Extend minicore with intrinsics and use it to replace #[rustc_intrinsic] in tests #140037

hbina · 2025-04-19T10:36:26Z

rustbot · 2025-04-19T10:36:30Z

rustbot has assigned @Mark-Simulacrum.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

hbina · 2025-04-19T10:38:51Z

I am getting this issue

hbina085@DESKTOP-PGERPVO ~/g/rustlang (hbina-update-use-minicore) [1]> reset && python x.py test /home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs
Building bootstrap
    Finished `dev` profile [unoptimized] target(s) in 0.05s
Building stage0 library artifacts (x86_64-unknown-linux-gnu)
    Finished `release` profile [optimized + debuginfo] target(s) in 0.05s
Building compiler artifacts (stage0 -> stage1, x86_64-unknown-linux-gnu)
    Finished `release` profile [optimized + debuginfo] target(s) in 0.25s
Creating a sysroot for stage1 compiler (use `rustup toolchain link 'name' build/host/stage1`)
Building stage0 tool lld-wrapper (x86_64-unknown-linux-gnu)
    Finished `release` profile [optimized + debuginfo] target(s) in 0.09s
Building stage1 library artifacts (x86_64-unknown-linux-gnu)
    Finished `release` profile [optimized + debuginfo] target(s) in 0.02s
Building stage0 tool compiletest (x86_64-unknown-linux-gnu)
    Finished `release` profile [optimized + debuginfo] target(s) in 0.16s
Testing stage1 compiletest suite=assembly mode=assembly (x86_64-unknown-linux-gnu)

running 539 tests
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii  88/539
ii.iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii 176/539
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii 264/539
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii 352/539
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii 440/539
iiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiiii 528/539
iiiiiiii
[assembly] tests/assembly/rust-abi-arg-attr.rs#loongarch64 ... F

[assembly] tests/assembly/rust-abi-arg-attr.rs#riscv64-zbb ... F

[assembly] tests/assembly/rust-abi-arg-attr.rs#riscv64 ... F


failures:

---- [assembly] tests/assembly/rust-abi-arg-attr.rs#loongarch64 stdout ----

error in revision `loongarch64`: verification with 'FileCheck' failed
status: exit status: 1
command: "/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/ci-llvm/bin/FileCheck" "--input-file" "/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.loongarch64/rust-abi-arg-attr.s" "/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs" "--check-prefix=CHECK" "--check-prefix" "loongarch64" "--allow-unused-prefixes" "--dump-input-context" "100"
stdout: none
--- stderr -------------------------------
/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs:34:17: error: CHECK-NEXT: expected string not found in input
 // CHECK-NEXT: .cfi_startproc
                ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.loongarch64/rust-abi-arg-attr.s:6:18: note: scanning from here
issue_114508_u32:
                 ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.loongarch64/rust-abi-arg-attr.s:15:30: note: possible intended match here
 .section .text.issue_114508_i32,"ax",@progbits
                             ^
/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs:54:17: error: CHECK-NEXT: expected string not found in input
 // CHECK-NEXT: .cfi_startproc
                ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.loongarch64/rust-abi-arg-attr.s:19:18: note: scanning from here
issue_114508_i32:
                 ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.loongarch64/rust-abi-arg-attr.s:29:17: note: possible intended match here
 .section ".note.GNU-stack","",@progbits
                ^

Input file: /home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.loongarch64/rust-abi-arg-attr.s
Check file: /home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs

-dump-input=help explains the following input dump.

