Re: [RFC PATCH v2 2/5] scsi: ufs: Add UFS-feature layer
From: Alim Akhtar
Date: Wed Jun 17 2020 - 04:57:57 EST
On Wed, Jun 17, 2020 at 12:27 PM Avri Altman <Avri.Altman@xxxxxxx> wrote:
>
> >
> > Hi, Bean
> > >
> > > On Mon, 2020-06-15 at 16:23 +0900, Daejun Park wrote:
> > > > +void ufsf_scan_features(struct ufs_hba *hba)
> > > > +{
> > > > + int ret;
> > > > +
> > > > + init_waitqueue_head(&hba->ufsf.sdev_wait);
> > > > + atomic_set(&hba->ufsf.slave_conf_cnt, 0);
> > > > +
> > > > + if (hba->dev_info.wspecversion >= HPB_SUPPORTED_VERSION &&
> > > > + (hba->dev_info.b_ufs_feature_sup & UFS_DEV_HPB_SUPPORT))
> > >
> > > How about removing this check "(hba->dev_info.wspecversion >=
> > > HPB_SUPPORTED_VERSION" since ufs with lower version than v3.1 can add
> > > HPB feature by FFU,
> > > if (hba->dev_info.b_ufs_feature_sup &UFS_FEATURE_SUPPORT_HPB_BIT) is
> > > enough.
> > OK, changing it seems no problem. But I want to know what other people
> > think
> > about this version checking code.
> HPB1.0 isn't part of ufs3.1, but published only later.
> Allowing earlier versions will required to quirk the descriptor sizes.
> I see Bean's point here, but I vote for adding it in a single quirk, when the time comes.
>
I second Avri here, older devices need a quirk to handle, let do that
as a separate patch.
> Thanks,
> Avri
--
Regards,
Alim