updated behavior to match POSIX

Author: Sriya Pratipati

libc/src/wchar/CMakeLists.txt

@@ -76,6 +76,7 @@ add_entrypoint_object(
     libc.hdr.types.wchar_t
     libc.src.__support.common
     libc.src.__support.macros.config
+    libc.src.__support.libc_errno
     libc.src.__support.wchar.mbrtowc
     libc.src.__support.wchar.mbstate
 )

libc/src/wchar/mbtowc.cpp

@@ -11,6 +11,7 @@
 #include "hdr/types/size_t.h"
 #include "hdr/types/wchar_t.h"
 #include "src/__support/common.h"
+#include "src/__support/libc_errno.h"
 #include "src/__support/macros/config.h"
 #include "src/__support/wchar/mbrtowc.h"
 #include "src/__support/wchar/mbstate.h"
@@ -20,12 +21,17 @@ namespace LIBC_NAMESPACE_DECL {
 LLVM_LIBC_FUNCTION(int, mbtowc,
                    (wchar_t *__restrict pwc, const char *__restrict s,
                     size_t n)) {
+  // returns 0 since UTF-8 encoding is not state-dependent
   if (s == nullptr)
     return 0;
   internal::mbstate internal_mbstate;
-  auto ret = internal::mbrtowc(pwc, s, n, &internal_mbstate);
+  // temp ptr to use if pwc is nullptr
+  wchar_t buf[1];
+  auto ret =
+      internal::mbrtowc(pwc == nullptr ? buf : pwc, s, n, &internal_mbstate);
   if (!ret.has_value() || static_cast<int>(ret.value()) == -2) {
     // Encoding failure
+    libc_errno = EILSEQ;
     return -1;
   }
   return static_cast<int>(ret.value());

libc/test/src/wchar/mbtowc_test.cpp

@@ -7,6 +7,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "hdr/types/wchar_t.h"
+#include "src/__support/libc_errno.h"
 #include "src/wchar/mbtowc.h"
 #include "test/UnitTest/Test.h"
 
@@ -20,6 +21,7 @@ TEST(LlvmLibcMBToWC, OneByte) {
   // Should fail since we have not read enough
   n = LIBC_NAMESPACE::mbtowc(dest, ch, 0);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
 }
 
 TEST(LlvmLibcMBToWC, TwoByte) {
@@ -36,6 +38,7 @@ TEST(LlvmLibcMBToWC, TwoByte) {
   // Should fail after trying to read next byte too
   n = LIBC_NAMESPACE::mbtowc(dest, ch + 1, 1);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
 }
 
 TEST(LlvmLibcMBToWC, ThreeByte) {
@@ -49,6 +52,7 @@ TEST(LlvmLibcMBToWC, ThreeByte) {
   // Should fail since we have not read enough
   n = LIBC_NAMESPACE::mbtowc(dest, ch, 2);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
 }
 
 TEST(LlvmLibcMBToWC, FourByte) {
@@ -63,13 +67,15 @@ TEST(LlvmLibcMBToWC, FourByte) {
   // Should fail since we have not read enough
   n = LIBC_NAMESPACE::mbtowc(dest, ch, 2);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
 }
 
 TEST(LlvmLibcMBToWC, InvalidByte) {
   const char ch[1] = {static_cast<char>(0x80)};
   wchar_t dest[2];
   int n = LIBC_NAMESPACE::mbtowc(dest, ch, 1);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
 }
 
 TEST(LlvmLibcMBToWC, InvalidMultiByte) {
@@ -80,6 +86,8 @@ TEST(LlvmLibcMBToWC, InvalidMultiByte) {
   // Trying to push all 4 should error
   int n = LIBC_NAMESPACE::mbtowc(dest, ch, 4);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
+
   // Trying to push the second and third should correspond to null wc
   n = LIBC_NAMESPACE::mbtowc(dest, ch + 1, 2);
   ASSERT_EQ(n, 0);
@@ -95,6 +103,7 @@ TEST(LlvmLibcMBToWC, InvalidLastByte) {
   // Trying to push all 4 should error
   int n = LIBC_NAMESPACE::mbtowc(dest, ch, 4);
   ASSERT_EQ(n, -1);
+  ASSERT_EQ(static_cast<int>(libc_errno), EILSEQ);
 }
 
 TEST(LlvmLibcMBToWC, ValidTwoByteWithExtraRead) {
@@ -130,3 +139,13 @@ TEST(LlvmLibcMBToWC, NullString) {
   n = LIBC_NAMESPACE::mbtowc(dest, ch, 1);
   ASSERT_EQ(n, 0);
 }
+
+TEST(LlvmLibcMBToWC, NullWCPtr) {
+  const char ch[2] = {
+      static_cast<char>(0xC2),
+      static_cast<char>(0x8E),
+  };
+  // a null destination should still return the number of read bytes
+  int n = LIBC_NAMESPACE::mbtowc(nullptr, ch, 2);
+  ASSERT_EQ(n, 2);
+}