There already existed a smoke test for the link content length,
but it appears that there exists some corruptions pernicious enough
to replace the file content with zeros, and keeping the same length.
--repair-path now goes as far as checking the content of the link,
making it true to its name and actually repairing the path for such
coruption cases.
This was already accidentally disabled in ba87b08. It also no longer
appears to be beneficial, and in fact slow things down, e.g. when
evaluating a NixOS system configuration:
elapsed time: median = 3.8170 mean = 3.8202 stddev = 0.0195 min = 3.7894 max = 3.8600 [rejected, p=0.00000, Δ=0.36929±0.02513]
Add a `_NIX_TRACE_BUILT_OUTPUTS` environment variable that can be set to
a filename in which the result of each build will be logged.
This is intentionally crude and undocumented as it’s only meant to be a
temporary thing to assess the usefulness of CA derivations.
Any other use would need a cleaner re-implementation first.
Make the build of unresolved derivations return the same status as the
resolved one, except in the case of an `AlreadyValid` in which case it
will return `ResolvesToAlreadyValid` to mean that the outputs of the unresolved
derivation weren’t known, but the resolved one is.
I downloaded Nix tonight, and immediately broke it by accidentally removing the default binary caching.
After figuring this out, I also failed to fix it properly, due to using the wrong key for Nix's default binary cache
If the diagnostic message would have been clearer about what/where a "signature" for a "substituter" is + comes from, it probably would have saved me a few hours.
Maybe we can save other noobs the same pain?
Because the manual is generated from default values which are themselves
generated from various sources (cpuid, bios settings (kvm), number of
cores). This commit hides non-reproducible settings from the manual
output.
No matter what, we need to resize the buffer to not have any scratch
space after we do the `read`. In the end of file case, `got` will be 0
from it's initial value.
Before, we forgot to resize in the EOF case with the break. Yes, we know
we didn't recieve any data in that case, but we still have the scatch
space to undo.
Co-Authored-By: Will Fancher <Will.Fancher@Obsidian.Systems>
This doesn't fix the bug, but makes the code less difficult to read.
Also improve the comments, now that it is clear what part is needed in
each code path.
For a typical desktop system (~2K packages) we can easily get 100K
entries in RealisationsRefs. Without indices query for RealisationsRefs
requires linear scan.
RealisationsRefs(referrer)
--------------------------
Inefficiency is seen as a 100% CPU load of nix-daemon for the following
scenario:
$ nix edit -f . bash # add unused environment variable, like FOO="1"
# populate RealisationsRefs, build fresh system
$ nix build -f nixos system --arg config '{ contentAddressedByDefault = true; }'
$ nix edit -f . bash # add unused environment variable, like FOO="2"
$ time nix build -f nixos system --arg config '{ contentAddressedByDefault = true; }'
In this case `bash `will be rebuilt a few times and then rest of CPU
time is spent on scanning RealisationsRefs table (about 5 CPU-minutes
on my machine).
Before the change:
$ time nix build -f nixos system ... # step 4 above
real 34m3,613s
user 0m5,232s
sys 0m0,758s
Of all this time about 29.5 minutes are taken by nix-daemon's CPU time.
After the change:
$ time nix build -f nixos system ... # step 4 above
real 4m50,061s
user 0m5,038s
sys 0m0,677s
Of all this time about 1 minute is taken by nix-daemon's CPU time.
Most of the time is spent polling for non-existent realisations on
cache-nixos.org.
Realisations(outputPath)
------------------------
After running CA system for two weeks I got ~1M entries in Realisations
table. `nix-collect-garbage` became very slow (seemingly 100 path deletions
per second). It happens due to a slow cascading delete from Realisations
triggered by deletion from ValidPaths.
The fix is to add an index on primary key from ValidPaths(id) that
triggers cascading deletions.
Before the change:
$ time nix-collect-garbage -d --max-freed 100G
<interrupted before finish, took too long>
real 23m32.411s
user 17m49.679s
sys 4m50.609s
Most of time was spent in re-scanning Realisations table on each path deletion.
After the change:
$ time nix-collect-garbage -d --max-freed 100G
real 8m43.226s
user 6m16.317s
sys 1m40.188s
Time is spent scanning sqlite indices and in kernel when unlinking directories.
Doing it as a side-effect of calling LocalStore::makeStoreWritable()
is very ugly.
Also, make sure that stopping the progress bar joins the update
thread, otherwise that thread should be unshared as well.
Since 4806f2f6b0, we can't have paths with
references passed to builtins.{path,filterSource}. This prevents many cases
of those functions called on IFD outputs from working. Resolve this by
passing the references found in the original path to the added path.
When setting flake-local options (with the `nixConfig` field), forward
these options to the daemon in case we’re using one.
This is necessary in particular for options like `binary-caches` or
`post-build-hook` to make sense.
Fix <343239fc8a (r44356843)>
Rather than having them plain strings scattered through the whole
codebase, create an enum containing all the known experimental features.
This means that
- Nix can now `warn` when an unkwown experimental feature is passed
(making it much nicer to spot typos and spot deprecated features)
- It’s now easy to remove a feature altogether (once the feature isn’t
experimental anymore or is dropped) by just removing the field for the
enum and letting the compiler point us to all the now invalid usages
of it.
Currently machine specification (`/etc/nix/machine`) parser fails
with a vague exception if the file had incorrect format.
This commit adds verbose exceptions and unit-tests for the parser.
This ensures any started processes can't write to /nix/store (except
during builds). This partially reverts 01d07b1e, which happened because
of #2646.
The problem was only happening after nix downloads anything, causing
me to suspect the download thread. The problem turns out to be:
"A process can't join a new mount namespace if it is sharing
filesystem-related attributes with another process", in this case this
process is the curl thread.
Ideally, we might kill it before spawning the shell process, but it's
inside a static variable in the getFileTransfer() function. So
instead, stop it from sharing FS state using unshare(). A strategy
such as the one from #5057 (single-threaded chroot helper binary) is
also very much on the table.
Fixes#4337.
This fixes a bug in the garbage collector where if a path
/nix/store/abcd-foo is valid, but we do a
isValidPath("/nix/store/abcd-foo.lock") first, then a negative entry
for /nix/store/abcd is added to pathInfoCache, so /nix/store/abcd-foo
is subsequently considered invalid and deleted.
