GNU logs - #38376, boring messages

Message sent to guix-patches@HIDDEN:

Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Carl Dong <contact@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Mon, 25 Nov 2019 21:33:44 +0000
Resent-Message-ID: <handler.38376.B.157470528021693 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
To: 38376 <at> debbugs.gnu.org
Cc: Carl Dong <contact@HIDDEN>
Reply-To: Carl Dong <contact@HIDDEN>
Date: Mon, 25 Nov 2019 18:01:48 +0000
From: Carl Dong <contact@HIDDEN>
Message-ID: <20191125180144.1060289-1-contact@HIDDEN>
Feedback-ID: a8j8tDUaJ4AYuDVBywMTwsJebN4w8TVXadJLsJb8td3t3dZi9RdXFlPaQvoFKnI9KgXySsPXcRkajVyY0cGTcA==:Ext:ProtonMail
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: list
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>

Prior to this commit, we had

1. Patched out a lot of search path detection in clang's codebase which
didn't make sense in Guix
2. Used -DC_INCLUDE_DIRS=3D to point clang to our libc, which, as a side
effect, causes the search path detection code in clang to return early,
and skip some search path detection

This was fine when running clang normally, but when a user supplies
clang with --sysroot, clang _should_ do the search path detection. This
commit fixes runing clang with --sysroot.

More details included as comments in the
gnu/packages/patches/clang-*-libc-search-path.patch files.

* gnu/packages/patches/clang-3.5-libc-search-path.patch: Restore search
  path detection code relative to $SYSROOT. Add @C_EXTRA_INCLUDE_DIRS@, a
  C_INCLUDE_DIRS replacement which does not return early.
* gnu/packages/patches/clang-3.8-libc-search-path.patch: Same as above.
* gnu/packages/patches/clang-6.0-libc-search-path.patch: Same as above.
* gnu/packages/patches/clang-7.0-libc-search-path.patch: Same as above.
* gnu/packages/llvm.scm: Substitute @C_EXTRA_INCLUDE_DIRS@ instead of
  using -DC_INCLUDE_DIRS=3D.
---
 gnu/packages/llvm.scm                         | 13 ++-
 .../patches/clang-3.5-libc-search-path.patch  | 74 ++++++++-------
 .../patches/clang-3.8-libc-search-path.patch  | 70 ++++++++-------
 .../patches/clang-6.0-libc-search-path.patch  | 77 ++++++++--------
 .../patches/clang-7.0-libc-search-path.patch  | 89 +++++++++----------
 5 files changed, 169 insertions(+), 154 deletions(-)

diff --git a/gnu/packages/llvm.scm b/gnu/packages/llvm.scm
index 082e6e96ca..383180e140 100644
--- a/gnu/packages/llvm.scm
+++ b/gnu/packages/llvm.scm
@@ -190,12 +190,7 @@ compiler.  In LLVM this library is called \"compiler-r=
t\".")
=20
              ;; Find libgcc_s, crtbegin.o, and crtend.o.
              (string-append "-DGCC_INSTALL_PREFIX=3D"
-                            (assoc-ref %build-inputs "gcc-lib"))
-
-             ;; Use a sane default include directory.
-             (string-append "-DC_INCLUDE_DIRS=3D"
-                            (assoc-ref %build-inputs "libc")
-                            "/include"))
+                            (assoc-ref %build-inputs "gcc-lib")))
=20
        ;; Don't use '-g' during the build to save space.
        #:build-type "Release"
@@ -207,6 +202,7 @@ compiler.  In LLVM this library is called \"compiler-rt=
\".")
                      (let ((libc (assoc-ref inputs "libc"))
                            (compiler-rt (assoc-ref inputs "clang-runtime")=
)
                            (gcc (assoc-ref inputs "gcc"))
+                           (kernel-headers (assoc-ref inputs "kernel-heade=
rs"))
                            (version
                             (string->number
                              ,(version-major (package-version clang-runtim=
e)))))
@@ -232,7 +228,10 @@ compiler.  In LLVM this library is called \"compiler-r=
t\".")
                             ;; Make sure libc's libdir is on the search pa=
th, to
                             ;; allow crt1.o & co. to be found.
                             (("@GLIBC_LIBDIR@")
-                             (string-append libc "/lib"))))
+                             (string-append libc "/lib"))
+                            (("@C_EXTRA_INCLUDE_DIRS@")
+                             (string-append libc "/include" ":"
+                                            kernel-headers "/include"))))
                          (else
                           (substitute* "lib/Driver/Tools.cpp"
                             ;; Patch the 'getLinuxDynamicLinker' function =
so that
diff --git a/gnu/packages/patches/clang-3.5-libc-search-path.patch b/gnu/pa=
ckages/patches/clang-3.5-libc-search-path.patch
index 50e4480239..54fb9be7d8 100644
--- a/gnu/packages/patches/clang-3.5-libc-search-path.patch
+++ b/gnu/packages/patches/clang-3.5-libc-search-path.patch
@@ -8,7 +8,7 @@ to make sure Clang also works on non-GuixSD systems.
=20
 --- cfe-3.6.0.src/lib/Driver/ToolChains.cpp=092015-02-18 22:03:07.00000000=
0 +0100
 +++ cfe-3.6.0.src/lib/Driver/ToolChains.cpp=092015-06-19 16:37:20.45970104=
4 +0200
-@@ -2931,6 +2931,9 @@ Linux::Linux(const Driver &D, const llvm
+@@ -3040,6 +3040,9 @@
 =20
    Linker =3D GetLinkerPath();
 =20
@@ -18,7 +18,7 @@ to make sure Clang also works on non-GuixSD systems.
    Distro Distro =3D DetectDistro(Arch);
 =20
    if (IsOpenSUSE(Distro) || IsUbuntu(Distro)) {
-@@ -2973,6 +2976,7 @@ Linux::Linux(const Driver &D, const llvm
+@@ -3082,6 +3085,7 @@
 =20
    if (IsOpenSUSE(Distro))
      ExtraOpts.push_back("--enable-new-dtags");
@@ -26,41 +26,49 @@ to make sure Clang also works on non-GuixSD systems.
 =20
    // The selection of paths to try here is designed to match the patterns=
 which
    // the GCC driver itself uses, as this is part of the GCC-compatible dr=
iver.
-@@ -3043,14 +3047,12 @@ Linux::Linux(const Driver &D, const llvm
-     addPathIfExists(D.Dir + "/../" + OSLibDir, Paths);
-   }
+@@ -3194,6 +3198,10 @@
 =20
--  addPathIfExists(SysRoot + "/lib/" + MultiarchTriple, Paths);
--  addPathIfExists(SysRoot + "/lib/../" + OSLibDir, Paths);
--  addPathIfExists(SysRoot + "/usr/lib/" + MultiarchTriple, Paths);
--  addPathIfExists(SysRoot + "/usr/lib/../" + OSLibDir, Paths);
--
-   // Try walking via the GCC triple path in case of biarch or multiarch G=
CC
-   // installations with strange symlinks.
-   if (GCCInstallation.isValid()) {
-+    // The following code would end up adding things like
-+    // "/usr/lib/x86_64-unknown-linux-gnu/../../lib64" to the search path=
.
-+#if 0
-     addPathIfExists(SysRoot + "/usr/lib/" + GCCInstallation.getTriple().s=
tr() +
-                     "/../../" + OSLibDir, Paths);
-=20
-@@ -3060,6 +3062,7 @@ Linux::Linux(const Driver &D, const llvm
-       addPathIfExists(GCCInstallation.getInstallPath() +
-                       BiarchSibling.gccSuffix(), Paths);
-     }
-+#endif
-=20
-     // See comments above on the multilib variant for details of why this=
 is
-     // included even from outside the sysroot.
-@@ -3083,8 +3086,9 @@ Linux::Linux(const Driver &D, const llvm
-   if (StringRef(D.Dir).startswith(SysRoot))
-     addPathIfExists(D.Dir + "/../lib", Paths);
-=20
--  addPathIfExists(SysRoot + "/lib", Paths);
--  addPathIfExists(SysRoot + "/usr/lib", Paths);
+   addPathIfExists(SysRoot + "/lib", Paths);
+   addPathIfExists(SysRoot + "/usr/lib", Paths);
++
 +  // Add libc's lib/ directory to the search path, so that crt1.o, crti.o=
,
 +  // and friends can be found.
 +  addPathIfExists("@GLIBC_LIBDIR@", Paths);
  }
 =20
  bool Linux::HasNativeLLVMSupport() const {
+@@ -3384,6 +3392,34 @@
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/include");
+=20
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/usr/include");
++
++  // Check for configure-time "extra" C include directories. When constru=
cting a
++  // toolchain, @C_EXTRA_INCLUDE_DIRS@ should be replaced with something =
like a
++  // colon-separated string of the include dirs of libc and kernel header=
s.
++  //
++  // The reason why we use this mechanism instead of C_INCLUDE_DIRS above=
 is
++  // because when a user supplies clang with --sysroot, the normal expect=
ation
++  // is that clang will detect and add the proper $SYSROOT/$MULTIARCHINCL=
,
++  // $SYSROOT/include, and $SYSROOT/usr/include to its list of search pat=
hs.
++  // However, if C_INCLUDE_DIRS is not empty, this function will return e=
arly
++  // and not attempt to add the aforementioned search paths, which is not
++  // desirable.
++  //
++  // By adding our configure-time "extra" C include directories here, aft=
er
++  // we've added $SYSROOT/include and $SYSROOT/usr/include, we make sure =
that IF
++  // --sysroot is supplied on the command line, we pick up the expected s=
earch
++  // paths in the $SYSROOT, and that they come before our configure-time =
"extra"
++  // C include directories.
++  StringRef CExtraIncludeDirs("@C_EXTRA_INCLUDE_DIRS@");
++  if (CExtraIncludeDirs !=3D "") {
++      SmallVector<StringRef, 5> dirs;
++      CExtraIncludeDirs.split(dirs, ":");
++      for (StringRef dir : dirs) {
++          StringRef Prefix =3D
++              llvm::sys::path::is_absolute(dir) ? StringRef(SysRoot) : ""=
;
++          addExternCSystemInclude(DriverArgs, CC1Args, Prefix + dir);
++      }
++  }
+ }
+=20
+ /// \brief Helper to add the variant paths of a libstdc++ installation.
diff --git a/gnu/packages/patches/clang-3.8-libc-search-path.patch b/gnu/pa=
ckages/patches/clang-3.8-libc-search-path.patch
index 0f7d0a4add..243d167755 100644
--- a/gnu/packages/patches/clang-3.8-libc-search-path.patch
+++ b/gnu/packages/patches/clang-3.8-libc-search-path.patch
@@ -29,41 +29,49 @@ changes in clang 3.8.
 =20
    // The selection of paths to try here is designed to match the patterns=
 which
    // the GCC driver itself uses, as this is part of the GCC-compatible dr=
iver.
-@@ -3771,14 +3775,12 @@
-     addPathIfExists(D, D.Dir + "/../" + OSLibDir, Paths);
-   }
+@@ -3817,6 +3821,10 @@
 =20
--  addPathIfExists(D, SysRoot + "/lib/" + MultiarchTriple, Paths);
--  addPathIfExists(D, SysRoot + "/lib/../" + OSLibDir, Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib/" + MultiarchTriple, Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib/../" + OSLibDir, Paths);
--
-   // Try walking via the GCC triple path in case of biarch or multiarch G=
CC
-   // installations with strange symlinks.
-   if (GCCInstallation.isValid()) {
-+    // The following code would end up adding things like
-+    // "/usr/lib/x86_64-unknown-linux-gnu/../../lib64" to the search path=
.
-+#if 0
-     addPathIfExists(D,
-                     SysRoot + "/usr/lib/" + GCCInstallation.getTriple().s=
tr() +
-                         "/../../" + OSLibDir,
-@@ -3791,6 +3793,7 @@
-                              BiarchSibling.gccSuffix(),
-                       Paths);
-     }
-+#endif
-=20
-     // See comments above on the multilib variant for details of why this=
 is
-     // included even from outside the sysroot.
-@@ -3815,8 +3818,9 @@
-   if (StringRef(D.Dir).startswith(SysRoot))
-     addPathIfExists(D, D.Dir + "/../lib", Paths);
-=20
--  addPathIfExists(D, SysRoot + "/lib", Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib", Paths);
+   addPathIfExists(D, SysRoot + "/lib", Paths);
+   addPathIfExists(D, SysRoot + "/usr/lib", Paths);
++
 +  // Add libc's lib/ directory to the search path, so that crt1.o, crti.o=
,
 +  // and friends can be found.
 +  addPathIfExists(D, "@GLIBC_LIBDIR@", Paths);
  }
 =20
  bool Linux::HasNativeLLVMSupport() const { return true; }
+@@ -4026,6 +4034,34 @@
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/include");
+=20
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/usr/include");
++
++  // Check for configure-time "extra" C include directories. When constru=
cting a
++  // toolchain, @C_EXTRA_INCLUDE_DIRS@ should be replaced with something =
like a
++  // colon-separated string of the include dirs of libc and kernel header=
s.
++  //
++  // The reason why we use this mechanism instead of C_INCLUDE_DIRS above=
 is
++  // because when a user supplies clang with --sysroot, the normal expect=
ation
++  // is that clang will detect and add the proper $SYSROOT/$MULTIARCHINCL=
,
++  // $SYSROOT/include, and $SYSROOT/usr/include to its list of search pat=
hs.
++  // However, if C_INCLUDE_DIRS is not empty, this function will return e=
arly
++  // and not attempt to add the aforementioned search paths, which is not
++  // desirable.
++  //
++  // By adding our configure-time "extra" C include directories here, aft=
er
++  // we've added $SYSROOT/include and $SYSROOT/usr/include, we make sure =
that IF
++  // --sysroot is supplied on the command line, we pick up the expected s=
earch
++  // paths in the $SYSROOT, and that they come before our configure-time =
"extra"
++  // C include directories.
++  StringRef CExtraIncludeDirs("@C_EXTRA_INCLUDE_DIRS@");
++  if (CExtraIncludeDirs !=3D "") {
++      SmallVector<StringRef, 5> dirs;
++      CExtraIncludeDirs.split(dirs, ":");
++      for (StringRef dir : dirs) {
++          StringRef Prefix =3D
++              llvm::sys::path::is_absolute(dir) ? StringRef(SysRoot) : ""=
;
++          addExternCSystemInclude(DriverArgs, CC1Args, Prefix + dir);
++      }
++  }
+ }
+=20
+=20
diff --git a/gnu/packages/patches/clang-6.0-libc-search-path.patch b/gnu/pa=
ckages/patches/clang-6.0-libc-search-path.patch
index a62e8063c2..d80c798467 100644
--- a/gnu/packages/patches/clang-6.0-libc-search-path.patch
+++ b/gnu/packages/patches/clang-6.0-libc-search-path.patch
@@ -8,18 +8,17 @@ to make sure Clang also works on non-GuixSD systems.
=20
 --- cfe-6.0.0.src/lib/Driver/ToolChains/Linux.cpp
 +++ cfe-6.0.0.src/lib/Driver/ToolChains/Linux.cpp
-@@ -207,7 +207,9 @@
-   PPaths.push_back(Twine(GCCInstallation.getParentLibPath() + "/../" +
+@@ -208,6 +208,9 @@
                           GCCInstallation.getTriple().str() + "/bin")
                         .str());
--
+=20
 +  // Comment out the distro-specific tweaks so that they don't bite when
 +  // using Guix on a foreign distro.
 +#if 0
    Distro Distro(D.getVFS());
 =20
    if (Distro.IsAlpineLinux()) {
-@@ -255,6 +257,7 @@
+@@ -255,6 +258,7 @@
 =20
    if (IsAndroid || Distro.IsOpenSUSE())
      ExtraOpts.push_back("--enable-new-dtags");
@@ -27,41 +26,49 @@ to make sure Clang also works on non-GuixSD systems.
 =20
    // The selection of paths to try here is designed to match the patterns=
 which
    // the GCC driver itself uses, as this is part of the GCC-compatible dr=
iver.
-@@ -329,14 +332,12 @@
-     addPathIfExists(D, D.Dir + "/../" + OSLibDir, Paths);
-   }
-=20
--  addPathIfExists(D, SysRoot + "/lib/" + MultiarchTriple, Paths);
--  addPathIfExists(D, SysRoot + "/lib/../" + OSLibDir, Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib/" + MultiarchTriple, Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib/../" + OSLibDir, Paths);
--
-   // Try walking via the GCC triple path in case of biarch or multiarch G=
CC
-   // installations with strange symlinks.
-   if (GCCInstallation.isValid()) {
-+    // The following code would end up adding things like
-+    // "/usr/lib/x86_64-unknown-linux-gnu/../../lib64" to the search path=
.
-+#if 0
-     addPathIfExists(D,
-                     SysRoot + "/usr/lib/" + GCCInstallation.getTriple().s=
tr() +
-                         "/../../" + OSLibDir,
-@@ -349,6 +350,7 @@
-                              BiarchSibling.gccSuffix(),
-                       Paths);
-     }
-+#endif
+@@ -375,6 +379,10 @@
 =20
-     // See comments above on the multilib variant for details of why this=
 is
-     // included even from outside the sysroot.
-@@ -373,8 +375,9 @@
-   if (StringRef(D.Dir).startswith(SysRoot))
-     addPathIfExists(D, D.Dir + "/../lib", Paths);
-=20
--  addPathIfExists(D, SysRoot + "/lib", Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib", Paths);
+   addPathIfExists(D, SysRoot + "/lib", Paths);
+   addPathIfExists(D, SysRoot + "/usr/lib", Paths);
++
 +  // Add libc's lib/ directory to the search path, so that crt1.o, crti.o=
,
 +  // and friends can be found.
 +  addPathIfExists(D, "@GLIBC_LIBDIR@", Paths);
  }
 =20
  bool Linux::HasNativeLLVMSupport() const { return true; }
+@@ -710,6 +718,34 @@
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/include");
+=20
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/usr/include");
++
++  // Check for configure-time "extra" C include directories. When constru=
cting a
++  // toolchain, @C_EXTRA_INCLUDE_DIRS@ should be replaced with something =
like a
++  // colon-separated string of the include dirs of libc and kernel header=
s.
++  //
++  // The reason why we use this mechanism instead of C_INCLUDE_DIRS above=
 is
++  // because when a user supplies clang with --sysroot, the normal expect=
ation
++  // is that clang will detect and add the proper $SYSROOT/$MULTIARCHINCL=
,
++  // $SYSROOT/include, and $SYSROOT/usr/include to its list of search pat=
hs.
++  // However, if C_INCLUDE_DIRS is not empty, this function will return e=
arly
++  // and not attempt to add the aforementioned search paths, which is not
++  // desirable.
++  //
++  // By adding our configure-time "extra" C include directories here, aft=
er
++  // we've added $SYSROOT/include and $SYSROOT/usr/include, we make sure =
that IF
++  // --sysroot is supplied on the command line, we pick up the expected s=
earch
++  // paths in the $SYSROOT, and that they come before our configure-time =
"extra"
++  // C include directories.
++  StringRef CExtraIncludeDirs("@C_EXTRA_INCLUDE_DIRS@");
++  if (CExtraIncludeDirs !=3D "") {
++      SmallVector<StringRef, 5> dirs;
++      CExtraIncludeDirs.split(dirs, ":");
++      for (StringRef dir : dirs) {
++          StringRef Prefix =3D
++              llvm::sys::path::is_absolute(dir) ? StringRef(SysRoot) : ""=
;
++          addExternCSystemInclude(DriverArgs, CC1Args, Prefix + dir);
++      }
++  }
+ }
+=20
+ static std::string DetectLibcxxIncludePath(StringRef base) {
diff --git a/gnu/packages/patches/clang-7.0-libc-search-path.patch b/gnu/pa=
ckages/patches/clang-7.0-libc-search-path.patch
index 07ff8c90bd..445b5d17ac 100644
--- a/gnu/packages/patches/clang-7.0-libc-search-path.patch
+++ b/gnu/packages/patches/clang-7.0-libc-search-path.patch
@@ -8,7 +8,7 @@ to make sure Clang also works on non-GuixSD systems.
=20
 --- a/lib/Driver/ToolChains/Linux.cpp
 +++ b/lib/Driver/ToolChains/Linux.cpp
-@@ -225,7 +225,9 @@ Linux::Linux(const Driver &D, const llvm::Triple &Trip=
le, const ArgList &Args)
+@@ -225,7 +225,9 @@
    PPaths.push_back(Twine(GCCInstallation.getParentLibPath() + "/../" +
                           GCCInstallation.getTriple().str() + "/bin")
                         .str());
@@ -19,7 +19,7 @@ to make sure Clang also works on non-GuixSD systems.
    Distro Distro(D.getVFS());
 =20
    if (Distro.IsAlpineLinux()) {
-@@ -284,6 +286,7 @@ Linux::Linux(const Driver &D, const llvm::Triple &Trip=
le, const ArgList &Args)
+@@ -284,6 +286,7 @@
 =20
    if (IsAndroid || Distro.IsOpenSUSE())
      ExtraOpts.push_back("--enable-new-dtags");
@@ -27,56 +27,49 @@ to make sure Clang also works on non-GuixSD systems.
 =20
    // The selection of paths to try here is designed to match the patterns=
 which
    // the GCC driver itself uses, as this is part of the GCC-compatible dr=
iver.
-@@ -342,7 +345,7 @@ Linux::Linux(const Driver &D, const llvm::Triple &Trip=
le, const ArgList &Args)
-     // the cross. Note that GCC does include some of these directories in=
 some
-     // configurations but this seems somewhere between questionable and s=
imply
-     // a bug.
--    if (StringRef(LibPath).startswith(SysRoot)) {
-+    if (0) {
-       addPathIfExists(D, LibPath + "/" + MultiarchTriple, Paths);
-       addPathIfExists(D, LibPath + "/../" + OSLibDir, Paths);
-     }
-@@ -361,6 +364,8 @@ Linux::Linux(const Driver &D, const llvm::Triple &Trip=
le, const ArgList &Args)
-   addPathIfExists(D, SysRoot + "/lib/" + MultiarchTriple, Paths);
-   addPathIfExists(D, SysRoot + "/lib/../" + OSLibDir, Paths);
+@@ -431,6 +434,10 @@
 =20
-+  // This requires the commented distro tweaks above.
-+#if 0
-   if (IsAndroid) {
-     // Android sysroots contain a library directory for each supported OS
-     // version as well as some unversioned libraries in the usual multiar=
ch
-@@ -389,10 +394,14 @@ Linux::Linux(const Driver &D, const llvm::Triple &Tr=
iple, const ArgList &Args)
-     addPathIfExists(D, SysRoot + "/" + OSLibDir + "/" + ABIName, Paths);
-     addPathIfExists(D, SysRoot + "/usr/" + OSLibDir + "/" + ABIName, Path=
s);
-   }
-+#endif
-=20
-   // Try walking via the GCC triple path in case of biarch or multiarch G=
CC
-   // installations with strange symlinks.
-   if (GCCInstallation.isValid()) {
-+    // The following code would end up adding things like
-+    // "/usr/lib/x86_64-unknown-linux-gnu/../../lib64" to the search path=
.
-+#if 0
-     addPathIfExists(D,
-                     SysRoot + "/usr/lib/" + GCCInstallation.getTriple().s=
tr() +
-                         "/../../" + OSLibDir,
-@@ -405,6 +414,7 @@ Linux::Linux(const Driver &D, const llvm::Triple &Trip=
le, const ArgList &Args)
-                              BiarchSibling.gccSuffix(),
-                       Paths);
-     }
-+#endif
-=20
-     // See comments above on the multilib variant for details of why this=
 is
-     // included even from outside the sysroot.
-@@ -429,8 +439,9 @@ Linux::Linux(const Driver &D, const llvm::Triple &Trip=
le, const ArgList &Args)
-   if (StringRef(D.Dir).startswith(SysRoot))
-     addPathIfExists(D, D.Dir + "/../lib", Paths);
-=20
--  addPathIfExists(D, SysRoot + "/lib", Paths);
--  addPathIfExists(D, SysRoot + "/usr/lib", Paths);
+   addPathIfExists(D, SysRoot + "/lib", Paths);
+   addPathIfExists(D, SysRoot + "/usr/lib", Paths);
++
 +  // Add libc's lib/ directory to the search path, so that crt1.o, crti.o=
,
 +  // and friends can be found.
 +  addPathIfExists(D, "@GLIBC_LIBDIR@", Paths);
  }
 =20
  bool Linux::HasNativeLLVMSupport() const { return true; }
+@@ -794,6 +801,34 @@
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/include");
+=20
+   addExternCSystemInclude(DriverArgs, CC1Args, SysRoot + "/usr/include");
++
++  // Check for configure-time "extra" C include directories. When constru=
cting a
++  // toolchain, @C_EXTRA_INCLUDE_DIRS@ should be replaced with something =
like a
++  // colon-separated string of the include dirs of libc and kernel header=
s.
++  //
++  // The reason why we use this mechanism instead of C_INCLUDE_DIRS above=
 is
++  // because when a user supplies clang with --sysroot, the normal expect=
ation
++  // is that clang will detect and add the proper $SYSROOT/$MULTIARCHINCL=
,
++  // $SYSROOT/include, and $SYSROOT/usr/include to its list of search pat=
hs.
++  // However, if C_INCLUDE_DIRS is not empty, this function will return e=
arly
++  // and not attempt to add the aforementioned search paths, which is not
++  // desirable.
++  //
++  // By adding our configure-time "extra" C include directories here, aft=
er
++  // we've added $SYSROOT/include and $SYSROOT/usr/include, we make sure =
that IF
++  // --sysroot is supplied on the command line, we pick up the expected s=
earch
++  // paths in the $SYSROOT, and that they come before our configure-time =
"extra"
++  // C include directories.
++  StringRef CExtraIncludeDirs("@C_EXTRA_INCLUDE_DIRS@");
++  if (CExtraIncludeDirs !=3D "") {
++      SmallVector<StringRef, 5> dirs;
++      CExtraIncludeDirs.split(dirs, ":");
++      for (StringRef dir : dirs) {
++          StringRef Prefix =3D
++              llvm::sys::path::is_absolute(dir) ? StringRef(SysRoot) : ""=
;
++          addExternCSystemInclude(DriverArgs, CC1Args, Prefix + dir);
++      }
++  }
+ }
+=20
+ static std::string DetectLibcxxIncludePath(StringRef base) {
--=20
2.24.0

Message sent:

Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-Mailer: MIME-tools 5.505 (Entity 5.505)
Content-Type: text/plain; charset=utf-8
X-Loop: help-debbugs@HIDDEN
From: help-debbugs@HIDDEN (GNU bug Tracking System)
To: Carl Dong <contact@HIDDEN>
Subject: bug#38376: Acknowledgement ([PATCH] gnu: clang: Fix search path
 detection with --sysroot.)
Message-ID: <handler.38376.B.157470528021693.ack <at> debbugs.gnu.org>
References: <20191125180144.1060289-1-contact@HIDDEN>
X-Gnu-PR-Message: ack 38376
X-Gnu-PR-Package: guix-patches
X-Gnu-PR-Keywords: patch
Reply-To: 38376 <at> debbugs.gnu.org
Date: Mon, 25 Nov 2019 21:33:44 +0000

Thank you for filing a new bug report with debbugs.gnu.org.

This is an automatically generated reply to let you know your message
has been received.

Your message is being forwarded to the package maintainers and other
interested parties for their attention; they will reply in due course.

Your message has been sent to the package maintainer(s):
 guix-patches@HIDDEN

If you wish to submit further information on this problem, please
send it to 38376 <at> debbugs.gnu.org.

Please do not send mail to help-debbugs@HIDDEN unless you wish
to report a problem with the Bug-tracking system.

