[dpdk-dev] [PATCH v4 1/3] Add Intel FPGA BUS Lib Code

[Xu, Rosen]

Hello Shreyansh,

> -----Original Message-----
> From: Shreyansh Jain [mailto:shreyansh.jain at nxp.com]
> Sent: Tuesday, April 03, 2018 17:25
> To: Xu, Rosen <rosen.xu at intel.com>
> Cc: dev at dpdk.org; Doherty, Declan <declan.doherty at intel.com>;
> Richardson, Bruce <bruce.richardson at intel.com>; Yigit, Ferruh
> <ferruh.yigit at intel.com>; Ananyev, Konstantin
> <konstantin.ananyev at intel.com>; Zhang, Tianfei <tianfei.zhang at intel.com>;
> Wu, Hao <hao.wu at intel.com>; gaetan.rivet at 6wind.com
> Subject: Re: [dpdk-dev] [PATCH v4 1/3] Add Intel FPGA BUS Lib Code
> 
> Hello Rosen,
> 
> On Saturday 31 March 2018 09:33 PM, Rosen Xu wrote:
> > Signed-off-by: Rosen Xu <rosen.xu at intel.com>
> > ---
> >   config/common_base                          |   5 +
> >   drivers/bus/Makefile                        |   1 +
> >   drivers/bus/ifpga/Makefile                  |  33 ++
> >   drivers/bus/ifpga/ifpga_bus.c               | 517
> ++++++++++++++++++++++++++++
> >   drivers/bus/ifpga/ifpga_common.c            | 141 ++++++++
> >   drivers/bus/ifpga/ifpga_common.h            |  22 ++
> >   drivers/bus/ifpga/ifpga_logs.h              |  31 ++
> >   drivers/bus/ifpga/rte_bus_ifpga.h           | 175 ++++++++++
> >   drivers/bus/ifpga/rte_bus_ifpga_version.map |   8 +
> >   mk/rte.app.mk                               |   2 +
> >   10 files changed, 935 insertions(+)
> >   create mode 100644 drivers/bus/ifpga/Makefile
> >   create mode 100644 drivers/bus/ifpga/ifpga_bus.c
> >   create mode 100644 drivers/bus/ifpga/ifpga_common.c
> >   create mode 100644 drivers/bus/ifpga/ifpga_common.h
> >   create mode 100644 drivers/bus/ifpga/ifpga_logs.h
> >   create mode 100644 drivers/bus/ifpga/rte_bus_ifpga.h
> >   create mode 100644 drivers/bus/ifpga/rte_bus_ifpga_version.map
> >
> > diff --git a/config/common_base b/config/common_base index
> > ad03cf4..49f6b09 100644
> > --- a/config/common_base
> > +++ b/config/common_base
> > @@ -134,6 +134,11 @@ CONFIG_RTE_LIBRTE_PCI_BUS=y
> >   CONFIG_RTE_LIBRTE_VDEV_BUS=y
> >
> >   #
> > +# Compile the Intel FPGA bus
> > +#
> > +CONFIG_RTE_LIBRTE_IFPGA_BUS=y
> > +
> > +#
> >   # Compile ARK PMD
> >   #
> >   CONFIG_RTE_LIBRTE_ARK_PMD=y
> > diff --git a/drivers/bus/Makefile b/drivers/bus/Makefile index
> > 7ef2593..55d2dfe 100644
> > --- a/drivers/bus/Makefile
> > +++ b/drivers/bus/Makefile
> > @@ -7,5 +7,6 @@ DIRS-$(CONFIG_RTE_LIBRTE_DPAA_BUS) += dpaa
> >   DIRS-$(CONFIG_RTE_LIBRTE_FSLMC_BUS) += fslmc
> >   DIRS-$(CONFIG_RTE_LIBRTE_PCI_BUS) += pci
> >   DIRS-$(CONFIG_RTE_LIBRTE_VDEV_BUS) += vdev
> > +DIRS-$(CONFIG_RTE_LIBRTE_IFPGA_BUS) += ifpga
> 
> When I attempted to compile the above using SHARED_LIB=y, this is what I
> got:
> 
> --->8---
>    LD librte_bus_ifpga.so.1.1
> ifpga_bus.o: In function `rte_ifpga_parse':
> ifpga_bus.c:(.text+0x2eb): undefined reference to `rte_rawdevs'
> ifpga_bus.o: In function `rte_ifpga_scan':
> ifpga_bus.c:(.text+0x53e): undefined reference to `rte_kvargs_parse'
> ifpga_bus.c:(.text+0x55e): undefined reference to `rte_kvargs_count'
> ifpga_bus.c:(.text+0x580): undefined reference to `rte_kvargs_process'
> ifpga_bus.c:(.text+0x5e0): undefined reference to `rte_rawdevs'
> ifpga_bus.c:(.text+0x746): undefined reference to `rte_kvargs_free'
> ifpga_bus.c:(.text+0x7b8): undefined reference to `rte_pci_addr_cmp'
> ifpga_bus.c:(.text+0x858): undefined reference to `rte_kvargs_parse'
> ifpga_bus.c:(.text+0x878): undefined reference to `rte_kvargs_count'
> ifpga_bus.c:(.text+0x89c): undefined reference to `rte_kvargs_process'
> ifpga_bus.c:(.text+0x8b8): undefined reference to `rte_kvargs_count'
> ifpga_bus.c:(.text+0x8dc): undefined reference to `rte_kvargs_process'
> ifpga_bus.c:(.text+0x8f8): undefined reference to `rte_kvargs_count'
> ifpga_bus.c:(.text+0x91c): undefined reference to `rte_kvargs_process'
> ifpga_bus.c:(.text+0x981): undefined reference to `rte_kvargs_free'
> ifpga_common.o: In function `ifpga_pci_addr_cmp':
> ifpga_common.c:(.text+0x2c5): undefined reference to
> `rte_eal_compare_pci_addr'
> collect2: error: ld returned 1 exit status
> /home/shreyansh/build/DPDK/00_dpdk/mk/rte.lib.mk:98: recipe for target
> 'librte_bus_ifpga.so.1.1' failed
> make[6]: *** [librte_bus_ifpga.so.1.1] Error 1
> /home/shreyansh/build/DPDK/00_dpdk/mk/rte.subdir.mk:35: recipe for
> target 'ifpga' failed
> make[5]: *** [ifpga] Error 2
> --->8---

