linux-kernel - Re: [PATCH v2 1/4] PCI: qcom-ep: Mark BAR0/BAR2 as 64bit BARs and BAR1/BAR3 as RESERVED

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: <2024113005-oboe-widow-d61e@gregkh>
Date: Sat, 30 Nov 2024 17:17:58 +0100
From: Greg KH <gregkh@...uxfoundation.org>
To: Bjorn Helgaas <helgaas@...nel.org>
Cc: Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>, kw@...ux.com,
	arnd@...db.de, lpieralisi@...nel.org, shuah@...nel.org,
	kishon@...nel.org, aman1.gupta@...sung.com, p.rajanbabu@...sung.com,
	linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org,
	bhelgaas@...gle.com, linux-arm-msm@...r.kernel.org, robh@...nel.org,
	linux-kselftest@...r.kernel.org, stable+noautosel@...nel.org
Subject: Re: [PATCH v2 1/4] PCI: qcom-ep: Mark BAR0/BAR2 as 64bit BARs and
 BAR1/BAR3 as RESERVED

On Fri, Nov 29, 2024 at 01:55:37PM -0600, Bjorn Helgaas wrote:
> On Fri, Nov 29, 2024 at 02:54:12PM +0530, Manivannan Sadhasivam wrote:
> > On all Qcom endpoint SoCs, BAR0/BAR2 are 64bit BARs by default and software
> > cannot change the type. So mark the those BARs as 64bit BARs and also mark
> > the successive BAR1/BAR3 as RESERVED BARs so that the EPF drivers cannot
> > use them.
> 
> "Default" implies an initial setting that can be changed, but you say
> "by default" and also "software cannot change the type."  Can they be
> anything *other* than 64-bit BARs?
> 
> If they're hardwired to be 64-bit BARs, I would just say that.
> 
> > Cc: stable+noautosel@...nel.org # depends on patch introducing only_64bit flag
> 
> If stable maintainers need to act on this, do they need to search for
> the patch introducing only_64bit flag?  That seems onerous; is there a
> SHA1 that would make it easier?
> 
> > Fixes: f55fee56a631 ("PCI: qcom-ep: Add Qualcomm PCIe Endpoint controller driver")

Yes, having a sha1 that has that "feature addition" would be great,
surely that isn't 6.13-rc1, is it?

thanks,

greg k-h