mrq/bitsandbytes-rocm
1
0
Fork 1
Code Issues 1 Pull Requests Packages Projects Releases Wiki Activity

Fix merge conflict

Browse Source
This commit is contained in:
0cc4m 2023-02-16 18:57:58 +01:00
commit 403557388d
14 changed files with 506 additions and 1065 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
Makefile
View File

@ -20,6 +20,7 @@ CSRC := $(ROOT_DIR)/csrc
BUILD_DIR:= $(ROOT_DIR)/build
FILES_CUDA := $(CSRC)/ops.cu $(CSRC)/kernels.cu
# FILES_HIP := $(CSRC)/ops.cu $(CSRC)/kernels.cu
FILES_CPP := $(CSRC)/common.cpp $(CSRC)/cpu_ops.cpp $(CSRC)/pythonInterface.c
INCLUDE := -I $(CUDA_HOME)/include -I $(ROOT_DIR)/csrc -I $(CONDA_PREFIX)/include -I $(ROOT_DIR)/include
@ -43,6 +44,7 @@ CC_CUDA10x += -gencode arch=compute_75,code=sm_75
CC_CUDA110 := -gencode arch=compute_75,code=sm_75
CC_CUDA110 += -gencode arch=compute_80,code=sm_80
# CC_CUDA11x := -gencode arch=compute_52,code=sm_52
CC_CUDA11x := -gencode arch=compute_75,code=sm_75
CC_CUDA11x += -gencode arch=compute_80,code=sm_80
CC_CUDA11x += -gencode arch=compute_86,code=sm_86
@ -51,6 +53,7 @@ CC_CUDA11x += -gencode arch=compute_86,code=sm_86
CC_cublasLt110 := -gencode arch=compute_75,code=sm_75
CC_cublasLt110 += -gencode arch=compute_80,code=sm_80
# CC_cublasLt111 := -gencode arch=compute_52,code=sm_52
CC_cublasLt111 := -gencode arch=compute_75,code=sm_75
CC_cublasLt111 += -gencode arch=compute_80,code=sm_80
CC_cublasLt111 += -gencode arch=compute_86,code=sm_86
@ -107,6 +110,17 @@ cuda12x: $(BUILD_DIR) env
cpuonly: $(BUILD_DIR) env
$(GPP) -std=c++14 -shared -fPIC -I $(ROOT_DIR)/csrc -I $(ROOT_DIR)/include $(FILES_CPP) -o ./bitsandbytes/libbitsandbytes_cpu.so
HIP_INCLUDE := -I $(ROOT_DIR)/csrc -I $(ROOT_DIR)/include
# -I /opt/rocm-5.3.0/hipcub/include
HIP_LIB := -L/opt/rocm-5.3.0/lib -L/opt/rocm-5.3.0/llvm/bin/../lib/clang/15.0.0/lib/linux -L/usr/lib/gcc/x86_64-linux-gnu/11 -L/usr/lib/gcc/x86_64-linux-gnu/11/../../../../lib64 -L/lib/x86_64-linux-gnu -L/lib/../lib64 -L/usr/lib/x86_64-linux-gnu -L/usr/lib/../lib64 -L/lib -L/usr/lib -lgcc_s -lgcc -lpthread -lm -lrt -lamdhip64 -lhipblas -lhipsparse -lclang_rt.builtins-x86_64 -lstdc++ -lm -lgcc_s -lgcc -lc -lgcc_s -lgcc
hip: $(BUILD_DIR)
/usr/bin/hipcc -std=c++14 -c -fPIC --amdgpu-target=gfx1030 $(HIP_INCLUDE) -o $(BUILD_DIR)/ops.o -D NO_CUBLASLT $(CSRC)/ops.cu
/usr/bin/hipcc -std=c++14 -c -fPIC --amdgpu-target=gfx1030 $(HIP_INCLUDE) -o $(BUILD_DIR)/kernels.o -D NO_CUBLASLT $(CSRC)/kernels.cu
# /usr/bin/hipcc -fPIC -static $(BUILD_DIR)/ops.o $(BUILD_DIR)/kernels.o -o $(BUILD_DIR)/link.so
$(GPP) -std=c++14 -D__HIP_PLATFORM_AMD__ -DBUILD_CUDA -shared -fPIC -I /opt/rocm/include $(HIP_INCLUDE) $(BUILD_DIR)/ops.o $(BUILD_DIR)/kernels.o $(FILES_CPP) $(HIP_LIB) -o ./bitsandbytes/libbitsandbytes_hip_nocublaslt.so
env:
@echo "ENVIRONMENT"
@echo "============================"
@ -124,9 +138,9 @@ $(BUILD_DIR):
mkdir -p build
mkdir -p dependencies
$(ROOT_DIR)/dependencies/cub:
git clone https://github.com/NVlabs/cub $(ROOT_DIR)/dependencies/cub
cd dependencies/cub; git checkout 1.11.0
# $(ROOT_DIR)/dependencies/cub:
# git clone https://github.com/NVlabs/cub $(ROOT_DIR)/dependencies/cub
# cd dependencies/cub; git checkout 1.11.0
clean:
rm build/*

1
bitsandbytes/__init__.py
View File

@ -3,7 +3,6 @@
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
from . import cuda_setup, utils
from .autograd._functions import (
MatmulLtState,
bmm_cublas,

16
bitsandbytes/__main__.py
View File

@ -28,23 +28,7 @@ print()
from . import COMPILED_WITH_CUDA, PACKAGE_GITHUB_URL
from .cuda_setup.env_vars import to_be_ignored
from .cuda_setup.main import get_compute_capabilities, get_cuda_lib_handle
print_header("POTENTIALLY LIBRARY-PATH-LIKE ENV VARS")
for k, v in os.environ.items():
if "/" in v and not to_be_ignored(k, v):
print(f"'{k}': '{v}'")
print_header("")
print(
"\nWARNING: Please be sure to sanitize sensible info from any such env vars!\n"
)
print_header("OTHER")
print(f"{COMPILED_WITH_CUDA = }")
cuda = get_cuda_lib_handle()
print(f"COMPUTE_CAPABILITIES_PER_GPU = {get_compute_capabilities(cuda)}")
print_header("")
print_header("DEBUG INFO END")
print_header("")

54
bitsandbytes/cextension.py
View File

@ -5,7 +5,59 @@ import torch
from pathlib import Path
from warnings import warn
from bitsandbytes.cuda_setup.main import CUDASetup
class CUDASetup(object):
_instance = None
def __init__(self):
raise RuntimeError("Call get_instance() instead")
def generate_instructions(self):
self.add_log_entry('CUDA SETUP: Something unexpected happened. Please compile from source:')
self.add_log_entry('git clone git@github.com:TimDettmers/bitsandbytes.git')
self.add_log_entry('cd bitsandbytes')
self.add_log_entry("<make_cmd here, commented out>")
self.add_log_entry('python setup.py install')
def initialize(self):
self.has_printed = False
self.lib = None
self.run_cuda_setup()
def run_cuda_setup(self):
self.initialized = True
self.cuda_setup_log = []
binary_name = "libbitsandbytes_hip_nocublaslt.so"
package_dir = Path(__file__).parent
binary_path = package_dir / binary_name
try:
if not binary_path.exists():
raise Exception('CUDA SETUP: Setup Failed!')
else:
self.add_log_entry(f"CUDA SETUP: Loading binary {binary_path}...")
self.lib = ct.cdll.LoadLibrary(binary_path)
except Exception as ex:
self.add_log_entry(str(ex))
self.print_log_stack()
def add_log_entry(self, msg, is_warning=False):
self.cuda_setup_log.append((msg, is_warning))
def print_log_stack(self):
for msg, is_warning in self.cuda_setup_log:
if is_warning:
warn(msg)
else:
print(msg)
@classmethod
def get_instance(cls):
if cls._instance is None:
cls._instance = cls.__new__(cls)
cls._instance.initialize()
return cls._instance
setup = CUDASetup.get_instance()

0
bitsandbytes/cuda_setup/__init__.py
View File

51
bitsandbytes/cuda_setup/env_vars.py
View File

@ -1,51 +0,0 @@
import os
from typing import Dict
def to_be_ignored(env_var: str, value: str) -> bool:
ignorable = {
"PWD", # PWD: this is how the shell keeps track of the current working dir
"OLDPWD",
"SSH_AUTH_SOCK", # SSH stuff, therefore unrelated
"SSH_TTY",
"HOME", # Linux shell default
"TMUX", # Terminal Multiplexer
"XDG_DATA_DIRS", # XDG: Desktop environment stuff
"XDG_RUNTIME_DIR",
"MAIL", # something related to emails
"SHELL", # binary for currently invoked shell
"DBUS_SESSION_BUS_ADDRESS", # hardware related
"PATH", # this is for finding binaries, not libraries
"LESSOPEN", # related to the `less` command
"LESSCLOSE",
"_", # current Python interpreter
}
return env_var in ignorable
def might_contain_a_path(candidate: str) -> bool:
return "/" in candidate
def is_active_conda_env(env_var: str) -> bool:
return "CONDA_PREFIX" == env_var
def is_other_conda_env_var(env_var: str) -> bool:
return "CONDA" in env_var
def is_relevant_candidate_env_var(env_var: str, value: str) -> bool:
return is_active_conda_env(env_var) or (
might_contain_a_path(value) and not
is_other_conda_env_var(env_var) and not
to_be_ignored(env_var, value)
)
def get_potentially_lib_path_containing_env_vars() -> Dict[str, str]:
return {
env_var: value
for env_var, value in os.environ.items()
if is_relevant_candidate_env_var(env_var, value)
}

