Prepare pgrx 0.11.3 #1496
Merged
Prepare pgrx 0.11.3 #1496
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
The rule utils header contains useful fonctions such as `deparse_expression()`.
this explains the version match requirements of `cargo-pgrx` and also breaks up the initial "one step" into two steps with slightly more explanation.
Discussed was also adding `libpq.h` but that would expand development requirements to Kerberos development headers and that isn't a good situation for existing users.
…1448) ureq defines `default = ["tls", "gzip"]`, so "tls" feature (rustls) is currently always enabled even when native-tls is preferred, i.e. both rustls and native-tls is built if openssl headers are available and the user cannot opt-out from using rustls. This also disallows building cargo-pgrx on platforms not supported by the ring crate (dependency of rustls). This should also solve #1430 if `cargo-pgrx` is built with `--no-default-features --features native-tls`.
I first settled down to write up something using `rustls-native_certs` but then I actually noticed ureq simply enables this with a feature. This should fully solve #1430 for both rustls and native-tls.
Closes #164
Rather than `cargo-pgrx` using just the `rustc` minor version when checking for a toolchain version mismatch, this now uses the full `cargo` version string. `cargo` uses the same version number as `rustc`, so it's a sufficient stand in for checking the version of `rustc`. By checking the full version string instead of only the minor version, we can now detect a mismatch between two nightly versions of rust issued in the same 6 week period. Example: `cargo 1.76.0-nightly (9b13310ca 2023-11-24)` vs `cargo 1.76.0-nightly (1aa9df1a5 2023-12-12)` These detect as the same minor version, but using the new full string check they will correctly show up as different and flag the mismatch error. Also: the error message now alerts people how they can override the version check when necessary.
- `allow(unused)` is redundant with `pub` - `link_name` is redundant with `no_mangle`
#911 introduced using the wrong signature and this UB was missed in review amidst the thousands of lines of changes. Signed-off-by: usamoi <usamoi@outlook.com>
I use `ereport!` because I like to include a `detail` entry but it comes
with some inconveniences. When I would write the following snippet in my
`Err` match arms, the compiler would complain about mismatch type output
from the match arms thus requiring the inclusion of `unreachable!`
```rust
pgrx::ereport!(
PgLogLevel::ERROR,
PgSqlErrorCode::ERRCODE_INTERNAL_ERROR,
format!("{} failed", stringify!(snowflake_pool_stats)),
err.to_string()
);
// necessary to satisfy the compiler
unreachable!()
```
After looking at the macro src for `ereport!` I realized why I started
writing `PgLogLevel::ERROR` rather than simply `ERROR`:
the `ERROR` case didn't support a `detail` argument so let's add that.
Adds header files for multixact.h and its dependencies.
There's already a pgrx::varlena module but utils/varlena.h adds functions to manipulate GUC variables and identifiers. Includes utils/sortsupport.h as a transitive inclusion, along with varatt.h which is already transitively included.
This documents our current state of support, and adds a note for people hoping for 32-bit support: it unfortunately would not be cheap, neither in terms of the incurred technical debt, nor in terms of running the additional build and test infra to test on 32-bit CPUs, which is not something GitHub Actions helps with.
Tweaking owo-color features required due to transitive deps. Addresses a Windows-only, deeply hypothetical security alert for atty.
workingjubilee
force-pushed
the
prepare-0.11.3
branch
from
720cae0
to
c1bcf74
Compare
`fs::canonicalize` calls `realpath` which requires the dest path to exist. It doesn't when installing the extension for the first time (and a new version of the extension for subsequent times). So...best we can [1] do is trust that @eeeebbbbrrrr 's canonicalization works. 🚀 [1]: https://stackoverflow.com/questions/68231306/stdfscanonicalize-for-files-that-dont-exist
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.
The pgrx 0.11.3 release addresses a few UB risks in pgrx, updates its dependencies on many points, and includes many additional headers. It should also now be easier to use cargo-pgrx on more-complicated network configurations.
New Bindings!
New bindings added thanks to
ForeignDataWrapperRelationIdOID definitions #1432include/multixact.h#1486"...wait, that's UB?"
Two UB fixes!
cargoversion string in the version match check. #1443Ergonomics
ereport!macro in Allowdetailargument and trailing commas inereport!#1472Less transport-level security problems in cargo-pgrx
Together these should mean it's possible to actually use cargo-pgrx on whatever your network configuration is, but you might have to use
cargo install --no-default-features --features native-tlsto install with native-tls (which, on Linux, means OpenSSL). By default, you will use rustls.Many dependency updates
These address some largely-hypothetical security risks, but one is particularly important: the bindgen update means we now should be compatible with some aarch64 builds that might have failed.