(where "referrers" includes the reverse of derivation outputs and
derivers). Now we do a full traversal to look if we can reach any
root. If not, all paths reached can be deleted.
The garbage collector no longer blocks other processes from
adding/building store paths or adding GC roots. To prevent the
collector from deleting store paths just added by another process,
processes need to connect to the garbage collector via a Unix domain
socket to register new temporary roots.
This reverts some parts of commit
8430a8f086 which was trying to rethrow
some exceptions while we weren’t in the context of a `catch` block,
causing some weird “terminate called without an active exception”
errors.
Fix#5368
In https://github.com/NixOS/nix/pull/5350 we noticed link failures
pkgsStatic.nixUnstable. Adding explicit dependency on libutil fixes
libstore-tests linking.
When I stop a download with Ctrl-C in a `nix repl` of a flake, the REPL
refuses to do any other downloads:
nix-repl> builtins.getFlake "nix-serve"
[0.0 MiB DL] downloading 'https://api.github.com/repos/edolstra/nix-serve/tarball/e9828a9e01a14297d15ca41 error: download of 'e9828a9e01' was interrupted
[0.0 MiB DL]
nix-repl> builtins.getFlake "nix-serve"
error: interrupted by the user
[0.0 MiB DL]
To fix this issue, two changes were necessary:
* Reset the global `_isInterrupted` variable: only because a single
operation was aborted, it should still be possible to continue the
session.
* Recreate a `fileTransfer`-instance if the current one was shut down by
an abort.
9c766a40cb broke logging from the
daemon, because commonChildInit is called when starting the build hook
in a vfork, so it ends up resetting the parent's logger. So don't
vfork.
It might be best to get rid of vfork altogether, but that may cause
problems, e.g. when we call an external program like git from the
evaluator.
Before the changes when building the whole system with
`contentAddressedByDefault = true;` we get many noninformative messages:
$ nix build -f nixos system --keep-going
...
warning: rewriting hashes in '/nix/store/...-clang-11.1.0.drv.chroot/nix/store/...-11.1.0'; cross fingers
warning: rewriting hashes in '/nix/store/...-clang-11.1.0.drv.chroot/nix/store/...-11.1.0-dev'; cross fingers
warning: rewriting hashes in '/nix/store/...-clang-11.1.0.drv.chroot/nix/store/...-11.1.0-python'; cross fingers
error: 2 dependencies of derivation '/nix/store/...-hub-2.14.2.drv' failed to build
warning: rewriting hashes in '/nix/store/...-subversion-1.14.1.drv.chroot/nix/store/...-subversion-1.14.1-dev'; cross fingers
warning: rewriting hashes in '/nix/store/...-subversion-1.14.1.drv.chroot/nix/store/...-subversion-1.14.1-man'; cross fingers
...
Let's downgrade these messages down to debug().
I had started the trend of doing `std::visit` by value (because a type
error once mislead me into thinking that was the only form that
existed). While the optomizer in principle should be able to deal with
extra coppying or extra indirection once the lambdas inlined, sticking
with by reference is the conventional default. I hope this might even
improve performance.
This actually bit me quite recently in `nixpkgs` because I assumed that
`nix-build --check` would also error out if hashes don't match anymore[1]
and so I wrongly assumed that I couldn't reproduce the mismatch error.
The fix is rather simple, during the output registration a so-called
`delayedException` is instantiated e.g. if a FOD hash-mismatch occurs.
However, in case of `nix-build --check` (or `--rebuild` in case of `nix
build`), the code-path where this exception is thrown will never be
reached.
By adding that check to the if-clause that causes an early exit in case
of `bmCheck`, the issue is gone. Also added a (previously failing)
test-case to demonstrate the problem.
[1] https://github.com/NixOS/nixpkgs/pull/139238, the underlying issue
was that `nix-prefetch-git` returns different hashes than `fetchgit`
because the latter one fetches submodules by default.
This is important if the remote side *does* execute
nix-store/nix-daemon successfully, but stdout is polluted
(e.g. because the remote user's bashrc script prints something to
stdout). In that case we have to shutdown the write side to force the
remote nix process to exit.
Instead of
error: serialised integer 7161674624452356180 is too large for type 'j'
we now get
error: 'nix-store --serve' protocol mismatch from 'sshtest@localhost', got 'This account is currently not available.'
Fixes https://github.com/NixOS/nixpkgs/issues/37287.
Before this commit, the dns lookup in preloadNSS would still go through
nscd. This did not have the effect of loading the nss_dns.so as expected
(nss_dns.so being out of reach from within the sandbox).
Should LOCALDOMAIN environment variable be defined, nss will completely
avoid nscd and will do its dns resolution on its own.
By temporarly setting LOCALDOMAIN variable before calling in NSS, we can
force NSS to load the shared libraries as expected.
Fixes#5089
Signed-off-by: Arthur Gautier <baloo@superbaloo.net>
Store paths are only allowed to contain a limited subset of the
alphabet, which doesn’t include `!`. So don’t create lockfiles that
contain this `!` character as that would otherwise confuse (and break)
the gc.
Fix#5176
When doing e.g.
nix-build -A package --keep-failed --option \
builders \
'ssh://mfhydra?remote-store=/home/bosch/store x86_64-linux - 10 4 big-parallel'
this doesn't work properly because this build-setting is ignored.
I changed this behavior by passing the `settings.keepFailed` through the
serve-protocol to remote machines to make sure that I can introspect the
build-directory (which is particularly helpful when I have to look at a
`config.log` from a failed build for instance).
Use `$(libdir)` while installing .pc files looks like a more generic
solution. For example, it will work for distributions like RHEL or
Fedora where .pc files are installed in `/usr/lib64/pkgconfig`.
This replaces the O(n) search complexity in our insert code with a
lookup of O(log n). It also makes removing waitees easier as we can use
the extract method provided by the set class.
Previously the code ensures that the isBase32 array would only be
initialised once in a single-threaded context. If two threads happen to
call the function before the initialisation was completed both of them
would have completed the initialization step. This allowed for a
race-condition where one thread might be done with the initialization
but the other thread sets all the fields to false again. For a brief
moment the base32 detection would then produce false-negatives.
