[llvm-bugs] [Bug 45926] New: [AMDGPU][MC][GFX8+] Instructions v_sub_u16_e64, v_subrev_u16_e64 and v_add_u16_e64 do not support clamp modifier

[via llvm-bugs]

https://bugs.llvm.org/show_bug.cgi?id=45926

            Bug ID: 45926
           Summary: [AMDGPU][MC][GFX8+] Instructions v_sub_u16_e64,
                    v_subrev_u16_e64 and v_add_u16_e64 do not support
                    clamp modifier
           Product: libraries
           Version: trunk
          Hardware: All
                OS: All
            Status: NEW
          Severity: enhancement
          Priority: P
         Component: Backend: AMDGPU
          Assignee: unassignedbugs at nondot.org
          Reporter: dpreobrazhensky at luxoft.com
                CC: llvm-bugs at lists.llvm.org

According to sp3 documentation, both _e64 and _sdwa variants of these opcodes
should support clamp. However llvm assembler supports clamp for _sdwa variants
only.

These opcodes were renamed to v_*_nc_u16 in gfx10 - they should also be
corrected to support clamp.

Examples of failed tests:

    v_sub_u16_e64 v5, v1, v2 clamp    // gfx8, gfx9
    v_sub_nc_u16 v5, v1, v2 clamp     // gfx10

An example of a passed test:

    v_sub_u16_sdwa v5, v1, v2 clamp   // gfx8, gfx9

-- 
You are receiving this mail because:
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-bugs/attachments/20200514/70937560/attachment.html>