def push_keys(*mirrors, **kwargs):
"""
Upload pgp public keys to the given mirrors
"""
keys = kwargs.get('keys')
regenerate_index = kwargs.get('regenerate_index', False)
tmpdir = kwargs.get('tmpdir')
remove_tmpdir = False
keys = spack.util.gpg.public_keys(*(keys or []))
try:
for mirror in mirrors:
push_url = getattr(mirror, 'push_url', mirror)
keys_url = url_util.join(push_url, _build_cache_relative_path,
_build_cache_keys_relative_path)
keys_local = url_util.local_file_path(keys_url)
verb = 'Writing' if keys_local else 'Uploading'
tty.debug('{0} public keys to {1}'.format(
verb, url_util.format(push_url)))
if keys_local: # mirror is local, don't bother with the tmpdir
prefix = keys_local
mkdirp(keys_local)
else:
# A tmp dir is created for the first mirror that is non-local.
# On the off-hand chance that all the mirrors are local, then
# we can avoid the need to create a tmp dir.
if tmpdir is None:
tmpdir = tempfile.mkdtemp()
remove_tmpdir = True
prefix = tmpdir
for fingerprint in keys:
tty.debug(' ' + fingerprint)
filename = fingerprint + '.pub'
export_target = os.path.join(prefix, filename)
spack.util.gpg.export_keys(export_target, fingerprint)
# If mirror is local, the above export writes directly to the
# mirror (export_target points directly to the mirror).
#
# If not, then export_target is a tmpfile that needs to be
# uploaded to the mirror.
if not keys_local:
spack.util.web.push_to_url(export_target,
url_util.join(
keys_url, filename),
keep_original=False)
if regenerate_index:
if keys_local:
generate_key_index(keys_url)
else:
generate_key_index(keys_url, tmpdir)
finally:
if remove_tmpdir:
shutil.rmtree(tmpdir)
def get_spec(spec=None, force=False):
"""
Check if spec.yaml exists on mirrors and return it if it does
"""
global _cached_specs
urls = set()
if spec is None:
return {}
specfile_name = tarball_name(spec, '.spec.yaml')
if not spack.mirror.MirrorCollection():
tty.debug("No Spack mirrors are currently configured")
return {}
if _cached_specs and spec in _cached_specs:
return _cached_specs
for mirror in spack.mirror.MirrorCollection().values():
fetch_url_build_cache = url_util.join(mirror.fetch_url,
_build_cache_relative_path)
mirror_dir = url_util.local_file_path(fetch_url_build_cache)
if mirror_dir:
tty.debug('Finding buildcaches in {0}'.format(mirror_dir))
link = url_util.join(fetch_url_build_cache, specfile_name)
urls.add(link)
else:
tty.debug('Finding buildcaches at {0}'.format(
url_util.format(fetch_url_build_cache)))
link = url_util.join(fetch_url_build_cache, specfile_name)
urls.add(link)
return try_download_specs(urls=urls, force=force)
def get_specs(force=False):
"""
Get spec.yaml's for build caches available on mirror
"""
global _cached_specs
if _cached_specs:
tty.debug("Using previously-retrieved specs")
return _cached_specs
if not spack.mirror.MirrorCollection():
tty.warn("No Spack mirrors are currently configured")
return {}
urls = set()
for mirror in spack.mirror.MirrorCollection().values():
fetch_url_build_cache = url_util.join(mirror.fetch_url,
_build_cache_relative_path)
mirror_dir = url_util.local_file_path(fetch_url_build_cache)
if mirror_dir:
tty.msg("Finding buildcaches in %s" % mirror_dir)
if os.path.exists(mirror_dir):
files = os.listdir(mirror_dir)
for file in files:
if re.search('spec.yaml', file):
link = url_util.join(fetch_url_build_cache, file)
urls.add(link)
else:
tty.msg("Finding buildcaches at %s" %
url_util.format(fetch_url_build_cache))
p, links = web_util.spider(
url_util.join(fetch_url_build_cache, 'index.html'))
for link in links:
if re.search("spec.yaml", link):
urls.add(link)
_cached_specs = []
for link in urls:
with Stage(link, name="build_cache", keep=True) as stage:
if force and os.path.exists(stage.save_filename):
os.remove(stage.save_filename)
if not os.path.exists(stage.save_filename):
try:
stage.fetch()
except fs.FetchError:
continue
with open(stage.save_filename, 'r') as f:
# read the spec from the build cache file. All specs
# in build caches are concrete (as they are built) so
# we need to mark this spec concrete on read-in.
spec = Spec.from_yaml(f)
spec._mark_concrete()
_cached_specs.append(spec)
return _cached_specs
def get_keys(install=False, trust=False, force=False, mirrors=None):
"""Get pgp public keys available on mirror with suffix .pub
"""
mirror_collection = (mirrors or spack.mirror.MirrorCollection())
if not mirror_collection:
tty.die("Please add a spack mirror to allow " +
"download of build caches.")
for mirror in mirror_collection.values():
fetch_url = mirror.fetch_url
keys_url = url_util.join(fetch_url, _build_cache_relative_path,
_build_cache_keys_relative_path)
keys_index = url_util.join(keys_url, 'index.json')
tty.debug('Finding public keys in {0}'.format(
url_util.format(fetch_url)))
try:
_, _, json_file = web_util.read_from_url(keys_index)
json_index = sjson.load(codecs.getreader('utf-8')(json_file))
except (URLError, web_util.SpackWebError) as url_err:
if web_util.url_exists(keys_index):
err_msg = [
'Unable to find public keys in {0},',
' caught exception attempting to read from {1}.',
]
tty.error(''.join(err_msg).format(url_util.format(fetch_url),
url_util.format(keys_index)))
tty.debug(url_err)
continue
for fingerprint, key_attributes in json_index['keys'].items():
link = os.path.join(keys_url, fingerprint + '.pub')
with Stage(link, name="build_cache", keep=True) as stage:
if os.path.exists(stage.save_filename) and force:
os.remove(stage.save_filename)
if not os.path.exists(stage.save_filename):
try:
stage.fetch()
except fs.FetchError:
continue
tty.debug('Found key {0}'.format(fingerprint))
if install:
if trust:
spack.util.gpg.trust(stage.save_filename)
tty.debug('Added this key to trusted keys.')
else:
tty.debug('Will not add this key to trusted keys.'
'Use -t to install all downloaded keys')
def update_index(mirror_url, update_keys=False):
mirror = spack.mirror.MirrorCollection().lookup(mirror_url)
outdir = url_util.format(mirror.push_url)
bindist.generate_package_index(
url_util.join(outdir, bindist.build_cache_relative_path()))
if update_keys:
keys_url = url_util.join(outdir, bindist.build_cache_relative_path(),
bindist.build_cache_keys_relative_path())
bindist.generate_key_index(keys_url)
def get_keys(install=False, trust=False, force=False):
"""
Get pgp public keys available on mirror
with suffix .key or .pub
"""
if not spack.mirror.MirrorCollection():
tty.die("Please add a spack mirror to allow " +
"download of build caches.")
keys = set()
for mirror in spack.mirror.MirrorCollection().values():
fetch_url_build_cache = url_util.join(mirror.fetch_url,
_build_cache_relative_path)
mirror_dir = url_util.local_file_path(fetch_url_build_cache)
if mirror_dir:
tty.msg("Finding public keys in %s" % mirror_dir)
files = os.listdir(str(mirror_dir))
for file in files:
if re.search(r'\.key', file) or re.search(r'\.pub', file):
link = url_util.join(fetch_url_build_cache, file)
keys.add(link)
else:
tty.msg("Finding public keys at %s" %
url_util.format(fetch_url_build_cache))
# For s3 mirror need to request index.html directly
p, links = web_util.spider(url_util.join(fetch_url_build_cache,
'index.html'),
depth=1)
for link in links:
if re.search(r'\.key', link) or re.search(r'\.pub', link):
keys.add(link)
for link in keys:
with Stage(link, name="build_cache", keep=True) as stage:
if os.path.exists(stage.save_filename) and force:
os.remove(stage.save_filename)
if not os.path.exists(stage.save_filename):
try:
stage.fetch()
except fs.FetchError:
continue
tty.msg('Found key %s' % link)
if install:
if trust:
Gpg.trust(stage.save_filename)
tty.msg('Added this key to trusted keys.')
else:
tty.msg('Will not add this key to trusted keys.'
'Use -t to install all downloaded keys')
def get_specs(allarch=False):
"""
Get spec.yaml's for build caches available on mirror
"""
global _cached_specs
arch = architecture.Arch(architecture.platform(), 'default_os',
'default_target')
if not spack.mirror.MirrorCollection():
tty.debug("No Spack mirrors are currently configured")
return {}
for mirror in spack.mirror.MirrorCollection().values():
fetch_url_build_cache = url_util.join(mirror.fetch_url,
_build_cache_relative_path)
tty.debug('Finding buildcaches at {0}'.format(
url_util.format(fetch_url_build_cache)))
index_url = url_util.join(fetch_url_build_cache, 'index.json')
try:
_, _, file_stream = web_util.read_from_url(index_url,
'application/json')
index_object = codecs.getreader('utf-8')(file_stream).read()
except (URLError, web_util.SpackWebError) as url_err:
tty.error('Failed to read index {0}'.format(index_url))
tty.debug(url_err)
# Continue on to the next mirror
continue
tmpdir = tempfile.mkdtemp()
index_file_path = os.path.join(tmpdir, 'index.json')
with open(index_file_path, 'w') as fd:
fd.write(index_object)
db_root_dir = os.path.join(tmpdir, 'db_root')
db = spack_db.Database(None,
db_dir=db_root_dir,
enable_transaction_locking=False)
db._read_from_file(index_file_path)
spec_list = db.query_local(installed=False)
for indexed_spec in spec_list:
spec_arch = architecture.arch_for_spec(indexed_spec.architecture)
if (allarch is True or spec_arch == arch):
_cached_specs.add(indexed_spec)
return _cached_specs
def download_tarball(spec):
"""
Download binary tarball for given package into stage area
Return True if successful
"""
if not spack.mirror.MirrorCollection():
tty.die("Please add a spack mirror to allow " +
"download of pre-compiled packages.")
tarball = tarball_path_name(spec, '.spack')
for mirror in spack.mirror.MirrorCollection().values():
url = url_util.join(mirror.fetch_url, _build_cache_relative_path,
tarball)
# stage the tarball into standard place
stage = Stage(url, name="build_cache", keep=True)
stage.create()
try:
stage.fetch()
return stage.save_filename
except fs.FetchError:
continue
return None
def generate_package_index(cache_prefix):
"""Create the build cache index page.
Creates (or replaces) the "index.html" page at the location given in
cache_prefix. This page contains a link for each binary package (*.yaml)
and signing key (*.key) under cache_prefix.
"""
tmpdir = tempfile.mkdtemp()
try:
index_html_path = os.path.join(tmpdir, 'index.html')
file_list = (entry for entry in web_util.list_url(cache_prefix)
if (entry.endswith('.yaml') or entry.endswith('.key')))
with open(index_html_path, 'w') as f:
f.write(
BUILD_CACHE_INDEX_TEMPLATE.format(
title='Spack Package Index',
path_list='\n'.join(
BUILD_CACHE_INDEX_ENTRY_TEMPLATE.format(path=path)
for path in file_list)))
web_util.push_to_url(index_html_path,
url_util.join(cache_prefix, 'index.html'),
keep_original=False,
extra_args={'ContentType': 'text/html'})
finally:
shutil.rmtree(tmpdir)
def generate_key_index(key_prefix, tmpdir=None):
"""Create the key index page.