--=20
38376: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=3D38376
GNU Bug Tracking System
Contact help-debbugs@HIDDEN with problems

Message sent to guix-patches@HIDDEN:

Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Mathieu Othacehe <m.othacehe@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Tue, 26 Nov 2019 07:16:01 +0000
Resent-Message-ID: <handler.38376.B38376.15747525304962 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
To: Carl Dong <contact@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
References: <20191125180144.1060289-1-contact@HIDDEN>
User-agent: mu4e 1.2.0; emacs 26.3
From: Mathieu Othacehe <m.othacehe@HIDDEN>
In-reply-to: <20191125180144.1060289-1-contact@HIDDEN>
Date: Tue, 26 Nov 2019 08:15:20 +0100
Message-ID: <8736ebt8fr.fsf@HIDDEN>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: list
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>

Hello Carl,

> This was fine when running clang normally, but when a user supplies
> clang with --sysroot, clang _should_ do the search path detection. This
> commit fixes runing clang with --sysroot.

Thanks for your patch. Do you think this could be done in a way that can
be upstreamed? In lib/Driver/ToolChains/Gnu.cpp, there are some Gentoo
specific stuff, maybe they would agree to integrate some Guix/nix specific
stuff too.

WDYT?

Mathieu

Message sent to guix-patches@HIDDEN:

Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Carl Dong <contact@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Tue, 26 Nov 2019 17:06:02 +0000
Resent-Message-ID: <handler.38376.B38376.157478792320078 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
To: Mathieu Othacehe <m.othacehe@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
Reply-To: Carl Dong <contact@HIDDEN>
Date: Tue, 26 Nov 2019 17:05:05 +0000
From: Carl Dong <contact@HIDDEN>
Message-ID: <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
In-Reply-To: <8736ebt8fr.fsf@HIDDEN>
References: <20191125180144.1060289-1-contact@HIDDEN>
 <8736ebt8fr.fsf@HIDDEN>
Feedback-ID: a8j8tDUaJ4AYuDVBywMTwsJebN4w8TVXadJLsJb8td3t3dZi9RdXFlPaQvoFKnI9KgXySsPXcRkajVyY0cGTcA==:Ext:ProtonMail
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: list
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>

Hi Mathieu!

I think that=E2=80=99s an excellent idea but unfortunately I don=E2=80=
=99t have the bandwidth for it currently=E2=80=A6 I=E2=80=99ll put it on my=
 TODO list though :-)
Please do let me know if there=E2=80=99s anything amiss in the patch as it =
stands right now!

Cheers,
Carl Dong

> On Nov 26, 2019, at 2:15 AM, Mathieu Othacehe <m.othacehe@HIDDEN> wrot=
e:
>=20
>=20
>=20
> Hello Carl,
>=20
>> This was fine when running clang normally, but when a user supplies
>> clang with --sysroot, clang _should_ do the search path detection. This
>> commit fixes runing clang with --sysroot.
>=20
> Thanks for your patch. Do you think this could be done in a way that can
> be upstreamed? In lib/Driver/ToolChains/Gnu.cpp, there are some Gentoo
> specific stuff, maybe they would agree to integrate some Guix/nix specifi=
c
> stuff too.
>=20
> WDYT?
>=20
> Mathieu
>=20
>=20
>=20

Message sent to guix-patches@HIDDEN:

X-Loop: help-debbugs@HIDDEN
Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Mathieu Othacehe <m.othacehe@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Mon, 02 Dec 2019 14:04:02 +0000
Resent-Message-ID: <handler.38376.B38376.157529543519146 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
X-GNU-PR-Message: followup 38376
X-GNU-PR-Package: guix-patches
X-GNU-PR-Keywords: patch
To: Carl Dong <contact@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
Received: via spool by 38376-submit <at> debbugs.gnu.org id=B38376.157529543519146
          (code B ref 38376); Mon, 02 Dec 2019 14:04:02 +0000
Received: (at 38376) by debbugs.gnu.org; 2 Dec 2019 14:03:55 +0000
Received: from localhost ([127.0.0.1]:37264 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1ibmIs-0004yk-Ii
	for submit <at> debbugs.gnu.org; Mon, 02 Dec 2019 09:03:54 -0500
Received: from mail-wr1-f65.google.com ([209.85.221.65]:38152)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <m.othacehe@HIDDEN>) id 1ibmIq-0004yW-Ld
 for 38376 <at> debbugs.gnu.org; Mon, 02 Dec 2019 09:03:53 -0500
Received: by mail-wr1-f65.google.com with SMTP id y17so5782150wrh.5
 for <38376 <at> debbugs.gnu.org>; Mon, 02 Dec 2019 06:03:52 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;
 h=references:user-agent:from:to:cc:subject:in-reply-to:date
 :message-id:mime-version:content-transfer-encoding;
 bh=hamd7aoWIrSiXv2onD+yHfydvpC6/2OmD7Wz1oJ5psA=;
 b=QJNL+gm8Es6qY2Z9Dvd9tX1jaImySOnzXRF/pb09pTL92S4qv7Spzu9bh0TjaFxu6e
 6pJqILIZCCn8PxNfqTkWTCGVtZJz85wfBW2PK0cfF/BPxeWOiEGprsUR8nNGQZFelhbI
 q9AUgwOsL3/yn+KPxWO8E9wo2Pm+sqZ4v/L82uzMxm00aT7mGtDtj/etQQsBRsiIDWfA
 xwvVdY8m0fM9SJqAQXUU5C3zUwIK+/EK4ilE59QujPyB97Hmu3GDyyxvdLoPbN66pDnp
 m+b+wWp6FpDrftkxjY9myOeOpodofzEoVvIOrGhudF0ZrOYbFZ+7ILA6Xa0ev6GgYNtk
 Qp0Q==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
 d=1e100.net; s=20161025;
 h=x-gm-message-state:references:user-agent:from:to:cc:subject
 :in-reply-to:date:message-id:mime-version:content-transfer-encoding;
 bh=hamd7aoWIrSiXv2onD+yHfydvpC6/2OmD7Wz1oJ5psA=;
 b=eS968iIhdUMGuhiXGjdVSQQbOwd1g6SroXW8E1tNL0XdKKnKtLnH9ByWHOW7M6tSMh
 LcQcCGcH0x9eWKQJmTXk8uL65mlT9Mc5DKIC2mB68rmdfVFPNNx+H0hTAOC59FFfm+vv
 O36TRMWMsFSHQqIOe4uydZrD3m52/SHroyo3UZCG6QRVID3jl6HVNcgbVeatRmg3FfNL
 Rv/g1+wChAHJE9FR0KsdJ5j9c8gLvuvFvaHwjd2ceHWcETiNyrU2H0YlHURWg1H1tWlF
 9bHLGBRvqL0A8vGwe36xAEKzTw5c9urTlU8chsK5baaHvGSbjsLdALyiRLA0XHeymaWS
 IkEA==
X-Gm-Message-State: APjAAAWdHQ4+tVhRbRyDUvuv8NbThSBX5PpRs/XRtZPPmDdhmQTiy34e
 fjEhP5Arws/VmYnsndzFXdvmVsR8
X-Google-Smtp-Source: APXvYqxQxUfi/he4ms/xsuVWOVF9a3fKXAjL9B2j33ERTgAIMgKo7Wg9P5bp+mVeGYVXPrVMtsXoLw==
X-Received: by 2002:adf:f604:: with SMTP id t4mr13102017wrp.33.1575295426536; 
 Mon, 02 Dec 2019 06:03:46 -0800 (PST)
Received: from meru (lfbn-ann-1-237-90.w86-200.abo.wanadoo.fr. [86.200.196.90])
 by smtp.gmail.com with ESMTPSA id a14sm6792956wrx.81.2019.12.02.06.03.45
 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256);
 Mon, 02 Dec 2019 06:03:45 -0800 (PST)
References: <20191125180144.1060289-1-contact@HIDDEN>
 <8736ebt8fr.fsf@HIDDEN> <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
User-agent: mu4e 1.2.0; emacs 26.3
From: Mathieu Othacehe <m.othacehe@HIDDEN>
In-reply-to: <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
Date: Mon, 02 Dec 2019 15:03:44 +0100
Message-ID: <8736e2lt8f.fsf@HIDDEN>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.1 (/)
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.1 (-)


Hello Carl,

> Please do let me know if there=E2=80=99s anything amiss in the patch as i=
t stands right now!

Sorry for the delay! Well in the current form we end up with 4 big
patches to clang + 2 that are missing for clang-8 and clang-9.

This makes 6 patches, it will keep increasing. So maintaing those
patches will be very hard. I'd like to find another way to proceed.

Could we instead patch "clang-from-llvm" procedure to fix sysroot
support?

Mathieu

Message sent to guix-patches@HIDDEN:

X-Loop: help-debbugs@HIDDEN
Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Carl Dong <contact@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Mon, 02 Dec 2019 18:11:02 +0000
Resent-Message-ID: <handler.38376.B38376.157531026015371 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
X-GNU-PR-Message: followup 38376
X-GNU-PR-Package: guix-patches
X-GNU-PR-Keywords: patch
To: Mathieu Othacehe <m.othacehe@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
Reply-To: Carl Dong <contact@HIDDEN>
Received: via spool by 38376-submit <at> debbugs.gnu.org id=B38376.157531026015371
          (code B ref 38376); Mon, 02 Dec 2019 18:11:02 +0000
Received: (at 38376) by debbugs.gnu.org; 2 Dec 2019 18:11:00 +0000
Received: from localhost ([127.0.0.1]:38914 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1ibq9z-0003zr-OS
	for submit <at> debbugs.gnu.org; Mon, 02 Dec 2019 13:10:59 -0500
Received: from mail-40134.protonmail.ch ([185.70.40.134]:63531)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <contact@HIDDEN>) id 1ibq9w-0003zc-RT
 for 38376 <at> debbugs.gnu.org; Mon, 02 Dec 2019 13:10:57 -0500
Date: Mon, 02 Dec 2019 18:10:36 +0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=carldong.me;
 s=protonmail; t=1575310250;
 bh=17an/Xl1EEQgp56jehBfsWMGKC/tIhcKow2BGlirLKQ=;
 h=Date:To:From:Cc:Reply-To:Subject:In-Reply-To:References:
 Feedback-ID:From;
 b=D9MjBsPWRW9reXcRp91mRb28FtQLfJ8ntWOaRQ8gM4wC59j3DXE8QcY4Sg0vNvPZz
 wE+IXEfeD5R6cQBTroDA34rh/Dt1FGZHbOCjMYmFsnfveRtS1rk7//57Of1bHe7TmE
 diJiY9dJxMx71mOoypqKPgaCHaQszlvAxJ+GmS3g=
From: Carl Dong <contact@HIDDEN>
Message-ID: <A72CBE3E-CB34-4694-BC3F-07517168B693@HIDDEN>
In-Reply-To: <8736e2lt8f.fsf@HIDDEN>
References: <20191125180144.1060289-1-contact@HIDDEN>
 <8736ebt8fr.fsf@HIDDEN>
 <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
 <8736e2lt8f.fsf@HIDDEN>
Feedback-ID: a8j8tDUaJ4AYuDVBywMTwsJebN4w8TVXadJLsJb8td3t3dZi9RdXFlPaQvoFKnI9KgXySsPXcRkajVyY0cGTcA==:Ext:ProtonMail
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Status: No, score=-1.2 required=7.0 tests=ALL_TRUSTED,DKIM_SIGNED,
 DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF autolearn=ham
 autolearn_force=no version=3.4.2
X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on mail.protonmail.ch
X-Spam-Score: -0.0 (/)
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.0 (-)

> This makes 6 patches, it will keep increasing. So maintaing those
> patches will be very hard. I'd like to find another way to proceed.
I agree it would be very nice if we can avoid maintaining these patches=
=E2=80=A6

> Could we instead patch "clang-from-llvm" procedure to fix sysroot
> support?
Do you have anything particular in mind? I can=E2=80=99t think of any clean=
 ways to do this off the top of my head right now unfortunately :-(

Cheers,
Carl Dong

Message sent to guix-patches@HIDDEN:

X-Loop: help-debbugs@HIDDEN
Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Mathieu Othacehe <m.othacehe@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Mon, 02 Dec 2019 19:39:02 +0000
Resent-Message-ID: <handler.38376.B38376.157531549124474 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
X-GNU-PR-Message: followup 38376
X-GNU-PR-Package: guix-patches
X-GNU-PR-Keywords: patch
To: Carl Dong <contact@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
Received: via spool by 38376-submit <at> debbugs.gnu.org id=B38376.157531549124474
          (code B ref 38376); Mon, 02 Dec 2019 19:39:02 +0000
Received: (at 38376) by debbugs.gnu.org; 2 Dec 2019 19:38:11 +0000
Received: from localhost ([127.0.0.1]:38965 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1ibrWN-0006Mg-BZ
	for submit <at> debbugs.gnu.org; Mon, 02 Dec 2019 14:38:11 -0500
Received: from mail-wr1-f65.google.com ([209.85.221.65]:46432)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <m.othacehe@HIDDEN>) id 1ibrWL-0006MT-HW
 for 38376 <at> debbugs.gnu.org; Mon, 02 Dec 2019 14:38:10 -0500
Received: by mail-wr1-f65.google.com with SMTP id z7so594011wrl.13
 for <38376 <at> debbugs.gnu.org>; Mon, 02 Dec 2019 11:38:09 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;
 h=references:user-agent:from:to:cc:subject:in-reply-to:date
 :message-id:mime-version:content-transfer-encoding;
 bh=UWnPlDqmumqlI1lThRlYlr+28uAAFIs6zCCL03Slw+U=;
 b=uZOsPXioCOK07gVSPUcmpnSW65G3D+LiqPsozLCjR9HCb58h6ATvVFBJVSZwW+3wTP
 yQ1yMod2p8158/TF52qb12cI1eqDO8v9KCM7pK21tj5zpPcJgiC5JUno/z1W6URvUQhc
 tkKwM9a8XRpWfLKQ+p5AUff3QjNQG1qMs9ANz5EWAezRi78/uf/Ifu5MsCiGnl6GrWvb
 axXa/bPOSxJo8QJSF+eXB/PmkdL06juu1r6dh/vWsNW+L/QhtFtrkAThPabiOt1fvKnK
 0WiokblSYCUHjliV135gDnF3/JqYPOavMYHGHPNl9ShX+QoxBSwtAUkfrzQzzAtUEkFL
 BlVA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
 d=1e100.net; s=20161025;
 h=x-gm-message-state:references:user-agent:from:to:cc:subject
 :in-reply-to:date:message-id:mime-version:content-transfer-encoding;
 bh=UWnPlDqmumqlI1lThRlYlr+28uAAFIs6zCCL03Slw+U=;
 b=UHCCGO3eTTRLSohVgMRPShNtFoH/gEcuL8GYD7sGMGKMKDupdmDFQYklaCLV6CMi5a
 1lzz2MiFlKEx7zpH4z3aHAZ5lFKBPrG46Tmkd6n/s1VLxT29YQhcXh+kt5A+gzI/rg/x
 UnxMpDJir+OERt8r0UBWPCZ69ydhE7mlwuNVJRZlCEHeF/eudzIWZY6jgHPFCDAqEwp5
 KERuOHj6n1mwWr9KljQXO8Km9Fwtx58ZvcUQFkbFZ8N/YFi75gpKoFC7wAogMsYC2pHe
 Uh0bGJY9/BYEPXETa/4XaRtrlGpM4Ll6XykxdruLrXJy2WrTT5KWITj9ROGSZC9P2C1M
 3ZDQ==
X-Gm-Message-State: APjAAAXpUrzschD0v3x56wV0JGA34NiuOoVBVnhZAKc2ZKcLEdnoWWYW
 eW8YZKOGT2gtPhQG0JtOLmUpUPMT
X-Google-Smtp-Source: APXvYqwuHMM/7zarFowAa6suGniPtFvpn0R5zWZDHeGCv/o8kiU5qKj0QQqWPx7JEq3s6gy7KNQTUg==
X-Received: by 2002:adf:e2cc:: with SMTP id d12mr715199wrj.168.1575315483293; 
 Mon, 02 Dec 2019 11:38:03 -0800 (PST)
Received: from cervin ([2a01:e0a:fa:a50:9921:bdb7:f49c:a63])
 by smtp.gmail.com with ESMTPSA id e8sm551678wrt.7.2019.12.02.11.38.01
 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256);
 Mon, 02 Dec 2019 11:38:02 -0800 (PST)
References: <20191125180144.1060289-1-contact@HIDDEN>
 <8736ebt8fr.fsf@HIDDEN> <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
 <8736e2lt8f.fsf@HIDDEN> <A72CBE3E-CB34-4694-BC3F-07517168B693@HIDDEN>
User-agent: mu4e 1.2.0; emacs 26.3
From: Mathieu Othacehe <m.othacehe@HIDDEN>
In-reply-to: <A72CBE3E-CB34-4694-BC3F-07517168B693@HIDDEN>
Date: Mon, 02 Dec 2019 20:38:01 +0100
Message-ID: <877e3ewmau.fsf@HIDDEN>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Score: -0.1 (/)
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.1 (-)


