[PATCH] D68230: [UpdateCCTestChecks] Allow asm in output with --allow-asm

[David Greene via Phabricator via llvm-commits]

greened added a comment.

In D68230#1690948 <https://reviews.llvm.org/D68230#1690948>, @RKSimon wrote:

> It then becomes a problem of maintenance - if a developer is just building/testing the llvm project for backend work are they responsible for keeping that test project passing? If not then it prevents us using CI bots to test and breaks are likely to linger for long lengths of time.


I've been uncomfortable with the notion of people working on pieces of the compiler but not testing the entire compiler.  As a data point, does gcc development work that way?  I know there was a long discussion about the monorepo proposal and people wanting to just take a slice of the compiler and only worry about that but I don't think that's feasible.  Back when LLVM was just an optimizer and code generator one didn't have to worry about other pieces but now we have an integrated toolset and it seems rather unsustainable to let people make changes without testing.

If we're not doing end-to-end testing then that is a large hole in our testing infrastructure and thus reduces our confidence in the final product.  If we are going to do end-to-end testing then we need to do it and have everyone be responsible for it.

I'll start a conversation on the mailing list.  In the meantime, is there any reason not to merge this change?  Just because someone *could* generate asm tests from C source doesn't mean the project has to accept the test.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D68230/new/

https://reviews.llvm.org/D68230