Avinash Kethineedi
f6ef19f5a9
* Update copyright information and add SPDX license identifier * Update AUTHORS * Remove `sos_tests`
336 lines
12 KiB
Python
336 lines
12 KiB
Python
###############################################################################
|
|
# Copyright (c) Advanced Micro Devices, Inc. All rights reserved.
|
|
#
|
|
# SPDX-License-Identifier: MIT
|
|
#
|
|
# Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
# of this software and associated documentation files (the "Software"), to
|
|
# deal in the Software without restriction, including without limitation the
|
|
# rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
|
|
# sell copies of the Software, and to permit persons to whom the Software is
|
|
# furnished to do so, subject to the following conditions:
|
|
#
|
|
# The above copyright notice and this permission notice shall be included in
|
|
# all copies or substantial portions of the Software.
|
|
#
|
|
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
|
# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
|
|
# IN THE SOFTWARE.
|
|
###############################################################################
|
|
|
|
import os
|
|
|
|
types = [
|
|
("float", "float"),
|
|
("double", "double"),
|
|
("char", "char"),
|
|
("signed char", "schar"),
|
|
("short", "short"),
|
|
("int", "int"),
|
|
("long", "long"),
|
|
("long long", "longlong"),
|
|
("unsigned char", "uchar"),
|
|
("unsigned short", "ushort"),
|
|
("unsigned int", "uint"),
|
|
("unsigned long", "ulong"),
|
|
("unsigned long long", "ulonglong"),
|
|
]
|
|
|
|
|
|
def put_api(T, TNAME):
|
|
return (
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_ctx_{TNAME}_put(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_{TNAME}_put(\n"
|
|
f" {T} *dest, const {T} *source, size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_ctx_{TNAME}_put(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_{TNAME}_put({T} *dest,\n"
|
|
f" const {T} *source, size_t nelems, int pe);\n\n"
|
|
)
|
|
|
|
|
|
def generate_put_api():
|
|
expanded_code = """
|
|
/**
|
|
* @name SHMEM_PUT
|
|
* @brief Writes contiguous data of \p nelems elements from \p source on the
|
|
* calling PE to \p dest at \p pe. The caller will block until the operation
|
|
* completes locally (it is safe to reuse \p source). The caller must
|
|
* call into rocshmem_quiet() if remote completion is required.
|
|
*
|
|
* This function can be called from divergent control paths at per-thread
|
|
* granularity. However, performance may be improved if the caller can
|
|
* coalesce contiguous messages and elect a leader thread to call into the
|
|
* ROCSHMEM function.
|
|
*
|
|
* @param[in] ctx Context with which to perform this operation.
|
|
* @param[in] dest Destination address. Must be an address on the symmetric
|
|
* heap.
|
|
* @param[in] source Source address. Must be an address on the symmetric heap.
|
|
* @param[in] nelems Size of the transfer in number of elements.
|
|
* @param[in] pe PE of the remote process.
|
|
*
|
|
* @return void.
|
|
*/\n"""
|
|
for type_, tname_ in types:
|
|
expanded_code += put_api(type_, tname_)
|
|
|
|
return expanded_code
|
|
|
|
|
|
def get_api(T, TNAME):
|
|
return (
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_ctx_{TNAME}_get(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_{TNAME}_get(\n"
|
|
f" {T} *dest, const {T} *source, size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_ctx_{TNAME}_get(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_{TNAME}_get({T} *dest,\n"
|
|
f" const {T} *source, size_t nelems, int pe);\n\n"
|
|
)
|
|
|
|
|
|
def generate_get_api():
|
|
expanded_code = """
|
|
/**
|
|
* @name SHMEM_GET
|
|
* @brief Reads contiguous data of \p nelems elements from \p source on \p pe
|
|
* to \p dest on the calling PE. The calling work-group will block until the
|
|
* operation completes (data has been placed in \p dest).
|
|
*
|
|
* This function can be called from divergent control paths at per-thread
|
|
* granularity. However, performance may be improved if the caller can
|
|
* coalesce contiguous messages and elect a leader thread to call into the
|
|
* ROCSHMEM function.
|
|
*
|
|
* @param[in] ctx Context with which to perform this operation.
|
|
* @param[in] dest Destination address. Must be an address on the symmetric
|
|
* heap.
|
|
* @param[in] source Source address. Must be an address on the symmetric heap.
