> | We have explicit tests for the ARM behavior; I don't see any reason to > | duplicate them on arbitrary other tests, and it's likely that they contain > | subtle target-dependencies anyway. Explicit triples are always best. > > It's an awkward question. On the one hand, clearly these tests are testing > behaviour that ought to be target independent, so we ought to be able to > have a set of tests that pick a particular ABI and just test that. On the > other hand, the reasons for these tests is to catch non-obvious breakage > early so it would be beneficial > to be testing them on various native platforms. (I'm looking at some of the > other failures and while I can't be definitive at this stage it looks like > there may be issues that show up on ARM in addition to pure ABI > differences.) > > In the end it probably comes down to deciding a trade-off: does the > additional lines of checking obscure what's going on more or less than it > increases the likelihood of catching platform specific (particularly ARM) > issues? I don't have a good feel for that.
If different architectures have different abi, and we test that in that way what those patches do (write different CHECK-ARCH), I am afraid that those test cases could be hard to read. ;) Regards, chenwj -- Wei-Ren Chen (陳韋任) Computer Systems Lab, Institute of Information Science, Academia Sinica, Taiwan (R.O.C.) Tel:886-2-2788-3799 #1667 Homepage: http://people.cs.nctu.edu.tw/~chenwj _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
