Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <0d78c8b7-848d-4e7e-b009-c9930e9e0c02@kernel.org>
Date:   Tue, 8 Aug 2023 13:54:22 +0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.14.0
Subject: Re: [PATCH v2 4/5] net: ti: icssg-prueth: add packet timestamping and
 ptp support
Content-Language: en-US
To:     MD Danish Anwar <danishanwar@ti.com>,
        Randy Dunlap <rdunlap@infradead.org>,
        Simon Horman <simon.horman@corigine.com>,
        Vignesh Raghavendra <vigneshr@ti.com>,
        Andrew Lunn <andrew@lunn.ch>,
        Richard Cochran <richardcochran@gmail.com>,
        Conor Dooley <conor+dt@kernel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@linaro.org>,
        Rob Herring <robh+dt@kernel.org>,
        Paolo Abeni <pabeni@redhat.com>,
        Jakub Kicinski <kuba@kernel.org>,
        Eric Dumazet <edumazet@google.com>,
        "David S. Miller" <davem@davemloft.net>
Cc:     nm@ti.com, srk@ti.com, linux-kernel@vger.kernel.org,
        devicetree@vger.kernel.org, netdev@vger.kernel.org,
        linux-omap@vger.kernel.org, linux-arm-kernel@lists.infradead.org
References: <20230807110048.2611456-1-danishanwar@ti.com>
 <20230807110048.2611456-5-danishanwar@ti.com>
From:   Roger Quadros <rogerq@kernel.org>
In-Reply-To: <20230807110048.2611456-5-danishanwar@ti.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi Danish,

On 07/08/2023 14:00, MD Danish Anwar wrote:
> From: Roger Quadros <rogerq@ti.com>
> 
> Add packet timestamping TS and PTP PHC clock support.
> 
> For AM65x and AM64x:
>  - IEP1 is not used
>  - IEP0 is configured in shadow mode with 1ms cycle and shared between
> Linux and FW. It provides time and TS in number cycles, so special
> conversation in ns is required.
>  - IEP0 shared between PRUeth ports.
>  - IEP0 supports PPS, periodic output.
>  - IEP0 settime() and enabling PPS required FW interraction.
>  - RX TS provided with each packet in CPPI5 descriptor.
>  - TX TS returned through separate ICSSG hw queues for each port. TX TS
> readiness is signaled by INTC IRQ. Only one packet at time can be requested
> for TX TS.
> 
> Signed-off-by: Roger Quadros <rogerq@ti.com>
> Co-developed-by: Grygorii Strashko <grygorii.strashko@ti.com>
> Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
> Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
> Signed-off-by: MD Danish Anwar <danishanwar@ti.com>
> ---
>  drivers/net/ethernet/ti/icssg/icss_iep.c      |   2 +-
>  drivers/net/ethernet/ti/icssg/icss_iep.h      |   3 +-
>  drivers/net/ethernet/ti/icssg/icssg_ethtool.c |  21 +
>  drivers/net/ethernet/ti/icssg/icssg_prueth.c  | 416 +++++++++++++++++-
>  drivers/net/ethernet/ti/icssg/icssg_prueth.h  |  28 +-
>  5 files changed, 463 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/net/ethernet/ti/icssg/icss_iep.c b/drivers/net/ethernet/ti/icssg/icss_iep.c
> index cc0ee113a2c5..455c803dea36 100644
> --- a/drivers/net/ethernet/ti/icssg/icss_iep.c
> +++ b/drivers/net/ethernet/ti/icssg/icss_iep.c
> @@ -224,7 +224,7 @@ static u64 icss_iep_gettime(struct icss_iep *iep,
>  	unsigned long flags;
>  
>  	if (iep->ops && iep->ops->gettime)
> -		return iep->ops->gettime(iep->clockops_data);
> +		return iep->ops->gettime(iep->clockops_data, sts);
>  
>  	/* use local_irq_x() to make it work for both RT/non-RT */
>  	local_irq_save(flags);
> diff --git a/drivers/net/ethernet/ti/icssg/icss_iep.h b/drivers/net/ethernet/ti/icssg/icss_iep.h
> index 4f9d4b6bb5d1..9c7f4d0a0916 100644
> --- a/drivers/net/ethernet/ti/icssg/icss_iep.h
> +++ b/drivers/net/ethernet/ti/icssg/icss_iep.h
> @@ -13,12 +13,13 @@
>  #include <linux/regmap.h>
>  
>  struct icss_iep;
> +extern const struct icss_iep_clockops prueth_iep_clockops;

Why do you need to do this?

>  
>  /* Firmware specific clock operations */
>  struct icss_iep_clockops {
>  	void (*settime)(void *clockops_data, u64 ns);
>  	void (*adjtime)(void *clockops_data, s64 delta);
> -	u64 (*gettime)(void *clockops_data);
> +	u64 (*gettime)(void *clockops_data, struct ptp_system_timestamp *sts);
>  	int (*perout_enable)(void *clockops_data,
>  			     struct ptp_perout_request *req, int on,
>  			     u64 *cmp);


Can we please squash all the above IEP driver changes in the patch that introduces IEP driver?
So this patch only deals with icssg_prueth driver.

> diff --git a/drivers/net/ethernet/ti/icssg/icssg_ethtool.c b/drivers/net/ethernet/ti/icssg/icssg_ethtool.c
> index 02c312f01d10..a27ec1dcc8d5 100644
> --- a/drivers/net/ethernet/ti/icssg/icssg_ethtool.c
> +++ b/drivers/net/ethernet/ti/icssg/icssg_ethtool.c
> @@ -109,6 +109,26 @@ static void emac_get_ethtool_stats(struct net_device *ndev,
>  			*(data++) = emac->stats[i];
>  }
>  
> +static int emac_get_ts_info(struct net_device *ndev,
> +			    struct ethtool_ts_info *info)
> +{
> +	struct prueth_emac *emac = netdev_priv(ndev);
> +
> +	info->so_timestamping =
> +		SOF_TIMESTAMPING_TX_HARDWARE |
> +		SOF_TIMESTAMPING_TX_SOFTWARE |
> +		SOF_TIMESTAMPING_RX_HARDWARE |
> +		SOF_TIMESTAMPING_RX_SOFTWARE |
> +		SOF_TIMESTAMPING_SOFTWARE |
> +		SOF_TIMESTAMPING_RAW_HARDWARE;
> +
> +	info->phc_index = icss_iep_get_ptp_clock_idx(emac->iep);
> +	info->tx_types = BIT(HWTSTAMP_TX_OFF) | BIT(HWTSTAMP_TX_ON);
> +	info->rx_filters = BIT(HWTSTAMP_FILTER_NONE) | BIT(HWTSTAMP_FILTER_ALL);
> +
> +	return 0;
> +}
> +
>  static int emac_set_channels(struct net_device *ndev,
>  			     struct ethtool_channels *ch)
>  {
> @@ -176,6 +196,7 @@ const struct ethtool_ops icssg_ethtool_ops = {
>  	.get_sset_count = emac_get_sset_count,
>  	.get_ethtool_stats = emac_get_ethtool_stats,
>  	.get_strings = emac_get_strings,
> +	.get_ts_info = emac_get_ts_info,
>  	.get_channels = emac_get_channels,
>  	.set_channels = emac_set_channels,
>  	.get_link_ksettings = emac_get_link_ksettings,
> diff --git a/drivers/net/ethernet/ti/icssg/icssg_prueth.c b/drivers/net/ethernet/ti/icssg/icssg_prueth.c
> index 47b941fb0198..b82a718fd602 100644
> --- a/drivers/net/ethernet/ti/icssg/icssg_prueth.c
> +++ b/drivers/net/ethernet/ti/icssg/icssg_prueth.c
> @@ -56,6 +56,8 @@
>  /* CTRLMMR_ICSSG_RGMII_CTRL register bits */
>  #define ICSSG_CTRL_RGMII_ID_MODE                BIT(24)
>  
> +#define IEP_DEFAULT_CYCLE_TIME_NS	1000000	/* 1 ms */
> +
>  static void prueth_cleanup_rx_chns(struct prueth_emac *emac,
>  				   struct prueth_rx_chn *rx_chn,
>  				   int max_rflows)
> @@ -471,6 +473,37 @@ static int prueth_dma_rx_push(struct prueth_emac *emac,
>  					desc_rx, desc_dma);
>  }
>  
> +static u64 icssg_ts_to_ns(u32 hi_sw, u32 hi, u32 lo, u32 cycle_time_ns)
> +{
> +	u32 iepcount_lo, iepcount_hi, hi_rollover_count;
> +	u64 ns;
> +
> +	iepcount_lo = lo & GENMASK(19, 0);
> +	iepcount_hi = (hi & GENMASK(11, 0)) << 12 | lo >> 20;
> +	hi_rollover_count = hi >> 11;
> +
> +	ns = ((u64)hi_rollover_count) << 23 | (iepcount_hi + hi_sw);
> +	ns = ns * cycle_time_ns + iepcount_lo;
> +
> +	return ns;
> +}
> +
> +static void emac_rx_timestamp(struct prueth_emac *emac,
> +			      struct sk_buff *skb, u32 *psdata)
> +{
> +	struct skb_shared_hwtstamps *ssh;
> +	u64 ns;
> +
> +	u32 hi_sw = readl(emac->prueth->shram.va +
> +			  TIMESYNC_FW_WC_COUNT_HI_SW_OFFSET_OFFSET);
> +	ns = icssg_ts_to_ns(hi_sw, psdata[1], psdata[0],
> +			    IEP_DEFAULT_CYCLE_TIME_NS);
> +
> +	ssh = skb_hwtstamps(skb);
> +	memset(ssh, 0, sizeof(*ssh));
> +	ssh->hwtstamp = ns_to_ktime(ns);
> +}
> +
>  static int emac_rx_packet(struct prueth_emac *emac, u32 flow_id)
>  {
>  	struct prueth_rx_chn *rx_chn = &emac->rx_chns;
> @@ -480,6 +513,7 @@ static int emac_rx_packet(struct prueth_emac *emac, u32 flow_id)
>  	struct sk_buff *skb, *new_skb;
>  	dma_addr_t desc_dma, buf_dma;
>  	void **swdata;
> +	u32 *psdata;
>  	int ret;
>  
>  	ret = k3_udma_glue_pop_rx_chn(rx_chn->rx_chn, flow_id, &desc_dma);
> @@ -497,6 +531,11 @@ static int emac_rx_packet(struct prueth_emac *emac, u32 flow_id)
>  	swdata = cppi5_hdesc_get_swdata(desc_rx);
>  	skb = *swdata;
>  
> +	psdata = cppi5_hdesc_get_psdata(desc_rx);
> +	/* RX HW timestamp */
> +	if (emac->rx_ts_enabled)
> +		emac_rx_timestamp(emac, skb, psdata);
> +
>  	cppi5_hdesc_get_obuf(desc_rx, &buf_dma, &buf_dma_len);
>  	k3_udma_glue_rx_cppi5_to_dma_addr(rx_chn->rx_chn, &buf_dma);
>  	pkt_len = cppi5_hdesc_get_pktlen(desc_rx);
> @@ -557,6 +596,82 @@ static void prueth_rx_cleanup(void *data, dma_addr_t desc_dma)
>  	dev_kfree_skb_any(skb);
>  }
>  
> +static int emac_get_tx_ts(struct prueth_emac *emac,
> +			  struct emac_tx_ts_response *rsp)
> +{
> +	struct prueth *prueth = emac->prueth;
> +	int slice = prueth_emac_slice(emac);
> +	int addr;
> +
> +	addr = icssg_queue_pop(prueth, slice == 0 ?
> +			       ICSSG_TS_POP_SLICE0 : ICSSG_TS_POP_SLICE1);
> +	if (addr < 0)
> +		return addr;
> +
> +	memcpy_fromio(rsp, prueth->shram.va + addr, sizeof(*rsp));
> +	/* return buffer back for to pool */
> +	icssg_queue_push(prueth, slice == 0 ?
> +			 ICSSG_TS_PUSH_SLICE0 : ICSSG_TS_PUSH_SLICE1, addr);
> +
> +	return 0;
> +}
> +
> +static void tx_ts_work(struct prueth_emac *emac)
> +{
> +	struct skb_shared_hwtstamps ssh;
> +	struct emac_tx_ts_response tsr;
> +	struct sk_buff *skb;
> +	int timeout = 10;
> +	int ret = 0;
> +	u32 hi_sw;
> +	u64 ns;
> +
> +	if (!test_bit(__STATE_TX_TS_IN_PROGRESS, &emac->state)) {
> +		netdev_err(emac->ndev, "unexpected TS response\n");
> +		return;
> +	}
> +
> +	skb = emac->tx_ts_skb;
> +	while (timeout-- > 0) {
> +		/* wait for response or timeout */
> +		ret = emac_get_tx_ts(emac, &tsr);
> +		if (!ret)
> +			break;
> +		usleep_range(10, 20);
> +	}
> +
> +	if (ret) {
> +		netdev_err(emac->ndev, "TX timestamp timeout\n");
> +		goto error;
> +	}
> +
> +	if (tsr.cookie != emac->tx_ts_cookie) {
> +		netdev_err(emac->ndev, "TX TS cookie mismatch 0x%x:0x%x\n",
> +			   tsr.cookie, emac->tx_ts_cookie);
> +		goto error;
> +	}
> +
> +	hi_sw = readl(emac->prueth->shram.va +
> +		      TIMESYNC_FW_WC_COUNT_HI_SW_OFFSET_OFFSET);
> +	ns = icssg_ts_to_ns(hi_sw, tsr.hi_ts, tsr.lo_ts,
> +			    IEP_DEFAULT_CYCLE_TIME_NS);
> +
> +	emac->tx_ts_cookie++;
> +	memset(&ssh, 0, sizeof(ssh));
> +	ssh.hwtstamp = ns_to_ktime(ns);
> +	clear_bit_unlock(__STATE_TX_TS_IN_PROGRESS, &emac->state);
> +
> +	skb_tstamp_tx(skb, &ssh);
> +	dev_consume_skb_any(skb);
> +
> +	return;
> +
> +error:
> +	dev_kfree_skb_any(skb);
> +	emac->tx_ts_skb = NULL;
> +	clear_bit_unlock(__STATE_TX_TS_IN_PROGRESS, &emac->state);
> +}
> +
>  /**
>   * emac_ndo_start_xmit - EMAC Transmit function
>   * @skb: SKB pointer
> @@ -577,6 +692,7 @@ static enum netdev_tx emac_ndo_start_xmit(struct sk_buff *skb, struct net_device
>  	struct prueth_tx_chn *tx_chn;
>  	dma_addr_t desc_dma, buf_dma;
>  	int i, ret = 0, q_idx;
> +	bool in_tx_ts = 0;
>  	void **swdata;
>  	u32 pkt_len;
>  	u32 *epib;
> @@ -608,6 +724,19 @@ static enum netdev_tx emac_ndo_start_xmit(struct sk_buff *skb, struct net_device
>  	epib = first_desc->epib;
>  	epib[0] = 0;
>  	epib[1] = 0;
> +	if (skb_shinfo(skb)->tx_flags & SKBTX_HW_TSTAMP &&
> +	    emac->tx_ts_enabled) {
> +		/* We currently support only one TX HW timestamp at a time */

If I remember right. There was a patch to get rid of this limitation.
