pmbootstrap/pmb/parse/depends.py

187 lines
7.1 KiB
Python
Raw Permalink Normal View History

2023-01-22 18:11:10 +00:00
# Copyright 2023 Oliver Smith
# SPDX-License-Identifier: GPL-3.0-or-later
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
import logging
import pmb.chroot
import pmb.chroot.apk
import pmb.helpers.pmaports
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
import pmb.parse.apkindex
import pmb.parse.arch
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
def package_from_aports(args, pkgname_depend):
"""
:returns: None when there is no aport, or a dict with the keys pkgname,
depends, version. The version is the combined pkgver and pkgrel.
"""
# Get the aport
aport = pmb.helpers.pmaports.find(args, pkgname_depend, False)
if not aport:
return None
# Parse its version
apkbuild = pmb.parse.apkbuild(f"{aport}/APKBUILD")
pkgname = apkbuild["pkgname"]
version = apkbuild["pkgver"] + "-r" + apkbuild["pkgrel"]
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
# Return the dict
logging.verbose(
f"{pkgname_depend}: provided by: {pkgname}-{version} in {aport}")
return {"pkgname": pkgname,
"depends": apkbuild["depends"],
"version": version}
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
def package_provider(args, pkgname, pkgnames_install, suffix="native"):
"""
:param pkgnames_install: packages to be installed
:returns: a block from the apkindex: {"pkgname": "...", ...}
or None (no provider found)
"""
# Get all providers
arch = pmb.parse.arch.from_chroot_suffix(args, suffix)
providers = pmb.parse.apkindex.providers(args, pkgname, arch, False)
# 0. No provider
if len(providers) == 0:
return None
# 1. Only one provider
logging.verbose(f"{pkgname}: provided by: {', '.join(providers)}")
if len(providers) == 1:
return list(providers.values())[0]
# 2. Provider with the same package name
if pkgname in providers:
logging.verbose(f"{pkgname}: choosing package of the same name as "
"provider")
return providers[pkgname]
# 3. Pick a package that will be installed anyway
for provider_pkgname, provider in providers.items():
if provider_pkgname in pkgnames_install:
logging.verbose(f"{pkgname}: choosing provider '{provider_pkgname}"
"', because it will be installed anyway")
return provider
# 4. Pick a package that is already installed
installed = pmb.chroot.apk.installed(args, suffix)
for provider_pkgname, provider in providers.items():
if provider_pkgname in installed:
logging.verbose(f"{pkgname}: choosing provider '{provider_pkgname}"
f"', because it is installed in the '{suffix}' "
"chroot already")
return provider
# 5. Pick an explicitly selected provider
provider_pkgname = args.selected_providers.get(pkgname, "")
if provider_pkgname in providers:
logging.verbose(f"{pkgname}: choosing provider '{provider_pkgname}', "
"because it was explicitly selected.")
return providers[provider_pkgname]
# 6. Pick the provider(s) with the highest priority
providers = pmb.parse.apkindex.provider_highest_priority(
providers, pkgname)
pmb.parse: Respect provider_priority if multiple providers exist (MR 1945) Alpine APKBUILDs have the concept of "provider priorities" that affect the choice of the provider to install when multiple packages provide a virtual package. One use case for this is to allow installation of different firmware versions. bq-paella can run unsigned firmware, therefore you have the choice between using the original firmware from the manufacturer, or a slightly newer version from Qualcomm for the Dragonboard 410c. We add provides="firmware-qcom-msm8916-wcnss" (the "virtual package") to both firmware-qcom-db410c-wcnss and firmware-bq-picmt-wcnss. At this point, attempting to install "firmware-qcom-msm8916-wcnss" would still fail with apk. (Because it does not know which provider to install.) To pick a default we can set e.g. provider_priority=100 for firmware-qcom-db410c-wcnss (the slightly newer version). In that case, firmware-qcom-db410c-wcnss should be installed by default. However, the user can choose to do "apk add firmware-bq-picmt-wcnss" to override the default choice in case of problems. In that case, the conflicting firmware-qcom-db410c-wcnss will be automatically removed. At the moment, pmbootstrap does not respect the "provider_priority" at all. In the above case, it would always install "firmware-bq-picmt-wcnss" during "pmbootstrap install" since that has the shortest name. Extend the pmbootstrap code to pick a provider with the highest priority (if any of the providers has a priority set).
2020-06-08 20:10:12 +00:00
if len(providers) == 1:
return list(providers.values())[0]
# 7. Pick the shortest provider. (Note: Normally apk would fail here!)
return pmb.parse.apkindex.provider_shortest(providers, pkgname)
def package_from_index(args, pkgname_depend, pkgnames_install, package_aport,
suffix="native"):
"""
:returns: None when there is no aport and no binary package, or a dict with
the keys pkgname, depends, version from either the aport or the
binary package provider.