|
|
* @param[in] nelems Size of the transfer in bytes.
|
|
* @param[in] pe PE of the remote process.
|
|
*
|
|
* @return void.
|
|
*/\n"""
|
|
for type_, tname_ in types:
|
|
expanded_code += get_api(type_, tname_)
|
|
|
|
return expanded_code
|
|
|
|
|
|
def p_api(T, TNAME):
|
|
return (
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_ctx_{TNAME}_p(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, {T} value,\n"
|
|
f" int pe);\n"
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_{TNAME}_p(\n"
|
|
f" {T} *dest, {T} value, int pe);\n"
|
|
f"__host__ void rocshmem_ctx_{TNAME}_p(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, {T} value,\n"
|
|
f" int pe);\n"
|
|
f"__host__ void rocshmem_{TNAME}_p(\n"
|
|
f" {T} *dest, {T} value, int pe);\n\n"
|
|
)
|
|
|
|
|
|
def generate_p_api():
|
|
expanded_code = """
|
|
/**
|
|
* @name SHMEM_P
|
|
* @brief Writes a single value to \p dest at \p pe PE to \p dst at \p pe.
|
|
* The caller must call into rocshmem_quiet() if remote completion is
|
|
* required.
|
|
*
|
|
* This function can be called from divergent control paths at per-thread
|
|
* granularity. However, performance may be improved if the caller can
|
|
* coalesce contiguous messages and elect a leader thread to call into the
|
|
* ROCSHMEM function.
|
|
*
|
|
* @param[in] ctx Context with which to perform this operation.
|
|
* @param[in] dest Destination address. Must be an address on the symmetric
|
|
* heap.
|
|
* @param[in] value Value to write to dest at \p pe.
|
|
* @param[in] pe PE of the remote process.
|
|
*
|
|
* @return void.
|
|
*/\n"""
|
|
for type_, tname_ in types:
|
|
expanded_code += p_api(type_, tname_)
|
|
|
|
return expanded_code
|
|
|
|
|
|
def g_api(T, TNAME):
|
|
return (
|
|
f"__device__ ATTR_NO_INLINE {T} rocshmem_ctx_{TNAME}_g(\n"
|
|
f" rocshmem_ctx_t ctx, const {T} *source, int pe);\n"
|
|
f"__device__ ATTR_NO_INLINE {T} rocshmem_{TNAME}_g(\n"
|
|
f" const {T} *source, int pe);\n"
|
|
f"__host__ {T} rocshmem_ctx_{TNAME}_g(\n"
|
|
f" rocshmem_ctx_t ctx, const {T} *source, int pe);\n"
|
|
f"__host__ {T} rocshmem_{TNAME}_g(\n"
|
|
f" const {T} *source, int pe);\n\n"
|
|
)
|
|
|
|
|
|
def generate_g_api():
|
|
expanded_code = """
|
|
/**
|
|
* @name SHMEM_G
|
|
* @brief reads and returns single value from \p source at \p pe.
|
|
* The calling work-group/thread will block until the operation completes.
|
|
*
|
|
* This function can be called from divergent control paths at per-thread
|
|
* granularity. However, performance may be improved if the caller can
|
|
* coalesce contiguous messages and elect a leader thread to call into the
|
|
* ROCSHMEM function.
|
|
*
|
|
* @param[in] ctx Context with which to perform this operation.
|
|
* @param[in] source Source address. Must be an address on the symmetric
|
|
* heap.
|
|
* @param[in] pe PE of the remote process.
|
|
*
|
|
* @return the value read from remote \p source at \p pe.
|
|
*/\n"""
|
|
for type_, tname_ in types:
|
|
expanded_code += g_api(type_, tname_)
|
|
|
|
return expanded_code
|
|
|
|
|
|
def put_nbi_api(T, TNAME):
|
|
return (
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_ctx_{TNAME}_put_nbi(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_{TNAME}_put_nbi(\n"
|
|
f" {T} *dest, const {T} *source, size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_ctx_{TNAME}_put_nbi(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_{TNAME}_put_nbi(\n"
|
|
f" {T} *dest, const {T} *source, size_t nelems, int pe);\n\n"
|
|
)
|
|
|
|
|
|
def generate_put_nbi_api():
|
|
expanded_code = """
|
|
/**
|
|
* @name SHMEM_PUT_NBI
|
|
* @brief Writes contiguous data of \p nelems elements from \p source on the
|
|
* calling PE to \p dest on \p pe. The operation is not blocking. The caller
|
|
* will return as soon as the request is posted. The caller must call
|
|
* rocshmem_quiet() on the same context if completion notification is
|
|
* required.
