From: Bjorn Helgaas <helgaas@kernel.org>
To: Damien Dejean <dam.dejean@gmail.com>
Cc: linux-pci@vger.kernel.org
Subject: Re: Old Asus doesn't seem to support MSI
Date: Wed, 7 Jun 2023 16:53:26 -0500 [thread overview]
Message-ID: <20230607215326.GA1176980@bhelgaas> (raw)
In-Reply-To: <CAErgN1CYEA6a954vHHv3hkJe1jDY0eyEX0DPKf75nyd4_tD3UA@mail.gmail.com>
On Wed, Jun 07, 2023 at 06:24:34PM +0200, Damien Dejean wrote:
> You'll find the results of the two commands in attachment. I ran them
> without the pci=nomsi option set just in case.
Thanks!
The kernel does check the ACPI table [1] to see whether MSI is
supported, and I don't see the "system doesn't support MSI" message
[2] in your dmesg log, so I don't think that's the problem. It is
possible that the chipset doesn't support MSI but the BIOS neglected
to set the ACPI_FADT_NO_MSI bit in ACPI, although Windows looks at the
same table and should be similarly broken in that case.
I poked around on the web for similar reports of needing "pci=nomsi"
for Asus X73SL or Asus F70SL, but I really didn't find anything. The
linux-hardware database [3] claims this chipset is used in quite a few
systems, so I would think we'd see some reports of problems if the
chipset were broken. Unfortunately that database doesn't have enough
information to see whether systems use MSI (no "lspci -vv" output) and
I don't see a way to contact the reporters to find out.
We do have an existing quirk [4] that disables MSI for the [1039:0761]
chipset, and maybe we could add an entry for the [1039:0671] that you
have. The fact that I couldn't find any reports makes me worry that
doing that would make it work for you, but unnecessarily disable MSI
for others.
It's also conceivable that MSI used to work in older kernels, but we
broke something by v5.10. Do you know whether any old kernels ever
worked without "pci=nomsi"?
What exactly is the symptom you see without "pci=nomsi"? I know we
turned on more AER reporting recently, which exposed issues on some
machines. I think using "pci=nomsi" basically disables the AER
reporting, too. If the problem you see is related to AER, you can
turn that off with "pci=noaer", and you may not need "pci=nomsi".
Obviously my goal is that you should not need either one!
Bjorn
[1] https://uefi.org/specs/ACPI/6.5/05_ACPI_Software_Programming_Model.html?highlight=fadt#ia-pc-boot-architecture-flags
[2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/pci/pci-acpi.c?id=v6.3#n1475
[3] https://linux-hardware.org/?view=search&vendorid=1039&deviceid=0671#list
[4] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/pci/quirks.c?id=v6.3#n2557
next prev parent reply other threads:[~2023-06-07 21:54 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-06-03 15:52 Old Asus doesn't seem to support MSI Damien Dejean
2023-06-06 22:12 ` Bjorn Helgaas
2023-06-07 16:24 ` Damien Dejean
2023-06-07 21:53 ` Bjorn Helgaas [this message]
2023-06-08 7:57 ` Damien Dejean
2023-06-08 16:21 ` Bjorn Helgaas
2023-06-08 21:15 ` Damien Dejean
2023-06-08 21:53 ` Bjorn Helgaas
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20230607215326.GA1176980@bhelgaas \
--to=helgaas@kernel.org \
--cc=dam.dejean@gmail.com \
--cc=linux-pci@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).