Can you please pick and squash it with this patch?

> +		if (!test_and_set_bit_lock(__STATE_TX_TS_IN_PROGRESS,
> +					   &emac->state)) {
> +			skb_shinfo(skb)->tx_flags |= SKBTX_IN_PROGRESS;
> +			/* Request TX timestamp */
> +			epib[0] = emac->tx_ts_cookie;
> +			epib[1] = 0x80000000;	/* TX TS request */
> +			emac->tx_ts_skb = skb_get(skb);
> +			in_tx_ts = 1;
> +		}
> +	}
>  
>  	/* set dst tag to indicate internal qid at the firmware which is at
>  	 * bit8..bit15. bit0..bit7 indicates port num for directed
> @@ -629,7 +758,7 @@ static enum netdev_tx emac_ndo_start_xmit(struct sk_buff *skb, struct net_device
>  		if (!next_desc) {
>  			netdev_err(ndev,
>  				   "tx: failed to allocate frag. descriptor\n");
> -			goto free_desc_stop_q_busy;
> +			goto free_desc_stop_q_busy_cleanup_tx_ts;
>  		}
>  
>  		buf_dma = skb_frag_dma_map(tx_chn->dma_dev, frag, 0, frag_size,
> @@ -638,7 +767,7 @@ static enum netdev_tx emac_ndo_start_xmit(struct sk_buff *skb, struct net_device
>  			netdev_err(ndev, "tx: Failed to map skb page\n");
>  			k3_cppi_desc_pool_free(tx_chn->desc_pool, next_desc);
>  			ret = NETDEV_TX_OK;
> -			goto drop_free_descs;
> +			goto cleanup_tx_ts;
>  		}
>  
>  		cppi5_hdesc_reset_hbdesc(next_desc);
> @@ -682,6 +811,13 @@ static enum netdev_tx emac_ndo_start_xmit(struct sk_buff *skb, struct net_device
>  
>  	return NETDEV_TX_OK;
>  
> +cleanup_tx_ts:
> +	if (in_tx_ts) {
> +		dev_kfree_skb_any(emac->tx_ts_skb);
> +		emac->tx_ts_skb = NULL;
> +		clear_bit_unlock(__STATE_TX_TS_IN_PROGRESS, &emac->state);
> +	}
> +
>  drop_free_descs:
>  	prueth_xmit_free(tx_chn, first_desc);
>  
> @@ -694,7 +830,12 @@ static enum netdev_tx emac_ndo_start_xmit(struct sk_buff *skb, struct net_device
>  
>  	return ret;
>  
> -free_desc_stop_q_busy:
> +free_desc_stop_q_busy_cleanup_tx_ts:
> +	if (in_tx_ts) {
> +		dev_kfree_skb_any(emac->tx_ts_skb);
> +		emac->tx_ts_skb = NULL;
> +		clear_bit_unlock(__STATE_TX_TS_IN_PROGRESS, &emac->state);
> +	}
>  	prueth_xmit_free(tx_chn, first_desc);
>  
>  drop_stop_q_busy:
> @@ -717,6 +858,16 @@ static void prueth_tx_cleanup(void *data, dma_addr_t desc_dma)
>  	dev_kfree_skb_any(skb);
>  }
>  
> +static irqreturn_t prueth_tx_ts_irq(int irq, void *dev_id)
> +{
> +	struct prueth_emac *emac = dev_id;
> +
> +	/* currently only TX timestamp is being returned */
> +	tx_ts_work(emac);
> +
> +	return IRQ_HANDLED;
> +}
> +
>  static irqreturn_t prueth_rx_irq(int irq, void *dev_id)
>  {
>  	struct prueth_emac *emac = dev_id;
> @@ -992,6 +1143,139 @@ static int emac_phy_connect(struct prueth_emac *emac)
>  	return 0;
>  }
>  
> +static u64 prueth_iep_gettime(void *clockops_data, struct ptp_system_timestamp *sts)
> +{
> +	u32 hi_rollover_count, hi_rollover_count_r;
> +	struct prueth_emac *emac = clockops_data;
> +	struct prueth *prueth = emac->prueth;
> +	void __iomem *fw_hi_r_count_addr;
> +	void __iomem *fw_count_hi_addr;
> +	u32 iepcount_hi, iepcount_hi_r;
> +	unsigned long flags;
> +	u32 iepcount_lo;
> +	u64 ts = 0;
> +
> +	fw_count_hi_addr = prueth->shram.va + TIMESYNC_FW_WC_COUNT_HI_SW_OFFSET_OFFSET;
> +	fw_hi_r_count_addr = prueth->shram.va + TIMESYNC_FW_WC_HI_ROLLOVER_COUNT_OFFSET;
> +
> +	local_irq_save(flags);
> +	do {
> +		iepcount_hi = icss_iep_get_count_hi(emac->iep);
> +		iepcount_hi += readl(fw_count_hi_addr);
> +		hi_rollover_count = readl(fw_hi_r_count_addr);
> +		ptp_read_system_prets(sts);
> +		iepcount_lo = icss_iep_get_count_low(emac->iep);
> +		ptp_read_system_postts(sts);
> +
> +		iepcount_hi_r = icss_iep_get_count_hi(emac->iep);
> +		iepcount_hi_r += readl(fw_count_hi_addr);
> +		hi_rollover_count_r = readl(fw_hi_r_count_addr);
> +	} while ((iepcount_hi_r != iepcount_hi) ||
> +		 (hi_rollover_count != hi_rollover_count_r));
> +	local_irq_restore(flags);
> +
> +	ts = ((u64)hi_rollover_count) << 23 | iepcount_hi;
> +	ts = ts * (u64)IEP_DEFAULT_CYCLE_TIME_NS + iepcount_lo;
> +
> +	return ts;
> +}
> +
> +static void prueth_iep_settime(void *clockops_data, u64 ns)
> +{
> +	struct icssg_setclock_desc __iomem *sc_descp;
> +	struct prueth_emac *emac = clockops_data;
> +	struct icssg_setclock_desc sc_desc;
> +	u64 cyclecount;
> +	u32 cycletime;
> +	int timeout;
> +
> +	if (!emac->fw_running)
> +		return;
> +
> +	sc_descp = emac->prueth->shram.va + TIMESYNC_FW_WC_SETCLOCK_DESC_OFFSET;
> +
> +	cycletime = IEP_DEFAULT_CYCLE_TIME_NS;
> +	cyclecount = ns / cycletime;
> +
> +	memset(&sc_desc, 0, sizeof(sc_desc));
> +	sc_desc.margin = cycletime - 1000;
> +	sc_desc.cyclecounter0_set = cyclecount & GENMASK(31, 0);
> +	sc_desc.cyclecounter1_set = (cyclecount & GENMASK(63, 32)) >> 32;
> +	sc_desc.iepcount_set = ns % cycletime;
> +	sc_desc.CMP0_current = cycletime - 4; //Count from 0 to (cycle time)-4
> +
> +	memcpy_toio(sc_descp, &sc_desc, sizeof(sc_desc));
> +
> +	writeb(1, &sc_descp->request);
> +
> +	timeout = 5;	/* fw should take 2-3 ms */
> +	while (timeout--) {
> +		if (readb(&sc_descp->acknowledgment))
> +			return;
> +
> +		usleep_range(500, 1000);
> +	}
> +
> +	dev_err(emac->prueth->dev, "settime timeout\n");
> +}
> +
> +static int prueth_perout_enable(void *clockops_data,
> +				struct ptp_perout_request *req, int on,
> +				u64 *cmp)
> +{
> +	struct prueth_emac *emac = clockops_data;
> +	u32 reduction_factor = 0, offset = 0;
> +	struct timespec64 ts;
> +	u64 ns_period;
> +
> +	if (!on)
> +		return 0;
> +
> +	/* Any firmware specific stuff for PPS/PEROUT handling */
> +	ts.tv_sec = req->period.sec;
> +	ts.tv_nsec = req->period.nsec;
> +	ns_period = timespec64_to_ns(&ts);
> +
> +	/* f/w doesn't support period less than cycle time */
> +	if (ns_period < IEP_DEFAULT_CYCLE_TIME_NS)
> +		return -ENXIO;
> +
> +	reduction_factor = ns_period / IEP_DEFAULT_CYCLE_TIME_NS;
> +	offset = ns_period % IEP_DEFAULT_CYCLE_TIME_NS;
> +
> +	/* f/w requires at least 1uS within a cycle so CMP
> +	 * can trigger after SYNC is enabled
> +	 */
> +	if (offset < 5 * NSEC_PER_USEC)
> +		offset = 5 * NSEC_PER_USEC;
> +
> +	/* if offset is close to cycle time then we will miss
> +	 * the CMP event for last tick when IEP rolls over.