412
bitsandbytes/cuda_setup/main.py
View File

@ -1,412 +0,0 @@
"""
extract factors the build is dependent on:
[X] compute capability
[ ] TODO: Q - What if we have multiple GPUs of different makes?
- CUDA version
- Software:
- CPU-only: only CPU quantization functions (no optimizer, no matrix multipl)
- CuBLAS-LT: full-build 8-bit optimizer
- no CuBLAS-LT: no 8-bit matrix multiplication (`nomatmul`)
evaluation:
- if paths faulty, return meaningful error
- else:
- determine CUDA version
- determine capabilities
- based on that set the default path
"""
import ctypes as ct
import os
import errno
import torch
from warnings import warn
from pathlib import Path
from typing import Set, Union
from .env_vars import get_potentially_lib_path_containing_env_vars
CUDA_RUNTIME_LIB: str = "libcudart.so"
class CUDASetup:
_instance = None
def __init__(self):
raise RuntimeError("Call get_instance() instead")
def generate_instructions(self):
if self.cuda is None:
self.add_log_entry('CUDA SETUP: Problem: The main issue seems to be that the main CUDA library was not detected.')
self.add_log_entry('CUDA SETUP: Solution 1): Your paths are probably not up-to-date. You can update them via: sudo ldconfig.')
self.add_log_entry('CUDA SETUP: Solution 2): If you do not have sudo rights, you can do the following:')
self.add_log_entry('CUDA SETUP: Solution 2a): Find the cuda library via: find / -name libcuda.so 2>/dev/null')
self.add_log_entry('CUDA SETUP: Solution 2b): Once the library is found add it to the LD_LIBRARY_PATH: export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:FOUND_PATH_FROM_2a')
self.add_log_entry('CUDA SETUP: Solution 2c): For a permanent solution add the export from 2b into your .bashrc file, located at ~/.bashrc')
return
if self.cudart_path is None:
self.add_log_entry('CUDA SETUP: Problem: The main issue seems to be that the main CUDA runtime library was not detected.')
self.add_log_entry('CUDA SETUP: Solution 1: To solve the issue the libcudart.so location needs to be added to the LD_LIBRARY_PATH variable')
self.add_log_entry('CUDA SETUP: Solution 1a): Find the cuda runtime library via: find / -name libcudart.so 2>/dev/null')
self.add_log_entry('CUDA SETUP: Solution 1b): Once the library is found add it to the LD_LIBRARY_PATH: export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:FOUND_PATH_FROM_1a')
self.add_log_entry('CUDA SETUP: Solution 1c): For a permanent solution add the export from 1b into your .bashrc file, located at ~/.bashrc')
self.add_log_entry('CUDA SETUP: Solution 2: If no library was found in step 1a) you need to install CUDA.')
self.add_log_entry('CUDA SETUP: Solution 2a): Download CUDA install script: wget https://github.com/TimDettmers/bitsandbytes/blob/main/cuda_install.sh')
self.add_log_entry('CUDA SETUP: Solution 2b): Install desired CUDA version to desired location. The syntax is bash cuda_install.sh CUDA_VERSION PATH_TO_INSTALL_INTO.')
self.add_log_entry('CUDA SETUP: Solution 2b): For example, "bash cuda_install.sh 113 ~/local/" will download CUDA 11.3 and install into the folder ~/local')
return
make_cmd = f'CUDA_VERSION={self.cuda_version_string}'
if len(self.cuda_version_string) < 3:
make_cmd += ' make cuda92'
elif self.cuda_version_string == '110':
make_cmd += ' make cuda110'
elif self.cuda_version_string[:2] == '11' and int(self.cuda_version_string[2]) > 0:
make_cmd += ' make cuda11x'
elif self.cuda_version_string == '100':
self.add_log_entry('CUDA SETUP: CUDA 10.0 not supported. Please use a different CUDA version.')
self.add_log_entry('CUDA SETUP: Before you try again running bitsandbytes, make sure old CUDA 10.0 versions are uninstalled and removed from $LD_LIBRARY_PATH variables.')
return
has_cublaslt = is_cublasLt_compatible(self.cc)
if not has_cublaslt:
make_cmd += '_nomatmul'
self.add_log_entry('CUDA SETUP: Something unexpected happened. Please compile from source:')
self.add_log_entry('git clone git@github.com:TimDettmers/bitsandbytes.git')
self.add_log_entry('cd bitsandbytes')
self.add_log_entry(make_cmd)
self.add_log_entry('python setup.py install')
def initialize(self):
if not getattr(self, 'initialized', False):
self.has_printed = False
self.lib = None
self.initialized = False
def run_cuda_setup(self):
self.initialized = True
self.cuda_setup_log = []
binary_name, cudart_path, cuda, cc, cuda_version_string = evaluate_cuda_setup()
self.cudart_path = cudart_path
self.cuda = cuda
self.cc = cc
self.cuda_version_string = cuda_version_string
package_dir = Path(__file__).parent.parent
binary_path = package_dir / binary_name
try:
if not binary_path.exists():
self.add_log_entry(f"CUDA SETUP: Required library version not found: {binary_name}. Maybe you need to compile it from source?")
legacy_binary_name = "libbitsandbytes_cpu.so"
self.add_log_entry(f"CUDA SETUP: Defaulting to {legacy_binary_name}...")
binary_path = package_dir / legacy_binary_name
if not binary_path.exists() or torch.cuda.is_available():
self.add_log_entry('')
self.add_log_entry('='*48 + 'ERROR' + '='*37)
self.add_log_entry('CUDA SETUP: CUDA detection failed! Possible reasons:')
self.add_log_entry('1. CUDA driver not installed')
self.add_log_entry('2. CUDA not installed')
self.add_log_entry('3. You have multiple conflicting CUDA libraries')
self.add_log_entry('4. Required library not pre-compiled for this bitsandbytes release!')
self.add_log_entry('CUDA SETUP: If you compiled from source, try again with `make CUDA_VERSION=DETECTED_CUDA_VERSION` for example, `make CUDA_VERSION=113`.')
self.add_log_entry('CUDA SETUP: The CUDA version for the compile might depend on your conda install. Inspect CUDA version via `conda list | grep cuda`.')
self.add_log_entry('='*80)
self.add_log_entry('')
self.generate_instructions()
self.print_log_stack()
raise Exception('CUDA SETUP: Setup Failed!')
self.lib = ct.cdll.LoadLibrary(binary_path)
else:
self.add_log_entry(f"CUDA SETUP: Loading binary {binary_path}...")
self.lib = ct.cdll.LoadLibrary(binary_path)
except Exception as ex:
self.add_log_entry(str(ex))
self.print_log_stack()
def add_log_entry(self, msg, is_warning=False):
self.cuda_setup_log.append((msg, is_warning))
def print_log_stack(self):
for msg, is_warning in self.cuda_setup_log:
if is_warning:
warn(msg)
else:
print(msg)
@classmethod
def get_instance(cls):
if cls._instance is None:
cls._instance = cls.__new__(cls)
cls._instance.initialize()
return cls._instance
def is_cublasLt_compatible(cc):
has_cublaslt = False
if cc is not None:
cc_major, cc_minor = cc.split('.')
if int(cc_major) < 7 or (int(cc_major) == 7 and int(cc_minor) < 5):
cuda_setup.add_log_entry("WARNING: Compute capability < 7.5 detected! Only slow 8-bit matmul is supported for your GPU!", is_warning=True)
else:
has_cublaslt = True
return has_cublaslt
def extract_candidate_paths(paths_list_candidate: str) -> Set[Path]:
return {Path(ld_path) for ld_path in paths_list_candidate.split(":") if ld_path}
def remove_non_existent_dirs(candidate_paths: Set[Path]) -> Set[Path]:
existent_directories: Set[Path] = set()
for path in candidate_paths:
try:
if path.exists():
existent_directories.add(path)
except OSError as exc:
if exc.errno != errno.ENAMETOOLONG:
raise exc
non_existent_directories: Set[Path] = candidate_paths - existent_directories
if non_existent_directories:
CUDASetup.get_instance().add_log_entry("WARNING: The following directories listed in your path were found to "
f"be non-existent: {non_existent_directories}", is_warning=True)
return existent_directories
def get_cuda_runtime_lib_paths(candidate_paths: Set[Path]) -> Set[Path]:
return {
path / CUDA_RUNTIME_LIB
for path in candidate_paths
if (path / CUDA_RUNTIME_LIB).is_file()
}
def resolve_paths_list(paths_list_candidate: str) -> Set[Path]:
"""
Searches a given environmental var for the CUDA runtime library,
i.e. `libcudart.so`.
"""
return remove_non_existent_dirs(extract_candidate_paths(paths_list_candidate))
def find_cuda_lib_in(paths_list_candidate: str) -> Set[Path]:
return get_cuda_runtime_lib_paths(
resolve_paths_list(paths_list_candidate)
)
def warn_in_case_of_duplicates(results_paths: Set[Path]) -> None:
if len(results_paths) > 1:
warning_msg = (
f"Found duplicate {CUDA_RUNTIME_LIB} files: {results_paths}.. "
"We'll flip a coin and try one of these, in order to fail forward.\n"
"Either way, this might cause trouble in the future:\n"
"If you get `CUDA error: invalid device function` errors, the above "
"might be the cause and the solution is to make sure only one "
f"{CUDA_RUNTIME_LIB} in the paths that we search based on your env.")
CUDASetup.get_instance().add_log_entry(warning_msg, is_warning=True)
def determine_cuda_runtime_lib_path() -> Union[Path, None]:
"""
Searches for a cuda installations, in the following order of priority:
1. active conda env
2. LD_LIBRARY_PATH
3. any other env vars, while ignoring those that
- are known to be unrelated (see `bnb.cuda_setup.env_vars.to_be_ignored`)
- don't contain the path separator `/`
If multiple libraries are found in part 3, we optimistically try one,
while giving a warning message.
"""
candidate_env_vars = get_potentially_lib_path_containing_env_vars()
if "CONDA_PREFIX" in candidate_env_vars:
conda_libs_path = Path(candidate_env_vars["CONDA_PREFIX"]) / "lib"
conda_cuda_libs = find_cuda_lib_in(str(conda_libs_path))
warn_in_case_of_duplicates(conda_cuda_libs)
if conda_cuda_libs:
return next(iter(conda_cuda_libs))
CUDASetup.get_instance().add_log_entry(f'{candidate_env_vars["CONDA_PREFIX"]} did not contain '
f'{CUDA_RUNTIME_LIB} as expected! Searching further paths...', is_warning=True)
if "LD_LIBRARY_PATH" in candidate_env_vars:
lib_ld_cuda_libs = find_cuda_lib_in(candidate_env_vars["LD_LIBRARY_PATH"])
if lib_ld_cuda_libs:
return next(iter(lib_ld_cuda_libs))
warn_in_case_of_duplicates(lib_ld_cuda_libs)
CUDASetup.get_instance().add_log_entry(f'{candidate_env_vars["LD_LIBRARY_PATH"]} did not contain '
f'{CUDA_RUNTIME_LIB} as expected! Searching further paths...', is_warning=True)
remaining_candidate_env_vars = {
env_var: value for env_var, value in candidate_env_vars.items()
if env_var not in {"CONDA_PREFIX", "LD_LIBRARY_PATH"}
}
cuda_runtime_libs = set()
for env_var, value in remaining_candidate_env_vars.items():
cuda_runtime_libs.update(find_cuda_lib_in(value))
if len(cuda_runtime_libs) == 0:
CUDASetup.get_instance().add_log_entry('CUDA_SETUP: WARNING! libcudart.so not found in any environmental path. Searching /usr/local/cuda/lib64...')
cuda_runtime_libs.update(find_cuda_lib_in('/usr/local/cuda/lib64'))
warn_in_case_of_duplicates(cuda_runtime_libs)
return next(iter(cuda_runtime_libs)) if cuda_runtime_libs else None
def check_cuda_result(cuda, result_val):
# 3. Check for CUDA errors
if result_val != 0:
error_str = ct.c_char_p()
cuda.cuGetErrorString(result_val, ct.byref(error_str))
if error_str.value is not None:
CUDASetup.get_instance().add_log_entry(f"CUDA exception! Error code: {error_str.value.decode()}")
else:
CUDASetup.get_instance().add_log_entry(f"Unknown CUDA exception! Please check your CUDA install. It might also be that your GPU is too old.")
# https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART____VERSION.html#group__CUDART____VERSION
def get_cuda_version(cuda, cudart_path):
if cuda is None: return None
try:
cudart = ct.CDLL(cudart_path)
except OSError:
CUDASetup.get_instance().add_log_entry(f'ERROR: libcudart.so could not be read from path: {cudart_path}!')
return None
version = ct.c_int()
try:
check_cuda_result(cuda, cudart.cudaRuntimeGetVersion(ct.byref(version)))
except AttributeError as e:
CUDASetup.get_instance().add_log_entry(f'ERROR: {str(e)}')
CUDASetup.get_instance().add_log_entry(f'CUDA SETUP: libcudart.so path is {cudart_path}')
CUDASetup.get_instance().add_log_entry(f'CUDA SETUP: Is seems that your cuda installation is not in your path. See https://github.com/TimDettmers/bitsandbytes/issues/85 for more information.')
version = int(version.value)
major = version//1000
minor = (version-(major*1000))//10
if major < 11:
CUDASetup.get_instance().add_log_entry('CUDA SETUP: CUDA version lower than 11 are currently not supported for LLM.int8(). You will be only to use 8-bit optimizers and quantization routines!!')
return f'{major}{minor}'
def get_cuda_lib_handle():
# 1. find libcuda.so library (GPU driver) (/usr/lib)
try:
cuda = ct.CDLL("libcuda.so")
except OSError:
CUDASetup.get_instance().add_log_entry('CUDA SETUP: WARNING! libcuda.so not found! Do you have a CUDA driver installed? If you are on a cluster, make sure you are on a CUDA machine!')
return None
check_cuda_result(cuda, cuda.cuInit(0))
return cuda
def get_compute_capabilities(cuda):
"""
1. find libcuda.so library (GPU driver) (/usr/lib)
init_device -> init variables -> call function by reference
2. call extern C function to determine CC
(https://docs.nvidia.com/cuda/cuda-driver-api/group__CUDA__DEVICE__DEPRECATED.html)
3. Check for CUDA errors
https://stackoverflow.com/questions/14038589/what-is-the-canonical-way-to-check-for-errors-using-the-cuda-runtime-api
# bits taken from https://gist.github.com/f0k/63a664160d016a491b2cbea15913d549
"""
nGpus = ct.c_int()
cc_major = ct.c_int()
cc_minor = ct.c_int()
device = ct.c_int()
check_cuda_result(cuda, cuda.cuDeviceGetCount(ct.byref(nGpus)))
ccs = []
for i in range(nGpus.value):
check_cuda_result(cuda, cuda.cuDeviceGet(ct.byref(device), i))
ref_major = ct.byref(cc_major)
ref_minor = ct.byref(cc_minor)
# 2. call extern C function to determine CC
check_cuda_result(cuda, cuda.cuDeviceComputeCapability(ref_major, ref_minor, device))
ccs.append(f"{cc_major.value}.{cc_minor.value}")
return ccs
# def get_compute_capability()-> Union[List[str, ...], None]: # FIXME: error
def get_compute_capability(cuda):
"""
Extracts the highest compute capbility from all available GPUs, as compute
capabilities are downwards compatible. If no GPUs are detected, it returns
None.
"""
if cuda is None: return None
# TODO: handle different compute capabilities; for now, take the max
ccs = get_compute_capabilities(cuda)
if ccs: return ccs[-1]
def evaluate_cuda_setup():
if 'BITSANDBYTES_NOWELCOME' not in os.environ or str(os.environ['BITSANDBYTES_NOWELCOME']) == '0':
print('')
print('='*35 + 'BUG REPORT' + '='*35)
print('Welcome to bitsandbytes. For bug reports, please submit your error trace to: https://github.com/TimDettmers/bitsandbytes/issues')
print('='*80)
if not torch.cuda.is_available(): return 'libsbitsandbytes_cpu.so', None, None, None, None
cuda_setup = CUDASetup.get_instance()
cudart_path = determine_cuda_runtime_lib_path()
cuda = get_cuda_lib_handle()
cc = get_compute_capability(cuda)
cuda_version_string = get_cuda_version(cuda, cudart_path)
failure = False
if cudart_path is None:
failure = True
cuda_setup.add_log_entry("WARNING: No libcudart.so found! Install CUDA or the cudatoolkit package (anaconda)!", is_warning=True)
else:
cuda_setup.add_log_entry(f"CUDA SETUP: CUDA runtime path found: {cudart_path}")
if cc == '' or cc is None:
failure = True
cuda_setup.add_log_entry("WARNING: No GPU detected! Check your CUDA paths. Proceeding to load CPU-only library...", is_warning=True)
else:
cuda_setup.add_log_entry(f"CUDA SETUP: Highest compute capability among GPUs detected: {cc}")
if cuda is None:
failure = True
else:
cuda_setup.add_log_entry(f'CUDA SETUP: Detected CUDA version {cuda_version_string}')
# 7.5 is the minimum CC vor cublaslt
has_cublaslt = is_cublasLt_compatible(cc)
# TODO:
# (1) CUDA missing cases (no CUDA installed by CUDA driver (nvidia-smi accessible)
# (2) Multiple CUDA versions installed
# we use ls -l instead of nvcc to determine the cuda version
# since most installations will have the libcudart.so installed, but not the compiler
if failure:
binary_name = "libbitsandbytes_cpu.so"
elif has_cublaslt:
binary_name = f"libbitsandbytes_cuda{cuda_version_string}.so"
else:
"if not has_cublaslt (CC < 7.5), then we have to choose _nocublaslt.so"
binary_name = f"libbitsandbytes_cuda{cuda_version_string}_nocublaslt.so"
return binary_name, cudart_path, cuda, cc, cuda_version_string

