From: Frank Li <Frank.li@oss.nxp.com>
To: Larisa Grigore <larisa.grigore@oss.nxp.com>
Cc: "NXP S32 Linux Team" <s32@nxp.com>,
"Alim Akhtar" <alim.akhtar@samsung.com>,
"Avri Altman" <avri.altman@sandisk.com>,
"Bart Van Assche" <bvanassche@acm.org>,
"Rob Herring" <robh@kernel.org>,
"Krzysztof Kozlowski" <krzk+dt@kernel.org>,
"Conor Dooley" <conor+dt@kernel.org>,
"Frank Li" <Frank.Li@nxp.com>,
"Sascha Hauer" <s.hauer@pengutronix.de>,
"Pengutronix Kernel Team" <kernel@pengutronix.de>,
"Fabio Estevam" <festevam@gmail.com>,
"James E.J. Bottomley" <James.Bottomley@hansenpartnership.com>,
"Martin K. Petersen" <mkp@kernel.org>,
"Sai Krishna Potthuri" <sai.krishna.potthuri@amd.com>,
"Ajay Neeli" <ajay.neeli@amd.com>,
"Manivannan Sadhasivam" <mani@kernel.org>,
"Pedro Sousa " <pedrom.sousa@synopsys.com>,
linux-scsi@vger.kernel.org, devicetree@vger.kernel.org,
linux-kernel@vger.kernel.org, imx@lists.linux.dev,
linux-arm-kernel@lists.infradead.org,
linux-arm-msm@vger.kernel.org, clizzi@redhat.com,
aruizrui@redhat.com, eballetb@redhat.com, echanude@redhat.com
Subject: Re: [PATCH v3 03/10] ufs: host: Add common Hibern8 TX FSM polling helper
Date: Wed, 16 Sep 2026 11:22:40 -0500 [thread overview]
Message-ID: <aqrCUDrYaIsL39zn@SMW015318> (raw)
In-Reply-To: <20260916084411.941297-4-larisa.grigore@oss.nxp.com>
On Wed, Sep 16, 2026 at 10:44:04AM +0200, Larisa Grigore wrote:
> Factor out the common logic used to poll TX_FSM_STATE until the TX lanes
> enter Hibern8 into ufshcd_check_hibern8().
>
> The HiSilicon and Qualcomm drivers currently implement similar checks
> using the UniPro TX_FSM_STATE attribute after link/PHY bring-up. Move
> this logic to a common helper in ufshcd so it can be shared across host
> controller drivers.
>
> Switch the HiSilicon and Qualcomm drivers to use the new helper and drop
> their local implementations. While at it, normalize the "TX lane failed to
> reach Hibern8" error path to return -ETIMEDOUT instead of the raw
> TX_FSM_STATE value (qcom) or -1 (hisi), so all callers get a proper errno.
> This is an intentional, harmless change of the returned error value; the
> pass/fail behaviour at the call sites is unchanged.
>
> Unlike the initial implementation, the timeout error is now reported only
> once by ufshcd_dme_check_tx_hibern8() after its final check, instead of
> per lane.
>
> This also prepares for reusing the same UniPro-specific Hibern8 check in
> a subsequent commit.
>
> Signed-off-by: Larisa Grigore <larisa.grigore@oss.nxp.com>
> ---
> drivers/ufs/core/ufshcd.c | 86 +++++++++++++++++++++++++++++++++++++
> drivers/ufs/host/ufs-hisi.c | 48 +--------------------
> drivers/ufs/host/ufs-qcom.c | 42 +-----------------
> include/ufs/ufshcd.h | 3 ++
> 4 files changed, 91 insertions(+), 88 deletions(-)
>
> diff --git a/drivers/ufs/core/ufshcd.c b/drivers/ufs/core/ufshcd.c
> index 2ba244cf40ac..f34b7fe54a16 100644
> --- a/drivers/ufs/core/ufshcd.c
> +++ b/drivers/ufs/core/ufshcd.c
> @@ -4448,6 +4448,92 @@ int ufshcd_dme_get_attr(struct ufs_hba *hba, u32 attr_sel,
> }
> EXPORT_SYMBOL_GPL(ufshcd_dme_get_attr);
>
> +/**
> + * ufshcd_poll_tx_hibern8_lanes - Check TX_FSM_STATE of all TX lanes once
> + * @hba: host controller instance
> + * @num_lanes: number of TX lanes to check
> + *
> + * Read TX_FSM_STATE for every lane and verify it reached Hibern8.