The experimental features are, well, experimental, and shouldn’t be
carelessly and transparently enabled.
Besides, some (`ca-derivations` at least) need to be enabled at startup
in order to work properly.
So it’s better to just require that daemon be started with the right
`experimental-features` option.
Fix#5017
This adds a new store operation 'addMultipleToStore' that reads a
number of NARs and ValidPathInfos from a Source, allowing any number
of store paths to be copied in a single call. This is much faster on
high-latency links when copying a lot of small files, like .drv
closures.
For example, on a connection with an 50 ms delay:
Before:
$ nix copy --to 'unix:///tmp/proxy-socket?root=/tmp/dest-chroot' \
/nix/store/90jjw94xiyg5drj70whm9yll6xjj0ca9-hello-2.10.drv \
--derivation --no-check-sigs
real 0m57.868s
user 0m0.103s
sys 0m0.056s
After:
real 0m0.690s
user 0m0.017s
sys 0m0.011s
With this, we don't have to copy the entire .drv closure to the
destination store ahead of time (or at all). Instead, buildPaths()
reads .drv files from the eval store and copies inputSrcs to the
destination store if it needs to build a derivation.
Issue #5025.
In particular, this now works:
$ nix path-info --eval-store auto --store https://cache.nixos.org nixpkgs#hello
Previously this would fail as it would try to upload the hello .drv to
cache.nixos.org. Now the .drv is instantiated in the local store, and
then we check for the existence of the outputs in cache.nixos.org.
Some people want to avoid using registries at all on their system; Instead
of having to add --no-registries to every command, this commit allows to
set use-registries = false in the config. --no-registries is still allowed
everywhere it was allowed previously, but is now deprecated.
Co-authored-by: Eelco Dolstra <edolstra@gmail.com>
- This can legitimately happen (for example because of a non-determinism
causing a build-time dependency to be kept or not as a runtime
reference)
- Because of older Nix versions, it can happen that we encounter a
realisation with an (erroneously) empty set of dependencies, in which
case we don’t want to fail, but just warn the user and try to fix it.
Fill `NIX_CONFIG` with the value of the current Nix configuration before
calling the post-build-hook.
That way the whole configuration (including the possible
`experimental-features`, a possibly `--store` option or whatever) will
be made available to the hook
'--delete-older-than 10' deletes the generations older than a single day, and '--delete-older-than 12m' deletes all generations older than 12 days.
This changes makes it throw on those invalid inputs, and gives an example of a valid input.
Add an access-control list to the realisations in recursive-nix (similar
to the already existing one for store paths), so that we can build
content-addressed derivations in the restricted store.
Fix#4353
Previously, the build system used uname(1) output when it wanted to
check the operating system it was being built for, which meant that it
didn't take into-account cross-compilation when the build and host
operating systems were different.
To fix this, instead of consulting uname output, we consult the host
triple, specifically the third "kernel" part.
For "kernel"s with stable ABIs, like Linux or Cygwin, we can use a
simple ifeq to test whether we're compiling for that system, but for
other platforms, like Darwin, FreeBSD, or Solaris, we have to use a
more complicated check to take into account the version numbers at the
end of the "kernel"s. I couldn't find a way to just strip these
version numbers in GNU Make without shelling out, which would be even
more ugly IMO. Because these checks differ between kernels, and the
patsubst ones are quite fiddly, I've added variables for each host OS
we might want to check to make them easier to reuse.
This way no derivation has to expect that these files are in the `cwd`
during the build. This is problematic for `nix-shell` where these files
would have to be inserted into the nix-shell's `cwd` which can become
problematic with e.g. recursive `nix-shell`.
To remain backwards-compatible, the location inside the build sandbox
will be kept, however using these files directly should be deprecated
from now on.
This is needed to push the adoption of structured attrs[1] forward. It's
now checked if a `__json` exists in the environment-map of the derivation
to be openend in a `nix-shell`.
Derivations with structured attributes enabled also make use of a file
named `.attrs.json` containing every environment variable represented as
JSON which is useful for e.g. `exportReferencesGraph`[2]. To
provide an environment similar to the build sandbox, `nix-shell` now
adds a `.attrs.json` to `cwd` (which is mostly equal to the one in the
build sandbox) and removes it using an exit hook when closing the shell.
To avoid leaking internals of the build-process to the `nix-shell`, the
entire logic to generate JSON and shell code for structured attrs was
moved into the `ParsedDerivation` class.
[1] https://nixos.mayflower.consulting/blog/2020/01/20/structured-attrs/
[2] https://nixos.org/manual/nix/unstable/expressions/advanced-attributes.html#advanced-attributes
Useful when we're using a daemon with a chroot store, e.g.
$ NIX_DAEMON_SOCKET_PATH=/tmp/chroot/nix/var/nix/daemon-socket/socket nix-daemon --store /tmp/chroot
Then the client can now connect with
$ nix build --store unix:///tmp/chroot/nix/var/nix/daemon-socket/socket?root=/tmp/chroot nixpkgs#hello
That doesn’t really make sense with CA derivations (and wasn’t even
really correct before because of FO derivations, though that probably
didn’t matter much in practice)
Make ca-derivations require a `ca-derivations` machine feature, and
ca-aware builders expose it.
That way, a network of builders can mix ca-aware and non-ca-aware
machines, and the scheduler will send them in the right place.
When adding a path to the local store (via `LocalStore::addToStore`),
ensure that the `ca` field of the provided `ValidPathInfo` does indeed
correspond to the content of the path.
Otherwise any untrusted user (or any binary cache) can add arbitrary
content-addressed paths to the store (as content-addressed paths don’t
need a signature).
Linux is (as far as I know) the only mainstream operating system that
requires linking with libdl for dlopen. On BSD, libdl doesn't exist,
so on non-FreeBSD BSDs linking will currently fail. On macOS, it's
apparently just a symlink to libSystem (macOS libc), presumably
present for compatibility with things that assume Linux.
So the right thing to do here is to only add -ldl on Linux, not to add
it for everything that isn't FreeBSD.
Only considers the closure in term of `Realisation`, ignores all the
opaque inputs.
Dunno whether that’s the nicest solution, need to think it through a bit
Align all the worker protocol with `buildDerivation` which inlines the
realisations as one opaque json blob.
That way we don’t have to bother changing the remote store protocol
when the definition of `Realisation` changes, as long as we keep the
json backwards-compatible
Align all the worker protocol with `buildDerivation` which inlines the
realisations as one opaque json blob.
That way we don’t have to bother changing the remote store protocol
when the definition of `Realisation` changes, as long as we keep the
json backwards-compatible
Move the `closure` logic of `computeFSClosure` to its own (templated) function.
This doesn’t bring much by itself (except for the ability to properly
test the “closure” functionality independently from the rest), but it
allows reusing it (in particular for the realisations which will require
a very similar closure computation)
For whatever reason, many programs trying to access SystemVersion.plist
also open SystemVersionCompat.plist; this includes Python code and
coreutils’ `cat(1)` (but not the native macOS `/bin/cat`). Illustratory
`dtruss(1m)` output:
open("/System/Library/CoreServices/SystemVersion.plist\0", 0x0, 0x0) = 3 0
open("/System/Library/CoreServices/SystemVersionCompat.plist\0", 0x0, 0x0) = 4 0
I assume this is a Big Sur change relating to the 10.16.x/11.x
version compatibility divide and that it’s something along the lines of
a hook inside libSystem.
Fixes a lot of sandboxed package builds under Big Sur.
When we don’t have enough free job slots to run a goal, we put it in
the waitForBuildSlot list & unlock its output locks. This will
continue from where we left off (tryLocalBuild). However, we need the
locks to get reacquired when/if the goal ever restarts. So, we need to
send it back through tryToBuild to get reqacquire those locks.
I think this bug was introduced in
https://github.com/NixOS/nix/pull/4570. It leads to some builds
starting without proper locks.
Similar to the nar-info disk cache (and using the same db).
This makes rebuilds muuch faster.
- This works regardless of the ca-derivations experimental feature.
I could modify the logic to not touch the db if the flag isn’t there,
but given that this is a trash-able local cache, it doesn’t seem to be
really worth it.
- We could unify the `NARs` and `Realisation` tables to only have one
generic kv table. This is left as an exercise to the reader.
- I didn’t update the cache db version number as the new schema just
adds a new table to the previous one, so the db will be transparently
migrated and is backwards-compatible.
Fix#4746
I guess I misunderstood John's initial explanation about why wildcards
for outputs are sent to older stores[1]. My `nix-daemon` from 2021-03-26
also has version 1.29, but misses the wildcard[2]. So bumping seems to
be the right call.
[1] https://github.com/NixOS/nix/pull/4759#issuecomment-830812464
[2] 255d145ba7
Starting in macOS 11, the on-disk dylib bundles are no longer available,
but nixpkgs needs to be able to keep compatibility with older versions
that require `/usr/lib/libSystem.B.dylib` in `__impureHostDeps`. Allow
it to keep backwards compatibility with these versions by marking these
dependencies as optional.
Fixes#4658.
They are equivalent according to
<https://spec.commonmark.org/0.29/#hard-line-breaks>,
and the trailing spaces tend to be a pain (because the make git
complain, editors tend to want to remove them − the `.editorconfig`
actually specifies that − etc..).
If there were many top-level goals (which are not destroyed until the
very end), commands like
$ nix copy --to 'ssh://localhost?remote-store=/tmp/nix' \
/run/current-system --no-check-sigs --substitute-on-destination
could fail with "Too many open files". So now we do some explicit
cleanup from amDone(). It would be cleaner to separate goals from
their temporary internal state, but that would be a bigger refactor.
This avoids an ambiguity where the `StorePathWithOutputs { drvPath, {}
}` could mean "build `brvPath`" or "substitute `drvPath`" depending on
context.
It also brings the internals closer in line to the new CLI, by
generalizing the `Buildable` type is used there and makes that
distinction already.
In doing so, relegate `StorePathWithOutputs` to being a type just for
backwards compatibility (CLI and RPC).
These are by no means part of the notion of a store, but rather are
things that happen to use stores. (Or put another way, there's no way
we'd make them virtual methods any time soon.) It's better to move them
out of that too-big class then.
Also, this helps us remove StorePathWithOutputs from the Store interface
altogether next commit.
A few versioning mistakes were corrected:
- In 27b5747ca7, Daemon protocol had some
version `>= 0xc` that should have been `>= 0x1c`, or `28` since the
other conditions used decimal.
- In a2b69660a9, legacy SSH gated new CAS
info on version 6, but version 5 in the server. It is now 6
everywhere.
Additionally, legacy ssh was sending over more metadata than the daemon
one was. The daemon now sends that data too.
CC @regnat
Co-authored-by: Cole Helbling <cole.e.helbling@outlook.com>
I guess the rationale behind the old name wath that
`pathInfoIsTrusted(info)` returns `true` iff we would need to `blindly`
trust the path (because it has no valid signature and `requireSigs` is
set), but I find it to be a really confusing footgun because it's quite
natural to give it the opposite meaning.
What happened was that Nix was trying to unconditionally mount these
paths in fixed-output derivations, but since the outer derivation was
pure, those paths did not exist. The solution is to only mount those
paths when they exist.
This separates the scheduling logic (including simple hook pathway) from
the local-store needing code.
This should be the final split for now. I'm reasonably happy with how
it's turning out, even before I'm done moving code into
`local-derivation-goal`. Benefits:
1. This will help "witness" that the hook case is indeed a lot simpler,
and also compensate for the increased complexity that comes from
content-addressed derivation outputs.
2. It also moves us ever so slightly towards a world where we could use
off-the-shelf storage or sandboxing, since `local-derivation-goal`
would be gutted in those cases, but `derivation-goal` should remain
nearly the same.
The new `#if 0` in the new files will be deleted in the following
commit. I keep it here so if it turns out more stuff can be moved over,
it's easy to do so in a way that preserves ordering --- and thus
prevents conflicts.
N.B.
```sh
git diff HEAD^^ --color-moved --find-copies-harder --patience --stat
```
makes nicer output.
This is already used by Hydra, and is very useful when materializing
a remote builder list from service discovery. This allows the service
discovery tool to only sync one file instead of two.
This is technically a breaking change, since attempting to set plugin
files after the first non-flag argument will now throw an error. This
is acceptable given the relative lack of stability in a plugin
interface and the need to tie the knot somewhere once plugins can
actually define new subcommands.