Input was:
<<<<<<
           1:  .file "rust_abi_arg_attr.eff6563d25fa939-cgu.0" 
           2:  .section .text.issue_114508_u32,"ax",@progbits 
           3:  .globl issue_114508_u32 
           4:  .p2align 5 
           5:  .type issue_114508_u32,@function 
           6: issue_114508_u32: 
next:34'0                      X error: no match found
           7:  sltu $a2, $a1, $a0 
next:34'0     ~~~~~~~~~~~~~~~~~~~~
           8:  masknez $a1, $a1, $a2 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~
           9:  maskeqz $a0, $a0, $a2 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~
          10:  or $a0, $a0, $a1 
next:34'0     ~~~~~~~~~~~~~~~~~~
          11:  ret 
next:34'0     ~~~~~
          12: .Lfunc_end0: 
next:34'0     ~~~~~~~~~~~~~
          13:  .size issue_114508_u32, .Lfunc_end0-issue_114508_u32 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          14:  
next:34'0     ~
          15:  .section .text.issue_114508_i32,"ax",@progbits 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:34'1                                  ?                   possible intended match
          16:  .globl issue_114508_i32 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
          17:  .p2align 5 
next:34'0     ~~~~~~~~~~~~
          18:  .type issue_114508_i32,@function 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          19: issue_114508_i32: 
next:34'0     ~~~~~~~~~~~~~~~~~
next:54'0                      X error: no match found
          20:  slt $a2, $a1, $a0 
next:54'0     ~~~~~~~~~~~~~~~~~~~
          21:  masknez $a1, $a1, $a2 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~
          22:  maskeqz $a0, $a0, $a2 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~
          23:  or $a0, $a0, $a1 
next:54'0     ~~~~~~~~~~~~~~~~~~
          24:  ret 
next:54'0     ~~~~~
          25: .Lfunc_end1: 
next:54'0     ~~~~~~~~~~~~~
          26:  .size issue_114508_i32, .Lfunc_end1-issue_114508_i32 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          27:  
next:54'0     ~
          28:  .ident "rustc version 1.88.0-dev" 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          29:  .section ".note.GNU-stack","",@progbits 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:54'1                     ?                         possible intended match
>>>>>>
------------------------------------------


---- [assembly] tests/assembly/rust-abi-arg-attr.rs#riscv64-zbb stdout ----

error in revision `riscv64-zbb`: verification with 'FileCheck' failed
status: exit status: 1
command: "/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/ci-llvm/bin/FileCheck" "--input-file" "/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64-zbb/rust-abi-arg-attr.s" "/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs" "--check-prefix=CHECK" "--check-prefix" "riscv64-zbb" "--allow-unused-prefixes" "--dump-input-context" "100"
stdout: none
--- stderr -------------------------------
/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs:34:17: error: CHECK-NEXT: expected string not found in input
 // CHECK-NEXT: .cfi_startproc
                ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64-zbb/rust-abi-arg-attr.s:8:18: note: scanning from here
issue_114508_u32:
                 ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64-zbb/rust-abi-arg-attr.s:14:30: note: possible intended match here
 .section .text.issue_114508_i32,"ax",@progbits
                             ^
/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs:54:17: error: CHECK-NEXT: expected string not found in input
 // CHECK-NEXT: .cfi_startproc
                ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64-zbb/rust-abi-arg-attr.s:18:18: note: scanning from here
issue_114508_i32:
                 ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64-zbb/rust-abi-arg-attr.s:25:17: note: possible intended match here
 .section ".note.GNU-stack","",@progbits
                ^

Input file: /home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64-zbb/rust-abi-arg-attr.s
Check file: /home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs

-dump-input=help explains the following input dump.

Input was:
<<<<<<
           1:  .attribute 4, 16 
           2:  .attribute 5, "rv64i2p1_m2p0_a2p1_f2p2_d2p2_c2p0_zicsr2p0_zifencei2p0_zmmul1p0_zaamo1p0_zalrsc1p0_zbb1p0" 
           3:  .file "rust_abi_arg_attr.eff6563d25fa939-cgu.0" 
           4:  .section .text.issue_114508_u32,"ax",@progbits 
           5:  .globl issue_114508_u32 
           6:  .p2align 1 
           7:  .type issue_114508_u32,@function 
           8: issue_114508_u32: 
next:34'0                      X error: no match found
           9:  maxu a0, a0, a1 
next:34'0     ~~~~~~~~~~~~~~~~~
          10:  ret 
next:34'0     ~~~~~
          11: .Lfunc_end0: 
next:34'0     ~~~~~~~~~~~~~
          12:  .size issue_114508_u32, .Lfunc_end0-issue_114508_u32 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          13:  
next:34'0     ~
          14:  .section .text.issue_114508_i32,"ax",@progbits 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:34'1                                  ?                   possible intended match
          15:  .globl issue_114508_i32 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
          16:  .p2align 1 
next:34'0     ~~~~~~~~~~~~
          17:  .type issue_114508_i32,@function 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          18: issue_114508_i32: 
next:34'0     ~~~~~~~~~~~~~~~~~
next:54'0                      X error: no match found
          19:  max a0, a0, a1 
next:54'0     ~~~~~~~~~~~~~~~~
          20:  ret 
next:54'0     ~~~~~
          21: .Lfunc_end1: 
next:54'0     ~~~~~~~~~~~~~
          22:  .size issue_114508_i32, .Lfunc_end1-issue_114508_i32 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          23:  
next:54'0     ~
          24:  .ident "rustc version 1.88.0-dev" 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          25:  .section ".note.GNU-stack","",@progbits 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:54'1                     ?                         possible intended match
>>>>>>
------------------------------------------