I have fixed it to modify Makefile in my PATCH v5.

> Essentially, you are dependent on librte_kvargs but you have not declared
> that in your Makefile. Same for lib_rawdev.

I have declared it in my PATCH v5.

> >
> >   include $(RTE_SDK)/mk/rte.subdir.mk
> > diff --git a/drivers/bus/ifpga/Makefile b/drivers/bus/ifpga/Makefile
> > new file mode 100644 index 0000000..1b569af
> > --- /dev/null
> > +++ b/drivers/bus/ifpga/Makefile
> > @@ -0,0 +1,33 @@
> > +# SPDX-License-Identifier: BSD-3-Clause # Copyright(c) 2018 Intel
> > +Corporation
> > +
> > +include $(RTE_SDK)/mk/rte.vars.mk
> > +
> > +#
> > +# library name
> > +#
> > +LIB = librte_bus_ifpga.a
> > +
> > +CFLAGS += -O3
> > +CFLAGS += $(WERROR_FLAGS)
> > +
> > +# versioning export map
> > +EXPORT_MAP := rte_bus_ifpga_version.map
> > +
> > +# library version
> > +LIBABIVER := 1
> > +
> > +VPATH += $(SRCDIR)/base
> > +
> > +SRCS-y += \
> > +        ifpga_bus.c \
> > +        ifpga_common.c
> > +
> > +LDLIBS += -lrte_eal
> > +
> > +#
> > +# Export include files
> > +#
> > +SYMLINK-y-include += rte_bus_ifpga.h
> > +
> > +include $(RTE_SDK)/mk/rte.lib.mk
> > diff --git a/drivers/bus/ifpga/ifpga_bus.c
> > b/drivers/bus/ifpga/ifpga_bus.c new file mode 100644 index
> > 0000000..eba2615
> > --- /dev/null
> > +++ b/drivers/bus/ifpga/ifpga_bus.c
> > @@ -0,0 +1,517 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2010-2018 Intel Corporation  */
> > +
> 
> [...]
> 
> > +
> > +	if (rawdev->dev_ops &&
> > +		rawdev->dev_ops->dev_start &&
> > +		rawdev->dev_ops->dev_start(rawdev))
> > +		goto free_dev;
> > +	if (path) {
> > +		strncpy(afu_pr_conf.bs_path, path, strlen(path));
> 
> strncpy with demarcation based on source is not right. strlen(path) can be
> larger than afu_pr_conf.bs_path

