[PATCH] D118450: [RuntimeDyld] Handle SPARC

[Rainer Orth via Phabricator via llvm-commits]

ro added a comment.

In D118450#3630074 <https://reviews.llvm.org/D118450#3630074>, @ro wrote:

> 



>> Setting `HaveJIT` to false is a good interim solution.
>
> I could give this a try (at least for the benefit of the Solaris/sparcv9 buildbot), but I vaguely recall having tried it and still getting quite a number of `MCJIT`-related failures.

I just gave it a quick whirl: it introduced a new failure

  FAIL: Clang :: Interpreter/code-undo.cpp

although that test is guarded by `host-supports-jit`.  The remaining tests seem to persist, too, although this is hard to see since gtest sharding produced useless test names like

  -FAIL: Clang-Unit :: Interpreter/./ClangReplInterpreterTests/2/7
  +FAIL: Clang-Unit :: Interpreter/./ClangReplInterpreterTests/2/8

and

  -FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/67/126
  -FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/68/126
  -FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/69/126
  -FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/70/126
  +FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/68/127
  +FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/69/127
  +FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/70/127
  +FAIL: LLVM-Unit :: ExecutionEngine/Orc/./OrcJITTests/71/127

This claimed optimization seems like a nightmare to me.  I'll fire off a debug build to investigate.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D118450/new/

https://reviews.llvm.org/D118450