All of lore.kernel.org
 help / color / mirror / Atom feed
From: Mostafa Saleh <smostafa@google.com>
To: Eric Auger <eric.auger@redhat.com>
Cc: qemu-arm@nongnu.org, peter.maydell@linaro.org,
	qemu-devel@nongnu.org, jean-philippe@linaro.org,
	alex.bennee@linaro.org, maz@kernel.org, nicolinc@nvidia.com,
	julien@xen.org, richard.henderson@linaro.org,
	marcin.juszkiewicz@linaro.org
Subject: Re: [RFC PATCH v3 07/18] hw/arm/smmuv3: Translate CD and TT using stage-2 table
Date: Thu, 16 May 2024 16:11:51 +0000	[thread overview]
Message-ID: <ZkYwR0JeHGhwLm6c@google.com> (raw)
In-Reply-To: <b5808f6e-cead-4291-8274-079ad11a922d@redhat.com>

Hi Eric,

On Wed, May 15, 2024 at 03:15:02PM +0200, Eric Auger wrote:
> Hi Mostafa,
> 
> On 4/29/24 05:23, Mostafa Saleh wrote:
> > According to ARM SMMU architecture specification (ARM IHI 0070 F.b),
> > In "5.2 Stream Table Entry":
> >  [51:6] S1ContextPtr
> >  If Config[1] == 1 (stage 2 enabled), this pointer is an IPA translated by
> >  stage 2 and the programmed value must be within the range of the IAS.
> >
> > In "5.4.1 CD notes":
> >  The translation table walks performed from TTB0 or TTB1 are always performed
> >  in IPA space if stage 2 translations are enabled.
> >
> > This patch implements translation of the S1 context descriptor pointer and
> > TTBx base addresses through the S2 stage (IPA -> PA)
> >
> > smmuv3_do_translate() is updated to have one arg which is translation
> > class, this is useful for:
> s/for/to?
Will do.
> >  - Decide wether a translation is stage-2 only or use the STE config.
> >  - Populate the class in case of faults, WALK_EABT is lefat as it as
> left unchanged?
Yup, that's a typo.
> >    it is always triggered from TT access so no need to use the input
> >    class.
> >
> > In case for stage-2 only translation, which only used in nesting, the
> in case of S2 translation used in the contexted of a nested translation, ...
Will do.
> > stage and asid are saved and restored before and after calling
> > smmu_translate().
> >
> > Translating CD or TTBx can fail for the following reasons:
> > 1) Large address size: This is described in
> >    (3.4.3 Address sizes of SMMU-originated accesses)
> >    - For CD ptr larger than IAS, for SMMUv3.1, it can trigger either
> >      C_BAD_STE or Translation fault, we implement the latter as it
> >      requires no extra code.
> >    - For TTBx, if larger than the effective stage 1 output address size, it
> >      triggers C_BAD_CD.
> >
> > 2) Faults from PTWs (7.3 Event records)
> >    - F_ADDR_SIZE: large address size after first level causes stage 2 Address
> >      Size fault (Also in 3.4.3 Address sizes of SMMU-originated accesses)
> >    - F_PERMISSION: Same as an address translation. However, when
> >      CLASS == CD, the access is implicitly Data and a read.
> >    - F_ACCESS: Same as an address translation.
> >    - F_TRANSLATION: Same as an address translation.
> >    - F_WALK_EABT: Same as an address translation.
> >   These are already implemented in the PTW logic, so no extra handling
> >   required.
> >
> > As, there is multiple locations where the address is calculated from
> > cached entry, a new macro is introduced CACHED_ENTRY_TO_ADDR.
> >
> > Signed-off-by: Mostafa Saleh <smostafa@google.com>
> > ---
> >  hw/arm/smmuv3.c              | 76 ++++++++++++++++++++++++++++++------
> >  include/hw/arm/smmu-common.h |  3 ++
> >  2 files changed, 66 insertions(+), 13 deletions(-)
> >
> > diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
> > index cc61708160..cc61c82321 100644
> > --- a/hw/arm/smmuv3.c
> > +++ b/hw/arm/smmuv3.c
> > @@ -337,14 +337,33 @@ static int smmu_get_ste(SMMUv3State *s, dma_addr_t addr, STE *buf,
> >  
> >  }
> >  
> > +static SMMUTranslationStatus smmuv3_do_translate(SMMUv3State *s, hwaddr addr,
> > +                                                 SMMUTransCfg *cfg,
> > +                                                 SMMUEventInfo *event,
> > +                                                 IOMMUAccessFlags flag,
> > +                                                 SMMUTLBEntry **out_entry,
> > +                                                 SMMUTranslationClass class);
> >  /* @ssid > 0 not supported yet */
> > -static int smmu_get_cd(SMMUv3State *s, STE *ste, uint32_t ssid,
> > -                       CD *buf, SMMUEventInfo *event)
> > +static int smmu_get_cd(SMMUv3State *s, STE *ste, SMMUTransCfg *cfg,
> > +                       uint32_t ssid, CD *buf, SMMUEventInfo *event)
> >  {
> >      dma_addr_t addr = STE_CTXPTR(ste);
> >      int ret, i;
> > +    SMMUTranslationStatus status;
> > +    SMMUTLBEntry *entry;
> >  
> >      trace_smmuv3_get_cd(addr);
> > +
> > +    if (cfg->stage == SMMU_NESTED) {
> > +        status = smmuv3_do_translate(s, addr, cfg, event,
> > +                                     IOMMU_RO, &entry, SMMU_CLASS_CD);
> > +        if (status != SMMU_TRANS_SUCCESS) {
> So I guess you rely on event being populated by above CD S2 translate().
> it does not need to be patched, correct?
> May be worth a comment.
Yes, only the class is different, I will add a comment.
> > +            return -EINVAL;
> > +        }
> > +
> > +        addr = CACHED_ENTRY_TO_ADDR(entry, addr);
> > +    }
> > +
> >      /* TODO: guarantee 64-bit single-copy atomicity */
> >      ret = dma_memory_read(&address_space_memory, addr, buf, sizeof(*buf),
> >                            MEMTXATTRS_UNSPECIFIED);
> > @@ -659,10 +678,13 @@ static int smmu_find_ste(SMMUv3State *s, uint32_t sid, STE *ste,
> >      return 0;
> >  }
> >  
> > -static int decode_cd(SMMUTransCfg *cfg, CD *cd, SMMUEventInfo *event)
> > +static int decode_cd(SMMUv3State *s, SMMUTransCfg *cfg,
> > +                     CD *cd, SMMUEventInfo *event)
> >  {
> >      int ret = -EINVAL;
> >      int i;
