[libc] add stpcpy and stpncpy

[michaelrj]

Summary

Adds an implementation for stpcpy and stpncpy, which are posix extension
functions.

[sivachandra]

libc/src/string/stpcpy.cpp

31

Not sure if this is a valid argument for mempcpy. As in, we might want mempcpy to be instrumented really.

libc/src/string/stpncpy.cpp

18

Nit: Instead of initializing to 0 here, do it in the for statement ...

21

... like this:

for (i = 0; ...)

24

Nit: This can be

size_t end = i;

26

Can this for loop be replaced with a call to bzero?

[michaelrj]

respond to comments

libc/src/string/stpcpy.cpp

31

I was figuring that since mempcpy is effectively just a call to memcpy it should be treated the same, but I have no strong opinion one way or another.

[sivachandra]

libc/src/string/stpncpy.cpp

22

Remove the braces around single statement for blocks.

libc/test/src/string/stpncpy_test.cpp

23

Is this comment correct?

[lntue]

libc/src/string/stpncpy.cpp

25

Do you still need both i and end?

[michaelrj]

fix reference to CPP utils and address comments

libc/config/linux/x86_64/entrypoints.txt

set(TARGET_LIBC_ENTRYPOINTS
     libc.src.string.memccpy
     libc.src.string.memchr
     libc.src.string.memcmp
     libc.src.string.memcpy
     libc.src.string.memmove
     libc.src.string.mempcpy
     libc.src.string.memrchr
     libc.src.string.memset
+    libc.src.string.stpcpy
+    libc.src.string.stpncpy
     libc.src.string.strcat
     libc.src.string.strchr
     libc.src.string.strcmp
     libc.src.string.strcpy
     libc.src.string.strcspn
     libc.src.string.strlen
     libc.src.string.strncat
     libc.src.string.strncmp

libc/spec/posix.td

[
         FunctionSpec<
             "mempcpy",
             RetValSpec<VoidPtr>,
             [ArgSpec<VoidRestrictedPtr>,
              ArgSpec<ConstVoidRestrictedPtr>,
              ArgSpec<SizeTType>]
         >,
         FunctionSpec<
+            "stpcpy",
+            RetValSpec<CharPtr>,
+            [ArgSpec<RestrictedCharPtr>,
+             ArgSpec<ConstRestrictedCharPtr>]
+        >,
+        FunctionSpec<
+            "stpncpy",
+            RetValSpec<CharPtr>,
+            [ArgSpec<RestrictedCharPtr>,
+             ArgSpec<ConstRestrictedCharPtr>,
+             ArgSpec<SizeTType>]
+        >,
+        FunctionSpec<
             "strnlen",
              RetValSpec<SizeTType>,
              [ArgSpec<ConstCharPtr>, ArgSpec<SizeTType>]
         >,
         FunctionSpec<
             "strtok_r",
             RetValSpec<CharPtr>,
             [ArgSpec<RestrictedCharPtr>,

libc/src/string/CMakeLists.txt

add_entrypoint_object(
   memrchr
   SRCS
     memrchr.cpp
   HDRS
     memrchr.h
 )
 
 add_entrypoint_object(
+  stpcpy
+  SRCS
+    stpcpy.cpp
+  HDRS
+    stpcpy.h
+  DEPENDS
+    .mempcpy
+    .string_utils
+)
+
+add_entrypoint_object(
+  stpncpy
+  SRCS
+    stpncpy.cpp
+  HDRS
+    stpncpy.h
+)
+
+add_entrypoint_object(
   strcat
   SRCS
     strcat.cpp
   HDRS
     strcat.h
   DEPENDS
     .strcpy
     .string_utils

libc/src/string/stpcpy.h

+//===-- Implementation header for stpcpy ------------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_SRC_STRING_STPCPY_H
+#define LLVM_LIBC_SRC_STRING_STPCPY_H
+
+namespace __llvm_libc {
+
+char *stpcpy(char *__restrict dest, const char *__restrict src);
+
+} // namespace __llvm_libc
+
+#endif // LLVM_LIBC_SRC_STRING_STPCPY_H