> + *
> + * Return: 0 if all lanes are in Hibern8, -EAGAIN if any lane is not (yet)
> + * in Hibern8, or a negative errno (e.g. -ETIMEDOUT from a hard UIC command
> + * timeout) if the attribute read fails.
> + */
> +static int ufshcd_poll_tx_hibern8_lanes(struct ufs_hba *hba,
> + unsigned int num_lanes)
> +{
> + u32 tx_fsm_val = 0;
> + unsigned int i;
> + int err;
> +
> + for (i = 0; i < num_lanes; i++) {
> + err = ufshcd_dme_get(hba,
> + UIC_ARG_MIB_SEL(TX_FSM_STATE,
> + UIC_ARG_MPHY_TX_GEN_SEL_INDEX(i)),
> + &tx_fsm_val);
> + if (err) {
> + dev_err(hba->dev,
> + "%s: unable to get TX_FSM_STATE for lane %u, err %d\n",
> + __func__, i, err);
> + return err;
> + }
> +
> + if (tx_fsm_val != TX_STATE_HIBERN8)
> + return -EAGAIN;
> + }
> +
> + return 0;
> +}
> +
> +/**
> + * ufshcd_dme_check_tx_hibern8 - Check if all TX lanes entered Hibern8 state
> + * @hba: host controller instance
> + * @num_lanes: number of TX lanes to check
> + * @timeout_ms: timeout in milliseconds for all lanes
> + *
> + * Return: 0 on success, negative errno on failure.
> + */
> +int ufshcd_dme_check_tx_hibern8(struct ufs_hba *hba, unsigned int num_lanes,
> + unsigned int timeout_ms)
> +{
> + unsigned long timeout;
> + int err;
> +
> + if (!num_lanes)
> + return -EINVAL;
> +
> + timeout = jiffies + msecs_to_jiffies(timeout_ms);
> +
> + do {
> + err = ufshcd_poll_tx_hibern8_lanes(hba, num_lanes);
> + /*
> + * -EAGAIN means the lanes are simply not in HIBERN8 yet, so
> + * keep polling. Any other error is a real failure and must
> + * be returned immediately instead of being retried.
> + */
> + if (err != -EAGAIN)
> + return err;
> +
> + /* sleep for max. 200us */
> + usleep_range(100, 200);
> + } while (time_before(jiffies, timeout));
> +
> + /*
> + * We might have been scheduled out for long during polling, so do
> + * one final check before reporting timeout.
> + */
> + err = ufshcd_poll_tx_hibern8_lanes(hba, num_lanes);
> + if (err == -EAGAIN) {
> + dev_err(hba->dev,
> + "%s: timeout waiting for TX lanes to enter HIBERN8\n",
> + __func__);
> + err = -ETIMEDOUT;
> + }
why not use read read_poll_timeout()
read_poll_timeout(ufshcd_poll_tx_hibern8_lanes, err, err != -EAGAIN, timeout_ms,
0, hba, numlanes);
Frank
> +
> + return err;
> +}
> +EXPORT_SYMBOL_GPL(ufshcd_dme_check_tx_hibern8);
> +
> /**
> * ufshcd_dme_rmw - get modify set a DME attribute
> * @hba: per adapter instance
> diff --git a/drivers/ufs/host/ufs-hisi.c b/drivers/ufs/host/ufs-hisi.c
> index bd223bda1ce2..b84075dfe00d 100644
> --- a/drivers/ufs/host/ufs-hisi.c
> +++ b/drivers/ufs/host/ufs-hisi.c
> @@ -22,50 +22,6 @@
> #include <ufs/ufshci.h>
> #include <ufs/ufs_quirks.h>
>
> -static int ufs_hisi_check_hibern8(struct ufs_hba *hba)
> -{
> - int err = 0;
> - u32 tx_fsm_val_0 = 0;
> - u32 tx_fsm_val_1 = 0;
> - unsigned long timeout = jiffies + msecs_to_jiffies(HBRN8_POLL_TOUT_MS);
> -
> - do {
> - err = ufshcd_dme_get(hba, UIC_ARG_MIB_SEL(TX_FSM_STATE, 0),
> - &tx_fsm_val_0);
> - err |= ufshcd_dme_get(hba,
> - UIC_ARG_MIB_SEL(TX_FSM_STATE, 1), &tx_fsm_val_1);
> - if (err || (tx_fsm_val_0 == TX_STATE_HIBERN8 &&
> - tx_fsm_val_1 == TX_STATE_HIBERN8))
> - break;
> -
> - /* sleep for max. 200us */
> - usleep_range(100, 200);
> - } while (time_before(jiffies, timeout));
> -
> - /*
> - * we might have scheduled out for long during polling so
> - * check the state again.
> - */
> - if (time_after(jiffies, timeout)) {
> - err = ufshcd_dme_get(hba, UIC_ARG_MIB_SEL(TX_FSM_STATE, 0),
> - &tx_fsm_val_0);
> - err |= ufshcd_dme_get(hba,
> - UIC_ARG_MIB_SEL(TX_FSM_STATE, 1), &tx_fsm_val_1);
> - }
> -
> - if (err) {
> - dev_err(hba->dev, "%s: unable to get TX_FSM_STATE, err %d\n",
> - __func__, err);
> - } else if (tx_fsm_val_0 != TX_STATE_HIBERN8 ||
> - tx_fsm_val_1 != TX_STATE_HIBERN8) {
> - err = -1;
> - dev_err(hba->dev, "%s: invalid TX_FSM_STATE, lane0 = %d, lane1 = %d\n",
> - __func__, tx_fsm_val_0, tx_fsm_val_1);
> - }
> -
> - return err;
> -}
> -
> static void ufs_hisi_clk_init(struct ufs_hba *hba)
> {
> struct ufs_hisi_host *host = ufshcd_get_variant(hba);
> @@ -224,9 +180,7 @@ static int ufs_hisi_link_startup_pre_change(struct ufs_hba *hba)
>
> /* Unipro VS_mphy_disable */
> ufshcd_dme_set(hba, UIC_ARG_MIB_SEL(0xD0C1, 0x0), 0x0);
> - err = ufs_hisi_check_hibern8(hba);
> - if (err)
> - dev_err(hba->dev, "ufs_hisi_check_hibern8 error\n");
> + err = ufshcd_dme_check_tx_hibern8(hba, 2, HBRN8_POLL_TOUT_MS);
>
> if (!(host->caps & UFS_HISI_CAP_PHY10nm))
> ufshcd_writel(hba, UFS_HCLKDIV_NORMAL_VALUE, UFS_REG_HCLKDIV);
> diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c
> index 1e492dac8d93..815ca82a0878 100644
> --- a/drivers/ufs/host/ufs-qcom.c
> +++ b/drivers/ufs/host/ufs-qcom.c
> @@ -382,46 +382,6 @@ static int ufs_qcom_init_lane_clks(struct ufs_qcom_host *host)
> return 0;
> }
>
> -static int ufs_qcom_check_hibern8(struct ufs_hba *hba)
> -{
> - int err;
> - u32 tx_fsm_val;
> - unsigned long timeout = jiffies + msecs_to_jiffies(HBRN8_POLL_TOUT_MS);
> -
> - do {
> - err = ufshcd_dme_get(hba,
> - UIC_ARG_MIB_SEL(TX_FSM_STATE,
> - UIC_ARG_MPHY_TX_GEN_SEL_INDEX(0)),
> - &tx_fsm_val);
> - if (err || tx_fsm_val == TX_STATE_HIBERN8)
> - break;
> -
> - /* sleep for max. 200us */
> - usleep_range(100, 200);
> - } while (time_before(jiffies, timeout));
> -
> - /*
> - * we might have scheduled out for long during polling so
> - * check the state again.
