linux-kernel - Re: [PATCH v7 4/5] rust: add find_bit_benchmark

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <aBCJQQWnQ-TRMECr@google.com>
Date: Tue, 29 Apr 2025 08:09:37 +0000
From: Alice Ryhl <aliceryhl@...gle.com>
To: Yury Norov <yury.norov@...il.com>
Cc: Burak Emir <bqe@...gle.com>, Boqun Feng <boqun.feng@...il.com>, 
	Rasmus Villemoes <linux@...musvillemoes.dk>, Viresh Kumar <viresh.kumar@...aro.org>, 
	Miguel Ojeda <ojeda@...nel.org>, Alex Gaynor <alex.gaynor@...il.com>, Gary Guo <gary@...yguo.net>, 
	"Björn Roy Baron" <bjorn3_gh@...tonmail.com>, Benno Lossin <benno.lossin@...ton.me>, 
	Andreas Hindborg <a.hindborg@...nel.org>, Trevor Gross <tmgross@...ch.edu>, 
	rust-for-linux@...r.kernel.org, linux-kernel@...r.kernel.org, 
	Rong Xu <xur@...gle.com>
Subject: Re: [PATCH v7 4/5] rust: add find_bit_benchmark_rust module.

On Mon, Apr 28, 2025 at 09:21:40AM -0400, Yury Norov wrote:
> > By the way, if you add assert_eq!(bitmap.len(), BITMAP_LEN) before the
> > loop you may get the bounds checks optimized out.
> 
> That sounds cheating, isn't?
> 
> I think nobody will reject this series because of +15% in performance
> test, neither +25%, or whatever reasonable number.
> 
> Let's just measure fair numbers and deliver clear maintainable code. 
> There's a single user for bitmaps in rust so far, and it's you. So
> you're to call if performance is bad for you, not me. I just want
> to make sure that your numbers are withing the sane boundaries.

Right, well, in Binder's case the relevant performance comparison is a
single call to the bitmap's next_zero_bit vs following a sorted linked
list to find the first index that isn't used by an element in the list,
so I'm pretty confident that the bitmap will win no matter what.

Alice