On Thu, Sep 08, 2022 at 10:20:47AM +0200, Ard Biesheuvel wrote: > On Wed, 7 Sept 2022 at 18:57, Andy Shevchenko > <andriy.shevchenko@xxxxxxxxxxxxxxx> wrote: ... > > for_each_acpi_dev_match(adev, hid, NULL, -1) { > > - if (adev->pnp.unique_id && !strcmp(adev->pnp.unique_id, uid)) > > + ret = acpi_dev_uid_to_integer(adev, &uid); > > + if (ret == -ENODATA && node->acpi.uid == 0) > > break; > > - if (!adev->pnp.unique_id && node->acpi.uid == 0) > > + if (ret == 0 && node->acpi.uid == uid) > > Is it necessary to reorder the conditions here? I.e., why not Code-wise there should be not much difference which does not affect the flow, I think I moved it to be closer to the pattern "let's handle errors first", but in this case I'm fine with your proposal. > > + ret = acpi_dev_uid_to_integer(adev, &uid); > > + if (ret == 0 && node->acpi.uid == uid) > > break; > > + if (ret == -ENODATA && node->acpi.uid == 0) > > break; > > ? > > With that fixed, > > Reviewed-by: Ard Biesheuvel <ardb@xxxxxxxxxx> Thanks! -- With Best Regards, Andy Shevchenko