linux-kernel - Re: [PATCHv2 00/12] TDX: Enable Dynamic PAMT

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

Message-ID: <d21a66fe-d2ce-46cc-b89e-b60b03eae3da@intel.com>
Date: Wed, 13 Aug 2025 16:31:29 -0700
From: Dave Hansen <dave.hansen@...el.com>
To: "Edgecombe, Rick P" <rick.p.edgecombe@...el.com>,
 "seanjc@...gle.com" <seanjc@...gle.com>
Cc: "Gao, Chao" <chao.gao@...el.com>,
 "Yamahata, Isaku" <isaku.yamahata@...el.com>, "x86@...nel.org"
 <x86@...nel.org>, "kas@...nel.org" <kas@...nel.org>,
 "bp@...en8.de" <bp@...en8.de>,
 "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
 "Zhao, Yan Y" <yan.y.zhao@...el.com>, "Huang, Kai" <kai.huang@...el.com>,
 "mingo@...hat.com" <mingo@...hat.com>,
 "pbonzini@...hat.com" <pbonzini@...hat.com>,
 "linux-coco@...ts.linux.dev" <linux-coco@...ts.linux.dev>,
 "dave.hansen@...ux.intel.com" <dave.hansen@...ux.intel.com>,
 "tglx@...utronix.de" <tglx@...utronix.de>,
 "kvm@...r.kernel.org" <kvm@...r.kernel.org>
Subject: Re: [PATCHv2 00/12] TDX: Enable Dynamic PAMT

On 8/13/25 15:43, Edgecombe, Rick P wrote:
> I redid the test. Boot 10 TDs with 16GB of ram, run userspace to fault in memory
> from 4 threads until OOM, then shutdown. TDs were split between two sockets. It
> ended up with 1136 contentions of the global lock, 4ms waiting.

4ms out of how much CPU time?

Also, contention is *NOT* necessarily bad here. Only _false_ contention.

The whole point of the lock is to ensure that there aren't two different
CPUs trying to do two different things to the same PAMT range at the
same time.

If there are, one of them *HAS* to wait. It can wait lots of different
ways, but it has to wait. That wait will show up as spinlock contention.

Even if the global lock went away, that 4ms of spinning might still be
there.