drm/i915: Refactor confusing __intel_gt_reset()
Subsystems (2)
PATCH 1/3 drm/i915: Refactor confusing __intel_gt_reset()
__intel_gt_reset() is really for resetting engines though the name might suggest something else. So add two helper functions to remove confusions with no functional changes.
--- drivers/gpu/drm/i915/gt/intel_engine_cs.c | 2 +- .../drm/i915/gt/intel_execlists_submission.c | 2 +- drivers/gpu/drm/i915/gt/intel_gt.c | 2 +- drivers/gpu/drm/i915/gt/intel_gt_pm.c | 2 +- drivers/gpu/drm/i915/gt/intel_reset.c | 43 ++++++++++++++----- drivers/gpu/drm/i915/gt/intel_reset.h | 3 +- drivers/gpu/drm/i915/gt/selftest_reset.c | 2 +- drivers/gpu/drm/i915/i915_driver.c | 2 +- 8 files changed, 41 insertions(+), 17 deletions(-)
drivers/gpu/drm/i915/gt/intel_engine_cs.c +1 -1
@@ -678,7 +678,7 @@ void intel_engines_release(struct intel_gt *gt)*/ GEM_BUG_ON (intel_gt_pm_is_awake (gt));if (!INTEL_INFO (gt->i915)->gpu_reset_clobbers_display) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt);/* Decouple the backend; but keep the layout for late GPU resets */ for_each_engine (engine, gt, id) {
drivers/gpu/drm/i915/gt/intel_execlists_submission.c +1 -1
@@ -2898,7 +2898,7 @@ static void enable_error_interrupt(struct intel_engine_cs *engine)drm_err (&engine->i915->drm,"engine '%s' resumed still in error: %08x\n" , engine->name, status); -__intel_gt_reset (engine->gt, engine->mask); +intel_gt_reset_engine (engine); }/*
drivers/gpu/drm/i915/gt/intel_gt.c +1 -1
@@ -832,7 +832,7 @@ void intel_gt_driver_unregister(struct intel_gt *gt)/* Scrub all HW state upon release */ with_intel_runtime_pm(gt->uncore->rpm, wakeref) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt); }void intel_gt_driver_release (struct intel_gt *gt)
drivers/gpu/drm/i915/gt/intel_gt_pm.c +1 -1
@@ -159,7 +159,7 @@ static bool reset_engines(struct intel_gt *gt)if (INTEL_INFO (gt->i915)->gpu_reset_clobbers_display)return false; -return __intel_gt_reset (gt,ALL_ENGINES ) ==0 ; +return intel_gt_reset_all_engines (gt) ==0 ; }static void gt_sanitize (struct intel_gt *gt,bool force)
drivers/gpu/drm/i915/gt/intel_reset.c +33 -10
81 lines not on this page · raw message
drivers/gpu/drm/i915/gt/intel_reset.h +2 -1
@@ -54,7 +54,8 @@ int intel_gt_terminally_wedged(struct intel_gt *gt);void intel_gt_set_wedged_on_init (struct intel_gt *gt);void intel_gt_set_wedged_on_fini (struct intel_gt *gt); -int __intel_gt_reset (struct intel_gt *gt,intel_engine_mask_t engine_mask); +int intel_gt_reset_engine (struct intel_engine_cs *engine); +int intel_gt_reset_all_engines (struct intel_gt *gt);int intel_reset_guc (struct intel_gt *gt);
drivers/gpu/drm/i915/gt/selftest_reset.c +1 -1
@@ -281,7 +281,7 @@ static int igt_atomic_reset(void *arg) awake =reset_prepare (gt); p->critical_section_begin (); - err =__intel_gt_reset (gt,ALL_ENGINES ); + err =intel_gt_reset_all_engines (gt); p->critical_section_end ();reset_finish (gt, awake);
drivers/gpu/drm/i915/i915_driver.c +1 -1
@@ -202,7 +202,7 @@ static void sanitize_gpu(struct drm_i915_private *i915)unsigned int i;for_each_gt (gt, i915, i) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt); } }
Signature · 1 line
2.42.0
PATCH 2/3 drm/i915 Rename intel_engine_reset to intel_gt_engine_recover
intel_engine_reset() not only reset a engine but also tries to recover it so give it a proper name without any functional changes.
--- .../drm/i915/gem/selftests/i915_gem_context.c | 2 +- .../drm/i915/gt/intel_execlists_submission.c | 2 +- drivers/gpu/drm/i915/gt/intel_reset.c | 4 ++-- drivers/gpu/drm/i915/gt/intel_reset.h | 4 ++-- drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 20 +++++++++---------- drivers/gpu/drm/i915/gt/selftest_mocs.c | 4 ++-- drivers/gpu/drm/i915/gt/selftest_reset.c | 2 +- .../gpu/drm/i915/gt/selftest_workarounds.c | 6 +++--- 8 files changed, 22 insertions(+), 22 deletions(-)
drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c +1 -1
@@ -1171,7 +1171,7 @@ __sseu_finish(const char *name,int ret =0 ;if (flags &TEST_RESET ) { - ret =intel_engine_reset (ce->engine,"sseu" ); + ret =intel_gt_engine_recover (ce->engine,"sseu" );if (ret) goto out; }
drivers/gpu/drm/i915/gt/intel_execlists_submission.c +1 -1
@@ -2404,7 +2404,7 @@ static void execlists_reset(struct intel_engine_cs *engine, const char *msg)ring_set_paused (engine,1 );/* Freeze the current request in place */ execlists_capture (engine); -intel_engine_reset (engine, msg); +intel_gt_engine_recover (engine, msg);tasklet_enable (&engine->sched_engine->tasklet);clear_and_wake_up_bit (bit, lock);
drivers/gpu/drm/i915/gt/intel_reset.c +2 -2
@@ -1348,7 +1348,7 @@ int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *msg) }/** -* intel_engine_reset - reset GPU engine to recover from a hang +* intel_gt_engine_recover - reset GPU engine to recover from a hang * @engine: engine to reset * @msg: reason for GPU reset; or NULL for no drm_notice() *
@@ -1360,7 +1360,7 @@ int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *msg)* - reset engine (which will force the engine to idle) * - re-init/configure engine */ -int intel_engine_reset (struct intel_engine_cs *engine,const char *msg) +int intel_gt_engine_recover (struct intel_engine_cs *engine,const char *msg) {int err;
drivers/gpu/drm/i915/gt/intel_reset.h +2 -2
@@ -31,8 +31,8 @@ void intel_gt_handle_error(struct intel_gt *gt,void intel_gt_reset (struct intel_gt *gt,intel_engine_mask_t stalled_mask,const char *reason); -int intel_engine_reset (struct intel_engine_cs *engine, -const char *reason); +int intel_gt_engine_recover (struct intel_engine_cs *engine, +const char *reason);int __intel_engine_reset_bh (struct intel_engine_cs *engine,const char *reason);
drivers/gpu/drm/i915/gt/selftest_hangcheck.c +10 -10
@@ -495,9 +495,9 @@ static int igt_reset_nop_engine(void *arg)i915_request_add (rq); } - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) { -pr_err ("intel_engine_reset(%s) failed, err:%d\n" , +pr_err ("intel_gt_engine_recover(%s) failed, err:%d\n" , engine->name, err);break ; }
@@ -574,7 +574,7 @@ static int igt_reset_fail_engine(void *arg) >->reset.flags));force_reset_timeout (engine); - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );cancel_reset_timeout (engine);if (err ==0 )/* timeouts only generated on gen8+ */ goto skip;
@@ -623,9 +623,9 @@ static int igt_reset_fail_engine(void *arg) }if (count &1 ) { - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) { -GEM_TRACE_ERR ("intel_engine_reset(%s) failed, err:%d\n" , +GEM_TRACE_ERR ("intel_gt_engine_recover(%s) failed, err:%d\n" , engine->name, err);GEM_TRACE_DUMP ();i915_request_put (last);
@@ -633,10 +633,10 @@ static int igt_reset_fail_engine(void *arg) } }else {force_reset_timeout (engine); - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );cancel_reset_timeout (engine);if (err != -ETIMEDOUT ) { -pr_err ("intel_engine_reset(%s) did not fail, err:%d\n" , +pr_err ("intel_gt_engine_recover(%s) did not fail, err:%d\n" , engine->name, err);i915_request_put (last);break ;
@@ -765,9 +765,9 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active) }if (!using_guc) { - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) { -pr_err ("intel_engine_reset(%s) failed, err:%d\n" , +pr_err ("intel_gt_engine_recover(%s) failed, err:%d\n" , engine->name, err); goto skip; }
@@ -1085,7 +1085,7 @@ static int __igt_reset_engines(struct intel_gt *gt, }if (!using_guc) { - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) {pr_err ("i915_reset_engine(%s:%s): failed, err=%d\n" , engine->name, test_name, err);
drivers/gpu/drm/i915/gt/selftest_mocs.c +2 -2
@@ -336,7 +336,7 @@ static int active_engine_reset(struct intel_context *ce, err =request_add_spin (rq, &spin);if (err ==0 && !using_guc) - err =intel_engine_reset (ce->engine, reason); + err =intel_gt_engine_recover (ce->engine, reason);/* Ensure the reset happens and kills the engine */ if (err ==0 )
@@ -356,7 +356,7 @@ static int __live_mocs_reset(struct live_mocs *mocs,if (intel_has_reset_engine (gt)) {if (!using_guc) { - err =intel_engine_reset (ce->engine,"mocs" ); + err =intel_gt_engine_recover (ce->engine,"mocs" );if (err)return err;
drivers/gpu/drm/i915/gt/selftest_reset.c +1 -1
@@ -115,7 +115,7 @@ __igt_reset_stolen(struct intel_gt *gt, }else {for_each_engine (engine, gt, id) {if (mask & engine->mask) -intel_engine_reset (engine,NULL ); +intel_gt_engine_recover (engine,NULL ); } }
drivers/gpu/drm/i915/gt/selftest_workarounds.c +3 -3
@@ -256,7 +256,7 @@ static int do_device_reset(struct intel_engine_cs *engine)static int do_engine_reset (struct intel_engine_cs *engine) { -return intel_engine_reset (engine,"live_workarounds" ); +return intel_gt_engine_recover (engine,"live_workarounds" ); }static int do_guc_reset (struct intel_engine_cs *engine)
@@ -1282,7 +1282,7 @@ live_engine_reset_workarounds(void *arg) goto err; } - ret =intel_engine_reset (engine,"live_workarounds:idle" ); + ret =intel_gt_engine_recover (engine,"live_workarounds:idle" );if (ret) {pr_err ("%s: Reset failed while idle\n" , engine->name); goto err;
@@ -1320,7 +1320,7 @@ live_engine_reset_workarounds(void *arg) }if (!using_guc) { - ret =intel_engine_reset (engine,"live_workarounds:active" ); + ret =intel_gt_engine_recover (engine,"live_workarounds:active" );if (ret) {pr_err ("%s: Reset failed on an active spinner\n" , engine->name);
Signature · 1 line
2.42.0
Re: [PATCH 2/3] drm/i915 Rename intel_engine_reset to intel_gt_engine_recover
Quoted 3 lines · Nirmoy Das
intel_engine_reset()not only reset a engine but also tries to recover it so give it a proper name without any functional changes.
