drm/i915: Flush TLBs before releasing backing store

We need to flush TLBs before releasing backing store otherwise userspace
is able to encounter stale entries if a) it is not declaring access to
certain buffers and b) it races with the backing store release from a
such undeclared execution already executing on the GPU in parallel.

The approach taken is to mark any buffer objects which were ever bound
to the GPU and to trigger a serialized TLB flush when their backing
store is released.

Alternatively the flushing could be done on VMA unbind, at which point
we would be able to ascertain whether there is potential a parallel GPU
execution (which could race), but essentially it boils down to paying
the cost of TLB flushes potentially needlessly at VMA unbind time (when
the backing store is not known to be going away so not needed for
safety), versus potentially needlessly at backing store relase time
(since we at that point cannot tell whether there is anything executing
on the GPU which uses that object).

Thereforce simplicity of implementation has been chosen for now with
scope to benchmark and refine later as required.

Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reported-by: Sushma Venkatesh Reddy <sushma.venkatesh.reddy@intel.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: Dave Airlie <airlied@redhat.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Jon Bloomfield <jon.bloomfield@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: stable@vger.kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

9 files changed, 161 insertions, 4 deletions

diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
index 4b4829eb16c2..0dd107dcecc2 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
+++ b/drivers/gpu/drm/i915/gem/i915_gem_object_types.h
@@ -311,6 +311,7 @@ struct drm_i915_gem_object {
 #define I915_BO_READONLY          BIT(6)
 #define I915_TILING_QUIRK_BIT     7 /* unknown swizzling; do not release! */
 #define I915_BO_PROTECTED         BIT(8)
+#define I915_BO_WAS_BOUND_BIT     9
 	/**
 	 * @mem_flags - Mutable placement-related flags
 	 *
diff --git a/drivers/gpu/drm/i915/gem/i915_gem_pages.c b/drivers/gpu/drm/i915/gem/i915_gem_pages.c
index 9f429ed6e78a..a50f884973bc 100644
--- a/drivers/gpu/drm/i915/gem/i915_gem_pages.c
+++ b/drivers/gpu/drm/i915/gem/i915_gem_pages.c
@@ -10,6 +10,8 @@
 #include "i915_gem_lmem.h"
 #include "i915_gem_mman.h"
 
+#include "gt/intel_gt.h"
+
 void __i915_gem_object_set_pages(struct drm_i915_gem_object *obj,
 				 struct sg_table *pages,
 				 unsigned int sg_page_sizes)
@@ -221,6 +223,14 @@ __i915_gem_object_unset_pages(struct drm_i915_gem_object *obj)
 	__i915_gem_object_reset_page_iter(obj);
 	obj->mm.page_sizes.phys = obj->mm.page_sizes.sg = 0;
 
+	if (test_and_clear_bit(I915_BO_WAS_BOUND_BIT, &obj->flags)) {
+		struct drm_i915_private *i915 = to_i915(obj->base.dev);
+		intel_wakeref_t wakeref;
+
+		with_intel_runtime_pm_if_active(&i915->runtime_pm, wakeref)
+			intel_gt_invalidate_tlbs(to_gt(i915));
+	}
+
 	return pages;
 }
 
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.c b/drivers/gpu/drm/i915/gt/intel_gt.c
index f98f0fb21efb..35d0fcd3a86c 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.c
+++ b/drivers/gpu/drm/i915/gt/intel_gt.c
@@ -29,6 +29,8 @@ void __intel_gt_init_early(struct intel_gt *gt, struct drm_i915_private *i915)
 {
 	spin_lock_init(&gt->irq_lock);
 
+	mutex_init(&gt->tlb_invalidate_lock);
+
 	INIT_LIST_HEAD(&gt->closed_vma);
 	spin_lock_init(&gt->closed_lock);
 
@@ -912,3 +914,109 @@ void intel_gt_info_print(const struct intel_gt_info *info,
 
 	intel_sseu_dump(&info->sseu, p);
 }
+
+struct reg_and_bit {
+	i915_reg_t reg;
+	u32 bit;
+};
+
+static struct reg_and_bit
+get_reg_and_bit(const struct intel_engine_cs *engine, const bool gen8,
+		const i915_reg_t *regs, const unsigned int num)
+{
+	const unsigned int class = engine->class;
+	struct reg_and_bit rb = { };
+
+	if (drm_WARN_ON_ONCE(&engine->i915->drm,
+			     class >= num || !regs[class].reg))
+		return rb;
+
+	rb.reg = regs[class];
+	if (gen8 && class == VIDEO_DECODE_CLASS)
+		rb.reg.reg += 4 * engine->instance; /* GEN8_M2TCR */
+	else
+		rb.bit = engine->instance;
+
+	rb.bit = BIT(rb.bit);
+
+	return rb;
+}
+
+void intel_gt_invalidate_tlbs(struct intel_gt *gt)
+{
+	static const i915_reg_t gen8_regs[] = {
+		[RENDER_CLASS]			= GEN8_RTCR,
+		[VIDEO_DECODE_CLASS]		= GEN8_M1TCR, /* , GEN8_M2TCR */
+		[VIDEO_ENHANCEMENT_CLASS]	= GEN8_VTCR,
+		[COPY_ENGINE_CLASS]		= GEN8_BTCR,
+	};
+	static const i915_reg_t gen12_regs[] = {
+		[RENDER_CLASS]			= GEN12_GFX_TLB_INV_CR,
+		[VIDEO_DECODE_CLASS]		= GEN12_VD_TLB_INV_CR,
+		[VIDEO_ENHANCEMENT_CLASS]	= GEN12_VE_TLB_INV_CR,
+		[COPY_ENGINE_CLASS]		= GEN12_BLT_TLB_INV_CR,
+	};
+	struct drm_i915_private *i915 = gt->i915;
+	struct intel_uncore *uncore = gt->uncore;
+	struct intel_engine_cs *engine;
+	enum intel_engine_id id;
+	const i915_reg_t *regs;
+	unsigned int num = 0;
+
+	if (I915_SELFTEST_ONLY(gt->awake == -ENODEV))
+		return;
+
+	if (GRAPHICS_VER(i915) == 12) {
+		regs = gen12_regs;
+		num = ARRAY_SIZE(gen12_regs);
+	} else if (GRAPHICS_VER(i915) >= 8 && GRAPHICS_VER(i915) <= 11) {
+		regs = gen8_regs;
+		num = ARRAY_SIZE(gen8_regs);
+	} else if (GRAPHICS_VER(i915) < 8) {
+		return;
+	}
+
+	if (drm_WARN_ONCE(&i915->drm, !num,
+			  "Platform does not implement TLB invalidation!"))
+		return;
+
+	GEM_TRACE("\n");
+
+	assert_rpm_wakelock_held(&i915->runtime_pm);
+
+	mutex_lock(&gt->tlb_invalidate_lock);
+	intel_uncore_forcewake_get(uncore, FORCEWAKE_ALL);
+
+	for_each_engine(engine, gt, id) {
+		/*
+		 * HW architecture suggest typical invalidation time at 40us,
+		 * with pessimistic cases up to 100us and a recommendation to
+		 * cap at 1ms. We go a bit higher just in case.
+		 */
+		const unsigned int timeout_us = 100;
+		const unsigned int timeout_ms = 4;
+		struct reg_and_bit rb;
+
+		rb = get_reg_and_bit(engine, regs == gen8_regs, regs, num);
+		if (!i915_mmio_reg_offset(rb.reg))
+			continue;
+
+		intel_uncore_write_fw(uncore, rb.reg, rb.bit);
+		if (__intel_wait_for_register_fw(uncore,
+						 rb.reg, rb.bit, 0,
+						 timeout_us, timeout_ms,
+						 NULL))
+			drm_err_ratelimited(&gt->i915->drm,
+					    "%s TLB invalidation did not complete in %ums!\n",
+					    engine->name, timeout_ms);
+	}
+
+	/*
+	 * Use delayed put since a) we mostly expect a flurry of TLB
+	 * invalidations so it is good to avoid paying the forcewake cost and
+	 * b) it works around a bug in Icelake which cannot cope with too rapid
+	 * transitions.
+	 */
+	intel_uncore_forcewake_put_delayed(uncore, FORCEWAKE_ALL);
+	mutex_unlock(&gt->tlb_invalidate_lock);
+}
diff --git a/drivers/gpu/drm/i915/gt/intel_gt.h b/drivers/gpu/drm/i915/gt/intel_gt.h
index 3ace129eb2af..a913fb6ffec3 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt.h
@@ -91,4 +91,6 @@ void intel_gt_info_print(const struct intel_gt_info *info,
 
 void intel_gt_watchdog_work(struct work_struct *work);
 
+void intel_gt_invalidate_tlbs(struct intel_gt *gt);
+
 #endif /* __INTEL_GT_H__ */
diff --git a/drivers/gpu/drm/i915/gt/intel_gt_types.h b/drivers/gpu/drm/i915/gt/intel_gt_types.h
index 14216cc471b1..f20687796490 100644
--- a/drivers/gpu/drm/i915/gt/intel_gt_types.h
+++ b/drivers/gpu/drm/i915/gt/intel_gt_types.h
@@ -73,6 +73,8 @@ struct intel_gt {
 
 	struct intel_uc uc;
 
+	struct mutex tlb_invalidate_lock;
+
 	struct i915_wa_list wa_list;
 
 	struct intel_gt_timelines {
diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h
index 971d601fe751..c32420cb8ed5 100644
--- a/drivers/gpu/drm/i915/i915_reg.h
+++ b/drivers/gpu/drm/i915/i915_reg.h
@@ -2721,6 +2721,12 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define   GAMT_CHKN_DISABLE_DYNAMIC_CREDIT_SHARING	(1 << 28)
 #define   GAMT_CHKN_DISABLE_I2M_CYCLE_ON_WR_PORT	(1 << 24)
 
