mirror of
https://github.com/llvm/llvm-project.git
synced 2025-04-29 22:56:06 +00:00
10cef708a7
Occupancy is expressed as waves per SIMD. This means that we need to take into account the number of SIMDs per "CU" or, to be more precise, the number of SIMDs over which a workgroup may be distributed. getOccupancyWithLocalMemSize was wrong because it didn't take SIMDs into account at all. At the same time, we need to take into account that WGP mode offers access to a larger total amount of LDS, since this can affect how non-power-of-two LDS allocations are rounded. To make this work consistently, we distinguish between (available) local memory size and addressable local memory size (which is always limited by 64kB on gfx10+, even with WGP mode). This change results in a massive amount of test churn. A lot of it is caused by the fact that the default work group size is 1024, which means that (due to rounding effects) the default occupancy on older hardware is 8 instead of 10, which affects scheduling via register pressure estimates. I've adjusted most tests by just running the UTC tools, but in some cases I manually changed the work group size to 32 or 64 to make sure that work group size chunkiness has no effect. Differential Revision: https://reviews.llvm.org/D139468
47 lines
1.7 KiB
C++
47 lines
1.7 KiB
C++
//===-- R600Subtarget.cpp - R600 Subtarget Information --------------------===//
|
|
//
|
|
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
|
|
// See https://llvm.org/LICENSE.txt for license information.
|
|
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
//
|
|
/// \file
|
|
/// Implements the R600 specific subclass of TargetSubtarget.
|
|
//
|
|
//===----------------------------------------------------------------------===//
|
|
|
|
#include "R600Subtarget.h"
|
|
#include "MCTargetDesc/R600MCTargetDesc.h"
|
|
|
|
using namespace llvm;
|
|
|
|
#define DEBUG_TYPE "r600-subtarget"
|
|
|
|
#define GET_SUBTARGETINFO_TARGET_DESC
|
|
#define GET_SUBTARGETINFO_CTOR
|
|
#include "R600GenSubtargetInfo.inc"
|
|
|
|
R600Subtarget::R600Subtarget(const Triple &TT, StringRef GPU, StringRef FS,
|
|
const TargetMachine &TM)
|
|
: R600GenSubtargetInfo(TT, GPU, /*TuneCPU*/ GPU, FS), AMDGPUSubtarget(TT),
|
|
InstrInfo(*this),
|
|
FrameLowering(TargetFrameLowering::StackGrowsUp, getStackAlignment(), 0),
|
|
TLInfo(TM, initializeSubtargetDependencies(TT, GPU, FS)),
|
|
InstrItins(getInstrItineraryForCPU(GPU)) {
|
|
AddressableLocalMemorySize = LocalMemorySize;
|
|
}
|
|
|
|
R600Subtarget &R600Subtarget::initializeSubtargetDependencies(const Triple &TT,
|
|
StringRef GPU,
|
|
StringRef FS) {
|
|
SmallString<256> FullFS("+promote-alloca,");
|
|
FullFS += FS;
|
|
ParseSubtargetFeatures(GPU, /*TuneCPU*/ GPU, FullFS);
|
|
|
|
HasMulU24 = getGeneration() >= EVERGREEN;
|
|
HasMulI24 = hasCaymanISA();
|
|
|
|
return *this;
|
|
}
|