Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1232324 > unrolled thread

Re: [PATCH] PCI: disable MSI on SiS 761

Started byBjorn Helgaas <helgaas@kernel.org>
First post2015-09-24 20:30 +0200
Last post2015-09-24 21:30 +0200
Articles 2 — 2 participants

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Re: [PATCH] PCI: disable MSI on SiS 761 Bjorn Helgaas <helgaas@kernel.org> - 2015-09-24 20:30 +0200
    Re: [PATCH] PCI: disable MSI on SiS 761 Ondrej Zary <linux@rainbow-software.org> - 2015-09-24 21:30 +0200

#1232324 — Re: [PATCH] PCI: disable MSI on SiS 761

FromBjorn Helgaas <helgaas@kernel.org>
Date2015-09-24 20:30 +0200
SubjectRe: [PATCH] PCI: disable MSI on SiS 761
Message-ID<qcjrj-2AE-5@gated-at.bofh.it>
Hi Ondrej,

On Fri, Sep 11, 2015 at 11:12:17PM +0200, Ondrej Zary wrote:
> MSI is broken on SiS 761 chipset at least on PC Chips A31G board.
> No interrupts are delivered once MSI is enabled for a device. This causes
> hang on X11 start with a nVidia card installed (with nouveau driver).
> 
> Disable MSI completely for this chipset.
> 
> Signed-off-by: Ondrej Zary <linux@rainbow-software.org>

This is a pretty old chipset, isn't it?  Google finds a reference to
SiS 761GX from 2005.  I wonder why we're only finding this problem
now.  Is there a problem report?  I wonder if the problem is really
elsewhere, and turning off MSI just avoids the real problem.

Bjorn

> ---
>  drivers/pci/quirks.c |    1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> index e9fd0e9..6caffc3 100644
> --- a/drivers/pci/quirks.c
> +++ b/drivers/pci/quirks.c
> @@ -2204,6 +2204,7 @@ DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_VT3336, quirk_disab
>  DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_VT3351, quirk_disable_all_msi);
>  DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_VT3364, quirk_disable_all_msi);
>  DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_VIA, PCI_DEVICE_ID_VIA_8380_0, quirk_disable_all_msi);
> +DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_SI, 0x0761, quirk_disable_all_msi);
>  
>  /* Disable MSI on chipsets that are known to not support it */
>  static void quirk_disable_msi(struct pci_dev *dev)
> -- 
> Ondrej Zary
> 
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]


#1232375

FromOndrej Zary <linux@rainbow-software.org>
Date2015-09-24 21:30 +0200
Message-ID<qckno-3W1-13@gated-at.bofh.it>
In reply to#1232324
On Thursday 24 September 2015 20:21:16 Bjorn Helgaas wrote:
> Hi Ondrej,
>
> On Fri, Sep 11, 2015 at 11:12:17PM +0200, Ondrej Zary wrote:
> > MSI is broken on SiS 761 chipset at least on PC Chips A31G board.
> > No interrupts are delivered once MSI is enabled for a device. This causes
> > hang on X11 start with a nVidia card installed (with nouveau driver).
> >
> > Disable MSI completely for this chipset.
> >
> > Signed-off-by: Ondrej Zary <linux@rainbow-software.org>
>
> This is a pretty old chipset, isn't it?  Google finds a reference to
> SiS 761GX from 2005.  I wonder why we're only finding this problem
> now.  Is there a problem report?  I wonder if the problem is really
> elsewhere, and turning off MSI just avoids the real problem.

Yes, it's an old chipset and board (Socket 754) and not very common. Nouveau 
was not usable back then, MSI support was added to Nouveau in 2010 and 
enabled by default only in 2013. Maybe I'm the only one who tested MSI on 
this board/chipset...

The facts are:
 - zero interrupts in /proc/interrupts with MSI enabled and X11 hangs
 - everything works fine with MSI disabled
 - tested with two different nVidia cards (don't have any other PCIe card to 
test)

I guess that either the MSI support in the chipset is broken or the BIOS does 
not set up some secret chipset register(s) correctly.

-- 
Ondrej Zary
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [standalone]


Back to top | Article view | linux.kernel


csiph-web