+#define GEN8_RTCR	_MMIO(0x4260)
+#define GEN8_M1TCR	_MMIO(0x4264)
+#define GEN8_M2TCR	_MMIO(0x4268)
+#define GEN8_BTCR	_MMIO(0x426c)
+#define GEN8_VTCR	_MMIO(0x4270)
+
 #if 0
 #define PRB0_TAIL	_MMIO(0x2030)
 #define PRB0_HEAD	_MMIO(0x2034)
@@ -2819,6 +2825,11 @@ static inline bool i915_mmio_reg_valid(i915_reg_t reg)
 #define   FAULT_VA_HIGH_BITS		(0xf << 0)
 #define   FAULT_GTT_SEL			(1 << 4)
 
+#define GEN12_GFX_TLB_INV_CR	_MMIO(0xced8)
+#define GEN12_VD_TLB_INV_CR	_MMIO(0xcedc)
+#define GEN12_VE_TLB_INV_CR	_MMIO(0xcee0)
+#define GEN12_BLT_TLB_INV_CR	_MMIO(0xcee4)
+
 #define GEN12_AUX_ERR_DBG		_MMIO(0x43f4)
 
 #define FPGA_DBG		_MMIO(0x42300)
diff --git a/drivers/gpu/drm/i915/i915_vma.c b/drivers/gpu/drm/i915/i915_vma.c
index 29a858c53bdd..c0d6d5526abe 100644
--- a/drivers/gpu/drm/i915/i915_vma.c
+++ b/drivers/gpu/drm/i915/i915_vma.c
@@ -457,6 +457,9 @@ int i915_vma_bind(struct i915_vma *vma,
 		vma->ops->bind_vma(vma->vm, NULL, vma, cache_level, bind_flags);
 	}
 
+	if (vma->obj)
+		set_bit(I915_BO_WAS_BOUND_BIT, &vma->obj->flags);
+
 	atomic_or(bind_flags, &vma->flags);
 	return 0;
 }
diff --git a/drivers/gpu/drm/i915/intel_uncore.c b/drivers/gpu/drm/i915/intel_uncore.c
index fc25ebf1a593..778da3179b3c 100644
--- a/drivers/gpu/drm/i915/intel_uncore.c
+++ b/drivers/gpu/drm/i915/intel_uncore.c
@@ -724,7 +724,8 @@ void intel_uncore_forcewake_get__locked(struct intel_uncore *uncore,
 }
 
 static void __intel_uncore_forcewake_put(struct intel_uncore *uncore,
-					 enum forcewake_domains fw_domains)
+					 enum forcewake_domains fw_domains,
+					 bool delayed)
 {
 	struct intel_uncore_forcewake_domain *domain;
 	unsigned int tmp;
@@ -739,7 +740,11 @@ static void __intel_uncore_forcewake_put(struct intel_uncore *uncore,
 			continue;
 		}
 
-		fw_domains_put(uncore, domain->mask);
+		if (delayed &&
+		    !(domain->uncore->fw_domains_timer & domain->mask))
+			fw_domain_arm_timer(domain);
+		else
+			fw_domains_put(uncore, domain->mask);
 	}
 }
 
@@ -760,7 +765,20 @@ void intel_uncore_forcewake_put(struct intel_uncore *uncore,
 		return;
 
 	spin_lock_irqsave(&uncore->lock, irqflags);
-	__intel_uncore_forcewake_put(uncore, fw_domains);
+	__intel_uncore_forcewake_put(uncore, fw_domains, false);
+	spin_unlock_irqrestore(&uncore->lock, irqflags);
+}
+
+void intel_uncore_forcewake_put_delayed(struct intel_uncore *uncore,
+					enum forcewake_domains fw_domains)
+{
+	unsigned long irqflags;
+
+	if (!uncore->fw_get_funcs)
+		return;
+
+	spin_lock_irqsave(&uncore->lock, irqflags);
+	__intel_uncore_forcewake_put(uncore, fw_domains, true);
 	spin_unlock_irqrestore(&uncore->lock, irqflags);
 }
 
@@ -802,7 +820,7 @@ void intel_uncore_forcewake_put__locked(struct intel_uncore *uncore,
 	if (!uncore->fw_get_funcs)
 		return;
 
-	__intel_uncore_forcewake_put(uncore, fw_domains);
+	__intel_uncore_forcewake_put(uncore, fw_domains, false);
 }
 
 void assert_forcewakes_inactive(struct intel_uncore *uncore)
diff --git a/drivers/gpu/drm/i915/intel_uncore.h b/drivers/gpu/drm/i915/intel_uncore.h
index 210fe2a71612..2a15b2b2e2fc 100644
--- a/drivers/gpu/drm/i915/intel_uncore.h
+++ b/drivers/gpu/drm/i915/intel_uncore.h
@@ -246,6 +246,8 @@ void intel_uncore_forcewake_get(struct intel_uncore *uncore,
 				enum forcewake_domains domains);
 void intel_uncore_forcewake_put(struct intel_uncore *uncore,
 				enum forcewake_domains domains);
+void intel_uncore_forcewake_put_delayed(struct intel_uncore *uncore,
+					enum forcewake_domains domains);
 void intel_uncore_forcewake_flush(struct intel_uncore *uncore,
 				  enum forcewake_domains fw_domains);