Skip to content

CTFE: simplify ConstValue by not checking for alignment - #63079

Merged
bors merged 5 commits into
rust-lang:masterfrom
RalfJung:ctfe-no-align
Aug 5, 2019
Merged

CTFE: simplify ConstValue by not checking for alignment#63079
bors merged 5 commits into
rust-lang:masterfrom
RalfJung:ctfe-no-align

Conversation

@RalfJung

@RalfJungRalfJung commented Jul 28, 2019•
edited
Loading

Copy link
Copy Markdown
Member

I hope the test suite actually covers the problematic cases here?

r? @oli-obk

Fixes#61952

@rust-highfiverust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jul 28, 2019
@RalfJungRalfJung changed the title CTFE: simplify Value type by not checking for alignmentCTFE: simplify ConstValue by not checking for alignmentJul 28, 2019
RalfJung
RalfJung commented Jul 28, 2019
View reviewed changes
Comment threadsrc/librustc_codegen_llvm/common.rs Outdated
@rust-highfive

This comment has been minimized.

Sign in to view
@oli-obk

oli-obk commented Jul 28, 2019

Copy link
Copy Markdown
Contributor

What do you think about removing the align field of Mplace in const eval by making it an assoc type of the Engine?

@RalfJung

RalfJung commented Jul 28, 2019

Copy link
Copy Markdown
MemberAuthor

What do you think about removing the align field of Mplace in const eval by making it an assoc type of the Engine?

Hm. Not sure how well that code would factor.

With #63075, we actually don't need the align field in places any more. However, my expectation is that for raw pointers we might eventually want to relax the rule that says "when you do *ptr, even if you are just computing a field offset, the ptr must be aligned and dereferencable". So I am reluctant to remove all that infrastructure if we have to bring it back later.

@oli-obk

oli-obk commented Jul 28, 2019

Copy link
Copy Markdown
Contributor

I don't want to remove the field. Miri will keep using it. Const eval doesn't need it though. If you do UB in const eval, we don't promise we catch it.

@oli-obk

oli-obk commented Jul 28, 2019

Copy link
Copy Markdown
Contributor

We should probably find a way to make miri's const_eval query run with all miri-checks on.

@RalfJung

RalfJung commented Jul 28, 2019•
edited
Loading

Copy link
Copy Markdown
MemberAuthor

Miri will keep using it.

With #63075 the field is effectively dead in Miri, though. For now. I think.

@oli-obk

oli-obk commented Jul 28, 2019

Copy link
Copy Markdown
Contributor

Ah, ok. We can revisit once that is resolved then. Can you make sure that it's documented $somewhere in the process, so that we actually do revisit?

r=me with travis green

@RalfJung

RalfJung commented Jul 28, 2019

Copy link
Copy Markdown
MemberAuthor

Const eval doesn't need it though.

Agreed. But factoring that through the Machine trait will be somewhat messy: MPlaceTy and PlaceTy would get another generic parameter; and the check_ptr_access method in Memory would end up split across both sides of the interface.

@RalfJungRalfJung mentioned this pull request Jul 28, 2019
Closed
@RalfJung

RalfJung commented Jul 28, 2019

Copy link
Copy Markdown
MemberAuthor

@bors r=oli-obk

@bors

bors commented Jul 28, 2019

Copy link
Copy Markdown
Collaborator

📌 Commit cfb13b0eeb620dee597a99c4100b1464f589b731 has been approved by oli-obk

@borsbors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jul 28, 2019
Centril
Centril reviewed Jul 28, 2019
View reviewed changes
Comment threadsrc/test/ui/consts/const-eval/ub-ref.rs Outdated
@bors

bors commented Jul 29, 2019

Copy link
Copy Markdown
Collaborator

⌛ Testing commit cfb13b0eeb620dee597a99c4100b1464f589b731 with merge 4034b4c3188a41efb8be7f415d0252d232da06b4...

@bors

bors commented Jul 29, 2019

Copy link
Copy Markdown
Collaborator

💔 Test failed - checks-azure

@rust-highfive

This comment has been minimized.

Sign in to view
@borsbors added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. labels Jul 29, 2019
@RalfJung

RalfJung commented Jul 29, 2019

Copy link
Copy Markdown
MemberAuthor

@oli-obk@Centril I made validity check alignment even when the machine does not. I am not very happy with check_ptr_access_align as method name though.

(Also that CI failure looks spurious.)

@bors

This comment has been minimized.

Sign in to view
@RalfJung

RalfJung commented Aug 2, 2019

Copy link
Copy Markdown
MemberAuthor

Rebased. @oli-obk what do you think of checking alignment in validity even when the machine otherwise ignores alignment (as is now implemented here)?

@oli-obk

oli-obk commented Aug 5, 2019

Copy link
Copy Markdown
Contributor

I think it's very reasonable. We can always weaken validity checks if we need to. Hardening them is not easy to do.

@bors r+

@bors

bors commented Aug 5, 2019

Copy link
Copy Markdown
Collaborator

📌 Commit 0cf4329 has been approved by oli-obk

@borsbors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Aug 5, 2019
@bors

bors commented Aug 5, 2019

Copy link
Copy Markdown
Collaborator

⌛ Testing commit 0cf4329 with merge 4be0675...

bors added a commit that referenced this pull request Aug 5, 2019
@bors
Auto merge of#63079- RalfJung:ctfe-no-align, r=oli-obk
4be0675
CTFE: simplify ConstValue by not checking for alignment
I hope the test suite actually covers the problematic cases here?
r? @oli-obkFixes#61952
@bors

bors commented Aug 5, 2019

Copy link
Copy Markdown
Collaborator

☀️ Test successful - checks-azure
Approved by: oli-obk
Pushing 4be0675 to master...

@borsbors added the merged-by-bors This PR was explicitly merged by bors. label Aug 5, 2019
@bors
bors merged commit 0cf4329 into rust-lang:masterAug 5, 2019
@RalfJung
RalfJung deleted the ctfe-no-align branch August 5, 2019 16:26
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@oli-obkoli-obkoli-obk left review comments
+1 more reviewer
@CentrilCentrilCentril left review comments
Reviewers whose approvals may not affect merge requirements

Assignees

@oli-obkoli-obk

Labels

merged-by-borsThis PR was explicitly merged by bors.S-waiting-on-borsStatus: Waiting on bors to run and complete tests. Bors will change the label on completion.

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Const eval does alignment checks

5 participants

@RalfJung@rust-highfive@oli-obk@bors@Centril