Creates (or replaces) the "index.json" page at the location given in
key_prefix. This page contains an entry for each key (.pub) under
key_prefix.
"""
tty.debug(' '.join(('Retrieving key.pub files from',
url_util.format(key_prefix), 'to build key index')))
fingerprints = (entry[:-4]
for entry in web_util.list_url(key_prefix, recursive=False)
if entry.endswith('.pub'))
keys_local = url_util.local_file_path(key_prefix)
if keys_local:
target = os.path.join(keys_local, 'index.json')
else:
target = os.path.join(tmpdir, 'index.json')
index = {
'keys':
dict((fingerprint, {}) for fingerprint in sorted(set(fingerprints)))
}
with open(target, 'w') as f:
sjson.dump(index, f)
if not keys_local:
web_util.push_to_url(target,
url_util.join(key_prefix, 'index.json'),
keep_original=False,
extra_args={'ContentType': 'application/json'})
def generate_package_index(cache_prefix):
"""Create the build cache index page.
Creates (or replaces) the "index.json" page at the location given in
cache_prefix. This page contains a link for each binary package (*.yaml)
and public key (*.key) under cache_prefix.
"""
tmpdir = tempfile.mkdtemp()
db_root_dir = os.path.join(tmpdir, 'db_root')
db = spack_db.Database(None, db_dir=db_root_dir,
enable_transaction_locking=False,
record_fields=['spec', 'ref_count'])
file_list = (
entry
for entry in web_util.list_url(cache_prefix)
if entry.endswith('.yaml'))
tty.debug('Retrieving spec.yaml files from {0} to build index'.format(
cache_prefix))
for file_path in file_list:
try:
yaml_url = url_util.join(cache_prefix, file_path)
tty.debug('fetching {0}'.format(yaml_url))
_, _, yaml_file = web_util.read_from_url(yaml_url)
yaml_contents = codecs.getreader('utf-8')(yaml_file).read()
# yaml_obj = syaml.load(yaml_contents)
# s = Spec.from_yaml(yaml_obj)
s = Spec.from_yaml(yaml_contents)
db.add(s, None)
except (URLError, web_util.SpackWebError) as url_err:
tty.error('Error reading spec.yaml: {0}'.format(file_path))
tty.error(url_err)
try:
index_json_path = os.path.join(db_root_dir, 'index.json')
with open(index_json_path, 'w') as f:
db._write_to_file(f)
web_util.push_to_url(
index_json_path,
url_util.join(cache_prefix, 'index.json'),
keep_original=False,
extra_args={'ContentType': 'application/json'})
finally:
shutil.rmtree(tmpdir)
def buildcache_update_index(args):
"""Update a buildcache index."""
outdir = '.'
if args.mirror_url:
outdir = args.mirror_url
mirror = spack.mirror.MirrorCollection().lookup(outdir)
outdir = url_util.format(mirror.push_url)
bindist.generate_package_index(
url_util.join(outdir, bindist.build_cache_relative_path()))
if args.keys:
keys_url = url_util.join(outdir, bindist.build_cache_relative_path(),
bindist.build_cache_keys_relative_path())
bindist.generate_key_index(keys_url)
def get_specs(force=False, allarch=False):
"""
Get spec.yaml's for build caches available on mirror
"""
arch = architecture.Arch(architecture.platform(), 'default_os',
'default_target')
arch_pattern = ('([^-]*-[^-]*-[^-]*)')
if not allarch:
arch_pattern = '(%s-%s-[^-]*)' % (arch.platform, arch.os)
regex_pattern = '%s(.*)(spec.yaml$)' % (arch_pattern)
arch_re = re.compile(regex_pattern)
if not spack.mirror.MirrorCollection():
tty.debug("No Spack mirrors are currently configured")
return {}
urls = set()
for mirror in spack.mirror.MirrorCollection().values():
fetch_url_build_cache = url_util.join(mirror.fetch_url,
_build_cache_relative_path)
mirror_dir = url_util.local_file_path(fetch_url_build_cache)
if mirror_dir:
tty.msg("Finding buildcaches in %s" % mirror_dir)
if os.path.exists(mirror_dir):
files = os.listdir(mirror_dir)
for file in files:
m = arch_re.search(file)
if m:
link = url_util.join(fetch_url_build_cache, file)
urls.add(link)
else:
tty.msg("Finding buildcaches at %s" %
url_util.format(fetch_url_build_cache))
p, links = web_util.spider(
url_util.join(fetch_url_build_cache, 'index.html'))
for link in links:
m = arch_re.search(link)
if m:
urls.add(link)
return try_download_specs(urls=urls, force=force)
def s3_open(self, req):
orig_url = req.get_full_url()
from botocore.exceptions import ClientError
try:
url, headers, stream = _s3_open(orig_url)
return urllib_response.addinfourl(stream, headers, url)
except ClientError as err:
# if no such [KEY], but [KEY]/index.html exists,
# return that, instead.
if err.response['Error']['Code'] == 'NoSuchKey':
try:
_, headers, stream = _s3_open(
url_util.join(orig_url, 'index.html'))
return urllib_response.addinfourl(
stream, headers, orig_url)
except ClientError as err2:
if err.response['Error']['Code'] == 'NoSuchKey':
# raise original error
raise urllib_error.URLError(err)
raise urllib_error.URLError(err2)
raise urllib_error.URLError(err)
def fetch(self, mirror_only=False):
"""Downloads an archive or checks out code from a repository."""
fetchers = []
if not mirror_only:
fetchers.append(self.default_fetcher)
# TODO: move mirror logic out of here and clean it up!
# TODO: Or @alalazo may have some ideas about how to use a
# TODO: CompositeFetchStrategy here.
self.skip_checksum_for_mirror = True
if self.mirror_paths:
# Join URLs of mirror roots with mirror paths. Because
# urljoin() will strip everything past the final '/' in
# the root, so we add a '/' if it is not present.
urls = []
for mirror in spack.mirror.MirrorCollection().values():
for rel_path in self.mirror_paths:
urls.append(url_util.join(mirror.fetch_url, rel_path))
# If this archive is normally fetched from a tarball URL,
# then use the same digest. `spack mirror` ensures that
# the checksum will be the same.
digest = None
expand = True
extension = None
if isinstance(self.default_fetcher, fs.URLFetchStrategy):
digest = self.default_fetcher.digest
expand = self.default_fetcher.expand_archive
extension = self.default_fetcher.extension
# Have to skip the checksum for things archived from
# repositories. How can this be made safer?
self.skip_checksum_for_mirror = not bool(digest)
# Add URL strategies for all the mirrors with the digest
for url in urls:
fetchers.insert(
0,
fs.from_url_scheme(url,
digest,
expand=expand,
extension=extension))
if self.default_fetcher.cachable:
for rel_path in reversed(list(self.mirror_paths)):
cache_fetcher = spack.caches.fetch_cache.fetcher(
rel_path, digest, expand=expand, extension=extension)
fetchers.insert(0, cache_fetcher)
def generate_fetchers():
for fetcher in fetchers:
yield fetcher
# The search function may be expensive, so wait until now to
# call it so the user can stop if a prior fetcher succeeded
if self.search_fn and not mirror_only:
dynamic_fetchers = self.search_fn()
for fetcher in dynamic_fetchers:
yield fetcher
for fetcher in generate_fetchers():
try:
fetcher.set_stage(self)
self.fetcher = fetcher
self.fetcher.fetch()
break
except spack.fetch_strategy.NoCacheError:
# Don't bother reporting when something is not cached.
continue
except spack.error.SpackError as e:
tty.msg("Fetching from %s failed." % fetcher)
tty.debug(e)
continue
else:
err_msg = "All fetchers failed for %s" % self.name
self.fetcher = self.default_fetcher
raise fs.FetchError(err_msg, None)
def _spider(url, collect_nested):
"""Fetches URL and any pages it links to.
Prints out a warning only if the root can't be fetched; it ignores
errors with pages that the root links to.
Args:
url (str): url being fetched and searched for links
collect_nested (bool): whether we want to collect arguments
for nested spidering on the links found in this url
Returns:
A tuple of:
- pages: dict of pages visited (URL) mapped to their full text.
- links: set of links encountered while visiting the pages.
- spider_args: argument for subsequent call to spider
"""
pages = {} # dict from page URL -> text content.
links = set() # set of all links seen on visited pages.
subcalls = []
try:
response_url, _, response = read_from_url(url, 'text/html')
if not response_url or not response:
return pages, links, subcalls
page = codecs.getreader('utf-8')(response).read()
pages[response_url] = page
# Parse out the links in the page
link_parser = LinkParser()
link_parser.feed(page)
while link_parser.links:
raw_link = link_parser.links.pop()
abs_link = url_util.join(response_url,
raw_link.strip(),
resolve_href=True)
links.add(abs_link)
# Skip stuff that looks like an archive
if any(raw_link.endswith(s) for s in ALLOWED_ARCHIVE_TYPES):
continue
# Skip already-visited links
if abs_link in _visited:
continue
# If we're not at max depth, follow links.
if collect_nested:
subcalls.append((abs_link, ))
_visited.add(abs_link)
except URLError as e:
tty.debug(str(e))
if hasattr(e, 'reason') and isinstance(e.reason, ssl.SSLError):
tty.warn("Spack was unable to fetch url list due to a "
"certificate verification problem. You can try "
"running spack -k, which will not check SSL "
"certificates. Use this at your own risk.")
except HTMLParseError as e:
# This error indicates that Python's HTML parser sucks.
msg = "Got an error parsing HTML."
