Commit 157d2c7f authored by Chris Wilson's avatar Chris Wilson Committed by Jani Nikula

drm/i915: Stop automatically retiring requests after a GPU hang

Following a GPU hang, we break out of the request loop in order to
unlock the struct_mutex for use by the GPU reset. However, if we retire
all the requests at that moment, we cannot identify the guilty request
after performing the reset.

v2: Not automatically retiring requests forces us to recheck for
available ringspace.

Fixes: f4457ae7 ("drm/i915: Prevent leaking of -EIO from i915_wait_request()")
Testcase: igt/gem_reset_stats/ban-*
Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Tested-by: default avatarMika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: default avatarMika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/1463137042-9669-4-git-send-email-chris@chris-wilson.co.uk
(cherry picked from commit e075a32f)
Signed-off-by: default avatarJani Nikula <jani.nikula@intel.com>
parent 92dcc67c
...@@ -1456,7 +1456,10 @@ i915_wait_request(struct drm_i915_gem_request *req) ...@@ -1456,7 +1456,10 @@ i915_wait_request(struct drm_i915_gem_request *req)
if (ret) if (ret)
return ret; return ret;
__i915_gem_request_retire__upto(req); /* If the GPU hung, we want to keep the requests to find the guilty. */
if (req->reset_counter == i915_reset_counter(&dev_priv->gpu_error))
__i915_gem_request_retire__upto(req);
return 0; return 0;
} }
...@@ -1513,7 +1516,8 @@ i915_gem_object_retire_request(struct drm_i915_gem_object *obj, ...@@ -1513,7 +1516,8 @@ i915_gem_object_retire_request(struct drm_i915_gem_object *obj,
else if (obj->last_write_req == req) else if (obj->last_write_req == req)
i915_gem_object_retire__write(obj); i915_gem_object_retire__write(obj);
__i915_gem_request_retire__upto(req); if (req->reset_counter == i915_reset_counter(&req->i915->gpu_error))
__i915_gem_request_retire__upto(req);
} }
/* A nonblocking variant of the above wait. This is a highly dangerous routine /* A nonblocking variant of the above wait. This is a highly dangerous routine
......
...@@ -2456,6 +2456,8 @@ int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords) ...@@ -2456,6 +2456,8 @@ int intel_ring_begin(struct drm_i915_gem_request *req, int num_dwords)
return ret; return ret;
intel_ring_update_space(ringbuf); intel_ring_update_space(ringbuf);
if (unlikely(ringbuf->space < wait_bytes))
return -EAGAIN;
} }
if (unlikely(need_wrap)) { if (unlikely(need_wrap)) {
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment