On Nov 27, 2024 / 10:55, Hans de Goede wrote: > Hi, > > Seems I was a bit too quick with reviewing at a second > look I have found a small issue with this patch. > > See my comment below. > > On 27-Nov-24 7:00 AM, Shin'ichiro Kawasaki wrote: [...] > > @@ -167,7 +163,12 @@ static int p2sb_cache_resources(void) > > pci_bus_read_config_dword(bus, devfn_p2sb, P2SBC, &value); > > p2sb_hidden_by_bios = value & P2SBC_HIDE; > > > > - ret = p2sb_scan_and_cache(bus, devfn_p2sb); > > + /* > > + * If the BIOS does not hide the P2SB device then its resources > > + * are accesilble. Cache them only if the P2SB device is hidden. > > + */ > > + if (p2sb_hidden_by_bios) > > + ret = p2sb_scan_and_cache(bus, devfn_p2sb); > > ret will be returned uninitialized now when p2sb_hidden_by_bios is false, > so this patch also needs to initialize ret to 0 when declaring it. Ah, right. Will fix it in v4. I compile tested with KCFLAGS=-Wall and expected it would catch such mistakes, but it didn't. I found that -Wmaybe-uninitialized does the check. Will use this check for my future patches. > > With this fixed you can keep my Reviewed-by. Thanks! Will send out v4 soon.
