drm/i915/tgl: Suspend pre-parser across GTT invalidations

Before we execute a batch, we must first issue any and all TLB
invalidations so that batch picks up the new page table entries.
Tigerlake's preparser is weakening our post-sync CS_STALL inside the
invalidate pipe-control and allowing the loading of the batch buffer
before we have setup its page table (and so it loads the wrong page and
executes indefinitely).

The igt_cs_tlb indicates that this issue can only be observed on rcs,
even though the preparser is common to all engines. Alternatively, we
could do TLB shootdown via mmio on updating the GTT.

By inserting the pre-parser disable inside EMIT_INVALIDATE, we will also
accidentally fixup execution that writes into subsequent batches, such
as gem_exec_whisper and even relocations performed on the GPU. We should
be careful not to allow this disable to become baked into the uABI! The
issue is that if userspace relies on our disabling of the HW
optimisation, when we are ready to enable that optimisation, userspace
will then be broken...

Testcase: igt/i915_selftests/live_gtt/igt_cs_tlb
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=111753
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Acked-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190919151811.9526-1-chris@chris-wilson.co.uk

1 files changed, 73 insertions, 1 deletions

diff --git a/drivers/gpu/drm/i915/gt/intel_lrc.c b/drivers/gpu/drm/i915/gt/intel_lrc.c
index a99166a2d2eb..49ed9230a2cb 100644
--- a/drivers/gpu/drm/i915/gt/intel_lrc.c
+++ b/drivers/gpu/drm/i915/gt/intel_lrc.c
@@ -2807,6 +2807,78 @@ static int gen11_emit_flush_render(struct i915_request *request,
 	return 0;
 }
 
+static u32 preparser_disable(bool state)
+{
+	return MI_ARB_CHECK | 1 << 8 | state;
+}
+
+static int gen12_emit_flush_render(struct i915_request *request,
+				   u32 mode)
+{
+	const u32 scratch_addr =
+		intel_gt_scratch_offset(request->engine->gt,
+					INTEL_GT_SCRATCH_FIELD_RENDER_FLUSH);
+
+	if (mode & EMIT_FLUSH) {
+		u32 flags = 0;
+		u32 *cs;
+
+		flags |= PIPE_CONTROL_TILE_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_DEPTH_CACHE_FLUSH;
+		flags |= PIPE_CONTROL_DC_FLUSH_ENABLE;
+		flags |= PIPE_CONTROL_FLUSH_ENABLE;
+
+		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+		flags |= PIPE_CONTROL_QW_WRITE;
+
+		flags |= PIPE_CONTROL_CS_STALL;
+
+		cs = intel_ring_begin(request, 6);
+		if (IS_ERR(cs))
+			return PTR_ERR(cs);
+
+		cs = gen8_emit_pipe_control(cs, flags, scratch_addr);
+		intel_ring_advance(request, cs);
+	}
+
+	if (mode & EMIT_INVALIDATE) {
+		u32 flags = 0;
+		u32 *cs;
+
+		flags |= PIPE_CONTROL_COMMAND_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_TLB_INVALIDATE;
+		flags |= PIPE_CONTROL_INSTRUCTION_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_TEXTURE_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_VF_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_CONST_CACHE_INVALIDATE;
+		flags |= PIPE_CONTROL_STATE_CACHE_INVALIDATE;
+
+		flags |= PIPE_CONTROL_GLOBAL_GTT_IVB;
+		flags |= PIPE_CONTROL_QW_WRITE;
+
+		flags |= PIPE_CONTROL_CS_STALL;
+
+		cs = intel_ring_begin(request, 8);
+		if (IS_ERR(cs))
+			return PTR_ERR(cs);
+
+		/*
+		 * Prevent the pre-parser from skipping past the TLB
+		 * invalidate and loading a stale page for the batch
+		 * buffer / request payload.
+		 */
+		*cs++ = preparser_disable(true);
+
+		cs = gen8_emit_pipe_control(cs, flags, scratch_addr);
+
+		*cs++ = preparser_disable(false);
+		intel_ring_advance(request, cs);
+	}
+
+	return 0;
+}
+
 /*
  * Reserve space for 2 NOOPs at the end of each request to be
  * used as a workaround for not being allowed to do lite
@@ -3072,7 +3144,7 @@ static void rcs_submission_override(struct intel_engine_cs *engine)
 {
 	switch (INTEL_GEN(engine->i915)) {
 	case 12:
-		engine->emit_flush = gen11_emit_flush_render;
+		engine->emit_flush = gen12_emit_flush_render;
 		engine->emit_fini_breadcrumb = gen12_emit_fini_breadcrumb_rcs;
 		break;
 	case 11: