This otherwise does not eval `:tested` any more, which means no nixos
channel updates.
Regression comes from 0eb6d0735f (#57751)
which added an assertion stopping the use of `autoResize` when the
filesystem cannot be resized automatically.
The `| tee` invocation always masked the return value of the
switch-to-configuration test.
```
~ $ false | tee && echo "oh no"
oh no
```
The added wrapper script will still output everything to stderr, while
passing failures to the test harness.
This reverts commit de86af48faa03a824917ac90f4776481c7ce9e54.
(Manual revert due to conflicts.)
See #54509
The patch is causing overlayfs to misbehave.
See #49441 for an earlier attempt, which was subsequently reverted. I am
assuming that doubling the time will be sufficient if the machine is
overloaded since so many of the tests already pass at 5 minutes, while
still not holding back failures for needlessly long.
This, paired with the previous commit, ensures the channel won't be held
back from a kernel upgrade and a non-building sd image, while still
having a new-kernel variant available.
This is because it will not eval properly with `hydra-eval-jobs`.
```
$ ...hydra/result/bin/hydra-eval-jobs \
--arg nixpkgs '{ outPath = ./.; revCount = 123; shortRev = "4567"; }' \
-I "$PWD" \
nixos/release-combined.nix
```
It fails with:
```
Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS
```
This was previously removed in 74c4e30842.
This will allow hydra to build iso and sd images for aarch64-linux, and
share a common channel with the x86-based platforms.
Change introduced in #50672.
The patch does not apply on older protobuf like protobuf3_1
```
$ nix-build -E 'with import ./. { }; python3Packages.protobuf.override { protobuf = protobuf3_1; }'
unpacking sources
unpacking source archive /nix/store/1zdyl0cxaa8ha2v1zp75zzdjd6j99d0m-source
source root is source
setting SOURCE_DATE_EPOCH to timestamp 315619200 of file source/util/python/BUILD
patching sources
applying patch /nix/store/yagx7hvylnnjq7lxbcia0y5lq1r736w3-0a59054c30e4f0ba10f10acfc1d7f3814c63e1a7.patch
patching file google/protobuf/pyext/descriptor.cc
Hunk #1 succeeded at 55 (offset -1 lines).
patching file google/protobuf/pyext/descriptor_containers.cc
patching file google/protobuf/pyext/descriptor_pool.cc
Hunk #1 succeeded at 47 (offset -1 lines).
patching file google/protobuf/pyext/extension_dict.cc
Hunk #1 FAILED at 53.
1 out of 1 hunk FAILED -- saving rejects to file google/protobuf/pyext/extension_dict.cc.rej
patching file google/protobuf/pyext/message.cc
Hunk #1 succeeded at 82 (offset 3 lines).
Hunk #2 succeeded at 1425 (offset -104 lines).
```
Since the patch isn't necessary on python versions older than 3.7, let's
only apply it for version 3.7. This means that most things using older
protobuf implementation will now be able to build when using an older
pythonPackage set (as is most probably the case anyway).
This still leaves protobuf 3.1 using packages hanging, but the errors
will be localized to those that would be breaking anyway with the
upgrade to 3.7 as default python.
Between 2b450377bf and the current
revision, the semantics behind "platforms" changed, and removing the
"aarch64-linux" string doesn't work anymore to filter it out.
Instead, blacklist the platform using the (comparatively) new
badPlatforms.
It looks like it would be a trivial fix with
`-Wno-shift-negative-value`, but in the end it fails with:
```
[100%] Linking C executable aften
libaften_static.a(a52enc.o): In function `aften_encode_init':
a52enc.c:(.text+0x303c): undefined reference to `apply_simd_restrictions'
collect2: error: ld returned 1 exit status
```
So it looks like it's not simply a warning issue.
This is to try and squeeze more lost space from the image, so that hydra
starts building it again.
The fsck previous to the resize2fs is required so resize2fs works.
The one afterwards is a sanity check.
Using `-M` from resize2fs will not give much saved space due to a known
(in the manual) issue.
```
[samueldr@aarch64:~/nixpkgs]$ ls -lh result-*/*/*.img
-r--r--r-- 1 root root 2.2G Jan 1 1970 result-original/sd-image/nixos-sd-image-18.09.git.a7fd431-aarch64-linux.img
-r--r--r-- 1 root root 2.1G Jan 1 1970 result-M/sd-image/nixos-sd-image-18.09.git.a7fd431-aarch64-linux.img
-r--r--r-- 1 root root 1.9G Jan 1 1970 result-slimmed/sd-image/nixos-sd-image-18.09.git.a7fd431-aarch64-linux.img
```
```
[samueldr@aarch64:~/nixpkgs]$ nix path-info -S ./result-original
/nix/store/c8k9n78gylx293rjh762fr05a069kxp2-nixos-sd-image-18.09.git.a7fd431-aarch64-linux.img 3844125000
[samueldr@aarch64:~/nixpkgs]$ nix path-info -S ./result-slimmed
/nix/store/962238skj5mnzhrsmjy23dyzmxk77sp4-nixos-sd-image-18.09.git.a7fd431-aarch64-linux.img 3447473208
```
```
b Batch mode. Optimized for huge recursive copies, but less secure if a crash happens during the copy.
```
It seems the "less secure if a crash happens" does not need a crash to
happen.
With batch mode:
```
/[...]/.
Start (0) does not point to parent (___)
```
For pretty much everything copied in.
Without batch mode, everything passes `fsck`.
See #51150
```
b Batch mode. Optimized for huge recursive copies, but less secure if a crash happens during the copy.
```
It seems the "less secure if a crash happens" does not need a crash to
happen.
With batch mode:
```
/[...]/.
Start (0) does not point to parent (___)
```
For pretty much everything copied in.
Without batch mode, everything passes `fsck`.
See #51150
Samuel Dionne-Riel