lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <AM5PR0401MB254500B1B76786E5C18626AD9A910@AM5PR0401MB2545.eurprd04.prod.outlook.com>
Date:   Mon, 19 Dec 2016 04:46:01 +0000
From:   Bharat Bhushan <bharat.bhushan@....com>
To:     Stuart Yoder <stuart.yoder@....com>,
        "gregkh@...uxfoundation.org" <gregkh@...uxfoundation.org>
CC:     "devel@...verdev.osuosl.org" <devel@...verdev.osuosl.org>,
        "agraf@...e.de" <agraf@...e.de>,
        Ruxandra Ioana Radulescu <ruxandra.radulescu@....com>,
        "arnd@...db.de" <arnd@...db.de>,
        Haiying Wang <haiying.wang@....com>,
        "Roy Pledge" <roy.pledge@....com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        Leo Li <leoyang.li@....com>,
        "Catalin Horghidan" <catalin.horghidan@....com>,
        Roy Pledge <roy.pledge@....com>,
        Stuart Yoder <stuart.yoder@....com>,
        Laurentiu Tudor <laurentiu.tudor@....com>
Subject: RE: [upstream-release] [PATCH v5 5/8] bus: fsl-mc: dpio: add QBMan
        portal APIs for DPAA2



> -----Original Message-----
> From: upstream-release-bounces@...ux.freescale.net [mailto:upstream-
> release-bounces@...ux.freescale.net] On Behalf Of Stuart Yoder
> Sent: Friday, December 16, 2016 10:01 PM
> To: gregkh@...uxfoundation.org
> Cc: devel@...verdev.osuosl.org; agraf@...e.de; Ruxandra Ioana Radulescu
> <ruxandra.radulescu@....com>; arnd@...db.de; Haiying Wang
> <haiying.wang@....com>; Roy Pledge <roy.pledge@....com>; linux-
> kernel@...r.kernel.org; Leo Li <leoyang.li@....com>; Catalin Horghidan
> <catalin.horghidan@....com>; Roy Pledge <roy.pledge@....com>; Stuart
> Yoder <stuart.yoder@....com>; Laurentiu Tudor
> <laurentiu.tudor@....com>
> Subject: [upstream-release] [PATCH v5 5/8] bus: fsl-mc: dpio: add QBMan
> portal APIs for DPAA2
> 
> From: Roy Pledge <Roy.Pledge@....com>
> 
> Add QBman APIs for frame queue and buffer pool operations.
> 
> Signed-off-by: Roy Pledge <roy.pledge@....com>
> Signed-off-by: Haiying Wang <haiying.wang@....com>
> Signed-off-by: Stuart Yoder <stuart.yoder@....com>
> ---
> 
> Notes:
>     -v5
>        -fixed typo that caused a compile error
>     -v4
>        -adjust file location to be in drivers/staging
>        -updated copyright
>        -added definition for static dequeue token value
>        -fixed bug in SDQCR #define
>        -added missing #include guard in qbman-portal.h
>        -added #define for QMAN_REV_MASK
>        -whitespace, alignment cleanup
>     -v3
>        -replace hardcoded dequeue token with a #define and check that
>         token when checking for a new result (bug fix suggested by
>         Ioana Radulescu)
>     -v2
>        -fix bug in buffer release command, by setting bpid field
>        -handle error (NULL) return value from qbman_swp_mc_complete()
>        -error message cleanup
>        -fix bug in sending management commands where the verb was
>         properly initialized
> 
>  drivers/staging/fsl-mc/bus/dpio/Makefile       |    2 +-
>  drivers/staging/fsl-mc/bus/dpio/qbman-portal.c | 1033
> ++++++++++++++++++++++++
>  drivers/staging/fsl-mc/bus/dpio/qbman-portal.h |  469 +++++++++++
>  3 files changed, 1503 insertions(+), 1 deletion(-)
>  create mode 100644 drivers/staging/fsl-mc/bus/dpio/qbman-portal.c
>  create mode 100644 drivers/staging/fsl-mc/bus/dpio/qbman-portal.h
> 
> diff --git a/drivers/staging/fsl-mc/bus/dpio/Makefile b/drivers/staging/fsl-
> mc/bus/dpio/Makefile
> index 128befc..6588498 100644
> --- a/drivers/staging/fsl-mc/bus/dpio/Makefile
> +++ b/drivers/staging/fsl-mc/bus/dpio/Makefile
> @@ -6,4 +6,4 @@ subdir-ccflags-y := -Werror
> 
>  obj-$(CONFIG_FSL_MC_DPIO) += fsl-mc-dpio.o
> 
> -fsl-mc-dpio-objs := dpio.o
> +fsl-mc-dpio-objs := dpio.o qbman-portal.o
> diff --git a/drivers/staging/fsl-mc/bus/dpio/qbman-portal.c
> b/drivers/staging/fsl-mc/bus/dpio/qbman-portal.c
> new file mode 100644
> index 0000000..4949102
> --- /dev/null
> +++ b/drivers/staging/fsl-mc/bus/dpio/qbman-portal.c
> @@ -0,0 +1,1033 @@
> +/*
> + * Copyright (C) 2014-2016 Freescale Semiconductor, Inc.
> + * Copyright 2016 NXP
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions are
> met:
> + *     * Redistributions of source code must retain the above copyright
> + *       notice, this list of conditions and the following disclaimer.
> + *     * Redistributions in binary form must reproduce the above copyright
> + *       notice, this list of conditions and the following disclaimer in the
> + *       documentation and/or other materials provided with the distribution.
> + *     * Neither the name of Freescale Semiconductor nor the
> + *       names of its contributors may be used to endorse or promote
> products
> + *       derived from this software without specific prior written permission.
> + *
> + * ALTERNATIVELY, this software may be distributed under the terms of the
> + * GNU General Public License ("GPL") as published by the Free Software
> + * Foundation, either version 2 of that License or (at your option) any
> + * later version.
> + *
> + * THIS SOFTWARE IS PROVIDED BY Freescale Semiconductor ``AS IS'' AND
> ANY
> + * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
> THE IMPLIED
> + * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
> PURPOSE ARE
> + * DISCLAIMED. IN NO EVENT SHALL Freescale Semiconductor BE LIABLE FOR
> ANY
> + * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
> CONSEQUENTIAL DAMAGES
> + * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
> GOODS OR SERVICES;
> + * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> HOWEVER CAUSED AND
> + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
> OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF
> THE USE OF THIS
> + * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + */
> +
> +#include <asm/cacheflush.h>
> +#include <linux/io.h>
> +#include <linux/slab.h>
> +#include "../../include/dpaa2-global.h"
> +
> +#include "qbman-portal.h"
> +
> +#define QMAN_REV_4000   0x04000000
> +#define QMAN_REV_4100   0x04010000
> +#define QMAN_REV_4101   0x04010001
> +#define QMAN_REV_MASK   0xffff0000
> +
> +/* All QBMan command and result structures use this "valid bit" encoding */
> +#define QB_VALID_BIT ((u32)0x80)

Can we define like 
#define QB_VALID_BIT		0x80U

> +
> +/* QBMan portal management command codes */
> +#define QBMAN_MC_ACQUIRE       0x30
> +#define QBMAN_WQCHAN_CONFIGURE 0x46
> +
> +/* CINH register offsets */
> +#define QBMAN_CINH_SWP_EQAR    0x8c0
> +#define QBMAN_CINH_SWP_DQPI    0xa00
> +#define QBMAN_CINH_SWP_DCAP    0xac0
> +#define QBMAN_CINH_SWP_SDQCR   0xb00
> +#define QBMAN_CINH_SWP_RAR     0xcc0
> +#define QBMAN_CINH_SWP_ISR     0xe00
> +#define QBMAN_CINH_SWP_IER     0xe40
> +#define QBMAN_CINH_SWP_ISDR    0xe80
> +#define QBMAN_CINH_SWP_IIR     0xec0
> +
> +/* CENA register offsets */
> +#define QBMAN_CENA_SWP_EQCR(n) (0x000 + ((u32)(n) << 6))
> +#define QBMAN_CENA_SWP_DQRR(n) (0x200 + ((u32)(n) << 6))
> +#define QBMAN_CENA_SWP_RCR(n)  (0x400 + ((u32)(n) << 6))
> +#define QBMAN_CENA_SWP_CR      0x600
> +#define QBMAN_CENA_SWP_RR(vb)  (0x700 + ((u32)(vb) >> 1))
> +#define QBMAN_CENA_SWP_VDQCR   0x780
> +
> +/* Reverse mapping of QBMAN_CENA_SWP_DQRR() */
> +#define QBMAN_IDX_FROM_DQRR(p) (((unsigned long)p & 0x1ff) >> 6)
> +
> +/* Define token used to determine if response written to memory is valid
> */
> +#define QMAN_DQ_TOKEN_VALID 1
> +
> +/* SDQCR attribute codes */
> +#define QB_SDQCR_FC_SHIFT   29
> +#define QB_SDQCR_FC_MASK    0x1
> +#define QB_SDQCR_DCT_SHIFT  24
> +#define QB_SDQCR_DCT_MASK   0x3
> +#define QB_SDQCR_TOK_SHIFT  16
> +#define QB_SDQCR_TOK_MASK   0xff
> +#define QB_SDQCR_SRC_SHIFT  0
> +#define QB_SDQCR_SRC_MASK   0xffff
> +
> +/* opaque token for static dequeues */
> +#define QMAN_SDQCR_TOKEN    0xbb
> +
> +enum qbman_sdqcr_dct {
> +	qbman_sdqcr_dct_null = 0,
> +	qbman_sdqcr_dct_prio_ics,
> +	qbman_sdqcr_dct_active_ics,
> +	qbman_sdqcr_dct_active
> +};
> +
> +enum qbman_sdqcr_fc {
> +	qbman_sdqcr_fc_one = 0,
> +	qbman_sdqcr_fc_up_to_3 = 1

 " = 1" can be avoided.

