linux-kernel - Re: [PATCH v3] PCI: Sanitise firmware BAR assignments behind a PCI-PCI bridge

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:   Fri, 23 Sep 2022 01:38:41 +0100 (BST)
From:   "Maciej W. Rozycki" <macro@...am.me.uk>
To:     Bjorn Helgaas <helgaas@...nel.org>
cc:     Bjorn Helgaas <bhelgaas@...gle.com>, linux-pci@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] PCI: Sanitise firmware BAR assignments behind a
 PCI-PCI bridge

On Wed, 21 Sep 2022, Bjorn Helgaas wrote:

> >  I have trimmed the change description down as you requested and left the 
> > change proper unmodified, as discussed in my earlier response.
> 
> I think this is great.  It shouldn't have taken me this long, so
> thanks for persevering.

 No worries, owing to various distractions it took me way too long to 
iterate over this change too.

> I think we can use pci_upstream_bridge() as below.  Let me know if
> not.

 Based on how the helper has been documented I think you are right.  Also 
I have verified your branch with my hardware and it still works.

> Here it is as I applied to pci/resource for v6.1:

 Thank you!  I find your rewritten change description a pleasure to read.

 I guess at this stage we'll never find out what the exact configuration 
was that has lead to commit 351fc6d1a517 ("PCI: Fix starting basis for 
resource requests").  At least it does not stand in the way.

 What a mess it was with the firmware side of the earlier PCI systems even 
once they've sorted the teething problems of the hardware side!  I still 
need to figure out how to get PCI interrupt routing to automagically work 
in the I/O APIC mode with this x86 machine and its broken MP-table which 
reports PCI interrupts as ISA interrupts, and then INTA only (at least our 
PIRQ router code now handles PCI-to-PCI bridges in the PIC mode).  And it 
would have been so easy to get it right even with a fixed table (the BIOS 
seems to build the table dynamically for no good reason even though the 
wiring is fixed in hardware)!

  Maciej