drm/msm/adreno: adreno_gpu: Use suspend() instead of idle() on load e…

…rror

commit 3eeca5e upstream.

The adreno_load_gpu() path is guarded by an error check on
adreno_load_fw(). This function is responsible for loading
Qualcomm-only-signed binaries (e.g. SQE and GMU FW for A6XX), but it
does not take the vendor-signed ZAP blob into account.

By embedding the SQE (and GMU, if necessary) firmware into the
initrd/kernel, we can trigger and unfortunate path that would not bail
out early and proceed with gpu->hw_init(). That will fail, as the ZAP
loader path will not find the firmware and return back to
adreno_load_gpu().

This error path involves pm_runtime_put_sync() which then calls idle()
instead of suspend(). This is suboptimal, as it means that we're not
going through the clean shutdown sequence. With at least A619_holi, this
makes the GPU not wake up until it goes through at least one more
start-fail-stop cycle. The pm_runtime_put_sync that appears in the error
path actually does not guarantee that because of the earlier enabling of
runtime autosuspend.

Fix that by using pm_runtime_put_sync_suspend to force a clean shutdown.

Test cases:
1. All firmware baked into kernel
2. error loading ZAP fw in initrd -> load from rootfs at DE start

Both succeed on A619_holi (SM6375) and A630 (SDM845).

Fixes: 0d997f9 ("drm/msm/adreno: fix runtime PM imbalance at gpu load")
Signed-off-by: Konrad Dybcio <konrad.dybcio@linaro.org>
Reviewed-by: Johan Hovold <johan+linaro@kernel.org>
Patchwork: https://patchwork.freedesktop.org/patch/530001/
Link: https://lore.kernel.org/r/20230330231517.2747024-1-konrad.dybcio@linaro.org
Signed-off-by: Rob Clark <robdclark@chromium.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

drivers/gpu/drm/msm/adreno/adreno_device.c

@@ -465,7 +465,7 @@ struct msm_gpu *adreno_load_gpu(struct drm_device *dev)
 	return gpu;
 
 err_put_rpm:
-	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_put_sync_suspend(&pdev->dev);
 err_disable_rpm:
 	pm_runtime_disable(&pdev->dev);