Not seeing what the difference is. If this was a super low level function (with an __ prefix for example) then one might expect it to literally just poke the reset register and leave the engine in a dead state. But as a high level function, I think it is reasonable to expect a reset function to 'recover' the entity being reset.
Also, many of the callers are tests that are explicitly testing reset. So now the tests all talk about attempting resets, resets failing, etc. but around a call to 'recover' instead of 'reset', which seems confusing.
John.
Quoted 213 lines
Signed-off-by Nirmoy Das <[email protected]>--- .../drm/i915/gem/selftests/i915_gem_context.c | 2 +- .../drm/i915/gt/intel_execlists_submission.c | 2 +- drivers/gpu/drm/i915/gt/intel_reset.c | 4 ++-- drivers/gpu/drm/i915/gt/intel_reset.h | 4 ++-- drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 20 +++++++++---------- drivers/gpu/drm/i915/gt/selftest_mocs.c | 4 ++-- drivers/gpu/drm/i915/gt/selftest_reset.c | 2 +- .../gpu/drm/i915/gt/selftest_workarounds.c | 6 +++--- 8 files changed, 22 insertions(+), 22 deletions(-)drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c +1
-1@@ -1171,7 +1171,7 @@ __sseu_finish(const char *name,int ret =0 ;if (flags &TEST_RESET ) { - ret =intel_engine_reset (ce->engine,"sseu" ); + ret =intel_gt_engine_recover (ce->engine,"sseu" );if (ret) goto out; }drivers/gpu/drm/i915/gt/intel_execlists_submission.c +1
-1@@ -2404,7 +2404,7 @@ static void execlists_reset(struct intel_engine_cs *engine, const char *msg)ring_set_paused (engine,1 );/* Freeze the current request in place */ execlists_capture (engine); -intel_engine_reset (engine, msg); +intel_gt_engine_recover (engine, msg);tasklet_enable (&engine->sched_engine->tasklet);clear_and_wake_up_bit (bit, lock);drivers/gpu/drm/i915/gt/intel_reset.c +2
-2@@ -1348,7 +1348,7 @@ int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *msg) }/** -* intel_engine_reset - reset GPU engine to recover from a hang +* intel_gt_engine_recover - reset GPU engine to recover from a hang * @engine: engine to reset * @msg: reason for GPU reset; or NULL for no drm_notice() * @@ -1360,7 +1360,7 @@ int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *msg)* - reset engine (which will force the engine to idle) * - re-init/configure engine */ -int intel_engine_reset (struct intel_engine_cs *engine,const char *msg) +int intel_gt_engine_recover (struct intel_engine_cs *engine,const char *msg) {int err;drivers/gpu/drm/i915/gt/intel_reset.h +2
-2@@ -31,8 +31,8 @@ void intel_gt_handle_error(struct intel_gt *gt,void intel_gt_reset (struct intel_gt *gt,intel_engine_mask_t stalled_mask,const char *reason); -int intel_engine_reset (struct intel_engine_cs *engine, -const char *reason); +int intel_gt_engine_recover (struct intel_engine_cs *engine, +const char *reason);int __intel_engine_reset_bh (struct intel_engine_cs *engine,const char *reason);drivers/gpu/drm/i915/gt/selftest_hangcheck.c +10
-10@@ -495,9 +495,9 @@ static int igt_reset_nop_engine(void *arg)i915_request_add (rq); } - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) { -pr_err ("intel_engine_reset(%s) failed, err:%d\n" , +pr_err ("intel_gt_engine_recover(%s) failed, err:%d\n" , engine->name, err);break ; }@@ -574,7 +574,7 @@ static int igt_reset_fail_engine(void *arg) >->reset.flags));force_reset_timeout (engine); - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );cancel_reset_timeout (engine);if (err ==0 )/* timeouts only generated on gen8+ */ goto skip;@@ -623,9 +623,9 @@ static int igt_reset_fail_engine(void *arg) }if (count &1 ) { - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) { -GEM_TRACE_ERR ("intel_engine_reset(%s) failed, err:%d\n" , +GEM_TRACE_ERR ("intel_gt_engine_recover(%s) failed, err:%d\n" , engine->name, err);GEM_TRACE_DUMP ();i915_request_put (last);@@ -633,10 +633,10 @@ static int igt_reset_fail_engine(void *arg) } }else {force_reset_timeout (engine); - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );cancel_reset_timeout (engine);if (err != -ETIMEDOUT ) { -pr_err ("intel_engine_reset(%s) did not fail, err:%d\n" , +pr_err ("intel_gt_engine_recover(%s) did not fail, err:%d\n" , engine->name, err);i915_request_put (last);break ;@@ -765,9 +765,9 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active) }if (!using_guc) { - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) { -pr_err ("intel_engine_reset(%s) failed, err:%d\n" , +pr_err ("intel_gt_engine_recover(%s) failed, err:%d\n" , engine->name, err); goto skip; }@@ -1085,7 +1085,7 @@ static int __igt_reset_engines(struct intel_gt *gt, }if (!using_guc) { - err =intel_engine_reset (engine,NULL ); + err =intel_gt_engine_recover (engine,NULL );if (err) {pr_err ("i915_reset_engine(%s:%s): failed, err=%d\n" , engine->name, test_name, err);drivers/gpu/drm/i915/gt/selftest_mocs.c +2
-2@@ -336,7 +336,7 @@ static int active_engine_reset(struct intel_context *ce, err =request_add_spin (rq, &spin);if (err ==0 && !using_guc) - err =intel_engine_reset (ce->engine, reason); + err =intel_gt_engine_recover (ce->engine, reason);/* Ensure the reset happens and kills the engine */ if (err ==0 )@@ -356,7 +356,7 @@ static int __live_mocs_reset(struct live_mocs *mocs,if (intel_has_reset_engine (gt)) {if (!using_guc) { - err =intel_engine_reset (ce->engine,"mocs" ); + err =intel_gt_engine_recover (ce->engine,"mocs" );if (err)return err;drivers/gpu/drm/i915/gt/selftest_reset.c +1
-1@@ -115,7 +115,7 @@ __igt_reset_stolen(struct intel_gt *gt, }else {for_each_engine (engine, gt, id) {if (mask & engine->mask) -intel_engine_reset (engine,NULL ); +intel_gt_engine_recover (engine,NULL ); } }drivers/gpu/drm/i915/gt/selftest_workarounds.c +3
-3@@ -256,7 +256,7 @@ static int do_device_reset(struct intel_engine_cs *engine)static int do_engine_reset (struct intel_engine_cs *engine) { -return intel_engine_reset (engine,"live_workarounds" ); +return intel_gt_engine_recover (engine,"live_workarounds" ); }static int do_guc_reset (struct intel_engine_cs *engine)@@ -1282,7 +1282,7 @@ live_engine_reset_workarounds(void *arg) goto err; } - ret =intel_engine_reset (engine,"live_workarounds:idle" ); + ret =intel_gt_engine_recover (engine,"live_workarounds:idle" );if (ret) {pr_err ("%s: Reset failed while idle\n" , engine->name); goto err;@@ -1320,7 +1320,7 @@ live_engine_reset_workarounds(void *arg) }if (!using_guc) { - ret =intel_engine_reset (engine,"live_workarounds:active" ); + ret =intel_gt_engine_recover (engine,"live_workarounds:active" );if (ret) {pr_err ("%s: Reset failed on an active spinner\n" , engine->name);
Re: [PATCH 2/3] drm/i915 Rename intel_engine_reset to intel_gt_engine_recover
Hi John,
Quoted 7 lines · John Harrison
Quoted 3 lines · Nirmoy Das
intel_engine_reset()not only reset a engine but also tries to recover it so give it a proper name without any functional changes.Not seeing what the difference is. If this was a super low level function (with an __ prefix for example) then one might expect it to literally just poke the reset register and leave the engine in a dead state. But as a high level function, I think it is reasonable to expect a reset function to 'recover' the entity being reset.
