[Openmp-commits] [PATCH] D104804: [AMDGPU] Add gfx1035 target

James Henderson via Phabricator via Openmp-commits openmp-commits at lists.llvm.org
Fri Jun 25 00:43:30 PDT 2021


jhenderson added inline comments.


================
Comment at: llvm/test/Object/AMDGPU/elf-header-flags-mach.yaml:189
 
+# RUN: sed -e 's/<BITS>/64/' -e 's/<MACH>/AMDGCN_GFX1035/' %s | yaml2obj -o %t.o.AMDGCN_GFX1035
+# RUN: llvm-readobj -S --file-headers %t.o.AMDGCN_GFX1035 | FileCheck --check-prefixes=ELF-AMDGCN-ALL,ELF-AMDGCN-GFX1035 %s
----------------
MaskRay wrote:
> sed can be replaced by `FileCheck -D` (search examples in `test/tools/`).
> 
> The long list becomes unwieldy now. @jhenderson Suggestions on decreasing the number of RUN lines?
> sed can be replaced by `FileCheck -D` (search examples in `test/tools/`).

Is that true in this context? The `sed` command is controlling input to yaml2obj, not some FileCheck stuff.

> The long list becomes unwieldy now. @jhenderson Suggestions on decreasing the number of RUN lines?

I think you could use additional -D options for FileCheck. Something like the following:

```
# RUN: llvm-readobj ... | FileCheck --check-prefixes=ELF-ALL,ELF-AMDGCN -DNAME=EF_AMDGPU_MACH_AMDGCN_GFX1035 -DVAL=0x3D
# RUN: obj2yaml ... | FileCheck %s --check-prefixes=YAML-ALL,YAML-AMDGCN -DFLAG=EF_AMDGPU_MACH_AMDGCN_GFX1035
## Repeat for all the other values.

## NB: Some of these might be better with the -NEXT suffix.
# ELF-R600: Format: elf32-amdgpu
# ELF-R600: Arch: r600
# ELF-R600: AddressSize: 32bit

# ELF-AMDGCN: Format: elf64-amdgpu
# ELF-AMDGCN: Arch: amdgcn
# ELF-AMDGCN: AddressSize: 64bit

# ELF-ALL:      Flags [
# ELF-ALL-NEXT:   [[NAME]] ([[VAL]])
# ELF-ALL-NEXT: ]

# YAML-R600:   Class: ELFCLASS32
# YAML-AMDGCN: Class: ELFCLASS64
# YAML-ALL:    Flags: [ [[FLAG]] ]
```


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D104804/new/

https://reviews.llvm.org/D104804



More information about the Openmp-commits mailing list