[libc++] Implements filebuf unbuffered. (#76629)

When calling setbuf(nullptr, 0) before performing file operations it
should set the file to unbuffered mode. Currently the code avoids
buffering internally, but the underlying stream still can buffer.

This is addressed by disabling the buffering of the underlying stream.

Fixes: https://github.com/llvm/llvm-project/issues/60509

2 files changed, 179 insertions, 1 deletions

diff --git a/libcxx/include/fstream b/libcxx/include/fstream
index 776641b..7a084d1 100644
--- a/libcxx/include/fstream
+++ b/libcxx/include/fstream
@@ -308,6 +308,43 @@ private:
   state_type __st_;
   state_type __st_last_;
   ios_base::openmode __om_;
+  // There have been no file operations yet, which allows setting unbuffered
+  // I/O mode.
+  static const ios_base::openmode __no_io_operations = ios_base::trunc;
+  // Unbuffered I/O mode has been requested.
+  static const ios_base::openmode __use_unbuffered_io = ios_base::ate;
+  // Used to track the currently used mode and track whether the output should
+  // be unbuffered.
+  // [filebuf.virtuals]/12
+  //   If setbuf(0, 0) is called on a stream before any I/O has occurred on
+  //   that stream, the stream becomes unbuffered. Otherwise the results are
+  //   implementation-defined.
+  // This allows calling setbuf(0, 0)
+  // - before opening a file,
+  // - after opening a file, before
+  //   - a read
+  //   - a write
+  //   - a seek.
+  // Note that opening a file with ios_base::ate does a seek operation.
+  // Normally underflow, overflow, and sync change this flag to ios_base::in,
+  // ios_base_out, or 0.
+  //
+  // The ios_base::trunc and ios_base::ate flags are not used in __cm_. They
+  // are used to track the state of the unbuffered request. For readability
+  // they have the aliases __no_io_operations and __use_unbuffered_io
+  // respectively.
+  //
+  // The __no_io_operations and __use_unbuffered_io flags are used in the
+  // following way:
+  // - __no_io_operations is set upon construction to indicate the unbuffered
+  //   state can be set.
+  // - When requesting unbuffered output:
+  //   - If the file is open it sets the mode.
+  //   - Else places a request by adding the __use_unbuffered_io flag.
+  // - When a file is opened it checks whether both __no_io_operations and
+  //   __use_unbuffered_io are set. If so switches to unbuffered mode.
+  // - All file I/O operations change the mode effectively clearing the
+  //   __no_io_operations and __use_unbuffered_io flags.
   ios_base::openmode __cm_;
   bool __owns_eb_;
   bool __owns_ib_;
@@ -327,7 +364,13 @@ private:
       return nullptr;
 
     __om_ = __mode;
+    if (__cm_ == (__no_io_operations | __use_unbuffered_io)) {
+      std::setbuf(__file_, nullptr);
+      __cm_ = 0;
+    }
+
     if (__mode & ios_base::ate) {
+      __cm_ = 0;
       if (fseek(__file_, 0, SEEK_END)) {
         fclose(__file_);
         __file_ = nullptr;
@@ -337,6 +380,20 @@ private:
 
     return this;
   }
+
+  // If the file is already open, switch to unbuffered mode. Otherwise, record
+  // the request to use unbuffered mode so that we use that mode when we
+  // eventually open the file.
+  _LIBCPP_HIDE_FROM_ABI void __request_unbuffered_mode(char_type* __s, streamsize __n) {
+    if (__cm_ == __no_io_operations && __s == nullptr && __n == 0) {
+      if (__file_) {
+        std::setbuf(__file_, nullptr);
+        __cm_ = 0;
+      } else {
+        __cm_ = __no_io_operations | __use_unbuffered_io;
+      }
+    }
+  }
 };
 
 template <class _CharT, class _Traits>
@@ -352,7 +409,7 @@ basic_filebuf<_CharT, _Traits>::basic_filebuf()
       __st_(),
       __st_last_(),
       __om_(0),
-      __cm_(0),
+      __cm_(__no_io_operations),
       __owns_eb_(false),
       __owns_ib_(false),
       __always_noconv_(false) {
@@ -810,6 +867,7 @@ template <class _CharT, class _Traits>
 basic_streambuf<_CharT, _Traits>* basic_filebuf<_CharT, _Traits>::setbuf(char_type* __s, streamsize __n) {
   this->setg(nullptr, nullptr, nullptr);
   this->setp(nullptr, nullptr);
+  __request_unbuffered_mode(__s, __n);
   if (__owns_eb_)
     delete[] __extbuf_;
   if (__owns_ib_)
diff --git a/libcxx/test/std/input.output/file.streams/fstreams/filebuf.virtuals/setbuf.pass.cpp b/libcxx/test/std/input.output/file.streams/fstreams/filebuf.virtuals/setbuf.pass.cpp
new file mode 100644
index 0000000..8bcce28
--- /dev/null
+++ b/libcxx/test/std/input.output/file.streams/fstreams/filebuf.virtuals/setbuf.pass.cpp
@@ -0,0 +1,120 @@
+//===----------------------------------------------------------------------===//
+//
+// 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
+//
+//===----------------------------------------------------------------------===//
+
+// <fstream>
+
+// basic_streambuf<charT, traits>* setbuf(char_type* s, streamsize n) override;
+
+#include <fstream>
+#include <cstddef>
+#include <cassert>
+
+#include "test_macros.h"
+
+template <class CharT>
+static std::size_t file_size(const char* filename) {
+  FILE* f = std::fopen(filename, "rb");
+  std::fseek(f, 0, SEEK_END);
+  long result = std::ftell(f);
+  std::fclose(f);
+  return result;
+}
+
+// Helper class to expose some protected std::basic_filebuf<CharT> members.
+template <class CharT>
+struct filebuf : public std::basic_filebuf<CharT> {
+  CharT* base() { return this->pbase(); }
+  CharT* ptr() { return this->pptr(); }
+};
+
+template <class CharT>
+static void buffered_request() {
+  filebuf<CharT> buffer;
+
+  CharT b[10] = {0};
+  assert(buffer.pubsetbuf(b, 10) == &buffer);
+
+  buffer.open("test.dat", std::ios_base::out);
+  buffer.sputc(CharT('a'));
+  assert(b[0] == 'a');
+
+  buffer.close();
+  assert(file_size<CharT>("test.dat") == 1);
+}
+
+template <class CharT>
+static void unbuffered_request_before_open() {
+  filebuf<CharT> buffer;
+
+  assert(buffer.pubsetbuf(nullptr, 0) == &buffer);
+  assert(buffer.base() == nullptr);
+  assert(buffer.ptr() == nullptr);
+
+  buffer.open("test.dat", std::ios_base::out);
+  assert(buffer.base() == nullptr);
+  assert(buffer.ptr() == nullptr);
+
+  buffer.sputc(CharT('a'));
+  assert(buffer.base() == nullptr);
+  assert(buffer.ptr() == nullptr);
+
+  assert(file_size<CharT>("test.dat") == 1);
+}
+
+template <class CharT>
+static void unbuffered_request_after_open() {
+  filebuf<CharT> buffer;
+
+  buffer.open("test.dat", std::ios_base::out);
+
+  assert(buffer.pubsetbuf(nullptr, 0) == &buffer);
+  assert(buffer.base() == nullptr);
+  assert(buffer.ptr() == nullptr);
+
+  buffer.sputc(CharT('a'));
+  assert(buffer.base() == nullptr);
+  assert(buffer.ptr() == nullptr);
+
+  assert(file_size<CharT>("test.dat") == 1);
+}
+
+template <class CharT>
+static void unbuffered_request_after_open_ate() {
+  filebuf<CharT> buffer;
+
+  buffer.open("test.dat", std::ios_base::out | std::ios_base::ate);
+
+  assert(buffer.pubsetbuf(nullptr, 0) == &buffer);
+
+  buffer.sputc(CharT('a'));
+  assert(file_size<CharT>("test.dat") <= 1);
+  // on libc++ buffering is used by default.
+  LIBCPP_ASSERT(file_size<CharT>("test.dat") == 0);
+
+  buffer.close();
+  assert(file_size<CharT>("test.dat") == 1);
+}
+
+template <class CharT>
+static void test() {
+  buffered_request<CharT>();
+
+  unbuffered_request_before_open<CharT>();
+  unbuffered_request_after_open<CharT>();
+  unbuffered_request_after_open_ate<CharT>();
+}
+
+int main(int, char**) {
+  test<char>();
+
+#ifndef TEST_HAS_NO_WIDE_CHARACTERS
+  test<wchar_t>();
+#endif
+
+  return 0;
+}