-
Notifications
You must be signed in to change notification settings - Fork 12.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Function arguments should never get promoted #59724
Conversation
(rust_highfive has picked a reviewer for you, use r? to override) |
r? @eddyb |
@@ -639,7 +639,7 @@ impl<'a, 'tcx> Checker<'a, 'tcx> { | |||
per_local.insert(local); | |||
} | |||
} | |||
cx.per_local[IsNotPromotable].insert(local); | |||
cx.per_local[IsNotConst].insert(local); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think we should do this for all locals where if !temps[local].is_promotable()
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What I meant is that we should do this for all locals, not just temps.
That is, don't do it on a case-by-case basis in this match
, but rather have this after it:
if !temps[local].is_promotable() {
cx.per_local[IsNotConst].insert(local);
}
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
That's a change that I don't want to backport. are you alright with r+ing this PR and I'll address this immediately in a master-only PR?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
There's a risk of there being more issues we haven't hit due to this not being general enough.
This is not just code cleanup, I think right now this is kind of broken.
As an experiment, I tried backporting this change to the beta branch, to double-check that it actually addresses the regression. One potential issue is that the regression tests may not themselves be immediately backportable as written, because I do not think #59044 is in the beta branch. Just a warning. |
(commit ccc0cd8 itself does seem to address the key problem in the beta branch.) @oli-obk @eddyb Regarding the beta-nomination, I do not think we can afford to wait until the next (To be absolutely clear, I have no personal problem with backporting ccc0cd8 to beta. Though it might be good to explore whether the generalization suggested by @eddyb in their review is what really needs to be backported.) |
I think we should beta accept this without a full compiler team consensus as it's small and important to get fixed. There seems no alternative to me. I can create a beta backport that regenerates the ui output the way beta expects it. |
Yes; absolutely. I would really like to avoid a C-future-compatibility warning due to a failure to backport. |
@@ -639,15 +639,15 @@ impl<'a, 'tcx> Checker<'a, 'tcx> { | |||
per_local.insert(local); | |||
} | |||
} | |||
cx.per_local[IsNotPromotable].insert(local); | |||
cx.per_local[IsNotConst].insert(local); | |||
} | |||
|
|||
LocalKind::Var if mode == Mode::Fn => { | |||
cx.per_local[IsNotConst].insert(local); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Like, does this mean we'll try to promote a Var
in a const fn
?!
@@ -817,15 +817,15 @@ impl<'a, 'tcx> Checker<'a, 'tcx> { | |||
} | |||
} | |||
|
|||
// Ensure the `IsNotPromotable` qualification is preserved. | |||
// Ensure the `IsNotConst` qualification is preserved. | |||
// NOTE(eddyb) this is actually unnecessary right now, as | |||
// we never replace the local's qualif, but we might in | |||
// the future, and so it serves to catch changes that unset | |||
// important bits (in which case, asserting `contains` could | |||
// be replaced with calling `insert` to re-set the bit). | |||
if kind == LocalKind::Temp { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Again, here, this seems risky, why don't we do this for all locals?
I think relying on the Var
/ Temp
distinction was a bit of a mistake... I mean, here, as opposed to in the analysis in promote_consts
.
r=me with or without #59724 (comment) / #59724 (comment) resolved For runtime code, this PR is probably fine (it marks all args and vars as "not const" - not the return place, but that tends not to ever get borrowed or passed to a function), but for |
I agree with this and am going to mark the PR as beta-accepted accordingly |
I guess I shouldn't delay this needlessly: |
📌 Commit cc66b17 has been approved by |
⌛ Testing commit cc66b17 with merge f46819ad4835ed97d28d4f2b276314d63551fff7... |
💔 Test failed - checks-travis |
📌 Commit 01e8394 has been approved by |
@bors p=1 (needs to be backported) |
⌛ Testing commit 01e8394 with merge 4976e83b11d48330e77f7aef8c79666fbfffdb4f... |
@bors retry Yielding priority to the beta backports. |
⌛ Testing commit 01e8394 with merge b6978e58ef44ca7c3b33ccaff8185111416e30ad... |
@bors retry Yielding priority again to the beta backports. |
[beta] Rollup backports Cherry-picked: * #58021: Fix fallout from #57667 * #59599: Updated RELEASES.md for 1.34.0 * #59587: Remove #[doc(hidden)] from Error::type_id * #58994: Hide deprecation warnings inside derive expansions * #58015: Expand docs for `TryFrom` and `TryInto`. * #59770: ci: pin android emulator to 28.0.23 * #59704: ci: Update FreeBSD tarball downloads * #59257: Update CI configuration for building Redox libraries * #59724: Function arguments should never get promoted r? @ghost
Moving before the clippy PR, @bors p=15 |
⌛ Testing commit 01e8394 with merge 330db3bec5731d012cdcf05ddfe577bd0b828a65... |
@bors retry |
Your PR failed on Travis (raw log). Through arcane magic we have determined that the following fragments from the build log may contain information about the problem. Click to expand the log.
I'm a bot! I can only do what humans tell me to, so if this was not helpful or you have suggestions for improvements, please ping or otherwise contact |
Function arguments should never get promoted fixes #59469
☀️ Test successful - checks-travis, status-appveyor |
[beta] Rollup backports Cherry-picked: * #58021: Fix fallout from #57667 * #59599: Updated RELEASES.md for 1.34.0 * #59587: Remove #[doc(hidden)] from Error::type_id * #58994: Hide deprecation warnings inside derive expansions * #58015: Expand docs for `TryFrom` and `TryInto`. * #59770: ci: pin android emulator to 28.0.23 * #59704: ci: Update FreeBSD tarball downloads * #59257: Update CI configuration for building Redox libraries * #59724: Function arguments should never get promoted * #59499: Fix broken download link in the armhf-gnu image * #58330: Add rustdoc JS non-std tests * #58848: Prevent cache issues on version updates r? @ghost
fixes #59469