diff options
| author | Timur Kristóf <timur.kristof@gmail.com> | 2020-01-24 15:17:44 +0100 |
|---|---|---|
| committer | Marge Bot <eric+marge@anholt.net> | 2020-01-24 16:21:08 +0000 |
| commit | c787b8d2a16d5e2950f209b1fcbec6e6c0388845 (patch) | |
| tree | 701457f58ea4d208b2fac3d246add1dfc90fbb7d /src/amd/compiler/aco_insert_NOPs.cpp | |
| parent | 8a32f57fff56b3b94f1b5589feba38016f39427c (diff) | |
| download | mesa-c787b8d2a16d5e2950f209b1fcbec6e6c0388845.tar.gz | |
aco/gfx10: Fix VcmpxExecWARHazard mitigation.
The SOPP instruction shouldn't have a definition, and its block
should be set to -1 in order to prevent it from being recognized
as a branch.
Also fix a typo in the readme.
Fixes: d6dfce02d074d615a3b88a3fccd8ee8c7e13c010
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Tested-by: Marge Bot <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3552>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/merge_requests/3552>
Diffstat (limited to 'src/amd/compiler/aco_insert_NOPs.cpp')
| -rw-r--r-- | src/amd/compiler/aco_insert_NOPs.cpp | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/src/amd/compiler/aco_insert_NOPs.cpp b/src/amd/compiler/aco_insert_NOPs.cpp index b9eaaed96db..689d5e25acc 100644 --- a/src/amd/compiler/aco_insert_NOPs.cpp +++ b/src/amd/compiler/aco_insert_NOPs.cpp @@ -491,9 +491,9 @@ void handle_instruction_gfx10(Program *program, NOP_ctx_gfx10 &ctx, aco_ptr<Inst ctx.has_nonVALU_exec_read = false; /* Insert s_waitcnt_depctr instruction with magic imm to mitigate the problem */ - aco_ptr<SOPP_instruction> depctr{create_instruction<SOPP_instruction>(aco_opcode::s_waitcnt_depctr, Format::SOPP, 0, 1)}; + aco_ptr<SOPP_instruction> depctr{create_instruction<SOPP_instruction>(aco_opcode::s_waitcnt_depctr, Format::SOPP, 0, 0)}; depctr->imm = 0xfffe; - depctr->definitions[0] = Definition(sgpr_null, s1); + depctr->block = -1; new_instructions.emplace_back(std::move(depctr)); } else if (instr_writes_sgpr(instr)) { /* Any VALU instruction that writes an SGPR mitigates the problem */ |