Joey Hess [Tue, 18 Mar 2025 17:55:33 +0000 (13:55 -0400)]
checkPresent of compute remote checks inputs are available
If an input file has been lost from all repositories, it is no longer
possible to compute the output. This will avoid dropping content that
was computed in such a situation, as well as making git-annex fsck --from
the compute remote do its usual thing when content has gone missing.
This implementation avoids recursing forever if there is a cycle,
which should not be possible anyway.
Note the use of RemoteStateHandle as a constructor here suggests that
this may not handle sameas remotes right, since usually a
RemoteStateHandle is constructed using the sameas uuid for a sameas
remote. That assumes a compute remote can even have or be a sameas remote.
Which doesn't seem to make sense, so I have not thought through what might
happen here in detail.
Joey Hess [Mon, 17 Mar 2025 18:27:56 +0000 (14:27 -0400)]
decided addcomputed will not support annex.smallfiles
If it did, recompute would need to somehow support recomputing
non-annexed files.
And, annex.smallfiles is typically used for configuration files or
source code kind of things, where the user doesn't want it to be an
annexed file. Computed artifacts are not likely that kind of thing.
Also, git-annex importfeed is an example of something that does support
annex.addunlocked, but does not support annex.smallfiles.
Joey Hess [Wed, 12 Mar 2025 17:36:16 +0000 (13:36 -0400)]
recompute: stage new version of file in git
When writing doc/tips/computing_annexed_files.mdwn, I noticed
that a recompute --reproducible followed by a drop and a re-get did not
actually test if the file could be reproducible computed again.
Turns out that get and drop both operate on staged files. If there is an
unstaged modification in the work tree, that's ignored. Somewhat
surprisingly, other commands like info do operate on staged files. So
behavior is inconsistent, and fairly surprising really, when there are
unstaged modifications to files.
Probably this is rarely noticed because `git-annex add` is used to add a
new version of a file, and then it's staged. Or `git mv` is used to move
a file, rather than `mv` of a file over top of an existing file. So it's
uncommon to have an unstaged annexed file in a worktree.
It might be worth making things more consistent, but that's out of scope
for what I'm working on currently.
Also, I anticipate that supporting unlocked files with recompute will
require it to stage changes anyway.
So, make recompute stage the new version of the file.
I considered having recompute refuse to overwrite an existing staged
file. After all, whatever version was staged before will get lost when
the new version is staged over top of it. But, that's no different than
`git-annex addcomputed` being run with the name of an existing staged
file. Or `git-annex add` being run with a new file content when there is
an existing staged file. Or, for that matter, `git add` being ran with a
new content when there is an existing staged file.
Joey Hess [Tue, 11 Mar 2025 16:40:21 +0000 (12:40 -0400)]
buffer responses to compute programs in a TQueue
This avoids a potential problem where the program sends several INPUT
before reading responses, so flushing the respose to the pipe could
block. It's unlikely, but seemed worth making sure it can't happen.
Joey Hess [Tue, 11 Mar 2025 15:49:47 +0000 (11:49 -0400)]
avoid error on missing compute state in checkKey
This improves eg `git-annex move --to` a compute remote that does not
contain the key. Rather than erroring with "Missing compute state" when
it checks if the key is in the remote, it proceeds to trying to store to
it, which has a nice error message.
Joey Hess [Tue, 11 Mar 2025 15:46:31 +0000 (11:46 -0400)]
add INPUT-REQUIRED
Used by git-annex-compute-singularity to make addcomputed --fast work.
Also, simplified git-annex-compute-singularity; there is no need to hard
link the container into place. singularity does not care about the
extension of the container, so can just pass it the annex object file.
Joey Hess [Mon, 10 Mar 2025 16:52:10 +0000 (12:52 -0400)]
compute: disallow output files that are not regular files
Use case where this came up is a compute program using singularity,
where the process inside the container will be allowed to write to the temp
directory, so could make eg a /etc/shadow symlink, which could then be
used to exfiltrate that from the system to wherever the annex object
might be pushed to.
It seemed better to fix this once in git-annex rather than in any such
compute program.
Joey Hess [Fri, 7 Mar 2025 18:47:34 +0000 (14:47 -0400)]
compute: add response to OUTPUT
This allows rejecting output filenames that are outside the repository,
and also handles converting eg "-foo" to "./-foo" to prevent a command
that it's passed to interpreting the output filename as a dashed option.
Joey Hess [Tue, 4 Mar 2025 18:54:13 +0000 (14:54 -0400)]
safer git sha object filename
Rather than use the filename provided by INPUT, which could come from user
input, and so could be something that looks like a dashed parameter,
use a .git/object/<sha> filename.
This avoids user input passing through INPUT and back out, with the file
path then passed to a command, which could do something unexpected with
a dashed parameter, or other special parameter.
Added a note in the design about being careful of passing user input to
commands. They still have to be careful of that in general, just not in
this case.
Joey Hess [Tue, 4 Mar 2025 15:06:58 +0000 (11:06 -0400)]
compute remote: get input files from other remotes
This needed some refactoring to avoid cycles, since Remote.Compute
cannot import Remote.List. Instead, it uses Annex.remotes. Which must be
populated by something else, but we know it has been, because something
is using Remote.Compute, which it must have found in the remote list,
which populates that.
In Remote.Compute, keyPossibilities' is called with all loggedLocations,
without the trustExclude DeadTrusted that keyLocations does. There is
another cycle there. This may be a problem if a dead repository is still
a remote.
This is missing cycle prevention, and it's certianly possible to make 2
files in the compute remote co-depend on one-another. Hopefully not in a
real world situation, but it an attacker could certainly do it. Cycle
prevention will need to be added to this.
Joey Hess [Mon, 3 Mar 2025 20:07:04 +0000 (16:07 -0400)]
rename config to annex.security.allowed-compute-programs
And require for enable as well as autoenable.
It seemed asking for trouble for `git-annex enable foo` to use whatever
compute program is stored in the git config, without verifying that the
user wants that program to be used.
Note that it would be good to allow `git-annex enable foo program=...`
to be used without the program being in the git config. Not implemented yet
though.
Joey Hess [Mon, 3 Mar 2025 19:47:09 +0000 (15:47 -0400)]
autoenable security for compute special remote
Added annex.security.autoenable-compute-programs and only allow
autoenabling special remotes that use compute programs on that list.
The reason this is needed is a user might have some compute programs
that are less safe to use than others. They might want to use an unsafe
one only with one repository, where they are the only committer or other
committers are trusted. They might be ok with others being used by any
repository, and if so they can add them to the list.
Another reason would be a user who has installed a compute program by
accident. Eg, it might be included with git-annex at some point, or
pulled in by some dependency. That user doesn't necessarily want that
compute program to be used in an autoenabled special remote.
Joey Hess [Mon, 3 Mar 2025 15:59:04 +0000 (11:59 -0400)]
support git files as input to computations
Using GIT keys, like are used when exporting git files to special
remotes. Except here the GIT key refers to a file checked into the git
repo.
Note that, since the compute remote uses catObject to get the content,
a symlink that is checked into git does not get followed. This is important
for security, because following a symlink and adding the content to the
repo as an annex object would allow exfiltrating content from outside
the repository.
Instead, the behavior with a symlink is to run the computation on the
symlink target. This may turn out to be confusing, and it might be worth
addcomputed checking if the file in git is a symlink and erroring out.
Or it could follow symlinks as long as the destination is a file in the
repisitory.
Joey Hess [Thu, 27 Feb 2025 20:19:41 +0000 (16:19 -0400)]
record VURL key hashes when getting from compute remote
Like when getting from the web special remote, when the output of the
computation has changed, record the new hash of the content as an
equivilant key for the VURL key.
Still needs to be done for addcomputed and recompute.
Joey Hess [Thu, 27 Feb 2025 19:12:29 +0000 (15:12 -0400)]
many recompute improvements
I've lost track of them all, but it includes:
* Using the same key backend as was used in the original computation.
* Fixing bug that prevented updating the source file key in the compute
state
* Handling --reproducible and --unreproducible.
* recompute --original of a file using VURL, when the result is
different, but the key remains the same, makes the object file
be updated with the new content
* Detecting some other ways the program behavior can change, just for
completeness.
* Also adds --backend to addcomputed.