Hi again,
I have started with the patching and this is what I have got for now.
Some strcpy's could be changed to their strlcpy variant, but some just
depend on the size of the original string. For those cases I just
substituted by memcpy(). Maybe it is just better to do memcpy() at
every instance since that would preserve completely the semantics of
the strcpy().
I am having difficulty generating the library again after having built
it for the first time. I can do gmake gnatlib inside build-amd64 but
this is very hacky, plus I then cant figure out how to rebuild the package
in the fake-amd64 directory. Is there any documentation on this matter?
Index: 15/patches/patch-gcc_ada_adaint_c
===================================================================
RCS file: /cvs/ports/lang/gcc/15/patches/patch-gcc_ada_adaint_c,v
diff -u -p -r1.1.1.1 patch-gcc_ada_adaint_c
--- 15/patches/patch-gcc_ada_adaint_c 18 Aug 2025 19:49:19 -0000 1.1.1.1
+++ 15/patches/patch-gcc_ada_adaint_c 21 Mar 2026 19:15:52 -0000
@@ -1,16 +1,102 @@
Index: gcc/ada/adaint.c
--- gcc/ada/adaint.c.orig
+++ gcc/ada/adaint.c
-@@ -3714,6 +3714,12 @@ void __gnat_killprocesstree (int pid, int sig_num)
- */
+@@ -576,8 +576,8 @@ __gnat_try_lock (char *dir, char *file)
+ GNAT_STRUCT_STAT stat_result;
+ int fd;
+
+- sprintf (full_path, "%s%c%s", dir, DIR_SEPARATOR, file);
+- sprintf (temp_file, "%s%cTMP-%ld-%ld",
++ snprintf (full_path, sizeof(full_path), "%s%c%s", dir, DIR_SEPARATOR, file);
++ snprintf (temp_file, sizeof(temp_file), "%s%cTMP-%ld-%ld",
+ dir, DIR_SEPARATOR, (long)getpid(), (long)getppid ());
+
+ /* Create the temporary file and write the process number. */
+@@ -756,8 +756,8 @@ __gnat_os_filename (char *filename ATTRIBUTE_UNUSED,
+ strcpy (encoding, "encoding=utf8");
+ *e_length = strlen (encoding);
+ #else
+- strcpy (os_name, filename);
+ *o_length = strlen (filename);
++ memcpy (os_name, filename, *o_length + 1);
+ *e_length = 0;
+ #endif
}
+@@ -1278,7 +1278,7 @@ __gnat_tmp_name (char *tmp_filename)
+ strcpy (tmp_filename, "/tmp/gnat-XXXXXX");
+ #endif
+ else
+- sprintf (tmp_filename, "%s/gnat-XXXXXX", tmpdir);
++ snprintf (tmp_filename, __gnat_max_path_len, "%s/gnat-XXXXXX", tmpdir);
+
+ close (mkstemp(tmp_filename));
+ #elif defined (__vxworks) && !defined (VTHREADS)
+@@ -1373,8 +1373,8 @@ __gnat_readdir (DIR *dirp, char *buffer, int *len)
+
+ if (dirent != NULL)
+ {
+- strcpy (buffer, dirent->d_name);
+- *len = strlen (buffer);
++ *len = strlen (dirent->d_name);
++ memcpy (buffer, dirent->d_name, *len + 1);
+ return buffer;
+ }
+ else
+@@ -2936,7 +2936,8 @@ __gnat_locate_file_with_predicate (char *file_name, ch
+ int (*predicate)(char *))
+ {
+ char *ptr;
+- char *file_path = (char *) alloca (strlen (file_name) + 1);
++ int file_name_len = strlen(file_name);
++ char *file_path = (char *) alloca (file_name_len + 1);
+ int absolute;
+
+ /* Return immediately if file_name is empty */
+@@ -2950,7 +2951,7 @@ __gnat_locate_file_with_predicate (char *file_name, ch
+ if (*ptr == '"')
+ ptr++;
+
+- strcpy (file_path, ptr);
++ strlcpy (file_path, ptr, file_name_len + 1);
+
+ ptr = file_path + strlen (file_path) - 1;
+
+@@ -3015,7 +3016,7 @@ __gnat_locate_file_with_predicate (char *file_name, ch
+ if (!IS_DIRECTORY_SEPARATOR(*ptr))
+ *++ptr = DIR_SEPARATOR;
+
+- strcpy (++ptr, file_name);
++ memcpy (++ptr, file_name, strlen(file_name) + 1);
+
+ if (predicate (file_path))
+ return xstrdup (file_path);
+@@ -3112,13 +3113,14 @@ __gnat_locate_exec_on_path (char *exec_name,
+ #else
+ const char *path_val = getenv ("PATH");
++ int path_val_len = strlen(path_val);
+
+ /* If PATH is not defined, proceed with __gnat_locate_exec anyway, so we can
+ find files that contain directory names. */
+
+ if (path_val == NULL) path_val = "";
+- apath_val = (char *) alloca (strlen (path_val) + 1);
+- strcpy (apath_val, path_val);
++ apath_val = (char *) alloca (path_val_len + 1);
++ strlcpy (apath_val, path_val, path_val_len + 1);
+ #endif
+
+ return __gnat_locate_exec (exec_name, apath_val);
+@@ -3712,6 +3714,12 @@ void __gnat_killprocesstree (int pid, int sig_num)
+ The 5th and 6th words are the pid and the 7th and 8th the ppid.
