This fixes `cargo build --all-features` by sidestepping the feature
unification problem described in The Cargo Book[0].
Add `cargo hack --feature-powerset` to CI to enforce that this doesn't
regress (and that all combinations of features work).
Since error_in_core is nightly-only, use core-error and a fake std
module to allow aya-obj to build without std on stable.
[0] https://doc.rust-lang.org/cargo/reference/features.html#feature-unification
This fix the following issues:
- Previously the DATASEC name wasn't sanitized resulting on "Invalid
name" returned by old kernels.
- The newly created BTF struct had a size of 0 making old kernels refuse
it.
This was tested on Debian 10 with kernel 4.19.0-21.
The crate has few libstd dependencies. Since it should be platform-
independent in principle, making it no_std like the object crate would
seem reasonable.
However, the feature `error_in_core` is not yet stabilized, and the
thiserror crate currently offers no no_std support. When the feature
no_std is selected, we enable the `error_in_core` feature, switch to
thiserror-core and replace the HashMap with the one in hashbrown.
To split the crate into two, several changes were made:
1. Most `pub(crate)` are now `pub` to allow access from Aya;
2. Parts of BpfError are merged into, for example, RelocationError;
3. BTF part of Features is moved into the new crate;
4. `#![deny(missing_docs)]` is removed temporarily;
5. Some other code gets moved into the new crate, mainly:
- aya::{bpf_map_def, BtfMapDef, PinningType},
- aya::programs::{CgroupSock*AttachType},
The new crate is currenly allowing missing_docs. Member visibility
will be adjusted later to minimize exposure of implementation details.
Refs: #473
Kenjiro Nakayama
Alessandro Decina
Tamir Duberstein
Quentin JEROME
Mary
Shenghui Ye