> +};
> +
> +/* Portal Access */
> +
> +static inline u32 qbman_read_register(struct qbman_swp *p, u32 offset)
> +{
> +	return readl_relaxed(p->addr_cinh + offset);
> +}
> +
> +static inline void qbman_write_register(struct qbman_swp *p, u32 offset,
> +					u32 value)
> +{
> +	writel_relaxed(value, p->addr_cinh + offset);
> +}
> +
> +static inline void *qbman_get_cmd(struct qbman_swp *p, u32 offset)
> +{
> +	return p->addr_cena + offset;
> +}
> +
> +#define QBMAN_CINH_SWP_CFG   0xd00
> +
> +#define SWP_CFG_DQRR_MF_SHIFT 20
> +#define SWP_CFG_EST_SHIFT     16
> +#define SWP_CFG_WN_SHIFT      14
> +#define SWP_CFG_RPM_SHIFT     12
> +#define SWP_CFG_DCM_SHIFT     10
> +#define SWP_CFG_EPM_SHIFT     8
> +#define SWP_CFG_SD_SHIFT      5
> +#define SWP_CFG_SP_SHIFT      4
> +#define SWP_CFG_SE_SHIFT      3
> +#define SWP_CFG_DP_SHIFT      2
> +#define SWP_CFG_DE_SHIFT      1
> +#define SWP_CFG_EP_SHIFT      0
> +
> +static inline u32 qbman_set_swp_cfg(u8 max_fill, u8 wn,	u8 est, u8
> rpm, u8 dcm,
> +				    u8 epm, int sd, int sp, int se,
> +				    int dp, int de, int ep)
> +{
> +	return cpu_to_le32 (max_fill << SWP_CFG_DQRR_MF_SHIFT |
> +			    est << SWP_CFG_EST_SHIFT |
> +			    wn << SWP_CFG_WN_SHIFT |
> +			    rpm << SWP_CFG_RPM_SHIFT |
> +			    dcm << SWP_CFG_DCM_SHIFT |
> +			    epm << SWP_CFG_EPM_SHIFT |
> +			    sd << SWP_CFG_SD_SHIFT |
> +			    sp << SWP_CFG_SP_SHIFT |
> +			    se << SWP_CFG_SE_SHIFT |
> +			    dp << SWP_CFG_DP_SHIFT |
> +			    de << SWP_CFG_DE_SHIFT |
> +			    ep << SWP_CFG_EP_SHIFT);
> +}
> +
> +/**
> + * qbman_swp_init() - Create a functional object representing the given
> + *                    QBMan portal descriptor.
> + * @d: the given qbman swp descriptor
> + *
> + * Return qbman_swp portal for success, NULL if the object cannot
> + * be created.
> + */
> +struct qbman_swp *qbman_swp_init(const struct qbman_swp_desc *d)
> +{
> +	struct qbman_swp *p = kmalloc(sizeof(*p), GFP_KERNEL);
> +	u32 reg;
> +
> +	if (!p)
> +		return NULL;
> +	p->desc = d;
> +	p->mc.valid_bit = QB_VALID_BIT;
> +	p->sdq = 0;
> +	p->sdq |= qbman_sdqcr_dct_prio_ics << QB_SDQCR_DCT_SHIFT;
> +	p->sdq |= qbman_sdqcr_fc_up_to_3 << QB_SDQCR_FC_SHIFT;
> +	p->sdq |= QMAN_SDQCR_TOKEN << QB_SDQCR_TOK_SHIFT;
> +
> +	atomic_set(&p->vdq.available, 1);
> +	p->vdq.valid_bit = QB_VALID_BIT;
> +	p->dqrr.next_idx = 0;
> +	p->dqrr.valid_bit = QB_VALID_BIT;
> +
> +	if ((p->desc->qman_version & QMAN_REV_MASK) <
> QMAN_REV_4100) {
> +		p->dqrr.dqrr_size = 4;
> +		p->dqrr.reset_bug = 1;
> +	} else {
> +		p->dqrr.dqrr_size = 8;
> +		p->dqrr.reset_bug = 0;
> +	}
> +
> +	p->addr_cena = d->cena_bar;
> +	p->addr_cinh = d->cinh_bar;
> +
> +	reg = qbman_set_swp_cfg(p->dqrr.dqrr_size,
> +				1, /* Writes Non-cacheable */
> +				0, /* EQCR_CI stashing threshold */
> +				3, /* RPM: Valid bit mode, RCR in array mode
> */
> +				2, /* DCM: Discrete consumption ack mode
> */
> +				3, /* EPM: Valid bit mode, EQCR in array
> mode */
> +				0, /* mem stashing drop enable == FALSE */
> +				1, /* mem stashing priority == TRUE */
> +				0, /* mem stashing enable == FALSE */
> +				1, /* dequeue stashing priority == TRUE */
> +				0, /* dequeue stashing enable == FALSE */
> +				0); /* EQCR_CI stashing priority == FALSE */
> +
> +	qbman_write_register(p, QBMAN_CINH_SWP_CFG, reg);
> +	reg = qbman_read_register(p, QBMAN_CINH_SWP_CFG);
> +	if (!reg) {
> +		pr_err("qbman: the portal is not enabled!\n");
> +		return NULL;
> +	}
> +
> +	/*
> +	 * SDQCR needs to be initialized to 0 when no channels are
> +	 * being dequeued from or else the QMan HW will indicate an
> +	 * error.  The values that were calculated above will be
> +	 * applied when dequeues from a specific channel are enabled.
> +	 */
> +	qbman_write_register(p, QBMAN_CINH_SWP_SDQCR, 0);
> +	return p;
> +}
> +
> +/**
> + * qbman_swp_finish() - Create and destroy a functional object
> representing
> + *                      the given QBMan portal descriptor.
> + * @p: the qbman_swp object to be destroyed
> + */
> +void qbman_swp_finish(struct qbman_swp *p)
> +{
> +	kfree(p);
> +}
> +
> +/**
> + * qbman_swp_interrupt_read_status()
> + * @p: the given software portal
> + *
> + * Return the value in the SWP_ISR register.
> + */
> +u32 qbman_swp_interrupt_read_status(struct qbman_swp *p)
> +{
> +	return qbman_read_register(p, QBMAN_CINH_SWP_ISR);
> +}
> +
> +/**
> + * qbman_swp_interrupt_clear_status()
> + * @p: the given software portal
> + * @mask: The mask to clear in SWP_ISR register
> + */
> +void qbman_swp_interrupt_clear_status(struct qbman_swp *p, u32 mask)
> +{
> +	qbman_write_register(p, QBMAN_CINH_SWP_ISR, mask);
> +}
> +
> +/**
> + * qbman_swp_interrupt_get_trigger() - read interrupt enable register
> + * @p: the given software portal
> + *
> + * Return the value in the SWP_IER register.
