ChanServ changed the topic of #rust-embedded to: Welcome to the Rust Embedded IRC channel! Bridged to #rust-embedded:matrix.org and logged at https://libera.irclog.whitequark.org/rust-embedded, code of conduct at https://www.rust-lang.org/conduct.html
Noah[m] has quit [Quit: Idle timeout reached: 172800s]
i509vcb[m] has quit [Quit: Idle timeout reached: 172800s]
cr1901_ has joined #rust-embedded
M762spr[m] has quit [Quit: Idle timeout reached: 172800s]
cr1901 has quit [Ping timeout: 252 seconds]
cr1901_ is now known as cr1901
KevinPFleming[m] has joined #rust-embedded
<KevinPFleming[m]> yow... Rust 1.87 shrank the binary size rather significantly for my project
<KevinPFleming[m]> .text section shrank from 14,820 bytes to 12,870
Amit[m] has joined #rust-embedded
<Amit[m]> Kevin P. Fleming: what was your previous version?
<KevinPFleming[m]> 1.86.0, I have stable in rust-toolchain.toml
<Amit[m]> (I'm seeing that the older nightly we'd been using is marked rustc 1.87.0-nightly so likely already included code size optimizations, e.g. newer LLVM or something. No change for me.)
<JamesMunns[m]> 1.87 switched to llvm20 iirc
<Amit[m]> Those of you/us using RISC-V, expect even more dramatic improvements over the next while. RISC-V size optimizations in LLVM at the moment are still fairly atrocious, but it's actively being worked on and there's no magic needed. Just elbow grease from the right interested people, which seems to be happening.
<Amit[m]> Kevin P. Fleming: it's something I really enjoy about working within such an active ecosystem/community. Things that sucked a while ago are all of sudden good with no effort on my part! Like when we used to buy a new CPU and all of a sudden my embarrassingly slow code was 1/2 as slow :)
sroemer has joined #rust-embedded
sroemer has quit [Changing host]
sroemer has joined #rust-embedded
_whitelogger has joined #rust-embedded
pcs38 has joined #rust-embedded
Matt[m] has quit [Quit: Idle timeout reached: 172800s]
vollbrecht[m] has quit [Quit: Idle timeout reached: 172800s]
<Kaspar[m]> <KevinPFleming[m]> ".text section shrank from 14,8..." <- yeah, we're also seeing 5-10% code size reduction in the Ariel examples. Nice!
chrysn[m] has joined #rust-embedded
<chrysn[m]> I'd have to double-check, but I think this is also why when I recently rebuilt a program that shouldn't take 40 seconds to build, it was way faster today. (Could be bad measurements too, though.)
<jason-kairos[m]> do floating point types in rust have a well-know / fixed representation regardless of hardware?
<jason-kairos[m]> * do floating point types in rust have a well-known / fixed binary representation regardless of hardware?
GrantM11235[m] has joined #rust-embedded
<GrantM11235[m]> Yes
<GrantM11235[m]> >A 32-bit floating-point type (specifically, the “binary32” type defined in IEEE 754-2008).
<jason-kairos[m]> do those binary size improvements only apply to riscv or do they apply to arm cortex as well?
pcs38 has quit [Quit: leaving]
<KevinPFleming[m]> I'm using Cortex-M0.
ithinuel[m] has quit [Quit: Idle timeout reached: 172800s]
sroemer has quit [Quit: WeeChat 4.5.2]
i509vcb[m] has joined #rust-embedded
<i509vcb[m]> I think ARM is trying their hand at "how confusing can naming become": https://newsroom.arm.com/news/new-arm-product-naming-architecture
<thejpster[m]> Unclear if the new names are only for Compute Subsystems, or for the processors too.
<thejpster[m]> Or how we do a crate that works across named platforms when the underlying technology is the same, it’s just segmented by market.
<thejpster[m]> Lots of questions. Wait and see I guess.
pcs38 has joined #rust-embedded
LeandroMarceddu[ has joined #rust-embedded
<LeandroMarceddu[> Oh so it can get worse..
_whitelogger has joined #rust-embedded
pcs38 has quit [Quit: leaving]
jannic[m] has quit [Quit: Idle timeout reached: 172800s]
ivche_ has joined #rust-embedded
ivche has quit [Ping timeout: 272 seconds]
ivche_ is now known as ivche
rainingmessages has joined #rust-embedded