Files
clang-p2996/llvm/test/CodeGen/NVPTX/vector-select.ll
Andrew Savonichev 0f1b5f115a [NVPTX] Integrate ptxas to LIT tests
ptxas is a proprietary compiler from Nvidia that can compile PTX to
machine code (SASS). It has a lot of diagnostics to catch errors
in PTX, which can be used to verify PTX output from llc.

Set -DPXTAS_EXECUTABLE=/path/to/ptxas CMake option to enable it.
If this option is not set, then ptxas is substituted to true which
effectively disables all ptxas RUN lines.

LLVM_PTXAS_EXECUTABLE environment variable takes precedence over
the CMake option, and allows to override ptxas executable that is used for LIT
without complete re-configuration.

Differential Revision: https://reviews.llvm.org/D121727
2022-04-28 14:59:45 +03:00

17 lines
744 B
LLVM

; RUN: llc < %s -march=nvptx -mcpu=sm_20 %if ptxas %{ | %ptxas-verify %}
; RUN: llc < %s -march=nvptx64 -mcpu=sm_20 %if ptxas %{ | %ptxas-verify %}
; This test makes sure that vector selects are scalarized by the type legalizer.
; If not, type legalization will fail.
define void @foo(<2 x i32> addrspace(1)* %def_a, <2 x i32> addrspace(1)* %def_b, <2 x i32> addrspace(1)* %def_c) {
entry:
%tmp4 = load <2 x i32>, <2 x i32> addrspace(1)* %def_a
%tmp6 = load <2 x i32>, <2 x i32> addrspace(1)* %def_c
%tmp8 = load <2 x i32>, <2 x i32> addrspace(1)* %def_b
%0 = icmp sge <2 x i32> %tmp4, zeroinitializer
%cond = select <2 x i1> %0, <2 x i32> %tmp6, <2 x i32> %tmp8
store <2 x i32> %cond, <2 x i32> addrspace(1)* %def_c
ret void
}