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

[Long Li]

> -----Original Message-----
> From: Stephen Hemminger [mailto:stephen@xxxxxxxxxxxxxxxxxx]
> Sent: Monday, December 5, 2016 8:53 AM
> 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: [PATCH] pci-hyperv: use kmalloc to allocate hypercall params
> buffer
> 
> On Tue,  8 Nov 2016 14:04:38 -0800
> Long Li <longli@xxxxxxxxxxxxxxxxxxxxxx> wrote:
> 
> > +	spin_lock_irqsave(&hbus->retarget_msi_interrupt_lock, flags);
> > +
> > +	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);
> > +	spin_unlock_irqrestore(&hbus->retarget_msi_interrupt_lock, flags);
> 
> It looks like the additional locking here is being overly paranoid.
> The caller is already holding the irq descriptor lock. Look at fixup_irqs.

You are right. On my test machine, there are two possible places calling hv_irq_unmask(): request _irq() and handle_edge_irq(). They both have desc->lock held when calling .irq_unmask on the chip. A review of the IRQ code shows that desc->lock is always held while calling chip->irq_unmask().

Since the lock doesn't do any harm and it is not on performance code path, we can remove the lock in the upcoming patches.