> > +    SMMUTranslationStatus status;
> > +    SMMUTLBEntry *entry;
> >  
> >      if (!CD_VALID(cd) || !CD_AARCH64(cd)) {
> >          goto bad_cd;
> > @@ -713,9 +735,21 @@ static int decode_cd(SMMUTransCfg *cfg, CD *cd, SMMUEventInfo *event)
> >  
> >          tt->tsz = tsz;
> >          tt->ttb = CD_TTB(cd, i);
> > +
> >          if (tt->ttb & ~(MAKE_64BIT_MASK(0, cfg->oas))) {
> >              goto bad_cd;
> >          }
> > +
> > +        /* Translate the TTBx, from IPA to PA if nesting is enabled. */
> > +        if (cfg->stage == SMMU_NESTED) {
> > +            status = smmuv3_do_translate(s, tt->ttb, cfg, event, IOMMU_RO,
> > +                                         &entry, SMMU_CLASS_TT);
> > +            if (status != SMMU_TRANS_SUCCESS) {
> same here.
> > +                return -EINVAL;
> > +            }
> > +            tt->ttb = CACHED_ENTRY_TO_ADDR(entry, tt->ttb);
> > +        }
> > +
> >          tt->had = CD_HAD(cd, i);
> >          trace_smmuv3_decode_cd_tt(i, tt->tsz, tt->ttb, tt->granule_sz, tt->had);
> >      }
> > @@ -767,12 +801,12 @@ static int smmuv3_decode_config(IOMMUMemoryRegion *mr, SMMUTransCfg *cfg,
> >          return 0;
> >      }
> >  
> > -    ret = smmu_get_cd(s, &ste, 0 /* ssid */, &cd, event);
> > +    ret = smmu_get_cd(s, &ste, cfg, 0 /* ssid */, &cd, event);
> >      if (ret) {
> >          return ret;
> >      }
> >  
> > -    return decode_cd(cfg, &cd, event);
> > +    return decode_cd(s, cfg, &cd, event);
> >  }
> >  
> >  /**
> > @@ -832,13 +866,29 @@ static SMMUTranslationStatus smmuv3_do_translate(SMMUv3State *s, hwaddr addr,
> >                                                   SMMUTransCfg *cfg,
> >                                                   SMMUEventInfo *event,
> >                                                   IOMMUAccessFlags flag,
> > -                                                 SMMUTLBEntry **out_entry)
> > +                                                 SMMUTLBEntry **out_entry,
> > +                                                 SMMUTranslationClass class)
> >  {
> >      SMMUPTWEventInfo ptw_info = {};
> >      SMMUState *bs = ARM_SMMU(s);
> >      SMMUTLBEntry *cached_entry = NULL;
> > +    int asid, stage;
> > +    bool S2_only = class != SMMU_CLASS_IN;
> > +
> > +    if (S2_only) {
> Please add a comment explaining that class value is used to identify
> S2-only forced translation in the context of a nested translation.
> In that case we hackily override the original config to reach our goal
> and then restore the original config.
> That's pretty hacky. Let's see if any other reviewer has a better idea
> ;-) on my end I understand it and I can bear the trick :)
> 
Indeed, this is unclear, I will add a comment.
I thought about this case a lot, I found this to be the least intrusive
way while having tolerable readability, it'd be great if someone has a
better idea.

> 
> > +        asid = cfg->asid;
> > +        stage = cfg->stage;
> > +        cfg->asid = -1;
> > +        cfg->stage = SMMU_STAGE_2;
> > +    }
> >  
> >      cached_entry = smmu_translate(bs, cfg, addr, flag, &ptw_info);
> > +
> > +    if (S2_only) {
> > +        cfg->asid = asid;
> > +        cfg->stage = stage;
> > +    }
> > +
> >      if (!cached_entry) {
> >          /* All faults from PTW has S2 field. */
> >          event->u.f_walk_eabt.s2 = (ptw_info.stage == SMMU_STAGE_2);
> > @@ -855,7 +905,7 @@ static SMMUTranslationStatus smmuv3_do_translate(SMMUv3State *s, hwaddr addr,
> >                  event->type = SMMU_EVT_F_TRANSLATION;
> >                  event->u.f_translation.addr = addr;
> >                  event->u.f_translation.addr2 = ptw_info.addr;
> > -                event->u.f_translation.class = SMMU_CLASS_IN;
> > +                event->u.f_translation.class = class;
> >                  event->u.f_translation.rnw = flag & 0x1;
> >              }
> >              break;
> > @@ -864,7 +914,7 @@ static SMMUTranslationStatus smmuv3_do_translate(SMMUv3State *s, hwaddr addr,
> >                  event->type = SMMU_EVT_F_ADDR_SIZE;
> >                  event->u.f_addr_size.addr = addr;
> >                  event->u.f_addr_size.addr2 = ptw_info.addr;
> > -                event->u.f_addr_size.class = SMMU_CLASS_IN;
> > +                event->u.f_addr_size.class = class;
> >                  event->u.f_addr_size.rnw = flag & 0x1;
> >              }
> >              break;
> > @@ -873,7 +923,7 @@ static SMMUTranslationStatus smmuv3_do_translate(SMMUv3State *s, hwaddr addr,
> >                  event->type = SMMU_EVT_F_ACCESS;
> >                  event->u.f_access.addr = addr;
> >                  event->u.f_access.addr2 = ptw_info.addr;
> > -                event->u.f_access.class = SMMU_CLASS_IN;
> > +                event->u.f_access.class = class;
> >                  event->u.f_access.rnw = flag & 0x1;
> >              }
> >              break;
> > @@ -882,7 +932,7 @@ static SMMUTranslationStatus smmuv3_do_translate(SMMUv3State *s, hwaddr addr,
> >                  event->type = SMMU_EVT_F_PERMISSION;
> >                  event->u.f_permission.addr = addr;
> >                  event->u.f_permission.addr2 = ptw_info.addr;
> > -                event->u.f_permission.class = SMMU_CLASS_IN;
> > +                event->u.f_permission.class = class;
> >                  event->u.f_permission.rnw = flag & 0x1;
> >              }
> >              break;
> > @@ -943,15 +993,15 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
> >          goto epilogue;
> >      }
> >  
> > -    status = smmuv3_do_translate(s, addr, cfg, &event, flag, &cached_entry);
> > +    status = smmuv3_do_translate(s, addr, cfg, &event, flag,
> > +                                 &cached_entry, SMMU_CLASS_IN);
> >  
> >  epilogue:
> >      qemu_mutex_unlock(&s->mutex);
> >      switch (status) {
> >      case SMMU_TRANS_SUCCESS:
> >          entry.perm = cached_entry->entry.perm;
> > -        entry.translated_addr = cached_entry->entry.translated_addr +
> > -                                    (addr & cached_entry->entry.addr_mask);
> > +        entry.translated_addr = CACHED_ENTRY_TO_ADDR(cached_entry, addr);
> >          entry.addr_mask = cached_entry->entry.addr_mask;
> >          trace_smmuv3_translate_success(mr->parent_obj.name, sid, addr,
> >                                         entry.translated_addr, entry.perm,
> > diff --git a/include/hw/arm/smmu-common.h b/include/hw/arm/smmu-common.h
> > index 96eb017e50..09d3b9e734 100644
> > --- a/include/hw/arm/smmu-common.h
> > +++ b/include/hw/arm/smmu-common.h
> > @@ -37,6 +37,9 @@
> >  #define VMSA_IDXMSK(isz, strd, lvl)         ((1ULL << \
> >                                               VMSA_BIT_LVL(isz, strd, lvl)) - 1)
> >  
> > +#define CACHED_ENTRY_TO_ADDR(ent, addr)      (ent)->entry.translated_addr + \
> > +                                             ((addr) & (ent)->entry.addr_mask);
> > +
> nit; this could be introduced in a separate patch since you have a
> caller in smmuv3_translate(). This may help the reviewer to focus on the
> most important class related changes.

Sure, I can introduce a small patch before this one with this macro.

Thanks,
Mostafa

> Besides
> Reviewed-by: Eric Auger <eric.auger@redhat.com>
> 
> Eric
> 
> >  /*
> >   * Page table walk error types
> >   */
> 

  reply	other threads:[~2024-05-16 16:11 UTC|newest]

Thread overview: 57+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-29  3:23 [RFC PATCH v3 00/18] SMMUv3 nested translation support Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 01/18] hw/arm/smmu-common: Add missing size check for stage-1 Mostafa Saleh
2024-05-13 11:30   ` Eric Auger
2024-04-29  3:23 ` [RFC PATCH v3 02/18] hw/arm/smmu: Fix IPA for stage-2 events Mostafa Saleh
2024-05-13 11:47   ` Eric Auger
2024-05-16 14:43     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 03/18] hw/arm/smmuv3: Fix encoding of CLASS in events Mostafa Saleh
2024-05-15 12:27   ` Eric Auger
2024-05-16 14:50     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 04/18] hw/arm/smmu: Use enum for SMMU stage Mostafa Saleh
2024-05-15 20:25   ` Alex Bennée
2024-04-29  3:23 ` [RFC PATCH v3 05/18] hw/arm/smmu: Split smmuv3_translate() Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 06/18] hw/arm/smmu: Consolidate ASID and VMID types Mostafa Saleh
2024-05-15 12:41   ` Eric Auger
2024-06-17 14:55     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 07/18] hw/arm/smmuv3: Translate CD and TT using stage-2 table Mostafa Saleh
2024-05-15 13:15   ` Eric Auger
2024-05-16 16:11     ` Mostafa Saleh [this message]
2024-04-29  3:23 ` [RFC PATCH v3 08/18] hw/arm/smmu-common: Add support for nested TLB Mostafa Saleh
2024-05-15 13:48   ` Eric Auger
2024-05-16 15:20     ` Mostafa Saleh
2024-05-20  8:20       ` Eric Auger
2024-05-22 12:44         ` Mostafa Saleh
2024-06-17 14:56           ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 09/18] hw/arm/smmu-common: Rework TLB lookup for nesting Mostafa Saleh
2024-05-15 13:54   ` Eric Auger
2024-05-16 15:30     ` Mostafa Saleh
2024-05-20  8:27   ` Eric Auger
2024-05-22 12:47     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 10/18] hw/arm/smmu-common: Support nested translation Mostafa Saleh
2024-05-20  9:48   ` Eric Auger
2024-06-17 14:57     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 11/18] hw/arm/smmu: Support nesting in smmuv3_range_inval() Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 12/18] hw/arm/smmu: Support nesting in the rest of commands Mostafa Saleh
2024-05-20 10:24   ` Eric Auger
2024-06-17 14:58     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 13/18] hw/arm/smmuv3: Support nested SMMUs in smmuv3_notify_iova() Mostafa Saleh
2024-05-20 10:37   ` Eric Auger
2024-06-17 15:02     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 14/18] hw/arm/smmuv3: Support and advertise nesting Mostafa Saleh
2024-05-20 13:16   ` Eric Auger
2024-06-17 15:03     ` Mostafa Saleh
2024-04-29  3:23 ` [RFC PATCH v3 15/18] hw/arm/smmuv3: Advertise S2FWB Mostafa Saleh
2024-05-20 13:30   ` Eric Auger
2024-06-17 15:04     ` Mostafa Saleh
2024-04-29  3:24 ` [RFC PATCH v3 16/18] hw/arm/smmu: Refactor SMMU OAS Mostafa Saleh
2024-05-20 13:59   ` Eric Auger
2024-04-29  3:24 ` [RFC PATCH v3 17/18] hw/arm/smmuv3: Add property for OAS Mostafa Saleh
2024-05-21  9:32   ` Eric Auger
2024-06-27 11:50     ` Mostafa Saleh
2024-04-29  3:24 ` [RFC PATCH v3 18/18] hw/arm/virt: Set SMMU OAS based on CPU PARANGE Mostafa Saleh
2024-05-21  9:43   ` Eric Auger
2024-05-24 17:22   ` Julien Grall
2024-06-27 11:44     ` Mostafa Saleh
2024-05-13 13:57 ` [RFC PATCH v3 00/18] SMMUv3 nested translation support Julien Grall
2024-05-21  9:47 ` Eric Auger
2024-05-27 16:12   ` Mostafa Saleh

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=ZkYwR0JeHGhwLm6c@google.com \
    --to=smostafa@google.com \
    --cc=alex.bennee@linaro.org \
    --cc=eric.auger@redhat.com \
    --cc=jean-philippe@linaro.org \
    --cc=julien@xen.org \
    --cc=marcin.juszkiewicz@linaro.org \
    --cc=maz@kernel.org \
    --cc=nicolinc@nvidia.com \
    --cc=peter.maydell@linaro.org \
    --cc=qemu-arm@nongnu.org \
    --cc=qemu-devel@nongnu.org \
    --cc=richard.henderson@linaro.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.