libc/src/string/stpcpy.cpp

+//===-- Implementation of stpcpy ------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/string/stpcpy.h"
+#include "src/string/mempcpy.h"
+#include "src/string/string_utils.h"
+
+#include "src/__support/common.h"
+#include "src/__support/sanitizer.h"
+
+namespace __llvm_libc {
+
+LLVM_LIBC_FUNCTION(char *, stpcpy,
+                   (char *__restrict dest, const char *__restrict src)) {
+  size_t size = internal::string_length(src) + 1;
+  char *result =
+      reinterpret_cast<char *>(__llvm_libc::mempcpy(dest, src, size));
+
+  // In many libc uses, we do not want memcpy to be instrumented. Hence,
+  // we mark the destination as initialized.
+  //
+  // We do not want memcpy to be instrumented because compilers can potentially
+  // generate calls to memcpy. If the sanitizer business logic ends up with a
+  // compiler generated call to memcpy which is instrumented, then it will
+  // break the sanitizers.
+  SANITIZER_MEMORY_INITIALIZED(result, size);
+
+  if (result != nullptr) {
+    return result - 1;
+  }
+  return nullptr;
+}
+
+} // namespace __llvm_libc

libc/src/string/stpncpy.h

+//===-- Implementation header for stpncpy -----------------------*- C++ -*-===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#ifndef LLVM_LIBC_SRC_STRING_STPNCPY_H
+#define LLVM_LIBC_SRC_STRING_STPNCPY_H
+
+#include <stddef.h>
+
+namespace __llvm_libc {
+
+char *stpncpy(char *__restrict dest, const char *__restrict src, size_t n);
+
+} // namespace __llvm_libc
+
+#endif // LLVM_LIBC_SRC_STRING_STPNCPY_H

libc/src/string/stpncpy.cpp

+//===-- Implementation of stpncpy -----------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/string/stpncpy.h"
+
+#include "src/__support/common.h"
+
+namespace __llvm_libc {
+
+LLVM_LIBC_FUNCTION(char *, stpncpy,
+                   (char *__restrict dest, const char *__restrict src,
+                    size_t n)) {
+  size_t i = 0;
+  size_t strEnd = 0;
+  // Copy up until \0 is found.
+  for (; i < n && src[i] != '\0'; ++i) {
+    dest[i] = src[i];
+  }
+  strEnd = i;
+  // When n>strlen(src), n-strlen(src) \0 are appended.
+  for (; i < n; ++i)
+    dest[i] = '\0';
+  return dest + strEnd;
+}
+
+} // namespace __llvm_libc

libc/test/src/string/CMakeLists.txt

SUITE
     libc_string_unittests
   SRCS
     memrchr_test.cpp
   DEPENDS
     libc.src.string.memrchr
 )
 
 add_libc_unittest(
+  stpcpy_test
+  SUITE
+    libc_string_unittests
+  SRCS
+    stpcpy_test.cpp
+  DEPENDS
+    libc.src.string.stpcpy
+)
+
+add_libc_unittest(
+  stpncpy_test
+  SUITE
+    libc_string_unittests
+  SRCS
+    stpncpy_test.cpp
+  DEPENDS
+    libc.src.string.stpncpy
+)
+
+add_libc_unittest(
   strcat_test
   SUITE
     libc_string_unittests
   SRCS
     strcat_test.cpp
   DEPENDS
     libc.src.string.strcat
 )

libc/test/src/string/stpcpy_test.cpp

+//===-- Unittests for stpcpy ----------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/string/stpcpy.h"
+#include "utils/UnitTest/Test.h"
+
+#include "src/string/string_utils.h"
+
+TEST(LlvmLibcStpCpyTest, EmptySrc) {
+  const char *empty = "";
+  size_t srcSize = __llvm_libc::internal::string_length(empty);
+  char dest[4] = {'a', 'b', 'c', '\0'};
+
+  char *result = __llvm_libc::stpcpy(dest, empty);
+  ASSERT_EQ(dest + srcSize, result);
+  ASSERT_EQ(result[0], '\0');
+  ASSERT_STREQ(dest, empty);
+}
+
+TEST(LlvmLibcStpCpyTest, EmptyDest) {
+  const char *abc = "abc";
+  size_t srcSize = __llvm_libc::internal::string_length(abc);
+  char dest[4];
+
+  char *result = __llvm_libc::stpcpy(dest, abc);
+  ASSERT_EQ(dest + srcSize, result);
+  ASSERT_EQ(result[0], '\0');
+  ASSERT_STREQ(dest, abc);
+}
+
+TEST(LlvmLibcStpCpyTest, OffsetDest) {
+  const char *abc = "abc";
+  size_t srcSize = __llvm_libc::internal::string_length(abc);
+  char dest[7] = {'x', 'y', 'z'};
+
+  char *result = __llvm_libc::stpcpy(dest + 3, abc);
+  ASSERT_EQ(dest + 3 + srcSize, result);
+  ASSERT_EQ(result[0], '\0');
+  ASSERT_STREQ(dest, "xyzabc");
+}

libc/test/src/string/stpncpy_test.cpp

+//===-- Unittests for stpncpy ---------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "src/string/stpncpy.h"
+#include "utils/CPP/ArrayRef.h"
+#include "utils/UnitTest/Test.h"
+#include <stddef.h> // For size_t.
+
+class LlvmLibcStpncpyTest : public __llvm_libc::testing::Test {
+public:
+  void check_stpncpy(__llvm_libc::cpp::MutableArrayRef<char> dst,
+                     const __llvm_libc::cpp::ArrayRef<char> src, size_t n,
+                     const __llvm_libc::cpp::ArrayRef<char> expected,
+                     size_t expectedCopied) {
+    // Making sure we don't overflow buffer.
+    ASSERT_GE(dst.size(), n);
+    // Making sure stpncpy returns dst.
+    ASSERT_EQ(__llvm_libc::stpncpy(dst.data(), src.data(), n),
+              dst.data() + expectedCopied);
+    // Expected must be of the same size as dst.
+    ASSERT_EQ(dst.size(), expected.size());
+    // Expected and dst are the same.
+    for (size_t i = 0; i < expected.size(); ++i)
+      ASSERT_EQ(expected[i], dst[i]);
+  }
+};
+
+TEST_F(LlvmLibcStpncpyTest, Untouched) {
+  char dst[] = {'a', 'b'};
+  const char src[] = {'x', '\0'};
+  const char expected[] = {'a', 'b'};
+  check_stpncpy(dst, src, 0, expected, 0);
+}
+
+TEST_F(LlvmLibcStpncpyTest, CopyOne) {
+  char dst[] = {'a', 'b'};
+  const char src[] = {'x', 'y'};
+  const char expected[] = {'x', 'b'}; // no \0 is appended
+  check_stpncpy(dst, src, 1, expected, 1);
+}
+
+TEST_F(LlvmLibcStpncpyTest, CopyNull) {
+  char dst[] = {'a', 'b'};
+  const char src[] = {'\0', 'y'};
+  const char expected[] = {'\0', 'b'};
+  check_stpncpy(dst, src, 1, expected, 0);
+}
+
+TEST_F(LlvmLibcStpncpyTest, CopyPastSrc) {
+  char dst[] = {'a', 'b'};
+  const char src[] = {'\0', 'y'};
+  const char expected[] = {'\0', '\0'};
+  check_stpncpy(dst, src, 2, expected, 0);
+}
+
+TEST_F(LlvmLibcStpncpyTest, CopyTwoNoNull) {
+  char dst[] = {'a', 'b'};
+  const char src[] = {'x', 'y'};
+  const char expected[] = {'x', 'y'};
+  check_stpncpy(dst, src, 2, expected, 2);
+}
+
+TEST_F(LlvmLibcStpncpyTest, CopyTwoWithNull) {
+  char dst[] = {'a', 'b'};
+  const char src[] = {'x', '\0'};
+  const char expected[] = {'x', '\0'};
+  check_stpncpy(dst, src, 2, expected, 1);
+}

libc/test/src/string/strcpy_test.cpp

 //===-- Unittests for strcpy ----------------------------------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 
 #include "src/string/strcpy.h"
 #include "utils/UnitTest/Test.h"
 
+TEST(LlvmLibcStrCpyTest, EmptySrc) {
+  const char *empty = "";
+  char dest[4] = {'a', 'b', 'c', '\0'};
+
+  char *result = __llvm_libc::strcpy(dest, empty);
+  ASSERT_EQ(dest, result);
+  ASSERT_STREQ(dest, result);
+  ASSERT_STREQ(dest, empty);
+}
+
 TEST(LlvmLibcStrCpyTest, EmptyDest) {
   const char *abc = "abc";
   char dest[4];
 
   char *result = __llvm_libc::strcpy(dest, abc);
   ASSERT_EQ(dest, result);
   ASSERT_STREQ(dest, result);
   ASSERT_STREQ(dest, abc);