some scripts temporarily install packages to accomplish some work before
then removing them. the list of packages installed is kept in memory in a
variable.
a weakness of this design is that if a failure occurs or the user cancels,
and then following this the user re-runs `lb build`, letting it try to
pick up and recover from where it left off, that list of packages that had
been installed is lost, resulting in those packages that were installed
then being a permanent part of the chroot.
here we fix this weakness by backing up the list to a file, which is always
read from on removal. thus in a recovery situation, any packages still
installed from a situation like that just described, will be removed upon
the next use of `Remove_package()`.
this is not perfect, since we are having to wait for opportunistic
execution of the remove function. we need to find a suitable place for the
`Cleanup_temp_packages()` function to be used.
- doing so in `Init_config_data()` would not be suitable because we don't
hold the lock when that's run, even if we ignored the hijacking of that
function for taking such action...
- doing it in `Exit()` doesn't seem a good fit either.
- putting it explicitly in every build script just seems a little messy...
perhaps a local exit trap like for removing the lock...?
note that `binary_rootfs` skips running the remove function after installing
tooling, since it just throws the wrapper chroot away, which then leaves the
file around with stale data for the next remove instance to pick up, which
then does not actually remove it because it's not installed. this is not
ideal either... perhaps the optimisation should be removed from that script?
Gbp-Dch: Short
- the comparison with LB_ARCHITECTURE seems to be based upon an old idea
of it potentially being intended in the past to hold more than one
arch (when it's name as also plural), but it in fact only holds one.
- similarly LB_BOOTSTRAP_QEMU_ARCHITECTURE is singular, so the comparison
with the set of params is bad.
- storing $@ in a variable before then looping is not a good idea
- script name was not being passed to the warnings
- can simplify by just returning in the valid case
i'm very suspicious about the correctness of the qemu bits, but that
can wait for now.
Gbp-Dch: Short
the design choice from when EFI support was introduced was to change
`--bootloader` to `--bootloaders`, with users specifying their selection
of BIOS and EFI bootloaders together. at this time there were not even any
decent validation checks being performed, and invalid combinations could
cause some chaos.
since then proper validation was put in place, including checking that
only a single instance of each of BIOS and EFI bootloaders exists in the
selection.
here we tweak things such that we stick with the same option, but we split
the selection up such that we store the BIOS and EFI selections separately
within the saved config file, and offer it up to scripts to help simplify
those scripts.
we must however retain support for splitting from the combined option,
both because we still use it in the combined option, and for backwards
compatibility with older saved configs.
Gbp-Dch: Short
thus far, config bootloader validation only did the basic check that each
bootloader specified was a known and supported bootloader, it did not check
combinations.
it now checks combinations, and strips out the previous "bootloader role"
stuff.
the no-bootloaders warning is duplicated, covering two slightly different
situations (empty string, and whitespace string). this is anticipated to
be just temporary, with this just being the first step in better handling
bootloader selections.
Gbp-Dch: Short
The core purpose of `Prepare_config()` is to prepare the config by setting
defaults where not specified. It should not be involved itself in
validation of user options.
This can be an important distinction, for instance in `clean` where we
use this function, but avoid validation to potentially allow for an option
that deletes the config, that should work even in the face of an invalid
config.
Gbp-Dch: Ignore
while `lb build` ran the config validation check, spotting invalid configs
and stopping with an error, the major build stages if executed directly did
not, nor did the component stages.
here we move execution of the validation function into the common init
function, with an exported variable used to indicate that validation has
been performed. thus validation is performed no matter what part of the
build system you execute, but only once.
Gbp-Dch: Short
using `$VERSION` as part of the default `$LB_ISO_PREPARER` means that when
you simply run `lb config` once, this variable is stored as a part of the
string, and replaced on use, but if you run `lb config` twice, it gets
replaced with a fixed version, that is then used in all subsequent builds.
let's replace with a placeholder (`@LB_VERSION@`) that can be used both in
the default, or in user strings, and will be replaced on use only.
this means that subsequent builds will always reflect the actual version
of live-build used.