---- [assembly] tests/assembly/rust-abi-arg-attr.rs#riscv64 stdout ----

error in revision `riscv64`: verification with 'FileCheck' failed
status: exit status: 1
command: "/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/ci-llvm/bin/FileCheck" "--input-file" "/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64/rust-abi-arg-attr.s" "/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs" "--check-prefix=CHECK" "--check-prefix" "riscv64" "--allow-unused-prefixes" "--dump-input-context" "100"
stdout: none
--- stderr -------------------------------
/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs:34:17: error: CHECK-NEXT: expected string not found in input
 // CHECK-NEXT: .cfi_startproc
                ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64/rust-abi-arg-attr.s:8:18: note: scanning from here
issue_114508_u32:
                 ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64/rust-abi-arg-attr.s:16:30: note: possible intended match here
 .section .text.issue_114508_i32,"ax",@progbits
                             ^
/home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs:54:17: error: CHECK-NEXT: expected string not found in input
 // CHECK-NEXT: .cfi_startproc
                ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64/rust-abi-arg-attr.s:20:18: note: scanning from here
issue_114508_i32:
                 ^
/home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64/rust-abi-arg-attr.s:29:17: note: possible intended match here
 .section ".note.GNU-stack","",@progbits
                ^

Input file: /home/hbina085/git/rustlang/build/x86_64-unknown-linux-gnu/test/assembly/rust-abi-arg-attr.riscv64/rust-abi-arg-attr.s
Check file: /home/hbina085/git/rustlang/tests/assembly/rust-abi-arg-attr.rs

-dump-input=help explains the following input dump.

Input was:
<<<<<<
           1:  .attribute 4, 16 
           2:  .attribute 5, "rv64i2p1_m2p0_a2p1_f2p2_d2p2_c2p0_zicsr2p0_zifencei2p0_zmmul1p0_zaamo1p0_zalrsc1p0" 
           3:  .file "rust_abi_arg_attr.eff6563d25fa939-cgu.0" 
           4:  .section .text.issue_114508_u32,"ax",@progbits 
           5:  .globl issue_114508_u32 
           6:  .p2align 1 
           7:  .type issue_114508_u32,@function 
           8: issue_114508_u32: 
next:34'0                      X error: no match found
           9:  bltu a1, a0, .LBB0_2 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~
          10:  mv a0, a1 
next:34'0     ~~~~~~~~~~~
          11: .LBB0_2: 
next:34'0     ~~~~~~~~~
          12:  ret 
next:34'0     ~~~~~
          13: .Lfunc_end0: 
next:34'0     ~~~~~~~~~~~~~
          14:  .size issue_114508_u32, .Lfunc_end0-issue_114508_u32 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          15:  
next:34'0     ~
          16:  .section .text.issue_114508_i32,"ax",@progbits 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:34'1                                  ?                   possible intended match
          17:  .globl issue_114508_i32 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~
          18:  .p2align 1 
next:34'0     ~~~~~~~~~~~~
          19:  .type issue_114508_i32,@function 
next:34'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          20: issue_114508_i32: 
next:34'0     ~~~~~~~~~~~~~~~~~
next:54'0                      X error: no match found
          21:  blt a1, a0, .LBB1_2 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~
          22:  mv a0, a1 
next:54'0     ~~~~~~~~~~~
          23: .LBB1_2: 
next:54'0     ~~~~~~~~~
          24:  ret 
next:54'0     ~~~~~
          25: .Lfunc_end1: 
next:54'0     ~~~~~~~~~~~~~
          26:  .size issue_114508_i32, .Lfunc_end1-issue_114508_i32 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          27:  
next:54'0     ~
          28:  .ident "rustc version 1.88.0-dev" 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
          29:  .section ".note.GNU-stack","",@progbits 
next:54'0     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
next:54'1                     ?                         possible intended match
>>>>>>
------------------------------------------



failures:
    [assembly] tests/assembly/rust-abi-arg-attr.rs#loongarch64
    [assembly] tests/assembly/rust-abi-arg-attr.rs#riscv64-zbb
    [assembly] tests/assembly/rust-abi-arg-attr.rs#riscv64

test result: FAILED. 1 passed; 3 failed; 535 ignored; 0 measured; 0 filtered out; finished in 102.37ms

Some tests failed in compiletest suite=assembly mode=assembly host=x86_64-unknown-linux-gnu target=x86_64-unknown-linux-gnu
help: ignored 497 up-to-date tests; use `--force-rerun` to prevent this

From what I can gather, the problem is that the compiler optimizes out the function?
It seems that Rust compiles crates as a separate thing so when the final test binary is built, the metadata is gone so the test here doesn't work anymore.

jieyouxu · 2025-04-19T11:21:07Z

From what I can gather, the problem is that the compiler optimizes out the function?

Not quite. From what I can tell, the key instructions are still present (i.e. if you comment out the .cfi* directive CHECK-NEXTs the test passes for me locally), but it's missing the CFI directives.

E.g. loongarch64 revision (this PR)

issue_114508_u32:
	sltu	$a2, $a1, $a0
	masknez	$a1, $a1, $a2
	maskeqz	$a0, $a0, $a2
	or	$a0, $a0, $a1
	ret
.Lfunc_end0:
	.size	issue_114508_u32, .Lfunc_end0-issue_114508_u32

	.section	.text.issue_114508_i32,"ax",@progbits
	.globl	issue_114508_i32
	.p2align	5
	.type	issue_114508_i32,@function
issue_114508_i32:
	slt	$a2, $a1, $a0
	masknez	$a1, $a1, $a2
	maskeqz	$a0, $a0, $a2
	or	$a0, $a0, $a1
	ret
.Lfunc_end1:
	.size	issue_114508_i32, .Lfunc_end1-issue_114508_i32

	.ident	"rustc version 1.88.0-dev"
	.section	".note.GNU-stack","",@progbits

on master:

issue_114508_u32:
	.cfi_startproc
	sltu	$a2, $a1, $a0
	masknez	$a1, $a1, $a2
	maskeqz	$a0, $a0, $a2
	or	$a0, $a0, $a1
	ret
.Lfunc_end0:
	.size	issue_114508_u32, .Lfunc_end0-issue_114508_u32
	.cfi_endproc

	.section	.text.issue_114508_i32,"ax",@progbits
	.globl	issue_114508_i32
	.p2align	5
	.type	issue_114508_i32,@function
issue_114508_i32:
	.cfi_startproc
	slt	$a2, $a1, $a0
	masknez	$a1, $a1, $a2
	maskeqz	$a0, $a0, $a2
	or	$a0, $a0, $a1
	ret
.Lfunc_end1:
	.size	issue_114508_i32, .Lfunc_end1-issue_114508_i32
	.cfi_endproc

	.ident	"rustc version 1.88.0-dev"
	.section	".note.GNU-stack","",@progbits

cc @rcvalle do you happen to know why the CFI directives might be missing? FWIW, minicore is built as an rlib + -Cpanic=abort and linked to the test crate via --extern minicore=/path/to/libminicore.rlib.

jieyouxu · 2025-04-20T00:17:26Z

EDIT: bjorn3 told me

.cfi_startproc is expected whenever you compile with -Cpanic=unwind or with -Cpanic=abort + -Cforce-unwind-tables.

In this case I think it has to do with minicore being compiled with -Cpanic=abort, need to double-check how the assembly tests are built.

RalfJung · 2025-04-22T09:19:41Z

In this case I think it has to do with minicore being compiled with -Cpanic=abort, need to double-check how the assembly tests are built.

Maybe we just need to add -Cforce-unwind-tables to the minicore build flags then?

jieyouxu · 2025-04-22T09:56:04Z

Maybe we just need to add -Cforce-unwind-tables to the minicore build flags then?

That seems like a reasonable solution, I'll make that change separately and document these two preset flags in rdg.

jieyouxu · 2025-04-23T09:29:01Z

#140194 will have //@ add-core-stubs also imply -Cforce-unwind-tables=yes when building the test (not minicore itself).

…les, r=bjorn3 minicore: Have `//@ add-core-stubs` also imply `-Cforce-unwind-tables=yes` To preserve CFI directives in assembly tests, as `//@ add-core-stubs` already imply `-C panic=abort`. This is a blocker for rust-lang#140037 (comment). cc `@RalfJung` r? `@bjorn3`

…les, r=bjorn3 minicore: Have `//@ add-core-stubs` also imply `-Cforce-unwind-tables=yes` To preserve CFI directives in assembly tests, as `//@ add-core-stubs` already imply `-C panic=abort`. This is a blocker for rust-lang#140037 (comment). cc ``@RalfJung`` r? ``@bjorn3``

…les, r=bjorn3 minicore: Have `//@ add-core-stubs` also imply `-Cforce-unwind-tables=yes` To preserve CFI directives in assembly tests, as `//@ add-core-stubs` already imply `-C panic=abort`. This is a blocker for rust-lang#140037 (comment). cc ```@RalfJung``` r? ```@bjorn3```

Rollup merge of rust-lang#140194 - jieyouxu:minicore-force-unwind-tables, r=bjorn3 minicore: Have `//@ add-core-stubs` also imply `-Cforce-unwind-tables=yes` To preserve CFI directives in assembly tests, as `//@ add-core-stubs` already imply `-C panic=abort`. This is a blocker for rust-lang#140037 (comment). cc ```@RalfJung``` r? ```@bjorn3```

…orn3 minicore: Have `//@ add-core-stubs` also imply `-Cforce-unwind-tables=yes` To preserve CFI directives in assembly tests, as `//@ add-core-stubs` already imply `-C panic=abort`. This is a blocker for rust-lang/rust#140037 (comment). cc ```@RalfJung``` r? ```@bjorn3```

Mark-Simulacrum · 2025-05-04T15:59:07Z

@rustbot author

rustbot · 2025-05-04T15:59:11Z

Reminder, once the PR becomes ready for a review, use @rustbot ready.

hbina · 2025-05-05T14:34:02Z

@rustbot ready

hbina · 2025-05-05T14:35:23Z

Trying to trigger the CI

RalfJung · 2025-05-05T14:35:46Z

You closed this PR so apparently it's not ready? CI gets triggered automatically on pushes.

Note that to reopen a PR it must be in the same state as when you closed it, i.e. you have to push it back to whatever the current commit was at that time. Closing and reopening indeed triggers CI but that is rarely ever necessary.

hbina · 2025-05-05T14:36:23Z

Ah darn I force pushed :>

edit: I might have the branch in my backup...

hbina · 2025-05-05T22:51:47Z

Ok, now this is the error that I am getting
https://pastebin.com/QGcAWUH4
No idea what it means tho, so any pointers are appreciated

Another question, can you implement Ordering like this?

#[lang = "Ordering"]
#[repr(i8)]
pub enum Ordering {
    Less = 0,
    Equal = 1,
    Greater = 1,
}

because I got this error when I use -1

error: requires `panic_const_neg_overflow` lang_item
  --> /home/hbina085/git/rust-lang/rust/tests/auxiliary/minicore.rs:210:21
   |
LL |                     -self
   |                     ^^^^^
...
LL | impl_neg_trait!(isize, i8, i16, i32, i64, i128,);
   | ------------------------------------------------ in this macro invocation
   |
   = note: duplicate diagnostic emitted due to `-Z deduplicate-diagnostics=no`
   = note: this error originates in the macro `impl_neg_trait` (in Nightly builds, run with -Z macro-backtrace for more info)

And I can't find any usages of panic_const_neg_overflow...

hbina085@DESKTOP-PGERPVO ~/g/r/rust (hbina-update-use-minicore) [SIGINT]> rg 'panic_const_neg_overflow' --context 2 .                                                                                                        (base) 
./compiler/rustc_span/src/symbol.rs
1531-        panic_const_gen_fn_none_panic,
1532-        panic_const_mul_overflow,
1533:        panic_const_neg_overflow,
1534-        panic_const_rem_by_zero,
1535-        panic_const_rem_overflow,