Also, many of the callers are tests that are explicitly testing reset. So now the tests all talk about attempting resets, resets failing, etc. but around a call to 'recover' instead of 'reset', which seems confusing.
Didn't think about it, I will drop it.
Thanks,
Nirmoy
Quoted 213 lines
John.
Quoted 209 lines
Signed-off-by Nirmoy Das <[email protected]>---
.../drm/i915/gem/selftests/i915_gem_context.c| 2 +- .../drm/i915/gt/intel_execlists_submission.c | 2 +- drivers/gpu/drm/i915/gt/intel_reset.c | 4 ++-- drivers/gpu/drm/i915/gt/intel_reset.h | 4 ++-- drivers/gpu/drm/i915/gt/selftest_hangcheck.c | 20 +++++++++---------- drivers/gpu/drm/i915/gt/selftest_mocs.c | 4 ++-- drivers/gpu/drm/i915/gt/selftest_reset.c | 2 +- .../gpu/drm/i915/gt/selftest_workarounds.c | 6 +++---8 files changed, 22 insertions(+), 22 deletions(-)drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c +1
-1@@ -1171,7 +1171,7 @@ __sseu_finish(const char *name, int ret = 0; if (flags & TEST_RESET) { - ret = intel_engine_reset(ce->engine, "sseu"); + ret = intel_gt_engine_recover(ce->engine, "sseu"); if (ret) goto out; }drivers/gpu/drm/i915/gt/intel_execlists_submission.c +0
-0@@ -2404,7 +2404,7 @@ static void execlists_reset(struct intel_engine_cs *engine, const char *msg)
ring_set_paused(engine, 1); /* Freeze the current request in place */execlists_capture(engine);
-intel_engine_reset(engine, msg);
+intel_gt_engine_recover(engine, msg);tasklet_enable(&engine->sched_engine->tasklet);clear_and_wake_up_bit(bit, lock);drivers/gpu/drm/i915/gt/intel_reset.c +2
-2@@ -1348,7 +1348,7 @@ int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *msg) } /** -* intel_engine_reset - reset GPU engine to recover from a hang +* intel_gt_engine_recover - reset GPU engine to recover from a hang * @engine: engine to reset * @msg: reason for GPU reset; or NULL for no drm_notice() * @@ -1360,7 +1360,7 @@ int __intel_engine_reset_bh(structintel_engine_cs *engine, const char *msg) * - reset engine (which will force the engine to idle) * - re-init/configure engine */ -int intel_engine_reset (struct intel_engine_cs *engine,const char *msg) +int intel_gt_engine_recover (struct intel_engine_cs *engine,const char *msg) {int err;
diff --gita/drivers/gpu/drm/i915/gt/intel_reset.hb/drivers/gpu/drm/i915/gt/intel_reset.hindexc00de353075c..be984357bf27100644drivers/gpu/drm/i915/gt/intel_reset.h +2
-2@@ -31,8 +31,8 @@ void intel_gt_handle_error(struct intel_gt *gt, void intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask, const char *reason); -int intel_engine_reset (struct intel_engine_cs *engine, - const char *reason); +int intel_gt_engine_recover (struct intel_engine_cs *engine, + const char *reason); int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *reason);diff --git
a/drivers/gpu/drm/i915/gt/selftest_hangcheck.cb/drivers/gpu/drm/i915/gt/selftest_hangcheck.cindex9ce8ff1c04fe..9bfda3f2bd24100644drivers/gpu/drm/i915/gt/selftest_hangcheck.c +3
-3@@ -495,9 +495,9 @@ static int igt_reset_nop_engine(void *arg) i915_request_add(rq); } - err = intel_engine_reset(engine, NULL); + err = intel_gt_engine_recover(engine, NULL); if (err) { - pr_err("intel_engine_reset(%s) failed, err:%d\n", + pr_err("intel_gt_engine_recover(%s) failed, err:%d\n", engine->name, err); break; }@@ -574,7 +574,7 @@ static int igt_reset_fail_engine(void *arg) >->reset.flags)); force_reset_timeout(engine); - err = intel_engine_reset(engine, NULL); + err = intel_gt_engine_recover(engine, NULL); cancel_reset_timeout(engine);if (err == 0) /* timeouts only generated on gen8+ */ goto skip;
drivers/gpu/drm/i915/gt/selftest_hangcheck.c +2
-2@@ -623,9 +623,9 @@ static int igt_reset_fail_engine(void *arg) } if (count & 1) { - err = intel_engine_reset(engine, NULL); + err = intel_gt_engine_recover(engine, NULL); if (err) { - GEM_TRACE_ERR("intel_engine_reset(%s) failed, err:%d\n", + GEM_TRACE_ERR("intel_gt_engine_recover(%s) failed, err:%d\n", engine->name, err); GEM_TRACE_DUMP();
i915_request_put(last);drivers/gpu/drm/i915/gt/selftest_hangcheck.c +2
-2@@ -633,10 +633,10 @@ static int igt_reset_fail_engine(void *arg) } } else { force_reset_timeout(engine); - err = intel_engine_reset(engine, NULL); + err = intel_gt_engine_recover(engine, NULL); cancel_reset_timeout(engine); if (err != -ETIMEDOUT) { - pr_err("intel_engine_reset(%s) did not fail, err:%d\n", + pr_err("intel_gt_engine_recover(%s) did not fail, err:%d\n", engine->name, err);
i915_request_put(last);
break;drivers/gpu/drm/i915/gt/selftest_hangcheck.c +2
-2@@ -765,9 +765,9 @@ static int __igt_reset_engine(struct intel_gt *gt, bool active) } if (!using_guc) { - err = intel_engine_reset(engine, NULL); + err = intel_gt_engine_recover(engine, NULL); if (err) { - pr_err("intel_engine_reset(%s) failed, err:%d\n", + pr_err("intel_gt_engine_recover(%s) failed, err:%d\n", engine->name, err); goto skip;}
drivers/gpu/drm/i915/gt/selftest_hangcheck.c +1
-1@@ -1085,7 +1085,7 @@ static int __igt_reset_engines(struct intel_gt *gt, } if (!using_guc) { - err = intel_engine_reset(engine, NULL); + err = intel_gt_engine_recover(engine, NULL); if (err) { pr_err("i915_reset_engine(%s:%s): failed, err=%d\n", engine->name, test_name, err);drivers/gpu/drm/i915/gt/selftest_mocs.c +2
-2@@ -336,7 +336,7 @@ static int active_engine_reset(structintel_context *ce, err = request_add_spin(rq, &spin); if (err == 0 && !using_guc) - err = intel_engine_reset(ce->engine, reason); + err = intel_gt_engine_recover(ce->engine, reason); /* Ensure the reset happens and kills the engine */ if (err == 0)@@ -356,7 +356,7 @@ static int __live_mocs_reset(struct live_mocs *mocs, if (intel_has_reset_engine(gt)) { if (!using_guc) { - err = intel_engine_reset(ce->engine, "mocs"); + err = intel_gt_engine_recover(ce->engine, "mocs"); if (err) return err;diff --git
a/drivers/gpu/drm/i915/gt/selftest_reset.cb/drivers/gpu/drm/i915/gt/selftest_reset.cindex2cfc23c58e90..9eaa1aed9f58100644drivers/gpu/drm/i915/gt/selftest_reset.c +1
-1@@ -115,7 +115,7 @@ __igt_reset_stolen(struct intel_gt *gt, } else { for_each_engine(engine, gt, id) { if (mask & engine->mask) - intel_engine_reset(engine, NULL); + intel_gt_engine_recover(engine, NULL); } }diff --git
a/drivers/gpu/drm/i915/gt/selftest_workarounds.cb/drivers/gpu/drm/i915/gt/selftest_workarounds.cindex14a8b25b6204..eb7516c7cb56100644drivers/gpu/drm/i915/gt/selftest_workarounds.c +2
-1@@ -256,7 +256,7 @@ static int do_device_reset(struct intel_engine_cs *engine) static int do_engine_reset(struct intel_engine_cs *engine) { - return intel_engine_reset(engine, "live_workarounds"); + return intel_gt_engine_recover(engine, "live_workarounds"); } static int do_guc_reset(struct intel_engine_cs *engine)@@ -1282,7 +1282,7 @@ live_engine_reset_workarounds(void *arg) goto err; } - ret = intel_engine_reset(engine, "live_workarounds:idle"); + ret = intel_gt_engine_recover(engine,"live_workarounds:idle" ); if (ret) { pr_err("%s: Reset failed while idle\n", engine->name);goto err;
drivers/gpu/drm/i915/gt/selftest_workarounds.c +1
-1@@ -1320,7 +1320,7 @@ live_engine_reset_workarounds(void *arg) } if (!using_guc) { - ret = intel_engine_reset(engine, "live_workarounds:active"); + ret = intel_gt_engine_recover(engine,"live_workarounds:active" ); if (ret) { pr_err("%s: Reset failed on an active spinner\n",
engine->name);
PATCH 3/3 drm/i915: Fix gt reset with GuC submission disabled
In reply to Nirmoy DasCurrently intel_gt_reset() happens as follows:
reset_prepare() ---> Sends GDRST to GuC, GuC is in GS_MIA_IN_RESET do_reset()
intel_gt_reset_all_engines()
*_engine_reset_prepare() -->RESET_CTL expects running GuC
*_reset_engines()intel_gt_init_hw() --> GuC comes out of GS_MIA_IN_RESET with FW loaded.
Fix the issue by sanitizing the GuC only after resetting requested engines and before intel_gt_init_hw().
Note intel_uc_reset_finish() and intel_uc_reset() are nop when guc submission is disabled.
--- drivers/gpu/drm/i915/gt/intel_reset.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)
drivers/gpu/drm/i915/gt/intel_reset.c +14 -2
@@ -907,8 +907,17 @@ static intel_engine_mask_t reset_prepare(struct intel_gt *gt)intel_engine_mask_t awake =0 ;enum intel_engine_id id; -/* For GuC mode, ensure submission is disabled before stopping ring */ -intel_uc_reset_prepare (>->uc); +/** +* For GuC mode with submission enabled, ensure submission +* is disabled before stopping ring. +* +* For GuC mode with submission disabled, ensure that GuC is not +* sanitized, do that at the end in reset_finish(). reset_prepare() +* is followed by engine reset which in this mode requires GuC to +* be functional to process engine reset events. +*/ +if (intel_uc_uses_guc_submission (>->uc)) +intel_uc_reset_prepare (>->uc);for_each_engine (engine, gt, id) {if (intel_engine_pm_get_if_awake (engine))
@@ -1255,6 +1264,9 @@ void intel_gt_reset(struct intel_gt *gt,intel_overlay_reset (gt->i915); +/* sanitize uC after engine reset */ +if (!intel_uc_uses_guc_submission (>->uc)) +intel_uc_reset_prepare (>->uc);/* * Next we need to restore the context, but we don't use those * yet either...
Signature · 1 line
2.42.0
Re: [PATCH 3/3] drm/i915: Fix gt reset with GuC submission disabled
Quoted 6 lines · Nirmoy Das
Currently
intel_gt_reset()happens as follows:
reset_prepare()---> Sends GDRST to GuC, GuC is inGS_MIA_IN_RESETdo_reset()
intel_gt_reset_all_engines()*_engine_reset_prepare() -->RESET_CTL expects running GuC
Not technically correct. There is no direct connection between RESET_CTL and GuC.
Quoted 5 lines
*_reset_engines()
intel_gt_init_hw()--> GuC comes out ofGS_MIA_IN_RESETwith FW loaded.Fix the issue by sanitizing the GuC only after resetting requested engines and before
intel_gt_init_hw().
You never actually state what the issue is.
The problem is that there is a dedicated CSB FIFO going to GuC (and nothing else has access to it). If that FIFO fills up, the hardware will block on the next context switch until there is space. If no-one (i.e. GuC) is draining it, that means the system is effectively hung. If an engine is reset whilst actively executing a context, a CSB entry will be sent to say that the context has gone idle. Thus if you reset a very busy system and start with killing GuC before killing the engines and only then re-enabling GuC, you run the risk of generating more CSB entries than will fit in the FIFO and deadlocking. Whereas, if the system is idle then you can reset the engines as much as you like while GuC is dead and it won't be a problem.
Quoted 17 lines
Note
intel_uc_reset_finish()andintel_uc_reset()are nop when guc submission is disabled.Signed-off-by Nirmoy Das <[email protected]>--- drivers/gpu/drm/i915/gt/intel_reset.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)drivers/gpu/drm/i915/gt/intel_reset.c +0
-0@@ -907,8 +907,17 @@ static intel_engine_mask_t reset_prepare(struct intel_gt *gt)intel_engine_mask_t awake =0 ;enum intel_engine_id id;
drivers/gpu/drm/i915/gt/intel_reset.c static intel_engine_mask_t reset_prepare(struct intel_gt *gt) · line 917
Quoted 10 lines
drivers/gpu/drm/i915/gt/intel_reset.c +8
-2-/* For GuC mode, ensure submission is disabled before stopping ring */ -intel_uc_reset_prepare (>->uc); +/** +* For GuC mode with submission enabled, ensure submission +* is disabled before stopping ring. +* +* For GuC mode with submission disabled, ensure that GuC is not +* sanitized, do that at the end in reset_finish(). reset_prepare() +* is followed by engine reset which in this mode requires GuC to +* be functional to process engine reset events.
-> to process any CSB FIFO entries generated by the resets.
John.
Quoted 16 lines
drivers/gpu/drm/i915/gt/intel_reset.c +6
-0+*/ +if (intel_uc_uses_guc_submission (>->uc)) +intel_uc_reset_prepare (>->uc);for_each_engine (engine, gt, id) {if (intel_engine_pm_get_if_awake (engine))@@ -1255,6 +1264,9 @@ void intel_gt_reset(struct intel_gt *gt,intel_overlay_reset (gt->i915); +/* sanitize uC after engine reset */ +if (!intel_uc_uses_guc_submission (>->uc)) +intel_uc_reset_prepare (>->uc);/* * Next we need to restore the context, but we don't use those * yet either...
Re: [PATCH 3/3] drm/i915: Fix gt reset with GuC submission disabled
Hi John,
Quoted 9 lines · John Harrison
Quoted 6 lines · Nirmoy Das
Currently
intel_gt_reset()happens as follows:
reset_prepare()---> Sends GDRST to GuC, GuC is inGS_MIA_IN_RESETdo_reset()
intel_gt_reset_all_engines()*_engine_reset_prepare()-->RESET_CTL expects running GuCNot technically correct. There is no direct connection between
RESET_CTLand GuC.
Quoted 8 lines
Quoted 5 lines
*_reset_engines()intel_gt_init_hw()--> GuC comes out ofGS_MIA_IN_RESETwith FW loaded.Fix the issue by sanitizing the GuC only after resetting requested engines and before
intel_gt_init_hw().You never actually state what the issue is.
The problem is that there is a dedicated CSB FIFO going to GuC (and nothing else has access to it). If that FIFO fills up, the hardware will block on the next context switch until there is space. If no-one (i.e. GuC) is draining it, that means the system is effectively hung. If an engine is reset whilst actively executing a context, a CSB entry will be sent to say that the context has gone idle. Thus if you reset a very busy system and start with killing GuC before killing the engines and only then re-enabling GuC, you run the risk of generating more CSB entries than will fit in the FIFO and deadlocking. Whereas, if the system is idle then you can reset the engines as much as you like while GuC is dead and it won't be a problem.
I wasn't sure if I could talk about internal details so kept it to minimal. I will borrow above explanation and resend :)
Quoted 21 lines
Quoted 21 lines
Note
intel_uc_reset_finish()andintel_uc_reset()are nop when guc submission is disabled.Signed-off-by Nirmoy Das <[email protected]>---
drivers/gpu/drm/i915/gt/intel_reset.c| 16 ++++++++++++++--1 file changed, 14 insertions(+), 2 deletions(-)drivers/gpu/drm/i915/gt/intel_reset.c +0
-0@@ -907,8 +907,17 @@ static intel_engine_mask_t reset_prepare(struct intel_gt *gt) intel_engine_mask_t awake = 0; enum intel_engine_id id;- /* For GuC mode, ensure submission is disabled before stopping ring */ -
intel_uc_reset_prepare(>->uc);
+ /**
Quoted 8 lines
Quoted 7 lines
+ * For GuC mode with submission enabled, ensure submission + * is disabled before stopping ring.
+ *
+ * For GuC mode with submission disabled, ensure that GuC is not + * sanitized, do that at the end inreset_finish().reset_prepare()+ * is followed by engine reset which in this mode requires GuC to + * be functional to process engine reset events.-> to process any CSB FIFO entries generated by the resets.
I will add this.
Thanks,
Nirmoy
Quoted 17 lines
John.
Quoted 13 lines
+ */
+ if (intel_uc_uses_guc_submission(>->uc)) +intel_uc_reset_prepare(>->uc);for_each_engine(engine, gt, id){ if (intel_engine_pm_get_if_awake(engine))drivers/gpu/drm/i915/gt/intel_reset.c +0
-0@@ -1255,6 +1264,9 @@ void intel_gt_reset(struct intel_gt *gt, intel_overlay_reset(gt->i915);+ /* sanitize uC after engine reset */
+ if (!intel_uc_uses_guc_submission(>->uc))
+intel_uc_reset_prepare(>->uc);
/*
* Next we need to restore the context, but we don't use those
* yet either...
Patchwork
CI✗ Fi.CI.SPARSE: warning for series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset()
In reply to Nirmoy Das== Series Details ==
== Series Details ==
Series: series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset()
URL : https://patchwork.freedesktop.org/series/132616/
State : warning
== Summary ==
Error: dim sparse failed
Sparse version: v0.6.2
Fast mode used, each commit won't be checked separately.
Patchwork
CI✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset()
In reply to Nirmoy Das== Series Details ==
== Series Details ==
Series: series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset()
URL : https://patchwork.freedesktop.org/series/132616/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_14606 -> Patchwork_132616v1
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with Patchwork_132616v1 absolutely need to be verified manually.
If you think the reported changes have nothing to do with the changes
introduced in Patchwork_132616v1, please notify your bug team ("[email protected]") to allow them to document this new failure mode, which will reduce false positives in CI.
External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/index.html
Participating hosts (39 -> 36)
------------------------------
Additional (2): bat-adlm-1 fi-kbl-8809g
Missing (5): fi-kbl-7567u bat-dg1-7 fi-glk-j4005 fi-elk-e7500 bat-dg2-11
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in Patchwork_132616v1:
### IGT changes ###
#### Possible regressions ####
* igt@debugfs_test@read_all_entries:
- bat-atsm-1: NOTRUN -> [INCOMPLETE][1]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-atsm-1/igt@debugfs_test@read_all_entries.html
* igt@kms_busy@basic@modeset:
- fi-ilk-650: [PASS][2] -> [INCOMPLETE][3]
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/fi-ilk-650/igt@kms_busy@[email protected]
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/fi-ilk-650/igt@kms_busy@[email protected]
#### Suppressed ####
The following results come from untrusted machines, tests, or statuses. They do not affect the overall result.
* igt@kms_busy@basic@flip:
- {bat-mtlp-9}: [PASS][4] -> [ABORT][5]
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-mtlp-9/igt@kms_busy@[email protected]
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-mtlp-9/igt@kms_busy@[email protected]
* igt@kms_busy@basic@modeset:
- {bat-mtlp-9}: [PASS][6] -> [DMESG-WARN][7]
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-mtlp-9/igt@kms_busy@[email protected]
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-mtlp-9/igt@kms_busy@[email protected]
Known issues
------------
Here are the changes found in Patchwork_132616v1 that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@debugfs_test@basic-hwmon:
- fi-kbl-8809g: NOTRUN -> [SKIP][8]
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/fi-kbl-8809g/igt@[email protected]
- bat-adlm-1: NOTRUN -> [SKIP][9] ([i915#3826])
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-adlm-1/igt@[email protected]
* igt@fbdev@eof:
- bat-adlm-1: NOTRUN -> [SKIP][10] ([i915#2582]) +3 other tests skip
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-adlm-1/igt@[email protected]
* igt@fbdev@info:
- bat-adlm-1: NOTRUN -> [SKIP][11] ([i915#1849] / [i915#2582])
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-adlm-1/igt@[email protected]
* igt@gem_close_race@basic-threads:
- fi-kbl-8809g: NOTRUN -> [ABORT][12] ([i915#10875])
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/fi-kbl-8809g/igt@[email protected]
- bat-adlm-1: NOTRUN -> [DMESG-WARN][13] ([i915#10875])
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-adlm-1/igt@[email protected]
* igt@i915_module_load@load:
- bat-dg2-13: [PASS][14] -> [INCOMPLETE][15] ([i915#10875])
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-dg2-13/igt@[email protected]
[15]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-dg2-13/igt@[email protected]
- bat-dg2-8: [PASS][16] -> [INCOMPLETE][17] ([i915#10877])
[16]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-dg2-8/igt@[email protected]
[17]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-dg2-8/igt@[email protected]
* igt@kms_busy@basic@modeset:
- fi-cfl-guc: [PASS][18] -> [INCOMPLETE][19] ([i915#10056])
[18]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/fi-cfl-guc/igt@kms_busy@[email protected]
[19]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/fi-cfl-guc/igt@kms_busy@[email protected]
- bat-mtlp-8: [PASS][20] -> [DMESG-WARN][21] ([i915#10875])
[20]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-mtlp-8/igt@kms_busy@[email protected]
[21]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-mtlp-8/igt@kms_busy@[email protected]
- bat-adls-6: [PASS][22] -> [ABORT][23] ([i915#10875])
[22]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-adls-6/igt@kms_busy@[email protected]
[23]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-adls-6/igt@kms_busy@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@b-edp1:
- bat-jsl-1: [PASS][24] -> [INCOMPLETE][25] ([i915#10876])
[24]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-jsl-1/igt@kms_flip@[email protected]
[25]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-jsl-1/igt@kms_flip@[email protected]
#### Possible fixes ####
* igt@gem_close_race@basic-process:
- bat-dg2-9: [INCOMPLETE][26] ([i915#10875]) -> [PASS][27]
[26]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-dg2-9/igt@[email protected]
[27]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-dg2-9/igt@[email protected]
* igt@i915_module_load@load:
- bat-atsm-1: [INCOMPLETE][28] ([i915#10875]) -> [PASS][29]
[28]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/bat-atsm-1/igt@[email protected]
[29]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/bat-atsm-1/igt@[email protected]
* igt@kms_busy@basic@modeset:
- fi-bsw-n3050: [DMESG-WARN][30] ([i915#10875]) -> [PASS][31]
[30]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14606/fi-bsw-n3050/igt@kms_busy@[email protected]
[31]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/fi-bsw-n3050/igt@kms_busy@[email protected]
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[i915#10056]: https://gitlab.freedesktop.org/drm/intel/issues/10056
[i915#10875]: https://gitlab.freedesktop.org/drm/intel/issues/10875
[i915#10876]: https://gitlab.freedesktop.org/drm/intel/issues/10876
[i915#10877]: https://gitlab.freedesktop.org/drm/intel/issues/10877
[i915#1849]: https://gitlab.freedesktop.org/drm/intel/issues/1849
[i915#2582]: https://gitlab.freedesktop.org/drm/intel/issues/2582
[i915#3826]: https://gitlab.freedesktop.org/drm/intel/issues/3826
Build changes
-------------
* Linux: CI_DRM_14606 -> Patchwork_132616v1
CI-20190529: 20190529
CI_DRM_14606: 121bb3cb2e7ce2d6f5c35cea84c6ae308cf1da0a @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_7813: 66841b7d9024447be4f4f5449aaf4f021e6323e5 @ https://gitlab.freedesktop.org/drm/igt-gpu-tools.git
Patchwork_132616v1: 121bb3cb2e7ce2d6f5c35cea84c6ae308cf1da0a @ git://anongit.freedesktop.org/gfx-ci/linux
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v1/index.html
Patchwork
CI✗ Fi.CI.SPARSE: warning for series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset() (rev2)
In reply to Nirmoy Das== Series Details ==
== Series Details ==
Series: series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset() (rev2)
URL : https://patchwork.freedesktop.org/series/132616/
State : warning
== Summary ==
Error: dim sparse failed
Sparse version: v0.6.2
Fast mode used, each commit won't be checked separately.
Patchwork
CI✗ Fi.CI.BAT: failure for series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset() (rev2)
In reply to Nirmoy Das== Series Details ==
== Series Details ==
Series: series starting with [1/3] drm/i915: Refactor confusing __intel_gt_reset() (rev2)
URL : https://patchwork.freedesktop.org/series/132616/
State : failure
== Summary ==
CI Bug Log - changes from CI_DRM_14607 -> Patchwork_132616v2
====================================================
Summary
-------
**FAILURE**
Serious unknown changes coming with Patchwork_132616v2 absolutely need to be verified manually.
If you think the reported changes have nothing to do with the changes
introduced in Patchwork_132616v2, please notify your bug team ("[email protected]") to allow them to document this new failure mode, which will reduce false positives in CI.
External URL: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/index.html
Participating hosts (37 -> 40)
------------------------------
Additional (4): fi-glk-j4005 bat-kbl-2 bat-jsl-1 fi-elk-e7500
Missing (1): bat-dg2-11
Possible new issues
-------------------
Here are the unknown changes that may have been introduced in Patchwork_132616v2:
### IGT changes ###
#### Possible regressions ####
* igt@gem_exec_basic@basic@vecs0-smem:
- fi-rkl-11600: NOTRUN -> [DMESG-WARN][1]
[1]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-rkl-11600/igt@gem_exec_basic@[email protected]
* igt@kms_busy@basic@flip:
- fi-bsw-nick: [PASS][2] -> [INCOMPLETE][3]
[2]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-bsw-nick/igt@kms_busy@[email protected]
[3]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-bsw-nick/igt@kms_busy@[email protected]
* igt@kms_busy@basic@modeset:
- fi-ivb-3770: [PASS][4] -> [INCOMPLETE][5]
[4]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-ivb-3770/igt@kms_busy@[email protected]
[5]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-ivb-3770/igt@kms_busy@[email protected]
* igt@kms_cursor_legacy@basic-flip-after-cursor-atomic:
- bat-adln-1: [PASS][6] -> [DMESG-WARN][7]
[6]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-adln-1/igt@[email protected]
[7]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-adln-1/igt@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@b-dp1:
- fi-apl-guc: NOTRUN -> [INCOMPLETE][8]
[8]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-apl-guc/igt@kms_flip@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@b-hdmi-a1:
- fi-cfl-guc: NOTRUN -> [INCOMPLETE][9]
[9]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-cfl-guc/igt@kms_flip@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@b-hdmi-a2:
- fi-glk-j4005: NOTRUN -> [INCOMPLETE][10]
[10]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-glk-j4005/igt@kms_flip@[email protected]
#### Warnings ####
* igt@i915_module_load@load:
- bat-adlm-1: [INCOMPLETE][11] ([i915#10875]) -> [INCOMPLETE][12]
[11]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-adlm-1/igt@[email protected]
[12]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-adlm-1/igt@[email protected]
#### Suppressed ####
The following results come from untrusted machines, tests, or statuses. They do not affect the overall result.
* igt@kms_cursor_legacy@basic-flip-after-cursor-legacy:
- {bat-mtlp-9}: NOTRUN -> [ABORT][13]
[13]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-mtlp-9/igt@[email protected]
Known issues
------------
Here are the changes found in Patchwork_132616v2 that come from known issues:
### IGT changes ###
#### Issues hit ####
* igt@debugfs_test@basic-hwmon:
- bat-jsl-1: NOTRUN -> [SKIP][14] ([i915#9318])
[14]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-jsl-1/igt@[email protected]
* igt@gem_huc_copy@huc-copy:
- fi-elk-e7500: NOTRUN -> [SKIP][15] +5 other tests skip
[15]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-elk-e7500/igt@[email protected]
- bat-jsl-1: NOTRUN -> [SKIP][16] ([i915#2190])
[16]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-jsl-1/igt@[email protected]
- fi-glk-j4005: NOTRUN -> [SKIP][17] ([i915#2190])
[17]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-glk-j4005/igt@[email protected]
* igt@i915_module_load@load:
- bat-kbl-2: NOTRUN -> [INCOMPLETE][18] ([i915#10877])
[18]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-kbl-2/igt@[email protected]
* igt@kms_busy@basic@modeset:
- bat-dg1-7: [PASS][19] -> [ABORT][20] ([i915#10875])
[19]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-dg1-7/igt@kms_busy@[email protected]
[20]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-dg1-7/igt@kms_busy@[email protected]
- bat-dg2-14: [PASS][21] -> [DMESG-WARN][22] ([i915#10875])
[21]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-dg2-14/igt@kms_busy@[email protected]
[22]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-dg2-14/igt@kms_busy@[email protected]
* igt@kms_cursor_legacy@basic-busy-flip-before-cursor-atomic:
- fi-glk-j4005: NOTRUN -> [SKIP][23] +3 other tests skip
[23]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-glk-j4005/igt@[email protected]
- bat-adlp-9: NOTRUN -> [SKIP][24] ([i915#4103]) +1 other test skip
[24]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-adlp-9/igt@[email protected]
* igt@kms_cursor_legacy@basic-busy-flip-before-cursor-legacy:
- fi-cfl-guc: NOTRUN -> [SKIP][25] +2 other tests skip
[25]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-cfl-guc/igt@[email protected]
- bat-jsl-1: NOTRUN -> [SKIP][26] ([i915#4103]) +1 other test skip
[26]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-jsl-1/igt@[email protected]
- bat-arls-1: NOTRUN -> [SKIP][27] ([i915#10202]) +1 other test skip
[27]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-arls-1/igt@[email protected]
- fi-apl-guc: NOTRUN -> [SKIP][28] +2 other tests skip
[28]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-apl-guc/igt@[email protected]
* igt@kms_cursor_legacy@basic-flip-after-cursor-atomic:
- bat-arls-2: [PASS][29] -> [ABORT][30] ([i915#10875])
[29]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-arls-2/igt@[email protected]
[30]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-arls-2/igt@[email protected]
- bat-arls-1: NOTRUN -> [ABORT][31] ([i915#10875])
[31]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-arls-1/igt@[email protected]
* igt@kms_dsc@dsc-basic:
- bat-jsl-1: NOTRUN -> [SKIP][32] ([i915#3555] / [i915#9886])
[32]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-jsl-1/igt@[email protected]
- bat-adlp-9: NOTRUN -> [SKIP][33] ([i915#3555] / [i915#3840])
[33]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-adlp-9/igt@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@b-edp1:
- bat-jsl-1: NOTRUN -> [INCOMPLETE][34] ([i915#10876])
[34]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-jsl-1/igt@kms_flip@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@c-dp1:
- bat-adlp-9: NOTRUN -> [INCOMPLETE][35] ([i915#10876])
[35]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-adlp-9/igt@kms_flip@[email protected]
#### Possible fixes ####
* igt@kms_busy@basic@flip:
- {bat-mtlp-9}: [ABORT][36] -> [PASS][37]
[36]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-mtlp-9/igt@kms_busy@[email protected]
[37]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-mtlp-9/igt@kms_busy@[email protected]
- fi-apl-guc: [INCOMPLETE][38] -> [PASS][39]
[38]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-apl-guc/igt@kms_busy@[email protected]
[39]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-apl-guc/igt@kms_busy@[email protected]
- fi-pnv-d510: [ABORT][40] -> [PASS][41]
[40]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-pnv-d510/igt@kms_busy@[email protected]
[41]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-pnv-d510/igt@kms_busy@[email protected]
- {bat-rpls-4}: [ABORT][42] -> [PASS][43]
[42]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-rpls-4/igt@kms_busy@[email protected]
[43]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-rpls-4/igt@kms_busy@[email protected]
* igt@kms_busy@basic@modeset:
- fi-cfl-guc: [INCOMPLETE][44] ([i915#10056]) -> [PASS][45]
[44]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-cfl-guc/igt@kms_busy@[email protected]
[45]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-cfl-guc/igt@kms_busy@[email protected]
- {bat-mtlp-9}: [DMESG-WARN][46] -> [PASS][47]
[46]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-mtlp-9/igt@kms_busy@[email protected]
[47]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-mtlp-9/igt@kms_busy@[email protected]
- bat-arls-1: [ABORT][48] ([i915#10875]) -> [PASS][49]
[48]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/bat-arls-1/igt@kms_busy@[email protected]
[49]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/bat-arls-1/igt@kms_busy@[email protected]
* igt@kms_flip@basic-flip-vs-dpms@b-dp1:
- fi-cfl-8109u: [INCOMPLETE][50] -> [PASS][51]
[50]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-cfl-8109u/igt@kms_flip@[email protected]
[51]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-cfl-8109u/igt@kms_flip@[email protected]
#### Warnings ####
* igt@kms_busy@basic@modeset:
- fi-pnv-d510: [SKIP][52] -> [ABORT][53] ([i915#10875])
[52]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_14607/fi-pnv-d510/igt@kms_busy@[email protected]
[53]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/fi-pnv-d510/igt@kms_busy@[email protected]
{name}: This element is suppressed. This means it is ignored when computing
the status of the difference (SUCCESS, WARNING, or FAILURE).
[i915#10056]: https://gitlab.freedesktop.org/drm/intel/issues/10056
[i915#10202]: https://gitlab.freedesktop.org/drm/intel/issues/10202
[i915#10875]: https://gitlab.freedesktop.org/drm/intel/issues/10875
[i915#10876]: https://gitlab.freedesktop.org/drm/intel/issues/10876
[i915#10877]: https://gitlab.freedesktop.org/drm/intel/issues/10877
[i915#2190]: https://gitlab.freedesktop.org/drm/intel/issues/2190
[i915#3555]: https://gitlab.freedesktop.org/drm/intel/issues/3555
[i915#3840]: https://gitlab.freedesktop.org/drm/intel/issues/3840
[i915#4103]: https://gitlab.freedesktop.org/drm/intel/issues/4103
[i915#4213]: https://gitlab.freedesktop.org/drm/intel/issues/4213
[i915#9318]: https://gitlab.freedesktop.org/drm/intel/issues/9318
[i915#9886]: https://gitlab.freedesktop.org/drm/intel/issues/9886
Build changes
-------------
* Linux: CI_DRM_14607 -> Patchwork_132616v2
CI-20190529: 20190529
CI_DRM_14607: cabe88f47c1f688f4493de88acc532bf584efe3c @ git://anongit.freedesktop.org/gfx-ci/linux
IGT_7813: 66841b7d9024447be4f4f5449aaf4f021e6323e5 @ https://gitlab.freedesktop.org/drm/igt-gpu-tools.git
Patchwork_132616v2: cabe88f47c1f688f4493de88acc532bf584efe3c @ git://anongit.freedesktop.org/gfx-ci/linux
== Logs ==
For more details see: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_132616v2/index.html
Quoted 3 lines · Nirmoy Das
__intel_gt_reset()is really for resetting engines though the name might suggest something else. So add two helper functions to remove confusions with no functional changes.
Technically you only added one and just moved the other :). It already existed, it just wasn't being used everywhere that it could be!
Quoted 100 lines
Signed-off-by Nirmoy Das <[email protected]>--- drivers/gpu/drm/i915/gt/intel_engine_cs.c | 2 +- .../drm/i915/gt/intel_execlists_submission.c | 2 +- drivers/gpu/drm/i915/gt/intel_gt.c | 2 +- drivers/gpu/drm/i915/gt/intel_gt_pm.c | 2 +- drivers/gpu/drm/i915/gt/intel_reset.c | 43 ++++++++++++++----- drivers/gpu/drm/i915/gt/intel_reset.h | 3 +- drivers/gpu/drm/i915/gt/selftest_reset.c | 2 +- drivers/gpu/drm/i915/i915_driver.c | 2 +- 8 files changed, 41 insertions(+), 17 deletions(-)drivers/gpu/drm/i915/gt/intel_engine_cs.c +1
-1@@ -678,7 +678,7 @@ void intel_engines_release(struct intel_gt *gt)*/ GEM_BUG_ON (intel_gt_pm_is_awake (gt));if (!INTEL_INFO (gt->i915)->gpu_reset_clobbers_display) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt);/* Decouple the backend; but keep the layout for late GPU resets */ for_each_engine (engine, gt, id) {drivers/gpu/drm/i915/gt/intel_execlists_submission.c +1
-1@@ -2898,7 +2898,7 @@ static void enable_error_interrupt(struct intel_engine_cs *engine)drm_err (&engine->i915->drm,"engine '%s' resumed still in error: %08x\n" , engine->name, status); -__intel_gt_reset (engine->gt, engine->mask); +intel_gt_reset_engine (engine); }/* drivers/gpu/drm/i915/gt/intel_gt.c +1
-1@@ -832,7 +832,7 @@ void intel_gt_driver_unregister(struct intel_gt *gt)/* Scrub all HW state upon release */ with_intel_runtime_pm(gt->uncore->rpm, wakeref) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt); }void intel_gt_driver_release (struct intel_gt *gt)drivers/gpu/drm/i915/gt/intel_gt_pm.c +1
-1@@ -159,7 +159,7 @@ static bool reset_engines(struct intel_gt *gt)if (INTEL_INFO (gt->i915)->gpu_reset_clobbers_display)return false; -return __intel_gt_reset (gt,ALL_ENGINES ) ==0 ; +return intel_gt_reset_all_engines (gt) ==0 ; }static void gt_sanitize (struct intel_gt *gt,bool force)drivers/gpu/drm/i915/gt/intel_reset.c +19
-1@@ -764,7 +764,7 @@ wa_14015076503_end(struct intel_gt *gt, intel_engine_mask_t engine_mask) HECI_H_GS1_ER_PREP, 0); } -int __intel_gt_reset (struct intel_gt *gt,intel_engine_mask_t engine_mask) +static int __intel_gt_reset (struct intel_gt *gt,intel_engine_mask_t engine_mask) {const int retries = engine_mask ==ALL_ENGINES ?RESET_MAX_RETRIES :1 ;reset_func reset;@@ -795,6 +795,34 @@ int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask)return ret; } +/** +* intel_gt_reset_all_engines() - Reset all engines in the given gt. +* @gt: the GT to reset all engines for. +* +* This function resets all engines within the given gt. +* +* Returns: +* Zero on success, negative error code on failure. +*/ +int intel_gt_reset_all_engines (struct intel_gt *gt) +{ +return __intel_gt_reset (gt,ALL_ENGINES ); +} + +/** +* intel_gt_reset_engine() - Reset a specific engine within a gt. +* @engine: engine to be reset. +*
drivers/gpu/drm/i915/gt/intel_reset.c int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask) · line 825
Quoted 10 lines
drivers/gpu/drm/i915/gt/intel_reset.c +10
-0+* This function resets the specified engine within a gt. +* +* Returns: +* Zero on success, negative error code on failure. +*/ +int intel_gt_reset_engine (struct intel_engine_cs *engine) +{ +return __intel_gt_reset (engine->gt, engine->mask); +} +
You could have just dropped the 'static' from the existing copy of this function and added the new version next to it. That would make the diff simpler and therefore clearer. Unless you think there is a good reason to move the code up here?
John.
Quoted 86 lines
drivers/gpu/drm/i915/gt/intel_reset.c +4
-9bool intel_has_gpu_reset (const struct intel_gt *gt) {if (!gt->i915->params.reset)@@ -978,7 +1006,7 @@ static void __intel_gt_set_wedged(struct intel_gt *gt)/* Even if the GPU reset fails, it should still stop the engines */ if (!INTEL_INFO (gt->i915)->gpu_reset_clobbers_display) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt);for_each_engine (engine, gt, id) engine->submit_request = nop_submit_request;@@ -1089,7 +1117,7 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt)/* We must reset pending GPU events before restoring our submission */ ok = !HAS_EXECLISTS (gt->i915);/* XXX better agnosticism desired */ if (!INTEL_INFO (gt->i915)->gpu_reset_clobbers_display) - ok =__intel_gt_reset (gt,ALL_ENGINES ) ==0 ; + ok =intel_gt_reset_all_engines (gt) ==0 ;if (!ok) {/* * Warn CI about the unrecoverable wedged condition. @@ -1133,10 +1161,10 @@ static int do_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask) {int err, i; - err =__intel_gt_reset (gt,ALL_ENGINES ); + err =intel_gt_reset_all_engines (gt);for (i =0 ; err && i <RESET_MAX_RETRIES ; i++) {msleep (10 * (i +1 )); - err =__intel_gt_reset (gt,ALL_ENGINES ); + err =intel_gt_reset_all_engines (gt); }if (err)return err;@@ -1270,11 +1298,6 @@ void intel_gt_reset(struct intel_gt *gt, goto finish; } -static int intel_gt_reset_engine (struct intel_engine_cs *engine) -{ -return __intel_gt_reset (engine->gt, engine->mask); -} -int __intel_engine_reset_bh (struct intel_engine_cs *engine,const char *msg) {struct intel_gt *gt = engine->gt;drivers/gpu/drm/i915/gt/intel_reset.h +2
-1@@ -54,7 +54,8 @@ int intel_gt_terminally_wedged(struct intel_gt *gt);void intel_gt_set_wedged_on_init (struct intel_gt *gt);void intel_gt_set_wedged_on_fini (struct intel_gt *gt); -int __intel_gt_reset (struct intel_gt *gt,intel_engine_mask_t engine_mask); +int intel_gt_reset_engine (struct intel_engine_cs *engine); +int intel_gt_reset_all_engines (struct intel_gt *gt);int intel_reset_guc (struct intel_gt *gt);drivers/gpu/drm/i915/gt/selftest_reset.c +1
-1@@ -281,7 +281,7 @@ static int igt_atomic_reset(void *arg) awake =reset_prepare (gt); p->critical_section_begin (); - err =__intel_gt_reset (gt,ALL_ENGINES ); + err =intel_gt_reset_all_engines (gt); p->critical_section_end ();reset_finish (gt, awake);drivers/gpu/drm/i915/i915_driver.c +1
-1@@ -202,7 +202,7 @@ static void sanitize_gpu(struct drm_i915_private *i915)unsigned int i;for_each_gt (gt, i915, i) -__intel_gt_reset (gt,ALL_ENGINES ); +intel_gt_reset_all_engines (gt); } }
Hi John.
Quoted 5 lines · John Harrison
Quoted 3 lines · Nirmoy Das
__intel_gt_reset()is really for resetting engines though the name might suggest something else. So add two helper functions to remove confusions with no functional changes.Technically you only added one and just moved the other :). It already existed, it just wasn't being used everywhere that it could be!
I did have one more helper functions but I removed it in favor of intel_gt_reset_engine() but didn't change the commit message :/
Thanks for catching it. I will fix it.
Quoted 100 lines
Quoted 100 lines
Signed-off-by Nirmoy Das <[email protected]>---
drivers/gpu/drm/i915/gt/intel_engine_cs.c | 2 +- .../drm/i915/gt/intel_execlists_submission.c | 2 +- drivers/gpu/drm/i915/gt/intel_gt.c | 2 +-drivers/gpu/drm/i915/gt/intel_gt_pm.c | 2 +- drivers/gpu/drm/i915/gt/intel_reset.c | 43 ++++++++++++++----- drivers/gpu/drm/i915/gt/intel_reset.h | 3 +- drivers/gpu/drm/i915/gt/selftest_reset.c | 2 +- drivers/gpu/drm/i915/i915_driver.c | 2 +- 8 files changed, 41 insertions(+), 17 deletions(-)drivers/gpu/drm/i915/gt/intel_engine_cs.c +1
-1@@ -678,7 +678,7 @@ void intel_engines_release(struct intel_gt *gt)*/ GEM_BUG_ON(intel_gt_pm_is_awake(gt)); if (!INTEL_INFO(gt->i915)->gpu_reset_clobbers_display) - __intel_gt_reset(gt, ALL_ENGINES); + intel_gt_reset_all_engines(gt); /* Decouple the backend; but keep the layout for late GPU resets */ for_each_engine(engine, gt, id) {drivers/gpu/drm/i915/gt/intel_execlists_submission.c +1
-1@@ -2898,7 +2898,7 @@ static void enable_error_interrupt(struct intel_engine_cs *engine) drm_err(&engine->i915->drm, "engine '%s' resumed still in error: %08x\n", engine->name, status); - __intel_gt_reset(engine->gt, engine->mask); + intel_gt_reset_engine(engine); } /*drivers/gpu/drm/i915/gt/intel_gt.c +1
-1@@ -832,7 +832,7 @@ void intel_gt_driver_unregister(struct intel_gt *gt) /* Scrub all HW state upon release */ with_intel_runtime_pm(gt->uncore->rpm, wakeref) - __intel_gt_reset(gt, ALL_ENGINES); + intel_gt_reset_all_engines(gt); } void intel_gt_driver_release(struct intel_gt *gt)drivers/gpu/drm/i915/gt/intel_gt_pm.c +1
-0@@ -159,7 +159,7 @@ static bool reset_engines(struct intel_gt *gt) if (INTEL_INFO(gt->i915)->gpu_reset_clobbers_display) return false; - return __intel_gt_reset(gt, ALL_ENGINES) == 0; + return intel_gt_reset_all_engines(gt) == 0; } static void gt_sanitize(struct intel_gt *gt, bool force)drivers/gpu/drm/i915/gt/intel_reset.c +1
-0@@ -764,7 +764,7 @@ wa_14015076503_end(struct intel_gt *gt, intel_engine_mask_t engine_mask) HECI_H_GS1_ER_PREP, 0); } -int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask) +static int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask) {const int retries =
engine_mask==ALL_ENGINES?RESET_MAX_RETRIES: 1;reset_funcreset;drivers/gpu/drm/i915/gt/intel_reset.c +0
-0@@ -795,6 +795,34 @@ int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask) return ret; }+/**
+ *intel_gt_reset_all_engines()- Reset all engines in the given gt.
+ * @gt: the GT to reset all engines for.
+ *
+ * This function resets all engines within the given gt.
+ *
+ * Returns:
+ * Zero on success, negative error code on failure.
+ */drivers/gpu/drm/i915/gt/intel_reset.c +12
-0+int intel_gt_reset_all_engines (struct intel_gt *gt) +{ + return __intel_gt_reset(gt, ALL_ENGINES); +} + +/** +* intel_gt_reset_engine() - Reset a specific engine within a gt. +* @engine: engine to be reset. +* +* This function resets the specified engine within a gt. +* +* Returns:
Quoted 8 lines
drivers/gpu/drm/i915/gt/intel_reset.c int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask) · line 825
Quoted 7 lines
+ * Zero on success, negative error code on failure.
+ */drivers/gpu/drm/i915/gt/intel_reset.c +5
-0+int intel_gt_reset_engine (struct intel_engine_cs *engine) +{ + return __intel_gt_reset(engine->gt, engine->mask); +} +You could have just dropped the 'static' from the existing copy of this function and added the new version next to it. That would make the diff simpler and therefore clearer. Unless you think there is a good reason to move the code up here?
Yes, make sense, I will do that.
Thanks,
Nirmoy
Quoted 83 lines
John.
Quoted 79 lines
bool
intel_has_gpu_reset(const struct intel_gt *gt){
if (!gt->i915->params.reset)drivers/gpu/drm/i915/gt/intel_reset.c +1
-1@@ -978,7 +1006,7 @@ static void __intel_gt_set_wedged(struct intel_gt *gt) /* Even if the GPU reset fails, it should still stop the engines */ if (!INTEL_INFO(gt->i915)->gpu_reset_clobbers_display) - __intel_gt_reset(gt, ALL_ENGINES); + intel_gt_reset_all_engines(gt); for_each_engine(engine, gt, id) engine->submit_request = nop_submit_request;@@ -1089,7 +1117,7 @@ static bool __intel_gt_unset_wedged(struct intel_gt *gt) /* We must reset pending GPU events before restoring our submission */ok = !HAS_EXECLISTS(gt->i915); /* XXX better agnosticism desired */ if (!INTEL_INFO(gt->i915)->gpu_reset_clobbers_display) - ok =
__intel_gt_reset(gt, ALL_ENGINES)== 0;
+ ok =intel_gt_reset_all_engines(gt)== 0;
if (!ok) {
/*
* Warn CI about the unrecoverable wedged condition.drivers/gpu/drm/i915/gt/intel_reset.c +2
-1@@ -1133,10 +1161,10 @@ static int do_reset(struct intel_gt *gt, intel_engine_mask_t stalled_mask) { int err, i; - err = __intel_gt_reset(gt, ALL_ENGINES); + err = intel_gt_reset_all_engines(gt); for (i = 0; err && i < RESET_MAX_RETRIES; i++) { msleep(10 * (i + 1)); - err = __intel_gt_reset(gt, ALL_ENGINES); + err = intel_gt_reset_all_engines(gt); } if (err)return err;
drivers/gpu/drm/i915/gt/intel_reset.c +0
-4@@ -1270,11 +1298,6 @@ void intel_gt_reset(struct intel_gt *gt, goto finish; } -static int intel_gt_reset_engine(struct intel_engine_cs *engine) -{ - return __intel_gt_reset(engine->gt, engine->mask); -} - int __intel_engine_reset_bh(struct intel_engine_cs *engine, const char *msg) { struct intel_gt *gt = engine->gt;drivers/gpu/drm/i915/gt/intel_reset.h +2
-0@@ -54,7 +54,8 @@ int intel_gt_terminally_wedged(struct intel_gt *gt); void intel_gt_set_wedged_on_init(struct intel_gt *gt); void intel_gt_set_wedged_on_fini(struct intel_gt *gt); -int __intel_gt_reset(struct intel_gt *gt, intel_engine_mask_t engine_mask); +int intel_gt_reset_engine (struct intel_engine_cs *engine); +int intel_gt_reset_all_engines (struct intel_gt *gt); int intel_reset_guc(struct intel_gt *gt);diff --git
a/drivers/gpu/drm/i915/gt/selftest_reset.cb/drivers/gpu/drm/i915/gt/selftest_reset.cindexf40de408cd3a..2cfc23c58e90100644drivers/gpu/drm/i915/gt/selftest_reset.c +1
-0@@ -281,7 +281,7 @@ static int igt_atomic_reset(void *arg) awake = reset_prepare(gt); p->critical_section_begin(); - err = __intel_gt_reset(gt, ALL_ENGINES); + err = intel_gt_reset_all_engines(gt); p->critical_section_end(); reset_finish(gt, awake);drivers/gpu/drm/i915/i915_driver.c +1
-1@@ -202,7 +202,7 @@ static void sanitize_gpu(struct drm_i915_private *i915) unsigned int i; for_each_gt(gt, i915, i) - __intel_gt_reset(gt, ALL_ENGINES); + intel_gt_reset_all_engines(gt); } }