> + */
> +u32 qbman_swp_interrupt_get_trigger(struct qbman_swp *p)
> +{
> +	return qbman_read_register(p, QBMAN_CINH_SWP_IER);
> +}
> +
> +/**
> + * qbman_swp_interrupt_set_trigger() - enable interrupts for a swp
> + * @p: the given software portal
> + * @mask: The mask of bits to enable in SWP_IER
> + */
> +void qbman_swp_interrupt_set_trigger(struct qbman_swp *p, u32 mask)
> +{
> +	qbman_write_register(p, QBMAN_CINH_SWP_IER, mask);
> +}
> +
> +/**
> + * qbman_swp_interrupt_get_inhibit() - read interrupt mask register
> + * @p: the given software portal object
> + *
> + * Return the value in the SWP_IIR register.
> + */
> +int qbman_swp_interrupt_get_inhibit(struct qbman_swp *p)
> +{
> +	return qbman_read_register(p, QBMAN_CINH_SWP_IIR);
> +}
> +
> +/**
> + * qbman_swp_interrupt_set_inhibit() - write interrupt mask register
> + * @p: the given software portal object
> + * @mask: The mask to set in SWP_IIR register
> + */
> +void qbman_swp_interrupt_set_inhibit(struct qbman_swp *p, int inhibit)
> +{
> +	qbman_write_register(p, QBMAN_CINH_SWP_IIR, inhibit ? 0xffffffff
> : 0);
> +}
> +
> +/*
> + * Different management commands all use this common base layer of code
> to issue
> + * commands and poll for results.
> + */
> +
> +/*
> + * Returns a pointer to where the caller should fill in their management
> command
> + * (caller should ignore the verb byte)
> + */
> +void *qbman_swp_mc_start(struct qbman_swp *p)
> +{
> +	return qbman_get_cmd(p, QBMAN_CENA_SWP_CR);
> +}
> +
> +/*
> + * Commits merges in the caller-supplied command verb (which should not
> include
> + * the valid-bit) and submits the command to hardware
> + */
> +void qbman_swp_mc_submit(struct qbman_swp *p, void *cmd, u8
> cmd_verb)
> +{
> +	u8 *v = cmd;
> +
> +	dma_wmb();
> +	*v = cmd_verb | p->mc.valid_bit;
> +}
> +
> +/*
> + * Checks for a completed response (returns non-NULL if only if the
> response
> + * is complete).
> + */
> +void *qbman_swp_mc_result(struct qbman_swp *p)
> +{
> +	u32 *ret, verb;
> +
> +	ret = qbman_get_cmd(p, QBMAN_CENA_SWP_RR(p-
> >mc.valid_bit));
> +
> +	/* Remove the valid-bit - command completed if the rest is non-zero
> */
> +	verb = ret[0] & ~QB_VALID_BIT;
> +	if (!verb)
> +		return NULL;
> +	p->mc.valid_bit ^= QB_VALID_BIT;
> +	return ret;
> +}
> +
> +#define QB_ENQUEUE_CMD_OPTIONS_SHIFT    0
> +enum qb_enqueue_commands {
> +	enqueue_empty = 0,
> +	enqueue_response_always = 1,
> +	enqueue_rejects_to_fq = 2
> +};
> +
> +#define QB_ENQUEUE_CMD_ORP_ENABLE_SHIFT      2
> +#define QB_ENQUEUE_CMD_IRQ_ON_DISPATCH_SHIFT 3
> +#define QB_ENQUEUE_CMD_TARGET_TYPE_SHIFT     4
> +
> +/**
> + * qbman_eq_desc_clear() - Clear the contents of a descriptor to
> + *                         default/starting state.
> + */
> +void qbman_eq_desc_clear(struct qbman_eq_desc *d)
> +{
> +	memset(d, 0, sizeof(*d));
> +}
> +
> +/**
> + * qbman_eq_desc_set_no_orp() - Set enqueue descriptor without orp
> + * @d:                the enqueue descriptor.
> + * @response_success: 1 = enqueue with response always; 0 = enqueue
> with
> + *                    rejections returned on a FQ.
> + */
> +void qbman_eq_desc_set_no_orp(struct qbman_eq_desc *d, int
> respond_success)
> +{
> +	d->verb &= ~(1 << QB_ENQUEUE_CMD_ORP_ENABLE_SHIFT);
> +	if (respond_success)
> +		d->verb |= enqueue_response_always;
> +	else
> +		d->verb |= enqueue_rejects_to_fq;
> +}
> +
> +/*
> + * Exactly one of the following descriptor "targets" should be set. (Calling
> any
> + * one of these will replace the effect of any prior call to one of these.)
> + *   -enqueue to a frame queue
> + *   -enqueue to a queuing destination
> + */
> +
> +/**
> + * qbman_eq_desc_set_fq() - set the FQ for the enqueue command
> + * @d:    the enqueue descriptor
> + * @fqid: the id of the frame queue to be enqueued
> + */
> +void qbman_eq_desc_set_fq(struct qbman_eq_desc *d, u32 fqid)
> +{
> +	d->verb &= ~(1 << QB_ENQUEUE_CMD_TARGET_TYPE_SHIFT);
> +	d->tgtid = cpu_to_le32(fqid);
> +}
> +
> +/**
> + * qbman_eq_desc_set_qd() - Set Queuing Destination for the enqueue
> command
> + * @d:       the enqueue descriptor
> + * @qdid:    the id of the queuing destination to be enqueued
> + * @qd_bin:  the queuing destination bin
> + * @qd_prio: the queuing destination priority
> + */
> +void qbman_eq_desc_set_qd(struct qbman_eq_desc *d, u32 qdid,
> +			  u32 qd_bin, u32 qd_prio)
> +{
> +	d->verb |= 1 << QB_ENQUEUE_CMD_TARGET_TYPE_SHIFT;
> +	d->tgtid = cpu_to_le32(qdid);
> +	d->qdbin = cpu_to_le16(qd_bin);
> +	d->qpri = qd_prio;
> +}
> +
> +#define EQAR_IDX(eqar)     ((eqar) & 0x7)
> +#define EQAR_VB(eqar)      ((eqar) & 0x80)
> +#define EQAR_SUCCESS(eqar) ((eqar) & 0x100)
> +
> +/**
> + * qbman_swp_enqueue() - Issue an enqueue command
> + * @s:  the software portal used for enqueue
> + * @d:  the enqueue descriptor
> + * @fd: the frame descriptor to be enqueued
> + *
> + * Please note that 'fd' should only be NULL if the "action" of the
> + * descriptor is "orp_hole" or "orp_nesn".
> + *
> + * Return 0 for successful enqueue, -EBUSY if the EQCR is not ready.
> + */
> +int qbman_swp_enqueue(struct qbman_swp *s, const struct
> qbman_eq_desc *d,
> +		      const struct dpaa2_fd *fd)
> +{
> +	struct qbman_eq_desc *p;
> +	u32 eqar = qbman_read_register(s, QBMAN_CINH_SWP_EQAR);
> +
> +	if (!EQAR_SUCCESS(eqar))
> +		return -EBUSY;
> +
> +	p = qbman_get_cmd(s,
> QBMAN_CENA_SWP_EQCR(EQAR_IDX(eqar)));
> +	memcpy(&p->dca, &d->dca, 31);
> +	memcpy(&p->fd, fd, sizeof(*fd));
> +
> +	/* Set the verb byte, have to substitute in the valid-bit */
> +	dma_wmb();
> +	p->verb = d->verb | EQAR_VB(eqar);
> +
> +	return 0;
> +}
> +
> +/* Static (push) dequeue */
> +
> +/**
> + * qbman_swp_push_get() - Get the push dequeue setup
> + * @p:           the software portal object
> + * @channel_idx: the channel index to query
> + * @enabled:     returned boolean to show whether the push dequeue is
> enabled
> + *               for the given channel
> + */
> +void qbman_swp_push_get(struct qbman_swp *s, u8 channel_idx, int
> *enabled)
> +{
> +	u16 src = (s->sdq >> QB_SDQCR_SRC_SHIFT) &
> QB_SDQCR_SRC_MASK;
> +
> +	WARN_ON(channel_idx > 15);
> +	*enabled = src | (1 << channel_idx);
> +}
> +
> +/**
> + * qbman_swp_push_set() - Enable or disable push dequeue
> + * @p:           the software portal object
> + * @channel_idx: the channel index (0 to 15)
> + * @enable:      enable or disable push dequeue
> + */
> +void qbman_swp_push_set(struct qbman_swp *s, u8 channel_idx, int
> enable)
> +{
> +	u16 dqsrc;
> +
> +	WARN_ON(channel_idx > 15);
> +	if (enable)
> +		s->sdq |= 1 << channel_idx;
> +	else
> +		s->sdq &= ~(1 << channel_idx);
> +
> +	/* Read make the complete src map.  If no channels are enabled
> +	 * the SDQCR must be 0 or else QMan will assert errors
> +	 */
> +	dqsrc = (s->sdq >> QB_SDQCR_SRC_SHIFT) &
> QB_SDQCR_SRC_MASK;
> +	if (dqsrc != 0)
> +		qbman_write_register(s, QBMAN_CINH_SWP_SDQCR, s-
> >sdq);
> +	else
> +		qbman_write_register(s, QBMAN_CINH_SWP_SDQCR, 0);
> +}
> +
> +#define QB_VDQCR_VERB_DCT_SHIFT    0
> +#define QB_VDQCR_VERB_DT_SHIFT     2
> +#define QB_VDQCR_VERB_RLS_SHIFT    4
> +#define QB_VDQCR_VERB_WAE_SHIFT    5
> +
> +enum qb_pull_dt_e {
> +	qb_pull_dt_channel,
> +	qb_pull_dt_workqueue,
> +	qb_pull_dt_framequeue
> +};
> +
> +/**
> + * qbman_pull_desc_clear() - Clear the contents of a descriptor to
> + *                           default/starting state
> + * @d: the pull dequeue descriptor to be cleared
> + */
> +void qbman_pull_desc_clear(struct qbman_pull_desc *d)
> +{
> +	memset(d, 0, sizeof(*d));
> +}
> +
> +/**
> + * qbman_pull_desc_set_storage()- Set the pull dequeue storage
> + * @d:            the pull dequeue descriptor to be set
> + * @storage:      the pointer of the memory to store the dequeue result
> + * @storage_phys: the physical address of the storage memory
> + * @stash:        to indicate whether write allocate is enabled
> + *
> + * If not called, or if called with 'storage' as NULL, the result pull dequeues
> + * will produce results to DQRR. If 'storage' is non-NULL, then results are
> + * produced to the given memory location (using the DMA address which
> + * the caller provides in 'storage_phys'), and 'stash' controls whether or not
> + * those writes to main-memory express a cache-warming attribute.
> + */
> +void qbman_pull_desc_set_storage(struct qbman_pull_desc *d,
> +				 struct dpaa2_dq *storage,
> +				 dma_addr_t storage_phys,
> +				 int stash)
> +{
> +	/* save the virtual address */
> +	d->rsp_addr_virt = (u64)storage;
> +
> +	if (!storage) {
> +		d->verb &= ~(1 << QB_VDQCR_VERB_RLS_SHIFT);
> +		return;
> +	}
> +	d->verb |= 1 << QB_VDQCR_VERB_RLS_SHIFT;
> +	if (stash)
> +		d->verb |= 1 << QB_VDQCR_VERB_WAE_SHIFT;
> +	else
> +		d->verb &= ~(1 << QB_VDQCR_VERB_WAE_SHIFT);
> +
> +	d->rsp_addr = cpu_to_le64(storage_phys);
> +}
> +
> +/**
> + * qbman_pull_desc_set_numframes() - Set the number of frames to be
> dequeued
> + * @d:         the pull dequeue descriptor to be set
> + * @numframes: number of frames to be set, must be between 1 and 16,
> inclusive
> + */
> +void qbman_pull_desc_set_numframes(struct qbman_pull_desc *d, u8
> numframes)
> +{
> +	d->numf = numframes - 1;
> +}
> +
> +void qbman_pull_desc_set_token(struct qbman_pull_desc *d, u8 token)
> +{
> +	d->tok = token;
> +}
> +
> +/*
> + * Exactly one of the following descriptor "actions" should be set. (Calling
> any
> + * one of these will replace the effect of any prior call to one of these.)
> + * - pull dequeue from the given frame queue (FQ)
> + * - pull dequeue from any FQ in the given work queue (WQ)
> + * - pull dequeue from any FQ in any WQ in the given channel
> + */
> +
> +/**
> + * qbman_pull_desc_set_fq() - Set fqid from which the dequeue command
> dequeues
> + * @fqid: the frame queue index of the given FQ
> + */
> +void qbman_pull_desc_set_fq(struct qbman_pull_desc *d, u32 fqid)
> +{
> +	d->verb |= 1 << QB_VDQCR_VERB_DCT_SHIFT;
> +	d->verb |= qb_pull_dt_framequeue <<
> QB_VDQCR_VERB_DT_SHIFT;
> +	d->dq_src = cpu_to_le32(fqid);
> +}
> +
> +/**
> + * qbman_pull_desc_set_wq() - Set wqid from which the dequeue
> command dequeues
> + * @wqid: composed of channel id and wqid within the channel
> + * @dct:  the dequeue command type
> + */
> +void qbman_pull_desc_set_wq(struct qbman_pull_desc *d, u32 wqid,
> +			    enum qbman_pull_type_e dct)
> +{
> +	d->verb |= dct << QB_VDQCR_VERB_DCT_SHIFT;
> +	d->verb |= qb_pull_dt_workqueue << QB_VDQCR_VERB_DT_SHIFT;
> +	d->dq_src = cpu_to_le32(wqid);
> +}
> +
> +/**
> + * qbman_pull_desc_set_channel() - Set channelid from which the
> dequeue command
> + *                                 dequeues
> + * @chid: the channel id to be dequeued
> + * @dct:  the dequeue command type
> + */
> +void qbman_pull_desc_set_channel(struct qbman_pull_desc *d, u32 chid,
> +				 enum qbman_pull_type_e dct)
> +{
> +	d->verb |= dct << QB_VDQCR_VERB_DCT_SHIFT;
> +	d->verb |= qb_pull_dt_channel << QB_VDQCR_VERB_DT_SHIFT;
> +	d->dq_src = cpu_to_le32(chid);
> +}
> +
> +/**
> + * qbman_swp_pull() - Issue the pull dequeue command
> + * @s: the software portal object
> + * @d: the software portal descriptor which has been configured with
> + *     the set of qbman_pull_desc_set_*() calls
> + *
> + * Return 0 for success, and -EBUSY if the software portal is not ready
> + * to do pull dequeue.
> + */
> +int qbman_swp_pull(struct qbman_swp *s, struct qbman_pull_desc *d)
> +{
> +	struct qbman_pull_desc *p;
> +
> +	if (!atomic_dec_and_test(&s->vdq.available)) {
> +		atomic_inc(&s->vdq.available);
> +		return -EBUSY;
> +	}
> +	s->vdq.storage = (void *)d->rsp_addr_virt;
> +	d->tok = QMAN_DQ_TOKEN_VALID;
> +	p = qbman_get_cmd(s, QBMAN_CENA_SWP_VDQCR);
> +	*p = *d;
> +	dma_wmb();
> +
> +	/* Set the verb byte, have to substitute in the valid-bit */
> +	p->verb |= s->vdq.valid_bit;
> +	s->vdq.valid_bit ^= QB_VALID_BIT;
> +
> +	return 0;
> +}
> +
> +#define QMAN_DQRR_PI_MASK   0xf
> +
> +/**
> + * qbman_swp_dqrr_next() - Get an valid DQRR entry
> + * @s: the software portal object
> + *
> + * Return NULL if there are no unconsumed DQRR entries. Return a DQRR
> entry
> + * only once, so repeated calls can return a sequence of DQRR entries,
> without
> + * requiring they be consumed immediately or in any particular order.
> + */
> +const struct dpaa2_dq *qbman_swp_dqrr_next(struct qbman_swp *s)
> +{
> +	u32 verb;
> +	u32 response_verb;
> +	u32 flags;
> +	struct dpaa2_dq *p;
> +
> +	/* Before using valid-bit to detect if something is there, we have to
> +	 * handle the case of the DQRR reset bug...
> +	 */
> +	if (unlikely(s->dqrr.reset_bug)) {
> +		/*
> +		 * We pick up new entries by cache-inhibited producer index,
> +		 * which means that a non-coherent mapping would require
> us to
> +		 * invalidate and read *only* once that PI has indicated that
> +		 * there's an entry here. The first trip around the DQRR ring
> +		 * will be much less efficient than all subsequent trips around
> +		 * it...
