Age | Commit message (Collapse) | Author |
|
forceValue() were called after a value is copied effectively forcing only one of the copies keeping another copy not evaluated.
This resulted in its evaluation of the same lazy value more than once (the number of hits is not big though)
|
|
Before:
$ command time nix-prefetch-url https://cdn.kernel.org/pub/linux/kernel/v4.x/linux-4.17.6.tar.xz
1.19user 1.02system 0:41.96elapsed 5%CPU (0avgtext+0avgdata 182720maxresident)k
After:
1.38user 1.05system 0:39.73elapsed 6%CPU (0avgtext+0avgdata 16204maxresident)k
Note however that addToStore() can still take a lot of memory
(e.g. RemoteStore::addToStore() is constant space, but
LocalStore::addToStore() isn't; that's fixed by
https://github.com/edolstra/nix/commit/c94b4fc7ee0c7b322a5f3c7ee784063b47a11d98
though).
Fixes #1400.
|
|
Apparently, on macOS, 'long' != 'int64_t'.
https://hydra.nixos.org/build/77100756
|
|
|
|
Closes #2282.
|
|
repl: use `nix build` for building instead of `nix-store -r`
|
|
|
|
|
|
|
|
|
|
|
|
For example, this allows you to do run nix-daemon as a non-privileged
user:
eelco$ NIX_STATE_DIR=~/my-nix/nix/var nix-daemon --store ~/my-nix/
The NIX_STATE_DIR is still needed because settings.nixDaemonSocketFile
is not derived from settings.storeUri (and we can't derive it from the
store's state directory because we don't want to open the store in the
parent process).
|
|
|
|
This is primarily useful for processing Cargo.lock files.
|
|
Even on 32-bit systems, Value has enough space to hold a double.
|
|
|
|
search.cc: improve UX for `nix search`
|
|
As proposed in #1634 the `nix search` command could use some
improvements. Initially 0413aeb35d6ee869a98b6565781c1cf47dc80080 added
some basic sorting behavior using `std::map`, a next step would be an
improvement of the output.
This patch includes the following changes:
* Use `$PAGER` for outputs with `RunPager` from `shared.hh`:
The same behavior is defined for `nix-env --query`, furthermore it
makes searching huge results way easier.
* Simplified result blocks:
The new output is heavily inspired by the output from `nox`, the first
line shows the attribute path and the derivaiton name
(`attribute path (derivation name)`) and the description in the second
line.
|
|
progress bar!
|
|
Slightly nicer behavior when updates are somewhat far apart
(during a long linking step, perhaps) ensuring things
don't appear unresponsive.
If we wait the maximum amount for the update,
don't bother waiting another 50ms (for rate-limiting purposes)
and just check if we should quit.
This also ensures we'll notice the request to quit within 1s
if quit is signalled but there is not an udpate.
(I'm not sure if this happens or not)
|
|
|
|
Fixes #2225.
|
|
libstore/gc.cc: ignore ESRCH when reading symlinks in /proc
|
|
readlink is also affected by the problem fixed for regular files in
c567afe355df
|
|
I hate to make this such a large check but the lack of documentation means we really have no idea what's allowed. All of them reported so far have been within ".app/Contents" directories. That appears to be a safe starting point. However, I would not be surprised to also find more paths that are disallowed for instance in .framework or .bundle directories.
Fixes #2031
Fixes #2229
|
|
This makes 'nix copy' and 'nix path-info' work on .drv store
paths. Removing special treatment of .drv files seems the most
future-proof approach given the possible removal of .drv files in the
future.
Note that 'nix build' will still build (rather than substitute) .drv
paths due to the unfortunate overloading in Store::buildPaths().
|
|
This makes 'nix copy --to /mnt /nix/store/bla.drv' work.
|
|
The former is removed in C++17.
|
|
EvalState contains a few counters (e.g. nrValues) that increase
quickly enough that they end up being interpreted as pointers by the
garbage collector. Moving it to the heap makes them invisible to the
garbage collector.
This reduces the max RSS doing 100 evaluations of
nixos.tests.firefox.x86_64-linux.drvPath from 455 MiB to 292 MiB.
Note: ideally, allocations would be much further up in the 64-bit
address space to reduce the odds of an integer being misinterpreted as
a pointer. Maybe we can use some linker magic to move the .bss segment
to a higher address.
|
|
This makes it possible to build with -DGC_DEBUG.
|
|
This reduces the risk of object liveness misdetection. For example,
Glibc has an internal variable "mp_" that often points to a Boehm
object, keeping it alive unnecessarily. Since we don't store any
actual roots in global variables, we can just disable data segment
scanning.
With this, the max RSS doing 100 evaluations of
nixos.tests.firefox.x86_64-linux.drvPath went from 718 MiB to 455 MiB.
|
|
|
|
This prevents EvalState::resetFileCache() from parsing everything all
over again.
|
|
|
|
|
|
If a process disappears between the time /proc/[pid]/maps is opened and
the time it is read, the read() syscall will return ESRCH. This should be ignored.
|
|
runner logs.
|
|
|
|
|
|
|
|
|
|
I'm not sure if curl ever asks for enough data at once
for truncation to occur but better safe than sorry.
|
|
Don't say "download" when we mean "upload".
|
|
Implement --delete-generations + flag for keeping last N number of gens
|
|
|
|
|
|
|
|
E.g.
nix-build --store ~/my-nix/ -E 'import <nix/fetchurl.nix> { url = https://cache.nixos.org/nar/0nwi996rgq4b914qyx0mv2wq4k80hjac7xilikavagw7kxmn2iiv.nar.xz; sha256 = "0nwi996rgq4b914qyx0mv2wq4k80hjac7xilikavagw7kxmn2iiv"; }'
now runs in 17 MiB (was 70 MiB), while
nix-build --store ~/my-nix/ -E 'import <nix/fetchurl.nix> { url = https://cache.nixos.org/nar/0nwi996rgq4b914qyx0mv2wq4k80hjac7xilikavagw7kxmn2iiv.nar.xz; sha256 = "0d2fxljdih3nc5dqx41hjzic3141ajil94m8kdbpryq569dpsbvb"; unpack = true; }'
runs in 17 MiB (was 346 MiB).
|
|
|
|
This reduces memory consumption of
nix copy --from https://cache.nixos.org --to ~/my-nix /nix/store/95cwv4q54dc6giaqv6q6p4r02ia2km35-blender-2.79
from 176 MiB to 82 MiB. (The remaining memory is probably due to xz
decompression overhead.)
Issue https://github.com/NixOS/nix/issues/1681.
Issue https://github.com/NixOS/nix/issues/1969.
|