This field used to be a `BasicDerivation`, but this `BasicDerivation`
was downcasted to a `Derivation` when needed (implicitely or not), so we
might as well make it a full `Derivation` and upcast it when needed.
This also allows getting rid of a weird duplication in the way we
compute the static output hashes for the derivation. We had to
do it differently and in a different place depending on whether the
derivation was a full derivation or just a basic drv, but we can now do
it unconditionally on the full derivation.
Fix#4559
- Pass it the name of the outputs rather than their output paths (as
these don't exist for ca derivations)
- Get the built output paths from the remote builder
- Register the new received realisations
When performing distributed builds of machine learning packages, it
would be nice if builders without the required SIMD instructions can
be excluded as build nodes.
Since x86_64 has accumulated a large number of different instruction
set extensions, listing all possible extensions would be unwieldy.
AMD, Intel, Red Hat, and SUSE have recently defined four different
microarchitecture levels that are now part of the x86-64 psABI
supplement and will be used in glibc 2.33:
https://gitlab.com/x86-psABIs/x86-64-ABIhttps://lwn.net/Articles/844831/
This change uses libcpuid to detect CPU features and then uses them to
add the supported x86_64 levels to the additional system types. For
example on a Ryzen 3700X:
$ ~/aps/bin/nix -vv --version | grep "Additional system"
Additional system types: i686-linux, x86_64-v1-linux, x86_64-v2-linux, x86_64-v3-linux
That way we
1. Don't have to recompute them several times
2. Can compute them in a place where we know the type of the parent
derivation, meaning that we don't need the casting dance we had before
Once a build is done, get back to the original derivation, and register
all the newly built outputs for this derivation.
This allows Nix to work properly with derivations that don't have all
their build inputs available − thus allowing garbage collection and
(once it's implemented) binary substitution
In addition to being some ugly template trickery, it was also totally
useless as it was used in only one place where I could replace it by
just a few extra characters
Where a `RealisedPath` is a store path with its history, meaning either
an opaque path for stuff that has been directly added to the store, or a
`Realisation` for stuff that has been built by a derivation
This is a low-level refactoring that doesn't bring anything by itself
(except a few dozen extra lines of code :/ ), but raising the
abstraction level a bit is important on a number of levels:
- Commands like `nix build` have to query for the realisations after the
build is finished which is fragile (see
27905f12e4a7207450abe37c9ed78e31603b67e1 for example). Having them
oprate directly at the realisation level would avoid that
- Others like `nix copy` currently operate directly on (built) store
paths, but need a bit more information as they will need to register
the realisations on the remote side
Fix a mismatch in the errors thrown when a needed output was missing
from an input derivation that was leading to a wrong and quite misleading error
message
Don't only show the name of the output, but also the derivation to which
this output belongs (as otherwise it's very hard to track back what went
wrong)
Changes:
* The divider lines are gone. These were in practice a bit confusing,
in particular with --show-trace or --keep-going, since then there
were multiple lines, suggesting a start/end which wasn't the case.
* Instead, multi-line error messages are now indented to align with
the prefix (e.g. "error: ").
* The 'description' field is gone since we weren't really using it.
* 'hint' is renamed to 'msg' since it really wasn't a hint.
* The error is now printed *before* the location info.
* The 'name' field is no longer printed since most of the time it
wasn't very useful since it was just the name of the exception (like
EvalError). Ideally in the future this would be a unique, easily
googleable error ID (like rustc).
* "trace:" is now just "…". This assumes error contexts start with
something like "while doing X".
Example before:
error: --- AssertionError ---------------------------------------------------------------------------------------- nix
at: (7:7) in file: /home/eelco/Dev/nixpkgs/pkgs/applications/misc/hello/default.nix
6|
7| x = assert false; 1;
| ^
8|
assertion 'false' failed
----------------------------------------------------- show-trace -----------------------------------------------------
trace: while evaluating the attribute 'x' of the derivation 'hello-2.10'
at: (192:11) in file: /home/eelco/Dev/nixpkgs/pkgs/stdenv/generic/make-derivation.nix
191| // (lib.optionalAttrs (!(attrs ? name) && attrs ? pname && attrs ? version)) {
192| name = "${attrs.pname}-${attrs.version}";
| ^
193| } // (lib.optionalAttrs (stdenv.hostPlatform != stdenv.buildPlatform && !dontAddHostSuffix && (attrs ? name || (attrs ? pname && attrs ? version)))) {
Example after:
error: assertion 'false' failed
at: (7:7) in file: /home/eelco/Dev/nixpkgs/pkgs/applications/misc/hello/default.nix
6|
7| x = assert false; 1;
| ^
8|
… while evaluating the attribute 'x' of the derivation 'hello-2.10'
at: (192:11) in file: /home/eelco/Dev/nixpkgs/pkgs/stdenv/generic/make-derivation.nix
191| // (lib.optionalAttrs (!(attrs ? name) && attrs ? pname && attrs ? version)) {
192| name = "${attrs.pname}-${attrs.version}";
| ^
193| } // (lib.optionalAttrs (stdenv.hostPlatform != stdenv.buildPlatform && !dontAddHostSuffix && (attrs ? name || (attrs ? pname && attrs ? version)))) {
This change is to simplify [Trustix](https://github.com/tweag/trustix) indexing and makes it possible to reconstruct this URL regardless of the compression used.
In particular this means that 7c2e9ca597/contrib/nix/nar/nar.go (L61-L71) can be removed and only the bits that are required to establish trust needs to be published in the Trustix build logs.
With the `ca-derivation` experimental features, non-ca derivations used
to have their output paths returned as unknown as long as they weren't
built (because of a mistake in the code that systematically erased the
previous value)
Thanks @regnat and @edolstra for catching this and comming up with the
solution.
They way I had generalized those is wrong, because local settings for
non-local stores is confusing default. And due to the nature of C++
inheritance, fixing the defaults is more annoying than it should be.
Additionally, I thought we might just drop the check in the substitution
logic since `Store::addToStore` is now streaming, but @regnat rightfully
pointed out that as it downloads dependencies first, that would still be
too late, and also waste effort on possibly unneeded/unwanted
dependencies.