2
csrc/cpu_ops.h
View File

@ -7,4 +7,4 @@
void quantize_cpu(float *code, float *A, float *absmax, unsigned char *out, long long blocksize, long long n);
void dequantize_cpu(float *code, unsigned char *A, float *absmax, float *out, long long blocksize, long long n);
#endif
#endif

226
csrc/kernels.cu
View File

@ -3,15 +3,17 @@
// This source code is licensed under the MIT license found in the
// LICENSE file in the root directory of this source tree.
#include <hip/hip_runtime.h>
#include <kernels.cuh>
#include <cub/block/block_radix_sort.cuh>
#include <cub/warp/warp_reduce.cuh>
#include <cub/block/block_load.cuh>
#include <cub/block/block_discontinuity.cuh>
#include <cub/block/block_store.cuh>
#include <cub/block/block_reduce.cuh>
#include <cub/cub.cuh>
#include <math_constants.h>
#include <hipcub/hipcub.hpp>
#include <hipcub/block/block_radix_sort.hpp>
#include <hipcub/warp/warp_reduce.hpp>
#include <hipcub/block/block_load.hpp>
#include <hipcub/block/block_discontinuity.hpp>
#include <hipcub/block/block_store.hpp>
#include <hipcub/block/block_reduce.hpp>
#include <hip/hip_math_constants.h>
#define HLF_MAX 65504
#define TH 1024
@ -19,29 +21,29 @@
#define NUM_BLOCK 4096
// source: https://stackoverflow.com/questions/17399119/how-do-i-use-atomicmax-on-floating-point-values-in-cuda
__device__ float atomicMax(float* address, float val) {
int* address_as_i = reinterpret_cast<int*>(address);
int old = *address_as_i, assumed;
do {
assumed = old;
old = atomicCAS(
reinterpret_cast<int*>(address), assumed,
__float_as_int(fmaxf(val, __int_as_float(assumed))));
} while (assumed != old);
return __int_as_float(old);
}
// __device__ float atomicMax(float* address, float val) {
// int* address_as_i = reinterpret_cast<int*>(address);
// int old = *address_as_i, assumed;
// do {
// assumed = old;
// old = atomicCAS(
// reinterpret_cast<int*>(address), assumed,
// __float_as_int(fmaxf(val, __int_as_float(assumed))));
// } while (assumed != old);
// return __int_as_float(old);
// }
__device__ float atomicMin(float* address, float val) {
int* address_as_i = reinterpret_cast<int*>(address);
int old = *address_as_i, assumed;
do {
assumed = old;
old = atomicCAS(
reinterpret_cast<int*>(address), assumed,
__float_as_int(fminf(val, __int_as_float(assumed))));
} while (assumed != old);
return __int_as_float(old);
}
// __device__ float atomicMin(float* address, float val) {
// int* address_as_i = reinterpret_cast<int*>(address);
// int old = *address_as_i, assumed;
// do {
// assumed = old;
// old = atomicCAS(
// reinterpret_cast<int*>(address), assumed,
// __float_as_int(fminf(val, __int_as_float(assumed))));
// } while (assumed != old);
// return __int_as_float(old);
// }
template <int STOCHASTIC>
__device__ unsigned char dQuantize(float* smem_code, const float rand, float x)
@ -232,9 +234,9 @@ __global__ void kHistogramScatterAdd2D(float* histogram, int *index1, int *index
template<typename T, int BLOCK_SIZE, int NUM_MAX>
__global__ void kCompressMax(T * __restrict__ const A, T* out, unsigned char* out_idx, const int n)
{
typedef cub::WarpReduce<T> WarpReduce;
typedef hipcub::WarpReduce<T> WarpReduce;
__shared__ typename WarpReduce::TempStorage temp_storage;
typedef cub::BlockLoad<T, BLOCK_SIZE/8 , 8, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/8 , 8, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
__shared__ typename LoadT::TempStorage loadt;
const int warp_idx = threadIdx.x/32;
@ -282,8 +284,8 @@ __global__ void kCompressMax(T * __restrict__ const A, T* out, unsigned char* ou
for(int i = 0; i < 8; i++)
{
// 3. do warp reduction + broadcast back
warp_max = WarpReduce(temp_storage).Reduce(max1, cub::Max());
warp_max = cub::ShuffleIndex<32>(warp_max, 0, 0xffffffff);
warp_max = WarpReduce(temp_storage).Reduce(max1, hipcub::Max());
warp_max = hipcub::ShuffleIndex<32>(warp_max, 0, 0xffffffff);
// 4. Up-shift maxed value, write index into shared memory, replace with 2nd largest
if(warp_max == max1)
@ -297,7 +299,9 @@ __global__ void kCompressMax(T * __restrict__ const A, T* out, unsigned char* ou
max2 = -64000.0f;
}
__syncwarp();
// __syncwarp();
__syncthreads();
}
if(threadIdx.x % 32 < 8)
@ -324,8 +328,8 @@ __global__ void kEstimateQuantiles(T *__restrict__ const A, float *code, const f
T vals[NUM_ESTIMATE];
typedef cub::BlockRadixSort<T, THREADS_ESTIMATE, NUM_ESTIMATE, cub::NullType, 4, true, cub::BLOCK_SCAN_RAKING> BlockRadixSort;
typedef cub::BlockLoad<T, THREADS_ESTIMATE, NUM_ESTIMATE, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockRadixSort<T, THREADS_ESTIMATE, NUM_ESTIMATE, hipcub::NullType, 4, true, hipcub::BLOCK_SCAN_RAKING> BlockRadixSort;
typedef hipcub::BlockLoad<T, THREADS_ESTIMATE, NUM_ESTIMATE, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
__shared__ union {
typename LoadFloat::TempStorage loadf;
@ -391,8 +395,8 @@ __global__ void kQuantize(float * code, float * __restrict__ const A, unsigned c
unsigned char qvals[NUM];
//const int lane_id = threadIdx.x % 2;
typedef cub::BlockLoad<float, TH, NUM, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef cub::BlockStore<unsigned char, TH, NUM, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef hipcub::BlockLoad<float, TH, NUM, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockStore<unsigned char, TH, NUM, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
__shared__ typename LoadFloat::TempStorage loadf;
__shared__ typename StoreChar::TempStorage storec;
@ -442,10 +446,10 @@ __global__ void kQuantizeBlockwise(float * code, T * __restrict__ const A, float
float local_abs_max = 0.0f;
int local_rand_idx = 0;
typedef cub::BlockLoad<T, BLOCK_SIZE/NUM_PER_TH, NUM_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockStore<unsigned char, BLOCK_SIZE/NUM_PER_TH, NUM_PER_TH, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef cub::BlockReduce<float, BLOCK_SIZE/NUM_PER_TH> BlockReduce;
typedef cub::BlockLoad<float, BLOCK_SIZE/NUM_PER_TH, NUM_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/NUM_PER_TH, NUM_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockStore<unsigned char, BLOCK_SIZE/NUM_PER_TH, NUM_PER_TH, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/NUM_PER_TH> BlockReduce;
typedef hipcub::BlockLoad<float, BLOCK_SIZE/NUM_PER_TH, NUM_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
__shared__ typename LoadT::TempStorage loadt;
__shared__ typename LoadFloat::TempStorage loadf;
@ -473,7 +477,7 @@ __global__ void kQuantizeBlockwise(float * code, T * __restrict__ const A, float
for(int j = 0; j < NUM_PER_TH; j++)
local_abs_max = fmaxf(local_abs_max, fabsf((float)vals[j]));
local_abs_max = BlockReduce(reduce).Reduce(local_abs_max, cub::Max(), valid_items);
local_abs_max = BlockReduce(reduce).Reduce(local_abs_max, hipcub::Max(), valid_items);
if(threadIdx.x == 0)
smem_absmax_value[0] = local_abs_max;
@ -485,7 +489,9 @@ __global__ void kQuantizeBlockwise(float * code, T * __restrict__ const A, float
else
local_abs_max = smem_absmax_value[0];
__syncwarp();
// __syncwarp();
__syncthreads();
local_abs_max = 1.0f/local_abs_max;
@ -521,8 +527,8 @@ __global__ void kDequantizeBlockwise(float *code, unsigned char * A, float * abs
unsigned char qvals[NUM_PER_TH];
float local_abs_max = -FLT_MAX;
typedef cub::BlockLoad<unsigned char, THREADS, NUM_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef cub::BlockStore<T, THREADS, NUM_PER_TH, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
typedef hipcub::BlockLoad<unsigned char, THREADS, NUM_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef hipcub::BlockStore<T, THREADS, NUM_PER_TH, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
__shared__ typename LoadChar::TempStorage loadchar;
__shared__ typename StoreT::TempStorage storet;
@ -592,9 +598,9 @@ __global__ void kPreconditionOptimizer32bit2State(T* g, T* p,
const float correction1 = 1.0f/(1.0f - powf(beta1, step));
const float correction2 = 1.0f/(1.0f - powf(beta2, step));
typedef cub::BlockLoad<T, BLOCK_SIZE/NUM_VALS, NUM_VALS, cub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef cub::BlockLoad<float, BLOCK_SIZE/NUM_VALS, NUM_VALS, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef cub::BlockReduce<float, BLOCK_SIZE/NUM_VALS> BlockReduce;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/NUM_VALS, NUM_VALS, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef hipcub::BlockLoad<float, BLOCK_SIZE/NUM_VALS, NUM_VALS, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/NUM_VALS> BlockReduce;
__shared__ union {
typename Load::TempStorage load;
@ -643,7 +649,9 @@ __global__ void kPreconditionOptimizer32bit2State(T* g, T* p,
if(threadIdx.x == 0)
atomicAdd(&unorm[0], s1_vals[0]);
__syncwarp();
// __syncwarp();
__syncthreads();
}
}
@ -681,11 +689,11 @@ __global__ void kOptimizer32bit2State(T* g, T* p,
}
else{ update_scale = 1.0f; }
typedef cub::BlockLoad<T, TH, NUM_PER_THREAD, cub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef cub::BlockStore<T, TH, NUM_PER_THREAD, cub::BLOCK_STORE_WARP_TRANSPOSE> Store;
typedef hipcub::BlockLoad<T, TH, NUM_PER_THREAD, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef hipcub::BlockStore<T, TH, NUM_PER_THREAD, hipcub::BLOCK_STORE_WARP_TRANSPOSE> Store;
typedef cub::BlockLoad<float, TH, NUM_PER_THREAD, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef cub::BlockStore<float, TH, NUM_PER_THREAD, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreFloat;
typedef hipcub::BlockLoad<float, TH, NUM_PER_THREAD, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockStore<float, TH, NUM_PER_THREAD, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreFloat;
__shared__ union {
typename Load::TempStorage load;
@ -755,9 +763,9 @@ __global__ void kPreconditionOptimizer32bit1State(T* g, T* p,
float s1_vals[NUM_VALS];
typedef cub::BlockLoad<T, BLOCK_SIZE/NUM_VALS, NUM_VALS, cub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef cub::BlockLoad<float, BLOCK_SIZE/NUM_VALS, NUM_VALS, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef cub::BlockReduce<float, BLOCK_SIZE/NUM_VALS> BlockReduce;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/NUM_VALS, NUM_VALS, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef hipcub::BlockLoad<float, BLOCK_SIZE/NUM_VALS, NUM_VALS, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/NUM_VALS> BlockReduce;
__shared__ union {
typename Load::TempStorage load;
@ -813,7 +821,9 @@ __global__ void kPreconditionOptimizer32bit1State(T* g, T* p,
if(threadIdx.x == 0)
atomicAdd(&unorm[0], s1_vals[0]);
__syncwarp();
// __syncwarp();
__syncthreads();
}
}
@ -843,11 +853,11 @@ __global__ void kOptimizer32bit1State(T *g, T *p,
float s1_vals[NUM_PER_THREAD];
typedef cub::BlockLoad<T, TH, NUM_PER_THREAD, cub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef cub::BlockStore<T, TH, NUM_PER_THREAD, cub::BLOCK_STORE_WARP_TRANSPOSE> Store;
typedef hipcub::BlockLoad<T, TH, NUM_PER_THREAD, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> Load;
typedef hipcub::BlockStore<T, TH, NUM_PER_THREAD, hipcub::BLOCK_STORE_WARP_TRANSPOSE> Store;
typedef cub::BlockLoad<float, TH, NUM_PER_THREAD, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef cub::BlockStore<float, TH, NUM_PER_THREAD, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreFloat;
typedef hipcub::BlockLoad<float, TH, NUM_PER_THREAD, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadFloat;
typedef hipcub::BlockStore<float, TH, NUM_PER_THREAD, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreFloat;
__shared__ union {
typename Load::TempStorage load;
@ -939,9 +949,9 @@ kPreconditionOptimizerStatic8bit2State(T* p, T* __restrict__ const g, unsigned c
unsigned char m_c1[NUM8BIT];
unsigned char r_c2[NUM8BIT];
typedef cub::BlockLoad<T, NUM_THREADS, NUM8BIT, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockLoad<unsigned char, NUM_THREADS, NUM8BIT, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadUInt8;
typedef cub::BlockReduce<float, NUM_THREADS> BlockReduce;
typedef hipcub::BlockLoad<T, NUM_THREADS, NUM8BIT, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<unsigned char, NUM_THREADS, NUM8BIT, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadUInt8;
typedef hipcub::BlockReduce<float, NUM_THREADS> BlockReduce;
__shared__ union {
@ -1008,13 +1018,13 @@ kPreconditionOptimizerStatic8bit2State(T* p, T* __restrict__ const g, unsigned c
}
__syncthreads();
local_max_s1 = BlockReduce(temp_storage.reduce).Reduce(local_max_s1, cub::Max(), valid_items);
local_max_s1 = BlockReduce(temp_storage.reduce).Reduce(local_max_s1, hipcub::Max(), valid_items);
__syncthreads();
local_max_s2 = BlockReduce(temp_storage.reduce).Reduce(local_max_s2, cub::Max(), valid_items);
local_max_s2 = BlockReduce(temp_storage.reduce).Reduce(local_max_s2, hipcub::Max(), valid_items);
if(unorm != NULL)
{
__syncthreads();
local_unorm = BlockReduce(temp_storage.reduce).Reduce(local_unorm, cub::Sum(), valid_items);
local_unorm = BlockReduce(temp_storage.reduce).Reduce(local_unorm, hipcub::Sum(), valid_items);
}
if(threadIdx.x == 0)
@ -1068,11 +1078,11 @@ kOptimizerStatic8bit2State(T* p, T* const g, unsigned char* state1, unsigned cha
unsigned char c2s[NUM_PER_THREAD2];
T p_vals[NUM_PER_THREAD2];
T g_vals[NUM_PER_THREAD2];
typedef cub::BlockLoad<T, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockLoad<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef hipcub::BlockLoad<T, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef cub::BlockStore<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef cub::BlockStore<T, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
typedef hipcub::BlockStore<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef hipcub::BlockStore<T, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
__shared__ float smem_quantiles1[256];
__shared__ float smem_quantiles2[256];
@ -1176,9 +1186,9 @@ kPreconditionOptimizerStatic8bit1State(T* p, T* __restrict__ const g, unsigned c
T g_vals[NUM8BIT];
unsigned char m_c1[NUM8BIT];
typedef cub::BlockLoad<T, NUM_THREADS, NUM8BIT, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockLoad<unsigned char, NUM_THREADS, NUM8BIT, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadUInt8;
typedef cub::BlockReduce<float, NUM_THREADS> BlockReduce;
typedef hipcub::BlockLoad<T, NUM_THREADS, NUM8BIT, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<unsigned char, NUM_THREADS, NUM8BIT, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadUInt8;
typedef hipcub::BlockReduce<float, NUM_THREADS> BlockReduce;
__shared__ union {
@ -1229,12 +1239,12 @@ kPreconditionOptimizerStatic8bit1State(T* p, T* __restrict__ const g, unsigned c
}
__syncthreads();
local_max_s1 = BlockReduce(temp_storage.reduce).Reduce(local_max_s1, cub::Max(), valid_items);
local_max_s1 = BlockReduce(temp_storage.reduce).Reduce(local_max_s1, hipcub::Max(), valid_items);
if(threadIdx.x == 0){ atomicMax(&new_max1[0], local_max_s1); }
if(unorm != NULL)
{
__syncthreads();
local_unorm = BlockReduce(temp_storage.reduce).Reduce(local_unorm, cub::Sum(), valid_items);
local_unorm = BlockReduce(temp_storage.reduce).Reduce(local_unorm, hipcub::Sum(), valid_items);
if(threadIdx.x == 0){ atomicAdd(&unorm[0], local_unorm); }
}
@ -1271,11 +1281,11 @@ kOptimizerStatic8bit1State(T* p, T* const g, unsigned char* state1,
unsigned char c1s[NUM_PER_THREAD2];
T p_vals[NUM_PER_THREAD2];
T g_vals[NUM_PER_THREAD2];
typedef cub::BlockLoad<T, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockLoad<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef hipcub::BlockLoad<T, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef cub::BlockStore<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef cub::BlockStore<T, NUM_THREADS2, NUM_PER_THREAD2, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
typedef hipcub::BlockStore<unsigned char, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef hipcub::BlockStore<T, NUM_THREADS2, NUM_PER_THREAD2, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
__shared__ float smem_quantiles1[256];
@ -1353,8 +1363,8 @@ __global__ void kPercentileClipping(T * __restrict__ g, float *gnorm_vec, int st
const int n_full = (BLOCK_SIZE*(n/BLOCK_SIZE)) + (n % BLOCK_SIZE == 0 ? 0 : BLOCK_SIZE);
int valid_items = 0;
typedef cub::BlockReduce<float, BLOCK_SIZE/NUM_VALS> BlockReduce;
typedef cub::BlockLoad<T, BLOCK_SIZE/NUM_VALS, NUM_VALS, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/NUM_VALS> BlockReduce;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/NUM_VALS, NUM_VALS, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
__shared__ typename BlockReduce::TempStorage reduce;
@ -1426,16 +1436,16 @@ kOptimizerStatic8bit2StateBlockwise(T* p, T* __restrict__ const g, unsigned char
unsigned char c1s[N_PER_TH];
unsigned char c2s[N_PER_TH];
T g_vals[N_PER_TH];
typedef cub::BlockLoad<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockLoad<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef cub::BlockStore<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef cub::BlockStore<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
typedef hipcub::BlockStore<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef hipcub::BlockStore<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
__shared__ float smem_quantiles1[LANES][257];
__shared__ float smem_quantiles2[LANES][257];
typedef cub::BlockReduce<float, BLOCK_SIZE/N_PER_TH> BlockReduce1;
typedef cub::BlockReduce<float, BLOCK_SIZE/N_PER_TH> BlockReduce2;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/N_PER_TH> BlockReduce1;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/N_PER_TH> BlockReduce2;
__shared__ typename BlockReduce1::TempStorage reduce1;
__shared__ typename BlockReduce2::TempStorage reduce2;
__shared__ float smem_exchange1[1];
@ -1504,8 +1514,8 @@ kOptimizerStatic8bit2StateBlockwise(T* p, T* __restrict__ const g, unsigned char
// reduce: 2.51/1.60 -> 2.67/1.69
new_local_abs_max1 = BlockReduce1(reduce1).Reduce(new_local_abs_max1, cub::Max());
new_local_abs_max2 = BlockReduce2(reduce2).Reduce(new_local_abs_max2, cub::Max());
new_local_abs_max1 = BlockReduce1(reduce1).Reduce(new_local_abs_max1, hipcub::Max());
new_local_abs_max2 = BlockReduce2(reduce2).Reduce(new_local_abs_max2, hipcub::Max());
if(threadIdx.x == 0)
{
@ -1599,14 +1609,14 @@ kOptimizerStatic8bit1StateBlockwise(T* p, T* __restrict__ const g, unsigned char
T g_vals[N_PER_TH];
T p_vals[N_PER_TH];
typedef cub::BlockLoad<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef cub::BlockLoad<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef hipcub::BlockLoad<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadT;
typedef hipcub::BlockLoad<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_LOAD_WARP_TRANSPOSE> LoadChar;
typedef cub::BlockStore<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef cub::BlockStore<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, cub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
typedef hipcub::BlockStore<unsigned char, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreChar;
typedef hipcub::BlockStore<T, BLOCK_SIZE/N_PER_TH, N_PER_TH, hipcub::BLOCK_STORE_WARP_TRANSPOSE> StoreT;
__shared__ float smem_quantiles1[LANES][257];
typedef cub::BlockReduce<float, BLOCK_SIZE/N_PER_TH> BlockReduce1;
typedef hipcub::BlockReduce<float, BLOCK_SIZE/N_PER_TH> BlockReduce1;
__shared__ typename BlockReduce1::TempStorage reduce1;
__shared__ float smem_exchange1[1];
@ -1678,7 +1688,7 @@ kOptimizerStatic8bit1StateBlockwise(T* p, T* __restrict__ const g, unsigned char
// reduce: 2.51/1.60 -> 2.67/1.69
new_local_abs_max1 = BlockReduce1(reduce1).Reduce(new_local_abs_max1, cub::Max());
new_local_abs_max1 = BlockReduce1(reduce1).Reduce(new_local_abs_max1, hipcub::Max());
if(threadIdx.x == 0)
smem_exchange1[0] = new_local_abs_max1;
@ -1756,10 +1766,10 @@ template<typename T, int THREADS, int ITEMS_PER_THREAD, int TILE_ROWS, int TILE_
const int base_idx = (base_row*cols) + base_col;
const int items_per_load = ITEMS_PER_THREAD*THREADS;
typedef cub::BlockLoad<T, THREADS, ITEMS_PER_THREAD, cub::BLOCK_LOAD_VECTORIZE> LoadT;
typedef cub::BlockReduce<float, THREADS> BlockRowReduce;
typedef cub::BlockReduce<int, THREADS> BlockRowSum;
typedef cub::BlockExchange<float, THREADS, ITEMS_PER_THREAD> BlockExchange;
typedef hipcub::BlockLoad<T, THREADS, ITEMS_PER_THREAD, hipcub::BLOCK_LOAD_VECTORIZE> LoadT;
typedef hipcub::BlockReduce<float, THREADS> BlockRowReduce;
typedef hipcub::BlockReduce<int, THREADS> BlockRowSum;
typedef hipcub::BlockExchange<float, THREADS, ITEMS_PER_THREAD> BlockExchange;
__shared__ union {
typename BlockExchange::TempStorage exchange;
@ -1837,7 +1847,7 @@ template<typename T, int THREADS, int ITEMS_PER_THREAD, int TILE_ROWS, int TILE_
__syncthreads();
row_absmax = (float)BlockRowReduce(temp_storage.rowreduce).Reduce(local_data_fp32, cub::Max());
row_absmax = (float)BlockRowReduce(temp_storage.rowreduce).Reduce(local_data_fp32, hipcub::Max());
if(SPARSE_DECOMP)
{
__syncthreads();
@ -1945,8 +1955,8 @@ template <int ITEMS_PER_THREAD, int SUBTILE_ROWS, int THREADS>__global__ void kd
float local_rowStats[ITEMS_PER_THREAD];
__shared__ float smem_rowStats[SUBTILE_ROWS];
typedef cub::BlockLoad<int, THREADS, ITEMS_PER_THREAD, cub::BLOCK_LOAD_DIRECT> LoadInt32;
typedef cub::BlockExchange<int, THREADS, ITEMS_PER_THREAD> ExchangeInt32;
typedef hipcub::BlockLoad<int, THREADS, ITEMS_PER_THREAD, hipcub::BLOCK_LOAD_DIRECT> LoadInt32;
typedef hipcub::BlockExchange<int, THREADS, ITEMS_PER_THREAD> ExchangeInt32;
__shared__ typename LoadInt32::TempStorage loadint32;
__shared__ typename ExchangeInt32::TempStorage exchangeint32;
@ -2033,9 +2043,9 @@ template <int THREADS, int ITEMS_PER_THREAD, int TILE_ROWS, int TILE_COLS, int S
const int base_idx = (base_row*cols) + base_col;
const int items_per_load = ITEMS_PER_THREAD*THREADS;
typedef cub::BlockLoad<half, THREADS, ITEMS_PER_THREAD, cub::BLOCK_LOAD_VECTORIZE> LoadHalf;
typedef hipcub::BlockLoad<half, THREADS, ITEMS_PER_THREAD, hipcub::BLOCK_LOAD_VECTORIZE> LoadHalf;
__shared__ typename LoadHalf::TempStorage loadhalf;
typedef cub::BlockStore<char, THREADS, ITEMS_PER_THREAD, cub::BLOCK_STORE_VECTORIZE> StoreInt8;
typedef hipcub::BlockStore<char, THREADS, ITEMS_PER_THREAD, hipcub::BLOCK_STORE_VECTORIZE> StoreInt8;
__shared__ typename StoreInt8::TempStorage storeint8;
__shared__ float smem_row_stats[TILE_ROWS];
@ -2168,7 +2178,7 @@ template <int THREADS, int ITEMS_PER_THREAD, int TILE_ROWS, int TILE_COLS, int T
// so that we can have contiguous stores
__shared__ char smem_data[32*33*ITEMS_PER_THREAD];
char local_data[ITEMS_PER_THREAD];
typedef cub::BlockExchange<char, THREADS, ITEMS_PER_THREAD> BlockExchange;
typedef hipcub::BlockExchange<char, THREADS, ITEMS_PER_THREAD> BlockExchange;
// we load row after row from the base_position
// Load data row by row

4
csrc/kernels.cuh
View File

@ -3,8 +3,10 @@
// This source code is licensed under the MIT license found in the
// LICENSE file in the root directory of this source tree.
#include <hip/hip_runtime.h>
#include <float.h>
#include <ops.cuh>
#include "ops.cuh"
#ifndef kernels
#define kernels

