[Intel-gfx] [RFC PATCH 39/97] drm/i915/guc: Increase size of CTB buffers

[Matthew Brost]

On Mon, May 24, 2021 at 03:43:11PM +0200, Michal Wajdeczko wrote:
> 
> 
> On 06.05.2021 21:13, Matthew Brost wrote:
> > With the introduction of non-blocking CTBs more than one CTB can be in
> > flight at a time. Increasing the size of the CTBs should reduce how
> > often software hits the case where no space is available in the CTB
> > buffer.
> > 
> > Cc: John Harrison <john.c.harrison at intel.com>
> > Signed-off-by: Matthew Brost <matthew.brost at intel.com>
> > ---
> >  drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c | 11 ++++++++---
> >  1 file changed, 8 insertions(+), 3 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> > index 77dfbc94dcc3..d6895d29ed2d 100644
> > --- a/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> > +++ b/drivers/gpu/drm/i915/gt/uc/intel_guc_ct.c
> > @@ -63,11 +63,16 @@ static inline struct drm_device *ct_to_drm(struct intel_guc_ct *ct)
> >   *      +--------+-----------------------------------------------+------+
> >   *
> >   * Size of each `CT Buffer`_ must be multiple of 4K.
> > - * As we don't expect too many messages, for now use minimum sizes.
> > + * We don't expect too many messages in flight at any time, unless we are
> > + * using the GuC submission. In that case each request requires a minimum
> > + * 16 bytes which gives us a maximum 256 queue'd requests. Hopefully this
> 
> nit: all our CTB calculations are in dwords now, not bytes
> 

I can change the wording to DW sizes.

> > + * enough space to avoid backpressure on the driver. We increase the size
> > + * of the receive buffer (relative to the send) to ensure a G2H response
> > + * CTB has a landing spot.
> 
> hmm, but we are not checking G2H CTB yet
> will start doing it around patch 54/97
> so maybe this other patch should be introduced earlier ?
>

Yes, that patch is going to be pulled down to an earlier spot in the
series.
 
> >   */
> >  #define CTB_DESC_SIZE		ALIGN(sizeof(struct guc_ct_buffer_desc), SZ_2K)
> >  #define CTB_H2G_BUFFER_SIZE	(SZ_4K)
> > -#define CTB_G2H_BUFFER_SIZE	(SZ_4K)
> > +#define CTB_G2H_BUFFER_SIZE	(4 * CTB_H2G_BUFFER_SIZE)
> 
> in theory, we (host) should be faster than GuC, so G2H CTB shall be
> almost always empty, if this is not a case, maybe we should start
> monitoring what is happening and report some warnings if G2H is half full ?
>

Certainly some IGTs put some more pressure on the G2H channel than the
H2G channel at least I think. This is something we can tune over time
after this lands upstream. IMO a message at this point is overkill.

Matt
 
> >  
> >  #define MAX_US_STALL_CTB	1000000
> >  
> > @@ -753,7 +758,7 @@ static int ct_read(struct intel_guc_ct *ct, struct ct_incoming_msg **msg)
> >  	/* beware of buffer wrap case */
> >  	if (unlikely(available < 0))
> >  		available += size;
> > -	CT_DEBUG(ct, "available %d (%u:%u)\n", available, head, tail);
> > +	CT_DEBUG(ct, "available %d (%u:%u:%u)\n", available, head, tail, size);
> >  	GEM_BUG_ON(available < 0);
> >  
> >  	header = cmds[head];
> >