by Manivannan Sadhasivam

[permalink] [raw]

Subject: Re: [PATCH v5 07/10] scsi: ufs: ufs-qcom: Set initial PHY gear to max HS gear for HW ver 5 and newer

On Tue, Nov 28, 2023 at 07:01:27PM +0800, Can Guo wrote:
>
>
> On 11/28/2023 6:59 PM, Manivannan Sadhasivam wrote:
> > On Tue, Nov 28, 2023 at 03:58:42PM +0800, Can Guo wrote:
> > > Hi Mani,
> > >
> > > On 11/28/2023 2:00 PM, Manivannan Sadhasivam wrote:
> > > > On Thu, Nov 23, 2023 at 12:46:27AM -0800, Can Guo wrote:
> > > > > Set the initial PHY gear to max HS gear for hosts with HW ver 5 and newer.
> > > > >
> > > >
> > > > MAX_GEAR will be used for hosts with hw_ver.major >= 4
> > >
> > > I put it > 5 because I am not intent to touch any old targets which has
> > > proven working fine with starting with PHY gear HS_G2. If I put it >= 4,
> > > there would be many targets impacted by this change. I need to go back and
> > > test those platforms (HW ver == 4).
> > >
> >
> > This assumption will make the code hard to maintain. I think if you happen to
> > test it on atleast a couple of old targets it should be good since I do not see
> > how others can fail.
>
> Point taken. I will put it >= 4 in next version and test it on platforms
> like SM8350 and SM8450.
>

Appreciated, thanks!

- Mani

> Thanks,
> Can Guo.
>
> >
> > - Mani
> >
> > > Thanks,
> > > Can Guo.
> > >
> > > >
> > > > > This patch is not changing any functionalities or logic but only a
> > > > > preparation patch for the next patch in this series.
> > > > >
> > > > > Signed-off-by: Can Guo <[email protected]>
> > > > > ---
> > > > > drivers/ufs/host/ufs-qcom.c | 21 +++++++++++++++------
> > > > > 1 file changed, 15 insertions(+), 6 deletions(-)
> > > > >
> > > > > diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c
> > > > > index 6756f8d..7bbccf4 100644
> > > > > --- a/drivers/ufs/host/ufs-qcom.c
> > > > > +++ b/drivers/ufs/host/ufs-qcom.c
> > > > > @@ -1067,6 +1067,20 @@ static void ufs_qcom_advertise_quirks(struct ufs_hba *hba)
> > > > > hba->quirks |= UFSHCD_QUIRK_REINIT_AFTER_MAX_GEAR_SWITCH;
> > > > > }
> > > > > +static void ufs_qcom_set_phy_gear(struct ufs_qcom_host *host)
> > > > > +{
> > > > > + struct ufs_host_params *host_params = &host->host_params;
> > > > > +
> > > > > + host->phy_gear = host_params->hs_tx_gear;
> > > > > +
> > > > > + /*
> > > > > + * Power up the PHY using the minimum supported gear (UFS_HS_G2).
> > > > > + * Switching to max gear will be performed during reinit if supported.
> > > >
> > > > You need to reword this comment too.
> > > >
> > > > > + */
> > > > > + if (host->hw_ver.major < 0x5)
> > > >
> > > > As I mentioned above, MAX_GEAR will be used if hw_ver.major is >=4 in
> > > > ufs_qcom_get_hs_gear(). So this check should be (< 0x4).
> > > >
> > > > - Mani
> > > >
> > > > > + host->phy_gear = UFS_HS_G2;
> > > > > +}
> > > > > +
> > > > > static void ufs_qcom_set_host_params(struct ufs_hba *hba)
> > > > > {
> > > > > struct ufs_qcom_host *host = ufshcd_get_variant(hba);
> > > > > @@ -1303,6 +1317,7 @@ static int ufs_qcom_init(struct ufs_hba *hba)
> > > > > ufs_qcom_set_caps(hba);
> > > > > ufs_qcom_advertise_quirks(hba);
> > > > > ufs_qcom_set_host_params(hba);
> > > > > + ufs_qcom_set_phy_gear(host);
> > > > > err = ufs_qcom_ice_init(host);
> > > > > if (err)
> > > > > @@ -1320,12 +1335,6 @@ static int ufs_qcom_init(struct ufs_hba *hba)
> > > > > dev_warn(dev, "%s: failed to configure the testbus %d\n",
> > > > > __func__, err);
> > > > > - /*
> > > > > - * Power up the PHY using the minimum supported gear (UFS_HS_G2).
> > > > > - * Switching to max gear will be performed during reinit if supported.
> > > > > - */
> > > > > - host->phy_gear = UFS_HS_G2;
> > > > > -
> > > > > return 0;
> > > > > out_variant_clear:
> > > > > --
> > > > > 2.7.4
> > > > >
> > > >
> >

--
மணிவண்ணன் சதாசிவம்

2023-11-28 11:24:47

by Manivannan Sadhasivam

[permalink] [raw]

