On 14. 07. 21, 8:54, Jiri Slaby wrote:

quoted

@@ -3994,14 +3982,9 @@ pciserial_init_ports(struct pci_dev *dev, const 

struct pciserial_board *board)
      if (board->flags & FL_NOIRQ) {
          uart.port.irq = 0;
      } else {
-        if (pci_match_id(pci_use_msi, dev)) {
-            dev_dbg(&dev->dev, "Using MSI(-X) interrupts\n");
-            pci_set_master(dev);
-            rc = pci_alloc_irq_vectors(dev, 1, 1, PCI_IRQ_ALL_TYPES);
-        } else {
-            dev_dbg(&dev->dev, "Using legacy interrupts\n");
-            rc = pci_alloc_irq_vectors(dev, 1, 1, PCI_IRQ_LEGACY);
-        }
+        pci_set_master(dev);

But bus mastering is not about MSIs. I *think* it's still OK, but you 
need to document that in the commit log too.

Actually, why the commit which added this code turns on bus mastering?

Forget about this line, I wasn't woken enough. Of course, MSI (writes) 
to bus need bus mastering.

In any case, I'm still not sure what happens to devices which do not 
support MSI if we enable mastering on them?

-- 
js
suse labs