Gbp-Dch: Ignore
`DATE_UTC_OPTION` is set in `Prepare_config()` for use by scripts, even
though only a few scripts will actually use it, since it allows those
scripts to be cleaner. we may want to possibly extend this as a
`DATE_OPTIONS` variable perhaps as part of enabling proper reproduciblity.
Gbp-Dch: Short
- the definition of $PROGRAM as used in $USAGE strings defined in each
script has been broken for a long time, being simply "lb" when it needs
to be "lb COMMAND".
- `config` changed $PROGRAM to "lb config" thus its output was correct in
this regard unlike everything else, but with the switch to a more
"intelligent" `Man()` function recently, it means that instead of
`man lb config`, what was actually run was `man lb config config`,
which displayed the manpage, then on exiting with `q`, it showed some
sort of index line todo with a "config" search (no exact manpage
match?), for which you had to enter `ctrl+c` to get rid of.
this revises things to fix the issues, minimising change by changing
$PROGRAM to "lb COMMAND", with the frontend overriding this.
Gbp-Dch: Ignore
load the base set of files first, then any user arch/dist specific override
files second.
this is important since whenever config items get moved in the base files,
this can break the override files, if an option specified in one was moved
to a different common file that it loaded after.
Gbp-Dch: Ignore
- it's only used by the debootstrap script after alternatives were dropped
long ago, so let's move it, avoiding it being loaded for everything
else.
- there's no need to pass printing another message through it.
- there's little point in making the sid distinction if you happen to
decide to build sid, it's a given that it's less stable than stable.
really, is there any need for this at all?
Gbp-Dch: Ignore
This is considered to be more robust.
Two instances remain:
scripts/build/chroot_archives, line 257:
if [ "${LB_APT}" = "aptitude" ] && [ ! $(Chroot chroot "which aptitude") ]
The command is run inside a chroot where the environment might be special,
and would need further testing.
manpages/Makefile, line 42:
@if [ ! -x "$$(which po4a 2>/dev/null)" ]; \
I am insufficiently familiar with makefile syntax to edit this.
support for this has not existed since v2.0~a24-1 !!
until 4d22ca948a (v2.0~a24-1), `PROGRAM` was
defined as `PROGRAM="$(basename ${0})"`. then this commit simply got rid
of the definition.
it then flipped back and forth a bit in a few ways, but always sticking
to a simple fixed "live-build" type string after that commit, as it has
been ever since.
the code removed here has thus been junk since then, and the documentation
corrected here out of date since then.
the checking of `LB_CONFIG` is also possibly junk, but i know too little
about it. it was introduced in v1.0~a43-1 in commit
705a4178e7 with a changelog entry of
"Centraly evaluate LH_CONFIG for custom configuration file."
Gbp-Dch: Short
just as most scripts are skipped if their stagefile exists (indicating
that they have already been run to completion), including chroot
preparation scripts in install mode, this implements the same guard for
chroot prep remove mode, such that they exit early if their stagefile
does not exist, indicating that the modification has already been removed.
(also override-able by --force in the same way).
this basically just uses a tweaked copy of Check_stagefile().
Gbp-Dch: Short
no backwards compatibility hack for reading the old var from existing
saved config used because this was previously stored in the alternate
format config/build file.
Gbp-Dch: Short
no backwards compatibility hack for reading the old var from existing
saved config used because this was previously stored in the alternate
format config/build file.
Gbp-Dch: Short
back in v4.0~a6-1 a transition process was started to move the live-build
config to a new format. the new format was INI style, and required
parsing functions to read/write values, compared to the existing format
which was just shell script code setting variables.
this partial transition is the explanation for the existence of the
`New_configuration()` function, and understanding this is important to
understanding the purpose of it - it is not in fact intended for creating
a new configuration, it is just related to the new config format
transition.
the positives of the new format were that it was somewhat cleaner looking,
while the negative was the terrible relative efficiency.
the file `config/build` was created to hold options in this new format.
the transition was only ever completed for a handful of config options:
- architecture
- archive areas and parent archive areas
- live image name
- live image type
a 'configuration version' attribute was also saved, which is not used by
anything.
the bootstrap-mirror and parent-bootstrap-mirror attributes are pointlessly
stored in it seemingly resulting from work done in v4.0~a17-1. (they are
also stored in another config file from which the value is actually used).
it in fact seems to have been a source of confusion for Raphaël in
authoring 44b9b0a650, since the new
`[parent]-distribution-{chroot|binary}` options it introduced were stored
both in `config/bootstrap` and in `config/build`, while only used from the
former. i expect, understandably, that he thought that `config/build` was
just an information file.
Gbp-Dch: Short
thus it can correctly indicate success/fail status instead of always
indicating failure. when a user asks for usage with -u|--usage then we
should exit in success mode rather than failure as when usage in printed
in response to incorrect usage.
Gbp-Dch: Short
this means that the usage goes from:
lb {-h|--help|-u|--usage|-v|--version}
lb COMMAND [OPTIONS]
to:
lb {-h|--help|-u|--usage|-v|--version}
lb [FRONTEND_OPTIONS] COMMAND [COMMAND_OPTIONS]
though it is probably not worth is to update the description in the
manpages...? hmm...
so for instance this matters for color control with --color|--no-color
(you already had full control via environment vars). previously you could
do `lb COMMAND --no-color` to turn off colour, only to find that output
at the frontend level was still coloured (the option is processed at the
command context level, not the frontend), so you might try to instead use
`lb --no-color COMMAND`, only to find that this was not supported. Well
now it is, and used at the frontend level will fully control colour output
(after the command is processed anyway).
the full set of common options are thus available (except --force) at the
frontend level, and thus for instance all Echo_*() helpers used in the
frontend will work correctly after args are processed.
furthermore usage like `lb --color --help` will actually work. (not that
color is used there, but this previously would have failed with the
frontend treating the `--color` argument as the command; that's the point!)
Gbp-Dch: Short
this was previously not done in 8b109ffb96
to keep the renaming simple, but leaving the variable plural is a cause
for confusion.
since this property is stored in the INI style config/build config file
rather than a shell script based one, at the property there is already
singular, there was no need for a backwards compatibility hack.
Gbp-Dch: Short
being done in Prepare_config() meant that it would get saved into the
config stored on disk which is not what was intended.
being saved on disk is not a big problem, since the injection of the
dynamically determined option is always done on top of the value read
from the saved config anyway, so you'd always end up with two copies of
the option given to apt, the first being from the saved config, the second
being the more correct dynamically determined one, overriding the first.
so there is no functional change here, just better achieving what was
intended.
Gbp-Dch: Ignore
note that the bit of code removed from source_debian relies upon a
variable LB_TASKS which itself is an old leftover artefact from before
v4.0.
Gbp-Dch: Short
since everywhere where 'apt' is a permitted value, 'apt-get' is also, it
just wasn't listed in the option's documentation and thus was also not
listed in the new validation check.
Gbp-Dch: Short
80aa5ab611 implemented a hack to handle
replacement of LB_LINUX_FLAVOURS with LB_LINUX_FLAVOURS_WITH_ARCH in
config files, but implemented it in the wrong place.
adding a conditional conversion within the config file meant that the old
value would only be read from **new** config files that are created
obviously without it, including re-saved configs if `lb config` were
re-run with additional options (not recommended). any existing value in an
existing config file would actually be ignored.
the right place to read the old value was in the Set_defaults() function
(since renamed).
a second issue also existed with the hack, it failed to excape the `$`
and thus printed the existing value of $LB_LINUX_FLAVOURS into the
conditional check being constructed in the config file, instead of
printing the name of the variable. the check embedded into the config
file thus became this on an amd64 machine:
```
if [ -n "amd64" ]
then
LB_LINUX_FLAVOURS_WITH_ARCH="amd64"
fi
```
which is clearly not what was intended.
Gbp-Dch: Short
- add a validation check where an error will be printed
- replace the check done in the grub scripts with one that simple exits
if executed bypassing the validation check
Gbp-Dch: Short
the stuff in defaults.sh is only partially to do with setting default
config values, it also does other prep work like centralised comma
separated value to space separated conversion, and validation.
this is split into two parts to keep the diff simple, first we move the
two small functions from configuration.sh into default.sh...
Gbp-Dch: Short
it mostly applies defaults where a value does not exist, but does more in
some cases. the new name better reflects its usage and functionality.
Gbp-Dch: Short
whilst some parts of the codebase were set up to work with multiple types
specified, others did not work with it and would not necessarily be easy
to adjust. this thus makes some tweaks to adjust things accordingly.
- option renamed to singular form (maintaining backwards compatibility)
- a validation check has been added
- unnecessary glob style type references fixed
- checks with In_list changed to a direct singular comparison
- typo of type "netboot" written as just "net" fixed (though unreachable
so of no consequence; really the code could be removed but it's trivial)
Gbp-Dch: Short
this option actually only has an impact when used for `lb config`; it thus
was incorrectly accepted and captured by Arguments() and the documentation
was wrong.
going back through git history, this has been incorrect since the v0.99 days.
also, reorganised the list alphabetically (no-color was inserted in the
wrong place).
Gbp-Dch: Short
two parts of the code worked with both comma and space separated lists,
while two others only worked with comma separated.
swapping out commas with spaces when we setup the var in
Set_config_defaults() means that individual scripts no longer need to worry
about it and everything supports both; and that we can avoid the
IFS/OLDIFS mess.
Gbp-Dch: Short
i've been noticing that in the binary_syslinux stage a wholebunch of
packages get installed whilst installing isolinux, and then upon removing
the installed packages at the end of the script, apt is listing a large
number of those that were auto installed as no longer in use and can be
removed via `apt autoremove`. this then persists through package
installation and removal steps throughout remaining scripts.
adding `--auto-remove` to `apt-get remove --purge` and similarly
`--purge-unused` to `aptitude purge` instructs apt to get rid of the unused
auto-installed packages at the same time as removing specific package that
brought them in in the first place, and thus resolves the problem.
Gbp-Dch: Short
For consistency with install entries (both in live-build and
official Debian install discs).
Comparing with live-build created installer entries, grub-legacy
and grub2 both favour vga-788 for GUI entries and vga=normal for
test entries, whilst syslinux uses vga-788 for everything.
Gbp-Dch: Short
apt v2.0.1 introduced support for coloured E:/W:/N: labels. this adds
support to control it based upon our own colour control.
note that with utilities like dpkg we do not do this, but apt only uses
its new colour support automatically when `apt` is used directly, it is
not automatically enabled (per isatty()) for `apt-get`/`aptitude` (the
`apt` developer responsible for adding colour support in response to my
request for it told me that it was deliberately done like this per being
customary to not change behaviour of those tools for compatibility
reasons). colour errors/warnings are useful, so we want to turn it on for
our use of these tools where we can.
Gbp-Dch: Short
...from the `Set_config_defaults` function, to being done directly
in `build.sh` (the component which is also responsible for loading
functions, loaded at the start of every script, including the front
end).
thus the colouring decision will now correctly...
- apply to the frontend, such as to the 'root privileges needed'
error, the 'no such script' error, and the command name
colouring that I want to add (the most significant issue).
- apply to error messages generated by the `Arguments` and
`Read_conffiles` functions, which are called before
`Set_config_defaults` by scripts.
as things were, due to the comparison with "false", colour would
_always_ be used in these places (unless _COLOR_ERR=false or
_COLOR_OUT=false wrt. the new command highlight, were set in the
environment when executing a script throught the frontend).
this would not be a problem for normal terminal use of course,
besides being inconsistent where color were turned off, but would
be a bit of a problem if redirected to a file.
a re-evaluation of _COLOR is performed in `Set_config_defaults` to
adjust _COLOR_OUT and _COLOR_ERR where necessary, to correctly
respond to _COLOR being set in saved config files (disabled by
default but a user could always enable), after the point of config
files being loaded.
