On Thu, May 20, 2021 at 01:39:54PM +0530, Nava kishore Manne wrote:
> Add support for Xilinx Versal FPGA manager.
>
> PDI source type can be DDR, OCM, QSPI flash etc..
> But driver allocates memory always from DDR, Since driver supports only
> DDR source type.
>
> Signed-off-by: Appana Durga Kedareswara rao <appana.durga.rao@xxxxxxxxxx>
> Signed-off-by: Nava kishore Manne <nava.manne@xxxxxxxxxx>
> Reviewed-by: Moritz Fischer <mdf@xxxxxxxxxx>
> ---
> Changes for v2:
> -Updated the Fpga Mgr registrations call's
> to 5.11
> -Fixed some minor coding issues as suggested by
> Moritz.
> Changes for v3:
> -Rewritten the Versal fpga Kconfig contents.
> Changes for v4:
> -Rebased the changes on linux-next.
> No functional changes.
> Changes for v5:
> -None.
> Changes for v6:
> -None.
>
> drivers/fpga/Kconfig | 9 +++
> drivers/fpga/Makefile | 1 +
> drivers/fpga/versal-fpga.c | 117 +++++++++++++++++++++++++++++++++++++
> 3 files changed, 127 insertions(+)
> create mode 100644 drivers/fpga/versal-fpga.c
>
> diff --git a/drivers/fpga/Kconfig b/drivers/fpga/Kconfig
> index 33e15058d0dc..92c20b92357a 100644
> --- a/drivers/fpga/Kconfig
> +++ b/drivers/fpga/Kconfig
> @@ -234,4 +234,13 @@ config FPGA_MGR_ZYNQMP_FPGA
> to configure the programmable logic(PL) through PS
> on ZynqMP SoC.
>
> +config FPGA_MGR_VERSAL_FPGA
> + tristate "Xilinx Versal FPGA"
> + depends on ARCH_ZYNQMP || COMPILE_TEST
> + help
> + Select this option to enable FPGA manager driver support for
> + Xilinx Versal SoC. This driver uses the firmware interface to
> + configure the programmable logic(PL).
> +
> + To compile this as a module, choose M here.
> endif # FPGA
> diff --git a/drivers/fpga/Makefile b/drivers/fpga/Makefile
> index 18dc9885883a..0bff783d1b61 100644
> --- a/drivers/fpga/Makefile
> +++ b/drivers/fpga/Makefile
> @@ -18,6 +18,7 @@ obj-$(CONFIG_FPGA_MGR_TS73XX) += ts73xx-fpga.o
> obj-$(CONFIG_FPGA_MGR_XILINX_SPI) += xilinx-spi.o
> obj-$(CONFIG_FPGA_MGR_ZYNQ_FPGA) += zynq-fpga.o
> obj-$(CONFIG_FPGA_MGR_ZYNQMP_FPGA) += zynqmp-fpga.o
> +obj-$(CONFIG_FPGA_MGR_VERSAL_FPGA) += versal-fpga.o
> obj-$(CONFIG_ALTERA_PR_IP_CORE) += altera-pr-ip-core.o
> obj-$(CONFIG_ALTERA_PR_IP_CORE_PLAT) += altera-pr-ip-core-plat.o
>
> diff --git a/drivers/fpga/versal-fpga.c b/drivers/fpga/versal-fpga.c
> new file mode 100644
> index 000000000000..5744e44f981d
> --- /dev/null
> +++ b/drivers/fpga/versal-fpga.c
> @@ -0,0 +1,117 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * Copyright (C) 2019-2021 Xilinx, Inc.
> + */
> +
> +#include <linux/dma-mapping.h>
> +#include <linux/fpga/fpga-mgr.h>
> +#include <linux/io.h>
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/of_address.h>
> +#include <linux/string.h>
> +#include <linux/firmware/xlnx-zynqmp.h>
> +
> +/**
> + * struct versal_fpga_priv - Private data structure
> + * @dev: Device data structure
> + */
> +struct versal_fpga_priv {
> + struct device *dev;
> +};
Don't you have this pointer already? What device is this exactly and
why does it differ from the structure it currently lives in?
> +
> +static int versal_fpga_ops_write_init(struct fpga_manager *mgr,
> + struct fpga_image_info *info,
> + const char *buf, size_t size)
> +{
> + return 0;
> +}
If you don't need this, why include it?
> +
> +static int versal_fpga_ops_write(struct fpga_manager *mgr,
> + const char *buf, size_t size)
> +{
> + struct versal_fpga_priv *priv;
> + dma_addr_t dma_addr = 0;
> + char *kbuf;
> + int ret;
> +
> + priv = mgr->priv;
> +
> + kbuf = dma_alloc_coherent(priv->dev, size, &dma_addr, GFP_KERNEL);
> + if (!kbuf)
> + return -ENOMEM;
> +
> + memcpy(kbuf, buf, size);
> +
> + wmb(); /* ensure all writes are done before initiate FW call */
What "writes"? The memcpy above? Are you _SURE_ that really is correct
here? This feels wrong.
> +
> + ret = zynqmp_pm_load_pdi(PDI_SRC_DDR, dma_addr);
If this needs some sort of barrier, shouldn't it be in this call?
> +
> + dma_free_coherent(priv->dev, size, kbuf, dma_addr);
> +
> + return ret;
> +}
> +
> +static int versal_fpga_ops_write_complete(struct fpga_manager *mgr,
> + struct fpga_image_info *info)
> +{
> + return 0;
> +}
Again, why have it if it does nothing?
> +
> +static enum fpga_mgr_states versal_fpga_ops_state(struct fpga_manager *mgr)
> +{
> + return FPGA_MGR_STATE_UNKNOWN;
> +}
Again, is this needed? If so, then the fpga_manager core needs to be
fixed up :)
> +static const struct fpga_manager_ops versal_fpga_ops = {
> + .state = versal_fpga_ops_state,
> + .write_init = versal_fpga_ops_write_init,
> + .write = versal_fpga_ops_write,
> + .write_complete = versal_fpga_ops_write_complete,
> +};
> +
> +static int versal_fpga_probe(struct platform_device *pdev)
> +{
> + struct device *dev = &pdev->dev;
> + struct versal_fpga_priv *priv;
> + struct fpga_manager *mgr;
> + int ret;
> +
> + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL);
> + if (!priv)
> + return -ENOMEM;
> +
> + priv->dev = dev;
You save a pointer to a reference counted structure, without
incrementing the reference count. What could go wrong? :)
You are getting lucky here, but as stated above, why do you need this
pointer?
thanks,
greg k-h
