linux-kernel - [PATCH 0/2] x86/entry: Trade cycles

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-next>] [day] [month] [year] [list]

Message-Id: <20231120143344.584345121@infradead.org>
Date:   Mon, 20 Nov 2023 15:33:44 +0100
From:   Peter Zijlstra <peterz@...radead.org>
To:     x86@...nel.org
Cc:     linux-kernel@...r.kernel.org, peterz@...radead.org
Subject: [PATCH 0/2] x86/entry: Trade cycles

Hi,

Two little patches that trade a little performance. First patch optimizes
(although I couldn't get definite numbers showing it makes a difference) the
return to user path by avoiding some PTI specifics in the generic path.

Second patch then steals some of the won cycles by making a debug check
unconditional.

This came forth from a discussion with amluto who lamented we only had the
debug check conditional on DEBUG_ENTRY.