array-to-raw-elem cast: test that Retag covers entire array #85023
Conversation
|
(rust-highfive has picked a reviewer for you, use r? to override) |
rust-highfive
added
the
S-waiting-on-review
| Retag(_4); // scope 1 at $DIR/retag.rs:59:13: 59:19 | ||
| _3 = &raw mut (*_4); // scope 1 at $DIR/retag.rs:59:13: 59:19 | ||
| Retag([raw] _3); // scope 1 at $DIR/retag.rs:59:13: 59:19 | ||
| _2 = move _3 as *mut usize (Pointer(ArrayToPointer)); // scope 1 at $DIR/retag.rs:59:13: 59:33 |
There was a problem hiding this comment.
This is the key part of the test. I am not sure how to make sure this won't be bless'ed away...
There was a problem hiding this comment.
Hm. I wonder if we want something like LLVM's FileCheck for mir-opt tests... cc @rust-lang/wg-mir-opt do we have any standard practice here?
|
r=me in theory, though I agree this seems error-prone as-is. |
|
@oli-obk I recall you working on this infrastructure so that mir-opt tests can be blessed; is there any way to make sure that the test actually keeps testing what it is supposed to test though? The retag mir-opt test as it exists currently in master is a lot more fragile that I'd like it to be (and back when I added the test, it was less fragile, back then it did ensure that there'd be |
|
There is no existing infrastructure, but it would be possible to add. I think the hard part is knowing what we actually want. We could do something simple like requiring that a certain string exists in the mir dump, that could easily be added to compiletest by adding more magic comments like the ones for actually doing a dump. |
|
📌 Commit 2d2ed21 has been approved by |
bors
added
S-waiting-on-bors
|
☀️ Test successful - checks-actions |
Make sure that we
Retagbefore doing theArrayToPointercast.