# Pre-2.7.3 Pythons in particular have rather prickly HTML parsing.
if sys.version_info[:3] < (2, 7, 3):
msg += " Use Python 2.7.3 or newer for better HTML parsing."
tty.warn(msg, url, "HTMLParseError: " + str(e))
except Exception as e:
# Other types of errors are completely ignored,
# except in debug mode
tty.debug("Error in _spider: %s:%s" % (type(e), str(e)),
traceback.format_exc())
finally:
tty.debug("SPIDER: [url={0}]".format(url))
return pages, links, subcalls
def fetch(self, mirror_only=False, err_msg=None):
"""Retrieves the code or archive
Args:
mirror_only (bool): only fetch from a mirror
err_msg (str or None): the error message to display if all fetchers
fail or ``None`` for the default fetch failure message
"""
fetchers = []
if not mirror_only:
fetchers.append(self.default_fetcher)
# TODO: move mirror logic out of here and clean it up!
# TODO: Or @alalazo may have some ideas about how to use a
# TODO: CompositeFetchStrategy here.
self.skip_checksum_for_mirror = True
if self.mirror_paths:
# Join URLs of mirror roots with mirror paths. Because
# urljoin() will strip everything past the final '/' in
# the root, so we add a '/' if it is not present.
mirror_urls = {}
for mirror in spack.mirror.MirrorCollection().values():
for rel_path in self.mirror_paths:
mirror_url = url_util.join(mirror.fetch_url, rel_path)
mirror_urls[mirror_url] = {}
if mirror.get_access_pair("fetch") or \
mirror.get_access_token("fetch") or \
mirror.get_profile("fetch"):
mirror_urls[mirror_url] = {
"access_token": mirror.get_access_token("fetch"),
"access_pair": mirror.get_access_pair("fetch"),
"access_profile": mirror.get_profile("fetch"),
"endpoint_url": mirror.get_endpoint_url("fetch")
}
# If this archive is normally fetched from a tarball URL,
# then use the same digest. `spack mirror` ensures that
# the checksum will be the same.
digest = None
expand = True
extension = None
if isinstance(self.default_fetcher, fs.URLFetchStrategy):
digest = self.default_fetcher.digest
expand = self.default_fetcher.expand_archive
extension = self.default_fetcher.extension
# Have to skip the checksum for things archived from
# repositories. How can this be made safer?
self.skip_checksum_for_mirror = not bool(digest)
# Add URL strategies for all the mirrors with the digest
# Insert fetchers in the order that the URLs are provided.
for url in reversed(list(mirror_urls.keys())):
fetchers.insert(
0,
fs.from_url_scheme(url,
digest,
expand=expand,
extension=extension,
connection=mirror_urls[url]))
if self.default_fetcher.cachable:
for rel_path in reversed(list(self.mirror_paths)):
cache_fetcher = spack.caches.fetch_cache.fetcher(
rel_path, digest, expand=expand, extension=extension)
fetchers.insert(0, cache_fetcher)
def generate_fetchers():
for fetcher in fetchers:
yield fetcher
# The search function may be expensive, so wait until now to
# call it so the user can stop if a prior fetcher succeeded
if self.search_fn and not mirror_only:
dynamic_fetchers = self.search_fn()
for fetcher in dynamic_fetchers:
yield fetcher
def print_errors(errors):
for msg in errors:
tty.debug(msg)
errors = []
for fetcher in generate_fetchers():
try:
fetcher.stage = self
self.fetcher = fetcher
self.fetcher.fetch()
break
except spack.fetch_strategy.NoCacheError:
# Don't bother reporting when something is not cached.
continue
except spack.error.SpackError as e:
errors.append('Fetching from {0} failed.'.format(fetcher))
tty.debug(e)
continue
else:
print_errors(errors)
self.fetcher = self.default_fetcher
default_msg = 'All fetchers failed for {0}'.format(self.name)
raise fs.FetchError(err_msg or default_msg, None)
print_errors(errors)
def _spider(url, visited, root, depth, max_depth, raise_on_error):
"""Fetches URL and any pages it links to up to max_depth.
depth should initially be zero, and max_depth is the max depth of
links to follow from the root.
Prints out a warning only if the root can't be fetched; it ignores
errors with pages that the root links to.
Returns a tuple of:
- pages: dict of pages visited (URL) mapped to their full text.
- links: set of links encountered while visiting the pages.
"""
pages = {} # dict from page URL -> text content.
links = set() # set of all links seen on visited pages.
try:
response_url, _, response = read_from_url(url, 'text/html')
if not response_url or not response:
return pages, links
page = codecs.getreader('utf-8')(response).read()
pages[response_url] = page
# Parse out the links in the page
link_parser = LinkParser()
subcalls = []
link_parser.feed(page)
while link_parser.links:
raw_link = link_parser.links.pop()
abs_link = url_util.join(response_url,
raw_link.strip(),
resolve_href=True)
links.add(abs_link)
# Skip stuff that looks like an archive
if any(raw_link.endswith(suf) for suf in ALLOWED_ARCHIVE_TYPES):
continue
# Skip things outside the root directory
if not abs_link.startswith(root):
continue
# Skip already-visited links
if abs_link in visited:
continue
# If we're not at max depth, follow links.
if depth < max_depth:
subcalls.append((abs_link, visited, root, depth + 1, max_depth,
raise_on_error))
visited.add(abs_link)
if subcalls:
pool = NonDaemonPool(processes=len(subcalls))
try:
results = pool.map(_spider_wrapper, subcalls)
for sub_pages, sub_links in results:
pages.update(sub_pages)
links.update(sub_links)
finally:
pool.terminate()
pool.join()
except URLError as e:
tty.debug(e)
if hasattr(e, 'reason') and isinstance(e.reason, ssl.SSLError):
tty.warn("Spack was unable to fetch url list due to a certificate "
"verification problem. You can try running spack -k, "
"which will not check SSL certificates. Use this at your "
"own risk.")
if raise_on_error:
raise NoNetworkConnectionError(str(e), url)
except HTMLParseError as e:
# This error indicates that Python's HTML parser sucks.
msg = "Got an error parsing HTML."
# Pre-2.7.3 Pythons in particular have rather prickly HTML parsing.
if sys.version_info[:3] < (2, 7, 3):
msg += " Use Python 2.7.3 or newer for better HTML parsing."
tty.warn(msg, url, "HTMLParseError: " + str(e))
except Exception as e:
# Other types of errors are completely ignored, except in debug mode.
tty.debug("Error in _spider: %s:%s" % (type(e), e),
traceback.format_exc())
return pages, links
def build_tarball(spec,
outdir,
force=False,
rel=False,
unsigned=False,
allow_root=False,
key=None,
regenerate_index=False):
"""
Build a tarball from given spec and put it into the directory structure
used at the mirror (following <tarball_directory_name>).
"""
if not spec.concrete:
raise ValueError('spec must be concrete to build tarball')
# set up some paths
tmpdir = tempfile.mkdtemp()
cache_prefix = build_cache_prefix(tmpdir)
tarfile_name = tarball_name(spec, '.tar.gz')
tarfile_dir = os.path.join(cache_prefix, tarball_directory_name(spec))
tarfile_path = os.path.join(tarfile_dir, tarfile_name)
spackfile_path = os.path.join(cache_prefix,
tarball_path_name(spec, '.spack'))
remote_spackfile_path = url_util.join(
outdir, os.path.relpath(spackfile_path, tmpdir))
mkdirp(tarfile_dir)
if web_util.url_exists(remote_spackfile_path):
if force:
web_util.remove_url(remote_spackfile_path)
else:
raise NoOverwriteException(url_util.format(remote_spackfile_path))
# need to copy the spec file so the build cache can be downloaded
# without concretizing with the current spack packages
# and preferences
spec_file = os.path.join(spec.prefix, ".spack", "spec.yaml")
specfile_name = tarball_name(spec, '.spec.yaml')
specfile_path = os.path.realpath(os.path.join(cache_prefix, specfile_name))
remote_specfile_path = url_util.join(
outdir, os.path.relpath(specfile_path, os.path.realpath(tmpdir)))
if web_util.url_exists(remote_specfile_path):
if force:
web_util.remove_url(remote_specfile_path)
else:
raise NoOverwriteException(url_util.format(remote_specfile_path))
# make a copy of the install directory to work with
workdir = os.path.join(tempfile.mkdtemp(), os.path.basename(spec.prefix))
install_tree(spec.prefix, workdir, symlinks=True)
# create info for later relocation and create tar
write_buildinfo_file(spec.prefix, workdir, rel=rel)
# optionally make the paths in the binaries relative to each other
# in the spack install tree before creating tarball
if rel:
try:
make_package_relative(workdir, spec, allow_root)
except Exception as e:
shutil.rmtree(workdir)
shutil.rmtree(tarfile_dir)
shutil.rmtree(tmpdir)
tty.die(e)
else:
try:
make_package_placeholder(workdir, spec, allow_root)
except Exception as e:
shutil.rmtree(workdir)
shutil.rmtree(tarfile_dir)
shutil.rmtree(tmpdir)
tty.die(e)
# create compressed tarball of the install prefix
with closing(tarfile.open(tarfile_path, 'w:gz')) as tar:
tar.add(name='%s' % workdir,
arcname='%s' % os.path.basename(spec.prefix))
# remove copy of install directory
shutil.rmtree(workdir)
# get the sha256 checksum of the tarball
checksum = checksum_tarball(tarfile_path)
# add sha256 checksum to spec.yaml
with open(spec_file, 'r') as inputfile:
content = inputfile.read()
spec_dict = yaml.load(content)
bchecksum = {}
bchecksum['hash_algorithm'] = 'sha256'
bchecksum['hash'] = checksum
spec_dict['binary_cache_checksum'] = bchecksum
# Add original install prefix relative to layout root to spec.yaml.
# This will be used to determine is the directory layout has changed.
buildinfo = {}
buildinfo['relative_prefix'] = os.path.relpath(spec.prefix,
spack.store.layout.root)
spec_dict['buildinfo'] = buildinfo
spec_dict['full_hash'] = spec.full_hash()
tty.debug('The full_hash ({0}) of {1} will be written into {2}'.format(
spec_dict['full_hash'], spec.name,
url_util.format(remote_specfile_path)))
tty.debug(spec.tree())
with open(specfile_path, 'w') as outfile:
outfile.write(syaml.dump(spec_dict))
# sign the tarball and spec file with gpg
if not unsigned:
sign_tarball(key, force, specfile_path)
# put tarball, spec and signature files in .spack archive
with closing(tarfile.open(spackfile_path, 'w')) as tar:
tar.add(name='%s' % tarfile_path, arcname='%s' % tarfile_name)
tar.add(name='%s' % specfile_path, arcname='%s' % specfile_name)
if not unsigned:
tar.add(name='%s.asc' % specfile_path,
arcname='%s.asc' % specfile_name)
# cleanup file moved to archive
os.remove(tarfile_path)
if not unsigned:
os.remove('%s.asc' % specfile_path)
web_util.push_to_url(spackfile_path,
remote_spackfile_path,
keep_original=False)
web_util.push_to_url(specfile_path,
remote_specfile_path,
keep_original=False)
try:
# create an index.html for the build_cache directory so specs can be
# found
if regenerate_index:
generate_package_index(
url_util.join(outdir, os.path.relpath(cache_prefix, tmpdir)))
finally:
shutil.rmtree(tmpdir)
return None
def test_url_join_local_paths():
# Resolve local link against page URL
# wrong:
assert (url_util.join('s3://bucket/index.html',
'../other-bucket/document.txt') ==
's3://bucket/other-bucket/document.txt')
# correct - need to specify resolve_href=True:
assert (url_util.join(
's3://bucket/index.html',
'../other-bucket/document.txt',
resolve_href=True) == 's3://other-bucket/document.txt')
# same as above: make sure several components are joined together correctly
assert (url_util.join(
# with resolve_href=True, first arg is the base url; can not be
# broken up
's3://bucket/index.html',
# with resolve_href=True, remaining arguments are the components of
# the local href that needs to be resolved
'..',
'other-bucket',
'document.txt',
resolve_href=True) == 's3://other-bucket/document.txt')
# Append local path components to prefix URL
# wrong:
assert (url_util.join(
'https://mirror.spack.io/build_cache', 'my-package',
resolve_href=True) == 'https://mirror.spack.io/my-package')
# correct - Need to specify resolve_href=False:
assert (url_util.join(
'https://mirror.spack.io/build_cache',
'my-package',
resolve_href=False) == 'https://mirror.spack.io/build_cache/my-package'
)
# same as above; make sure resolve_href=False is default
assert (url_util.join(
'https://mirror.spack.io/build_cache',
'my-package') == 'https://mirror.spack.io/build_cache/my-package')
# same as above: make sure several components are joined together correctly
assert (url_util.join(
# with resolve_href=False, first arg is just a prefix. No
# resolution is done. So, there should be no difference between
# join('/a/b/c', 'd/e'),
# join('/a/b', 'c', 'd/e'),
# join('/a', 'b/c', 'd', 'e'), etc.
'https://mirror.spack.io',
'build_cache',
'my-package') == 'https://mirror.spack.io/build_cache/my-package')
# file:// URL path components are *NOT* canonicalized
spack_root = spack.paths.spack_root
join_result = url_util.join('/a/b/c', '$spack')
assert (join_result == 'file:///a/b/c/$spack') # not canonicalized
format_result = url_util.format(join_result)
# canoncalize by hand
expected = url_util.format(
os.path.abspath(os.path.join('/', 'a', 'b', 'c', '.' + spack_root)))
assert (format_result == expected)
# see test_url_join_absolute_paths() for more on absolute path components
join_result = url_util.join('/a/b/c', '/$spack')
assert (join_result == 'file:///$spack') # not canonicalized
format_result = url_util.format(join_result)
expected = url_util.format(spack_root)
assert (format_result == expected)
# For s3:// URLs, the "netloc" (bucket) is considered part of the path.
# Make sure join() can cross bucket boundaries in this case.
args = ['s3://bucket/a/b', 'new-bucket', 'c']
assert (url_util.join(*args) == 's3://bucket/a/b/new-bucket/c')
args.insert(1, '..')
