clang/0001-clang-MinGW-Improve-extend-the-gcc-sysroot-detection.patch

From efe3686d4a519b357cecb5aaf715f1e1b6f2d025 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Martin=20Storsj=C3=B6?= <martin@martin.st>
Date: Tue, 22 Nov 2022 14:38:54 +0200
Subject: [PATCH 1/2] [clang] [MinGW] Improve/extend the gcc/sysroot detection
 logic

There are three functions that try to detect the right implicit
sysroot and libgcc directory setup to use
- One which looks for mingw sysroots located in
  <clangbin>/../<sysrootname>
- One which looks for a mingw-targeting gcc executables in the PATH
- One which looks in the <gccroot>/lib/gcc directory to find the
  right one to use, and the right specific triple used for arch
  specific directories in the gcc/libstdc++ install

These have mostly tried to look for executables named
"<arch>-w64-mingw32-gcc" or "mingw32-gcc" or subdirectories
named "<arch>-w64-mingw32" or "mingw32".

In the case of findClangRelativeSysroot, it also has looked
for directories with the name of the actual triple. This
was added in deff7536278d355977171726124f83aa4bb95419,
with the intent of looking for a directory matching exactly
the user provided literal triple - however the triple here
is the normalized one, not the one provided by the user on
the command line.

Improve and unify this logic somewhat:
- Always first look for things based on the literal triple
  provided by the user.
- Secondly look for things based on the normalized triple
  (which usually ends up as e.g. x86_64-w64-windows-gnu),
  accessed via the Triple which is passed to the constructor
- Then look for the common triple form <arch>-w64-mingw32

The literal triple provided by the user is available via
Driver::getTargetTriple(), but computeTargetTriple() may
change e.g. the architecture of it, so we need to
reapply the effective architecture on the literal triple
spelling from Driver::getTargetTriple().

Do this consistently for all of findGcc, findClangRelativeSysroot
and findGccLibDir (while keeping the existing plain "mingw32"
cases in findGcc and findGccLibDir too).

Fedora 37 started shipping mingw sysroots targeting UCRT,
in addition to the traditional msvcrt.dll, and these use
triples in the form <arch>-w64-mingw32ucrt - see
https://fedoraproject.org/wiki/Changes/F37MingwUCRT.

Thus, in addition to the existing default tested triples,
try looking for triples in the form <arch>-w64-mingw32ucrt,
to automatically find the UCRT sysroots on Fedora 37.
By explicitly setting a specific target on the Clang command
line, the user can be more explicit with which flavour is
to be preferred.

This should fix the main issue in
https://github.com/llvm/llvm-project/issues/59001.

Differential Revision: https://reviews.llvm.org/D138692
---
 clang/lib/Driver/ToolChains/MinGW.cpp         | 68 ++++++++++++++-----
 clang/lib/Driver/ToolChains/MinGW.h           |  2 +-
 .../12.2.1/include-fixed/.keep                |  0
 .../sys-root/mingw/include/.keep              |  0
 .../sys-root/mingw/include/c++/backward/.keep |  0
 .../include/c++/x86_64-w64-mingw32ucrt/.keep  |  0
 clang/test/Driver/mingw-sysroot.cpp           | 13 ++++
 clang/test/Driver/mingw.cpp                   |  4 ++
 8 files changed, 68 insertions(+), 19 deletions(-)
 create mode 100644 clang/test/Driver/Inputs/mingw_fedora_tree/usr/lib/gcc/x86_64-w64-mingw32ucrt/12.2.1/include-fixed/.keep
 create mode 100644 clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/.keep
 create mode 100644 clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/backward/.keep
 create mode 100644 clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/x86_64-w64-mingw32ucrt/.keep

diff --git a/clang/lib/Driver/ToolChains/MinGW.cpp b/clang/lib/Driver/ToolChains/MinGW.cpp
index ae7c4c56bf9e..0ebacea6e87f 100644
--- a/clang/lib/Driver/ToolChains/MinGW.cpp
+++ b/clang/lib/Driver/ToolChains/MinGW.cpp
@@ -359,13 +359,26 @@ static bool findGccVersion(StringRef LibDir, std::string &GccLibDir,
   return Ver.size();
 }
 
