Intel-XE Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Souza, Jose" <jose.souza@intel.com>
To: "intel-xe@lists.freedesktop.org" <intel-xe@lists.freedesktop.org>,
	"tvrtko.ursulin@igalia.com" <tvrtko.ursulin@igalia.com>
Cc: "kernel-dev@igalia.com" <kernel-dev@igalia.com>
Subject: Re: [PATCH v7 11/24] drm/xe: Flush L3 when flushing render cache
Date: Fri, 27 Jun 2025 18:57:47 +0000	[thread overview]
Message-ID: <3bbea88dc4e95e6e976f0143a38ae40b813a9728.camel@intel.com> (raw)
In-Reply-To: <371c413dbaeb584cf56264bd7d1d44d0ca284ec5.camel@intel.com>

On Fri, 2025-06-27 at 11:23 -0700, José Roberto de Souza wrote:
> On Fri, 2025-06-27 at 14:33 +0100, Tvrtko Ursulin wrote:
> > I915 sets PIPE_CONTROL_FLUSH_L3 (bit 27) when flushing render caches but
> > interesting thing is Tigerlake PRM lists that bit as reserved.
> > 
> > Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
> > ---
> > Is xe missing this? Or has this been wrong for so long in i915? Or is this
> > an undocumented bit?
> > ---
> >  drivers/gpu/drm/xe/instructions/xe_gpu_commands.h |  1 +
> >  drivers/gpu/drm/xe/xe_ring_ops.c                  | 10 ++++++++++
> >  2 files changed, 11 insertions(+)
> > 
> > diff --git a/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h b/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h
> > index 78c0e87dbd37..27892984403c 100644
> > --- a/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h
> > +++ b/drivers/gpu/drm/xe/instructions/xe_gpu_commands.h
> > @@ -47,6 +47,7 @@
> >  
> >  #define   PIPE_CONTROL_COMMAND_CACHE_INVALIDATE		(1<<29)
> >  #define   PIPE_CONTROL_TILE_CACHE_FLUSH			(1<<28)
> > +#define   PIPE_CONTROL_FLUSH_L3                         (1<<27)
> 
> On spec this bit is Protected Memory Disable. I think what you want is bit 30.

That is also wrong on i915 oO.
Will give some testing here and submit a i915 patch.

> 
> >  #define   PIPE_CONTROL_AMFS_FLUSH			(1<<25)
> >  #define   PIPE_CONTROL_GLOBAL_GTT_IVB			(1<<24)
> >  #define   PIPE_CONTROL_LRI_POST_SYNC			BIT(23)
> > diff --git a/drivers/gpu/drm/xe/xe_ring_ops.c b/drivers/gpu/drm/xe/xe_ring_ops.c
> > index a1289f086191..8f655b6fe913 100644
> > --- a/drivers/gpu/drm/xe/xe_ring_ops.c
> > +++ b/drivers/gpu/drm/xe/xe_ring_ops.c
> > @@ -197,6 +197,16 @@ static int emit_render_cache_flush(struct xe_sched_job *job, u32 *dw, int i)
> >  	if (XE_WA(gt, 1409600907))
> >  		flags |= PIPE_CONTROL_DEPTH_STALL;
> >  
> > +	/*
> > +	 * L3 fabric flush is needed for AUX CCS invalidation
> > +	 * which happens as part of pipe-control so we can
> > +	 * ignore PIPE_CONTROL_FLUSH_L3. Also PIPE_CONTROL_FLUSH_L3
> > +	 * deals with Protected Memory which is not needed for
> > +	 * AUX CCS invalidation and lead to unwanted side effects.
> > +	 */
> > +	if (GRAPHICS_VERx100(xe) < 1270)
> > +		flags |= PIPE_CONTROL_FLUSH_L3;
> > +
> >  	if (lacks_render)
> >  		flags &= ~PIPE_CONTROL_3D_ARCH_FLAGS;
> >  	else if (job->q->class == XE_ENGINE_CLASS_COMPUTE)

  reply	other threads:[~2025-06-27 18:57 UTC|newest]

Thread overview: 29+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-06-27 13:33 [PATCH v7 00/24] AuxCCS handling and render compression modifiers Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 01/24] drm/xe: Consolidate LRC offset calculations Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 02/24] drm/xe: Generalize wa bb emission code Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 03/24] drm/xe: Rename utilisation workaround emission function Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 04/24] drm/xe: Return number of written dwords from workaround batch buffer emission Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 05/24] drm/xe: Allow specifying number of extra dwords at the end of wa bb emission Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 06/24] drm/xe: Add plumbing for indirect context workarounds Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 07/24] drm/xe/xelp: Implement Wa_16010904313 Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 08/24] drm/xe/xelp: Add Wa_18022495364 Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 09/24] drm/xe: Use emit_flush_imm_ggtt helper instead of open coding Tvrtko Ursulin
2025-06-27 21:57   ` Matthew Brost
2025-06-27 13:33 ` [PATCH v7 10/24] drm/xe/xelpg: Flush CCS when flushing caches Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 11/24] drm/xe: Flush L3 when flushing render cache Tvrtko Ursulin
2025-06-27 18:23   ` Souza, Jose
2025-06-27 18:57     ` Souza, Jose [this message]
2025-06-30 12:43       ` Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 12/24] drm/xe/xelp: Quiesce memory traffic before invalidating auxccs Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 13/24] drm/xe/xelp: Support auxccs invalidation on blitter Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 14/24] drm/xe/xelp: Use MI_FLUSH_DW_CCS on auxccs platforms Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 15/24] drm/xe/xelp: Wait for AuxCCS invalidation to complete Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 16/24] drm/xe/xelp: Add AuxCCS invalidation to the buffer migration path Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 17/24] drm/xe: Export xe_emit_aux_table_inv Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 18/24] drm/xe/xelp: Add AuxCCS invalidation to the indirect context workarounds Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 19/24] drm/xe: Use fb cached min alignment Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 20/24] drm/xe: Flush GGTT writes after populating DPT Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 21/24] drm/xe: Handle DPT in system memory Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 22/24] drm/xe: Force flush system memory AuxCCS framebuffers before scan out Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 23/24] drm/xe/display: Add support for AuxCCS Tvrtko Ursulin
2025-06-27 13:33 ` [PATCH v7 24/24] drm/i915/display: Expose AuxCCS frame buffer modifiers for Xe Tvrtko Ursulin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3bbea88dc4e95e6e976f0143a38ae40b813a9728.camel@intel.com \
    --to=jose.souza@intel.com \
    --cc=intel-xe@lists.freedesktop.org \
    --cc=kernel-dev@igalia.com \
    --cc=tvrtko.ursulin@igalia.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox