linux-kernel - Re: [PATCH] iommu/arm-smmu-v3: Poll for CMDQ drain completion more effectively

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20170503153738.GP8233@arm.com>
Date:   Wed, 3 May 2017 16:37:39 +0100
From:   Will Deacon <will.deacon@....com>
To:     Sunil Kovvuri <sunil.kovvuri@...il.com>
Cc:     Robin Murphy <robin.murphy@....com>,
        iommu@...ts.linux-foundation.org,
        LKML <linux-kernel@...r.kernel.org>,
        LAKML <linux-arm-kernel@...ts.infradead.org>,
        Robert Richter <robert.richter@...ium.com>, jcm@...hat.com,
        Sunil Goutham <sgoutham@...ium.com>, Geetha <gakula@...ium.com>
Subject: Re: [PATCH] iommu/arm-smmu-v3: Poll for CMDQ drain completion more
 effectively

On Wed, May 03, 2017 at 06:49:09PM +0530, Sunil Kovvuri wrote:
> On Thu, Apr 27, 2017 at 4:43 PM,  <sunil.kovvuri@...il.com> wrote:
> > From: Sunil Goutham <sgoutham@...ium.com>
> >
> > Modified polling on CMDQ consumer similar to how polling is done for TLB SYNC
> > completion in SMMUv2 driver. Code changes are done with reference to
> >
> > 8513c8930069 iommu/arm-smmu: Poll for TLB sync completion more effectively
> >
> > Poll timeout has been increased which addresses issue of 100us timeout not
> > sufficient, when command queue is full with TLB invalidation commands.
> >
> > Signed-off-by: Sunil Goutham <sgoutham@...ium.com>
> > Signed-off-by: Geetha <gakula@...ium.com>
> > ---
> >  drivers/iommu/arm-smmu-v3.c | 15 ++++++++++++---
> >  1 file changed, 12 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/iommu/arm-smmu-v3.c b/drivers/iommu/arm-smmu-v3.c
> > index d412bdd..34599d4 100644
> > --- a/drivers/iommu/arm-smmu-v3.c
> > +++ b/drivers/iommu/arm-smmu-v3.c
> > @@ -379,6 +379,9 @@
> >  #define CMDQ_SYNC_0_CS_NONE            (0UL << CMDQ_SYNC_0_CS_SHIFT)
> >  #define CMDQ_SYNC_0_CS_SEV             (2UL << CMDQ_SYNC_0_CS_SHIFT)
> >
> > +#define CMDQ_DRAIN_TIMEOUT_US          1000
> > +#define CMDQ_SPIN_COUNT                        10
> > +
> >  /* Event queue */
> >  #define EVTQ_ENT_DWORDS                        4
> >  #define EVTQ_MAX_SZ_SHIFT              7
> > @@ -737,7 +740,8 @@ static void queue_inc_prod(struct arm_smmu_queue *q)
> >   */
> >  static int queue_poll_cons(struct arm_smmu_queue *q, bool drain, bool wfe)
> >  {
> > -       ktime_t timeout = ktime_add_us(ktime_get(), ARM_SMMU_POLL_TIMEOUT_US);
> > +       ktime_t timeout = ktime_add_us(ktime_get(), CMDQ_DRAIN_TIMEOUT_US);
> > +       unsigned int spin_cnt, delay = 1;
> >
> >         while (queue_sync_cons(q), (drain ? !queue_empty(q) : queue_full(q))) {
> >                 if (ktime_compare(ktime_get(), timeout) > 0)
> > @@ -746,8 +750,13 @@ static int queue_poll_cons(struct arm_smmu_queue *q, bool drain, bool wfe)
> >                 if (wfe) {
> >                         wfe();
> >                 } else {
> > -                       cpu_relax();
> > -                       udelay(1);
> > +                       for (spin_cnt = 0;
> > +                            spin_cnt < CMDQ_SPIN_COUNT; spin_cnt++) {
> > +                               cpu_relax();
> > +                               continue;
> > +                       }
> > +                       udelay(delay);
> > +                       delay *= 2;
> >                 }
> >         }
> >
> > --
> > 2.7.4
> >
> 
> Sorry for the ignorance.
> Is there a patchwork where I can check current status of ARM IOMMU
> related patches ?
> 
> And is this patch accepted, if not any comments / feedback ?

Please be patient: the merge window is open and it's not been long since you
posted the patch, which looks pretty bonkers at first glance.

Will