assert (url_util.join(*args) == 's3://bucket/a/new-bucket/c')
args.insert(1, '..')
assert (url_util.join(*args) == 's3://bucket/new-bucket/c')
# new-bucket is now the "netloc" (bucket name)
args.insert(1, '..')
assert (url_util.join(*args) == 's3://new-bucket/c')
def test_url_join_absolute_paths():
# Handling absolute path components is a little tricky. To this end, we
# distinguish "absolute path components", from the more-familiar concept of
# "absolute paths" as they are understood for local filesystem paths.
#
# - All absolute paths are absolute path components. Joining a URL with
# these components has the effect of completely replacing the path of the
# URL with the absolute path. These components do not specify a URL
# scheme, so the scheme of the URL procuced when joining them depend on
# those provided by components that came before it (file:// assumed if no
# such scheme is provided).
# For eaxmple:
p = '/path/to/resource'
# ...is an absolute path
# http:// URL
assert (url_util.join('http://example.com/a/b/c',
p) == 'http://example.com/path/to/resource')
# s3:// URL
# also notice how the netloc is treated as part of the path for s3:// URLs
assert (url_util.join('s3://example.com/a/b/c',
p) == 's3://path/to/resource')
# - URL components that specify a scheme are always absolute path
# components. Joining a base URL with these components effectively
# discards the base URL and "resets" the joining logic starting at the
# component in question and using it as the new base URL.
# For eaxmple:
p = 'http://example.com/path/to'
# ...is an http:// URL
join_result = url_util.join(p, 'resource')
assert (join_result == 'http://example.com/path/to/resource')
# works as if everything before the http:// URL was left out
assert (url_util.join('literally', 'does', 'not', 'matter', p,
'resource') == join_result)
# It's important to keep in mind that this logic applies even if the
# component's path is not an absolute path!
# For eaxmple:
p = './d'
# ...is *NOT* an absolute path
# ...is also *NOT* an absolute path component
u = 'file://./d'
# ...is a URL
# The path of this URL is *NOT* an absolute path
# HOWEVER, the URL, itself, *is* an absolute path component
# (We just need...
cwd = os.getcwd()
# ...to work out what resource it points to)
# So, even though parse() assumes "file://" URL, the scheme is still
# significant in URL path components passed to join(), even if the base
# is a file:// URL.
path_join_result = 'file:///a/b/c/d'
assert (url_util.join('/a/b/c', p) == path_join_result)
assert (url_util.join('file:///a/b/c', p) == path_join_result)
url_join_result = 'file://{CWD}/d'.format(CWD=cwd)
assert (url_util.join('/a/b/c', u) == url_join_result)
assert (url_util.join('file:///a/b/c', u) == url_join_result)
# Finally, resolve_href should have no effect for how absolute path
# components are handled because local hrefs can not be absolute path
# components.
args = [
's3://does', 'not', 'matter', 'http://example.com', 'also', 'does',
'not', 'matter', '/path'
]
expected = 'http://example.com/path'
assert (url_util.join(*args, resolve_href=True) == expected)
assert (url_util.join(*args, resolve_href=False) == expected)
# resolve_href only matters for the local path components at the end of the
# argument list.
args[-1] = '/path/to/page'
args.extend(('..', '..', 'resource'))
assert (url_util.join(*args,
resolve_href=True) == 'http://example.com/resource')
assert (url_util.join(
*args, resolve_href=False) == 'http://example.com/path/resource')
def buildcache_sync(args):
""" Syncs binaries (and associated metadata) from one mirror to another.
Requires an active environment in order to know which specs to sync.
Args:
src (str): Source mirror URL
dest (str): Destination mirror URL
"""
# Figure out the source mirror
source_location = None
if args.src_directory:
source_location = args.src_directory
scheme = url_util.parse(source_location, scheme='<missing>').scheme
if scheme != '<missing>':
raise ValueError(
'"--src-directory" expected a local path; got a URL, instead')
# Ensure that the mirror lookup does not mistake this for named mirror
source_location = 'file://' + source_location
elif args.src_mirror_name:
source_location = args.src_mirror_name
result = spack.mirror.MirrorCollection().lookup(source_location)
if result.name == "<unnamed>":
raise ValueError('no configured mirror named "{name}"'.format(
name=source_location))
elif args.src_mirror_url:
source_location = args.src_mirror_url
scheme = url_util.parse(source_location, scheme='<missing>').scheme
if scheme == '<missing>':
raise ValueError(
'"{url}" is not a valid URL'.format(url=source_location))
src_mirror = spack.mirror.MirrorCollection().lookup(source_location)
src_mirror_url = url_util.format(src_mirror.fetch_url)
# Figure out the destination mirror
dest_location = None
if args.dest_directory:
dest_location = args.dest_directory
scheme = url_util.parse(dest_location, scheme='<missing>').scheme
if scheme != '<missing>':
raise ValueError(
'"--dest-directory" expected a local path; got a URL, instead')
# Ensure that the mirror lookup does not mistake this for named mirror
dest_location = 'file://' + dest_location
elif args.dest_mirror_name:
dest_location = args.dest_mirror_name
result = spack.mirror.MirrorCollection().lookup(dest_location)
if result.name == "<unnamed>":
raise ValueError('no configured mirror named "{name}"'.format(
name=dest_location))
elif args.dest_mirror_url:
dest_location = args.dest_mirror_url
scheme = url_util.parse(dest_location, scheme='<missing>').scheme
if scheme == '<missing>':
raise ValueError(
'"{url}" is not a valid URL'.format(url=dest_location))
dest_mirror = spack.mirror.MirrorCollection().lookup(dest_location)
dest_mirror_url = url_util.format(dest_mirror.fetch_url)
# Get the active environment
env = spack.cmd.require_active_env(cmd_name='buildcache sync')
tty.msg('Syncing environment buildcache files from {0} to {1}'.format(
src_mirror_url, dest_mirror_url))
build_cache_dir = bindist.build_cache_relative_path()
buildcache_rel_paths = []
tty.debug('Syncing the following specs:')
for s in env.all_specs():
tty.debug(' {0}{1}: {2}'.format('* ' if s in env.roots() else ' ',
s.name, s.dag_hash()))
buildcache_rel_paths.extend([
os.path.join(build_cache_dir,
bindist.tarball_path_name(s, '.spack')),
os.path.join(build_cache_dir,
bindist.tarball_name(s, '.spec.yaml')),
os.path.join(build_cache_dir, bindist.tarball_name(s, '.cdashid'))
])
tmpdir = tempfile.mkdtemp()
try:
for rel_path in buildcache_rel_paths:
src_url = url_util.join(src_mirror_url, rel_path)
local_path = os.path.join(tmpdir, rel_path)
dest_url = url_util.join(dest_mirror_url, rel_path)
tty.debug('Copying {0} to {1} via {2}'.format(
src_url, dest_url, local_path))
stage = Stage(src_url,
name="temporary_file",
path=os.path.dirname(local_path),
keep=True)
try:
stage.create()
stage.fetch()
web_util.push_to_url(local_path, dest_url, keep_original=True)
except fs.FetchError as e:
tty.debug(
'spack buildcache unable to sync {0}'.format(rel_path))
tty.debug(e)
finally:
stage.destroy()
finally:
shutil.rmtree(tmpdir)
def ci_rebuild(args):
"""This command represents a gitlab-ci job, corresponding to a single
release spec. As such it must first decide whether or not the spec it
has been assigned to build is up to date on the remote binary mirror.
If it is not (i.e. the full_hash of the spec as computed locally does
not match the one stored in the metadata on the mirror), this script
will build the package, create a binary cache for it, and then push all
related files to the remote binary mirror. This script also
communicates with a remote CDash instance to share status on the package
build process.
The spec to be built by this job is represented by essentially two
pieces of information: 1) a root spec (possibly already concrete, but
maybe still needing to be concretized) and 2) a package name used to
index that root spec (once the root is, for certain, concrete)."""
env = ev.get_env(args, 'ci rebuild', required=True)
yaml_root = ev.config_dict(env.yaml)