./compiler/rustc_hir/src/lang_items.rs
297-    PanicDivOverflow,        sym::panic_const_div_overflow, panic_const_div_overflow, Target::Fn, GenericRequirement::None;
298-    PanicRemOverflow,        sym::panic_const_rem_overflow, panic_const_rem_overflow, Target::Fn, GenericRequirement::None;
299:    PanicNegOverflow,        sym::panic_const_neg_overflow, panic_const_neg_overflow, Target::Fn, GenericRequirement::None;
300-    PanicShrOverflow,        sym::panic_const_shr_overflow, panic_const_shr_overflow, Target::Fn, GenericRequirement::None;
301-    PanicShlOverflow,        sym::panic_const_shl_overflow, panic_const_shl_overflow, Target::Fn, GenericRequirement::None;

./compiler/rustc_codegen_cranelift/example/mini_core.rs
492-    panic_const_div_overflow = "attempt to divide with overflow",
493-    panic_const_rem_overflow = "attempt to calculate the remainder with overflow",
494:    panic_const_neg_overflow = "attempt to negate with overflow",
495-    panic_const_shr_overflow = "attempt to shift right with overflow",
496-    panic_const_shl_overflow = "attempt to shift left with overflow",

./compiler/rustc_codegen_gcc/example/mini_core.rs
496-    panic_const_div_overflow = "attempt to divide with overflow",
497-    panic_const_rem_overflow = "attempt to calculate the remainder with overflow",
498:    panic_const_neg_overflow = "attempt to negate with overflow",
499-    panic_const_shr_overflow = "attempt to shift right with overflow",
500-    panic_const_shl_overflow = "attempt to shift left with overflow",

./library/core/src/panicking.rs
191-        panic_const_div_overflow = "attempt to divide with overflow",
192-        panic_const_rem_overflow = "attempt to calculate the remainder with overflow",
193:        panic_const_neg_overflow = "attempt to negate with overflow",
194-        panic_const_shr_overflow = "attempt to shift right with overflow",
195-        panic_const_shl_overflow = "attempt to shift left with overflow",

bors · 2025-05-05T23:31:00Z

☔ The latest upstream changes (presumably #140664) made this pull request unmergeable. Please resolve the merge conflicts.

RalfJung · 2025-05-06T06:49:26Z

src/tools/miri/cargo-miri/src/arg.rs

Please don't reformat files you're not even changing in this PR. All of the Miri and Clippy changes should be undone.

RalfJung · 2025-05-06T06:50:36Z

And I can't find any usages of panic_const_neg_overflow...

Yeah, this is a lang item and it is inserted into the code by the compiler.

RalfJung · 2025-05-06T06:53:31Z

The errors that look like this

2025-05-05T15:24:16.9701876Z /checkout/tests/codegen/unwind-abis/win64-unwind-abi.rs:18:11: error: CHECK: expected string not found in input
2025-05-05T15:24:16.9702420Z // CHECK: @rust_item_that_can_unwind() unnamed_addr #1 {
2025-05-05T15:24:16.9702697Z           ^
2025-05-05T15:24:16.9703207Z /checkout/obj/build/x86_64-unknown-linux-gnu/test/codegen/unwind-abis/win64-unwind-abi/win64-unwind-abi.ll:7:70: note: scanning from here
2025-05-05T15:24:16.9703858Z define win64cc void @rust_item_that_cannot_unwind() unnamed_addr #0 {
2025-05-05T15:24:16.9704213Z                                                                      ^
2025-05-05T15:24:16.9704840Z /checkout/obj/build/x86_64-unknown-linux-gnu/test/codegen/unwind-abis/win64-unwind-abi/win64-unwind-abi.ll:13:21: note: possible intended match here
2025-05-05T15:24:16.9705595Z define win64cc void @rust_item_that_can_unwind() unnamed_addr #0 {

are caused by the attribute indices in the LLVM IR shifting around. I think it's possible to write codegen tests that are robust to such index shifts but I don't know how... the easiest fix is to just change the expected numbers in the test. Ensure to replace all occurrences of an index with the new one!

RalfJung · 2025-05-06T06:54:32Z

Given how many issues you are running into, I'd suggest reducing the scope of this PR to only change 1 or 2 files to use minicore. That'll make it much easier.

rustbot assigned Mark-Simulacrum Apr 19, 2025

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. T-compiler Relevant to the compiler team, which will review and decide on the PR/issue. labels Apr 19, 2025

hbina changed the title ~~Forgot to add the stubs~~ Extend minicore with intrinsics and use it to replace #[rustc_intrinsic] in tests Apr 19, 2025

This comment has been minimized.

Sign in to view

jieyouxu mentioned this pull request Apr 23, 2025

minicore: Have //@ add-core-stubs also imply -Cforce-unwind-tables=yes #140194

Merged

rustbot added S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels May 4, 2025

hbina closed this May 5, 2025

hbina force-pushed the hbina-update-use-minicore branch from e480e1f to 54d024e Compare May 5, 2025 14:30

rustbot added S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. and removed S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. labels May 5, 2025

RalfJung removed the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label May 5, 2025

RalfJung added the S-waiting-on-author Status: This is awaiting some action (such as code changes or more information) from the author. label May 5, 2025

WIP

96b3dca

hbina reopened this May 5, 2025

rustbot added A-rustdoc-json Area: Rustdoc JSON backend A-test-infra-minicore Area: `minicore` test auxiliary and `//@ add-core-stubs` PG-exploit-mitigations Project group: Exploit mitigations labels May 5, 2025

This comment has been minimized.

Sign in to view

RalfJung reviewed May 6, 2025

View reviewed changes

Moved catch_unwind and size_of<T>` to minicore

2d895a5

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Extend minicore with intrinsics and use it to replace #[rustc_intrinsic] in tests #140037

Extend minicore with intrinsics and use it to replace #[rustc_intrinsic] in tests #140037

hbina commented Apr 19, 2025 •

edited

Loading

rustbot commented Apr 19, 2025

hbina commented Apr 19, 2025

This comment has been minimized.

jieyouxu commented Apr 19, 2025 •

edited

Loading

jieyouxu commented Apr 20, 2025

RalfJung commented Apr 22, 2025

jieyouxu commented Apr 22, 2025

jieyouxu commented Apr 23, 2025 •

edited

Loading

Mark-Simulacrum commented May 4, 2025

rustbot commented May 4, 2025

hbina commented May 5, 2025

hbina commented May 5, 2025

RalfJung commented May 5, 2025 •

edited

Loading

hbina commented May 5, 2025 •

edited

Loading

This comment has been minimized.

hbina commented May 5, 2025 •

edited

Loading

bors commented May 5, 2025

RalfJung May 6, 2025 •

edited

Loading

RalfJung commented May 6, 2025

RalfJung commented May 6, 2025

RalfJung commented May 6, 2025 •

edited

Loading

Extend minicore with intrinsics and use it to replace #[rustc_intrinsic] in tests #140037

Are you sure you want to change the base?

Extend minicore with intrinsics and use it to replace #[rustc_intrinsic] in tests #140037

Conversation

hbina commented Apr 19, 2025 • edited Loading

rustbot commented Apr 19, 2025

hbina commented Apr 19, 2025

This comment has been minimized.

jieyouxu commented Apr 19, 2025 • edited Loading

jieyouxu commented Apr 20, 2025

RalfJung commented Apr 22, 2025

jieyouxu commented Apr 22, 2025

jieyouxu commented Apr 23, 2025 • edited Loading

Mark-Simulacrum commented May 4, 2025

rustbot commented May 4, 2025

hbina commented May 5, 2025

hbina commented May 5, 2025

RalfJung commented May 5, 2025 • edited Loading

hbina commented May 5, 2025 • edited Loading

This comment has been minimized.

hbina commented May 5, 2025 • edited Loading

bors commented May 5, 2025

RalfJung May 6, 2025 • edited Loading

Choose a reason for hiding this comment

RalfJung commented May 6, 2025

RalfJung commented May 6, 2025

RalfJung commented May 6, 2025 • edited Loading

hbina commented Apr 19, 2025 •

edited

Loading

jieyouxu commented Apr 19, 2025 •

edited

Loading

jieyouxu commented Apr 23, 2025 •

edited

Loading

RalfJung commented May 5, 2025 •

edited

Loading

hbina commented May 5, 2025 •

edited

Loading

hbina commented May 5, 2025 •

edited

Loading

RalfJung May 6, 2025 •

edited

Loading

RalfJung commented May 6, 2025 •

edited

Loading