On Fri, Feb 03, 2017 at 02:45:29PM -0800, Daniele Ceraolo Spurio wrote:
> Fences are creted/checked before the pm ref is taken, so if we jump to
> pre_mutex_err we will uncorrectly call intel_runtime_pm_put.
>
> Fixes: fec0445caa27 (drm/i915: Support explicit fencing for execbuf)
> Testcase: igt/gem_exec_params
> Cc: Chris Wilson <chris@xxxxxxxxxxxxxxxxxx>
> Signed-off-by: Daniele Ceraolo Spurio <daniele.ceraolospurio@xxxxxxxxx>
Sigh. The tree I was using has this:
if (args->flags & I915_EXEC_FENCE_IN) {
in_fence = sync_file_get_fence(lower_32_bits(args->rsvd2));
if (!in_fence)
return -EINVAL;
}
if (args->flags & I915_EXEC_FENCE_OUT) {
out_fence_fd = get_unused_fd_flags(O_CLOEXEC);
if (out_fence_fd < 0) {
ret = out_fence_fd;
goto err_in_fence;
}
}
...
err_unlock:
mutex_unlock(&dev->struct_mutex);
err_rpm:
intel_runtime_pm_put(eb.i915);
eb_destroy(&eb);
if (out_fence_fd != -1)
put_unused_fd(out_fence_fd);
err_in_fence:
dma_fence_put(in_fence);
return ret;
}
Transforming the unwind sequence to match would be appreciated.
-Chris
--
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