I have changed to destination string in PATCH v5.

> I saw some comment from Gaetan in previous version, if I recall correctly.
> 
> > +		if (rawdev->dev_ops->firmware_load &&
> > +			rawdev->dev_ops->firmware_load(rawdev,
> > +					&afu_pr_conf)){
> > +			printf("firmware load error %d\n", ret);
> > +			goto free_dev;
> > +		}
> > +		afu_dev->id.uuid_low  = afu_pr_conf.afu_id.uuid_low;
> > +		afu_dev->id.uuid_high = afu_pr_conf.afu_id.uuid_high;
> > +	}
> > +
> > +	return afu_dev;
> > +
> > +free_dev:
> > +	free(afu_dev);
> > +end:
> > +	if (kvlist)
> > +		rte_kvargs_free(kvlist);
> > +	if (path)
> > +		free(path);
> > +
> > +	return NULL;
> > +}
> > +
> > +/*
> > + * Scan the content of the FPGA bus, and the devices in the devices
> > + * list
> > + */
> > +static int
> > +rte_ifpga_scan(void)
> > +{
> 
> [...]
> 
> > +
> > +static int
> > +ifpga_probe_all_drivers(struct rte_afu_device *afu_dev) {
> > +	struct rte_afu_driver *drv = NULL;
> > +	int rc;
> > +
> > +	if (afu_dev == NULL)
> > +		return -1;
> > +
> > +	/* Check if a driver is already loaded */
> > +	if (afu_dev->driver != NULL)
> > +		return 0;
> > +
> > +	TAILQ_FOREACH(drv, &rte_ifpga_bus.driver_list, next) {
> > +		rc = ifpga_probe_one_driver(drv, afu_dev);
> > +		if (rc < 0)
> > +			/* negative value is an error */
> > +			return -1;
> > +		if (rc > 0)
> > +			/* positive value means driver doesn't support it */
> > +			continue;
> > +		return 0;
> > +	}
> > +	return 1;
> > +}
> > +
> > +/*
> > + * Scan the content of the PCI bus, and call the probe() function for
> 
> I think you are not moving on PCI bus elements. You are looping on
> rte_ifpga_bus.

I have modified all PCI bus elements to FPGA BUS in my PATCH v5.
 
> > + * all registered drivers that have a matching entry in its id_table
> > + * for discovered devices.
> > + */
> > +static int
> > +rte_ifpga_probe(void)
> > +{
> > +	struct rte_ifpga_device *ifpga_dev;
> > +	struct rte_afu_device *afu_dev = NULL;
> > +	int ret = 0;
> > +
> > +	TAILQ_FOREACH(ifpga_dev, &rte_ifpga_bus.ifpga_list, next) {
> > +		TAILQ_FOREACH(afu_dev, &ifpga_dev->afu_list, next) {
> > +
> > +			if (afu_dev->device.driver)
> > +				continue;
> > +
> > +			ret = ifpga_probe_all_drivers(afu_dev);
> > +			if (ret < 0)
> > +				IFPGA_BUS_ERR("failed to initialize %s
> device\n",
> > +					rte_ifpga_device_name(afu_dev));
> > +		}
> > +	}
> > +
> > +	return 0;
> > +}
> > +
> 
> [...]
> 
> > diff --git a/drivers/bus/ifpga/ifpga_common.c
> > b/drivers/bus/ifpga/ifpga_common.c
> > new file mode 100644
> > index 0000000..124ffd2
> > --- /dev/null
> > +++ b/drivers/bus/ifpga/ifpga_common.c
> > @@ -0,0 +1,141 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2010-2018 Intel Corporation  */
> > +
> 
> [...]
> 
> > +
> > +}
> > +int ifpga_get_bdf_arg(const char *key __rte_unused,
> > +	const char *value, void *extra_args) { #define MAX_PATH_LEN 1024
> 
> Is this max len of a file path or a max len of the value string (BDF).
> Can you rename this?

I will rename it to IFPGA_MAX_BDF_LEN in my PATCH v5.

> Just a trivial comment, though.
> 
> [...]
> 
> > diff --git a/drivers/bus/ifpga/ifpga_logs.h
> > b/drivers/bus/ifpga/ifpga_logs.h new file mode 100644 index
> > 0000000..873e0a4
> > --- /dev/null
> > +++ b/drivers/bus/ifpga/ifpga_logs.h
> > @@ -0,0 +1,31 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2010-2018 Intel Corporation  */
> > +
> > +#ifndef _IFPGA_LOGS_H_
> > +#define _IFPGA_LOGS_H_
> > +
> > +#include <rte_log.h>
> > +
> > +extern int ifpga_bus_logtype;
> > +
> > +#define IFPGA_LOG(level, fmt, args...) \
> > +	rte_log(RTE_LOG_ ## level, ifpga_bus_logtype, "%s(): " fmt "\n", \
> > +		__func__, ##args)
> 
> I don't see macro above being used. Would you be using this in later patches?
> (But, i think they might have their own logging definitions).

Yes, they will be used in later patches.
For FPGA BUS is a common module, so we need to provide all macros will be used.
 
> > +
> > +#define IFPGA_BUS_LOG(level, fmt, args...) \
> > +	rte_log(RTE_LOG_ ## level, ifpga_bus_logtype, "%s(): " fmt "\n", \
> > +		__func__, ##args)
> > +
> > +#define IFPGA_BUS_FUNC_TRACE() IFPGA_BUS_LOG(DEBUG, ">>")
> > +
> > +#define IFPGA_BUS_DEBUG(fmt, args...) \
> > +	IFPGA_BUS_LOG(DEBUG, fmt, ## args)
> > +#define IFPGA_BUS_INFO(fmt, args...) \
> > +	IFPGA_BUS_LOG(INFO, fmt, ## args)
> > +#define IFPGA_BUS_ERR(fmt, args...) \
> > +	IFPGA_BUS_LOG(ERR, fmt, ## args)
> > +#define IFPGA_BUS_WARN(fmt, args...) \
> > +	IFPGA_BUS_LOG(WARNING, fmt, ## args)
> > +
> > +#endif /* _IFPGA_BUS_LOGS_H_ */
> > diff --git a/drivers/bus/ifpga/rte_bus_ifpga.h
> > b/drivers/bus/ifpga/rte_bus_ifpga.h
> > new file mode 100644
> > index 0000000..e22ab4e
> > --- /dev/null
> > +++ b/drivers/bus/ifpga/rte_bus_ifpga.h
> > @@ -0,0 +1,175 @@
> > +/* SPDX-License-Identifier: BSD-3-Clause
> > + * Copyright(c) 2010-2018 Intel Corporation  */
> > +
> > +#ifndef _RTE_BUS_IFPGA_H_
> > +#define _RTE_BUS_IFPGA_H_
> > +
> > +/**
> > + * @file
> > + *
> > + * RTE PCI Bus Interface
> 
> This is not a "RTE PCI Bus Interface" - It should be AFU/IFPA Bus interface file

I will modify it in PATCH v5.
 
> There are some comments which were given in early versions. Like this one.
> Please do respond to those individually if you are not fixing them.
> (Also, it is nice to get acknowledgment of those which are being fixed).

I have fix those comments and I also will reply it to those individually.
 
