with the file index could be used to
+ *   check if two handles refer to the same file.
+ */
+#ifndef _WIN32
+typedef struct stat                file_info_t;
+#else
+typedef BY_HANDLE_FILE_INFORMATION file_info_t;
+#endif
+
+int get_file_info(const char *fname, file_info_t *out_info)
+{
+#ifndef _WIN32
+    return stat(fname, out_info);
+#else
+    int rv = 1;
+    HANDLE h = CreateFile(fname, GENERIC_READ, FILE_SHARE_READ, NULL, OPEN_EXISTING,
+                          FILE_ATTRIBUTE_NORMAL|FILE_FLAG_BACKUP_SEMANTICS, NULL);
+
+    if (h != INVALID_HANDLE_VALUE) {
+        rv = !GetFileInformationByHandle(h, out_info);
+        CloseHandle(h);
+    }
+    return rv;
+#endif
+}
+
+int is_dir(file_info_t *info)
+{
+#ifndef _WIN32
+    return S_ISDIR(info->st_mode);
+#else
+    return (info->dwFileAttributes & FILE_ATTRIBUTE_DIRECTORY) ==
+           FILE_ATTRIBUTE_DIRECTORY;
+#endif
+}
+
+int is_same_file(const file_info_t *fi1, const file_info_t *fi2)
+{
+#ifndef _WIN32
+    return fi1->st_dev == fi2->st_dev &&
+           fi1->st_ino == fi2->st_ino;
+#else
+    return fi1->dwVolumeSerialNumber == fi2->dwVolumeSerialNumber &&
+           fi1->nFileIndexHigh       == fi2->nFileIndexHigh &&
+           fi1->nFileIndexLow        == fi2->nFileIndexLow;
+#endif
+}
+
+static void
+tcc_normalize_inc_dirs_aux(file_info_t *stats, size_t *pnum, char **path)
+{
+    size_t i, num = *pnum;
+    if (get_file_info(*path, &stats[num]) || !is_dir(&stats[num]))
+        goto remove;
+    for (i = 0; i < num; i++)
+        if (is_same_file(&stats[i], &stats[num]))
+            goto remove;
+    *pnum = num + 1;
+    return;
+ remove:
+    tcc_free(*path);
+    *path = 0;
+}
+
+/* Remove non-existent and duplicate directories from include paths. */
+ST_FUNC void tcc_normalize_inc_dirs(TCCState *s)
+{
+    file_info_t *stats =
+        tcc_malloc(((size_t)s->nb_sysinclude_paths + s->nb_include_paths) *
+                   sizeof(*stats));
+    size_t i, num = 0;
+    for (i = 0; i < s->nb_sysinclude_paths; i++)
+        tcc_normalize_inc_dirs_aux(stats, &num, &s->sysinclude_paths[i]);
+    for (i = 0; i < s->nb_include_paths; i++)
+        tcc_normalize_inc_dirs_aux(stats, &num, &s->include_paths[i]);
+    tcc_free(stats);
+}
+
 LIBTCCAPI int tcc_set_output_type(TCCState *s, int output_type)
 {
     s->output_type = output_type;
@@ -1573,6 +1663,8 @@ LIBTCCAPI int tcc_set_output_type(TCCState *s, int output_type)
     }
 #endif
 
+    if (s->normalize_inc_dirs)
+        tcc_normalize_inc_dirs(s);
     if (s->output_type == TCC_OUTPUT_PREPROCESS)
         print_defines();
 
@@ -1652,6 +1744,7 @@ static const FlagDef flag_defs[] = {
     { offsetof(TCCState, ms_extensions), 0, "ms-extensions" },
     { offsetof(TCCState, old_struct_init_code), 0, "old-struct-init-code" },
     { offsetof(TCCState, dollars_in_identifiers), 0, "dollars-in-identifiers" },
+    { offsetof(TCCState, normalize_inc_dirs), 0, "normalize-inc-dirs" },
 };
 
 /* set/reset a flag */
diff --git a/tcc.h b/tcc.h
index 3b98a8c..0200207 100644
--- a/tcc.h
+++ b/tcc.h
@@ -647,7 +647,8 @@ struct TCCState {
     int old_struct_init_code;	/* use old algorithm to init array in struct when there is no '{' used.
 				   Liuux 2.4.26 can't find initrd when compiled with a new algorithm */
     int dollars_in_identifiers;	/* allows '$' char in indentifiers */
-    
+    int normalize_inc_dirs;	/* remove non-existent or duplicate directories from include paths */
+
     /* warning switches */
     int warn_write_strings;
     int warn_unsupported;
diff --git a/tccpp.c b/tccpp.c
index 1803fe3..64c4919 100644
--- a/tccpp.c
+++ b/tccpp.c
@@ -1663,6 +1663,7 @@ ST_FUNC void preprocess(int is_bof)
                 /* search in all the include paths */
                 int j = i - 2, k = j - s1->nb_include_paths;
                 path = k < 0 ? s1->include_paths[j] : s1->sysinclude_paths[k];
+                if (path == 0) continue;
                 pstrcpy(buf1, sizeof(buf1), path);
                 pstrcat(buf1, sizeof(buf1), "/");
             }