The simple and correct thing to do is just make a store method for the
boolean logic, keeping all the setting and key stuff the way it was
before. That new method is both used by `LocalStore::addToStore` and the
substitution goal check. Perhaps we might eventually make it fancier,
e.g. sending the ValidPathInfo to remote stores for them to validate,
but this is good enough for now.
By default, once you enter x86_64 Rosetta 2, macOS will try to run
everything in x86_64. So an x86_64 Nix will still try to use x86_64
even when system = aarch64-darwin. To avoid this we can set
kern.curproc_arch_affinity sysctl. With kern.curproc_arch_affinity=0,
we ignore this preference.
This is based on how
https://opensource.apple.com/source/system_cmds/system_cmds-880.40.5/arch.tproj/arch.c.auto.html
works. Completely undocumented, but seems to work!
Note, you can verify this works with this impure Nix expression:
```
{
a = derivation {
name = "a";
system = "aarch64-darwin";
builder = "/bin/sh";
args = [ "-e" (builtins.toFile "builder" ''
[ "$(/usr/bin/arch)" = arm64 ]
[ "$(/usr/bin/arch -arch x86_64 /bin/sh -c /usr/bin/arch)" = i386 ]
[ "$(/usr/bin/arch -arch arm64 /bin/sh -c /usr/bin/arch)" = arm64 ]
/usr/bin/touch $out
'') ];
};
b = derivation {
name = "b";
system = "x86_64-darwin";
builder = "/bin/sh";
args = [ "-e" (builtins.toFile "builder" ''
[ "$(/usr/bin/arch)" = i386 ]
[ "$(/usr/bin/arch -arch x86_64 /bin/sh -c /usr/bin/arch)" = i386 ]
[ "$(/usr/bin/arch -arch arm64 /bin/sh -c /usr/bin/arch)" = arm64 ]
/usr/bin/touch $out
'') ];
};
}
```
This resolves#3810 by changing the behavior of `max-jobs = 0`, so
that specifying the option also avoids local building of derivations
with the attribute `preferLocalBuild = true`.
We embrace virtual the rest of the way, and get rid of the
`assert(false)` 0-param constructors.
We also list config base classes first, so the constructor order is
always:
1. all the configs
2. all the stores
Each in the same order
PRs #4370 and #4348 had a bad interaction in that the second broke the fist
one in a not trivial way.
The issue was that since #4348 the logic for detecting whether a
derivation output is already built requires some logic that was specific
to the `LocalStore`.
It happens though that most of this logic could be upstreamed to any `Store`,
which is what this commit does.
This ignore was here because `queryPartialDrvOutputMap` was used both
1. as a cache to avoid having to re-read the derivation (when gc-ing for
example), and
2. as the source of truth for ca realisations
The use-case 2. required it to be able to work even when the derivation
wasn't there anymore (see https://github.com/NixOS/nix/issues/4138).
However, this use-case is now handled by `queryRealisation`, meaning
that we can safely error out if the derivation isn't there anymore
`buildPaths` can be called even for stores where it's not defined in case it's
bound to be a no-op.
The “no-op detection” mechanism was only detecting the case wher `buildPaths`
was called on a set of (non-drv) paths that were already present on the store.
This commit extends this mechanism to also detect the case where `buildPaths`
is called on a set of derivation outputs which are already built on the store.
This only works with the ca-derivations flag. It could be possible to
extend this to also work without it, but it would add quite a bit of
complexity, and it's not used without it anyways.
Extend `FSAccessor::readFile` to allow not checking that the path is a
valid one, and rewrite `readInvalidDerivation` using this extended
`readFile`.
Several places in the code use `readInvalidDerivation`, either because
they need to read a derivation that has been written in the store but
not registered yet, or more generally to prevent a deadlock because
`readDerivation` tries to lock the state, so can't be called from a
place where the lock is already held.
However, `readInvalidDerivation` implicitely assumes that the store is a
`LocalFSStore`, which isn't always the case.
The concrete motivation for this is that it's required for `nix copy
--from someBinaryCache` to work, which is tremendously useful for the
tests.
Because of a too eager refactoring, `addTextToStore` used to throw an
error because the input wasn't a valid nar.
Partially revert that refactoring to wrap the text into a proper nar
(using `dumpString`) to make this method work again
Rather than storing the derivation outputs as `drvPath!outputName` internally,
store them as `drvHashModulo!outputName` (or `outputHash!outputName` for
fixed-output derivations).
This makes the storage slightly more opaque, but enables an earlier
cutoff in cases where a fixed-output dependency changes (but keeps the
same output hash) − same as what we already do for input-addressed
derivations.
Add a new table for tracking the derivation output mappings.
We used to hijack the `DerivationOutputs` table for that, but (despite its
name), it isn't a really good fit:
- Its entries depend on the drv being a valid path, making it play badly with
garbage collection and preventing us to copy a drv output without copying
the whole drv closure too;
- It dosen't guaranty that the output path exists;
By using a different table, we can experiment with a different schema better
suited for tracking the output mappings of CA derivations.
(incidentally, this also fixes#4138)
For each known realisation, store:
- its output
- its output path
This comes with a set of needed changes:
- New `realisations` module declaring the types needed for describing
these mappings
- New `Store::registerDrvOutput` method registering all the needed informations
about a derivation output (also replaces `LocalStore::linkDeriverToPath`)
- new `Store::queryRealisation` method to retrieve the informations for a
derivations
This introcudes some redundancy on the remote-store side between
`wopQueryDerivationOutputMap` and `wopQueryRealisation`.
However we might need to keep both (regardless of backwards compat)
because we sometimes need to get some infos for all the outputs of a
derivation (where `wopQueryDerivationOutputMap` is handy), but all the
stores can't implement it − because listing all the outputs of a
derivation isn't really possible for binary caches where the server
doesn't allow to list a directory.
In `nixStable` (2.3.7 to be precise) it's possible to connect to stores
using an IPv6 address:
nix ping-store --store ssh://root@2001:db8::1
This is also useful for `nixops(1)` where you could specify an IPv6
address in `deployment.targetHost`.
However, this behavior is broken on `nixUnstable` and fails with the
following error:
$ nix store ping --store ssh://root@2001:db8::1
don't know how to open Nix store 'ssh://root@2001:db8::1'
This happened because `openStore` from `libstore` uses the `parseURL`
function from `libfetchers` which expects a valid URL as defined in
RFC2732. However, this is unsupported by `ssh(1)`:
$ nix store ping --store 'ssh://root@[2001:db8::1]'
cannot connect to 'root@[2001:db8::1]'
This patch now allows both ways of specifying a store (`root@2001:db8::1`) and
also `root@[2001:db8::1]` since the latter one is useful to pass query
parameters to the remote store.
In order to achieve this, the following changes were made:
* The URL regex from `url-parts.hh` now allows an IPv6 address in the
form `2001:db8::1` and also `[2001:db8::1]`.
* In `libstore`, a new function named `extractConnStr` ensures that a
proper URL is passed to e.g. `ssh(1)`:
* If a URL looks like either `[2001:db8::1]` or `root@[2001:db8::1]`,
the brackets will be removed using a regex. No additional validation
is done here as only strings parsed by `parseURL` are expected.
* In any other case, the string will be left untouched.
* The rules above only apply for `LegacySSHStore` and `SSHStore` (a.k.a
`ssh://` and `ssh-ng://`).
Unresolved questions:
* I'm not really sure whether we want to allow both variants of IPv6
addresses in the URL parser. However it should be noted that both seem
to be possible according to RFC2732:
> This document incudes an update to the generic syntax for Uniform
> Resource Identifiers defined in RFC 2396 [URL]. It defines a syntax
> for IPv6 addresses and allows the use of "[" and "]" within a URI
> explicitly for this reserved purpose.
* Currently, it's not supported to specify a port number behind the
hostname, however it seems as this is not really supported by the URL
parser. Hence, this is probably out of scope here.
The `DerivationGoal` has a variable storing the “final” derivation
output paths that is used (amongst other things) to fill the environment
for the post build hook. However this variable wasn't set when the
build-hook is used, causing a crash when both hooks are used together.
Fix this by setting this variable (from the informations in the db) after a run
of the post build hook.
This reverts commit 1b1e076033.
Using `queryPartialDerivationOutputMap` assumes that the derivation
exists locally which isn't the case for remote builders.
Since 0744f7f, it is now useful to have cache.nixos.org in substituers
even if /nix/store is not the Nix Store Dir. This can always be
overridden via configuration, though.
When running universal binaries like /bin/bash, Darwin XNU will choose
which architecture of the binary to use based on "binary preferences".
This change sets that to the current platform for aarch64 and x86_64
builds. In addition it now uses posix_spawn instead of the usual
execve. Note, that this does not prevent the other architecture from
being run, just advises which to use.
Unfortunately, posix_spawnattr_setbinpref_np does not appear to be
inherited by child processes in x86_64 Rosetta 2 translations, meaning
that this will not always work as expected.
For example:
{
arm = derivation {
name = "test";
system = "aarch64-darwin";
builder = "/bin/bash";
args = [ "-e" (builtins.toFile "test" ''
set -x
/usr/sbin/sysctl sysctl.proc_translated
/usr/sbin/sysctl sysctl.proc_native
[ "$(/usr/bin/arch)" = arm64 ]
/usr/bin/touch $out
'') ];
};
rosetta = derivation {
name = "test";
system = "x86_64-darwin";
builder = "/bin/bash";
args = [ "-e" (builtins.toFile "test" ''
set -x
/usr/sbin/sysctl sysctl.proc_translated
/usr/sbin/sysctl sysctl.proc_native
[ "$(/usr/bin/arch)" = i386 ]
echo It works!
/usr/bin/touch $out
'') ];
};
}
`arm' fails on x86_64-compiled Nix, but `arm' and `rosetta' succeed on
aarch64-compiled Nix. I suspect there is a way to fix this since:
$ /usr/bin/arch -arch x86_64 /bin/bash \
-c '/usr/bin/arch -arch arm64e /bin/bash -c /usr/bin/arch'
arm64
seems to work correctly. We may need to wait for Apple to update
system_cmds in opensource.apple.com to find out how though.
macOS systems with ARM64 can utilize a translation layer at
/Library/Apple/usr/libexec/oah to run x86_64 binaries. This change
makes Nix recognize that and it to "extra-platforms". Note that there
are two cases here since Nix could be built for either x86_64 or
aarch64. In either case, we can switch to the other architecture.
Unfortunately there is not a good way to prevent aarch64 binaries from
being run in x86_64 contexts or vice versa - programs can always
execute programs for the other architecture.
If the build closure contains some CA derivations, then we can't know
ahead-of-time that we won't build anything as early-cutoff might come-in
at a laster stage
This fixes a bug I encountered where `nix-store -qR` will deadlock when
the `--include-outputs` flag is passed and `max-connections=1`.
The deadlock occurs because `RemoteStore::queryDerivationOutputs` takes
the only connection from the connection pool and uses it to check the
daemon version. If the version is new enough, it calls
`Store::queryDerivationOutputs`, which eventually calls
`RemoteStore::queryPartialDerivationOutputMap`, where we take another
connection from the connection pool to check the version again. Because
we still haven't released the connection from the caller, this waits for
a connection to be available, causing a deadlock.
This diff solves the issue by using `getProtocol` to check the protocol
version in the caller `RemoteStore::queryDerivationOutputs`, which
immediately frees the connection back to the pool before returning the
protocol version. That way we've already freed the connection by the
time we call `RemoteStore::queryPartialDerivationOutputMap`.
Until now, it was not possible to substitute missing paths from e.g.
`https://cache.nixos.org` on a remote server when building on it using
the new `ssh-ng` protocol.
This is because every store implementation except legacy `ssh://`
ignores the substitution flag passed to `Store::queryValidPaths` while
the `legacy-ssh-store` substitutes the remote store using
`cmdQueryValidPaths` when the remote store is opened with `nix-store
--serve`.
This patch slightly modifies the daemon protocol to allow passing an
integer value suggesting whether to substitute missing paths during
`wopQueryValidPaths`. To implement this on the daemon-side, the
substitution logic from `nix-store --serve` has been moved into a
protected method named `Store::substitutePaths` which gets currently
called from `LocalStore::queryValidPaths` and `Store::queryValidPaths`
if `maybeSubstitute` is `true`.