> > + */
> > +
> > +#ifdef __cplusplus
> > +extern "C" {
> > +#endif
> > +
> > +#include <rte_bus.h>
> > +#include <rte_pci.h>
> > +
> > +/** Name of Intel FPGA Bus */
> > +#define IFPGA_BUS_NAME ifpga
> > +
> > +/* Forward declarations */
> > +struct rte_ifpga_device;
> > +struct rte_afu_device;
> > +struct rte_afu_driver;
> > +
> > +/** List of Intel FPGA devices */
> > +TAILQ_HEAD(rte_ifpga_device_list, rte_ifpga_device);
> > +/** List of Intel AFU devices */
> > +TAILQ_HEAD(rte_afu_device_list, rte_afu_device);
> > +/** List of AFU drivers */
> > +TAILQ_HEAD(rte_afu_driver_list, rte_afu_driver);
> > +
> > +#define IFPGA_BUS_BITSTREAM_PATH_MAX_LEN 256
> > +
> > +struct rte_afu_uuid {
> > +	uint64_t uuid_low;
> > +	uint64_t uuid_high;
> > +} __attribute__ ((packed));
> > +
> > +#define IFPGA_BUS_DEV_PORT_MAX 4
> > +
> > +/**
> > + * A structure describing an ID for a AFU driver. Each driver
> > +provides a
> > + * table of these IDs for each device that it supports.
> > + */
> > +struct rte_afu_id {
> > +	struct rte_pci_addr pci_addr;
> > +	uint64_t uuid_low;
> > +	uint64_t uuid_high;
> > +	int      port;
> > +} __attribute__ ((packed));
> > +
> > +/**
> > + * A structure pr configuration AFU driver.
> > + */
> > +
> > +struct rte_afu_pr_conf {
> > +	struct rte_afu_id afu_id;
> > +	int pr_enable;
> > +	char		bs_path[IFPGA_BUS_BITSTREAM_PATH_MAX_LEN];
>              ^^^^^^^^^^^
> Some stray indentation issue, it seems

Do you means to change it  to:
char bs_path[IFPGA_BUS_BITSTREAM_PATH_MAX_LEN];?
If so I have fixed it in my PATCH v5.

> > +};
> > +
> > +#define AFU_PRI_STR_SIZE (PCI_PRI_STR_SIZE + 8)
> > +
> > +/**
> > + * A structure describing a fpga device.
> > + */
> > +struct rte_ifpga_device {
> > +	TAILQ_ENTRY(rte_ifpga_device) next;       /**< Next in device list. */
> > +	struct rte_pci_addr pci_addr;
> > +	struct rte_rawdev *rdev;
> > +	struct rte_afu_device_list afu_list;  /**< List of AFU devices */ };
> > +
> > +/**
> > + * A structure describing a AFU device.
> > + */
> > +struct rte_afu_device {
> > +	TAILQ_ENTRY(rte_afu_device) next;       /**< Next in device list. */
> > +	struct rte_device device;               /**< Inherit core device */
> > +	struct rte_rawdev *rawdev;    /**< Point Rawdev */
> > +	struct rte_ifpga_device *ifpga_dev;    /**< Point ifpga device */
> > +	struct rte_afu_id id;                   /**< AFU id within FPGA. */
> > +	uint32_t num_region;   /**< number of regions found */
> > +	struct rte_mem_resource mem_resource[PCI_MAX_RESOURCE];
> > +						/**< PCI Memory Resource
> */
> > +	struct rte_intr_handle intr_handle;     /**< Interrupt handle */
> > +	struct rte_afu_driver *driver;          /**< Associated driver */
> > +	char path[IFPGA_BUS_BITSTREAM_PATH_MAX_LEN];
> > +} __attribute__ ((packed));
> > +
> > +/**
> > + * @internal
> > + * Helper macro for drivers that need to convert to struct rte_afu_device.
> > + */
> > +#define RTE_DEV_TO_AFU(ptr) \
> > +	container_of(ptr, struct rte_afu_device, device)
> > +
> > +/**
> > + * Initialisation function for the driver called during PCI probing.
> 
> PCI Probing would have already been done through the PCI bus. I think this is
> probing of the AFU devices (based on the PCI already probed).

rte_ifpga_plug()/rte_ifpga_unplug() will use this macro.
 