>> Could we instead patch "clang-from-llvm" procedure to fix sysroot
>> support?
> Do you have anything particular in mind? I can=E2=80=99t think of any cle=
an ways to do this off the top of my head right now unfortunately :-(

No sadly, nothing in mind at the moment, but I'll have a look tomorrow!

Thanks,

Mathieu

Message sent to guix-patches@HIDDEN:

X-Loop: help-debbugs@HIDDEN
Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Carl Dong <contact@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Thu, 05 Dec 2019 18:27:02 +0000
Resent-Message-ID: <handler.38376.B38376.157557037831345 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
X-GNU-PR-Message: followup 38376
X-GNU-PR-Package: guix-patches
X-GNU-PR-Keywords: patch
To: Mathieu Othacehe <m.othacehe@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
Reply-To: Carl Dong <contact@HIDDEN>
Received: via spool by 38376-submit <at> debbugs.gnu.org id=B38376.157557037831345
          (code B ref 38376); Thu, 05 Dec 2019 18:27:02 +0000
Received: (at 38376) by debbugs.gnu.org; 5 Dec 2019 18:26:18 +0000
Received: from localhost ([127.0.0.1]:45172 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1icvpS-00089V-2p
	for submit <at> debbugs.gnu.org; Thu, 05 Dec 2019 13:26:18 -0500
Received: from mail1.protonmail.ch ([185.70.40.18]:16117)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <contact@HIDDEN>) id 1icvpQ-00089F-B8
 for 38376 <at> debbugs.gnu.org; Thu, 05 Dec 2019 13:26:17 -0500
Date: Thu, 05 Dec 2019 18:26:07 +0000
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=carldong.me;
 s=protonmail; t=1575570369;
 bh=4HxEqGq3MJLvkOmGvwzM0Ks8eGM9fWtKA5XCj7coiN4=;
 h=Date:To:From:Cc:Reply-To:Subject:In-Reply-To:References:
 Feedback-ID:From;
 b=T9UU2Ib9jJ0a07EsuAEG1Mx4kgX/R0DpDFTUnE/0LmS4ByrTqznwwKHoTB8vyosxb
 fr0I9WFRAlB+ayMpx21a6zL3iOUPX8p9+6olBuDfZd7cOIhp1SErXn71FeprmXrmOZ
 Azgspu23JBgeG3aEahb341mZoqLqCNzChc1VHvk8=
From: Carl Dong <contact@HIDDEN>
Message-ID: <33E84CBF-D451-4512-A515-1DA6A5E26F20@HIDDEN>
In-Reply-To: <877e3ewmau.fsf@HIDDEN>
References: <20191125180144.1060289-1-contact@HIDDEN>
 <8736ebt8fr.fsf@HIDDEN>
 <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
 <8736e2lt8f.fsf@HIDDEN>
 <A72CBE3E-CB34-4694-BC3F-07517168B693@HIDDEN>
 <877e3ewmau.fsf@HIDDEN>
Feedback-ID: a8j8tDUaJ4AYuDVBywMTwsJebN4w8TVXadJLsJb8td3t3dZi9RdXFlPaQvoFKnI9KgXySsPXcRkajVyY0cGTcA==:Ext:ProtonMail
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
X-Spam-Status: No, score=-1.2 required=7.0 tests=ALL_TRUSTED,DKIM_SIGNED,
 DKIM_VALID,DKIM_VALID_AU,DKIM_VALID_EF autolearn=ham
 autolearn_force=no version=3.4.2
X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on mail.protonmail.ch
X-Spam-Score: -0.7 (/)
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.7 (-)

Hey Mathieu,

Did you get a chance to think of something? If not, what do you think about=
 going with this right now and leaving an issue open for a cleaner way to p=
atch in the future?
Thanks again for all your help and insight.

