RE: [PATCH v2 2/9] scsi: ufshpb: Add host control mode support to rsp_upiu

From: Avri Altman
Date: Tue Feb 02 2021 - 06:56:08 EST

Next message: Oscar Salvador: "Re: [RFC][PATCH 08/13] mm/migrate: demote pages during reclaim"
Previous message: Greg KH: "Re: [PATCH v1 1/1] Firstly, as Andy mentioned, this should be smp_rmb() instead of rmb(). considering that TSYNC is a cross-thread situation, and rmb() is a mandatory barrier which should not be used to control SMP effects, since mandatory barriers impose unnecessary overhead on both SMP and UP systems, as kernel Documentation said."
In reply to: Greg KH: "Re: [PATCH v2 2/9] scsi: ufshpb: Add host control mode support to rsp_upiu"
Next in thread: Avri Altman: "[PATCH v2 4/9] scsi: ufshpb: Make eviction depends on region's reads"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

> On Tue, Feb 02, 2021 at 11:24:04AM +0000, Avri Altman wrote:
> >
> > >
> > > On Tue, Feb 02, 2021 at 10:30:00AM +0200, Avri Altman wrote:
> > > > diff --git a/drivers/scsi/ufs/ufshpb.h b/drivers/scsi/ufs/ufshpb.h
> > > > index afeb6365daf8..5ec4023db74d 100644
> > > > --- a/drivers/scsi/ufs/ufshpb.h
> > > > +++ b/drivers/scsi/ufs/ufshpb.h
> > > > @@ -48,6 +48,11 @@ enum UFSHPB_MODE {
> > > > HPB_DEVICE_CONTROL,
> > > > };
> > > >
> > > > +enum HPB_RGN_FLAGS {
> > > > + RGN_FLAG_UPDATE = 0,
> > > > + RGN_FLAG_DIRTY,
> > > > +};
> > > > +
> > > > enum UFSHPB_STATE {
> > > > HPB_PRESENT = 1,
> > > > HPB_SUSPEND,
> > > > @@ -109,6 +114,7 @@ struct ufshpb_region {
> > > >
> > > > /* below information is used by lru */
> > > > struct list_head list_lru_rgn;
> > > > + unsigned long rgn_flags;
> > >
> > > Why an unsigned long for a simple enumerated type? And why not make
> > > this "type safe" by explicitly saying this is an enumerated type
> > > variable?
> > I am using it for atomic bit operations.
>
> That's not obvious given you have an enumerated type above. Seems like
> an odd mix...
Done.
Will make it clear that those are bit indices.