Subject: Re: [PATCH v5 06/10] scsi: ufs: ufs-qcom: Limit HS-G5 Rate-A to hosts with HW version 5

On Tue, Nov 28, 2023 at 06:59:39PM +0800, Can Guo wrote:
>
>
> On 11/28/2023 6:55 PM, Manivannan Sadhasivam wrote:
> > On Tue, Nov 28, 2023 at 03:48:02PM +0800, Can Guo wrote:
> > > Hi Mani,
> > >
> > > On 11/28/2023 1:55 PM, Manivannan Sadhasivam wrote:
> > > > On Thu, Nov 23, 2023 at 12:46:26AM -0800, Can Guo wrote:
> > > > > Qcom UFS hosts, with HW ver 5, can only support up to HS-G5 Rate-A due to
> > > > > HW limitations. If the HS-G5 PHY gear is used, update host_params->hs_rate
> > > > > to Rate-A, so that the subsequent power mode changes shall stick to Rate-A.
> > > > >
> > > > > Signed-off-by: Can Guo <[email protected]>
> > > >
> > > > Reviewed-by: Manivannan Sadhasivam <[email protected]>
> > > >
> > > > One question below...
> > > >
> > > > > ---
> > > > > drivers/ufs/host/ufs-qcom.c | 18 +++++++++++++++++-
> > > > > 1 file changed, 17 insertions(+), 1 deletion(-)
> > > > >
> > > > > diff --git a/drivers/ufs/host/ufs-qcom.c b/drivers/ufs/host/ufs-qcom.c
> > > > > index 9613ad9..6756f8d 100644
> > > > > --- a/drivers/ufs/host/ufs-qcom.c
> > > > > +++ b/drivers/ufs/host/ufs-qcom.c
> > > > > @@ -442,9 +442,25 @@ static u32 ufs_qcom_get_hs_gear(struct ufs_hba *hba)
> > > > > static int ufs_qcom_power_up_sequence(struct ufs_hba *hba)
> > > > > {
> > > > > struct ufs_qcom_host *host = ufshcd_get_variant(hba);
> > > > > + struct ufs_host_params *host_params = &host->host_params;
> > > > > struct phy *phy = host->generic_phy;
> > > > > + enum phy_mode mode;
> > > > > int ret;
> > > > > + /*
> > > > > + * HW ver 5 can only support up to HS-G5 Rate-A due to HW limitations.
> > > > > + * If the HS-G5 PHY gear is used, update host_params->hs_rate to Rate-A,
> > > > > + * so that the subsequent power mode change shall stick to Rate-A.
> > > > > + */
> > > > > + if (host->hw_ver.major == 0x5) {
> > > > > + if (host->phy_gear == UFS_HS_G5)
> > > > > + host_params->hs_rate = PA_HS_MODE_A;
> > > > > + else
> > > > > + host_params->hs_rate = PA_HS_MODE_B;
> > > >
> > > > Is this 'else' part really needed? Since there wouldn't be any 2nd init, I think
> > > > we can skip that.
> > >
> > > We need it because, even there is only one init, if a UFS3.1 device is
> > > attached, phy_gear is given as UFS_HS_G4 in ufs_qcom_set_phy_gear(), hence
> > > we need to put the UFS at HS-G4 Rate B, not Rate A.
> > >
> >
> > But the default hs_rate is PA_HS_MODE_B only and the else condition would be not
> > needed for the 1st init.
>
> You are right, but still we need this in case the UFS device version is not
> populated, meaning dual init can also happen to SM8550. We need to apply the
> right hs_rate in case the 2nd init asks for HS_G4.
>

Hmm, yeah I missed that corner case. This is fine.

- Mani

> Thanks,
> Can Guo.
>
> >
> > - Mani
> >
> > > Thanks,
> > > Can Guo.
> > >
> > > >
> > > > - Mani
> > > >
> > > > > + }
> > > > > +
> > > > > + mode = host_params->hs_rate == PA_HS_MODE_B ? PHY_MODE_UFS_HS_B : PHY_MODE_UFS_HS_A;
> > > > > +
> > > > > /* Reset UFS Host Controller and PHY */
> > > > > ret = ufs_qcom_host_reset(hba);
> > > > > if (ret)
> > > > > @@ -459,7 +475,7 @@ static int ufs_qcom_power_up_sequence(struct ufs_hba *hba)
> > > > > return ret;
> > > > > }
> > > > > - phy_set_mode_ext(phy, PHY_MODE_UFS_HS_B, host->phy_gear);
> > > > > + phy_set_mode_ext(phy, mode, host->phy_gear);
> > > > > /* power on phy - start serdes and phy's power and clocks */
> > > > > ret = phy_power_on(phy);
> > > > > --
> > > > > 2.7.4
> > > > >
> > > >
> >

--
மணிவண்ணன் சதாசிவம்