On Wed, Aug 21, 2024 at 10:08:42AM -0700, Shashank Babu Chinta Venkata wrote:
> Refactor common code from RC(Root Complex) and EP(End Point)
> drivers and move them to a common driver. This acts as placeholder
> for common source code for both drivers, thus avoiding duplication.
Much of this seems to be replacing qcom_pcie_icc_opp_update() and
qcom_pcie_ep_icc_update() with qcom_pcie_common_icc_update().
That seems worthwhile and it would be helpful if the commit log called
that out so we'd know what to look for in the patch.
I think the qcom_pcie_common_icc_init() rework would be more
understandable if it were in its own patch and not mixed in here.
> +++ b/drivers/pci/controller/dwc/pcie-qcom-common.c
> @@ -0,0 +1,88 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (c) 2014-2015, 2020 The Linux Foundation. All rights reserved.
> + * Copyright (c) 2015, 2021 Linaro Limited.
> + * Copyright (c) 2024 Qualcomm Innovation Center, Inc. All rights reserved.
> + *
Spurious blank line.
> + */
> +struct icc_path *qcom_pcie_common_icc_get_resource(struct dw_pcie *pci, const char *path)
> +{
> + struct icc_path *icc_p;
> +
> + icc_p = devm_of_icc_get(pci->dev, path);
> + return icc_p;
return devm_of_icc_get(pci->dev, path);
> +}
> +EXPORT_SYMBOL_GPL(qcom_pcie_common_icc_get_resource);
> +
> +int qcom_pcie_common_icc_init(struct dw_pcie *pci, struct icc_path *icc, u32 bandwidth)
> +{
> + int ret;
> +
> + ret = icc_set_bw(icc, 0, bandwidth);
> + if (ret) {
> + dev_err(pci->dev, "Failed to set interconnect bandwidth: %d\n",
> + ret);
> + return ret;
> + }
The callers also check and log similar messages. I don't see the
point.
> +
> + return 0;
> +}
> +EXPORT_SYMBOL_GPL(qcom_pcie_common_icc_init);
These both seem of dubious value.
> +++ b/drivers/pci/controller/dwc/pcie-qcom-common.h
Do we need "-common" in the filename? Seems like "pcie-qcom.h" would
be enough. I *hope* we don't someday need both a "pcie-qcom.h and a
"pcie-qcom-common.h"; that seems like it would really be overkill.
Bjorn