> +		 */
> +		u8 pi = qbman_read_register(s, QBMAN_CINH_SWP_DQPI)
> &
> +			QMAN_DQRR_PI_MASK;
> +
> +		/* there are new entries if pi != next_idx */
> +		if (pi == s->dqrr.next_idx)
> +			return NULL;
> +
> +		/*
> +		 * if next_idx is/was the last ring index, and 'pi' is
> +		 * different, we can disable the workaround as all the ring
> +		 * entries have now been DMA'd to so valid-bit checking is
> +		 * repaired. Note: this logic needs to be based on next_idx
> +		 * (which increments one at a time), rather than on pi (which
> +		 * can burst and wrap-around between our snapshots of it).
> +		 */
> +		if (s->dqrr.next_idx == (s->dqrr.dqrr_size - 1)) {
> +			pr_debug("next_idx=%d, pi=%d, clear reset bug\n",
> +				 s->dqrr.next_idx, pi);
> +			s->dqrr.reset_bug = 0;
> +		}
> +		prefetch(qbman_get_cmd(s,
> +				       QBMAN_CENA_SWP_DQRR(s-
> >dqrr.next_idx)));
> +	}
> +
> +	p = qbman_get_cmd(s, QBMAN_CENA_SWP_DQRR(s-
> >dqrr.next_idx));
> +	verb = p->dq.verb;
> +
> +	/*
> +	 * If the valid-bit isn't of the expected polarity, nothing there. Note,
> +	 * in the DQRR reset bug workaround, we shouldn't need to skip
> these
> +	 * check, because we've already determined that a new entry is
> available
> +	 * and we've invalidated the cacheline before reading it, so the
> +	 * valid-bit behaviour is repaired and should tell us what we already
> +	 * knew from reading PI.
> +	 */
> +	if ((verb & QB_VALID_BIT) != s->dqrr.valid_bit) {
> +		prefetch(qbman_get_cmd(s,
> +				       QBMAN_CENA_SWP_DQRR(s-
> >dqrr.next_idx)));
> +		return NULL;
> +	}
> +	/*
> +	 * There's something there. Move "next_idx" attention to the next
> ring
> +	 * entry (and prefetch it) before returning what we found.
> +	 */
> +	s->dqrr.next_idx++;
> +	s->dqrr.next_idx &= s->dqrr.dqrr_size - 1; /* Wrap around */
> +	if (!s->dqrr.next_idx)
> +		s->dqrr.valid_bit ^= QB_VALID_BIT;
> +
> +	/*
> +	 * If this is the final response to a volatile dequeue command
> +	 * indicate that the vdq is available
> +	 */
> +	flags = p->dq.stat;
> +	response_verb = verb & QBMAN_RESULT_MASK;
> +	if ((response_verb == QBMAN_RESULT_DQ) &&
> +	    (flags & DPAA2_DQ_STAT_VOLATILE) &&
> +	    (flags & DPAA2_DQ_STAT_EXPIRED))
> +		atomic_inc(&s->vdq.available);
> +
> +	prefetch(qbman_get_cmd(s, QBMAN_CENA_SWP_DQRR(s-
> >dqrr.next_idx)));
> +
> +	return p;
> +}
> +
> +/**
> + * qbman_swp_dqrr_consume() -  Consume DQRR entries previously
> returned from
> + *                             qbman_swp_dqrr_next().
> + * @s: the software portal object
> + * @dq: the DQRR entry to be consumed
> + */
> +void qbman_swp_dqrr_consume(struct qbman_swp *s, const struct
> dpaa2_dq *dq)
> +{
> +	qbman_write_register(s, QBMAN_CINH_SWP_DCAP,
> QBMAN_IDX_FROM_DQRR(dq));
> +}
> +
> +/**
> + * qbman_result_has_new_result() - Check and get the dequeue response
> from the
> + *                                 dq storage memory set in pull dequeue command
> + * @s: the software portal object
> + * @dq: the dequeue result read from the memory
> + *
> + * Return 1 for getting a valid dequeue result, or 0 for not getting a valid
> + * dequeue result.
> + *
> + * Only used for user-provided storage of dequeue results, not DQRR. For
> + * efficiency purposes, the driver will perform any required endianness
> + * conversion to ensure that the user's dequeue result storage is in host-
> endian
> + * format. As such, once the user has called
> qbman_result_has_new_result() and
> + * been returned a valid dequeue result, they should not call it again on
> + * the same memory location (except of course if another dequeue
> command has
> + * been executed to produce a new result to that location).
> + */
> +int qbman_result_has_new_result(struct qbman_swp *s, const struct
> dpaa2_dq *dq)
> +{
> +	if (dq->dq.tok != QMAN_DQ_TOKEN_VALID)
> +		return 0;
> +
> +	/*
> +	 * Set token to be 0 so we will detect change back to 1
> +	 * next time the looping is traversed. Const is cast away here
> +	 * as we want users to treat the dequeue responses as read only.
> +	 */
> +	((struct dpaa2_dq *)dq)->dq.tok = 0;
> +
> +	/*
> +	 * Determine whether VDQCR is available based on whether the
> +	 * current result is sitting in the first storage location of
> +	 * the busy command.
> +	 */
> +	if (s->vdq.storage == dq) {
> +		s->vdq.storage = NULL;
> +		atomic_inc(&s->vdq.available);
> +	}
> +
> +	return 1;
> +}
> +
> +/**
> + * qbman_release_desc_clear() - Clear the contents of a descriptor to
> + *                              default/starting state.
> + */
> +void qbman_release_desc_clear(struct qbman_release_desc *d)
> +{
> +	memset(d, 0, sizeof(*d));
> +	d->verb = 1 << 5; /* Release Command Valid */
> +}
> +
> +/**
> + * qbman_release_desc_set_bpid() - Set the ID of the buffer pool to
> release to
> + */
> +void qbman_release_desc_set_bpid(struct qbman_release_desc *d, u16
> bpid)
> +{
> +	d->bpid = cpu_to_le16(bpid);
> +}
> +
> +/**
> + * qbman_release_desc_set_rcdi() - Determines whether or not the
> portal's RCDI
> + * interrupt source should be asserted after the release command is
> completed.
> + */
> +void qbman_release_desc_set_rcdi(struct qbman_release_desc *d, int
> enable)
> +{
> +	if (enable)
> +		d->verb |= 1 << 6;
> +	else
> +		d->verb &= ~(1 << 6);

Can we avoid hardcoded values, or a comment?

> +}
> +
> +#define RAR_IDX(rar)     ((rar) & 0x7)
> +#define RAR_VB(rar)      ((rar) & 0x80)
> +#define RAR_SUCCESS(rar) ((rar) & 0x100)
> +
> +/**
> + * qbman_swp_release() - Issue a buffer release command
> + * @s:           the software portal object
> + * @d:           the release descriptor
> + * @buffers:     a pointer pointing to the buffer address to be released
> + * @num_buffers: number of buffers to be released,  must be less than 8
> + *
> + * Return 0 for success, -EBUSY if the release command ring is not ready.
> + */
> +int qbman_swp_release(struct qbman_swp *s, const struct
> qbman_release_desc *d,
> +		      const u64 *buffers, unsigned int num_buffers)
> +{
> +	int i;
> +	struct qbman_release_desc *p;
> +	u32 rar;
> +
> +	if (!num_buffers || (num_buffers > 7))
> +		return -EINVAL;
> +
> +	rar = qbman_read_register(s, QBMAN_CINH_SWP_RAR);
> +	if (!RAR_SUCCESS(rar))
> +		return -EBUSY;
> +
> +	/* Start the release command */
> +	p = qbman_get_cmd(s, QBMAN_CENA_SWP_RCR(RAR_IDX(rar)));
> +	/* Copy the caller's buffer pointers to the command */
> +	for (i = 0; i < num_buffers; i++)
> +		p->buf[i] = cpu_to_le64(buffers[i]);
> +	p->bpid = d->bpid;
> +
> +	/*
> +	 * Set the verb byte, have to substitute in the valid-bit and the
> number
> +	 * of buffers.
> +	 */
> +	dma_wmb();
> +	p->verb = d->verb | RAR_VB(rar) | num_buffers;
> +
> +	return 0;
> +}
> +
> +struct qbman_acquire_desc {
> +	u8 verb;
> +	u8 reserved;
> +	u16 bpid;
> +	u8 num;
> +	u8 reserved2[59];
> +};
> +
> +struct qbman_acquire_rslt {
> +	u8 verb;
> +	u8 rslt;
> +	u16 reserved;
> +	u8 num;
> +	u8 reserved2[3];
> +	u64 buf[7];
> +};
> +
> +/**
> + * qbman_swp_acquire() - Issue a buffer acquire command
> + * @s:           the software portal object
> + * @bpid:        the buffer pool index
> + * @buffers:     a pointer pointing to the acquired buffer addresses
> + * @num_buffers: number of buffers to be acquired, must be less than 8
> + *
> + * Return 0 for success, or negative error code if the acquire command
> + * fails.
> + */
> +int qbman_swp_acquire(struct qbman_swp *s, u16 bpid, u64 *buffers,
> +		      unsigned int num_buffers)
> +{
> +	struct qbman_acquire_desc *p;
> +	struct qbman_acquire_rslt *r;
> +	int i;
> +
> +	if (!num_buffers || (num_buffers > 7))
> +		return -EINVAL;
> +
> +	/* Start the management command */
> +	p = qbman_swp_mc_start(s);
> +
> +	if (!p)
> +		return -EBUSY;
> +
> +	/* Encode the caller-provided attributes */
> +	p->bpid = cpu_to_le16(bpid);
> +	p->num = num_buffers;
> +
> +	/* Complete the management command */
> +	r = qbman_swp_mc_complete(s, p, QBMAN_MC_ACQUIRE);
> +	if (unlikely(!r)) {
> +		pr_err("qbman: acquire from BPID %d failed, no
> response\n",
> +		       bpid);
> +		return -EIO;
> +	}
> +
> +	/* Decode the outcome */
> +	WARN_ON((r->verb & 0x7f) != QBMAN_MC_ACQUIRE);
> +
> +	/* Determine success or failure */
> +	if (unlikely(r->rslt != QBMAN_MC_RSLT_OK)) {
> +		pr_err("qbman: acquire from BPID 0x%x failed,
> code=0x%02x\n",
> +		       bpid, r->rslt);
> +		return -EIO;
> +	}
> +
> +	WARN_ON(r->num > num_buffers);
> +
> +	/* Copy the acquired buffers to the caller's array */
> +	for (i = 0; i < r->num; i++)
> +		buffers[i] = le64_to_cpu(r->buf[i]);
> +
> +	return (int)r->num;
> +}
> +
> +struct qbman_alt_fq_state_desc {
> +	u8 verb;
> +	u8 reserved[3];
> +	u32 fqid;
> +	u8 reserved2[56];
> +};
> +
> +struct qbman_alt_fq_state_rslt {
> +	u8 verb;
> +	u8 rslt;
> +	u8 reserved[62];
> +};
> +
> +#define ALT_FQ_FQID_MASK 0x00FFFFFF
> +
> +int qbman_swp_alt_fq_state(struct qbman_swp *s, u32 fqid,
> +			   u8 alt_fq_verb)
> +{
> +	struct qbman_alt_fq_state_desc *p;
> +	struct qbman_alt_fq_state_rslt *r;
> +
> +	/* Start the management command */
> +	p = qbman_swp_mc_start(s);
> +	if (!p)
> +		return -EBUSY;
> +
> +	p->fqid = cpu_to_le32(fqid) & ALT_FQ_FQID_MASK;
> +
> +	/* Complete the management command */
> +	r = qbman_swp_mc_complete(s, p, alt_fq_verb);
> +	if (unlikely(!r)) {
> +		pr_err("qbman: mgmt cmd failed, no response
> (verb=0x%x)\n",
> +		       alt_fq_verb);
> +		return -EIO;
> +	}
> +
> +	/* Decode the outcome */
> +	WARN_ON(r->verb != alt_fq_verb);
> +
> +	/* Determine success or failure */
> +	if (unlikely(r->rslt != QBMAN_MC_RSLT_OK)) {
> +		pr_err("qbman: ALT FQID %d failed: verb = 0x%08x code =
> 0x%02x\n",
> +		       fqid, r->verb, r->rslt);
> +		return -EIO;
> +	}
> +
> +	return 0;
> +}
> +
> +struct qbman_cdan_ctrl_desc {
> +	u8 verb;
> +	u8 reserved;
> +	u16 ch;
> +	u8 we;
> +	u8 ctrl;
> +	u16 reserved2;
> +	u64 cdan_ctx;
> +	u8 reserved3[48];
> +
> +};
> +
> +struct qbman_cdan_ctrl_rslt {
> +	u8 verb;
> +	u8 rslt;
> +	u16 ch;
> +	u8 reserved[60];
> +};
> +
> +int qbman_swp_CDAN_set(struct qbman_swp *s, u16 channelid,
> +		       u8 we_mask, u8 cdan_en,
> +		       u64 ctx)
> +{
> +	struct qbman_cdan_ctrl_desc *p = NULL;
> +	struct qbman_cdan_ctrl_rslt *r = NULL;
> +
> +	/* Start the management command */
> +	p = qbman_swp_mc_start(s);
> +	if (!p)
> +		return -EBUSY;
> +
> +	/* Encode the caller-provided attributes */
> +	p->verb = 0;
> +	p->ch = cpu_to_le16(channelid);
> +	p->we = we_mask;
> +	if (cdan_en)
> +		p->ctrl = 1;
> +	else
> +		p->ctrl = 0;
> +	p->cdan_ctx = cpu_to_le64(ctx);
> +
> +	/* Complete the management command */
> +	r = qbman_swp_mc_complete(s, p,
> QBMAN_WQCHAN_CONFIGURE);
> +	if (unlikely(!r)) {
> +		pr_err("qbman: wqchan config failed, no response\n");
> +		return -EIO;
> +	}
> +
> +	WARN_ON((r->verb & 0x7f) != QBMAN_WQCHAN_CONFIGURE);
> +
> +	/* Determine success or failure */
> +	if (unlikely(r->rslt != QBMAN_MC_RSLT_OK)) {
> +		pr_err("qbman: CDAN cQID %d failed: code = 0x%02x\n",
> +		       channelid, r->rslt);
> +		return -EIO;
> +	}
> +
> +	return 0;
> +}
> diff --git a/drivers/staging/fsl-mc/bus/dpio/qbman-portal.h
> b/drivers/staging/fsl-mc/bus/dpio/qbman-portal.h
> new file mode 100644
> index 0000000..f0f21aa3
> --- /dev/null
> +++ b/drivers/staging/fsl-mc/bus/dpio/qbman-portal.h
> @@ -0,0 +1,469 @@
> +/*
> + * Copyright (C) 2014-2016 Freescale Semiconductor, Inc.
> + * Copyright 2016 NXP
> + *
> + * Redistribution and use in source and binary forms, with or without
> + * modification, are permitted provided that the following conditions are
> met:
> + *     * Redistributions of source code must retain the above copyright
> + *       notice, this list of conditions and the following disclaimer.
> + *     * Redistributions in binary form must reproduce the above copyright
> + *       notice, this list of conditions and the following disclaimer in the
> + *       documentation and/or other materials provided with the distribution.
> + *     * Neither the name of Freescale Semiconductor nor the
> + *       names of its contributors may be used to endorse or promote
> products
> + *       derived from this software without specific prior written permission.
> + *
> + * ALTERNATIVELY, this software may be distributed under the terms of the
> + * GNU General Public License ("GPL") as published by the Free Software
> + * Foundation, either version 2 of that License or (at your option) any
> + * later version.
> + *
> + * THIS SOFTWARE IS PROVIDED BY Freescale Semiconductor ``AS IS'' AND
> ANY
> + * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
> THE IMPLIED
> + * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
> PURPOSE ARE
> + * DISCLAIMED. IN NO EVENT SHALL Freescale Semiconductor BE LIABLE FOR
> ANY
> + * DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
> CONSEQUENTIAL DAMAGES
> + * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE
> GOODS OR SERVICES;
> + * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> HOWEVER CAUSED AND
> + * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
> OR TORT
> + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF
> THE USE OF THIS
> + * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
> + */
> +#ifndef __FSL_QBMAN_PORTAL_H
> +#define __FSL_QBMAN_PORTAL_H
> +
> +#include "../../include/dpaa2-fd.h"
> +
> +struct dpaa2_dq;
> +struct qbman_swp;
> +
> +/* qbman software portal descriptor structure */
> +struct qbman_swp_desc {
> +	void *cena_bar; /* Cache-enabled portal base address */
> +	void *cinh_bar; /* Cache-inhibited portal base address */
> +	u32 qman_version;
> +};
> +
> +#define QBMAN_SWP_INTERRUPT_EQRI 0x01
> +#define QBMAN_SWP_INTERRUPT_EQDI 0x02
> +#define QBMAN_SWP_INTERRUPT_DQRI 0x04
> +#define QBMAN_SWP_INTERRUPT_RCRI 0x08
> +#define QBMAN_SWP_INTERRUPT_RCDI 0x10
> +#define QBMAN_SWP_INTERRUPT_VDCI 0x20
> +
> +/* the structure for pull dequeue descriptor */
> +struct qbman_pull_desc {
> +	u8 verb;
> +	u8 numf;
> +	u8 tok;
> +	u8 reserved;
> +	u32 dq_src;
> +	u64 rsp_addr;
> +	u64 rsp_addr_virt;
> +	u8 padding[40];
> +};
> +
> +enum qbman_pull_type_e {
> +	/* dequeue with priority precedence, respect intra-class scheduling
> */
> +	qbman_pull_type_prio = 1,
> +	/* dequeue with active FQ precedence, respect ICS */
> +	qbman_pull_type_active,
> +	/* dequeue with active FQ precedence, no ICS */
> +	qbman_pull_type_active_noics
> +};
> +
> +/* Definitions for parsing dequeue entries */
> +#define QBMAN_RESULT_MASK      0x7f
> +#define QBMAN_RESULT_DQ        0x60
> +#define QBMAN_RESULT_FQRN      0x21
> +#define QBMAN_RESULT_FQRNI     0x22
> +#define QBMAN_RESULT_FQPN      0x24
> +#define QBMAN_RESULT_FQDAN     0x25
> +#define QBMAN_RESULT_CDAN      0x26
> +#define QBMAN_RESULT_CSCN_MEM  0x27
> +#define QBMAN_RESULT_CGCU      0x28
> +#define QBMAN_RESULT_BPSCN     0x29
> +#define QBMAN_RESULT_CSCN_WQ   0x2a
> +
> +/* QBMan FQ management command codes */
> +#define QBMAN_FQ_SCHEDULE	0x48
> +#define QBMAN_FQ_FORCE		0x49
> +#define QBMAN_FQ_XON		0x4d
> +#define QBMAN_FQ_XOFF		0x4e
> +
> +/* structure of enqueue descriptor */
> +struct qbman_eq_desc {
> +	u8 verb;
> +	u8 dca;
> +	u16 seqnum;
> +	u16 orpid;
> +	u16 reserved1;
> +	u32 tgtid;
> +	u32 tag;
> +	u16 qdbin;
> +	u8 qpri;
> +	u8 reserved[3];
> +	u8 wae;
> +	u8 rspid;
> +	u64 rsp_addr;
> +	u8 fd[32];
> +};
> +
> +/* buffer release descriptor */
> +struct qbman_release_desc {
> +	u8 verb;
> +	u8 reserved;
> +	u16 bpid;
> +	u32 reserved2;
> +	u64 buf[7];
> +};
> +
> +/* Management command result codes */
> +#define QBMAN_MC_RSLT_OK      0xf0
> +
> +#define CODE_CDAN_WE_EN    0x1
> +#define CODE_CDAN_WE_CTX   0x4
> +
> +/* portal data structure */
> +struct qbman_swp {
> +	const struct qbman_swp_desc *desc;
> +	void __iomem *addr_cena;
> +	void __iomem *addr_cinh;
> +
> +	/* Management commands */
> +	struct {
> +		u32 valid_bit; /* 0x00 or 0x80 */
> +	} mc;
> +
> +	/* Push dequeues */
> +	u32 sdq;
> +
> +	/* Volatile dequeues */
> +	struct {
> +		atomic_t available; /* indicates if a command can be sent */
> +		u32 valid_bit; /* 0x00 or 0x80 */
> +		struct dpaa2_dq *storage; /* NULL if DQRR */
> +	} vdq;
> +
> +	/* DQRR */
> +	struct {
> +		u32 next_idx;
> +		u32 valid_bit;
> +		u8 dqrr_size;
> +		int reset_bug; /* indicates dqrr reset workaround is needed
> */
> +	} dqrr;
> +};
> +
> +struct qbman_swp *qbman_swp_init(const struct qbman_swp_desc *d);
> +void qbman_swp_finish(struct qbman_swp *p);
> +u32 qbman_swp_interrupt_read_status(struct qbman_swp *p);
> +void qbman_swp_interrupt_clear_status(struct qbman_swp *p, u32 mask);
> +u32 qbman_swp_interrupt_get_trigger(struct qbman_swp *p);
> +void qbman_swp_interrupt_set_trigger(struct qbman_swp *p, u32 mask);
> +int qbman_swp_interrupt_get_inhibit(struct qbman_swp *p);
> +void qbman_swp_interrupt_set_inhibit(struct qbman_swp *p, int inhibit);
> +
> +void qbman_swp_push_get(struct qbman_swp *p, u8 channel_idx, int
> *enabled);
> +void qbman_swp_push_set(struct qbman_swp *p, u8 channel_idx, int
> enable);
> +
> +void qbman_pull_desc_clear(struct qbman_pull_desc *d);
> +void qbman_pull_desc_set_storage(struct qbman_pull_desc *d,
> +				 struct dpaa2_dq *storage,
> +				 dma_addr_t storage_phys,
> +				 int stash);
> +void qbman_pull_desc_set_numframes(struct qbman_pull_desc *d, u8
> numframes);
> +void qbman_pull_desc_set_fq(struct qbman_pull_desc *d, u32 fqid);
> +void qbman_pull_desc_set_wq(struct qbman_pull_desc *d, u32 wqid,
> +			    enum qbman_pull_type_e dct);
> +void qbman_pull_desc_set_channel(struct qbman_pull_desc *d, u32 chid,
> +				 enum qbman_pull_type_e dct);
> +
> +int qbman_swp_pull(struct qbman_swp *p, struct qbman_pull_desc *d);
> +
> +const struct dpaa2_dq *qbman_swp_dqrr_next(struct qbman_swp *s);
> +void qbman_swp_dqrr_consume(struct qbman_swp *s, const struct
> dpaa2_dq *dq);
> +
> +int qbman_result_has_new_result(struct qbman_swp *p, const struct
> dpaa2_dq *dq);
> +
> +void qbman_eq_desc_clear(struct qbman_eq_desc *d);
> +void qbman_eq_desc_set_no_orp(struct qbman_eq_desc *d, int
> respond_success);
> +void qbman_eq_desc_set_token(struct qbman_eq_desc *d, u8 token);
> +void qbman_eq_desc_set_fq(struct qbman_eq_desc *d, u32 fqid);
> +void qbman_eq_desc_set_qd(struct qbman_eq_desc *d, u32 qdid,
> +			  u32 qd_bin, u32 qd_prio);
> +
> +int qbman_swp_enqueue(struct qbman_swp *p, const struct
> qbman_eq_desc *d,
> +		      const struct dpaa2_fd *fd);
> +
> +void qbman_release_desc_clear(struct qbman_release_desc *d);
> +void qbman_release_desc_set_bpid(struct qbman_release_desc *d, u16
> bpid);
> +void qbman_release_desc_set_rcdi(struct qbman_release_desc *d, int
> enable);
> +
> +int qbman_swp_release(struct qbman_swp *s, const struct
> qbman_release_desc *d,
> +		      const u64 *buffers, unsigned int num_buffers);
> +int qbman_swp_acquire(struct qbman_swp *s, u16 bpid, u64 *buffers,
> +		      unsigned int num_buffers);
> +int qbman_swp_alt_fq_state(struct qbman_swp *s, u32 fqid,
> +			   u8 alt_fq_verb);
> +int qbman_swp_CDAN_set(struct qbman_swp *s, u16 channelid,
> +		       u8 we_mask, u8 cdan_en,
> +		       u64 ctx);
> +
> +void *qbman_swp_mc_start(struct qbman_swp *p);
> +void qbman_swp_mc_submit(struct qbman_swp *p, void *cmd, u8
> cmd_verb);
> +void *qbman_swp_mc_result(struct qbman_swp *p);
> +
> +/**
> + * qbman_result_is_DQ() - check if the dequeue result is a dequeue
> response
> + * @dq: the dequeue result to be checked
> + *
> + * DQRR entries may contain non-dequeue results, ie. notifications
> + */
> +static inline int qbman_result_is_DQ(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_DQ);
> +}
> +
> +/**
> + * qbman_result_is_SCN() - Check the dequeue result is notification or not
> + * @dq: the dequeue result to be checked
> + *
> + */
> +static inline int qbman_result_is_SCN(const struct dpaa2_dq *dq)
> +{
