On Wed, Jul 14, 2021 at 08:57:06AM +0200, Jiri Slaby wrote:
> On 13. 07. 21, 12:40, Andy Shevchenko wrote:
> > The loop can be refactored by using ARRAY_SIZE() instead of NULL terminator.
> > This reduces code base and makes it easier to read and understand.
> > + iobase = request_region(inta_addr[i], ITE_887x_IOSIZE, "ite887x");
>
> Irrelevant whitespace change.
>
> > if (iobase != NULL) {
> > /* write POSIO0R - speed | size | ioport */
> > pci_write_config_dword(dev, ITE_887x_POSIO0,
> > ITE_887x_POSIO_ENABLE | ITE_887x_POSIO_SPEED |
> > ITE_887x_POSIO_IOSIZE_32 | inta_addr[i]);
> > /* write INTCBAR - ioport */
> > - pci_write_config_dword(dev, ITE_887x_INTCBAR,
> > - inta_addr[i]);
> > + pci_write_config_dword(dev, ITE_887x_INTCBAR, inta_addr[i]);
>
> detto
>
> > ret = inb(inta_addr[i]);
> > if (ret != 0xff) {
> > /* ioport connected */
> > break;
> > }
> > release_region(iobase->start, ITE_887x_IOSIZE);
> > - iobase = NULL;
> > }
> > - i++;
I believe with Joe's suggestion I can improve entire body of this branch
perhaps in a separate patch. Do you prefer one or two patches?
> > if (!inta_addr[i]) {
>
> OOB access?
Yep, Dan reported the same. Missed during conversion. Will fix.
--
With Best Regards,
Andy Shevchenko