"""
# No binary package
provider = package_provider(args, pkgname_depend, pkgnames_install, suffix)
if not provider:
return package_aport
# Binary package outdated
if (package_aport and pmb.parse.version.compare(package_aport["version"],
provider["version"]) == 1):
logging.verbose(pkgname_depend + ": binary package is outdated")
return package_aport
# Binary up to date (#893: overrides aport, so we have sonames in depends)
if package_aport:
logging.verbose(pkgname_depend + ": binary package is"
" up to date, using binary dependencies"
" instead of the ones from the aport")
return provider
def recurse(args, pkgnames, suffix="native"):
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
"""
Find all dependencies of the given pkgnames.
:param suffix: the chroot suffix to resolve dependencies for. If a package
has multiple providers, we look at the installed packages in
the chroot to make a decision (see package_provider()).
:returns: list of pkgnames: consists of the initial pkgnames plus all
depends. Dependencies explicitly marked as conflicting are
prefixed with !.
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
"""
logging.debug(f"({suffix}) calculate depends of {', '.join(pkgnames)} "
"(pmbootstrap -v for details)")
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
# Iterate over todo-list until is is empty
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
todo = list(pkgnames)
required_by = {}
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
ret = []
while len(todo):
# Skip already passed entries
pkgname_depend = todo.pop(0)
if pkgname_depend in ret:
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
continue
# Check if the dependency is explicitly marked as conflicting
is_conflict = pkgname_depend.startswith("!")
pkgname_depend = pkgname_depend.lstrip("!")
# Get depends and pkgname from aports
pkgnames_install = list(ret) + todo
package = package_from_aports(args, pkgname_depend)
package = package_from_index(args, pkgname_depend, pkgnames_install,
package, suffix)
# Nothing found
if not package:
if is_conflict:
# This package was probably dropped from the repos, so we don't
# care if it doesn't exist since it's a conflicting depend that
# wouldn't be installed anyways.
continue
source = 'world'
if pkgname_depend in required_by:
source = ', '.join(required_by[pkgname_depend])
raise RuntimeError(f"Could not find dependency '{pkgname_depend}' "
"in checked out pmaports dir or any APKINDEX. "
f"Required by '{source}'. See: "
"https://postmarketos.org/depends")
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
# Determine pkgname
pkgname = package["pkgname"]
if is_conflict:
pkgname = f"!{pkgname}"
# Append to todo/ret (unless it is a duplicate)
if pkgname in ret:
logging.verbose(f"{pkgname}: already found")
else:
if not is_conflict:
depends = package["depends"]
logging.verbose(f"{pkgname}: depends on: {','.join(depends)}")
if depends:
todo += depends
for dep in depends:
if dep not in required_by:
required_by[dep] = set()
required_by[dep].add(pkgname_depend)
ret.append(pkgname)
Properly rebuild/install packages when something changed (Fix #120, #108, #131) (#129) TLDR: Always rebuild/install packages when something changed when executing "pmbootstrap install/initfs/flash", more speed in dependency resolution. --- pmbootstrap has already gotten some support for "timestamp based rebuilds", which modifies the logic for when packages should be rebuilt. It doesn't only consider packages outdated with old pkgver/pkgrel combinations, but also packages, where a source file has a newer timestamp, than the built package has. I've found out, that this can lead to more rebuilds than expected. For example, when you check out the pmbootstrap git repository again into another folder, although you have already built packages. Then all files have the timestamp of the checkout, and the packages will appear to be outdated. While this is not largely a concern now, this will become a problem once we have a binary package repository, because then the packages from the binary repo will always seem to be outdated, if you just freshly checked out the repository. To combat this, git gets asked if the files from the aport we're looking at are in sync with upstream, or not. Only when the files are not in sync with upstream and the timestamps of the sources are newer, a rebuild gets triggered from now on. In case this logic should fail, I've added an option during "pmbootstrap init" where you can enable or disable the "timestamp based rebuilds" option. In addition to that, this commit also works on fixing #120: packages do not get updated in "pmbootstrap install" after they have been rebuilt. For this to work, we specify all packages explicitly for abuild, instead of letting abuild do the resolving. This feature will also work with the "timestamp based rebuilds". This commit also fixes the working_dir argument in pmb.helpers.run.user, which was simply ignored before. Finally, the performance of the dependency resolution is faster again (when compared to the current version in master), because the parsed apkbuilds and finding the aport by pkgname gets cached during one pmbootstrap call (in args.cache, which also makes it easy to put fake data there in testcases). The new dependency resolution code can output lots of verbose messages for debugging by specifying the `-v` parameter. The meaning of that changed, it used to output the file names where log messages come from, but no one seemed to use that anyway.
2017-07-10 15:23:43 +00:00
return ret