Fixes#2770
This removes the extra-substituters and extra-sandbox-paths settings
and instead makes every array setting extensible by setting
"extra-<name> = <value>" in the configuration file or passing
"--<name> <value>" on the command line.
This makes it even clearer which of the two hashes was specified in the
nix files. Some may think that "wanted" and "got" is obvious, but:
"got" could mean "got in nix file" and "wanted" could mean "want to see in nix file".
In particular, this means that derivations can output derivations. But
that ramification isn't (yet!) useful as we would want, since there is
no way to have a dependent derivation that is itself a dependent
derivation.
Observed on Centos 7 when user namespaces are disabled:
DerivationGoal::startBuilder() throws an exception, ~DerivationGoal()
waits for the child process to exit, but the child process hangs
forever in drainFD(userNamespaceSync.readSide.get()) in
DerivationGoal::runChild(). Not sure why the SIGKILL doesn't get
through.
Issue #4092.
This change provides support for using access tokens with other
instances of GitHub and GitLab beyond just github.com and
gitlab.com (especially company-specific or foundation-specific
instances).
This change also provides the ability to specify the type of access
token being used, where different types may have different handling,
based on the forge type.
After 0ed946aa61, max-jobs setting (-j/--max-jobs)
stopped working.
The reason was that nrLocalBuilds (which compared to maxBuildJobs to figure
out whether the limit is reached or not) is not incremented yet when tryBuild
is started; So, the solution is to move the check to tryLocalBuild.
Closes https://github.com/nixos/nix/issues/3763
We don't need it yet, but we could/should in the future, and it's a
cost-free change since we already have the reference. I like it.
Co-authored-by: Robert Hensing <roberth@users.noreply.github.com>
Rather than showing an integer as the default, instead show the boolean
referenced in the description.
The nix.conf.5 manpage used to show "default: 0", which is unnecessarily
opaque and confusing (doesn't 0 mean false, even though the default is
true?); now it properly shows that the default is true.
`nix flake info` calls the github 'commits' API, which requires
authorization when the repository is private. Currently this request
fails with a 404.
This commit adds an authorization header when calling the 'commits' API.
It also changes the way that the 'tarball' API authenticates, moving the
user's token from a query parameter into the Authorization header.
The query parameter method is recently deprecated and will be disallowed
in November 2020. Using them today triggers a warning email.
Rework the `Store` hierarchy so that there's now one hierarchy for the
store configs and one for the implementations (where each implementation
extends the corresponding config). So a class hierarchy like
```
StoreConfig-------->Store
| |
v v
SubStoreConfig----->SubStore
| |
v v
SubSubStoreConfig-->SubSubStore
```
(with virtual inheritance to prevent DDD).
The advantage of this architecture is that we can now introspect the configuration of a store without having to instantiate the store itself
Add a new `init()` method to the `Store` class that is supposed to
handle all the effectful initialisation needed to set-up the store.
The constructor should remain side-effect free and just initialize the
c++ data structure.
The goal behind that is that we can create “dummy” instances of each
store to query static properties about it (the parameters it accepts for
example)
Directly register the store classes rather than a function to build an
instance of them.
This gives the possibility to introspect static members of the class or
choose different ways of instantiating them.
Add a fallback path in `queryPartialDerivationOutputMap` for daemons
that don't support it.
Also upstreams a couple methods from `SSHStore` to `RemoteStore` as this
is needed to handle the fallback path.
When deploying a Hydra instance with current Nix master, most builds
would not run because of errors like this:
queue monitor: error: --- Error --- hydra-queue-runner
error: --- UsageError --- nix-daemon
not a content address because it is not in the form '<prefix>:<rest>': /nix/store/...-somedrv
The last error message is from parseContentAddress, which expects a
colon-separated string, however what we got here is a store path.
Looking at the worker protocol, the following message sent to the Nix
daemon caused the error above:
0x1E -> wopQuerySubstitutablePathInfos
0x01 -> Number of paths
0x16 -> Length of string
"/nix/store/...-somedrv"
0x00 -> Length of string
""
Looking at writeStorePathCAMap, the store path is indeed the first field
that's transmitted. However, readStorePathCAMap expects it to be the
*second* field *on my machine*, since expression evaluation order is a
classic form of unspecified behaviour[1] in C++.
This has been introduced in https://github.com/NixOS/nix/pull/3689,
specifically in commit 66a62b3189.
[1]: https://en.wikipedia.org/wiki/Unspecified_behavior#Order_of_evaluation_of_subexpressions
Signed-off-by: aszlig <aszlig@nix.build>
If we resolve using the known path of a derivation whose output we
didn't have, we previously blew up. Now we just fail gracefully,
returning the map of all outputs unknown.
This means profiles outside of /nix/var/nix/profiles don't get
garbage-collected. It also means we don't need to scan
/nix/var/nix/profiles for GC roots anymore, except for compatibility
with previously existing generations.
Evidentally this was never implemented because Nix switched to using
`buildDerivation` exclusively before `build-remote.pl` was rewritten.
The `nix-copy-ssh` test (already) tests this.
Include a long comment explaining the policy. Perhaps this can be moved
to the manual at some point in the future.
Also bump the daemon protocol minor version, so clients can tell whether
`wopBuildDerivation` supports trustless CA derivation building. I hope
to take advantage of this in a follow-up PR to support trustless remote
building with the minimal sending of derivation closures.
This seems more correct. It also means one can specify the features a
store should support with --store and remote-store=..., which is useful.
I use this to clean up the build remotes test.
Before, processConnection wanted to know a user name and user id, and
`nix-daemon --stdio`, when it isn't proxying to an underlying daemon,
would just assume "root" and 0. But `nix-daemon --stdio` (no proxying)
shouldn't make guesses about who holds the other end of its standard
streams.
Now processConnection takes an "auth hook", so `nix-daemon` can provide
the appropriate policy and daemon.cc doesn't need to know or care what
it is.
Thanks @regnat for catching one of them. The other follows for many of
the same reasons. I'm find fixing others on a need-to-fix basis,
provided their are no regressions.
Las Safin
Sergei Trofimovich
Yorick
Maximilian Bosch
Alyssa Ross
Robert Hensing
adisbladis
Stéphan Kochen