From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Bjorn Helgaas <helgaas@kernel.org>, Yongji Xie <elohimes@gmail.com>
Cc: bhelgaas@google.com, linux-pci@vger.kernel.org,
linuxppc-dev@lists.ozlabs.org, alex.williamson@redhat.com,
gwshan@linux.vnet.ibm.com, aik@ozlabs.ru, mpe@ellerman.id.au,
paulus@samba.org, zhong@linux.vnet.ibm.com
Subject: Re: [PATCH v10 3/4] powerpc/powernv: Override pcibios_default_alignment() to force PCI devices to be page aligned
Date: Sat, 15 Apr 2017 07:52:54 +1000 [thread overview]
Message-ID: <1492206774.25766.14.camel@kernel.crashing.org> (raw)
In-Reply-To: <20170414155849.GB5864@bhelgaas-glaptop.roam.corp.google.com>
On Fri, 2017-04-14 at 10:58 -0500, Bjorn Helgaas wrote:
> On Mon, Apr 10, 2017 at 07:58:13PM +0800, Yongji Xie wrote:
> > This overrides pcibios_default_alignment() to set default alignment
> > to PAGE_SIZE for all PCI devices on PowerNV platform. Thus sub-page
> > BARs would not share a page and could be mapped into guest when
> > VFIO
> > passthrough them.
>
> Thanks for doing this. This looks like a reasonable strategy to me,
> but it would be good to get a powerpc ack for it.
I agree in principle. I'm surprised that PowerPC is the only one
interested here though, what about other platforms who want to use
KVM and PCI pass-through and use Linux to assign BARs ?
Cheers,
Ben.
> > Signed-off-by: Yongji Xie <elohimes@gmail.com>
> > ---
> > arch/powerpc/include/asm/machdep.h | 2 ++
> > arch/powerpc/kernel/pci-common.c | 8 ++++++++
> > arch/powerpc/platforms/powernv/pci-ioda.c | 7 +++++++
> > 3 files changed, 17 insertions(+)
> >
> > diff --git a/arch/powerpc/include/asm/machdep.h
> > b/arch/powerpc/include/asm/machdep.h
> > index 5011b69..a82c192 100644
> > --- a/arch/powerpc/include/asm/machdep.h
> > +++ b/arch/powerpc/include/asm/machdep.h
> > @@ -173,6 +173,8 @@ struct machdep_calls {
> > /* Called after scan and before resource survey */
> > void (*pcibios_fixup_phb)(struct pci_controller *hose);
> >
> > + resource_size_t (*pcibios_default_alignment)(struct
> > pci_dev *);
> > +
> > #ifdef CONFIG_PCI_IOV
> > void (*pcibios_fixup_sriov)(struct pci_dev *pdev);
> > resource_size_t (*pcibios_iov_resource_alignment)(struct
> > pci_dev *, int resno);
> > diff --git a/arch/powerpc/kernel/pci-common.c
> > b/arch/powerpc/kernel/pci-common.c
> > index ffda24a..ceda574 100644
> > --- a/arch/powerpc/kernel/pci-common.c
> > +++ b/arch/powerpc/kernel/pci-common.c
> > @@ -233,6 +233,14 @@ void pcibios_reset_secondary_bus(struct
> > pci_dev *dev)
> > pci_reset_secondary_bus(dev);
> > }
> >
> > +resource_size_t pcibios_default_alignment(struct pci_dev *pdev)
> > +{
> > + if (ppc_md.pcibios_default_alignment)
> > + return ppc_md.pcibios_default_alignment(pdev);
> > +
> > + return 0;
> > +}
> > +
> > #ifdef CONFIG_PCI_IOV
> > resource_size_t pcibios_iov_resource_alignment(struct pci_dev
> > *pdev, int resno)
> > {
> > diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c
> > b/arch/powerpc/platforms/powernv/pci-ioda.c
> > index e367382..354c852 100644
> > --- a/arch/powerpc/platforms/powernv/pci-ioda.c
> > +++ b/arch/powerpc/platforms/powernv/pci-ioda.c
> > @@ -3297,6 +3297,11 @@ static void pnv_pci_setup_bridge(struct
> > pci_bus *bus, unsigned long type)
> > }
> > }
> >
> > +static resource_size_t pnv_pci_default_alignment(struct pci_dev
> > *pdev)
> > +{
> > + return PAGE_SIZE;
> > +}
> > +
> > #ifdef CONFIG_PCI_IOV
> > static resource_size_t pnv_pci_iov_resource_alignment(struct
> > pci_dev *pdev,
> > int resno)
> > @@ -3830,6 +3835,8 @@ static void __init
> > pnv_pci_init_ioda_phb(struct device_node *np,
> > hose->controller_ops =
> > pnv_pci_ioda_controller_ops;
> > }
> >
> > + ppc_md.pcibios_default_alignment =
> > pnv_pci_default_alignment;
> > +
> > #ifdef CONFIG_PCI_IOV
> > ppc_md.pcibios_fixup_sriov =
> > pnv_pci_ioda_fixup_iov_resources;
> > ppc_md.pcibios_iov_resource_alignment =
> > pnv_pci_iov_resource_alignment;
> > --
> > 1.7.9.5
> >
next prev parent reply other threads:[~2017-04-14 21:53 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-04-10 11:58 [PATCH v10 0/4] PCI: Introduce a way to enforce all MMIO BARs not to share PAGE_SIZE Yongji Xie
2017-04-10 11:58 ` [PATCH v10 1/4] PCI: A fix for caculating bridge window's size and alignment Yongji Xie
2017-04-10 11:58 ` [PATCH v10 2/4] PCI: Add pcibios_default_alignment() for arch-specific alignment control Yongji Xie
2017-04-10 11:58 ` [PATCH v10 3/4] powerpc/powernv: Override pcibios_default_alignment() to force PCI devices to be page aligned Yongji Xie
2017-04-14 15:58 ` Bjorn Helgaas
2017-04-14 21:52 ` Benjamin Herrenschmidt [this message]
2017-04-15 16:36 ` Bjorn Helgaas
2017-04-15 22:06 ` Benjamin Herrenschmidt
2017-04-17 15:51 ` Bjorn Helgaas
2017-04-10 11:58 ` [PATCH v10 4/4] PCI: Don't extend device's size when using default alignment for all devices Yongji Xie
2017-04-14 22:54 ` Bjorn Helgaas
2017-04-17 6:27 ` Yongji Xie
2017-04-17 20:33 ` 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=1492206774.25766.14.camel@kernel.crashing.org \
--to=benh@kernel.crashing.org \
--cc=aik@ozlabs.ru \
--cc=alex.williamson@redhat.com \
--cc=bhelgaas@google.com \
--cc=elohimes@gmail.com \
--cc=gwshan@linux.vnet.ibm.com \
--cc=helgaas@kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mpe@ellerman.id.au \
--cc=paulus@samba.org \
--cc=zhong@linux.vnet.ibm.com \
/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).