[bug report] drm/dp: annotate implicit fall throughs
Dan Carpenter
dan.carpenter at oracle.com
Mon Aug 31 09:06:55 UTC 2020
On Fri, Aug 28, 2020 at 12:53:32PM +0300, Jani Nikula wrote:
> On Tue, 25 Aug 2020, Dan Carpenter <dan.carpenter at oracle.com> wrote:
> > Hello Mathieu Malaterre,
> >
> > The patch e9c0c874711b: "drm/dp: annotate implicit fall throughs"
> > from Jan 14, 2019, leads to the following static checker warning:
> >
> > drivers/gpu/drm/drm_dp_helper.c:495 drm_dp_downstream_max_bpc()
> > warn: ignoring unreachable code.
> >
> > drivers/gpu/drm/drm_dp_helper.c
> > 467 int drm_dp_downstream_max_bpc(const u8 dpcd[DP_RECEIVER_CAP_SIZE],
> > 468 const u8 port_cap[4])
> > 469 {
> > 470 int type = port_cap[0] & DP_DS_PORT_TYPE_MASK;
> > 471 bool detailed_cap_info = dpcd[DP_DOWNSTREAMPORT_PRESENT] &
> > 472 DP_DETAILED_CAP_INFO_AVAILABLE;
> > 473 int bpc;
> > 474
> > 475 if (!detailed_cap_info)
> > 476 return 0;
> > 477
> > 478 switch (type) {
> > 479 case DP_DS_PORT_TYPE_VGA:
> > 480 case DP_DS_PORT_TYPE_DVI:
> > 481 case DP_DS_PORT_TYPE_HDMI:
> > 482 case DP_DS_PORT_TYPE_DP_DUALMODE:
> > 483 bpc = port_cap[2] & DP_DS_MAX_BPC_MASK;
> > ^^^^^^^^^^^^^^^^^^
> > This is 0x3.
> >
> > 484
> > 485 switch (bpc) {
> > 486 case DP_DS_8BPC:
> > 487 return 8;
> > 488 case DP_DS_10BPC:
> > 489 return 10;
> > 490 case DP_DS_12BPC:
> > 491 return 12;
> > 492 case DP_DS_16BPC:
> > 493 return 16;
> > 494 }
> > 495 fallthrough;
> >
> > This fallthrough is impossible. Probably the way to work around the
> > bogus warning is the change the fallthough to "return 0; /* impossible */"
> > otherwise the fallthrough is sort of misleading...
>
> Won't that be unreachable as well?
>
> Maybe just add the default label to switch (bpc)?
A default switch is going to be unreachable as well... But adding a
return 0 is the older more traditional ways way of marking these so
people and checkers are used to it.
regards,
dan carpenter
More information about the dri-devel
mailing list