1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
|
# Copyright 2013-2023 Lawrence Livermore National Security, LLC and other
# Spack Project Developers. See the top-level COPYRIGHT file for details.
#
# SPDX-License-Identifier: (Apache-2.0 OR MIT)
from spack.package import *
class Zfp(CMakePackage, CudaPackage):
"""zfp is a compressed number format for multidimensional floating-point
and integer arrays.
zfp provides compressed-array classes that support high throughput
read and write random access to individual array elements. zfp also
supports serial and parallel (OpenMP and CUDA) compression of whole
arrays.
"""
# Package info
homepage = "https://zfp.llnl.gov"
url = "https://github.com/LLNL/zfp/releases/download/1.0.0/zfp-1.0.0.tar.gz"
git = "https://github.com/LLNL/zfp.git"
maintainers("lindstro", "GarrettDMorrison")
tags = ["radiuss", "e4s"]
license("BSD-3-Clause")
# Versions
version("develop", branch="develop")
version("1.0.0", sha256="0ea08ae3e50e3c92f8b8cf41ba5b6e2de8892bc4a4ca0c59b8945b6c2ab617c4")
version("0.5.5", sha256="fdf7b948bab1f4e5dccfe2c2048fd98c24e417ad8fb8a51ed3463d04147393c5")
version("0.5.4", sha256="746e17aaa401c67dcffd273d6e6f95c76adfbbd5cf523dcad56d09e9d3b71196")
version("0.5.3", sha256="a5d2f8e5b47a7c92e2a5775b82cbfb3a76c87d0ac83d25abb4ac10ea75a2856e")
version(
"0.5.2",
sha256="9c738ec525cc76b4bb80b2b3f7c9f07507eeda3a341470e5942cda97efbe9a4f",
url="https://github.com/LLNL/zfp/archive/0.5.2/zfp-0.5.2.tar.gz",
)
version(
"0.5.1",
sha256="f255dd1708c9ae4dc6a56dd2614e8b47a10d833c87fd349cbd47545a19c2b779",
url="https://github.com/LLNL/zfp/archive/0.5.1/zfp-0.5.1.tar.gz",
)
# Dependencies
depends_on("cmake@3.9.0:", type="build")
depends_on("cuda@7:", type=("build", "test", "run"), when="+cuda")
depends_on("python", type=("build", "test", "run"), when="+python")
depends_on("py-numpy", type=("build", "test", "run"), when="+python")
depends_on("py-cython", type="build", when="+python")
# Build targets
variant("shared", default=True, description="Build shared libraries")
variant("utilities", default=True, description="Build zfp utilities")
# Language bindings
variant("c", default=False, when="@0.5.4:", description="Enable compressed array C bindings")
variant("python", default=False, when="@0.5.5:", description="Enable Python bindings")
variant("fortran", default=False, when="@0.5.5:", description="Enable Fortran bindings")
# Execution policies
variant("openmp", default=False, when="@0.5.3:", description="Enable OpenMP execution")
variant("cuda", default=False, when="@0.5.4:", description="Enable CUDA execution")
# Advanced settings
variant("aligned", default=False, description="Enable aligned memory allocation")
variant(
"bsws",
default="64",
values=("8", "16", "32", "64"),
multi=False,
description="Bit stream word size: "
"use smaller for finer rate granularity. "
"Use 8 for H5Z-ZFP filter.",
)
variant(
"daz",
default=False,
when="@1.0.0:",
description="Denormals are zero: "
"Treat denormal-only blocks as containing "
"all zeroes",
)
variant(
"fasthash",
default=False,
when="@0.5.2:",
description="Use a faster but more collision prone hash function",
)
variant("profile", default=False, when="@0.5.2:", description="Count cache misses")
variant(
"round",
default="never",
values=("never", "first", "last"),
multi=False,
when="@1.0.0:",
description="EXPERIMENTAL: Set coefficient rounding method",
)
variant(
"strided", default=False, description="Enable strided access for progressive zfp streams"
)
variant(
"tight-error",
default=False,
when="@1.0.0:",
description="EXPERIMENTAL: Use tighter error bound when rounding is enabled",
)
variant("twoway", default=False, description="Use two-way skew-associative cache")
# Conflicts
conflicts(
"+tight-error",
when="round=never",
msg="Using zfp with tight error requires a rounding mode other than never",
)
# CMake options
def cmake_args(self):
spec = self.spec
args = [
self.define("BUILD_TESTING", self.run_tests),
self.define_from_variant("BUILD_SHARED_LIBS", "shared"),
self.define_from_variant("BUILD_UTILITIES", "utilities"),
self.define_from_variant("BUILD_CFP", "c"),
self.define_from_variant("BUILD_ZFPY", "python"),
self.define_from_variant("BUILD_ZFORP", "fortran"),
self.define_from_variant("ZFP_WITH_OPENMP", "openmp"),
self.define_from_variant("ZFP_WITH_CUDA", "cuda"),
self.define_from_variant("ZFP_WITH_ALIGNED_ALLOC", "aligned"),
self.define("ZFP_BIT_STREAM_WORD_SIZE", spec.variants["bsws"].value),
self.define_from_variant("ZFP_WITH_DAZ", "daz"),
self.define_from_variant("ZFP_WITH_CACHE_FAST_HASH", "fasthash"),
self.define_from_variant("ZFP_WITH_CACHE_PROFILE", "profile"),
self.define_from_variant("ZFP_WITH_BIT_STREAM_STRIDED", "strided"),
self.define_from_variant("ZFP_WITH_TIGHT_ERROR", "tight-error"),
self.define_from_variant("ZFP_WITH_CACHE_TWOWAY", "twoway"),
]
if "round" in spec.variants:
args.append(
"ZFP_ROUNDING_MODE=ZFP_ROUND_{0}".format(spec.variants["round"].value.upper())
)
if "+cuda" in spec:
args.append("-DCUDA_BIN_DIR={0}".format(spec["cuda"].prefix.bin))
if not spec.satisfies("cuda_arch=none"):
cuda_arch = spec.variants["cuda_arch"].value
args.append("-DCMAKE_CUDA_FLAGS=-arch sm_{0}".format(cuda_arch[0]))
return args
|