333
csrc/ops.cu
View File

@ -3,14 +3,17 @@
// This source code is licensed under the MIT license found in the
// LICENSE file in the root directory of this source tree.
#include <ops.cuh>
#include <kernels.cuh>
#include <cub/device/device_scan.cuh>
#include <limits>
#include <BinSearch.h>
#include <cassert>
#include <common.h>
#include <hip/hip_runtime.h>
#include "ops.cuh"
#include "kernels.cuh"
// #include <hipcub/device/device_scan.cuh>
#include <limits>
// #include <BinSearch.h>
#include <AAlloc.h>
#include <BinAlgo.h>
#include <cassert>
// #include <common.h>
using namespace BinSearch;
using std::cout;
@ -22,16 +25,16 @@ void histogramScatterAdd2D(float* histogram, int *index1, int *index2, float *sr
int num_blocks = n/threads;
num_blocks = n % threads == 0 ? num_blocks : num_blocks + 1;
kHistogramScatterAdd2D<<<num_blocks, 512>>>(histogram, index1, index2, src, maxidx1, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
template <typename T> void estimateQuantiles(T *A, float *code, float offset, int n)
{
int num_blocks = n/4096;
num_blocks = n % 4096 == 0 ? num_blocks : num_blocks + 1;
CUDA_CHECK_RETURN(cudaMemset(code, 0, 256*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(code, 0, 256*sizeof(float)));
kEstimateQuantiles<T><<<num_blocks, 512>>>(A, code, offset, std::numeric_limits<T>::max(), n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
void quantize(float *code, float *A, unsigned char *out, int n)
@ -39,7 +42,7 @@ void quantize(float *code, float *A, unsigned char *out, int n)
int num_blocks = n/1024;
num_blocks = n % 1024 == 0 ? num_blocks : num_blocks + 1;
kQuantize<<<num_blocks, 1024>>>(code, A, out, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
void dequantize(float *code, unsigned char *A, float *out, int n)
@ -47,7 +50,7 @@ void dequantize(float *code, unsigned char *A, float *out, int n)
int num_blocks = n/1024;
num_blocks = n % 1024 == 0 ? num_blocks : num_blocks + 1;
kDequantize<<<num_blocks, 1024>>>(code, A, out, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
template <typename T, int STOCHASTIC> void quantizeBlockwise(float * code, T *A, float *absmax, unsigned char *out, float *rand, int rand_offset, int blocksize, const int n)
@ -73,7 +76,7 @@ template <typename T, int STOCHASTIC> void quantizeBlockwise(float * code, T *A,
kQuantizeBlockwise<T, 64, 1, 0><<<num_blocks, 64>>>(code, A, absmax, out, rand, rand_offset, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
template<typename T> void dequantizeBlockwise(float *code, unsigned char *A, float *absmax, T *out, int blocksize, const int n)
@ -95,7 +98,7 @@ template<typename T> void dequantizeBlockwise(float *code, unsigned char *A, flo
else if(blocksize == 64)
kDequantizeBlockwise<T, 64, 64, 1><<<num_blocks, 64/1>>>(code, A, absmax, out, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
template<typename T, int OPTIMIZER> void optimizer32bit(T* g, T* p,
@ -110,12 +113,12 @@ template<typename T, int OPTIMIZER> void optimizer32bit(T* g, T* p,
case ADAM:
if(max_unorm > 0.0f)
{
CUDA_CHECK_RETURN(cudaMemset(unorm, 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(unorm, 0, 1*sizeof(float)));
kPreconditionOptimizer32bit2State<T, OPTIMIZER, 4096, 8><<<num_blocks, 512>>>(g, p, state1, state2, unorm, beta1, beta2, eps, weight_decay, step, lr, gnorm_scale, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
kOptimizer32bit2State<T, OPTIMIZER><<<num_blocks, 1024>>>(g, p, state1, state2, unorm, max_unorm, param_norm, beta1, beta2, eps, weight_decay, step, lr, gnorm_scale, skip_zeros, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
break;
case MOMENTUM:
case RMSPROP:
@ -123,13 +126,13 @@ template<typename T, int OPTIMIZER> void optimizer32bit(T* g, T* p,
if(max_unorm > 0.0f)
{
CUDA_CHECK_RETURN(cudaMemset(unorm, 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(unorm, 0, 1*sizeof(float)));
kPreconditionOptimizer32bit1State<T, OPTIMIZER, 4096, 8><<<num_blocks, 512>>>(g, p, state1, unorm, beta1, eps, weight_decay, step, lr, gnorm_scale, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
kOptimizer32bit1State<T, OPTIMIZER><<<num_blocks, 1024>>>(g, p, state1, unorm, max_unorm, param_norm, beta1, eps, weight_decay, step, lr, gnorm_scale, skip_zeros, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
break;
}
}
@ -147,28 +150,28 @@ template<typename T, int OPTIMIZER> void optimizerStatic8bit(T* p, T* g,
int num_blocks = n/4096;
num_blocks = n % 4096 == 0 ? num_blocks : num_blocks + 1;
if(max_unorm > 0.0f){ CUDA_CHECK_RETURN(cudaMemset(unorm, 0, 1*sizeof(float))); }
if(max_unorm > 0.0f){ CUDA_CHECK_RETURN(hipMemset(unorm, 0, 1*sizeof(float))); }
switch(OPTIMIZER)
{
case ADAM:
CUDA_CHECK_RETURN(cudaMemset(new_max1, 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(cudaMemset(new_max2, 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(new_max1, 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(new_max2, 0, 1*sizeof(float)));
kPreconditionOptimizerStatic8bit2State<T, OPTIMIZER><<<num_blocks, 256>>>(p, g, state1, state2, unorm, beta1, beta2, eps, step, quantiles1, quantiles2, max1, max2, new_max1, new_max2, gnorm_scale, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
kOptimizerStatic8bit2State<T, OPTIMIZER><<<num_blocks, 1024>>>(p, g, state1, state2, unorm, max_unorm, param_norm, beta1, beta2, eps, step, lr,
quantiles1, quantiles2, max1, max2, new_max1, new_max2, weight_decay, gnorm_scale, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
break;
case MOMENTUM:
case RMSPROP:
case ADAGRAD:
CUDA_CHECK_RETURN(cudaMemset(new_max1, 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(new_max1, 0, 1*sizeof(float)));
kPreconditionOptimizerStatic8bit1State<T, OPTIMIZER><<<num_blocks, 256>>>(p, g, state1, unorm, beta1, eps, step, quantiles1, max1, new_max1, weight_decay, gnorm_scale, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
kOptimizerStatic8bit1State<T, OPTIMIZER><<<num_blocks, 1024>>>(p, g, state1, unorm, max_unorm, param_norm, beta1, eps, step, lr,
quantiles1, max1, new_max1, weight_decay, gnorm_scale, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
break;
default:
break;
@ -193,7 +196,7 @@ template<typename T, int OPTIMIZER> void optimizerStatic8bitBlockwise(T* p, T* g
num_blocks = n % BLOCKSIZE_2STATE == 0 ? num_blocks : num_blocks + 1;
kOptimizerStatic8bit2StateBlockwise<T, OPTIMIZER, BLOCKSIZE_2STATE, NUM_2STATE><<<num_blocks, BLOCKSIZE_2STATE/NUM_2STATE>>>(p, g, state1, state2, beta1, beta2, eps, step, lr,
quantiles1, quantiles2, absmax1, absmax2, weight_decay, gnorm_scale, skip_zeros, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
break;
case MOMENTUM:
case RMSPROP:
@ -202,7 +205,7 @@ template<typename T, int OPTIMIZER> void optimizerStatic8bitBlockwise(T* p, T* g
num_blocks = n % BLOCKSIZE_1STATE == 0 ? num_blocks : num_blocks + 1;
kOptimizerStatic8bit1StateBlockwise<T, OPTIMIZER, BLOCKSIZE_1STATE, NUM_1STATE><<<num_blocks, BLOCKSIZE_1STATE/NUM_1STATE>>>(p, g, state1, beta1, beta2, eps, step, lr,
quantiles1, absmax1, weight_decay, gnorm_scale, skip_zeros, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
break;
}
}
@ -213,31 +216,39 @@ template<typename T> void percentileClipping(T * g, float *gnorm_vec, int step,
{
int num_blocks = n/2048;
num_blocks = n % 2048 == 0 ? num_blocks : num_blocks + 1;
CUDA_CHECK_RETURN(cudaMemset(&gnorm_vec[step % 100], 0, 1*sizeof(float)));
CUDA_CHECK_RETURN(hipMemset(&gnorm_vec[step % 100], 0, 1*sizeof(float)));
kPercentileClipping<T, 2048, 4><<<num_blocks, 512>>>(g, gnorm_vec, step, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
void gemmex(Context *context, bool transposeA, bool transposeB, int m, int n, int k, void *A, void *B, void *C, int lda, int ldb, int ldc)
{
const int falpha = 1;
const int fbeta = 0;
const void * alpha = &falpha;
const void * beta = &fbeta;
cublasStatus_t status;
cout << "" << endl;
cout << "=============================================" << endl;
cout << "ERROR: Your GPU does not support Int8 Matmul!" << endl;
cout << "=============================================" << endl;
cout << "" << endl;
assert(false);
status = cublasGemmEx(context->m_handle,
transposeA ? CUBLAS_OP_T : CUBLAS_OP_N,
transposeB ? CUBLAS_OP_T : CUBLAS_OP_N,
m, n, k,
alpha, A, CUDA_R_8I, lda, B, CUDA_R_8I, ldb, beta,
C, CUDA_R_32I, ldc,
CUDA_R_32I, CUBLAS_GEMM_DEFAULT_TENSOR_OP);
return ;
// const int falpha = 1;
// const int fbeta = 0;
// const void * alpha = &falpha;
// const void * beta = &fbeta;
// hipblasStatus_t status;
if (status != CUBLAS_STATUS_SUCCESS)
{
std::cout << "CUBLAS ERROR: Status " << status << std::endl;
}
// status = hipblasGemmEx(context->m_handle,
// transposeA ? HIPBLAS_OP_T : HIPBLAS_OP_N,
// transposeB ? HIPBLAS_OP_T : HIPBLAS_OP_N,
// m, n, k,
// alpha, A, HIPBLAS_R_8I, lda, B, HIPBLAS_R_8I, ldb, beta,
// C, HIPBLAS_R_32I, ldc,
// HIPBLAS_R_32I, CUBLAS_GEMM_DEFAULT_TENSOR_OP);
// if (status != HIPBLAS_STATUS_SUCCESS)
// {
// std::cout << "CUBLAS ERROR: Status " << status << std::endl;
// }
}
@ -248,7 +259,7 @@ void strided_gemmex(Context *context, bool transposeA, bool transposeB, int m, i
const int fbeta = 0;
const void * alpha = &falpha;
const void * beta = &fbeta;
cublasStatus_t status;
hipblasStatus_t status;
//cout << transposeA << transposeB << endl;
//printf("%i %i %i\n", m,n,k);
@ -256,15 +267,15 @@ void strided_gemmex(Context *context, bool transposeA, bool transposeB, int m, i
//printf("%i %i %i\n", strideA, strideB, strideC);
//printf("%i\n", batchCount);
status = cublasGemmStridedBatchedEx(context->m_handle,
transposeA ? CUBLAS_OP_T : CUBLAS_OP_N,
transposeB ? CUBLAS_OP_T : CUBLAS_OP_N,
status = hipblasGemmStridedBatchedEx(context->m_handle,
transposeA ? HIPBLAS_OP_T : HIPBLAS_OP_N,
transposeB ? HIPBLAS_OP_T : HIPBLAS_OP_N,
m, n, k,
alpha, A, CUDA_R_8I, lda, (long long int)strideA, B, CUDA_R_8I, ldb, (long long int)strideB, beta,
C, CUDA_R_32I, ldc, (long long int)strideC, batchCount,
CUDA_R_32I, CUBLAS_GEMM_DEFAULT);
alpha, A, HIPBLAS_R_8I, lda, (long long int)strideA, B, HIPBLAS_R_8I, ldb, (long long int)strideB, beta,
C, HIPBLAS_R_32I, ldc, (long long int)strideC, batchCount,
HIPBLAS_R_32I, HIPBLAS_GEMM_DEFAULT);
if (status != CUBLAS_STATUS_SUCCESS)
if (status != HIPBLAS_STATUS_SUCCESS)
{
std::cout << "CUBLAS ERROR: Status " << status << std::endl;
}
@ -276,42 +287,6 @@ int roundoff(int v, int d) {
}
#ifdef NO_CUBLASLT
#else
template<int ORDER> cublasLtOrder_t get_order()
{
switch(ORDER)
{
case ROW:
return CUBLASLT_ORDER_ROW;
break;
case COL:
return CUBLASLT_ORDER_COL;
break;
case COL32:
return CUBLASLT_ORDER_COL32;
break;
case COL_TURING:
return CUBLASLT_ORDER_COL4_4R2_8C;
break;
case COL_AMPERE:
return CUBLASLT_ORDER_COL32_2R_4R4;
break;
default:
break;
}
return CUBLASLT_ORDER_ROW;
}
template cublasLtOrder_t get_order<ROW>();
template cublasLtOrder_t get_order<COL>();
template cublasLtOrder_t get_order<COL32>();
template cublasLtOrder_t get_order<COL_TURING>();
template cublasLtOrder_t get_order<COL_AMPERE>();
#endif
template<int ORDER> int get_leading_dim(int dim1, int dim2)
{
switch(ORDER)
@ -345,47 +320,12 @@ template int get_leading_dim<COL32>(int dim1, int dim2);
template <typename T, int SRC, int TARGET, bool transpose, int DTYPE> void transform(cublasLtHandle_t ltHandle, T *A, T *out, int dim1, int dim2)
{
#ifdef NO_CUBLASLT
#else
cublasLtOrder_t orderA = get_order<SRC>();
cublasLtOrder_t orderOut = get_order<TARGET>();
int ldA = get_leading_dim<SRC>(dim1, dim2);
int ldOut = get_leading_dim<TARGET>(dim1, dim2);
cublasLtMatrixLayout_t A_desc = NULL, out_desc = NULL;
cublasLtMatrixTransformDesc_t A2Out_desc = NULL;
cublasOperation_t opTranspose = CUBLAS_OP_T;
float transformAlpha = 1.0f, transformBeta = 0.0f;
if(DTYPE == 8)
{
checkCublasStatus(cublasLtMatrixLayoutCreate(&A_desc, CUDA_R_8I, dim1, dim2, ldA));
checkCublasStatus(cublasLtMatrixLayoutCreate(&out_desc, CUDA_R_8I, dim1, dim2, ldOut));
}
else if(DTYPE == 32)
{
checkCublasStatus(cublasLtMatrixLayoutCreate(&A_desc, CUDA_R_32I, dim1, dim2, ldA));
checkCublasStatus(cublasLtMatrixLayoutCreate(&out_desc, CUDA_R_32I, dim1, dim2, ldOut));
}
else
{
printf("ERROR WRONG TYPE FOR TRANSFORM: %i\n", DTYPE);
}
checkCublasStatus(cublasLtMatrixLayoutSetAttribute(A_desc, CUBLASLT_MATRIX_LAYOUT_ORDER, &orderA, sizeof(orderA)));
checkCublasStatus(cublasLtMatrixLayoutSetAttribute(out_desc, CUBLASLT_MATRIX_LAYOUT_ORDER, &orderOut, sizeof(orderOut)));
checkCublasStatus(cublasLtMatrixTransformDescCreate(&A2Out_desc, CUDA_R_32F));
if(transpose){ checkCublasStatus(cublasLtMatrixTransformDescSetAttribute(A2Out_desc, CUBLASLT_MATRIX_TRANSFORM_DESC_TRANSA, &opTranspose, sizeof(opTranspose))); }
checkCublasStatus(cublasLtMatrixTransform(ltHandle, A2Out_desc, &transformAlpha, A, A_desc, &transformBeta, NULL, NULL, out, out_desc, 0));
if (A_desc) checkCublasStatus(cublasLtMatrixLayoutDestroy(A_desc));
if (out_desc) checkCublasStatus(cublasLtMatrixLayoutDestroy(out_desc));
if (A2Out_desc) checkCublasStatus(cublasLtMatrixTransformDescDestroy(A2Out_desc));
#endif
cout << "" << endl;
cout << "=============================================" << endl;
cout << "ERROR: Your GPU does not support Int8 Matmul!" << endl;
cout << "=============================================" << endl;
cout << "" << endl;
assert(false);
}
template void transform<int8_t, ROW, COL, false, 8>(cublasLtHandle_t ltHandle, int8_t *A, int8_t *out, int dim1, int dim2);
@ -399,7 +339,6 @@ template void transform<int32_t, COL32, ROW, false, 32>(cublasLtHandle_t ltHandl
template <int FORMATB, int DTYPE_OUT, int SCALE_ROWS> int igemmlt(cublasLtHandle_t ltHandle, int m, int n, int k, const int8_t *A, const int8_t *B, void *C, float *row_scale, int lda, int ldb, int ldc)
{
#ifdef NO_CUBLASLT
cout << "" << endl;
cout << "=============================================" << endl;
cout << "ERROR: Your GPU does not support Int8 Matmul!" << endl;
@ -408,62 +347,6 @@ template <int FORMATB, int DTYPE_OUT, int SCALE_ROWS> int igemmlt(cublasLtHandle
assert(false);
return 0;
#else
int has_error = 0;
cublasLtMatmulDesc_t matmulDesc = NULL;
cublasLtMatrixLayout_t Adesc = NULL, Bdesc = NULL, Cdesc = NULL;
cublasOperation_t opT = CUBLAS_OP_T;
cublasLtPointerMode_t alphaVec = CUBLASLT_POINTER_MODE_ALPHA_DEVICE_VECTOR_BETA_ZERO;
cublasLtOrder_t col32 = CUBLASLT_ORDER_COL32;
cublasLtOrder_t col_turing = CUBLASLT_ORDER_COL4_4R2_8C;
cublasLtOrder_t col_ampere = CUBLASLT_ORDER_COL32_2R_4R4;
has_error |= checkCublasStatus(cublasLtMatrixLayoutCreate(&Adesc, CUDA_R_8I, m, k, lda));
has_error |= checkCublasStatus(cublasLtMatrixLayoutCreate(&Bdesc, CUDA_R_8I, n, k, ldb));
has_error |= checkCublasStatus(cublasLtMatrixLayoutSetAttribute(Adesc, CUBLASLT_MATRIX_LAYOUT_ORDER, &col32, sizeof(col32)));
if(FORMATB == COL_TURING)
has_error |= checkCublasStatus(cublasLtMatrixLayoutSetAttribute(Bdesc, CUBLASLT_MATRIX_LAYOUT_ORDER, &col_turing, sizeof(col_turing)));
else
has_error |= checkCublasStatus(cublasLtMatrixLayoutSetAttribute(Bdesc, CUBLASLT_MATRIX_LAYOUT_ORDER, &col_ampere, sizeof(col_ampere)));
if(DTYPE_OUT == 32)
{
has_error |= checkCublasStatus(cublasLtMatmulDescCreate(&matmulDesc, CUBLAS_COMPUTE_32I, CUDA_R_32I));
has_error |= checkCublasStatus(cublasLtMatmulDescSetAttribute(matmulDesc, CUBLASLT_MATMUL_DESC_TRANSB, &opT, sizeof(opT)));
has_error |= checkCublasStatus(cublasLtMatrixLayoutCreate(&Cdesc, CUDA_R_32I, m, n, ldc));
has_error |= checkCublasStatus(cublasLtMatrixLayoutSetAttribute(Cdesc, CUBLASLT_MATRIX_LAYOUT_ORDER, &col32, sizeof(col32)));
int alpha = 1, beta = 0;
has_error |= checkCublasStatus(cublasLtMatmul(ltHandle, matmulDesc,&alpha, A, Adesc, B, Bdesc, &beta, (int32_t*)C, Cdesc, (int32_t*)C, Cdesc, NULL, NULL, 0, 0));
}
else
{
has_error |= checkCublasStatus(cublasLtMatmulDescCreate(&matmulDesc, CUBLAS_COMPUTE_32I, CUDA_R_32F));
has_error |= checkCublasStatus(cublasLtMatmulDescSetAttribute(matmulDesc, CUBLASLT_MATMUL_DESC_TRANSB, &opT, sizeof(opT)));
has_error |= checkCublasStatus(cublasLtMatrixLayoutCreate(&Cdesc, CUDA_R_8I, m, n, ldc));
has_error |= checkCublasStatus(cublasLtMatrixLayoutSetAttribute(Cdesc, CUBLASLT_MATRIX_LAYOUT_ORDER, &col32, sizeof(col32)));
if(!SCALE_ROWS)
{
float alpha = 1.0f, beta = 0.0f;
has_error |= checkCublasStatus(cublasLtMatmul(ltHandle, matmulDesc,&alpha, A, Adesc, B, Bdesc, &beta, (int8_t*)C, Cdesc, (int8_t*)C, Cdesc, NULL, NULL, 0, 0));
}
else
{
has_error |= checkCublasStatus(cublasLtMatmulDescSetAttribute(matmulDesc, CUBLASLT_MATMUL_DESC_POINTER_MODE, &alphaVec, sizeof(alphaVec)));
has_error |= checkCublasStatus(cublasLtMatmul(ltHandle, matmulDesc, row_scale, A, Adesc, B, Bdesc, NULL, (int8_t*)C, Cdesc, (int8_t*)C, Cdesc, NULL, NULL, 0, 0));
}
}
if (Cdesc) has_error |= checkCublasStatus(cublasLtMatrixLayoutDestroy(Cdesc));
if (Bdesc) has_error |= checkCublasStatus(cublasLtMatrixLayoutDestroy(Bdesc));
if (Adesc) has_error |= checkCublasStatus(cublasLtMatrixLayoutDestroy(Adesc));
if (matmulDesc) has_error |= checkCublasStatus(cublasLtMatmulDescDestroy(matmulDesc));
if(has_error == 1)
printf("error detected");
return has_error;
#endif
}
int fill_up_to_nearest_multiple(int value, int multiple)
@ -484,7 +367,7 @@ void dequant_mm_int32_fp16(int *A, float *rowStats, float *colStats, half *out,
assert(threads <= tilesize);
kdequant_mm_int32_fp16<4, 128, 512><<<num_blocks, threads>>>(A, rowStats, colStats, out, newRowStats, newcolStats, bias, numRows, numCols, tileCols, n);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
#define STATS_THREADS 64
@ -505,7 +388,7 @@ void getColRowStats(half * A, float *rowStats, float *colStats, int *nnz_count_r
kgetColRowStats<half, STATS_THREADS, STATS_ITEMS, STATS_ROWS, STATS_THREADS*STATS_ITEMS, 0><<<num_blocks, STATS_THREADS>>>(A, rowStats, colStats, nnz_count_row, nnz_threshold, rows, cols, tiledRows, tiledCols);
else if(nnz_threshold != 0.0)
kgetColRowStats<half, STATS_THREADS, STATS_ITEMS, STATS_ROWS, STATS_THREADS*STATS_ITEMS, 1><<<num_blocks, STATS_THREADS>>>(A, rowStats, colStats, nnz_count_row, nnz_threshold, rows, cols, tiledRows, tiledCols);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
@ -529,7 +412,7 @@ void doubleRowColQuant(half * A, float *rowStats, float *colStats, char *out_col
else
kDoubleRowColQuant<64, 4, 16, 64*4, 0><<<num_blocks, threads>>>(A, rowStats, colStats, out_col_normed, out_row_normed, rowidx, colidx, val, nnz_block_ptr, threshold, rows, cols, tiledCols);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
template <int FORMAT, int TRANSPOSE> void transformRowToFormat(char * A, char *out, int rows, int cols)
@ -573,69 +456,27 @@ template <int FORMAT, int TRANSPOSE> void transformRowToFormat(char * A, char *o
}
kTransformRowToFormat<256, 8, 32, 32*8, TRANSPOSE, FORMAT><<<num_blocks, threads>>>(A, out, rows, cols, tiledCols, outRows, outCols);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
void spmm_coo(cusparseHandle_t handle, int *A_rowidx, int *A_colidx, half *A_vals, int A_nnz, int A_rows, int A_cols, int B_cols, int ldb, half *B, int ldc, half* C, bool transposed_B)
void spmm_coo(hipsparseHandle_t handle, int *A_rowidx, int *A_colidx, half *A_vals, int A_nnz, int A_rows, int A_cols, int B_cols, int ldb, half *B, int ldc, half* C, bool transposed_B)
{
#ifdef NO_CUBLASLT
#else
cout << "" << endl;
cout << "=============================================" << endl;
cout << "ERROR: Your GPU does not support Int8 Matmul!" << endl;
cout << "=============================================" << endl;
cout << "" << endl;
assert(false);
cusparseSpMatDescr_t descA;
cusparseDnMatDescr_t descB, descC;
float alpha = 1.0f;
float beta = 0.0f;
void *dBuffer = NULL;
size_t bufferSize = 0;
CHECK_CUSPARSE( cusparseCreateCoo(&descA, A_rows, A_cols, A_nnz,
A_rowidx, A_colidx, A_vals,
CUSPARSE_INDEX_32I,
CUSPARSE_INDEX_BASE_ZERO, CUDA_R_16F) );
// Create dense matrix C
CHECK_CUSPARSE( cusparseCreateDnMat(&descC, A_rows, B_cols, ldc, C,
CUDA_R_16F, CUSPARSE_ORDER_ROW) );
// Create dense matrix B
if(transposed_B)
{
int tmp = A_cols;
A_cols = B_cols;
B_cols = tmp;
}
CHECK_CUSPARSE( cusparseCreateDnMat(&descB, A_cols, B_cols, ldb, B,
CUDA_R_16F, CUSPARSE_ORDER_ROW) );
// allocate an external buffer if needed
CHECK_CUSPARSE( cusparseSpMM_bufferSize(
handle,
CUSPARSE_OPERATION_NON_TRANSPOSE,
transposed_B ? CUSPARSE_OPERATION_TRANSPOSE : CUSPARSE_OPERATION_NON_TRANSPOSE,
&alpha, descA, descB, &beta, descC, CUDA_R_32F,
CUSPARSE_SPMM_ALG_DEFAULT, &bufferSize) );
CUDA_CHECK_RETURN( cudaMalloc(&dBuffer, bufferSize) );
// execute SpMM
CHECK_CUSPARSE( cusparseSpMM(handle,
CUSPARSE_OPERATION_NON_TRANSPOSE,
transposed_B ? CUSPARSE_OPERATION_TRANSPOSE : CUSPARSE_OPERATION_NON_TRANSPOSE,
&alpha, descA, descB, &beta, descC, CUDA_R_32F,
CUSPARSE_SPMM_ALG_DEFAULT, dBuffer));
// destroy matrix/vector descriptors
CHECK_CUSPARSE( cusparseDestroySpMat(descA) );
CHECK_CUSPARSE( cusparseDestroyDnMat(descB) );
CHECK_CUSPARSE( cusparseDestroyDnMat(descC) );
CUDA_CHECK_RETURN( cudaFree(dBuffer) );
#endif
return;
}
template <typename T, int BITS> void spmm_coo_very_sparse_naive(int *max_count, int *max_idx, int *offset_rowidx, int *rowidx, int *colidx, half *values, T *B, half *out, float *dequant_stats, int nnz_rows, int nnz, int rowsA, int rowsB, int colsB)
{
kspmm_coo_very_sparse_naive<T, 8, BITS><<<nnz_rows, 256>>>(max_count, max_idx, offset_rowidx, rowidx, colidx, values, B, out, dequant_stats, nnz, rowsA, rowsB, colsB);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
@ -658,7 +499,7 @@ template <int FORMAT> void extractOutliers(char * A, int *idx, char *out, int id
}
kExtractOutliers<FORMAT><<<num_blocks, threads>>>(A, idx, out, idx_size, rows, cols, tiledRows, tiledCols);
CUDA_CHECK_RETURN(cudaPeekAtLastError());
CUDA_CHECK_RETURN(hipPeekAtLastError());
}
//==============================================================

73
csrc/ops.cuh
View File

@ -11,30 +11,31 @@
#include <iostream>
#include <unistd.h>
#include <assert.h>
#include <cuda_runtime_api.h>
#include <cuda_fp16.h>
#include <cublas_v2.h>
#include <cublasLt.h>
#include <cusparse.h>
#include <hip/hip_runtime_api.h>
#include <hip/hip_fp16.h>
#include <hipblas.h>
// #include <cublasLt.h>
#include <hipsparse.h>
#include <vector>
#include <functional>
typedef struct cublasLtContext* cublasLtHandle_t;
#define CUDA_CHECK_RETURN(value) { \
cudaError_t _m_cudaStat = value; \
if (_m_cudaStat != cudaSuccess) { \
hipError_t _m_cudaStat = value; \
if (_m_cudaStat != hipSuccess) { \
fprintf(stderr, "Error %s at line %d in file %s\n", \
cudaGetErrorString(_m_cudaStat), __LINE__, __FILE__); \
hipGetErrorString(_m_cudaStat), __LINE__, __FILE__); \
exit(1); \
} }
#define THREADS_PER_BLOCKS (512)
#define CHECK_CUSPARSE(value) { \
cusparseStatus_t _m_cudaStat = value; \
if (_m_cudaStat != CUSPARSE_STATUS_SUCCESS) { \
fprintf(stderr, "Error %s at line %d in file %s\n", \
cusparseGetErrorString(_m_cudaStat), __LINE__, __FILE__); \
hipsparseStatus_t _m_cudaStat = value; \
if (_m_cudaStat != HIPSPARSE_STATUS_SUCCESS) { \
fprintf(stderr, "Error <sparse error> at line %d in file %s\n", \
__LINE__, __FILE__); \
exit(1); \
} }
@ -42,15 +43,15 @@
#define THREADS_PER_BLOCKS (512)
inline void checkCudaStatus(cudaError_t status) {
if (status != cudaSuccess) {
printf("cuda API failed with status %d: %s\n", status, cudaGetErrorString(status));
inline void checkCudaStatus(hipError_t status) {
if (status != hipSuccess) {
printf("cuda API failed with status %d: %s\n", status, hipGetErrorString(status));
throw std::logic_error("cuda API failed");
}
}
inline int checkCublasStatus(cublasStatus_t status) {
if (status != CUBLAS_STATUS_SUCCESS) {
inline int checkCublasStatus(hipblasStatus_t status) {
if (status != HIPBLAS_STATUS_SUCCESS) {
printf("cuBLAS API failed with status %d\n", status);
//throw std::logic_error("cuBLAS API failed");
return 1;
@ -84,40 +85,40 @@ typedef enum Transform_t
class Context
{
public:
cublasHandle_t m_handle;
hipblasHandle_t m_handle;
Context()
{
cublasHandle_t handle;
cublasCreate_v2(&handle);
hipblasHandle_t handle;
hipblasCreate(&handle);
m_handle = handle;
}
};
class ContextLt
{
public:
cublasLtHandle_t m_handle;
// class ContextLt
// {
// public:
// cublasLtHandle_t m_handle;
ContextLt()
{
cublasLtHandle_t handle;
cublasLtCreate(&handle);
m_handle = handle;
}
// ContextLt()
// {
// cublasLtHandle_t handle;
// cublasLtCreate(&handle);
// m_handle = handle;
// }
};
// };
class ContextCusparse
{
public:
cusparseHandle_t m_handle;
hipsparseHandle_t m_handle;
ContextCusparse()
{
cusparseHandle_t handle;
cusparseCreate(&handle);
hipsparseHandle_t handle;
hipsparseCreate(&handle);
m_handle = handle;
}
@ -170,7 +171,7 @@ void doubleRowColQuant(half * A, float *rowStats, float *colStats, char *out_col
template <int FORMAT, int TRANSPOSE> void transformRowToFormat(char * A, char *out, int rows, int cols);
void spmm_coo(cusparseHandle_t handle, int *A_rowidx, int *A_colidx, half *A_vals, int A_nnz, int A_rows, int A_cols, int B_cols, int ldb, half *B, int ldc, half* C, bool transposed_B);
void spmm_coo(hipsparseHandle_t handle, int *A_rowidx, int *A_colidx, half *A_vals, int A_nnz, int A_rows, int A_cols, int B_cols, int ldb, half *B, int ldc, half* C, bool transposed_B);
template <typename T, int BITS> void spmm_coo_very_sparse_naive(int *max_count, int *max_idx, int *offset_rowidx, int *rowidx, int *colidx, half *values, T *B, half *out, float *dequant_stats, int nnz_rows, int nnz, int rowsA, int rowsB, int colsB);

2
csrc/pythonInterface.c
View File

@ -275,7 +275,7 @@ extern "C"
{ transform_row2ampereT(A, out, rows, cols); }
void cspmm_coo(ContextCusparse *context, int *A_rowidx, int *A_colidx, half *A_vals, int A_nnz, int A_rows, int A_cols, int B_cols, int ldb, half *B, int ldc, half* C, bool transposed_B)
{ spmm_coo((cusparseHandle_t) context->m_handle, A_rowidx, A_colidx, A_vals, A_nnz, A_rows, A_cols, B_cols, ldb, B, ldc, C, transposed_B); }
{ spmm_coo((hipsparseHandle_t) context->m_handle, A_rowidx, A_colidx, A_vals, A_nnz, A_rows, A_cols, B_cols, ldb, B, ldc, C, transposed_B); }
void cspmm_coo_very_sparse_naive_fp16(int *max_count, int *max_idx, int *offset_rowidx, int *rowidx, int *colidx, half *values, half *B, half *out, float *dequant_stats, int nnz_rows, int nnz, int rowsA, int rowsB, int colsB)
{ spmm_coo_very_sparse_naive_fp16(max_count, max_idx, offset_rowidx, rowidx, colidx, values, B, out, dequant_stats, nnz_rows, nnz, rowsA, rowsB, colsB); }