> +	 * In normal mode, IEP tick is 4ns.
> +	 * In slow compensation it could be 0ns or 8ns at
> +	 * every slow compensation cycle.
> +	 */
> +	if (offset > IEP_DEFAULT_CYCLE_TIME_NS - 8)
> +		offset = IEP_DEFAULT_CYCLE_TIME_NS - 8;
> +
> +	/* we're in shadow mode so need to set upper 32-bits */
> +	*cmp = (u64)offset << 32;
> +
> +	writel(reduction_factor, emac->prueth->shram.va +
> +		TIMESYNC_FW_WC_SYNCOUT_REDUCTION_FACTOR_OFFSET);
> +
> +	writel(0, emac->prueth->shram.va +
> +		TIMESYNC_FW_WC_SYNCOUT_START_TIME_CYCLECOUNT_OFFSET);
> +
> +	return 0;
> +}
> +
> +const struct icss_iep_clockops prueth_iep_clockops = {
> +	.settime = prueth_iep_settime,
> +	.gettime = prueth_iep_gettime,
> +	.perout_enable = prueth_perout_enable,
> +};
> +
>  /**
>   * emac_ndo_open - EMAC device open
>   * @ndev: network adapter device
> @@ -1066,10 +1350,20 @@ static int emac_ndo_open(struct net_device *ndev)
>  
>  	icssg_mii_update_mtu(prueth->mii_rt, slice, ndev->max_mtu);
>  
> +	if (!prueth->emacs_initialized) {
> +		ret = icss_iep_init(emac->iep, &prueth_iep_clockops,
> +				    emac, IEP_DEFAULT_CYCLE_TIME_NS);
> +	}
> +
> +	ret = request_threaded_irq(emac->tx_ts_irq, NULL, prueth_tx_ts_irq,
> +				   IRQF_ONESHOT, dev_name(dev), emac);
> +	if (ret)
> +		goto stop;
> +
>  	/* Prepare RX */
>  	ret = prueth_prepare_rx_chan(emac, &emac->rx_chns, PRUETH_MAX_PKT_SIZE);
>  	if (ret)
> -		goto stop;
> +		goto free_tx_ts_irq;
>  
>  	ret = k3_udma_glue_enable_rx_chn(emac->rx_chns.rx_chn);
>  	if (ret)
> @@ -1102,6 +1396,8 @@ static int emac_ndo_open(struct net_device *ndev)
>  	prueth_reset_tx_chan(emac, i, false);
>  reset_rx_chn:
>  	prueth_reset_rx_chan(&emac->rx_chns, max_rx_flows, false);
> +free_tx_ts_irq:
> +	free_irq(emac->tx_ts_irq, emac);
>  stop:
>  	prueth_emac_stop(emac);
>  free_rx_irq:
> @@ -1173,6 +1469,14 @@ static int emac_ndo_stop(struct net_device *ndev)
>  	/* stop PRUs */
>  	prueth_emac_stop(emac);
>  
> +	if (prueth->emacs_initialized == 1)
> +		icss_iep_exit(emac->iep);
> +
> +	/* stop PRUs */
> +	prueth_emac_stop(emac);
> +
> +	free_irq(emac->tx_ts_irq, emac);
> +
>  	free_irq(emac->rx_chns.irq[rx_flow], emac);
>  	prueth_ndev_del_tx_napi(emac, emac->tx_ch_num);
>  	prueth_cleanup_tx_chns(emac);
> @@ -1235,8 +1539,79 @@ static void emac_ndo_set_rx_mode(struct net_device *ndev)
>  	queue_work(emac->cmd_wq, &emac->rx_mode_work);
>  }
>  
> +static int emac_set_ts_config(struct net_device *ndev, struct ifreq *ifr)
> +{
> +	struct prueth_emac *emac = netdev_priv(ndev);
> +	struct hwtstamp_config config;
> +
> +	if (copy_from_user(&config, ifr->ifr_data, sizeof(config)))
> +		return -EFAULT;
> +
> +	switch (config.tx_type) {
> +	case HWTSTAMP_TX_OFF:
> +		emac->tx_ts_enabled = 0;
> +		break;
> +	case HWTSTAMP_TX_ON:
> +		emac->tx_ts_enabled = 1;
> +		break;
> +	default:
> +		return -ERANGE;
> +	}
> +
> +	switch (config.rx_filter) {
> +	case HWTSTAMP_FILTER_NONE:
> +		emac->rx_ts_enabled = 0;
> +		break;
> +	case HWTSTAMP_FILTER_ALL:
> +	case HWTSTAMP_FILTER_SOME:
> +	case HWTSTAMP_FILTER_PTP_V1_L4_EVENT:
> +	case HWTSTAMP_FILTER_PTP_V1_L4_SYNC:
> +	case HWTSTAMP_FILTER_PTP_V1_L4_DELAY_REQ:
> +	case HWTSTAMP_FILTER_PTP_V2_L4_EVENT:
> +	case HWTSTAMP_FILTER_PTP_V2_L4_SYNC:
> +	case HWTSTAMP_FILTER_PTP_V2_L4_DELAY_REQ:
> +	case HWTSTAMP_FILTER_PTP_V2_L2_EVENT:
> +	case HWTSTAMP_FILTER_PTP_V2_L2_SYNC:
> +	case HWTSTAMP_FILTER_PTP_V2_L2_DELAY_REQ:
> +	case HWTSTAMP_FILTER_PTP_V2_EVENT:
> +	case HWTSTAMP_FILTER_PTP_V2_SYNC:
> +	case HWTSTAMP_FILTER_PTP_V2_DELAY_REQ:
> +	case HWTSTAMP_FILTER_NTP_ALL:
> +		emac->rx_ts_enabled = 1;
> +		config.rx_filter = HWTSTAMP_FILTER_ALL;
> +		break;
> +	default:
> +		return -ERANGE;
> +	}
> +
> +	return copy_to_user(ifr->ifr_data, &config, sizeof(config)) ?
> +		-EFAULT : 0;
> +}
> +
> +static int emac_get_ts_config(struct net_device *ndev, struct ifreq *ifr)
> +{
> +	struct prueth_emac *emac = netdev_priv(ndev);
> +	struct hwtstamp_config config;
> +
> +	config.flags = 0;
> +	config.tx_type = emac->tx_ts_enabled ? HWTSTAMP_TX_ON : HWTSTAMP_TX_OFF;
> +	config.rx_filter = emac->rx_ts_enabled ? HWTSTAMP_FILTER_ALL : HWTSTAMP_FILTER_NONE;
> +
> +	return copy_to_user(ifr->ifr_data, &config, sizeof(config)) ?
> +			    -EFAULT : 0;
> +}
> +
>  static int emac_ndo_ioctl(struct net_device *ndev, struct ifreq *ifr, int cmd)
>  {
> +	switch (cmd) {
> +	case SIOCGHWTSTAMP:
> +		return emac_get_ts_config(ndev, ifr);
> +	case SIOCSHWTSTAMP:
> +		return emac_set_ts_config(ndev, ifr);
> +	default:
> +		break;
> +	}
> +
>  	return phy_do_ioctl(ndev, ifr, cmd);
>  }
>  
> @@ -1316,6 +1691,7 @@ static int prueth_netdev_init(struct prueth *prueth,
>  	struct prueth_emac *emac;
>  	struct net_device *ndev;
>  	enum prueth_port port;
> +	const char *irq_name;
>  	enum prueth_mac mac;
>  
>  	port = prueth_node_port(eth_node);
> @@ -1355,6 +1731,15 @@ static int prueth_netdev_init(struct prueth *prueth,
>  
>  	emac->tx_ch_num = 1;
>  
> +	irq_name = "tx_ts0";
> +	if (emac->port_id == PRUETH_PORT_MII1)
> +		irq_name = "tx_ts1";
> +	emac->tx_ts_irq = platform_get_irq_byname_optional(prueth->pdev, irq_name);
> +	if (emac->tx_ts_irq < 0) {
> +		ret = dev_err_probe(prueth->dev, emac->tx_ts_irq, "could not get tx_ts_irq\n");
> +		goto free;
> +	}
> +
>  	SET_NETDEV_DEV(ndev, prueth->dev);
>  	spin_lock_init(&emac->lock);
>  	mutex_init(&emac->cmd_lock);
> @@ -1680,6 +2065,22 @@ static int prueth_probe(struct platform_device *pdev)
>  	dev_dbg(dev, "sram: pa %llx va %p size %zx\n", prueth->msmcram.pa,
>  		prueth->msmcram.va, prueth->msmcram.size);
>  
> +	prueth->iep0 = icss_iep_get_idx(np, 0);
> +	if (IS_ERR(prueth->iep0)) {
> +		ret = dev_err_probe(dev, PTR_ERR(prueth->iep0), "iep0 get failed\n");
> +		prueth->iep0 = NULL;
> +		goto free_pool;
> +	}
> +
> +	prueth->iep1 = icss_iep_get_idx(np, 1);
> +	if (IS_ERR(prueth->iep1)) {
> +		ret = dev_err_probe(dev, PTR_ERR(prueth->iep1), "iep1 get failed\n");
> +		icss_iep_put(prueth->iep0);
> +		prueth->iep0 = NULL;
> +		prueth->iep1 = NULL;
> +		goto free_pool;
> +	}
> +

As IEP1 is not used, why do you need to get it?

