Allow overriding the TLS model#45666
Conversation
|
r? @pnkfelix (rust_highfive has picked a reviewer for you, use r? to override) |
kennytm
added
the
S-waiting-on-review
|
At the moment |
|
Ok, I'm now caching the parsed TLS model in the crate context. This PR is ready for merging. |
|
I think @alexcrichton is more qualified to review this. |
|
I don't really know what a TLS model is but it seems fine to expose for now. Can this be placed behind feature gates, however, to require nightly and go through the normal stabilization process? |
|
I don't think we have a mechanism for putting a target json attribute under a feature gate. |
|
Also, how do I mark a -C option as unstable? |
|
That's ok to leave the target json as-is, those are "pretty unstable" as-is anyway. For the options you can move it to |
|
Done |
|
@bors: r+ |
|
📌 Commit fdf7ba2 has been approved by |
Allow overriding the TLS model This PR adds the ability to override the default "global-dynamic" TLS model with a more specific one through a target json option or a command-line option. This allows for better code generation in certain situations. This is similar to the `-ftls-model=` option in GCC and Clang.
|
☀️ Test successful - status-appveyor, status-travis |
8 participants
This PR adds the ability to override the default "global-dynamic" TLS model with a more specific one through a target json option or a command-line option. This allows for better code generation in certain situations.
This is similar to the
-ftls-model=option in GCC and Clang.