> - */
> - if (time_after(jiffies, timeout))
> - err = ufshcd_dme_get(hba,
> - UIC_ARG_MIB_SEL(TX_FSM_STATE,
> - UIC_ARG_MPHY_TX_GEN_SEL_INDEX(0)),
> - &tx_fsm_val);
> -
> - if (err) {
> - dev_err(hba->dev, "%s: unable to get TX_FSM_STATE, err %d\n",
> - __func__, err);
> - } else if (tx_fsm_val != TX_STATE_HIBERN8) {
> - err = tx_fsm_val;
> - dev_err(hba->dev, "%s: invalid TX_FSM_STATE = %d\n",
> - __func__, err);
> - }
> -
> - return err;
> -}
> -
> static void ufs_qcom_select_unipro_mode(struct ufs_qcom_host *host)
> {
> ufshcd_rmwl(host->hba, QUNIPRO_SEL, QUNIPRO_SEL, REG_UFS_CFG1);
> @@ -607,7 +567,7 @@ static int ufs_qcom_hce_enable_notify(struct ufs_hba *hba,
> break;
> case POST_CHANGE:
> /* check if UFS PHY moved from DISABLED to HIBERN8 */
> - err = ufs_qcom_check_hibern8(hba);
> + err = ufshcd_dme_check_tx_hibern8(hba, 1, HBRN8_POLL_TOUT_MS);
> ufs_qcom_enable_hw_clk_gating(hba);
> ufs_qcom_ice_enable(host);
> ufs_qcom_config_ice_allocator(host);
> diff --git a/include/ufs/ufshcd.h b/include/ufs/ufshcd.h
> index dfd302f2dc7c..0d9d0a26eb20 100644
> --- a/include/ufs/ufshcd.h
> +++ b/include/ufs/ufshcd.h
> @@ -1556,6 +1556,9 @@ extern int ufshcd_dme_set_attr(struct ufs_hba *hba, u32 attr_sel,
> u8 attr_set, u32 mib_val, u8 peer);
> extern int ufshcd_dme_get_attr(struct ufs_hba *hba, u32 attr_sel,
> u32 *mib_val, u8 peer);
> +extern int ufshcd_dme_check_tx_hibern8(struct ufs_hba *hba,
> + unsigned int num_lanes,
> + unsigned int timeout_ms);
> extern int ufshcd_change_power_mode(struct ufs_hba *hba,
> struct ufs_pa_layer_attr *pwr_mode,
> enum ufshcd_pmc_policy pmc_policy);
> --
> 2.43.0
>
>
next prev parent reply other threads:[~2026-09-16 16:22 UTC|newest]
Thread overview: 17+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-16 8:44 [PATCH v3 00/10] Add S32N79RDB UFS support Larisa Grigore
2026-09-16 8:44 ` [PATCH v3 01/10] ufs: unipro: Move Tx/Rx FSM state definitions to unipro.h Larisa Grigore
2026-09-16 16:14 ` Frank Li
2026-09-17 9:00 ` Larisa Ileana Grigore
2026-09-16 8:44 ` [PATCH v3 02/10] ufs: unipro: Add TX/RX FSM state attributes Larisa Grigore
2026-09-16 8:44 ` [PATCH v3 03/10] ufs: host: Add common Hibern8 TX FSM polling helper Larisa Grigore
2026-09-16 16:22 ` Frank Li [this message]
2026-09-17 9:00 ` Larisa Ileana Grigore
2026-09-16 8:44 ` [PATCH v3 04/10] scsi: ufs: Move Versal2 M-PHY CREG access helpers into ufshcd-dwc Larisa Grigore
2026-09-16 16:29 ` Frank Li
2026-09-16 8:44 ` [PATCH v3 05/10] scsi: ufs: dwc: Export common clock divider and link status helpers Larisa Grigore
2026-09-16 8:44 ` [PATCH v3 06/10] dt-bindings: ufs: Add NXP S32N79 UFS host controller Larisa Grigore
2026-09-16 16:34 ` Frank Li
2026-09-16 8:44 ` [PATCH v3 07/10] scsi: ufs: Add NXP S32N79 UFS host controller driver Larisa Grigore
2026-09-16 8:44 ` [PATCH v3 08/10] arm64: dts: freescale: s32n79: Add UFS host controller Larisa Grigore
2026-09-16 8:44 ` [PATCH v3 09/10] arm64: dts: freescale: s32n79-rdb: Enable UFS Larisa Grigore
2026-09-16 8:44 ` [PATCH v3 10/10] MAINTAINERS: Add NXP S32N7 UFS host controller entry Larisa Grigore
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=aqrCUDrYaIsL39zn@SMW015318 \
--to=frank.li@oss.nxp.com \
--cc=Frank.Li@nxp.com \
--cc=James.Bottomley@hansenpartnership.com \
--cc=ajay.neeli@amd.com \
--cc=alim.akhtar@samsung.com \
--cc=aruizrui@redhat.com \
--cc=avri.altman@sandisk.com \
--cc=bvanassche@acm.org \
--cc=clizzi@redhat.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=eballetb@redhat.com \
--cc=echanude@redhat.com \
--cc=festevam@gmail.com \
--cc=imx@lists.linux.dev \
--cc=kernel@pengutronix.de \
--cc=krzk+dt@kernel.org \
--cc=larisa.grigore@oss.nxp.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-arm-msm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-scsi@vger.kernel.org \
--cc=mani@kernel.org \
--cc=mkp@kernel.org \
--cc=pedrom.sousa@synopsys.com \
--cc=robh@kernel.org \
--cc=s.hauer@pengutronix.de \
--cc=s32@nxp.com \
--cc=sai.krishna.potthuri@amd.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®