+ See: /usr/include/sys/procfs.h (struct pstatus).
+ */
++}
++
+const char *
+fname_as_string(int pretty_p __attribute__((__unused__)))
+{
+ return NULL;
-+}
-+
- #ifdef __cplusplus
}
- #endif
+
+ #ifdef __cplusplus
Index: 15/patches/patch-gcc_ada_cstreams_c
===================================================================
RCS file: /cvs/ports/lang/gcc/15/patches/patch-gcc_ada_cstreams_c,v
diff -u -p -r1.1.1.1 patch-gcc_ada_cstreams_c
--- 15/patches/patch-gcc_ada_cstreams_c 18 Aug 2025 19:49:19 -0000 1.1.1.1
+++ 15/patches/patch-gcc_ada_cstreams_c 21 Mar 2026 19:15:52 -0000
@@ -10,3 +10,19 @@ Index: gcc/ada/cstreams.c
/* Use realpath function which resolves links and references to . and ..
on those Unix systems that support it. Note that GNU/Linux provides it
but
+@@ -228,12 +228,12 @@ __gnat_full_name (char *nam, char *buffer)
+ /* If the name returned is an absolute path, it is safe to append '/'
+ to the path and concatenate the name of the file. */
+ if (buffer[0] == '/')
+- strcat (buffer, "/");
++ strlcat (buffer, "/", __gnat_max_path_len);
+
+- strcat (buffer, nam);
++ strlcat (buffer, nam, __gnat_max_path_len);
+ }
+ else
+- strcpy (buffer, nam);
++ strlcpy (buffer, nam, __gnat_max_path_len);
+ #endif
+
+ return buffer;
Index: 15/patches/patch-gcc_ada_env_c
===================================================================
RCS file: 15/patches/patch-gcc_ada_env_c
diff -N 15/patches/patch-gcc_ada_env_c
--- /dev/null 1 Jan 1970 00:00:00 -0000
+++ 15/patches/patch-gcc_ada_env_c 21 Mar 2026 19:15:52 -0000
@@ -0,0 +1,12 @@
+Index: gcc/ada/env.c
+--- gcc/ada/env.c.orig
++++ gcc/ada/env.c
+@@ -109,7 +109,7 @@ __gnat_setenv (char *name, char *value)
+
+ expression = (char *) xmalloc (size * sizeof (char));
+
+- sprintf (expression, "%s=%s", name, value);
++ snprintf (expression, size, "%s=%s", name, value);
+ putenv (expression);
+ #if defined (__MINGW32__) || defined (__vxworks)
+ /* putenv for Windows and VxWorks 6 kernel modules makes a copy of the
Index: 15/patches/patch-libphobos_configure
===================================================================
RCS file: /cvs/ports/lang/gcc/15/patches/patch-libphobos_configure,v
diff -u -p -r1.2 patch-libphobos_configure
--- 15/patches/patch-libphobos_configure 7 Mar 2026 21:43:19 -0000
1.2
+++ 15/patches/patch-libphobos_configure 21 Mar 2026 19:15:52 -0000
@@ -48,14 +48,14 @@ Index: libphobos/configure
druntime_fiber_asm_external=yes
;;
esac
-@@ -16067,6 +16081,10 @@ if test -z "${DRUNTIME_CPU_POWERPC_TRUE}" && test -z "
- as_fn_error $? "conditional \"DRUNTIME_CPU_POWERPC\" was never defined.
- Usually this means the macro was only invoked conditionally." "$LINENO" 5
+@@ -16065,6 +16079,10 @@ Usually this means the macro was only invoked conditio
fi
-+if test -z "${DRUNTIME_CPU_SPARC64_TRUE}" && test -z
"${DRUNTIME_CPU_SPARC64_FALSE}"; then
-+ as_fn_error $? "conditional \"DRUNTIME_CPU_SPARC64\" was never defined.
+ if test -z "${DRUNTIME_CPU_POWERPC_TRUE}" && test -z
"${DRUNTIME_CPU_POWERPC_FALSE}"; then
+ as_fn_error $? "conditional \"DRUNTIME_CPU_POWERPC\" was never defined.
+Usually this means the macro was only invoked conditionally." "$LINENO" 5
+fi
- if test -z "${DRUNTIME_CPU_X86_TRUE}" && test -z "${DRUNTIME_CPU_X86_FALSE}";
then
- as_fn_error $? "conditional \"DRUNTIME_CPU_X86\" was never defined.
++if test -z "${DRUNTIME_CPU_SPARC64_TRUE}" && test -z
"${DRUNTIME_CPU_SPARC64_FALSE}"; then
++ as_fn_error $? "conditional \"DRUNTIME_CPU_SPARC64\" was never defined.
Usually this means the macro was only invoked conditionally." "$LINENO" 5
+ fi
+ if test -z "${DRUNTIME_CPU_X86_TRUE}" && test -z "${DRUNTIME_CPU_X86_FALSE}";
then
On Friday, March 20th, 2026 at 10:39 AM, Stuart Henderson
<[email protected]> wrote:
> On 2026/03/19 23:33, ZenitDS wrote:
> > Hello,
> >
> > for gcc ada toolchain, libgnat uses several strcpy and strcat that spawn
> > the annoying warnings of strcpy and strcat on each compilation. I think
> > it may be a good idea to add some patches to fix this situation with
> > strlcpy. I am willing to do it in case it is ok and could be merged.
> >
> > Cheers
> >
>
> As long as the patches aren't too intrusive I think this makes sense.
>
