Re: [PATCH] drm/i915: Distribute switch variables for initialization

[Kees Cook]

On Thu, Feb 20, 2020 at 12:21:14PM +0200, Jani Nikula wrote:
> On Wed, 19 Feb 2020, Kees Cook <keescook@xxxxxxxxxxxx> wrote:
> > Variables declared in a switch statement before any case statements
> > cannot be automatically initialized with compiler instrumentation (as
> > they are not part of any execution flow). With GCC's proposed automatic
> > stack variable initialization feature, this triggers a warning (and they
> > don't get initialized). Clang's automatic stack variable initialization
> > (via CONFIG_INIT_STACK_ALL=y) doesn't throw a warning, but it also
> > doesn't initialize such variables[1]. Note that these warnings (or silent
> > skipping) happen before the dead-store elimination optimization phase,
> > so even when the automatic initializations are later elided in favor of
> > direct initializations, the warnings remain.
> >
> > To avoid these problems, move such variables into the "case" where
> > they're used or lift them up into the main function body.
> >
> > drivers/gpu/drm/i915/display/intel_display.c: In function âcheck_digital_port_conflictsâ:
> > drivers/gpu/drm/i915/display/intel_display.c:12963:17: warning: statement will never be executed [-Wswitch-unreachable]
> > 12963 |    unsigned int port_mask;
> >       |                 ^~~~~~~~~
> >
> > drivers/gpu/drm/i915/intel_pm.c: In function âvlv_get_fifo_sizeâ:
> > drivers/gpu/drm/i915/intel_pm.c:474:7: warning: statement will never be executed [-Wswitch-unreachable]
> >   474 |   u32 dsparb, dsparb2, dsparb3;
> >       |       ^~~~~~
> > drivers/gpu/drm/i915/intel_pm.c: In function âvlv_atomic_update_fifoâ:
> > drivers/gpu/drm/i915/intel_pm.c:1997:7: warning: statement will never be executed [-Wswitch-unreachable]
> >  1997 |   u32 dsparb, dsparb2, dsparb3;
> >       |       ^~~~~~
> >
> > [1] https://bugs.llvm.org/show_bug.cgi?id=44916
> >
> > Signed-off-by: Kees Cook <keescook@xxxxxxxxxxxx>
> 
> Reviewed-by: Jani Nikula <jani.nikula@xxxxxxxxx>

Thanks!

> 
> If you look at i915/Makefile, you'll see that we don't shy away from
> enabling lots of extra warnings, and we run our CI with -Werror to keep
> it clean. It does not seem like -Wswitch-unreachable does me any good,
> though... is it new?

It's already enabled by default, but the GCC feature that tweaks it
doesn't exist yet. But it points out a problem that exists for Clang
today, but Clang doesn't actually warn on (yet). So this is a fix to
avoid the silent Clang problem and fix future warnings before they
happen.

-Kees

> 
> BR,
> Jani.
> 
> 
> > ---
> >  drivers/gpu/drm/i915/display/intel_display.c |    6 ++++--
> >  drivers/gpu/drm/i915/intel_pm.c              |    4 ++--
> >  2 files changed, 6 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_display.c b/drivers/gpu/drm/i915/display/intel_display.c
> > index 064dd99bbc49..c829cd26f99e 100644
> > --- a/drivers/gpu/drm/i915/display/intel_display.c
> > +++ b/drivers/gpu/drm/i915/display/intel_display.c
> > @@ -12960,14 +12960,15 @@ static bool check_digital_port_conflicts(struct intel_atomic_state *state)
> >  		WARN_ON(!connector_state->crtc);
> >  
> >  		switch (encoder->type) {
> > -			unsigned int port_mask;
> >  		case INTEL_OUTPUT_DDI:
> >  			if (WARN_ON(!HAS_DDI(to_i915(dev))))
> >  				break;
> >  			/* else, fall through */
> >  		case INTEL_OUTPUT_DP:
> >  		case INTEL_OUTPUT_HDMI:
> > -		case INTEL_OUTPUT_EDP:
> > +		case INTEL_OUTPUT_EDP: {
> > +			unsigned int port_mask;
> > +
> >  			port_mask = 1 << encoder->port;
> >  
> >  			/* the same port mustn't appear more than once */
> > @@ -12976,6 +12977,7 @@ static bool check_digital_port_conflicts(struct intel_atomic_state *state)
> >  
> >  			used_ports |= port_mask;
> >  			break;
> > +		}
> >  		case INTEL_OUTPUT_DP_MST:
> >  			used_mst_ports |=
> >  				1 << encoder->port;
> > diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c
> > index bd2d30ecc030..17d8833787c4 100644
> > --- a/drivers/gpu/drm/i915/intel_pm.c
> > +++ b/drivers/gpu/drm/i915/intel_pm.c
> > @@ -469,9 +469,9 @@ static void vlv_get_fifo_size(struct intel_crtc_state *crtc_state)
> >  	struct vlv_fifo_state *fifo_state = &crtc_state->wm.vlv.fifo_state;
> >  	enum pipe pipe = crtc->pipe;
> >  	int sprite0_start, sprite1_start;
> > +	u32 dsparb, dsparb2, dsparb3;
> >  
> >  	switch (pipe) {
> > -		u32 dsparb, dsparb2, dsparb3;
> >  	case PIPE_A:
> >  		dsparb = I915_READ(DSPARB);
> >  		dsparb2 = I915_READ(DSPARB2);
> > @@ -1969,6 +1969,7 @@ static void vlv_atomic_update_fifo(struct intel_atomic_state *state,
> >  	const struct vlv_fifo_state *fifo_state =
> >  		&crtc_state->wm.vlv.fifo_state;
> >  	int sprite0_start, sprite1_start, fifo_size;
> > +	u32 dsparb, dsparb2, dsparb3;
> >  
> >  	if (!crtc_state->fifo_changed)
> >  		return;
> > @@ -1994,7 +1995,6 @@ static void vlv_atomic_update_fifo(struct intel_atomic_state *state,
> >  	spin_lock(&uncore->lock);
> >  
> >  	switch (crtc->pipe) {
> > -		u32 dsparb, dsparb2, dsparb3;
> >  	case PIPE_A:
> >  		dsparb = intel_uncore_read_fw(uncore, DSPARB);
> >  		dsparb2 = intel_uncore_read_fw(uncore, DSPARB2);
> >
> 
> -- 
> Jani Nikula, Intel Open Source Graphics Center

-- 
Kees Cook