linux-kernel - Re: [PATCH] drivers/iommu: Ensure that the queue base address is successfully written during SMMU initialization.

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]

Date: Sat, 24 Feb 2024 00:20:10 +0800
From: "ni.liqiang" <ni_liqiang@....com>
To: will@...nel.org,
	danielmentz@...gle.com
Cc: iommu@...ts.linux.dev,
	jin.qi@....com.cn,
	joro@...tes.org,
	linux-arm-kernel@...ts.infradead.org,
	linux-kernel@...r.kernel.org,
	ni.liqiang@....com.cn,
	niliqiang.io@...il.com,
	robin.murphy@....com
Subject: Re: [PATCH] drivers/iommu: Ensure that the queue base address is successfully written during SMMU initialization.

> (Disclaimer: I don't know what a CCG port is)

CCG(CXL Gateway) is a part of the CMN700 Coherent Mesh Network. It plays
a crucial role in facilitating cross-die access in a multi-chip system.

> Hmmm. The part that doesn't make sense to me here is that migrating between
> CPUs implies context-switching, and we have a DSB on that path in
> __switch_to(). So why would adding barriers to the driver help? Maybe it
> just changes the timing?

This is very likely. Through our experiments, adding a delay before CMDQEN
does not reproduce the failure of writing to the CMDQ base register.

> I'm not sure what you're proposing, but I don't think Linux should be
> changed to accomodate this.

I am very grateful for the responses from both of you experts. We will
continue to check the current hardware configuration of the system
and attempt to fix this issue.
Once again, I express my thanks. Thank you.