377
include/Algo-Direct2.h
View File

@ -2,73 +2,75 @@
#include "Algo-Direct-Common.h"
namespace BinSearch {
namespace Details {
template <typename T, Algos A>
struct AlgoScalarBase<T, A, typename std::enable_if<DirectAux::IsDirect2<A>::value>::type> : DirectAux::DirectInfo<2, T, A>
namespace BinSearch
{
private:
typedef DirectAux::DirectInfo<2, T, A> base_t;
static const size_t Offset=2;
public:
AlgoScalarBase(const T* x, const uint32 n)
: base_t(x, n)
namespace Details
{
}
FORCE_INLINE uint32 scalar(T z) const
{
const T* px = base_t::data.xi;
const uint32* buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
uint32 bidx = base_t::fun_t::f(base_t::data.scaler, base_t::data.cst0, z);
uint32 iidx = buckets[bidx];
px += iidx;
if (z < *px)
--iidx;
if (z < *(px+1))
--iidx;
return iidx;
}
};
template <typename T, Algos A>
struct AlgoScalarBase<T, A, typename std::enable_if<DirectAux::IsDirect2<A>::value>::type> : DirectAux::DirectInfo<2, T, A>
{
private:
typedef DirectAux::DirectInfo<2, T, A> base_t;
static const size_t Offset = 2;
public:
AlgoScalarBase(const T *x, const uint32 n)
: base_t(x, n)
{
}
template <InstrSet I, typename T, Algos A>
struct AlgoVecBase<I, T, A, typename std::enable_if<DirectAux::IsDirect2<A>::value>::type> : AlgoScalarBase<T, A>
{
static const uint32 nElem = sizeof(typename InstrFloatTraits<I, T>::vec_t) / sizeof(T);
FORCE_INLINE uint32 scalar(T z) const
{
const T *px = base_t::data.xi;
const uint32 *buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
uint32 bidx = base_t::fun_t::f(base_t::data.scaler, base_t::data.cst0, z);
uint32 iidx = buckets[bidx];
px += iidx;
if (z < *px)
--iidx;
if (z < *(px + 1))
--iidx;
return iidx;
}
};
typedef FVec<I, T> fVec;
typedef IVec<SSE, T> i128;
template <InstrSet I, typename T, Algos A>
struct AlgoVecBase<I, T, A, typename std::enable_if<DirectAux::IsDirect2<A>::value>::type> : AlgoScalarBase<T, A>
{
static const uint32 nElem = sizeof(typename InstrFloatTraits<I, T>::vec_t) / sizeof(T);
struct Constants
{
fVec vscaler;
fVec vcst0;
IVec<I, T> one;
};
typedef FVec<I, T> fVec;
typedef IVec<SSE, T> i128;
private:
typedef AlgoScalarBase<T, A> base_t;
struct Constants
{
fVec vscaler;
fVec vcst0;
IVec<I, T> one;
};
FORCE_INLINE
//NO_INLINE
void resolve(const FVec<SSE, float>& vz, const IVec<SSE, float>& bidx, uint32 *pr) const
{
union U {
__m128i vec;
uint32 ui32[4];
} u;
private:
typedef AlgoScalarBase<T, A> base_t;
const uint32* buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const float *xi = base_t::data.xi;
FORCE_INLINE
// NO_INLINE
void resolve(const FVec<SSE, float> &vz, const IVec<SSE, float> &bidx, uint32 *pr) const
{
union U
{
__m128i vec;
uint32 ui32[4];
} u;
// read indices t
const double *p3 = reinterpret_cast<const double *>(&xi[(u.ui32[3] = buckets[bidx.get3()])]);
const double *p2 = reinterpret_cast<const double *>(&xi[(u.ui32[2] = buckets[bidx.get2()])]);
const double *p1 = reinterpret_cast<const double *>(&xi[(u.ui32[1] = buckets[bidx.get1()])]);
const double *p0 = reinterpret_cast<const double *>(&xi[(u.ui32[0] = buckets[bidx.get0()])]);
const uint32 *buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const float *xi = base_t::data.xi;
// read indices t
const double *p3 = reinterpret_cast<const double *>(&xi[(u.ui32[3] = buckets[bidx.get3()])]);
const double *p2 = reinterpret_cast<const double *>(&xi[(u.ui32[2] = buckets[bidx.get2()])]);
const double *p1 = reinterpret_cast<const double *>(&xi[(u.ui32[1] = buckets[bidx.get1()])]);
const double *p0 = reinterpret_cast<const double *>(&xi[(u.ui32[0] = buckets[bidx.get0()])]);
#if 0
// read pairs ( X(t-1), X(t) )
@ -87,65 +89,66 @@ private:
__m128 vxm = _mm_shuffle_ps(u01, u23, (0) + (1 << 2) + (0 << 4) + (1 << 6));
__m128 vxp = _mm_shuffle_ps(u01, u23, (2) + (3 << 2) + (2 << 4) + (3 << 6));
#else
__m128 xp23 = _mm_castpd_ps(_mm_set_pd(*p3, *p2));
__m128 xp01 = _mm_castpd_ps(_mm_set_pd(*p1, *p0));
__m128 vxm = _mm_shuffle_ps(xp01, xp23, (0) + (2 << 2) + (0 << 4) + (2 << 6));
__m128 vxp = _mm_shuffle_ps(xp01, xp23, (1) + (3 << 2) + (1 << 4) + (3 << 6));
__m128 xp23 = _mm_castpd_ps(_mm_set_pd(*p3, *p2));
__m128 xp01 = _mm_castpd_ps(_mm_set_pd(*p1, *p0));
__m128 vxm = _mm_shuffle_ps(xp01, xp23, (0) + (2 << 2) + (0 << 4) + (2 << 6));
__m128 vxp = _mm_shuffle_ps(xp01, xp23, (1) + (3 << 2) + (1 << 4) + (3 << 6));
#endif
IVec<SSE, float> i(u.vec);
IVec<SSE, float> vlem = vz < vxm;
IVec<SSE, float> vlep = vz < vxp;
i = i + vlem + vlep;
i.store(pr);
}
IVec<SSE, float> i(u.vec);
IVec<SSE, float> vlem = (vz < vxm);
IVec<SSE, float> vlep = (vz < vxp);
i = i + vlem + vlep;
i.store(pr);
}
FORCE_INLINE
//NO_INLINE
void resolve(const FVec<SSE, double>& vz, const IVec<SSE, float>& bidx, uint32 *pr) const
{
const uint32* buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const double *xi = base_t::data.xi;
FORCE_INLINE
// NO_INLINE
void resolve(const FVec<SSE, double> &vz, const IVec<SSE, float> &bidx, uint32 *pr) const
{
const uint32 *buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const double *xi = base_t::data.xi;
uint32 b1 = buckets[bidx.get1()];
uint32 b0 = buckets[bidx.get0()];
uint32 b1 = buckets[bidx.get1()];
uint32 b0 = buckets[bidx.get0()];
const double *p1 = &xi[b1];
const double *p0 = &xi[b0];
const double *p1 = &xi[b1];
const double *p0 = &xi[b0];
// read pairs ( X(t-1), X(t) )
__m128d vx1 = _mm_loadu_pd(p1);
__m128d vx0 = _mm_loadu_pd(p0);
// read pairs ( X(t-1), X(t) )
__m128d vx1 = _mm_loadu_pd(p1);
__m128d vx0 = _mm_loadu_pd(p0);
// build:
// { X(t(0)-1), X(t(1)-1) }
// { X(t(0)), X(t(1)) }
__m128d vxm = _mm_shuffle_pd(vx0, vx1, 0);
__m128d vxp = _mm_shuffle_pd(vx0, vx1, 3);
// build:
// { X(t(0)-1), X(t(1)-1) }
// { X(t(0)), X(t(1)) }
__m128d vxm = _mm_shuffle_pd(vx0, vx1, 0);
__m128d vxp = _mm_shuffle_pd(vx0, vx1, 3);
IVec<SSE, double> i(b1, b0);
IVec<SSE, double> vlem = (vz < vxm);
IVec<SSE, double> vlep = (vz < vxp);
i = i + vlem + vlep;
IVec<SSE, double> i(b1, b0);
IVec<SSE, double> vlem = (vz < vxm);
IVec<SSE, double> vlep = (vz < vxp);
i = i + vlem + vlep;
union {
__m128i vec;
uint32 ui32[4];
} u;
u.vec = i;
pr[0] = u.ui32[0];
pr[1] = u.ui32[2];
}
union
{
__m128i vec;
uint32 ui32[4];
} u;
u.vec = i;
pr[0] = u.ui32[0];
pr[1] = u.ui32[2];
}
#ifdef USE_AVX
FORCE_INLINE
//NO_INLINE
void resolve(const FVec<AVX, float>& vz, const IVec<AVX, float>& bidx, uint32 *pr) const
{
const uint32* buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const float *xi = base_t::data.xi;
FORCE_INLINE
// NO_INLINE
void resolve(const FVec<AVX, float> &vz, const IVec<AVX, float> &bidx, uint32 *pr) const
{
const uint32 *buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const float *xi = base_t::data.xi;
#if 0 // use gather instructions
#if 0 // use gather instructions
IVec<AVX,float> idxm;
idxm.setidx(buckets, bidx);
@ -159,21 +162,22 @@ private:
#else // do not use gather instrucions
union U {
__m256i vec;
uint32 ui32[8];
} u;
union U
{
__m256i vec;
uint32 ui32[8];
} u;
// read indices t
// read indices t
const double *p7 = reinterpret_cast<const double *>(&xi[(u.ui32[7] = buckets[bidx.get7()])]);
const double *p6 = reinterpret_cast<const double *>(&xi[(u.ui32[6] = buckets[bidx.get6()])]);
const double *p5 = reinterpret_cast<const double *>(&xi[(u.ui32[5] = buckets[bidx.get5()])]);
const double *p4 = reinterpret_cast<const double *>(&xi[(u.ui32[4] = buckets[bidx.get4()])]);
const double *p3 = reinterpret_cast<const double *>(&xi[(u.ui32[3] = buckets[bidx.get3()])]);
const double *p2 = reinterpret_cast<const double *>(&xi[(u.ui32[2] = buckets[bidx.get2()])]);
const double *p1 = reinterpret_cast<const double *>(&xi[(u.ui32[1] = buckets[bidx.get1()])]);
const double *p0 = reinterpret_cast<const double *>(&xi[(u.ui32[0] = buckets[bidx.get0()])]);
const double *p7 = reinterpret_cast<const double *>(&xi[(u.ui32[7] = buckets[bidx.get7()])]);
const double *p6 = reinterpret_cast<const double *>(&xi[(u.ui32[6] = buckets[bidx.get6()])]);
const double *p5 = reinterpret_cast<const double *>(&xi[(u.ui32[5] = buckets[bidx.get5()])]);
const double *p4 = reinterpret_cast<const double *>(&xi[(u.ui32[4] = buckets[bidx.get4()])]);
const double *p3 = reinterpret_cast<const double *>(&xi[(u.ui32[3] = buckets[bidx.get3()])]);
const double *p2 = reinterpret_cast<const double *>(&xi[(u.ui32[2] = buckets[bidx.get2()])]);
const double *p1 = reinterpret_cast<const double *>(&xi[(u.ui32[1] = buckets[bidx.get1()])]);
const double *p0 = reinterpret_cast<const double *>(&xi[(u.ui32[0] = buckets[bidx.get0()])]);
#if 0 // perform 8 loads in double precision
@ -210,96 +214,93 @@ private:
IVec<AVX, float> ip(u.vec);
#else // use __mm256_set_pd
#else // use __mm256_set_pd
// read pairs ( X(t-1), X(t) )
__m256 x0145 = _mm256_castpd_ps(_mm256_set_pd(*p5, *p4, *p1, *p0)); // { x0(t-1), x0(t), x1(t-1), x1(t), x4(t-1), x4(t), x5(t-1), x5(t) }
__m256 x2367 = _mm256_castpd_ps(_mm256_set_pd(*p7, *p6, *p3, *p2)); // { x2(t-1), x2(t), x3(t-1), x3(t), x6(t-1), x6(t), x7(t-1), x7(t) }
// read pairs ( X(t-1), X(t) )
__m256 x0145 = _mm256_castpd_ps(_mm256_set_pd(*p5, *p4, *p1, *p0)); // { x0(t-1), x0(t), x1(t-1), x1(t), x4(t-1), x4(t), x5(t-1), x5(t) }
__m256 x2367 = _mm256_castpd_ps(_mm256_set_pd(*p7, *p6, *p3, *p2)); // { x2(t-1), x2(t), x3(t-1), x3(t), x6(t-1), x6(t), x7(t-1), x7(t) }
// { x0(t-1), x1(t-1), x2(t-1), 3(t-1, x4(t-1), x5(t-1), x6(t-1), xt(t-1) }
FVec<AVX, float> vxm = _mm256_shuffle_ps(x0145, x2367, 0 + (2 << 2) + (0 << 4) + (2 << 6) );
// { x0(t), x1(t), x2(t), 3(t, x4(t), x5(t), x6(t), xt(t) }
FVec<AVX, float> vxp = _mm256_shuffle_ps(x0145, x2367, 1 + (3 << 2) + (1 << 4) + (3 << 6) );
// { x0(t-1), x1(t-1), x2(t-1), 3(t-1, x4(t-1), x5(t-1), x6(t-1), xt(t-1) }
FVec<AVX, float> vxm = _mm256_shuffle_ps(x0145, x2367, 0 + (2 << 2) + (0 << 4) + (2 << 6));
// { x0(t), x1(t), x2(t), 3(t, x4(t), x5(t), x6(t), xt(t) }
FVec<AVX, float> vxp = _mm256_shuffle_ps(x0145, x2367, 1 + (3 << 2) + (1 << 4) + (3 << 6));
IVec<AVX, float> ip(u.vec);
IVec<AVX, float> ip(u.vec);
#endif
#endif
IVec<AVX, float> vlem = vz < vxm;
IVec<AVX, float> vlep = vz < vxp;
ip = ip + vlem + vlep;
IVec<AVX, float> vlem = vz < vxm;
IVec<AVX, float> vlep = vz < vxp;
ip = ip + vlem + vlep;
ip.store(pr);
}
ip.store(pr);
}
FORCE_INLINE
// NO_INLINE
void resolve(const FVec<AVX, double> &vz, const IVec<SSE, float> &bidx, uint32 *pr) const
{
union
{
__m256i vec;
uint64 ui64[4];
} u;
const uint32 *buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const double *xi = base_t::data.xi;
FORCE_INLINE
//NO_INLINE
void resolve(const FVec<AVX, double>& vz, const IVec<SSE, float>& bidx, uint32 *pr) const
{
union {
__m256i vec;
uint64 ui64[4];
} u;
// read indices t
const double *p3 = &xi[(u.ui64[3] = buckets[bidx.get3()])];
const double *p2 = &xi[(u.ui64[2] = buckets[bidx.get2()])];
const double *p1 = &xi[(u.ui64[1] = buckets[bidx.get1()])];
const double *p0 = &xi[(u.ui64[0] = buckets[bidx.get0()])];
const uint32* buckets = reinterpret_cast<const uint32 *>(base_t::data.buckets);
const double *xi = base_t::data.xi;
// read pairs ( X(t-1), X(t) )
__m128d xp3 = _mm_loadu_pd(p3);
__m128d xp2 = _mm_loadu_pd(p2);
__m128d xp1 = _mm_loadu_pd(p1);
__m128d xp0 = _mm_loadu_pd(p0);
// read indices t
const double *p3 = &xi[(u.ui64[3] = buckets[bidx.get3()])];
const double *p2 = &xi[(u.ui64[2] = buckets[bidx.get2()])];
const double *p1 = &xi[(u.ui64[1] = buckets[bidx.get1()])];
const double *p0 = &xi[(u.ui64[0] = buckets[bidx.get0()])];
// build:
// { X(t(0)-1), X(t(1)-1), X(t(2)-1), X(t(3)-1) }
// { X(t(0)), X(t(1)), X(t(2)), X(t(3)) }
__m256d x02 = _mm256_insertf128_pd(_mm256_castpd128_pd256(xp0), xp2, 1);
__m256d x13 = _mm256_insertf128_pd(_mm256_castpd128_pd256(xp1), xp3, 1);
FVec<AVX, double> vxm = _mm256_unpacklo_pd(x02, x13);
FVec<AVX, double> vxp = _mm256_unpackhi_pd(x02, x13);
// read pairs ( X(t-1), X(t) )
__m128d xp3 = _mm_loadu_pd(p3);
__m128d xp2 = _mm_loadu_pd(p2);
__m128d xp1 = _mm_loadu_pd(p1);
__m128d xp0 = _mm_loadu_pd(p0);
// __m128d h01m = _mm_shuffle_pd(xp0, xp1, 0);
// __m128d h23m = _mm_shuffle_pd(xp2, xp3, 0);
// __m128d h01p = _mm_shuffle_pd(xp0, xp1, 3);
// __m128d h23p = _mm_shuffle_pd(xp2, xp3, 3);
// FVec<AVX, double> vxm = _mm256_insertf128_pd(_mm256_castpd128_pd256(h01m), h23m, 1);
// FVec<AVX, double> vxp = _mm256_insertf128_pd(_mm256_castpd128_pd256(h01p), h23p, 1);
// build:
// { X(t(0)-1), X(t(1)-1), X(t(2)-1), X(t(3)-1) }
// { X(t(0)), X(t(1)), X(t(2)), X(t(3)) }
__m256d x02 = _mm256_insertf128_pd(_mm256_castpd128_pd256(xp0), xp2, 1);
__m256d x13 = _mm256_insertf128_pd(_mm256_castpd128_pd256(xp1), xp3, 1);
FVec<AVX, double> vxm = _mm256_unpacklo_pd(x02,x13);
FVec<AVX, double> vxp = _mm256_unpackhi_pd(x02,x13);
// __m128d h01m = _mm_shuffle_pd(xp0, xp1, 0);
// __m128d h23m = _mm_shuffle_pd(xp2, xp3, 0);
// __m128d h01p = _mm_shuffle_pd(xp0, xp1, 3);
// __m128d h23p = _mm_shuffle_pd(xp2, xp3, 3);
// FVec<AVX, double> vxm = _mm256_insertf128_pd(_mm256_castpd128_pd256(h01m), h23m, 1);
// FVec<AVX, double> vxp = _mm256_insertf128_pd(_mm256_castpd128_pd256(h01p), h23p, 1);
IVec<AVX, double> i(u.vec);
IVec<AVX, double> vlem = vz < vxm;
IVec<AVX, double> vlep = vz < vxp;
i = i + vlem + vlep;
i.extractLo32s().store(pr);
}
IVec<AVX, double> i(u.vec);
IVec<AVX, double> vlem = vz < vxm;
IVec<AVX, double> vlep = vz < vxp;
i = i + vlem + vlep;
i.extractLo32s().store(pr);
}
#endif
public:
public:
AlgoVecBase(const T *x, const uint32 n) : base_t(x, n) {}
AlgoVecBase(const T* x, const uint32 n) : base_t(x, n) {}
void initConstants(Constants &cst) const
{
cst.vscaler.setN(base_t::data.scaler);
cst.vcst0.setN(base_t::data.cst0);
cst.one.setN(uint32(1));
}
void initConstants(Constants& cst) const
{
cst.vscaler.setN(base_t::data.scaler);
cst.vcst0.setN(base_t::data.cst0);
cst.one.setN(uint32(1));
}
void vectorial(uint32 *pr, const T *pz, const Constants& cst) const
{
fVec vz(pz);
resolve(vz, base_t::fun_t::f(cst.vscaler, cst.vcst0, vz), pr);
}
};
} // namespace Details
void vectorial(uint32 *pr, const T *pz, const Constants &cst) const
{
fVec vz(pz);
resolve(vz, base_t::fun_t::f(cst.vscaler, cst.vcst0, vz), pr);
}
};
} // namespace Details
} // namespace BinSearch