rust-cbindgen: 0.13.1 -> 0.13.2 #83247
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
jonringer
suggested changes
Mar 24, 2020
There was a problem hiding this comment.
several regressions
[5 built (2 failed), 43 copied (2027.9 MiB), 1152.9 MiB DL]
error: build of '/nix/store/8p4299inv57nma311vyw4b9jf8jy8pxa-env.drv' failed
https://github.com/NixOS/nixpkgs/pull/83247
5 package failed to build:
adoptopenjdk-icedtea-web firefox-esr-68-unwrapped xulrunner mozplugger slimerjs
2 package built:
rust-cbindgen thunderbird
|
On 11:00 24.03.20, Jon wrote:
```
[5 built (2 failed), 43 copied (2027.9 MiB), 1152.9 MiB DL]
error: build of '/nix/store/8p4299inv57nma311vyw4b9jf8jy8pxa-env.drv' failed
#83247
5 package failed to build:
adoptopenjdk-icedtea-web firefox-esr-68-unwrapped xulrunner mozplugger slimerjs
2 package built:
rust-cbindgen thunderbird
```
That firefox package actually compiled for me. Do you have a build log?
Maybe you did run out of space in the temporary build dir?
Also I managed to build all these 5 that failed for you:
/nix/store/lvk8g3488pamv7zmfd0d00hmncrwx9vy-adoptopenjdk-icedtea-web-1.8.3
/nix/store/ad3v1cjhbxdwzncak8g62z4cy2glm7m2-firefox-esr-unwrapped-68.6.0esr
/nix/store/n5bncnlq8m7qf5j6r9a7064k2fkxd89c-firefox-unwrapped-74.0
/nix/store/pim6v44qr10y0xvcadf7ygxbxhpw4ppw-mozplugger-2.1.6
/nix/store/88japdinxfqasm06p7358xjk43g32sjb-slimerjs-1.0.0
(xulrunner builds firefox 74)
|
|
As one data point, the last time we updated My suspicion is that the combination of the simultaneous firefox latest + firefox ESR rebuilds is causing your nixpkgs-review failure @jonringer, and if you just build them individually prior to running review you'll be OK. |
|
@GrahamcOfBorg build firefox-esr-68-unwrapped |
|
I discovered firefox build failures both in master and while building #84211, so I'm fairly confident this is a regression introduced by the bump of I pushed a revert in 09dde57. |
|
Strange, both @andir and I managed to build |
|
I just built $ git checkout 09dde57e93d010b786eed4796d53f34a6c6e3891^
$ nix-build -A firefox -A rust-cbindgen
/nix/store/4ipn59ydqk1xwy1l9adwkb0icz1ndhlb-firefox-74.0
/nix/store/q321xa7nd39jsgkpszpqlwdid6v56z45-rust-cbindgen-0.13.2I suspect that the build is just huge and vulnerable to the build host running out of memory and the OS killing something, resulting in the appearance of failure. |
|
Having built my own patched Firefox for ages, I see this sort of build failure all the time. A good signal it’s a memory issue is to build with fewer cores ( |
|
On 03:55 04.04.20, Alyssa Ross wrote:
Having built my own patched Firefox for ages, I see this sort of build
failure all the time. A good signal it’s a memory issue is to build
with fewer cores (`--option cores`). If it works with fewer cores, it
was probably a memory issue. (With other packages, this would be more
a sign of parallelism bugs, but with Firefox, it’s usually memory. We
can be somewhat confident of that by using a smaller but still large
number of cores. For example, I usually get a build failure with 48
cores, but it’s always fine with 24.)
For a long time that was an issue due to how the build system did
produces outputs. THese days it is likely just the amount of scratch
space and RAM that has to be available during the compilation. On a
system with 16GB RAM you can't build Firefox on tmpfs (at least not
straight forward/with the defaults).
|
|
In doing some further experimentation, I found that, even when
constrained to a single core, the rustc proccess for the gkrust crate
(which is where I've noticed Firefox usually fails) peaks at 6.5G
resident.
6.5G! For a single compiler invocation!
|
|
I ran the build on a AMD EPYC 7401P with 128GB, restricted to 1 core. |
|
I ran the build on a AMD EPYC 7401P with 128GB, restricted to 1
core. `/tmp` is not on tmpfs. I can't really believe this is purely
high memory usage.
Yeah I agree now. I thought to look in dmesg and didn't see the OOM
killer get involved. Are you also getting it on master? I am.
|
|
You're right - I can reproduce the build failure on master… This probably means the revert of I quickly checked, building master firefox failed on hydra due to receiving a SIGKILL. I propose reverting the revert for now, and tracking this separately. |
flokli
added a commit
that referenced
this pull request
Apr 4, 2020
This reverts commit 09dde57. Apparently, the rust-cbindgen bump wasn't the cause for the firefox build error reported in #83247 (we could reproduce the build error even after 09dde57 applied). For some reason it must have succeeded on hydra, as it's in the cache, tricking us in believing 76458f8 broke it initially. So the build seems flaky of some sort - we haven't yet determined whether it's luck, compiling with the right CPUs or something else. :-/ There's still some investigation to be done (#84283), but no need to keep an ineffective revert around.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Semi-automatic update generated by nixpkgs-update tools. This update was made based on information from https://github.com/eqrion/cbindgen/releases.
meta.description for rust-cbindgen is: "A project for generating C bindings from Rust code"
meta.homepage for rust-cbindgen is: "https://github.com/eqrion/cbindgen"
Updates performed:
Compare changes on GitHub
Checks done (click to expand)
built on NixOS
The tests defined in
passthru.tests, if any, passed0 of 0 passed binary check by having a zero exit code.
0 of 0 passed binary check by having the new version present in output.
found 0.13.2 with grep in /nix/store/q321xa7nd39jsgkpszpqlwdid6v56z45-rust-cbindgen-0.13.2
directory tree listing: https://gist.github.com/ed163e8a2abfd195cda4d878767731c8
du listing: https://gist.github.com/6d471dae1562cb8d3c5111bb853a1c45
Rebuild report (if merged into master) (click to expand)
36 total rebuild path(s)
10 package rebuild(s)
10 x86_64-linux rebuild(s)
10 i686-linux rebuild(s)
6 x86_64-darwin rebuild(s)
10 aarch64-linux rebuild(s)
First fifty rebuilds by attrpath
adoptopenjdk-icedtea-web
firefox-esr-68-unwrapped
firefox-unwrapped
firefoxPackages.firefox
firefoxPackages.firefox-esr-68
mozplugger
rust-cbindgen
slimerjs
thunderbird
xulrunner
Instructions to test this update (click to expand)
Either download from Cachix:
(r-ryantm's Cachix cache is only trusted for this store-path realization.)
For the Cachix download to work, your user must be in the
trusted-userslist or you can usesudosince root is effectively trusted.Or, build yourself:
After you've downloaded or built it, look at the files and if there are any, run the binaries:
cc @jtojnar @andir for testing.