From: Chris Wilson chris@chris-wilson.co.uk
commit 08f56f8f3799b2ed1c5ac7eed6d86a4926289655 upstream.
We still need to wait for the initial OA configuration to happen before we enable OA report writes to the OA buffer.
Reported-by: Lionel Landwerlin lionel.g.landwerlin@intel.com Fixes: 15d0ace1f876 ("drm/i915/perf: execute OA configuration from command stream") Closes: https://gitlab.freedesktop.org/drm/intel/issues/1356 Testcase: igt/perf/stream-open-close Signed-off-by: Chris Wilson chris@chris-wilson.co.uk Cc: Lionel Landwerlin lionel.g.landwerlin@intel.com Reviewed-by: Lionel Landwerlin lionel.g.landwerlin@intel.com Link: https://patchwork.freedesktop.org/patch/msgid/20200302085812.4172450-7-chris... (cherry picked from commit 4b4e973d5eb89244b67d3223b60f752d0479f253) Signed-off-by: Jani Nikula jani.nikula@intel.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org
--- drivers/gpu/drm/i915/i915_perf.c | 58 +++++++++++++++++++++++---------- drivers/gpu/drm/i915/i915_perf_types.h | 3 + 2 files changed, 43 insertions(+), 18 deletions(-)
--- a/drivers/gpu/drm/i915/i915_perf.c +++ b/drivers/gpu/drm/i915/i915_perf.c @@ -1950,9 +1950,10 @@ out: return i915_vma_get(oa_bo->vma); }
-static int emit_oa_config(struct i915_perf_stream *stream, - struct i915_oa_config *oa_config, - struct intel_context *ce) +static struct i915_request * +emit_oa_config(struct i915_perf_stream *stream, + struct i915_oa_config *oa_config, + struct intel_context *ce) { struct i915_request *rq; struct i915_vma *vma; @@ -1960,7 +1961,7 @@ static int emit_oa_config(struct i915_pe
vma = get_oa_vma(stream, oa_config); if (IS_ERR(vma)) - return PTR_ERR(vma); + return ERR_CAST(vma);
err = i915_vma_pin(vma, 0, 0, PIN_GLOBAL | PIN_HIGH); if (err) @@ -1983,13 +1984,17 @@ static int emit_oa_config(struct i915_pe err = rq->engine->emit_bb_start(rq, vma->node.start, 0, I915_DISPATCH_SECURE); + if (err) + goto err_add_request; + + i915_request_get(rq); err_add_request: i915_request_add(rq); err_vma_unpin: i915_vma_unpin(vma); err_vma_put: i915_vma_put(vma); - return err; + return err ? ERR_PTR(err) : rq; }
static struct intel_context *oa_context(struct i915_perf_stream *stream) @@ -1997,7 +2002,8 @@ static struct intel_context *oa_context( return stream->pinned_ctx ?: stream->engine->kernel_context; }
-static int hsw_enable_metric_set(struct i915_perf_stream *stream) +static struct i915_request * +hsw_enable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore;
@@ -2408,7 +2414,8 @@ static int lrc_configure_all_contexts(st return oa_configure_all_contexts(stream, regs, ARRAY_SIZE(regs)); }
-static int gen8_enable_metric_set(struct i915_perf_stream *stream) +static struct i915_request * +gen8_enable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; struct i915_oa_config *oa_config = stream->oa_config; @@ -2450,12 +2457,13 @@ static int gen8_enable_metric_set(struct */ ret = lrc_configure_all_contexts(stream, oa_config); if (ret) - return ret; + return ERR_PTR(ret);
return emit_oa_config(stream, oa_config, oa_context(stream)); }
-static int gen12_enable_metric_set(struct i915_perf_stream *stream) +static struct i915_request * +gen12_enable_metric_set(struct i915_perf_stream *stream) { struct intel_uncore *uncore = stream->uncore; struct i915_oa_config *oa_config = stream->oa_config; @@ -2488,7 +2496,7 @@ static int gen12_enable_metric_set(struc */ ret = gen12_configure_all_contexts(stream, oa_config); if (ret) - return ret; + return ERR_PTR(ret);
/* * For Gen12, performance counters are context @@ -2498,7 +2506,7 @@ static int gen12_enable_metric_set(struc if (stream->ctx) { ret = gen12_configure_oar_context(stream, true); if (ret) - return ret; + return ERR_PTR(ret); }
return emit_oa_config(stream, oa_config, oa_context(stream)); @@ -2693,6 +2701,20 @@ static const struct i915_perf_stream_ops .read = i915_oa_read, };
+static int i915_perf_stream_enable_sync(struct i915_perf_stream *stream) +{ + struct i915_request *rq; + + rq = stream->perf->ops.enable_metric_set(stream); + if (IS_ERR(rq)) + return PTR_ERR(rq); + + i915_request_wait(rq, 0, MAX_SCHEDULE_TIMEOUT); + i915_request_put(rq); + + return 0; +} + /** * i915_oa_stream_init - validate combined props for OA stream and init * @stream: An i915 perf stream @@ -2826,7 +2848,7 @@ static int i915_oa_stream_init(struct i9 stream->ops = &i915_oa_stream_ops; perf->exclusive_stream = stream;
- ret = perf->ops.enable_metric_set(stream); + ret = i915_perf_stream_enable_sync(stream); if (ret) { DRM_DEBUG("Unable to enable metric set\n"); goto err_enable; @@ -3144,7 +3166,7 @@ static long i915_perf_config_locked(stru return -EINVAL;
if (config != stream->oa_config) { - int err; + struct i915_request *rq;
/* * If OA is bound to a specific context, emit the @@ -3155,11 +3177,13 @@ static long i915_perf_config_locked(stru * When set globally, we use a low priority kernel context, * so it will effectively take effect when idle. */ - err = emit_oa_config(stream, config, oa_context(stream)); - if (err == 0) + rq = emit_oa_config(stream, config, oa_context(stream)); + if (!IS_ERR(rq)) { config = xchg(&stream->oa_config, config); - else - ret = err; + i915_request_put(rq); + } else { + ret = PTR_ERR(rq); + } }
i915_oa_config_put(config); --- a/drivers/gpu/drm/i915/i915_perf_types.h +++ b/drivers/gpu/drm/i915/i915_perf_types.h @@ -339,7 +339,8 @@ struct i915_oa_ops { * counter reports being sampled. May apply system constraints such as * disabling EU clock gating as required. */ - int (*enable_metric_set)(struct i915_perf_stream *stream); + struct i915_request * + (*enable_metric_set)(struct i915_perf_stream *stream);
/** * @disable_metric_set: Remove system constraints associated with using