# The following environment variables should defined in the CI
# infrastructre (or some other external source) in the case that the
# remote mirror is an S3 bucket. The AWS keys are used to upload
# buildcache entries to S3 using the boto3 api.
#
# AWS_ACCESS_KEY_ID
# AWS_SECRET_ACCESS_KEY
# S3_ENDPOINT_URL (only needed for non-AWS S3 implementations)
#
# If present, we will import the SPACK_SIGNING_KEY using the
# "spack gpg trust" command, so it can be used both for verifying
# dependency buildcache entries and signing the buildcache entry we create
# for our target pkg.
#
# SPACK_SIGNING_KEY
ci_artifact_dir = get_env_var('CI_PROJECT_DIR')
ci_pipeline_id = get_env_var('CI_PIPELINE_ID')
signing_key = get_env_var('SPACK_SIGNING_KEY')
root_spec = get_env_var('SPACK_ROOT_SPEC')
job_spec_pkg_name = get_env_var('SPACK_JOB_SPEC_PKG_NAME')
compiler_action = get_env_var('SPACK_COMPILER_ACTION')
cdash_build_name = get_env_var('SPACK_CDASH_BUILD_NAME')
related_builds = get_env_var('SPACK_RELATED_BUILDS_CDASH')
pr_env_var = get_env_var('SPACK_IS_PR_PIPELINE')
pr_mirror_url = get_env_var('SPACK_PR_MIRROR_URL')
gitlab_ci = None
if 'gitlab-ci' in yaml_root:
gitlab_ci = yaml_root['gitlab-ci']
if not gitlab_ci:
tty.die('spack ci rebuild requires an env containing gitlab-ci cfg')
enable_cdash = False
if 'cdash' in yaml_root:
enable_cdash = True
ci_cdash = yaml_root['cdash']
job_spec_buildgroup = ci_cdash['build-group']
cdash_base_url = ci_cdash['url']
cdash_project = ci_cdash['project']
proj_enc = urlencode({'project': cdash_project})
eq_idx = proj_enc.find('=') + 1
cdash_project_enc = proj_enc[eq_idx:]
cdash_site = ci_cdash['site']
tty.debug('cdash_base_url = {0}'.format(cdash_base_url))
tty.debug('cdash_project = {0}'.format(cdash_project))
tty.debug('cdash_project_enc = {0}'.format(cdash_project_enc))
tty.debug('cdash_build_name = {0}'.format(cdash_build_name))
tty.debug('cdash_site = {0}'.format(cdash_site))
tty.debug('related_builds = {0}'.format(related_builds))
tty.debug('job_spec_buildgroup = {0}'.format(job_spec_buildgroup))
remote_mirror_url = None
if 'mirrors' in yaml_root:
ci_mirrors = yaml_root['mirrors']
mirror_urls = [url for url in ci_mirrors.values()]
remote_mirror_url = mirror_urls[0]
if not remote_mirror_url:
tty.die('spack ci rebuild requires an env containing a mirror')
tty.debug('ci_artifact_dir = {0}'.format(ci_artifact_dir))
tty.debug('root_spec = {0}'.format(root_spec))
tty.debug('remote_mirror_url = {0}'.format(remote_mirror_url))
tty.debug('job_spec_pkg_name = {0}'.format(job_spec_pkg_name))
tty.debug('compiler_action = {0}'.format(compiler_action))
cdash_report_dir = os.path.join(ci_artifact_dir, 'cdash_report')
temp_dir = os.path.join(ci_artifact_dir, 'jobs_scratch_dir')
job_log_dir = os.path.join(temp_dir, 'logs')
spec_dir = os.path.join(temp_dir, 'specs')
local_mirror_dir = os.path.join(ci_artifact_dir, 'local_mirror')
build_cache_dir = os.path.join(local_mirror_dir, 'build_cache')
spack_is_pr_pipeline = True if pr_env_var == 'True' else False
pipeline_mirror_url = None
temp_storage_url_prefix = None
if 'temporary-storage-url-prefix' in gitlab_ci:
temp_storage_url_prefix = gitlab_ci['temporary-storage-url-prefix']
pipeline_mirror_url = url_util.join(
temp_storage_url_prefix, ci_pipeline_id)
enable_artifacts_mirror = False
if 'enable-artifacts-buildcache' in gitlab_ci:
enable_artifacts_mirror = gitlab_ci['enable-artifacts-buildcache']
if (enable_artifacts_mirror or (spack_is_pr_pipeline and
not enable_artifacts_mirror and not temp_storage_url_prefix)):
# If you explicitly enabled the artifacts buildcache feature, or
# if this is a PR pipeline but you did not enable either of the
# per-pipeline temporary storage features, we force the use of
# artifacts buildcache. Otherwise jobs will not have binary
# dependencies from previous stages available since we do not
# allow pushing binaries to the remote mirror during PR pipelines.
enable_artifacts_mirror = True
pipeline_mirror_url = 'file://' + local_mirror_dir
mirror_msg = 'artifact buildcache enabled, mirror url: {0}'.format(
pipeline_mirror_url)
tty.debug(mirror_msg)
# Clean out scratch directory from last stage
if os.path.exists(temp_dir):
shutil.rmtree(temp_dir)
if os.path.exists(cdash_report_dir):
shutil.rmtree(cdash_report_dir)
os.makedirs(job_log_dir)
os.makedirs(spec_dir)
job_spec_yaml_path = os.path.join(
spec_dir, '{0}.yaml'.format(job_spec_pkg_name))
job_log_file = os.path.join(job_log_dir, 'pipeline_log.txt')
cdash_build_id = None
cdash_build_stamp = None
with open(job_log_file, 'w') as log_fd:
os.dup2(log_fd.fileno(), sys.stdout.fileno())
os.dup2(log_fd.fileno(), sys.stderr.fileno())
current_directory = os.getcwd()
tty.debug('Current working directory: {0}, Contents:'.format(
current_directory))
directory_list = os.listdir(current_directory)
for next_entry in directory_list:
tty.debug(' {0}'.format(next_entry))
tty.debug('job concrete spec path: {0}'.format(job_spec_yaml_path))
if signing_key:
spack_ci.import_signing_key(signing_key)
can_sign = spack_ci.can_sign_binaries()
sign_binaries = can_sign and spack_is_pr_pipeline is False
can_verify = spack_ci.can_verify_binaries()
verify_binaries = can_verify and spack_is_pr_pipeline is False
spack_ci.configure_compilers(compiler_action)
spec_map = spack_ci.get_concrete_specs(
root_spec, job_spec_pkg_name, related_builds, compiler_action)
job_spec = spec_map[job_spec_pkg_name]
tty.debug('Here is the concrete spec: {0}'.format(job_spec))
with open(job_spec_yaml_path, 'w') as fd:
fd.write(job_spec.to_yaml(hash=ht.build_hash))
tty.debug('Done writing concrete spec')
# DEBUG
with open(job_spec_yaml_path) as fd:
tty.debug('Wrote spec file, read it back. Contents:')
tty.debug(fd.read())
# DEBUG the root spec
root_spec_yaml_path = os.path.join(spec_dir, 'root.yaml')
with open(root_spec_yaml_path, 'w') as fd:
fd.write(spec_map['root'].to_yaml(hash=ht.build_hash))
# TODO: Refactor the spack install command so it's easier to use from
# python modules. Currently we use "exe.which('spack')" to make it
# easier to install packages from here, but it introduces some
# problems, e.g. if we want the spack command to have access to the
# mirrors we're configuring, then we have to use the "spack" command
# to add the mirrors too, which in turn means that any code here *not*
# using the spack command does *not* have access to the mirrors.
spack_cmd = exe.which('spack')
mirrors_to_check = {
'ci_remote_mirror': remote_mirror_url,
}
def add_mirror(mirror_name, mirror_url):
m_args = ['mirror', 'add', mirror_name, mirror_url]
tty.debug('Adding mirror: spack {0}'.format(m_args))
mirror_add_output = spack_cmd(*m_args)
# Workaround: Adding the mirrors above, using "spack_cmd" makes
# sure they're available later when we use "spack_cmd" to install
# the package. But then we also need to add them to this dict
# below, so they're available in this process (we end up having to
# pass them to "bindist.get_mirrors_for_spec()")
mirrors_to_check[mirror_name] = mirror_url
tty.debug('spack mirror add output: {0}'.format(mirror_add_output))
# Configure mirrors
if pr_mirror_url:
add_mirror('ci_pr_mirror', pr_mirror_url)
if pipeline_mirror_url:
add_mirror(spack_ci.TEMP_STORAGE_MIRROR_NAME, pipeline_mirror_url)
tty.debug('listing spack mirrors:')
spack_cmd('mirror', 'list')
spack_cmd('config', 'blame', 'mirrors')
# Checks all mirrors for a built spec with a matching full hash
matches = bindist.get_mirrors_for_spec(
job_spec, full_hash_match=True, mirrors_to_check=mirrors_to_check,
index_only=False)
if matches:
# Got at full hash match on at least one configured mirror. All
# matches represent the fully up-to-date spec, so should all be
# equivalent. If artifacts mirror is enabled, we just pick one
# of the matches and download the buildcache files from there to
# the artifacts, so they're available to be used by dependent
# jobs in subsequent stages.
tty.debug('No need to rebuild {0}'.format(job_spec_pkg_name))
if enable_artifacts_mirror:
matching_mirror = matches[0]['mirror_url']
tty.debug('Getting {0} buildcache from {1}'.format(
job_spec_pkg_name, matching_mirror))
tty.debug('Downloading to {0}'.format(build_cache_dir))
buildcache.download_buildcache_files(
job_spec, build_cache_dir, True, matching_mirror)
else:
# No full hash match anywhere means we need to rebuild spec
# Build up common install arguments
install_args = [
'-d', '-v', '-k', 'install',
'--keep-stage',
'--require-full-hash-match',
]
if not verify_binaries:
install_args.append('--no-check-signature')
# Add arguments to create + register a new build on CDash (if
# enabled)
if enable_cdash:
tty.debug('Registering build with CDash')
(cdash_build_id,
cdash_build_stamp) = spack_ci.register_cdash_build(
cdash_build_name, cdash_base_url, cdash_project,
cdash_site, job_spec_buildgroup)
cdash_upload_url = '{0}/submit.php?project={1}'.format(
cdash_base_url, cdash_project_enc)
install_args.extend([
'--cdash-upload-url', cdash_upload_url,
'--cdash-build', cdash_build_name,
'--cdash-site', cdash_site,
'--cdash-buildstamp', cdash_build_stamp,
])
install_args.append(job_spec_yaml_path)
tty.debug('Installing {0} from source'.format(job_spec.name))
try:
tty.debug('spack install arguments: {0}'.format(
install_args))
spack_cmd(*install_args)
finally:
spack_ci.copy_stage_logs_to_artifacts(job_spec, job_log_dir)
# Create buildcache on remote mirror, either on pr-specific
# mirror or on mirror defined in spack environment
if spack_is_pr_pipeline:
buildcache_mirror_url = pr_mirror_url
else:
buildcache_mirror_url = remote_mirror_url
try:
spack_ci.push_mirror_contents(
env, job_spec, job_spec_yaml_path, buildcache_mirror_url,
cdash_build_id, sign_binaries)
except Exception as inst:
# If the mirror we're pushing to is on S3 and there's some
# permissions problem, for example, we can't just target
# that exception type here, since users of the
# `spack ci rebuild' may not need or want any dependency
# on boto3. So we use the first non-boto exception type
# in the heirarchy:
# boto3.exceptions.S3UploadFailedError
# boto3.exceptions.Boto3Error
# Exception
# BaseException
# object
err_msg = 'Error msg: {0}'.format(inst)
if 'Access Denied' in err_msg:
tty.msg('Permission problem writing to mirror')
tty.msg(err_msg)
# Create another copy of that buildcache in the per-pipeline
# temporary storage mirror (this is only done if either artifacts
# buildcache is enabled or a temporary storage url prefix is set)
spack_ci.push_mirror_contents(env, job_spec, job_spec_yaml_path,
pipeline_mirror_url, cdash_build_id,
sign_binaries)
# Relate this build to its dependencies on CDash (if enabled)
if enable_cdash:
spack_ci.relate_cdash_builds(
spec_map, cdash_base_url, cdash_build_id, cdash_project,
pipeline_mirror_url or pr_mirror_url or remote_mirror_url)
def ci_rebuild(args):
"""Check a single spec against the remote mirror, and rebuild it from
source if the mirror does not contain the full hash match of the spec
as computed locally. """
env = ev.get_env(args, 'ci rebuild', required=True)
