On 10/12/2019 17:49, Greg Kurz wrote: > On Tue, 10 Dec 2019 14:58:44 +0100 > Cédric Le Goater <c...@kaod.org> wrote: > >> Some PnvXScomInterface objects lie a bit deeper (PnvPBCQState) than > > I didn't find any trace of PnvPBCQState in the code... what is it ?
PHB4, which is not merged yet. C. > >> the first layer, so we need to loop on the whole object hierarchy to >> catch them. >> >> Signed-off-by: Cédric Le Goater <c...@kaod.org> >> --- >> hw/ppc/pnv_xscom.c | 7 ++++++- >> 1 file changed, 6 insertions(+), 1 deletion(-) >> >> diff --git a/hw/ppc/pnv_xscom.c b/hw/ppc/pnv_xscom.c >> index bed41840845e..006d87e970d9 100644 >> --- a/hw/ppc/pnv_xscom.c >> +++ b/hw/ppc/pnv_xscom.c >> @@ -326,7 +326,12 @@ int pnv_dt_xscom(PnvChip *chip, void *fdt, int >> root_offset) >> args.fdt = fdt; >> args.xscom_offset = xscom_offset; >> >> - object_child_foreach(OBJECT(chip), xscom_dt_child, &args); >> + /* >> + * Loop on the whole object hierarchy to catch all >> + * PnvXScomInterface objects which can lie a bit deeper the first > > s/deeper the first/deeper than the first/ > >> + * layer. >> + */ >> + object_child_foreach_recursive(OBJECT(chip), xscom_dt_child, &args); >> return 0; >> } >> >
