On 9/1/25 4:26 AM, Robin Dapp wrote:
Hi,
In a two-source gather we unconditionally overwrite target with the
first gather already. If op1 == target this clobbers the source operand
for the second gather. This patch uses a temporary in that case.
Regtested on rv64gcv_zvl512b.
Regards
Robin
PR target/121724
gcc/ChangeLog:
* config/riscv/riscv-v.cc (expand_vec_perm): Use temporary if
op1 and target overlap.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/autovec/pr121724.c: New test.
OK. Though I think you've got the wrong PR. 121724 is a C++
rejects-valid bug :-)
---
gcc/config/riscv/riscv-v.cc | 11 +++++--
.../gcc.target/riscv/rvv/autovec/pr121724.c | 30 +++++++++++++++++++
2 files changed, 38 insertions(+), 3 deletions(-)
create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/pr121724.c
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index daa1b37bfe5..eb9afffe2ac 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -3323,15 +3323,17 @@ expand_vec_perm (rtx target, rtx op0, rtx op1,
rtx sel)
mask_mode = get_mask_mode (data_mode);
rtx mask = gen_reg_rtx (mask_mode);
rtx max_sel = gen_const_vector_dup (sel_mode, nunits);
+ bool overlap = target == op1;
+ rtx tmp_target = overlap ? gen_reg_rtx (data_mode) : target;
Consider reg_overlap_mentioned_p. It may not really matter in practice
here, hence the "consider" :-)
jeff
