RE: [Resend] [PATCH] pci-hyperv: use kmalloc to allocate hypercall params buffer

[Long Li]

> -----Original Message-----
> From: devel [mailto:driverdev-devel-bounces@xxxxxxxxxxxxxxxxxxxxxx] On
> Behalf Of Long Li
> Sent: Tuesday, November 8, 2016 8:57 AM
> To: Greg KH <gregkh@xxxxxxxxxxxxxxxxxxx>
> Cc: linux-pci@xxxxxxxxxxxxxxx; Haiyang Zhang <haiyangz@xxxxxxxxxxxxx>;
> linux-kernel@xxxxxxxxxxxxxxx; Bjorn Helgaas <bhelgaas@xxxxxxxxxx>;
> devel@xxxxxxxxxxxxxxxxxxxxxx
> Subject: RE: [Resend] [PATCH] pci-hyperv: use kmalloc to allocate hypercall
> params buffer
> 
> This sender failed our fraud detection checks and may not be who they
> appear to be. Learn about spoofing at http://aka.ms/LearnAboutSpoofing
> 
> > -----Original Message-----
> > From: Greg KH [mailto:gregkh@xxxxxxxxxxxxxxxxxxx]
> > Sent: Monday, November 7, 2016 11:00 PM
> > To: Long Li <longli@xxxxxxxxxxxxx>
> > Cc: KY Srinivasan <kys@xxxxxxxxxxxxx>; Haiyang Zhang
> > <haiyangz@xxxxxxxxxxxxx>; Bjorn Helgaas <bhelgaas@xxxxxxxxxx>;
> > devel@xxxxxxxxxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx; linux-
> > pci@xxxxxxxxxxxxxxx
> > Subject: Re: [Resend] [PATCH] pci-hyperv: use kmalloc to allocate
> > hypercall params buffer
> >
> > On Tue, Nov 08, 2016 at 12:14:14AM -0800, Long Li wrote:
> > > From: Long Li <longli@xxxxxxxxxxxxx>
> > >
> > > hv_do_hypercall assumes that we pass a segment from a physically
> > continuous buffer. Buffer allocated on the stack may not work if
> > CONFIG_VMAP_STACK=y is set. Use kmalloc to allocate this buffer.
> >
> > Please wrap your changelog at 72 columns.
> >
> > >
> > > Signed-off-by: Long Li <longli@xxxxxxxxxxxxx>
> > > Reported-by: Haiyang Zhang <haiyangz@xxxxxxxxxxxxx>
> > > ---
> > >  drivers/pci/host/pci-hyperv.c | 24 +++++++++++++-----------
> > >  1 file changed, 13 insertions(+), 11 deletions(-)
> > >
> > > diff --git a/drivers/pci/host/pci-hyperv.c
> > > b/drivers/pci/host/pci-hyperv.c index 763ff87..97e6daf 100644
> > > --- a/drivers/pci/host/pci-hyperv.c
> > > +++ b/drivers/pci/host/pci-hyperv.c
> > > @@ -378,6 +378,7 @@ struct hv_pcibus_device {
> > >     struct msi_domain_info msi_info;
> > >     struct msi_controller msi_chip;
> > >     struct irq_domain *irq_domain;
> > > +   struct retarget_msi_interrupt retarget_msi_interrupt_params;
> >
> > Can you handle potentially unaligned accesses like this?  Is there
> > some lock preventing you from using this structure more than once at the
> same time?
> >
> > >  };
> > >
> > >  /*
> > > @@ -774,7 +775,7 @@ void hv_irq_unmask(struct irq_data *data)  {
> > >     struct msi_desc *msi_desc = irq_data_get_msi_desc(data);
> > >     struct irq_cfg *cfg = irqd_cfg(data);
> > > -   struct retarget_msi_interrupt params;
> > > +   struct retarget_msi_interrupt *params;
> > >     struct hv_pcibus_device *hbus;
> > >     struct cpumask *dest;
> > >     struct pci_bus *pbus;
> > > @@ -785,23 +786,24 @@ void hv_irq_unmask(struct irq_data *data)
> > >     pdev = msi_desc_to_pci_dev(msi_desc);
> > >     pbus = pdev->bus;
> > >     hbus = container_of(pbus->sysdata, struct hv_pcibus_device,
> > > sysdata);
> > > -
> > > -   memset(&params, 0, sizeof(params));
> > > -   params.partition_id = HV_PARTITION_ID_SELF;
> > > -   params.source = 1; /* MSI(-X) */
> > > -   params.address = msi_desc->msg.address_lo;
> > > -   params.data = msi_desc->msg.data;
> > > -   params.device_id = (hbus->hdev->dev_instance.b[5] << 24) |
> > > +   params = &hbus->retarget_msi_interrupt_params;
> > > +
> > > +   memset(params, 0, sizeof(*params));
> > > +   params->partition_id = HV_PARTITION_ID_SELF;
> > > +   params->source = 1; /* MSI(-X) */
> > > +   params->address = msi_desc->msg.address_lo;
> > > +   params->data = msi_desc->msg.data;
> > > +   params->device_id = (hbus->hdev->dev_instance.b[5] << 24) |
> > >                        (hbus->hdev->dev_instance.b[4] << 16) |
> > >                        (hbus->hdev->dev_instance.b[7] << 8) |
> > >                        (hbus->hdev->dev_instance.b[6] & 0xf8) |
> > >                        PCI_FUNC(pdev->devfn);
> > > -   params.vector = cfg->vector;
> > > +   params->vector = cfg->vector;
> > >
> > >     for_each_cpu_and(cpu, dest, cpu_online_mask)
> > > -           params.vp_mask |= (1ULL <<
> > vmbus_cpu_number_to_vp_number(cpu));
> > > +           params->vp_mask |= (1ULL <<
> > vmbus_cpu_number_to_vp_number(cpu));
> > >
> > > -   hv_do_hypercall(HVCALL_RETARGET_INTERRUPT, &params, NULL);
> > > +   hv_do_hypercall(HVCALL_RETARGET_INTERRUPT, params, NULL);
> >
> > As you only use this in one spot, why not just allocate it here and
> > then free it?  Why add it to the pcibus device structure?
> 
> Thanks Greg. I will send a V2.

Sorry forgot to address the reason why we don't just allocate the buffer here. This function cannot fail. So it's better to pre-allocate the buffer.

> 
> >
> > thanks,
> >
> > greg k-h
> _______________________________________________
> devel mailing list
> devel@xxxxxxxxxxxxxxxxxxxxxx
> https://na01.safelinks.protection.outlook.com/?url=http%3A%2F%2Fdriverd
> ev.linuxdriverproject.org%2Fmailman%2Flistinfo%2Fdriverdev-
> devel&data=02%7C01%7Clongli%40microsoft.com%7C6e28f5459da345cdb5e
> 408d407f836f0%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C636142
> 210041216500&sdata=vvVxGQUet7KMuRgs9%2BRbR8JE70rKF1AJo%2Fu3zx%
> 2FQNwY%3D&reserved=0