================
@@ -1938,11 +1942,22 @@ struct DarwinPlatform {
SourceKind Kind;
DarwinPlatformKind Platform;
DarwinEnvironmentKind Environment = DarwinEnvironmentKind::NativeEnvironment;
- VersionTuple NativeTargetVersion;
- std::string OSVersion;
- bool HasOSVersion = true, InferSimulatorFromArch = true;
+ // When compiling for a zippered target, this means both target &
+ // target variant is set on the command line, ZipperedOSVersion holds the
+ // OSVersion tied to the main target value.
+ VersionTuple ZipperedOSVersion;
+ // We allow multiple ways to set or default the OS
+ // version used for compilation. The ResolvedOSVersion always represents what
+ // will be used.
+ VersionTuple ResolvedOSVersion;
----------------
cyndyishida wrote:
I think `DarwinPlatform` is intended to be a lightweight type for representing
platform info from various arguments, while `Darwin::AddDeploymentTarget` is
the core logic for mapping it back to the target triple for CC1. It has direct
access to objects like the `Driver` to be able to report errors when attempting
to resolve the OS version. For now, I want to preserve that behavior.
However, to address the overarching concerns I have:
* changed `VersionTuple ResolvedOSVersion` to `std::optional<VersionTuple>
UnderlyingOSVersion`
* replaced `providedOSVersion` with `hasOSVersion` and simplified setting
UnderlyingOSVersion, to no longer worry about what existed at the construction
of the `DarwinPlatform` object.
* Introduce `takeOSVersion()` that should be called when the driver has
resolved the OSVersion as far as the `DarwinPlatform` object is concerned.
* renamed an unrelated `getOSVersion` to `getInferredOSVersion`
https://github.com/llvm/llvm-project/pull/142013
_______________________________________________
cfe-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
