[Clang][Driver] Add -mcpu=help to clang

[michaelmaitland]

Summary

Clang currently uses -mcpu=?. The ? causes errors on some shells such as zsh
since it is a special character. In order for it to work on shells such as zsh,
the option must be passed in quotes or escaped. This patch adds -mcpu=help as
another alias for --print-supported-cpus. In llc, -mcpu=help an alias to
print supported cpus.

[reames]

Code and description appear out of sync. (help != list) Personally, I like the help naming a lot better.

[michaelmaitland]

Change list -> help.

[michaelmaitland]

In D144914#4156567, @reames wrote:

Code and description appear out of sync. (help != list) Personally, I like the help naming a lot better.

Thanks for catching this. I meant to write help instead of list.

[reames]

LGTM

[MaskRay]

Sorry I just saw this but I am not sure this is a good idea. Why can't the user use --print-supported-cpus instead? The additional alias doesn't seem useful. If you can make GCC add this as well, it will be different.

[MaskRay]

FWIW I think this should be reverted.

[michaelmaitland]

In D144914#4159291, @MaskRay wrote:

Sorry I just saw this but I am not sure this is a good idea. Why can't the user use --print-supported-cpus instead? The additional alias doesn't seem useful. If you can make GCC add this as well, it will be different.

According to https://reviews.llvm.org/D63105, I believe the reason mcpu=? and mtune=? were added was:

This option is useful but may be hard to discover. Will something like -march=? and -mtune=? make the feature more discoverable?

I would then say this option is being made available for the same reason: discoverability. Since llc uses help instead of ?, I think that this clang option is easier to discover since it exists in other parts of llvm. Additionally, it comes with the added benefit that help does not contain special characters that require escaping. Maybe we revert this patch and remove the ? and only support help?

[MaskRay]

Since llc has -mcpu=help, it seems fine to have clang -mcpu=help. I made a bad suggestion there about -mcpu=?. Perhaps I should revert de94ac9357750fdba45e09eefa8f67a650ae6a64 . This -mcpu=help patch is good.

clang/docs/CommandGuide/clang.rst

.. option:: --print-supported-cpus
   Print out a list of supported processors for the given target (specified
   through ``--target=<architecture>`` or :option:`-arch` ``<architecture>``). If no
   target is specified, the system default target will be used.
 
 .. option:: -mcpu=?, -mtune=?
 
   Acts as an alias for :option:`--print-supported-cpus`.
 
+.. option:: -mcpu=help, -mtune=help
+
+  Acts as an alias for :option:`--print-supported-cpus`.
+
 .. option:: -march=<cpu>
 
   Specify that Clang should generate code for a specific processor family
   member and later.  For example, if you specify -march=i486, the compiler is
   allowed to generate instructions that are valid on i486 and later processors,
   but which may not exist on earlier ones.
 
 

clang/include/clang/Driver/Options.td

def darwin_target_variant : Separate<["-"], "darwin-target-variant">,
   HelpText<"Generate code for an additional runtime variant of the deployment target">;
 def print_supported_cpus : Flag<["-", "--"], "print-supported-cpus">,
   Group<CompileOnly_Group>, Flags<[CC1Option, CoreOption]>,
   HelpText<"Print supported cpu models for the given target (if target is not specified,"
            " it will print the supported cpus for the default target)">,
   MarshallingInfoFlag<FrontendOpts<"PrintSupportedCPUs">>;
 def mcpu_EQ_QUESTION : Flag<["-"], "mcpu=?">, Alias<print_supported_cpus>;
 def mtune_EQ_QUESTION : Flag<["-"], "mtune=?">, Alias<print_supported_cpus>;
+def mcpu_EQ_help : Flag<["-"], "mcpu=help">, Alias<print_supported_cpus>;
+def mtune_EQ_help : Flag<["-"], "mtune=help">, Alias<print_supported_cpus>;
 def time : Flag<["-"], "time">,
   HelpText<"Time individual commands">;
 def traditional_cpp : Flag<["-", "--"], "traditional-cpp">, Flags<[CC1Option]>,
   HelpText<"Enable some traditional CPP emulation">,
   MarshallingInfoFlag<LangOpts<"TraditionalCPP">>;
 def traditional : Flag<["-", "--"], "traditional">;
 def trigraphs : Flag<["-", "--"], "trigraphs">, Alias<ftrigraphs>,
   HelpText<"Process trigraph sequences">;

clang/test/Driver/print-supported-cpus.c

 // Test that --print-supported-cpus lists supported CPU models.
 
 // REQUIRES: x86-registered-target
 // REQUIRES: arm-registered-target
 
 // RUN: %clang --target=x86_64-unknown-linux-gnu --print-supported-cpus 2>&1 | \
 // RUN:   FileCheck %s --check-prefix=CHECK-X86
 
 // Test -mcpu=? and -mtune=? alises.
 // RUN: %clang --target=x86_64-unknown-linux-gnu -mcpu=? 2>&1 | \
 // RUN:   FileCheck %s --check-prefix=CHECK-X86
 
 // RUN: %clang --target=x86_64-unknown-linux-gnu -mtune=? -fuse-ld=dummy 2>&1 | \
 // RUN:   FileCheck %s --check-prefix=CHECK-X86
 
+// Test -mcpu=help and -mtune=help alises.
+// RUN: %clang --target=x86_64-unknown-linux-gnu -mcpu=help 2>&1 | \
+// RUN:   FileCheck %s --check-prefix=CHECK-X86
+
+// RUN: %clang --target=x86_64-unknown-linux-gnu -mtune=help -fuse-ld=dummy 2>&1 | \
+// RUN:   FileCheck %s --check-prefix=CHECK-X86
+
 // CHECK-NOT: warning: argument unused during compilation
 // CHECK-X86: Target: x86_64-unknown-linux-gnu
 // CHECK-X86: corei7
 // CHECK-X86: Use -mcpu or -mtune to specify the target's processor.
 
 // RUN: %clang --target=arm-unknown-linux-android --print-supported-cpus 2>&1 | \
 // RUN:   FileCheck %s --check-prefix=CHECK-ARM
 
 // CHECK-ARM: Target: arm-unknown-linux-android
 // CHECK-ARM: cortex-a73
 // CHECK-ARM: cortex-a75
 // CHECK-ARM: Use -mcpu or -mtune to specify the target's processor.