Cheers,
Carl Dong

> On Dec 2, 2019, at 2:38 PM, Mathieu Othacehe <m.othacehe@HIDDEN> wrote=
:
>=20
>=20
>=20
>>> Could we instead patch "clang-from-llvm" procedure to fix sysroot
>>> support?
>> Do you have anything particular in mind? I can=E2=80=99t think of any cl=
ean ways to do this off the top of my head right now unfortunately :-(
>=20
> No sadly, nothing in mind at the moment, but I'll have a look tomorrow!
>=20
> Thanks,
>=20
> Mathieu
>=20
>=20
>=20

Message sent to guix-patches@HIDDEN:

X-Loop: help-debbugs@HIDDEN
Subject: [bug#38376] [PATCH] gnu: clang: Fix search path detection with --sysroot.
Resent-From: Mathieu Othacehe <m.othacehe@HIDDEN>
Original-Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
Resent-CC: guix-patches@HIDDEN
Resent-Date: Sun, 08 Dec 2019 15:32:02 +0000
Resent-Message-ID: <handler.38376.B38376.157581907829437 <at> debbugs.gnu.org>
Resent-Sender: help-debbugs@HIDDEN
X-GNU-PR-Message: followup 38376
X-GNU-PR-Package: guix-patches
X-GNU-PR-Keywords: patch
To: Carl Dong <contact@HIDDEN>
Cc: 38376 <at> debbugs.gnu.org
Received: via spool by 38376-submit <at> debbugs.gnu.org id=B38376.157581907829437
          (code B ref 38376); Sun, 08 Dec 2019 15:32:02 +0000
Received: (at 38376) by debbugs.gnu.org; 8 Dec 2019 15:31:18 +0000
Received: from localhost ([127.0.0.1]:52062 helo=debbugs.gnu.org)
	by debbugs.gnu.org with esmtp (Exim 4.84_2)
	(envelope-from <debbugs-submit-bounces <at> debbugs.gnu.org>)
	id 1idyWk-0007ej-Ci
	for submit <at> debbugs.gnu.org; Sun, 08 Dec 2019 10:31:18 -0500
Received: from mail-wm1-f68.google.com ([209.85.128.68]:35254)
 by debbugs.gnu.org with esmtp (Exim 4.84_2)
 (envelope-from <m.othacehe@HIDDEN>) id 1idyWj-0007eU-GL
 for 38376 <at> debbugs.gnu.org; Sun, 08 Dec 2019 10:31:17 -0500
Received: by mail-wm1-f68.google.com with SMTP id c20so10740908wmb.0
 for <38376 <at> debbugs.gnu.org>; Sun, 08 Dec 2019 07:31:17 -0800 (PST)
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025;
 h=references:user-agent:from:to:cc:subject:in-reply-to:date
 :message-id:mime-version;
 bh=tSffrc1WCP9yzGtkd8cI0KnOjZkwZ9PzsaO7xfgbKlo=;
 b=dlsQMxPVB4jhQtcEk24RgecoqvtijSmv62vyUCouFH9iWWIoSpsNrRxwPNw/EoR6Kt
 UWvby5O3C2IZBqCBTRAwlItSjj3wrbn8TWAdjGWur0pXNHZMNFz66y2U6gAOYq12U25r
 wrY9QdaQxXxizZ5DENNPsfl7r9oHxrQCBPGG0gxsBV02MEYppMRp0/hnKL+ckK4mP2ly
 CHoFkiN0U1xO9QHsjB/c6x4X/dPCVG7oXFvnQw4NBkOEkg2cY/DOt9RFW7oQecNQ//yv
 QLZ90e806/LVI2kN3QuiUdNMVjDIC+haRZJeJHSkLQIMftpsepizZkYtjztcZtvk75YA
 jvvA==
X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
 d=1e100.net; s=20161025;
 h=x-gm-message-state:references:user-agent:from:to:cc:subject
 :in-reply-to:date:message-id:mime-version;
 bh=tSffrc1WCP9yzGtkd8cI0KnOjZkwZ9PzsaO7xfgbKlo=;
 b=EhAn7A52xiL8XH/JjId9p6L18Ehhar975mQ6jFf+MsmDtv1izlROjW6ZC2WPiOyK7p
 qpkQDdexekDFNaL9PxT6aLYyOT+cbg/VthYT3R6iXiB68noS5K34fZPEASN2IJXlyqlK
 JBycGuTW6CX3Yq/6Rxk/Uiok4YOpEnoEMq2Hk8BtM+IBn6OoUTYebze5ch1EixnfjAKc
 Dk4aMvrumyWlXu4Q7pgsuw0iIvb/4d5UtzMqYKh1updNunVDL7wsOFhcI19g8qBSmrMd
 PrBsRdu7u2NNaSdTOFot3/Z607MzX7bUtyJxKjseEmPrtIDzhzC0ERctaILyNZlXIJ4z
 6RSg==
X-Gm-Message-State: APjAAAUDNFtIk8BK+km4JNSHJJsi76/hwzKn0+QLMOqMG8r3KodRdlY+
 a1N/XK285KDe5d0/1ZidHQgmubBW
X-Google-Smtp-Source: APXvYqzHy6t/64YqbWPNC6R01QHiyVC3sSusBE9i1Lw2xA4HR5svq1pBHNizERy5IFl1t1h7yHf+1Q==
X-Received: by 2002:a1c:3941:: with SMTP id g62mr19907603wma.165.1575819071410; 
 Sun, 08 Dec 2019 07:31:11 -0800 (PST)
Received: from cervin ([2a01:e0a:fa:a50:cca0:e766:5a6d:364b])
 by smtp.gmail.com with ESMTPSA id l3sm23627928wrt.29.2019.12.08.07.31.09
 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256);
 Sun, 08 Dec 2019 07:31:10 -0800 (PST)
References: <20191125180144.1060289-1-contact@HIDDEN>
 <8736ebt8fr.fsf@HIDDEN> <379EB7C5-A387-4331-A48A-91C9083F3A73@HIDDEN>
 <8736e2lt8f.fsf@HIDDEN> <A72CBE3E-CB34-4694-BC3F-07517168B693@HIDDEN>
 <877e3ewmau.fsf@HIDDEN> <33E84CBF-D451-4512-A515-1DA6A5E26F20@HIDDEN>
User-agent: mu4e 1.2.0; emacs 26.3
From: Mathieu Othacehe <m.othacehe@HIDDEN>
In-reply-to: <33E84CBF-D451-4512-A515-1DA6A5E26F20@HIDDEN>
Date: Sun, 08 Dec 2019 16:31:08 +0100
Message-ID: <8736duj0lf.fsf@HIDDEN>
MIME-Version: 1.0
Content-Type: text/plain
X-Spam-Score: -0.1 (/)
X-BeenThere: debbugs-submit <at> debbugs.gnu.org
X-Mailman-Version: 2.1.18
Precedence: list
List-Id: <debbugs-submit.debbugs.gnu.org>
List-Unsubscribe: <https://debbugs.gnu.org/cgi-bin/mailman/options/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=unsubscribe>
List-Archive: <https://debbugs.gnu.org/cgi-bin/mailman/private/debbugs-submit/>
List-Post: <mailto:debbugs-submit <at> debbugs.gnu.org>
List-Help: <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=help>
List-Subscribe: <https://debbugs.gnu.org/cgi-bin/mailman/listinfo/debbugs-submit>, 
 <mailto:debbugs-submit-request <at> debbugs.gnu.org?subject=subscribe>
Errors-To: debbugs-submit-bounces <at> debbugs.gnu.org
Sender: "Debbugs-submit" <debbugs-submit-bounces <at> debbugs.gnu.org>
X-Spam-Score: -1.1 (-)


Hello Carl,

> Did you get a chance to think of something? If not, what do you think about going with this right now and leaving an issue open for a cleaner way to patch in the future?
> Thanks again for all your help and insight.

No sadly I didn't get time :(. I would prefer we at least wait to find a way to
factorize the different patches.

Thank you,

Mathieu

Last modified: Sun, 8 Dec 2019 15:45:01 UTC

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997 nCipher Corporation Ltd, 1994-97 Ian Jackson.