labath added a comment. In https://reviews.llvm.org/D46005#1077013, @zturner wrote:
> I thought the intention was going to be parallelize at file-granularity > rather than method granularity, since the whole point of grouping tests > together into classes is that they can share similar set up and tear down > which may be expensive to perform multiple times. If we're going to > parallelize at method-granularity, I would rather have one .py file per > method. Hm... good question. I don't think any of our current tests take advantage of this possibility (sharing set-up code). Well.. except if you don't count bringing up the entire dotest machinery as set-up code (which does take a non-zero amount of time, but hopefully that can be brought down once we move everything to lit). However, maybe this does need more discussion. FWIW, I think that we shouldn't enforce the one-test-method-per-file completely rigidly. It may make sense to separate out some of the things, but I think that our tests often need some similar utility functions, which are too specific to put in some generic library, and it's easier to share these if they are in the same file. Having multiple tests in the same file also reduces the boiler plate and avoids another round of modifying each test file. (I think even lit acknowledges the usefulness of having logically separate tests in the same file (CHECK-LABEL)). https://reviews.llvm.org/D46005 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