# Make sure the environment is "gitlab-enabled", or else there's nothing
# to do.
yaml_root = ev.config_dict(env.yaml)
gitlab_ci = None
if 'gitlab-ci' in yaml_root:
gitlab_ci = yaml_root['gitlab-ci']
if not gitlab_ci:
tty.die('spack ci rebuild requires an env containing gitlab-ci cfg')
# Grab the environment variables we need. These either come from the
# pipeline generation step ("spack ci generate"), where they were written
# out as variables, or else provided by GitLab itself.
pipeline_artifacts_dir = get_env_var('SPACK_ARTIFACTS_ROOT')
job_log_dir = get_env_var('SPACK_JOB_LOG_DIR')
repro_dir = get_env_var('SPACK_JOB_REPRO_DIR')
local_mirror_dir = get_env_var('SPACK_LOCAL_MIRROR_DIR')
concrete_env_dir = get_env_var('SPACK_CONCRETE_ENV_DIR')
ci_pipeline_id = get_env_var('CI_PIPELINE_ID')
ci_job_name = get_env_var('CI_JOB_NAME')
signing_key = get_env_var('SPACK_SIGNING_KEY')
root_spec = get_env_var('SPACK_ROOT_SPEC')
job_spec_pkg_name = get_env_var('SPACK_JOB_SPEC_PKG_NAME')
compiler_action = get_env_var('SPACK_COMPILER_ACTION')
cdash_build_name = get_env_var('SPACK_CDASH_BUILD_NAME')
related_builds = get_env_var('SPACK_RELATED_BUILDS_CDASH')
spack_pipeline_type = get_env_var('SPACK_PIPELINE_TYPE')
pr_mirror_url = get_env_var('SPACK_PR_MIRROR_URL')
remote_mirror_url = get_env_var('SPACK_REMOTE_MIRROR_URL')
# Construct absolute paths relative to current $CI_PROJECT_DIR
ci_project_dir = get_env_var('CI_PROJECT_DIR')
pipeline_artifacts_dir = os.path.join(ci_project_dir,
pipeline_artifacts_dir)
job_log_dir = os.path.join(ci_project_dir, job_log_dir)
repro_dir = os.path.join(ci_project_dir, repro_dir)
local_mirror_dir = os.path.join(ci_project_dir, local_mirror_dir)
concrete_env_dir = os.path.join(ci_project_dir, concrete_env_dir)
# Debug print some of the key environment variables we should have received
tty.debug('pipeline_artifacts_dir = {0}'.format(pipeline_artifacts_dir))
tty.debug('root_spec = {0}'.format(root_spec))
tty.debug('remote_mirror_url = {0}'.format(remote_mirror_url))
tty.debug('job_spec_pkg_name = {0}'.format(job_spec_pkg_name))
tty.debug('compiler_action = {0}'.format(compiler_action))
# Query the environment manifest to find out whether we're reporting to a
# CDash instance, and if so, gather some information from the manifest to
# support that task.
enable_cdash = False
if 'cdash' in yaml_root:
enable_cdash = True
ci_cdash = yaml_root['cdash']
job_spec_buildgroup = ci_cdash['build-group']
cdash_base_url = ci_cdash['url']
cdash_project = ci_cdash['project']
proj_enc = urlencode({'project': cdash_project})
eq_idx = proj_enc.find('=') + 1
cdash_project_enc = proj_enc[eq_idx:]
cdash_site = ci_cdash['site']
tty.debug('cdash_base_url = {0}'.format(cdash_base_url))
tty.debug('cdash_project = {0}'.format(cdash_project))
tty.debug('cdash_project_enc = {0}'.format(cdash_project_enc))
tty.debug('cdash_build_name = {0}'.format(cdash_build_name))
tty.debug('cdash_site = {0}'.format(cdash_site))
tty.debug('related_builds = {0}'.format(related_builds))
tty.debug('job_spec_buildgroup = {0}'.format(job_spec_buildgroup))
# Is this a pipeline run on a spack PR or a merge to develop? It might
# be neither, e.g. a pipeline run on some environment repository.
spack_is_pr_pipeline = spack_pipeline_type == 'spack_pull_request'
spack_is_develop_pipeline = spack_pipeline_type == 'spack_protected_branch'
tty.debug('Pipeline type - PR: {0}, develop: {1}'.format(
spack_is_pr_pipeline, spack_is_develop_pipeline))
# Figure out what is our temporary storage mirror: Is it artifacts
# buildcache? Or temporary-storage-url-prefix? In some cases we need to
# force something or pipelines might not have a way to propagate build
# artifacts from upstream to downstream jobs.
pipeline_mirror_url = None
temp_storage_url_prefix = None
if 'temporary-storage-url-prefix' in gitlab_ci:
temp_storage_url_prefix = gitlab_ci['temporary-storage-url-prefix']
pipeline_mirror_url = url_util.join(temp_storage_url_prefix,
ci_pipeline_id)
enable_artifacts_mirror = False
if 'enable-artifacts-buildcache' in gitlab_ci:
enable_artifacts_mirror = gitlab_ci['enable-artifacts-buildcache']
if (enable_artifacts_mirror
or (spack_is_pr_pipeline and not enable_artifacts_mirror
and not temp_storage_url_prefix)):
# If you explicitly enabled the artifacts buildcache feature, or
# if this is a PR pipeline but you did not enable either of the
# per-pipeline temporary storage features, we force the use of
# artifacts buildcache. Otherwise jobs will not have binary
# dependencies from previous stages available since we do not
# allow pushing binaries to the remote mirror during PR pipelines.
enable_artifacts_mirror = True
pipeline_mirror_url = 'file://' + local_mirror_dir
mirror_msg = 'artifact buildcache enabled, mirror url: {0}'.format(
pipeline_mirror_url)
tty.debug(mirror_msg)
# Whatever form of root_spec we got, use it to get a map giving us concrete
# specs for this job and all of its dependencies.
spec_map = spack_ci.get_concrete_specs(env, root_spec, job_spec_pkg_name,
related_builds, compiler_action)
job_spec = spec_map[job_spec_pkg_name]
job_spec_yaml_file = '{0}.yaml'.format(job_spec_pkg_name)
job_spec_yaml_path = os.path.join(repro_dir, job_spec_yaml_file)
# To provide logs, cdash reports, etc for developer download/perusal,
# these things have to be put into artifacts. This means downstream
# jobs that "need" this job will get those artifacts too. So here we
# need to clean out the artifacts we may have got from upstream jobs.
cdash_report_dir = os.path.join(pipeline_artifacts_dir, 'cdash_report')
if os.path.exists(cdash_report_dir):
shutil.rmtree(cdash_report_dir)
if os.path.exists(job_log_dir):
shutil.rmtree(job_log_dir)
if os.path.exists(repro_dir):
shutil.rmtree(repro_dir)
# Now that we removed them if they existed, create the directories we
# need for storing artifacts. The cdash_report directory will be
# created internally if needed.
os.makedirs(job_log_dir)
os.makedirs(repro_dir)
# Copy the concrete environment files to the repro directory so we can
# expose them as artifacts and not conflict with the concrete environment
# files we got as artifacts from the upstream pipeline generation job.
# Try to cast a slightly wider net too, and hopefully get the generated
# pipeline yaml. If we miss it, the user will still be able to go to the
# pipeline generation job and get it from there.
target_dirs = [concrete_env_dir, pipeline_artifacts_dir]
for dir_to_list in target_dirs:
for file_name in os.listdir(dir_to_list):
src_file = os.path.join(dir_to_list, file_name)
if os.path.isfile(src_file):
dst_file = os.path.join(repro_dir, file_name)
shutil.copyfile(src_file, dst_file)
# If signing key was provided via "SPACK_SIGNING_KEY", then try to
# import it.
if signing_key:
spack_ci.import_signing_key(signing_key)
# Depending on the specifics of this job, we might need to turn on the
# "config:install_missing compilers" option (to build this job spec
# with a bootstrapped compiler), or possibly run "spack compiler find"
# (to build a bootstrap compiler or one of its deps in a
# compiler-agnostic way), or maybe do nothing at all (to build a spec
# using a compiler already installed on the target system).
spack_ci.configure_compilers(compiler_action)
# Write this job's spec yaml into the reproduction directory, and it will
# also be used in the generated "spack install" command to install the spec
tty.debug('job concrete spec path: {0}'.format(job_spec_yaml_path))
with open(job_spec_yaml_path, 'w') as fd:
fd.write(job_spec.to_yaml(hash=ht.build_hash))
# Write the concrete root spec yaml into the reproduction directory
root_spec_yaml_path = os.path.join(repro_dir, 'root.yaml')
with open(root_spec_yaml_path, 'w') as fd:
fd.write(spec_map['root'].to_yaml(hash=ht.build_hash))
# Write some other details to aid in reproduction into an artifact
repro_file = os.path.join(repro_dir, 'repro.json')
repro_details = {
'job_name': ci_job_name,
'job_spec_yaml': job_spec_yaml_file,
'root_spec_yaml': 'root.yaml',
'ci_project_dir': ci_project_dir
}
with open(repro_file, 'w') as fd:
fd.write(json.dumps(repro_details))
# Write information about spack into an artifact in the repro dir
spack_info = spack_ci.get_spack_info()
spack_info_file = os.path.join(repro_dir, 'spack_info.txt')
with open(spack_info_file, 'w') as fd:
fd.write('\n{0}\n'.format(spack_info))
# If we decided there should be a temporary storage mechanism, add that
# mirror now so it's used when we check for a full hash match already
# built for this spec.
if pipeline_mirror_url:
spack.mirror.add(spack_ci.TEMP_STORAGE_MIRROR_NAME,
pipeline_mirror_url, cfg.default_modify_scope())
cdash_build_id = None
cdash_build_stamp = None
# Check configured mirrors for a built spec with a matching full hash
matches = bindist.get_mirrors_for_spec(job_spec,
full_hash_match=True,
index_only=False)
if matches:
# Got a full hash match on at least one configured mirror. All
# matches represent the fully up-to-date spec, so should all be
# equivalent. If artifacts mirror is enabled, we just pick one
# of the matches and download the buildcache files from there to
# the artifacts, so they're available to be used by dependent
# jobs in subsequent stages.
tty.msg('No need to rebuild {0}, found full hash match at: '.format(
job_spec_pkg_name))
for match in matches:
tty.msg(' {0}'.format(match['mirror_url']))
if enable_artifacts_mirror:
matching_mirror = matches[0]['mirror_url']
build_cache_dir = os.path.join(local_mirror_dir, 'build_cache')
tty.debug('Getting {0} buildcache from {1}'.format(
job_spec_pkg_name, matching_mirror))
tty.debug('Downloading to {0}'.format(build_cache_dir))
buildcache.download_buildcache_files(job_spec, build_cache_dir,
False, matching_mirror)
# Now we are done and successful
sys.exit(0)
# No full hash match anywhere means we need to rebuild spec
# Start with spack arguments
install_args = [base_arg for base_arg in CI_REBUILD_INSTALL_BASE_ARGS]
config = cfg.get('config')
if not config['verify_ssl']:
install_args.append('-k')
install_args.extend([
'install',
'--keep-stage',
'--require-full-hash-match',
])
can_verify = spack_ci.can_verify_binaries()
verify_binaries = can_verify and spack_is_pr_pipeline is False
if not verify_binaries:
install_args.append('--no-check-signature')
# If CDash reporting is enabled, we first register this build with
# the specified CDash instance, then relate the build to those of
# its dependencies.
if enable_cdash:
tty.debug('CDash: Registering build')
(cdash_build_id, cdash_build_stamp) = spack_ci.register_cdash_build(
cdash_build_name, cdash_base_url, cdash_project, cdash_site,
job_spec_buildgroup)
if cdash_build_id is not None:
cdash_upload_url = '{0}/submit.php?project={1}'.format(
cdash_base_url, cdash_project_enc)
install_args.extend([
'--cdash-upload-url',
cdash_upload_url,
'--cdash-build',
cdash_build_name,
'--cdash-site',
cdash_site,
'--cdash-buildstamp',
cdash_build_stamp,
])
tty.debug('CDash: Relating build with dependency builds')
spack_ci.relate_cdash_builds(
spec_map, cdash_base_url, cdash_build_id, cdash_project,
[pipeline_mirror_url, pr_mirror_url, remote_mirror_url])
