Re: [Intel-gfx] [Intel-xe] [PATCH v4] drm/i915: handle uncore spinlock when not available

[Rodrigo Vivi <rodrigo.vivi@intel.com>]

On Wed, Nov 29, 2023 at 03:24:33PM -0500, Coelho, Luciano wrote:
> On Wed, 2023-11-29 at 13:01 -0500, Rodrigo Vivi wrote:
> > On Wed, Nov 29, 2023 at 11:17:28AM +0200, Luca Coelho wrote:
> > > The uncore code may not always be available (e.g. when we build the
> > > display code with Xe), so we can't always rely on having the uncore's
> > > spinlock.
> > > 
> > > To handle this, split the spin_lock/unlock_irqsave/restore() into
> > > spin_lock/unlock() followed by a call to local_irq_save/restore() and
> > > create wrapper functions for locking and unlocking the uncore's
> > > spinlock.  In these functions, we have a condition check and only
> > > actually try to lock/unlock the spinlock when I915 is defined, and
> > > thus uncore is available.
> > > 
> > > This keeps the ifdefs contained in these new functions and all such
> > > logic inside the display code.
> > > 
> > > Cc: Tvrtko Ursulin <tvrto.ursulin@intel.com>
> > > Cc: Jani Nikula <jani.nikula@intel.com>
> > > Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
> > > Cc: Ville Syrj?l? <ville.syrjala@linux.intel.com>
> > > Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
> > > ---
> > > 
> > > In v2:
> > > 
> > >    * Renamed uncore_spin_*() to intel_spin_*()
> > >    * Corrected the order: save, lock, unlock, restore
> > > 
> > > In v3:
> > > 
> > >    * Undid the change to pass drm_i915_private instead of the lock
> > >      itself, since we would have to include i915_drv.h and that pulls
> > >      in a truckload of other includes.
> > > 
> > > In v4:
> > > 
> > >    * After a brief attempt to replace this with a different patch,
> > >      we're back to this one;
> > >    * Pass drm_i195_private again, and move the functions to
> > >      intel_vblank.c, so we don't need to include i915_drv.h in a
> > >      header file and it's already included in intel_vblank.c;
> > > 
> > >  drivers/gpu/drm/i915/display/intel_display.h |  1 +
> > >  drivers/gpu/drm/i915/display/intel_vblank.c  | 45 +++++++++++++++-----
> > >  2 files changed, 36 insertions(+), 10 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i915/display/intel_display.h b/drivers/gpu/drm/i915/display/intel_display.h
> > > index 8548f49e3972..5ff299bc4b87 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_display.h
> > > +++ b/drivers/gpu/drm/i915/display/intel_display.h
> > > @@ -29,6 +29,7 @@
> > >  
> > >  #include "i915_reg_defs.h"
> > >  #include "intel_display_limits.h"
> > > +#include "i915_drv.h"
> > 
> > please move this include to intel_vblank.c
> 
> Oops, this is a leftover of some tests I was making to see just how
> much worse things would get by adding this here.
> 
> Actually, why don't we move the drm_i915_private structure (and maybe
> others?) to a lighter header file than i915_drv.h? IMHO it's really
> annoying to have the forward declarations for it in many places just
> because we don't want to include the actual header.  When I want to
> find its global definition, cscope always returns tens of results
> because of the forward declarations... This is obviously orthogonal to
> the current patch.

yeah, I know. It is really inconvenient sometimes. I got used to run
cscope and then search for "struct something {" to find the right place.

But this inconvenience is actually smaller when compared to the compilation
time whenever a header gets modified and included by other headers. If I
remember correctly Jani did the initial assessment of compilation times
and started to move headers including out of other headers. He might
have more details/data on his findings.

> 
> 
> > >  enum drm_scaling_filter;
> > >  struct dpll;
> > > diff --git a/drivers/gpu/drm/i915/display/intel_vblank.c b/drivers/gpu/drm/i915/display/intel_vblank.c
> > > index 2cec2abf9746..d9625db82681 100644
> > > --- a/drivers/gpu/drm/i915/display/intel_vblank.c
> > > +++ b/drivers/gpu/drm/i915/display/intel_vblank.c
> > > @@ -265,6 +265,26 @@ int intel_crtc_scanline_to_hw(struct intel_crtc *crtc, int scanline)
> > >  	return (scanline + vtotal - crtc->scanline_offset) % vtotal;
> > >  }
> > >  
> > > +/*
> > > + * The uncore version of the spin lock functions is used to decide
> > > + * whether we need to lock the uncore lock or not.  This is only
> > > + * needed in i915, not in Xe.  Keep the decision-making centralized
> > > + * here.
> > 
> > maybe we could add brief mention that it is only needed because old hardware
> > that is not supported by Xe.
> 
> Good idea, I'll add it.
> 
> > 
> > > + */
> > > +static inline void intel_vblank_section_enter(struct drm_i915_private *i915)
> > 
> > let's avoid inline here.
> 
> Okay, I'll remove it.
> 
> 
> > > +{
> > > +#ifdef I915
> > > +	spin_lock(&i915->uncore.lock);
> > > +#endif
> > > +}
> > > +
> > > +static inline void intel_vblank_section_exit(struct drm_i915_private *i915)
> > 
> > and here
> 
> Okay.
> 
> 
> > With these changes:
> > 
> > Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
> 
> Thanks for the review, Rodrigo!
> 
> --
> Cheers,
> Luca.