You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
This is a continuation of code generation (#227) that focuses on the isolated JIT engine library. The idea is that the JIT should support the lates ORC JIT v2 in LLVM 20 and drop all legacy support/consideration. This is the most like a wholesale re-write as there is little in the way of functionality that hasn't changed for this.
Describe the solution you'd like
Full re-write for support of the LLVM20 ORC JIT v2. This MAY include a need for additional extension APIs. Attempts to implement the support and use it from the Kaleidoscope samples will inform the direction of this more.
Describe alternatives you've considered
There really is none, even the "legacy MCJIT" is not directly compatible and is layered on top of ORC JIT v2 now as far as I can tell. (The writing is clearly "on the wall" that the legacy JIT and APIs are not long for this world)
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
This is a continuation of code generation (#227) that focuses on the isolated JIT engine library. The idea is that the JIT should support the lates ORC JIT v2 in LLVM 20 and drop all legacy support/consideration. This is the most like a wholesale re-write as there is little in the way of functionality that hasn't changed for this.
Describe the solution you'd like
Full re-write for support of the LLVM20 ORC JIT v2. This MAY include a need for additional extension APIs. Attempts to implement the support and use it from the Kaleidoscope samples will inform the direction of this more.
Describe alternatives you've considered
There really is none, even the "legacy MCJIT" is not directly compatible and is layered on top of ORC JIT v2 now as far as I can tell. (The writing is clearly "on the wall" that the legacy JIT and APIs are not long for this world)
The text was updated successfully, but these errors were encountered: