[libc-commits] [libc] [libc] Make 'printf' converter output "(null)" instead of "null" (PR #85845)

[Joseph Huber via libc-commits]

https://github.com/jhuber6 updated https://github.com/llvm/llvm-project/pull/85845

>From 3536c9ed9528509d76a3fe83d99e7d316e0a6e82 Mon Sep 17 00:00:00 2001
From: Joseph Huber <huberjn at outlook.com>
Date: Tue, 19 Mar 2024 13:38:40 -0500
Subject: [PATCH] [libc] Make 'printf' converter output "(null)" instead of
 "null"

Summary:
Currently we print `null` for the null pointer in a `%s` expression.
Although it's not defined by the standard, other implementations choose
to use `(null)` to indicate this. We also currently print `(nullptr)` so
I think it's more consistent to use parens in both cases.
---
 libc/docs/dev/printf_behavior.rst             | 2 ++
 libc/docs/dev/undefined_behavior.rst          | 7 +++++++
 libc/src/stdio/printf_core/string_converter.h | 2 +-
 libc/test/src/stdio/sprintf_test.cpp          | 4 ++--
 4 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/libc/docs/dev/printf_behavior.rst b/libc/docs/dev/printf_behavior.rst
index 00d6c83f4b0d36..9548bfda57aa7d 100644
--- a/libc/docs/dev/printf_behavior.rst
+++ b/libc/docs/dev/printf_behavior.rst
@@ -1,3 +1,5 @@
+.. _printf_behavior:
+
 ====================================
 Printf Behavior Under All Conditions
 ====================================
diff --git a/libc/docs/dev/undefined_behavior.rst b/libc/docs/dev/undefined_behavior.rst
index 3a06b7007febd4..50e8bdde89ddd6 100644
--- a/libc/docs/dev/undefined_behavior.rst
+++ b/libc/docs/dev/undefined_behavior.rst
@@ -74,3 +74,10 @@ The POSIX.1 standard does not delineate the behavior consequent to invoking hsea
 Path without Leading Slashs in shm_open
 ----------------------------------------
 POSIX.1 leaves that when the name of a shared memory object does not begin with a slash, the behavior is implementation defined. In such cases, the shm_open in LLVM libc is implemented to behave as if the name began with a slash.
+
+Handling of NULL arguments to the 's' format specifier
+------------------------------------------------------
+The C standard does not specify behavior for ``printf("%s", NULL)``. We will
+print the string literal ``(null)`` unless using the 
+``LIBC_COPT_PRINTF_NO_NULLPTR_CHECKS`` option described in :ref:`printf 
+behavior<printf_behavior>`.
diff --git a/libc/src/stdio/printf_core/string_converter.h b/libc/src/stdio/printf_core/string_converter.h
index 04dc5a06da2226..9e05591079faa4 100644
--- a/libc/src/stdio/printf_core/string_converter.h
+++ b/libc/src/stdio/printf_core/string_converter.h
@@ -25,7 +25,7 @@ LIBC_INLINE int convert_string(Writer *writer, const FormatSection &to_conv) {
 
 #ifndef LIBC_COPT_PRINTF_NO_NULLPTR_CHECKS
   if (str_ptr == nullptr) {
-    str_ptr = "null";
+    str_ptr = "(null)";
   }
 #endif // LIBC_COPT_PRINTF_NO_NULLPTR_CHECKS
 
diff --git a/libc/test/src/stdio/sprintf_test.cpp b/libc/test/src/stdio/sprintf_test.cpp
index ab8bdb23b1187c..8dde95d02a96dc 100644
--- a/libc/test/src/stdio/sprintf_test.cpp
+++ b/libc/test/src/stdio/sprintf_test.cpp
@@ -121,8 +121,8 @@ TEST(LlvmLibcSPrintfTest, StringConv) {
 
 #ifndef LIBC_COPT_PRINTF_NO_NULLPTR_CHECKS
   written = LIBC_NAMESPACE::sprintf(buff, "%s", nullptr);
-  EXPECT_EQ(written, 4);
-  ASSERT_STREQ(buff, "null");
+  EXPECT_EQ(written, 6);
+  ASSERT_STREQ(buff, "(null)");
 #endif // LIBC_COPT_PRINTF_NO_NULLPTR_CHECKS
 }