On Thu, 18 Jun 2026 15:58:38 +0100 Peter Maydell <[email protected]> wrote:
> On Tue, 3 Mar 2026 at 09:26, Igor Mammedov <[email protected]> wrote: > > > > Comment about keeping legacy freq, is wrong to begin with > > (should be 62.5MHz), but that value also doesn't make > > sense anymore as the watchdog is used only by un-versioned > > SBSA board and the later has hard-coded it to 1GHz. > > > > Other potential user (arm/virt) also has system clock at 1GHz. > > It's not quite that simple. We use the back-compat 62.5MHz: > > * - for QEMU CPU types added before we standardized on 1GHz > * - for versioned machine types with a version of 9.0 or earlier > > sbsa-ref sets the same 1GHz frequency on both the wdt device > and also on the CPU objects it creates (because it doesn't have > to care about back-compat, so we were able to update it). > The virt board doesn't set the cpu cntfrq property, so it gets > 62.5MHz in the above situations. > > Changing the default value here is fine, though. > > (I haven't checked whether the subsequent patches in this > set make sure that when the virt board creates a watchdog > that it sets the clock-frequency here to match what the > CPU is is using.) Current patches don't set it explicitly as it matches current arm/virt. So for clock to match CPUs, it will need either to be explicitly set like sbsa does or we reuse compat props to fix it up to 62.5MHz on older machine types. > > -- PMM >