> > + */
> > +typedef int (afu_probe_t)(struct rte_afu_device *);
> > +
> > +/**
> > + * Uninitialisation function for the driver called during hotplugging.
> > + */
> > +typedef int (afu_remove_t)(struct rte_afu_device *);
> > +
> > +/**
> > + * A structure describing a PCI device.
>                                ^^^^
> Structure describing an AFU device.

Yes, it should be AFU device, and I have fixed it in my PATCH v5.

> > + */
> > +struct rte_afu_driver {
> > +	TAILQ_ENTRY(rte_afu_driver) next;       /**< Next afu driver. */
> > +	struct rte_driver driver;               /**< Inherit core driver. */
> > +	afu_probe_t *probe;                     /**< Device Probe function. */
> > +	afu_remove_t *remove;                   /**< Device Remove function. */
> > +	const struct rte_afu_uuid *id_table;    /**< AFU uuid within FPGA. */
> > +	uint32_t drv_flags;         /**< Flags contolling handling of device. */
> > +};
> > +
> > +/**
> > + * Structure describing the Intel FPGA bus  */ struct rte_ifpga_bus {
> > +	struct rte_bus bus;               /**< Inherit the generic class */
> > +	struct rte_ifpga_device_list ifpga_list;  /**< List of FPGA devices */
> > +	struct rte_afu_driver_list driver_list;  /**< List of FPGA drivers
> > +*/ };
> > +
> > +static inline const char *
> > +rte_ifpga_device_name(const struct rte_afu_device *afu) {
> > +	if (afu && afu->device.name)
> > +		return afu->device.name;
> > +	return NULL;
> > +}
> > +
> > +extern struct rte_ifpga_bus rte_ifpga_bus;
> > +
> > +/**
> > + * Register a ifpga afu device driver.
> > + *
> > + * @param driver
> > + *   A pointer to a rte_afu_driver structure describing the driver
> > + *   to be registered.
> > + */
> > +void rte_ifpga_driver_register(struct rte_afu_driver *driver);
> > +
> > +/**
> > + * Unregister a ifpga afu device driver.
> > + *
> > + * @param driver
> > + *   A pointer to a rte_afu_driver structure describing the driver
> > + *   to be unregistered.
> > + */
> > +void rte_ifpga_driver_unregister(struct rte_afu_driver *driver);
> > +
> > +#define RTE_PMD_REGISTER_AFU(nm, afudrv)\ RTE_INIT(afudrvinitfn_
> > +##afudrv);\ static const char *afudrvinit_ ## nm ## _alias;\ static
> > +void afudrvinitfn_ ##afudrv(void)\ {\
> > +	(afudrv).driver.name = RTE_STR(nm);\
> > +	(afudrv).driver.alias = afudrvinit_ ## nm ## _alias;\
> > +	rte_ifpga_driver_register(&afudrv);\
> > +} \
> > +RTE_PMD_EXPORT_NAME(nm, __COUNTER__)
> > +
> > +#define RTE_PMD_REGISTER_AFU_ALIAS(nm, alias)\ static const char
> > +*afudrvinit_ ## nm ## _alias = RTE_STR(alias)
> > +
> > +#endif /* _RTE_BUS_IFPGA_H_ */
> > diff --git a/drivers/bus/ifpga/rte_bus_ifpga_version.map
> > b/drivers/bus/ifpga/rte_bus_ifpga_version.map
> > new file mode 100644
> > index 0000000..4edc9c0
> > --- /dev/null
> > +++ b/drivers/bus/ifpga/rte_bus_ifpga_version.map
> > @@ -0,0 +1,8 @@
> > +DPDK_18.05 {
> > +	global:
> > +
> > +    rte_ifpga_driver_register;
> > +    rte_ifpga_driver_unregister;
> 
> Should be tab indented

Yes, I will fix them.

> > +
> > +	local: *;
> > +};
> 
> [...]
> 
> One suggestion:
> 
> I think a lot of comments were provided by Gaetan in the previous version. I
> see some that some of them are still not fixed in this version.
> 
> I suggest you individually reply to his comments if you are not going to fix,
> with reason. He put quite an effort to go through your patches.
> 
> That would help you track comments as well as not discount a reviewers
> effort.

OK, I will individually reply to Gaetan's comments.

> -
> Shreyansh