>  	/* setup netdev interfaces */
>  	if (eth0_node) {
>  		ret = prueth_netdev_init(prueth, eth0_node);
> @@ -1688,6 +2089,7 @@ static int prueth_probe(struct platform_device *pdev)
>  				      eth0_node->name);
>  			goto netdev_exit;
>  		}
> +		prueth->emac[PRUETH_MAC0]->iep = prueth->iep0;
>  	}
>  
>  	if (eth1_node) {
> @@ -1697,6 +2099,8 @@ static int prueth_probe(struct platform_device *pdev)
>  				      eth1_node->name);
>  			goto netdev_exit;
>  		}
> +
> +		prueth->emac[PRUETH_MAC1]->iep = prueth->iep0;
>  	}
>  
>  	/* register the network devices */
> @@ -1754,6 +2158,7 @@ static int prueth_probe(struct platform_device *pdev)
>  		prueth_netdev_exit(prueth, eth_node);
>  	}
>  
> +free_pool:
>  	gen_pool_free(prueth->sram_pool,
>  		      (unsigned long)prueth->msmcram.va, msmc_ram_size);
>  
> @@ -1798,6 +2203,9 @@ static void prueth_remove(struct platform_device *pdev)
>  		prueth_netdev_exit(prueth, eth_node);
>  	}
>  
> +	icss_iep_put(prueth->iep1);
> +	icss_iep_put(prueth->iep0);
> +
>  	gen_pool_free(prueth->sram_pool,
>  		      (unsigned long)prueth->msmcram.va,
>  		      MSMC_RAM_SIZE);
> diff --git a/drivers/net/ethernet/ti/icssg/icssg_prueth.h b/drivers/net/ethernet/ti/icssg/icssg_prueth.h
> index a8ce4d01ef16..67ddbff40108 100644
> --- a/drivers/net/ethernet/ti/icssg/icssg_prueth.h
> +++ b/drivers/net/ethernet/ti/icssg/icssg_prueth.h
> @@ -35,6 +35,7 @@
>  #include <net/devlink.h>
>  
>  #include "icssg_config.h"
> +#include "icss_iep.h"
>  #include "icssg_switch_map.h"
>  
>  #define PRUETH_MAX_MTU          (2000 - ETH_HLEN - ETH_FCS_LEN)
> @@ -117,6 +118,10 @@ struct prueth_rx_chn {
>  	char name[32];
>  };
>  
> +enum prueth_state_flags {
> +	__STATE_TX_TS_IN_PROGRESS,
> +};
> +
>  /* There are 4 Tx DMA channels, but the highest priority is CH3 (thread 3)
>   * and lower three are lower priority channels or threads.
>   */
> @@ -139,6 +144,9 @@ struct prueth_emac {
>  	struct device_node *phy_node;
>  	phy_interface_t phy_if;
>  	enum prueth_port port_id;
> +	struct icss_iep *iep;
> +	unsigned int rx_ts_enabled : 1;
> +	unsigned int tx_ts_enabled : 1;
>  
>  	/* DMA related */
>  	struct prueth_tx_chn tx_chns[PRUETH_MAX_TX_QUEUES];
> @@ -151,6 +159,14 @@ struct prueth_emac {
>  	spinlock_t lock;	/* serialize access */
>  
>  	unsigned long state;
> +	/* TX HW Timestamping */
> +	u32 tx_ts_cookie;
> +	struct sk_buff *tx_ts_skb;
> +	int tx_ts_irq;
> +
> +	u8 cmd_seq;
> +	/* shutdown related */
> +	u32 cmd_data[4];
>  	struct completion cmd_complete;
>  	/* Mutex to serialize access to firmware command interface */
>  	struct mutex cmd_lock;
> @@ -193,6 +209,8 @@ struct prueth_pdata {
>   * @pdata: pointer to platform data for ICSSG driver
>   * @icssg_hwcmdseq: seq counter or HWQ messages
>   * @emacs_initialized: num of EMACs/ext ports that are up/running
> + * @iep0: pointer to IEP0 device
> + * @iep1: pointer to IEP1 device
>   */
>  struct prueth {
>  	struct device *dev;
> @@ -214,8 +232,16 @@ struct prueth {
>  	struct platform_device *pdev;
>  	struct prueth_pdata pdata;
>  	u8 icssg_hwcmdseq;
> -
>  	int emacs_initialized;
> +	struct icss_iep *iep0;
> +	struct icss_iep *iep1;

Do you need iep1?

> +};
> +
> +struct emac_tx_ts_response {
> +	u32 reserved[2];
> +	u32 cookie;
> +	u32 lo_ts;
> +	u32 hi_ts;
>  };
>  
>  /* get PRUSS SLICE number from prueth_emac */

-- 
cheers,
-roger