hlrs-software-stack/spack
7
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

Fixed some issues with naming

Browse source
This commit is contained in:
Mario Melara 2015-11-02 11:08:08 -08:00
parent 9a91da9ccd
commit eb2cf1698f
1 changed files with 8 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

23
lib/spack/spack/architecture.py
View file

@ -49,13 +49,6 @@ def __init__(self):
class Target(object):
""" This is the processor type e.g. cray-ivybridge """
# Front end or back end target. Target needs to know which one this is
# Should autodetect from the machine
# features of a target
# - a module name
# -a compiler finding strategy
# -a name
# architecture classes handling the aliasing for front-end, back-end and default
def __init__(self,name, module_name=None):
self.name = name # case of cray "ivybridge" but if it's x86_64
@ -67,7 +60,8 @@ def compiler_strategy(self):
return "MODULES"
else:
return "PATH"
class Architecture(object):
""" Abstract class that each type of Architecture will subclass. Will return a instance of it once it
is returned
@ -85,17 +79,19 @@ def __init__(self, name):
def add_target(self, name, target):
self.targets[name] = target
def target(self, name):
"""This is a getter method for the target dictionary that handles defaulting based
on the values provided by default, front-end, and back-end. This can be overwritten
by a subclass for which we want to provide further aliasing options.
"""
if name == 'default':
name = default
name = self.default
elif name == 'front_end' or name == 'fe':
name = front-end
name = self.front_end
elif name == 'back_end' or name == 'be':
name = back-end
name = self.back_end
return self.targets[name]
@classmethod
@ -148,10 +144,7 @@ def get_sys_type_from_uname():
return None
def get_sys_type_from_config_file():
""" Should read in a sys_type from the config yaml file. This should be the first thing looked at since
The user can specify that the architecture is a cray-xc40. A template yaml should be created when spack
is installed. Similar to .spackconfig
"""
spack_home_dir = os.environ["HOME"] + "/.spack"
yaml_file = os.path.join(spack_home_dir, 'architecture.yaml')
try: