Based on the discussion in #3, it seems that when you declare a distributed_slice within a library it automatically gets prefixed with the crate's name, thus the duplicate check shouldn't be necessary.
It appears however that the inclusion of multiple semvers of the crate can result in duplicate declarations.
Unfortunately, the duplicate check has two unwanted consequences: an extra .linkm2_ section as well as a potential panic!.
I tried adding println!("cargo:rustc-link-arg-bins=/MERGE:.linkme_=.data,.linkm2_=.rdata") to the build.rs script in an attempt to get rid of the section(s), but that didn't seem to have any effect.
It would be helpful to have a feature flag, like no_dupcheck, to conditionally opt-out of the duplicate checks introduced in #53 when creating a release build.
Also, is there any way to prevent multiple semvers from causing duplicate slices in the first place?
Based on the discussion in #3, it seems that when you declare a
distributed_slicewithin a library it automatically gets prefixed with the crate's name, thus the duplicate check shouldn't be necessary. It appears however that the inclusion of multiple semvers of the crate can result in duplicate declarations.Unfortunately, the duplicate check has two unwanted consequences: an extra
.linkm2_section as well as a potentialpanic!. I tried addingprintln!("cargo:rustc-link-arg-bins=/MERGE:.linkme_=.data,.linkm2_=.rdata")to thebuild.rsscript in an attempt to get rid of the section(s), but that didn't seem to have any effect.It would be helpful to have a feature flag, like
no_dupcheck, to conditionally opt-out of the duplicate checks introduced in #53 when creating areleasebuild.Also, is there any way to prevent multiple semvers from causing duplicate slices in the first place?