> +	return !qbman_result_is_DQ(dq);
> +}
> +
> +/* FQ Data Availability */
> +static inline int qbman_result_is_FQDAN(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_FQDAN);
> +}
> +
> +/* Channel Data Availability */
> +static inline int qbman_result_is_CDAN(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_CDAN);
> +}
> +
> +/* Congestion State Change */
> +static inline int qbman_result_is_CSCN(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_CSCN_WQ);
> +}
> +
> +/* Buffer Pool State Change */
> +static inline int qbman_result_is_BPSCN(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_BPSCN);
> +}
> +
> +/* Congestion Group Count Update */
> +static inline int qbman_result_is_CGCU(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_CGCU);
> +}
> +
> +/* Retirement */
> +static inline int qbman_result_is_FQRN(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_FQRN);
> +}
> +
> +/* Retirement Immediate */
> +static inline int qbman_result_is_FQRNI(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_FQRNI);
> +}
> +
> + /* Park */
> +static inline int qbman_result_is_FQPN(const struct dpaa2_dq *dq)
> +{
> +	return ((dq->dq.verb & QBMAN_RESULT_MASK) ==
> QBMAN_RESULT_FQPN);
> +}
> +
> +/**
> + * qbman_result_SCN_state() - Get the state field in State-change
> notification
> + */
> +static inline u8 qbman_result_SCN_state(const struct dpaa2_dq *scn)
> +{
> +	return scn->scn.state;
> +}
> +
> +#define SCN_RID_MASK 0x00FFFFFF
> +
> +/**
> + * qbman_result_SCN_rid() - Get the resource id in State-change
> notification
> + */
> +static inline u32 qbman_result_SCN_rid(const struct dpaa2_dq *scn)
> +{
> +	return le32_to_cpu(scn->scn.rid_tok) & SCN_RID_MASK;
> +}
> +
> +/**
> + * qbman_result_SCN_ctx() - Get the context data in State-change
> notification
> + */
> +static inline u64 qbman_result_SCN_ctx(const struct dpaa2_dq *scn)
> +{
> +	return le64_to_cpu(scn->scn.ctx);
> +}
> +
> +/**
> + * qbman_swp_fq_schedule() - Move the fq to the scheduled state
> + * @s:    the software portal object
> + * @fqid: the index of frame queue to be scheduled
> + *
> + * There are a couple of different ways that a FQ can end up parked state,
> + * This schedules it.
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_fq_schedule(struct qbman_swp *s, u32 fqid)
> +{
> +	return qbman_swp_alt_fq_state(s, fqid, QBMAN_FQ_SCHEDULE);
> +}
> +
> +/**
> + * qbman_swp_fq_force() - Force the FQ to fully scheduled state
> + * @s:    the software portal object
> + * @fqid: the index of frame queue to be forced
> + *
> + * Force eligible will force a tentatively-scheduled FQ to be fully-scheduled
> + * and thus be available for selection by any channel-dequeuing behaviour
> (push
> + * or pull). If the FQ is subsequently "dequeued" from the channel and is
> still
> + * empty at the time this happens, the resulting dq_entry will have no FD.
> + * (qbman_result_DQ_fd() will return NULL.)
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_fq_force(struct qbman_swp *s, u32 fqid)
> +{
> +	return qbman_swp_alt_fq_state(s, fqid, QBMAN_FQ_FORCE);
> +}
> +
> +/**
> + * qbman_swp_fq_xon() - sets FQ flow-control to XON
> + * @s:    the software portal object
> + * @fqid: the index of frame queue
> + *
> + * This setting doesn't affect enqueues to the FQ, just dequeues.
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_fq_xon(struct qbman_swp *s, u32 fqid)
> +{
> +	return qbman_swp_alt_fq_state(s, fqid, QBMAN_FQ_XON);
> +}
> +
> +/**
> + * qbman_swp_fq_xoff() - sets FQ flow-control to XOFF
> + * @s:    the software portal object
> + * @fqid: the index of frame queue
> + *
> + * This setting doesn't affect enqueues to the FQ, just dequeues.
> + * XOFF FQs will remain in the tenatively-scheduled state, even when
> + * non-empty, meaning they won't be selected for scheduled dequeuing.
> + * If a FQ is changed to XOFF after it had already become truly-scheduled
> + * to a channel, and a pull dequeue of that channel occurs that selects
> + * that FQ for dequeuing, then the resulting dq_entry will have no FD.
> + * (qbman_result_DQ_fd() will return NULL.)
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_fq_xoff(struct qbman_swp *s, u32 fqid)
> +{
> +	return qbman_swp_alt_fq_state(s, fqid, QBMAN_FQ_XOFF);
> +}
> +
> +/* If the user has been allocated a channel object that is going to generate
> + * CDANs to another channel, then the qbman_swp_CDAN* functions will
> be
> + * necessary.
> + *
> + * CDAN-enabled channels only generate a single CDAN notification, after
> which
> + * they need to be reenabled before they'll generate another. The idea is
> + * that pull dequeuing will occur in reaction to the CDAN, followed by a
> + * reenable step. Each function generates a distinct command to hardware,
> so a
> + * combination function is provided if the user wishes to modify the
> "context"
> + * (which shows up in each CDAN message) each time they reenable, as a
> single
> + * command to hardware.
> + */
> +
> +/**
> + * qbman_swp_CDAN_set_context() - Set CDAN context
> + * @s:         the software portal object
> + * @channelid: the channel index
> + * @ctx:       the context to be set in CDAN
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_CDAN_set_context(struct qbman_swp *s, u16
> channelid,
> +					     u64 ctx)
> +{
> +	return qbman_swp_CDAN_set(s, channelid,
> +				  CODE_CDAN_WE_CTX,
> +				  0, ctx);
> +}
> +
> +/**
> + * qbman_swp_CDAN_enable() - Enable CDAN for the channel
> + * @s:         the software portal object
> + * @channelid: the index of the channel to generate CDAN
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_CDAN_enable(struct qbman_swp *s, u16
> channelid)
> +{
> +	return qbman_swp_CDAN_set(s, channelid,
> +				  CODE_CDAN_WE_EN,
> +				  1, 0);
> +}
> +
> +/**
> + * qbman_swp_CDAN_disable() - disable CDAN for the channel
> + * @s:         the software portal object
> + * @channelid: the index of the channel to generate CDAN
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_CDAN_disable(struct qbman_swp *s, u16
> channelid)
> +{
> +	return qbman_swp_CDAN_set(s, channelid,
> +				  CODE_CDAN_WE_EN,
> +				  0, 0);
> +}
> +
> +/**
> + * qbman_swp_CDAN_set_context_enable() - Set CDAN contest and
> enable CDAN
> + * @s:         the software portal object
> + * @channelid: the index of the channel to generate CDAN
> + * @ctx:i      the context set in CDAN
> + *
> + * Return 0 for success, or negative error code for failure.
> + */
> +static inline int qbman_swp_CDAN_set_context_enable(struct qbman_swp
> *s,
> +						    u16 channelid,
> +						    u64 ctx)
> +{
> +	return qbman_swp_CDAN_set(s, channelid,
> +				  CODE_CDAN_WE_EN |
> CODE_CDAN_WE_CTX,
> +				  1, ctx);
> +}
> +
> +/* Wraps up submit + poll-for-result */
> +static inline void *qbman_swp_mc_complete(struct qbman_swp *swp,
> void *cmd,
> +					  u8 cmd_verb)
> +{
> +	int loopvar = 1000;
> +
> +	qbman_swp_mc_submit(swp, cmd, cmd_verb);
> +
> +	do {
> +		cmd = qbman_swp_mc_result(swp);
> +	} while (cmd == NULL && loopvar--);
> +
> +	WARN_ON(!loopvar);
> +
> +	return cmd;
> +}
> +
> +#endif /* __FSL_QBMAN_PORTAL_H */
> --
> 1.9.0
> 
> _______________________________________________
> upstream-release mailing list
> upstream-release@...ux.freescale.net
> http://linux.freescale.net/mailman/listinfo/upstream-release

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