|
|
*
|
|
* This function can be called from divergent control paths at per-thread
|
|
* granularity. However, performance may be improved if the caller can
|
|
* coalesce contiguous messages and elect a leader thread to call into the
|
|
* ROCSHMEM function.
|
|
*
|
|
* @param[in] ctx Context with which to perform this operation.
|
|
* @param[in] dest Destination address. Must be an address on the symmetric
|
|
heap.
|
|
* @param[in] source Source address. Must be an address on the symmetric heap.
|
|
* @param[in] nelems Size of the transfer in bytes.
|
|
* @param[in] pe PE of the remote process.
|
|
*
|
|
* @return void.
|
|
*/\n"""
|
|
for type_, tname_ in types:
|
|
expanded_code += put_nbi_api(type_, tname_)
|
|
|
|
return expanded_code
|
|
|
|
|
|
def get_nbi_api(T, TNAME):
|
|
return (
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_ctx_{TNAME}_get_nbi(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__device__ ATTR_NO_INLINE void rocshmem_{TNAME}_get_nbi(\n"
|
|
f" {T} *dest, const {T} *source, size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_ctx_{TNAME}_get_nbi(\n"
|
|
f" rocshmem_ctx_t ctx, {T} *dest, const {T} *source,\n"
|
|
f" size_t nelems, int pe);\n"
|
|
f"__host__ void rocshmem_{TNAME}_get_nbi({T} *dest,\n"
|
|
f" const {T} *source, size_t nelems, int pe);\n\n"
|
|
)
|
|
|
|
|
|
def generate_get_nbi_api():
|
|
expanded_code = """
|
|
/**
|
|
* @name SHMEM_GET_NBI
|
|
* @brief Reads contiguous data of \p nelems elements from \p source on \p pe
|
|
* to \p dest on the calling PE. The operation is not blocking. The caller will
|
|
* return as soon as the request is posted. The caller must call
|
|
* rocshmem_quiet() on the same context if completion notification is
|
|
* required.
|
|
*
|
|
* This function can be called from divergent control paths at per-thread
|
|
* granularity. However, performance may be improved if the caller can
|
|
* coalesce contiguous messages and elect a leader thread to call into the
|
|
* ROCSHMEM function.
|
|
*
|
|
* @param[in] ctx Context with which to perform this operation.
|
|
* @param[in] dest Destination address. Must be an address on the symmetric
|
|
* heap.
|
|
* @param[in] source Source address. Must be an address on the symmetric heap.
|
|
* @param[in] nelems Size of the transfer in bytes.
|
|
* @param[in] pe PE of the remote process.
|
|
*
|
|
* @return void.
|
|
*/\n"""
|
|
for type_, tname_ in types:
|
|
expanded_code += get_nbi_api(type_, tname_)
|
|
|
|
return expanded_code
|
|
|
|
|
|
def write_to_file(filename, content):
|
|
with open(filename, 'w') as file:
|
|
file.write(content)
|
|
|
|
|
|
def generate_RMA_header(output_dir, copyright):
|
|
expanded_code = copyright
|
|
|
|
expanded_code += """
|
|
#ifndef LIBRARY_INCLUDE_ROCSHMEM_RMA_HPP
|
|
#define LIBRARY_INCLUDE_ROCSHMEM_RMA_HPP
|
|
|
|
namespace rocshmem {
|
|
"""
|
|
|
|
expanded_code += (
|
|
generate_put_api() +
|
|
generate_p_api() +
|
|
generate_get_api() +
|
|
generate_g_api() +
|
|
generate_put_nbi_api() +
|
|
generate_get_nbi_api()
|
|
)
|
|
|
|
expanded_code += """
|
|
} // namespace rocshmem
|
|
|
|
#endif // LIBRARY_INCLUDE_ROCSHMEM_RMA_HPP
|
|
"""
|
|
|
|
output_file = os.path.join(
|
|
output_dir, 'rocshmem_RMA.hpp'
|
|
)
|
|
|
|
write_to_file(output_file, expanded_code)
|