llvm/llvm-project

missed fold, sext(fptosi x to i32) => fptosi x to i64

Open

#85,268 opened on Mar 14, 2024

View on GitHub
 (13 comments) (0 reactions) (1 assignee)C++ (10,782 forks)batch import
floating-pointgood first issuellvm:codegenmissed-optimization

Repository metrics

Stars
 (26,378 stars)
PR merge metrics
 (Avg merge 1d 2h) (1,000 merged PRs in 30d)

Description

https://alive2.llvm.org/ce/z/bFW9vZ online Alive timed out for this one, you'll need local alive2 with larger smt-to.

define i64 @src(float %0) {
if.end27:
%1 = fptosi float %0 to i32
%2 = sext i32 %1 to i64
ret i64 %2

sink: ; No predecessors!
unreachable
}

define i64 @tgt(float %0) {
if.end27:
%1 = fptosi float %0 to i64
ret i64 %1

sink: ; No predecessors!
unreachable
}

no brainer, same pattern applies to fptoui.

Contributor guide

Research direction
Explore the LLVM InstCombine pass to add a fold for sext(fptosi x to i32) to fptosi x to i64, and similarly for fptoui. Study existing similar folds for patterns like sext(trunc) and fptosi to larger types.
Tech stack
None
Domain
backend
Issue type
Bug
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
Half day
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
Active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
Clear
Prerequisites
C++LLVMGit
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
40