_COLOR can still be controlled from the environment just as before,
overriding both _COLOR_OUT and _COLOR_ERR.
note that this does not address the fact that --color|--no-color
do not work in the frontend and thus will not impact the colouring
of to-be-introduced command highlighting. this needs to be
addressed separately.
Gbp-Dch: Short
progress-linux, as discussed in MR #142 ([1]) is a little known distro,
which appears to be little more than a personal project of the original
author of live-build.
given that, the expense of maintaining all of these old hacks for it
cannot be justified. it is not known whether or not live-build is even
used with respect to it since the author abandoned live-build some
years ago.
also, at least one past change in live-build possibly broke progress-linux
compatibility anyway, which would have required progress-linux users of
live-build to use a custom progress-linux config, or a progress-linux
fork of live-build, and there is no knowing how much of the hacks in this
"upstream" codebase any user of progress-linux currently relies upon.
and again, progress-linux appears to just be a personal project of
Daniel's, with afaik very little userbase. (Daniel seems to be the only
developer working on the project which speaks to how small it is).
[1]: https://salsa.debian.org/live-team/live-build/-/merge_requests/142
Gbp-Dch: Short
seems to have been obsolete since all the way back at v1.0~a22-1.
history:
- in 0d0de885e3 it was renamed from
LIVE_ROOT to LH_ROOT, but also seems to have become completely
obsolete at this point, and thus mistakenly not actually removed.
before this it seems to have been used to hold the base directory of
live-build from which paths were constructed, but then this use was
removed making it redundant, but it remained in the code.
- 83bc63f725 renamed to LB_ROOT.
- a79a5bea10 dropped setting the variable
only if not already set, in favour of always setting it depending upon
LB_MODE. but still it remained unused.
Gbp-Dch: Short
the frontend handles -h|--help directly and correctly redirects to the
man page.
component scripts however fail to load the correct manpage because they
are being directed to `man <script>` instead of `man lb script`.
(affects the top level commands and major build stages which actually have
man pages; the low level components don't and so will always fail anyway).
this is handled for every script in build.sh. this is not stored in the
saved config or anything, so no need to re-evaluate in
`Set_config_defaults`. this just seems to completely pointless.
Gbp-Dch: Short
this function takes one or more required stage fileS _plural_, and exits
if any are missing (or at least it does now after the refactor).
let's rename it to make things more clear
Gbp-Dch: Short
the name of the stage is already printed earlier in the output prior to
the error here being printed. so the error really does not need to include
the script name itself.
now having investigated my suspicions of the functionality and use of
Require_stagefile(), i conclude that it has been fundamentally broken
all the way back to v1.0~a8-1 (or at least usage of it since v1.0.1-2).
gah. (╯°□°)╯︵ ┻━┻
----
very early on in the history of live-build this function took the name of
a _single_ stage file only and did `exit 1` should the file not be found.
this was simple and clearly accomplished "what was on the tin", so to
speak.
in bd1a5ddc82 (2007, 1.0~a8-1) things got
weird. it was modified to support a list of multiple files. but instead of
being written to cause failure if _any_ of the listed files were missing
as perhaps one might expect, it was instead written to fail only if all
files were missing!
if you jump to the conclusion that i'm talking about a simple flipped
logic from a lack or otherwise of a `!` here, you'd be mistaken; there is
a comment inside the function that could not be more clear about what was
intended by the author - "Find at least one of the required stages"! this
makes me thoroughly confused about what they were thinking.
as we'll get to, this was fundamentally flawed (or at least its later use
was), but furthermore there were other notable issues introduced at this
point (but don't worry too much about these, they've all been addressed):
- `NAME` was modified in the loop, using the existing value, but nothing
initially set it...
- the setting of `NAME` seems related to its use in the subsequent error
output, yet they are logically separated; it is only set if a file
exists, while the error is only printed if none exist.
- it is pointlessly using a messy `CONTINUE="true"` based mechanism,
when it could just `return 0`.
