linux-kernel - Re: [PATCH v3 1/1] PCI: Fix bug resulting in double hpmemsize being assigned to MMIO window

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: <20191119133828.GA171084@google.com>
Date:   Tue, 19 Nov 2019 07:38:28 -0600
From:   Bjorn Helgaas <helgaas@...nel.org>
To:     Nicholas Johnson <nicholas.johnson-opensource@...look.com.au>
Cc:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "linux-pci@...r.kernel.org" <linux-pci@...r.kernel.org>,
        "mika.westerberg@...ux.intel.com" <mika.westerberg@...ux.intel.com>,
        "corbet@....net" <corbet@....net>,
        "benh@...nel.crashing.org" <benh@...nel.crashing.org>,
        "logang@...tatee.com" <logang@...tatee.com>
Subject: Re: [PATCH v3 1/1] PCI: Fix bug resulting in double hpmemsize being
 assigned to MMIO window

On Tue, Nov 19, 2019 at 03:17:04AM +0000, Nicholas Johnson wrote:
> I did just discover linux-next and I built it. Should I be doing this 
> more often to help find regressions?

Yes, if you build and run linux-next, that's a great service because
it helps find problems before they appear in mainline.

> I will now concentrate on fixing the problem where pci=nocrs does not 
> ignore the bus resource. One motherboard I own gives 00-7e or similar, 
> instead of 00-ff. The nocrs does not help, and I had to patch the kernel 
> myself. Only acpi=off fixes the problem, while knocking out SMT (MADT), 
> IOMMU (DMAR) and the ability to suspend without crashing.
> 
> If you disagree that nocrs should override bus resource, then let me 
> know and I will not attempt this.

I guess the problem is that with "pci=nocrs", we ignore the MMIO and
I/O port resources from _CRS, but we still pay attention to bus number
resources in _CRS?  That does sound like it would be unexpected
behavior.

I *would* like to see the complete dmesg log because these _CRS
methods are pretty reliable because Windows relies on them as well, so
problems are frequently a result of Linux defects.  If we can fix
Linux or automatically work around issues so users don't have to use
"pci=nocrs" explicitly, that's the best.

Bjorn