[PATCH v2 1/7] iommu: Do not call pci_dev_reset_iommu_done() unless reset succeeds

[Nicolin Chen]

On Wed, Mar 18, 2026 at 07:21:53AM +0000, Tian, Kevin wrote:
> > From: Nicolin Chen <nicolinc at nvidia.com>
> > Sent: Wednesday, March 18, 2026 3:16 AM
> > 
> > @@ -977,7 +978,15 @@ int pci_dev_acpi_reset(struct pci_dev *dev, bool
> > probe)
> >  		ret = -ENOTTY;
> >  	}
> > 
> > -	pci_dev_reset_iommu_done(dev);
> > +	/*
> > +	 * The reset might be invoked to recover a serious error. E.g. when
> > the
> > +	 * ATC failed to invalidate its stale entries, which can result in data
> > +	 * corruption. Thus, do not unblock ATS until a successful reset.
> > +	 */
> > +	if (!ret || !pci_ats_supported(dev))
> > +		pci_dev_reset_iommu_done(dev);
> > +	else
> > +		pci_warn(dev, "Reset failed. Blocking ATS to protect
> > memory\n");
> >  	return ret;
> 
> let's pass the reset status to pci_dev_reset_iommu_done() then
> put above detail inside. 

Yea, that's cleaner. I will make a change:
@@ -4014,7 +4014,7 @@ EXPORT_SYMBOL_GPL(pci_dev_reset_iommu_prepare);
  * Note that, although unlikely, there is a risk that re-attaching domains might
  * fail due to some unexpected happening like OOM.
  */
-void pci_dev_reset_iommu_done(struct pci_dev *pdev)
+void pci_dev_reset_iommu_done(struct pci_dev *pdev, bool reset_succeeds)
 {
        struct iommu_group *group = pdev->dev.iommu_group;
        const struct iommu_ops *ops;
@@ -4025,6 +4025,11 @@ void pci_dev_reset_iommu_done(struct pci_dev *pdev)
                return;
        ops = dev_iommu_ops(&pdev->dev);

+       if (!reset_succeeds) {
+               pci_err(pdev, "Reset failed. Blocking ATS to protect memory\n");
+               return;
+       }
+
        guard(mutex)(&group->mutex);

        /* pci_dev_reset_iommu_prepare() was bypassed for the device */


Thanks
Nicolin