- it did not handle correctly the bad use case of no params having been
supplied.
it doesn't seem to have been entirely thought through, despite its
pervasive use throughout the build system.
note that no change was made in that commit to make actual use of the
new multi-param support. it would not be used until about a year later.
the function has remained largely untouched since then. in
c68c0a2708 a notable change was made to add
an initial setting of `NAME`, which partially addressed one of the above
issues. but it did not really address the issue the change was meant to
solve, since the `NAME` as printed in the error was now the name of the
script when what was really wanted was the name of the stagefile. this was
finally fixed properly in d54990695f.
however the weirdly pointless setting of `NAME` persisted in the loop.
finally i personally just refactored the function in the commit prior to
this one, retaining the same functionality but addressing the remaining
of the above minor implementation issues.
looking at usage of the new functionality introduced in
bd1a5ddc82, it does not seem to have been
until 0cbbde2b96 (2008, almost a year after
it was made possible) that changes were made to finally start making use
of the ability to pass more than one filename at a time to the function,
and it would appear that perhaps the author forgot what it actually was
that the function accomplished when used with multiple params, and failed
to double check.
in this first use of multiple parameters, this commit went from passing
single file names to individual calls to the function to passing the files
in one single call, in a commit the purpose of which was described as
simply tidying things up. it was most certainly not intended to change
stage requirements.
unfortunately, a change in requirements did occur as a result since the
new usage of the function was not accomplishing the same as before. this
change completely broke the stage requirements protection mechanism such
that only a single one of the listed stages needed to have completed for
the check to pass, rather than all as expected.
this flaw made it into release v1.0.1-2 and it has existed every since.
in the very next commit from that one,
6204dc0e6d things got even worse. here we
see the config stage being specified commonly as the first stage listed,
which is still the case today. this means that ever since this commit,
if you've already got a config before building (which you inevitably do,
especially after some later commits introduced automatically creating it
if missing), then all other stage requirements are simply ignored.
so it seems pretty damn clear that this function is accomplishing
completely the wrong objective. it _should_ be checking that _all_ files
given to it exist, not just one or more. ¯\_(ツ)_/¯
this FINALLY addresses this mistake.
(not that i wish to berate the author; i've made silly mistakes of my own
before)
- count of params is available as $#, we don't need the pipe-to-wc logic.
- the whole 'CONTINUE' based logic is silly, we can just return once one
of the files is found.
- setting of 'NAME' in the loop was completely pointless.
- the error message for multiple files was not very clear just injecting
a sequence of words into a sentence.
- it did not work correctly if no arguments were given (bad usage)
note, you might question whether the functionality of this function is
correct, as did I; this is tackled in a followup commit whilst this
commit retains the existing functionality!
Gbp-Dch: Short
as suggested by Raphaël
rather than have fixed stagefile filename strings at all in the scripts,
use `$(basename $0)` to use the name of the script (which is the same for
almost all cases anyway, and the stage files are supposed to be almost
exclusively unique per-script). we can thus simplify things by determining
the filename for most use cases within the functions themselves.
this does change the file used by a couple of scripts, affecting backwards
compatibility of executing live-build upon an existing partially or fully
completed build:
- binary_grub-pc used "binary_grub"
- chroot_includes used "includes.chroot"
care had to be taken for the following cases:
- there are some cases like bootstrap_cache, source_debian and
bootstrap_debootstrap which are dealing with more than one file, and/or
otherwise a filename that is not specific to the script itself exactly,
or should not be based upon its name.
- some cases like chroot_cache, bootstrap_cache and
chroot_install-packages need to append something to the end of the name
depending upon which pass/action mode the script is being executed with.
- furthermore in the bootstrap_cache case one of the filenames is used
within the bootstrap_debootstrap and thus needs very careful handling
to be certain that a change in filename of bootstrap_cache does not
break bootstrap_debootstrap.
Gbp-Dch: Short
- avoid all need to pass ".build/" path in stage file names into the
functions
- add a helper to remove a stage file (required to complete the above
properly)
- avoid duplicating filenames within scripts which makes them prone to
mistakes (some instances of which I've actually encountered and had
to fix)
Gbp-Dch: Short
Ryan Finnie
Lyndon Brown
johnraff
Raphaël Hertzog
Raphaël Hertzog