[PATCH] D106191: [clang] Option control afn flag

[Masoud Ataei via Phabricator via llvm-commits]

masoud.ataei marked 4 inline comments as done.
masoud.ataei added a comment.

Sorry that it took me so long to reply reviews. Thank you for reviewing this patch.



================
Comment at: clang/include/clang/Driver/Options.td:1732-1733
   NegFlag<SetFalse>>;
-def fapprox_func : Flag<["-"], "fapprox-func">, Group<f_Group>, Flags<[CC1Option, NoDriverOption]>,
-  MarshallingInfoFlag<LangOpts<"ApproxFunc">>, ImpliedByAnyOf<[menable_unsafe_fp_math.KeyPath]>;
 defm finite_math_only : BoolFOption<"finite-math-only",
----------------
bmahjour wrote:
> So this option already exists and seems to behave the way we want it to. Does anyone know why it was made `NoDriverOption`?
> 
> ```
> > cat fp.c
> #include <math.h>
> void foo(float *f1, float *f2)
> {
>   *f1 = sin(*f2) + *f1;
> }
> > clang -c fp.c -S -emit-llvm -mllvm -disable-llvm-optzns -O3 -Xclang -fapprox-func && grep afn fp.ll
>   %call = call afn double @sin(double %conv) #2
>   %add = fadd afn double %call, %conv1
> ```
> 
> Could we just expose it as a supported option and call it done. ie make it more like `fhonor_nans` below but without introducing a new function attribute:
> 
> ```def fapprox_func : Flag<["-"], "fapprox-func">, Group<f_Group>;```
> 
> so that instead of having `-Xclang -fapprox-func ` in the command above we could just have `-fapprox-func `?
With the change that I proposed for `approx_func` option, we don't need to pass it with `-Xclang`. So, in your example, we can simply remove `-Xclang` and it will work the same:
```
> clang -c fp.c -S -emit-llvm -mllvm -disable-llvm-optzns -O3 -fapprox-func && grep afn fp.ll
  %call = call afn double @sin(double %conv) #2
  %add = fadd afn double %call, %conv1
```
(even no need for ` -mllvm -disable-llvm-optzns -O3`)

Example for `-fapprox-func` and `-fno-approx-func`:
```
$ clang -c fp.c -S -emit-llvm -fapprox-func && grep afn fp.ll
  %call = call afn double @sin(double %conv) #2
  %add = fadd afn double %call, %conv1
$ clang -c fp.c -S -emit-llvm -fno-approx-func && grep afn fp.ll
$
```


================
Comment at: clang/lib/CodeGen/CGCall.cpp:1818
+    if (LangOpts.ApproxFunc)
+      FuncAttrs.addAttribute("approx-func-fp-math", "true");
     if (LangOpts.UnsafeFPMath)
----------------
Regarding the attributes discussion below, here is place that attributes are added. 


================
Comment at: clang/test/CodeGen/afn-flag-test.c:10
+  // CHECK-AFN:      %{{.*}} = call afn double @{{.*}}exp{{.*}}(double %{{.*}})
+  // CHECK-AFN:      attributes #0 ={{.*}} "approx-func-fp-math"="true" {{.*}}
+
----------------
bmahjour wrote:
> can we avoid these attributes?
I understand that attributes are less desirable now in LLVM. But all other options for fast-math flags (`-fhonor-nans`, `-fhonor_infinities`, ...) add the attributes too. So, to be consistent with others I added this attribute too. 

Example: This will add `"no-nans-fp-math"="true"` attribute along with `nnan` flag. 
```
$ clang -c fp.c -S -emit-llvm -fno-honor-nans
```



================
Comment at: clang/test/CodeGen/afn-flag-test.c:13
+  // CHECK-NO-AFN:   %{{.*}} = call double @{{.*}}exp{{.*}}(double %{{.*}})
+  // CHECK-NO-AFN-NOT:  attributes #0 ={{.*}} "approx-func-fp-math"="true" {{.*}}
+}
----------------
bmahjour wrote:
> avoid attributes.
Same as above.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D106191/new/

https://reviews.llvm.org/D106191