[WIP] Use assume rather than range metadata

[dianqk]

Fixes #122726.

Currently, range can only be used in load, call and invoke instructions. Due to SROA being run before other passes, even in the simplest IR, LLVM cannot infer that %i1 is 0. Most of the range metadata will be dropped before they work.

define noundef i32 @src(i32 noundef %arg) {
  %i = alloca i32, align 4
  store i32 %arg, ptr %i, align 4
  %i1 = load i32, ptr %i, align 4, !range !0
  ret i32 %i1
}

https://alive2.llvm.org/ce/z/MjsH9b

r? @ghost

[dianqk]

@bors try @rust-timer queue

[bors]

⌛ Trying commit f50d551 with merge 4c17eaa...

[rust-log-analyzer]

The job x86_64-gnu-llvm-17 failed! Check out the build log: (web) (plain)

Click to see the possible cause of the failure (guessed by this bot)

#16 exporting to docker image format
#16 sending tarball 44.0s done
#16 DONE 62.0s
##[endgroup]
Setting extra environment values for docker:  --env ENABLE_GCC_CODEGEN=1 --env GCC_EXEC_PREFIX=/usr/lib/gcc/
[CI_JOB_NAME=x86_64-gnu-llvm-17]
##[group]Clock drift check
  local time: Tue Mar 19 14:05:17 UTC 2024
  network time: Tue, 19 Mar 2024 14:05:17 GMT
  network time: Tue, 19 Mar 2024 14:05:17 GMT
##[endgroup]
sccache: Starting the server...
##[group]Configure the build
configure: processing command line
configure: 
configure: build.configure-args := ['--build=x86_64-unknown-linux-gnu', '--llvm-root=/usr/lib/llvm-17', '--enable-llvm-link-shared', '--set', 'rust.thin-lto-import-instr-limit=10', '--set', 'change-id=99999999', '--enable-verbose-configure', '--enable-sccache', '--disable-manage-submodules', '--enable-locked-deps', '--enable-cargo-native-static', '--set', 'rust.codegen-units-std=1', '--set', 'dist.compression-profile=balanced', '--dist-compression-formats=xz', '--disable-dist-src', '--release-channel=nightly', '--enable-debug-assertions', '--enable-overflow-checks', '--enable-llvm-assertions', '--set', 'rust.verify-llvm-ir', '--set', 'rust.codegen-backends=llvm,cranelift,gcc', '--set', 'llvm.static-libstdcpp', '--enable-new-symbol-mangling']
configure: target.x86_64-unknown-linux-gnu.llvm-config := /usr/lib/llvm-17/bin/llvm-config
configure: llvm.link-shared     := True
configure: rust.thin-lto-import-instr-limit := 10
configure: change-id            := 99999999
---
failures:

---- [ui] tests/ui/abi/stack-protector.rs#no-ssp stdout ----

error in revision `no-ssp`: test run failed!
status: signal: 4 (SIGILL) (core dumped)
command: cd "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/abi/stack-protector.no-ssp" && RUST_TEST_THREADS="16" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/abi/stack-protector.no-ssp/a"
status: exit status: 0
status: exit status: 0
stdout: pwned!
------------------------------------------
stderr: none



---- [ui] tests/ui/abi/stack-protector.rs#ssp stdout ----

error in revision `ssp`: test run failed!
status: signal: 4 (SIGILL) (core dumped)
command: cd "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/abi/stack-protector.ssp" && RUST_TEST_THREADS="16" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/abi/stack-protector.ssp/a"
--- stdout -------------------------------
status: signal: 6 (SIGABRT) (core dumped)
stdout: 
stderr: *** stack smashing detected ***: terminated
stderr: none


---- [ui] tests/ui/backtrace.rs stdout ----
---- [ui] tests/ui/backtrace.rs stdout ----

error: test run failed!
status: signal: 4 (SIGILL) (core dumped)
command: cd "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/backtrace" && RUST_TEST_THREADS="16" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/backtrace/a"
stderr: none


---- [ui] tests/ui/std-backtrace.rs stdout ----
---- [ui] tests/ui/std-backtrace.rs stdout ----

error: test run failed!
status: signal: 4 (SIGILL) (core dumped)
command: cd "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/std-backtrace" && RUST_TEST_THREADS="16" "/checkout/obj/build/x86_64-unknown-linux-gnu/test/ui/std-backtrace/a"
stderr: none

[bors]

☀️ Try build successful - checks-actions
Build commit: 4c17eaa (4c17eaae24ad25a8e12b6a6a79f95266853fd829)

[rust-timer]

Finished benchmarking commit (4c17eaa): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

@bors rollup=never
@rustbot label: -S-waiting-on-perf +perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	2.4%	[0.3%, 4.7%]	32
Regressions ❌ (secondary)	4.4%	[0.2%, 64.0%]	40
Improvements ✅ (primary)	-1.4%	[-25.2%, -0.2%]	91
Improvements ✅ (secondary)	-0.7%	[-1.8%, -0.1%]	27
All ❌✅ (primary)	-0.4%	[-25.2%, 4.7%]	123

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	2.5%	[1.3%, 3.7%]	2
Regressions ❌ (secondary)	2.1%	[2.1%, 2.1%]	1
Improvements ✅ (primary)	-5.3%	[-5.3%, -5.3%]	1
Improvements ✅ (secondary)	-3.8%	[-3.8%, -3.8%]	1
All ❌✅ (primary)	-0.1%	[-5.3%, 3.7%]	3

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	2.5%	[1.3%, 3.6%]	23
Regressions ❌ (secondary)	11.1%	[2.4%, 48.6%]	11
Improvements ✅ (primary)	-9.2%	[-26.8%, -1.7%]	12
Improvements ✅ (secondary)	-3.0%	[-3.5%, -2.1%]	3
All ❌✅ (primary)	-1.5%	[-26.8%, 3.6%]	35

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.3%	[0.0%, 2.1%]	55
Regressions ❌ (secondary)	1.3%	[0.0%, 3.3%]	87
Improvements ✅ (primary)	-6.0%	[-14.0%, -0.1%]	27
Improvements ✅ (secondary)	-0.8%	[-0.8%, -0.8%]	1
All ❌✅ (primary)	-1.8%	[-14.0%, 2.1%]	82

Bootstrap: 669.996s -> 691.729s (3.24%)
Artifact size: 312.80 MiB -> 312.82 MiB (0.00%)

[dianqk]

I would prefer to add range on arguments with the support of LLVM 19.