# A compiler action of 'FIND_ANY' means we are building a bootstrap
# compiler or one of its deps.
# TODO: when compilers are dependencies, we should include --no-add
if compiler_action != 'FIND_ANY':
install_args.append('--no-add')
# TODO: once we have the concrete spec registry, use the DAG hash
# to identify the spec to install, rather than the concrete spec
# yaml file.
install_args.extend(['-f', job_spec_yaml_path])
tty.debug('Installing {0} from source'.format(job_spec.name))
tty.debug('spack install arguments: {0}'.format(install_args))
# Write the install command to a shell script
with open('install.sh', 'w') as fd:
fd.write('#!/bin/bash\n\n')
fd.write('\n# spack install command\n')
fd.write(' '.join(['"{0}"'.format(i) for i in install_args]))
fd.write('\n')
st = os.stat('install.sh')
os.chmod('install.sh', st.st_mode | stat.S_IEXEC)
install_copy_path = os.path.join(repro_dir, 'install.sh')
shutil.copyfile('install.sh', install_copy_path)
# Run the generated install.sh shell script as if it were being run in
# a login shell.
try:
install_process = subprocess.Popen(['bash', '-l', './install.sh'])
install_process.wait()
install_exit_code = install_process.returncode
except (ValueError, subprocess.CalledProcessError, OSError) as inst:
tty.error('Encountered error running install script')
tty.error(inst)
# Now do the post-install tasks
tty.debug('spack install exited {0}'.format(install_exit_code))
# If a spec fails to build in a spack develop pipeline, we add it to a
# list of known broken full hashes. This allows spack PR pipelines to
# avoid wasting compute cycles attempting to build those hashes.
if install_exit_code == INSTALL_FAIL_CODE and spack_is_develop_pipeline:
tty.debug('Install failed on develop')
if 'broken-specs-url' in gitlab_ci:
broken_specs_url = gitlab_ci['broken-specs-url']
dev_fail_hash = job_spec.full_hash()
broken_spec_path = url_util.join(broken_specs_url, dev_fail_hash)
tty.msg('Reporting broken develop build as: {0}'.format(
broken_spec_path))
tmpdir = tempfile.mkdtemp()
empty_file_path = os.path.join(tmpdir, 'empty.txt')
broken_spec_details = {
'broken-spec': {
'job-url': get_env_var('CI_JOB_URL'),
'pipeline-url': get_env_var('CI_PIPELINE_URL'),
'concrete-spec-yaml': job_spec.to_dict(hash=ht.full_hash)
}
}
try:
with open(empty_file_path, 'w') as efd:
efd.write(syaml.dump(broken_spec_details))
web_util.push_to_url(empty_file_path,
broken_spec_path,
keep_original=False,
extra_args={'ContentType': 'text/plain'})
except Exception as err:
# If we got some kind of S3 (access denied or other connection
# error), the first non boto-specific class in the exception
# hierarchy is Exception. Just print a warning and return
msg = 'Error writing to broken specs list {0}: {1}'.format(
broken_spec_path, err)
tty.warn(msg)
finally:
shutil.rmtree(tmpdir)
# We generated the "spack install ..." command to "--keep-stage", copy
# any logs from the staging directory to artifacts now
spack_ci.copy_stage_logs_to_artifacts(job_spec, job_log_dir)
# Create buildcache on remote mirror, either on pr-specific mirror or
# on the main mirror defined in the gitlab-enabled spack environment
if spack_is_pr_pipeline:
buildcache_mirror_url = pr_mirror_url
else:
buildcache_mirror_url = remote_mirror_url
# If the install succeeded, create a buildcache entry for this job spec
# and push it to one or more mirrors. If the install did not succeed,
# print out some instructions on how to reproduce this build failure
# outside of the pipeline environment.
if install_exit_code == 0:
can_sign = spack_ci.can_sign_binaries()
sign_binaries = can_sign and spack_is_pr_pipeline is False
# Create buildcache in either the main remote mirror, or in the
# per-PR mirror, if this is a PR pipeline
if buildcache_mirror_url:
spack_ci.push_mirror_contents(env, job_spec, job_spec_yaml_path,
buildcache_mirror_url, sign_binaries)
if cdash_build_id:
tty.debug('Writing cdashid ({0}) to remote mirror: {1}'.format(
cdash_build_id, buildcache_mirror_url))
spack_ci.write_cdashid_to_mirror(cdash_build_id, job_spec,
buildcache_mirror_url)
# Create another copy of that buildcache in the per-pipeline
# temporary storage mirror (this is only done if either
# artifacts buildcache is enabled or a temporary storage url
# prefix is set)
if pipeline_mirror_url:
spack_ci.push_mirror_contents(env, job_spec, job_spec_yaml_path,
pipeline_mirror_url, sign_binaries)
if cdash_build_id:
tty.debug('Writing cdashid ({0}) to remote mirror: {1}'.format(
cdash_build_id, pipeline_mirror_url))
spack_ci.write_cdashid_to_mirror(cdash_build_id, job_spec,
pipeline_mirror_url)
# If this is a develop pipeline, check if the spec that we just built is
# on the broken-specs list. If so, remove it.
if spack_is_develop_pipeline and 'broken-specs-url' in gitlab_ci:
broken_specs_url = gitlab_ci['broken-specs-url']
just_built_hash = job_spec.full_hash()
broken_spec_path = url_util.join(broken_specs_url, just_built_hash)
if web_util.url_exists(broken_spec_path):
tty.msg('Removing {0} from the list of broken specs'.format(
broken_spec_path))
try:
web_util.remove_url(broken_spec_path)
except Exception as err:
# If we got some kind of S3 (access denied or other connection
# error), the first non boto-specific class in the exception
# hierarchy is Exception. Just print a warning and return
msg = 'Error removing {0} from broken specs list: {1}'.format(
broken_spec_path, err)
tty.warn(msg)
else:
tty.debug('spack install exited non-zero, will not create buildcache')
api_root_url = get_env_var('CI_API_V4_URL')
ci_project_id = get_env_var('CI_PROJECT_ID')
ci_job_id = get_env_var('CI_JOB_ID')
repro_job_url = '{0}/projects/{1}/jobs/{2}/artifacts'.format(
api_root_url, ci_project_id, ci_job_id)
# Control characters cause this to be printed in blue so it stands out
reproduce_msg = """
\033[34mTo reproduce this build locally, run:
spack ci reproduce-build {0} [--working-dir <dir>]
If this project does not have public pipelines, you will need to first:
export GITLAB_PRIVATE_TOKEN=<generated_token>
... then follow the printed instructions.\033[0;0m
""".format(repro_job_url)
print(reproduce_msg)
# Tie job success/failure to the success/failure of building the spec
return install_exit_code
def generate_gitlab_ci_yaml(env,
print_summary,
output_file,
prune_dag=False,
check_index_only=False,
run_optimizer=False,
use_dependencies=False):
# FIXME: What's the difference between one that opens with 'spack'
# and one that opens with 'env'? This will only handle the former.
with spack.concretize.disable_compiler_existence_check():
env.concretize()
yaml_root = ev.config_dict(env.yaml)
if 'gitlab-ci' not in yaml_root:
tty.die('Environment yaml does not have "gitlab-ci" section')
gitlab_ci = yaml_root['gitlab-ci']
build_group = None
enable_cdash_reporting = False
cdash_auth_token = None
if 'cdash' in yaml_root:
enable_cdash_reporting = True
ci_cdash = yaml_root['cdash']
build_group = ci_cdash['build-group']
cdash_url = ci_cdash['url']
cdash_project = ci_cdash['project']
cdash_site = ci_cdash['site']
if 'SPACK_CDASH_AUTH_TOKEN' in os.environ:
tty.verbose("Using CDash auth token from environment")
cdash_auth_token = os.environ.get('SPACK_CDASH_AUTH_TOKEN')
is_pr_pipeline = (os.environ.get('SPACK_IS_PR_PIPELINE',
'').lower() == 'true')
spack_pr_branch = os.environ.get('SPACK_PR_BRANCH', None)
pr_mirror_url = None
if spack_pr_branch:
pr_mirror_url = url_util.join(SPACK_PR_MIRRORS_ROOT_URL,
spack_pr_branch)
if 'mirrors' not in yaml_root or len(yaml_root['mirrors'].values()) < 1:
tty.die('spack ci generate requires an env containing a mirror')
ci_mirrors = yaml_root['mirrors']
mirror_urls = [url for url in ci_mirrors.values()]
enable_artifacts_buildcache = False
if 'enable-artifacts-buildcache' in gitlab_ci:
enable_artifacts_buildcache = gitlab_ci['enable-artifacts-buildcache']
rebuild_index_enabled = True
if 'rebuild-index' in gitlab_ci and gitlab_ci['rebuild-index'] is False:
rebuild_index_enabled = False
temp_storage_url_prefix = None
if 'temporary-storage-url-prefix' in gitlab_ci:
temp_storage_url_prefix = gitlab_ci['temporary-storage-url-prefix']
bootstrap_specs = []
phases = []
if 'bootstrap' in gitlab_ci:
for phase in gitlab_ci['bootstrap']:
try:
phase_name = phase.get('name')
strip_compilers = phase.get('compiler-agnostic')
except AttributeError:
phase_name = phase
strip_compilers = False
phases.append({
'name': phase_name,
'strip-compilers': strip_compilers,
})
for bs in env.spec_lists[phase_name]:
bootstrap_specs.append({
'spec': bs,
'phase-name': phase_name,
'strip-compilers': strip_compilers,
})
phases.append({
'name': 'specs',
'strip-compilers': False,
})
# Add this mirror if it's enabled, as some specs might be up to date
# here and thus not need to be rebuilt.
if pr_mirror_url:
add_pr_mirror(pr_mirror_url)