-void toolchains::MinGW::findGccLibDir() {
-  llvm::SmallVector<llvm::SmallString<32>, 2> SubdirNames;
+static llvm::Triple getLiteralTriple(const Driver &D, const llvm::Triple &T) {
+  llvm::Triple LiteralTriple(D.getTargetTriple());
+  // The arch portion of the triple may be overridden by -m32/-m64.
+  LiteralTriple.setArchName(T.getArchName());
+  return LiteralTriple;
+}
+
+void toolchains::MinGW::findGccLibDir(const llvm::Triple &LiteralTriple) {
+  llvm::SmallVector<llvm::SmallString<32>, 5> SubdirNames;
+  SubdirNames.emplace_back(LiteralTriple.str());
+  SubdirNames.emplace_back(getTriple().str());
   SubdirNames.emplace_back(getTriple().getArchName());
-  SubdirNames[0] += "-w64-mingw32";
+  SubdirNames.back() += "-w64-mingw32";
+  SubdirNames.emplace_back(getTriple().getArchName());
+  SubdirNames.back() += "-w64-mingw32ucrt";
   SubdirNames.emplace_back("mingw32");
-  if (SubdirName.empty())
-    SubdirName = std::string(SubdirNames[0].str());
+  if (SubdirName.empty()) {
+    SubdirName = getTriple().getArchName();
+    SubdirName += "-w64-mingw32";
+  }
   // lib: Arch Linux, Ubuntu, Windows
   // lib64: openSUSE Linux
   for (StringRef CandidateLib : {"lib", "lib64"}) {
@@ -380,10 +393,17 @@ void toolchains::MinGW::findGccLibDir() {
   }
 }
 
-static llvm::ErrorOr<std::string> findGcc(const llvm::Triple &T) {
-  llvm::SmallVector<llvm::SmallString<32>, 2> Gccs;
+static llvm::ErrorOr<std::string> findGcc(const llvm::Triple &LiteralTriple,
+                                          const llvm::Triple &T) {
+  llvm::SmallVector<llvm::SmallString<32>, 5> Gccs;
+  Gccs.emplace_back(LiteralTriple.str());
+  Gccs.back() += "-gcc";
+  Gccs.emplace_back(T.str());
+  Gccs.back() += "-gcc";
   Gccs.emplace_back(T.getArchName());
-  Gccs[0] += "-w64-mingw32-gcc";
+  Gccs.back() += "-w64-mingw32-gcc";
+  Gccs.emplace_back(T.getArchName());
+  Gccs.back() += "-w64-mingw32ucrt-gcc";
   Gccs.emplace_back("mingw32-gcc");
   // Please do not add "gcc" here
   for (StringRef CandidateGcc : Gccs)
@@ -393,12 +413,15 @@ static llvm::ErrorOr<std::string> findGcc(const llvm::Triple &T) {
 }
 
 static llvm::ErrorOr<std::string>
-findClangRelativeSysroot(const Driver &D, const llvm::Triple &T,
-                         std::string &SubdirName) {
-  llvm::SmallVector<llvm::SmallString<32>, 2> Subdirs;
+findClangRelativeSysroot(const Driver &D, const llvm::Triple &LiteralTriple,
+                         const llvm::Triple &T, std::string &SubdirName) {
+  llvm::SmallVector<llvm::SmallString<32>, 4> Subdirs;
+  Subdirs.emplace_back(LiteralTriple.str());
   Subdirs.emplace_back(T.str());
   Subdirs.emplace_back(T.getArchName());
-  Subdirs[1] += "-w64-mingw32";
+  Subdirs.back() += "-w64-mingw32";
+  Subdirs.emplace_back(T.getArchName());
+  Subdirs.back() += "-w64-mingw32ucrt";
   StringRef ClangRoot = llvm::sys::path::parent_path(D.getInstalledDir());
   StringRef Sep = llvm::sys::path::get_separator();
   for (StringRef CandidateSubdir : Subdirs) {
@@ -418,14 +441,16 @@ toolchains::MinGW::MinGW(const Driver &D, const llvm::Triple &Triple,
 
   // The sequence for detecting a sysroot here should be kept in sync with
   // the testTriple function below.
+  llvm::Triple LiteralTriple = getLiteralTriple(D, getTriple());
   if (getDriver().SysRoot.size())
     Base = getDriver().SysRoot;
   // Look for <clang-bin>/../<triplet>; if found, use <clang-bin>/.. as the
   // base as it could still be a base for a gcc setup with libgcc.
-  else if (llvm::ErrorOr<std::string> TargetSubdir =
-               findClangRelativeSysroot(getDriver(), getTriple(), SubdirName))
+  else if (llvm::ErrorOr<std::string> TargetSubdir = findClangRelativeSysroot(
+               getDriver(), LiteralTriple, getTriple(), SubdirName))
     Base = std::string(llvm::sys::path::parent_path(TargetSubdir.get()));
-  else if (llvm::ErrorOr<std::string> GPPName = findGcc(getTriple()))
+  else if (llvm::ErrorOr<std::string> GPPName =
+               findGcc(LiteralTriple, getTriple()))
     Base = std::string(llvm::sys::path::parent_path(
         llvm::sys::path::parent_path(GPPName.get())));
   else
@@ -433,7 +458,7 @@ toolchains::MinGW::MinGW(const Driver &D, const llvm::Triple &Triple,
         llvm::sys::path::parent_path(getDriver().getInstalledDir()));
 
   Base += llvm::sys::path::get_separator();
-  findGccLibDir();
+  findGccLibDir(LiteralTriple);
   // GccLibDir must precede Base/lib so that the
   // correct crtbegin.o ,cetend.o would be found.
   getFilePaths().push_back(GccLibDir);
@@ -576,6 +601,12 @@ void toolchains::MinGW::printVerboseInfo(raw_ostream &OS) const {
 // /usr/include/c++/4.8/backward
 // /usr/x86_64-w64-mingw32/include
 
+// Fedora
+// /usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/x86_64-w64-mingw32ucrt
+// /usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/backward
+// /usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include
+// /usr/lib/gcc/x86_64-w64-mingw32ucrt/12.2.1/include-fixed
+
 void toolchains::MinGW::AddClangSystemIncludeArgs(const ArgList &DriverArgs,
                                                   ArgStringList &CC1Args) const {
   if (DriverArgs.hasArg(options::OPT_nostdinc))
@@ -667,10 +698,11 @@ static bool testTriple(const Driver &D, const llvm::Triple &Triple,
   std::string SubdirName;
   if (D.SysRoot.size())
     return true;
+  llvm::Triple LiteralTriple = getLiteralTriple(D, Triple);
   if (llvm::ErrorOr<std::string> TargetSubdir =
-          findClangRelativeSysroot(D, Triple, SubdirName))
+          findClangRelativeSysroot(D, LiteralTriple, Triple, SubdirName))
     return true;
-  if (llvm::ErrorOr<std::string> GPPName = findGcc(Triple))
+  if (llvm::ErrorOr<std::string> GPPName = findGcc(LiteralTriple, Triple))
     return true;
   // If we neither found a colocated sysroot or a matching gcc executable,
   // conclude that we can't know if this is the correct spelling of the triple.
diff --git a/clang/lib/Driver/ToolChains/MinGW.h b/clang/lib/Driver/ToolChains/MinGW.h
index f15f99dc8a8c..bc17d273e776 100644
--- a/clang/lib/Driver/ToolChains/MinGW.h
+++ b/clang/lib/Driver/ToolChains/MinGW.h
@@ -108,7 +108,7 @@ private:
   std::string SubdirName;
   mutable std::unique_ptr<tools::gcc::Preprocessor> Preprocessor;
   mutable std::unique_ptr<tools::gcc::Compiler> Compiler;
-  void findGccLibDir();
+  void findGccLibDir(const llvm::Triple &LiteralTriple);
 
   bool NativeLLVMSupport;
 };
diff --git a/clang/test/Driver/Inputs/mingw_fedora_tree/usr/lib/gcc/x86_64-w64-mingw32ucrt/12.2.1/include-fixed/.keep b/clang/test/Driver/Inputs/mingw_fedora_tree/usr/lib/gcc/x86_64-w64-mingw32ucrt/12.2.1/include-fixed/.keep
new file mode 100644
index 000000000000..e69de29bb2d1
diff --git a/clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/.keep b/clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/.keep
new file mode 100644
index 000000000000..e69de29bb2d1
diff --git a/clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/backward/.keep b/clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/backward/.keep
new file mode 100644
index 000000000000..e69de29bb2d1
diff --git a/clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/x86_64-w64-mingw32ucrt/.keep b/clang/test/Driver/Inputs/mingw_fedora_tree/usr/x86_64-w64-mingw32ucrt/sys-root/mingw/include/c++/x86_64-w64-mingw32ucrt/.keep
new file mode 100644
index 000000000000..e69de29bb2d1
diff --git a/clang/test/Driver/mingw-sysroot.cpp b/clang/test/Driver/mingw-sysroot.cpp
index 1cb51b3aea2d..985be9e5f2ac 100644
--- a/clang/test/Driver/mingw-sysroot.cpp
+++ b/clang/test/Driver/mingw-sysroot.cpp
@@ -14,6 +14,11 @@
 // RUN: ln -s %S/Inputs/mingw_ubuntu_posix_tree/usr/x86_64-w64-mingw32 %T/testroot-clang/x86_64-w64-mingw32
 // RUN: ln -s %S/Inputs/mingw_arch_tree/usr/i686-w64-mingw32 %T/testroot-clang/i686-w64-mingw32
 
+// RUN: rm -rf %T/testroot-custom-triple
+// RUN: mkdir -p %T/testroot-custom-triple/bin
+// RUN: ln -s %clang %T/testroot-custom-triple/bin/x86_64-w64-mingw32foo-clang
+// RUN: ln -s %clang %T/testroot-custom-triple/bin/clang
+// RUN: ln -s %S/Inputs/mingw_ubuntu_posix_tree/usr/x86_64-w64-mingw32 %T/testroot-custom-triple/x86_64-w64-mingw32foo
 
 // If we find a gcc in the path with the right triplet prefix, pick that as
 // sysroot:
@@ -51,3 +56,11 @@
 
 // RUN: env "PATH=%T/testroot-gcc/bin:%PATH%" %T/testroot-clang/bin/x86_64-w64-mingw32-clang --target=x86_64-w64-mingw32 -m32 -rtlib=compiler-rt -stdlib=libstdc++ --sysroot="" -c -### %s 2>&1 | FileCheck -check-prefix=CHECK_TESTROOT_CLANG_I686 %s
 // CHECK_TESTROOT_CLANG_I686: "{{[^"]+}}/testroot-clang{{/|\\\\}}i686-w64-mingw32{{/|\\\\}}include"
+
+
+// If the user calls clang with a custom literal triple, make sure this maps
+// to sysroots with the matching spelling.
+
+// RUN: %T/testroot-custom-triple/bin/x86_64-w64-mingw32foo-clang -rtlib=compiler-rt -stdlib=libstdc++ --sysroot="" -c -### %s 2>&1 | FileCheck -check-prefix=CHECK_TESTROOT_CUSTOM_TRIPLE %s
+// RUN: %T/testroot-custom-triple/bin/clang --target=x86_64-w64-mingw32foo -rtlib=compiler-rt -stdlib=libstdc++ --sysroot="" -c -### %s 2>&1 | FileCheck -check-prefix=CHECK_TESTROOT_CUSTOM_TRIPLE %s
+// CHECK_TESTROOT_CUSTOM_TRIPLE: "{{[^"]+}}/testroot-custom-triple{{/|\\\\}}x86_64-w64-mingw32foo{{/|\\\\}}include"
diff --git a/clang/test/Driver/mingw.cpp b/clang/test/Driver/mingw.cpp
index b37503e475d8..bde952fb3565 100644
--- a/clang/test/Driver/mingw.cpp
+++ b/clang/test/Driver/mingw.cpp
@@ -39,6 +39,10 @@
 // CHECK_MINGW_OPENSUSE_TREE: "[[BASE]]/Inputs/mingw_opensuse_tree/usr{{/|\\\\}}x86_64-w64-mingw32/sys-root/mingw/include"
 
 
+// RUN: %clang -target x86_64-pc-windows-gnu -rtlib=platform -stdlib=libstdc++ -c -### --sysroot=%S/Inputs/mingw_fedora_tree/usr %s 2>&1 | FileCheck -check-prefix=CHECK_MINGW_FEDORA_TREE %s
+// CHECK_MINGW_FEDORA_TREE: "[[BASE:[^"]+]]/Inputs/mingw_fedora_tree/usr{{/|\\\\}}x86_64-w64-mingw32ucrt/sys-root/mingw/include"
+
+
 // RUN: %clang -target i686-pc-windows-gnu -rtlib=platform -stdlib=libstdc++ -c -### --sysroot=%S/Inputs/mingw_arch_tree/usr %s 2>&1 | FileCheck -check-prefix=CHECK_MINGW_ARCH_TREE %s
 // CHECK_MINGW_ARCH_TREE: "[[BASE:[^"]+]]/Inputs/mingw_arch_tree/usr{{/|\\\\}}i686-w64-mingw32{{/|\\\\}}include{{/|\\\\}}c++{{/|\\\\}}5.1.0"
 // CHECK_MINGW_ARCH_TREE: "[[BASE]]/Inputs/mingw_arch_tree/usr{{/|\\\\}}i686-w64-mingw32{{/|\\\\}}include{{/|\\\\}}c++{{/|\\\\}}5.1.0{{/|\\\\}}i686-w64-mingw32"
-- 
2.38.1