# Speed up staging by first fetching binary indices from all mirrors
# (including the per-PR mirror we may have just added above).
bindist.binary_index.update()
staged_phases = {}
try:
for phase in phases:
phase_name = phase['name']
with spack.concretize.disable_compiler_existence_check():
staged_phases[phase_name] = stage_spec_jobs(
env.spec_lists[phase_name],
check_index_only=check_index_only)
finally:
# Clean up PR mirror if enabled
if pr_mirror_url:
remove_pr_mirror()
all_job_names = []
output_object = {}
job_id = 0
stage_id = 0
stage_names = []
max_length_needs = 0
max_needs_job = ''
before_script, after_script = None, None
for phase in phases:
phase_name = phase['name']
strip_compilers = phase['strip-compilers']
main_phase = is_main_phase(phase_name)
spec_labels, dependencies, stages = staged_phases[phase_name]
for stage_jobs in stages:
stage_name = 'stage-{0}'.format(stage_id)
stage_names.append(stage_name)
stage_id += 1
for spec_label in stage_jobs:
spec_record = spec_labels[spec_label]
root_spec = spec_record['rootSpec']
pkg_name = pkg_name_from_spec_label(spec_label)
release_spec = root_spec[pkg_name]
runner_attribs = find_matching_config(release_spec, gitlab_ci)
if not runner_attribs:
tty.warn('No match found for {0}, skipping it'.format(
release_spec))
continue
tags = [tag for tag in runner_attribs['tags']]
variables = {}
if 'variables' in runner_attribs:
variables.update(runner_attribs['variables'])
image_name = None
image_entry = None
if 'image' in runner_attribs:
build_image = runner_attribs['image']
try:
image_name = build_image.get('name')
entrypoint = build_image.get('entrypoint')
image_entry = [p for p in entrypoint]
except AttributeError:
image_name = build_image
job_script = [
'spack env activate --without-view .',
'spack ci rebuild',
]
if 'script' in runner_attribs:
job_script = [s for s in runner_attribs['script']]
before_script = None
if 'before_script' in runner_attribs:
before_script = [
s for s in runner_attribs['before_script']
]
after_script = None
if 'after_script' in runner_attribs:
after_script = [s for s in runner_attribs['after_script']]
osname = str(release_spec.architecture)
job_name = get_job_name(phase_name, strip_compilers,
release_spec, osname, build_group)
compiler_action = 'NONE'
if len(phases) > 1:
compiler_action = 'FIND_ANY'
if is_main_phase(phase_name):
compiler_action = 'INSTALL_MISSING'
job_vars = {
'SPACK_ROOT_SPEC':
format_root_spec(root_spec, main_phase, strip_compilers),
'SPACK_JOB_SPEC_PKG_NAME':
release_spec.name,
'SPACK_COMPILER_ACTION':
compiler_action,
'SPACK_IS_PR_PIPELINE':
str(is_pr_pipeline),
}
job_dependencies = []
if spec_label in dependencies:
if enable_artifacts_buildcache:
# Get dependencies transitively, so they're all
# available in the artifacts buildcache.
dep_jobs = [
d for d in release_spec.traverse(deptype=all,
root=False)
]
else:
# In this case, "needs" is only used for scheduling
# purposes, so we only get the direct dependencies.
dep_jobs = []
for dep_label in dependencies[spec_label]:
dep_pkg = pkg_name_from_spec_label(dep_label)
dep_root = spec_labels[dep_label]['rootSpec']
dep_jobs.append(dep_root[dep_pkg])
job_dependencies.extend(
format_job_needs(phase_name, strip_compilers, dep_jobs,
osname, build_group, prune_dag,
spec_labels,
enable_artifacts_buildcache))
rebuild_spec = spec_record['needs_rebuild']
# This next section helps gitlab make sure the right
# bootstrapped compiler exists in the artifacts buildcache by
# creating an artificial dependency between this spec and its
# compiler. So, if we are in the main phase, and if the
# compiler we are supposed to use is listed in any of the
# bootstrap spec lists, then we will add more dependencies to
# the job (that compiler and maybe it's dependencies as well).
if is_main_phase(phase_name):
spec_arch_family = (release_spec.architecture.target.
microarchitecture.family)
compiler_pkg_spec = compilers.pkg_spec_for_compiler(
release_spec.compiler)
for bs in bootstrap_specs:
c_spec = bs['spec']
bs_arch = c_spec.architecture
bs_arch_family = (
bs_arch.target.microarchitecture.family)
if (c_spec.satisfies(compiler_pkg_spec)
and bs_arch_family == spec_arch_family):
# We found the bootstrap compiler this release spec
# should be built with, so for DAG scheduling
# purposes, we will at least add the compiler spec
# to the jobs "needs". But if artifact buildcache
# is enabled, we'll have to add all transtive deps
# of the compiler as well.
# Here we check whether the bootstrapped compiler
# needs to be rebuilt. Until compilers are proper
# dependencies, we artificially force the spec to
# be rebuilt if the compiler targeted to build it
# needs to be rebuilt.
bs_specs, _, _ = staged_phases[bs['phase-name']]
c_spec_key = spec_deps_key(c_spec)
rbld_comp = bs_specs[c_spec_key]['needs_rebuild']
rebuild_spec = rebuild_spec or rbld_comp
# Also update record so dependents do not fail to
# add this spec to their "needs"
spec_record['needs_rebuild'] = rebuild_spec
dep_jobs = [c_spec]
if enable_artifacts_buildcache:
dep_jobs = [
d for d in c_spec.traverse(deptype=all)
]
job_dependencies.extend(
format_job_needs(bs['phase-name'],
bs['strip-compilers'],
dep_jobs, str(bs_arch),
build_group, prune_dag,
bs_specs,
enable_artifacts_buildcache))
else:
debug_msg = ''.join([
'Considered compiler {0} for spec ',
'{1}, but rejected it either because it was ',
'not the compiler required by the spec, or ',
'because the target arch families of the ',
'spec and the compiler did not match'
]).format(c_spec, release_spec)
tty.debug(debug_msg)
if prune_dag and not rebuild_spec:
continue
job_vars['SPACK_SPEC_NEEDS_REBUILD'] = str(rebuild_spec)
if enable_cdash_reporting:
cdash_build_name = get_cdash_build_name(
release_spec, build_group)
all_job_names.append(cdash_build_name)
related_builds = [] # Used for relating CDash builds
if spec_label in dependencies:
related_builds = ([
spec_labels[d]['spec'].name
for d in dependencies[spec_label]
])
job_vars['SPACK_CDASH_BUILD_NAME'] = cdash_build_name
job_vars['SPACK_RELATED_BUILDS_CDASH'] = ';'.join(
sorted(related_builds))
variables.update(job_vars)
artifact_paths = [
'jobs_scratch_dir',
'cdash_report',
]
if enable_artifacts_buildcache:
bc_root = 'local_mirror/build_cache'
artifact_paths.extend([
os.path.join(bc_root, p) for p in [
bindist.tarball_name(release_spec, '.spec.yaml'),
bindist.tarball_name(release_spec, '.cdashid'),
bindist.tarball_directory_name(release_spec),
]
])
job_object = {
'stage': stage_name,
'variables': variables,
'script': job_script,
'tags': tags,
'artifacts': {
'paths': artifact_paths,
'when': 'always',
},
'needs': sorted(job_dependencies, key=lambda d: d['job']),
'retry': {
'max': 2,
'when': JOB_RETRY_CONDITIONS,
},
'interruptible': True
}
length_needs = len(job_dependencies)
if length_needs > max_length_needs:
max_length_needs = length_needs
max_needs_job = job_name
if before_script:
job_object['before_script'] = before_script
if after_script:
job_object['after_script'] = after_script
if image_name:
job_object['image'] = image_name
if image_entry is not None:
job_object['image'] = {
'name': image_name,
'entrypoint': image_entry,
}
output_object[job_name] = job_object
job_id += 1
if print_summary:
for phase in phases:
phase_name = phase['name']
tty.msg('Stages for phase "{0}"'.format(phase_name))
phase_stages = staged_phases[phase_name]
print_staging_summary(*phase_stages)
tty.debug('{0} build jobs generated in {1} stages'.format(
job_id, stage_id))
if job_id > 0:
tty.debug('The max_needs_job is {0}, with {1} needs'.format(
max_needs_job, max_length_needs))
# Use "all_job_names" to populate the build group for this set
if enable_cdash_reporting and cdash_auth_token:
try:
populate_buildgroup(all_job_names, build_group, cdash_project,
cdash_site, cdash_auth_token, cdash_url)
except (SpackError, HTTPError, URLError) as err:
tty.warn('Problem populating buildgroup: {0}'.format(err))
else:
tty.warn('Unable to populate buildgroup without CDash credentials')
service_job_config = None
if 'service-job-attributes' in gitlab_ci:
service_job_config = gitlab_ci['service-job-attributes']
default_attrs = [
'image',
'tags',
'variables',
'before_script',
# 'script',
'after_script',
]
if job_id > 0:
if temp_storage_url_prefix:
# There were some rebuild jobs scheduled, so we will need to
# schedule a job to clean up the temporary storage location
# associated with this pipeline.
stage_names.append('cleanup-temp-storage')
cleanup_job = {}
if service_job_config:
copy_attributes(default_attrs, service_job_config, cleanup_job)
cleanup_job['stage'] = 'cleanup-temp-storage'
cleanup_job['script'] = [
'spack -d mirror destroy --mirror-url {0}/$CI_PIPELINE_ID'.
format(temp_storage_url_prefix)
]
cleanup_job['when'] = 'always'
output_object['cleanup'] = cleanup_job
if rebuild_index_enabled:
# Add a final job to regenerate the index
stage_names.append('stage-rebuild-index')
final_job = {}
if service_job_config:
copy_attributes(default_attrs, service_job_config, final_job)
index_target_mirror = mirror_urls[0]
if is_pr_pipeline:
index_target_mirror = pr_mirror_url
final_job['stage'] = 'stage-rebuild-index'
final_job['script'] = [
'spack buildcache update-index --keys -d {0}'.format(
index_target_mirror)
]
final_job['when'] = 'always'
output_object['rebuild-index'] = final_job
output_object['stages'] = stage_names
# Capture the version of spack used to generate the pipeline, transform it
# into a value that can be passed to "git checkout", and save it in a
# global yaml variable
spack_version = spack.main.get_version()
version_to_clone = None
v_match = re.match(r"^\d+\.\d+\.\d+$", spack_version)
if v_match:
version_to_clone = 'v{0}'.format(v_match.group(0))
else:
v_match = re.match(r"^[^-]+-[^-]+-([a-f\d]+)$", spack_version)
if v_match:
version_to_clone = v_match.group(1)
else:
version_to_clone = spack_version
output_object['variables'] = {
'SPACK_VERSION': spack_version,
'SPACK_CHECKOUT_VERSION': version_to_clone,
}
if pr_mirror_url:
output_object['variables']['SPACK_PR_MIRROR_URL'] = pr_mirror_url
sorted_output = {}
for output_key, output_value in sorted(output_object.items()):
sorted_output[output_key] = output_value
# TODO(opadron): remove this or refactor
if run_optimizer:
import spack.ci_optimization as ci_opt
sorted_output = ci_opt.optimizer(sorted_output)
# TODO(opadron): remove this or refactor
if use_dependencies:
import spack.ci_needs_workaround as cinw
sorted_output = cinw.needs_to_dependencies(sorted_output)
else:
# No jobs were generated
tty.debug('No specs to rebuild, generating no-op job')
noop_job = {}
if service_job_config:
copy_attributes(default_attrs, service_job_config, noop_job)
if 'script' not in noop_job:
noop_job['script'] = [
'echo "All specs already up to date, nothing to rebuild."',
]
sorted_output = {'no-specs-to-rebuild': noop_job}
with open(output_file, 'w') as outf:
outf.write(syaml.dump_config(sorted_output, default_flow_style=True))
