From: Eric Auger <eric.auger@redhat.com>
To: Mostafa Saleh <smostafa@google.com>, qemu-devel@nongnu.org
Cc: jean-philippe@linaro.org, peter.maydell@linaro.org,
qemu-arm@nongnu.org, richard.henderson@linaro.org
Subject: Re: [RFC PATCH v2 06/11] hw/arm/smmuv3: Make TLB lookup work for stage-2
Date: Mon, 20 Mar 2023 17:05:31 +0100 [thread overview]
Message-ID: <45961bdc-cd3a-c35a-8f10-bb9a52dd2e33@redhat.com> (raw)
In-Reply-To: <20230226220650.1480786-7-smostafa@google.com>
Hi Mostafa,
On 2/26/23 23:06, Mostafa Saleh wrote:
> Right now, either stage-1 or stage-2 are supported, this simplifies
> how we can deal with TLBs.
> This patch makes TLB lookup work if stage-2 is enabled instead of
> stage-1.
> TLB lookup is done before a PTW, if a valid entry is found we won't
> do the PTW.
> To be able to do TLB lookup, we need the correct tagging info, as
> granularity and input size, so we get this based on the supported
> translation stage. The TLB entries are added correctly from each
> stage PTW.
>
> When nested translation is supported, this would need to change, for
> example if we go with a combined TLB implementation, we would need to
> use the min of the granularities in TLB.
>
> As stage-2 shouldn't be tagged by ASID, it will be set to -1 if S1P
> is not enabled.
>
> Signed-off-by: Mostafa Saleh <smostafa@google.com>
> ---
> Changes in v2:
> - check if S1 is enabled(not supported) when reading S1 TT.
> ---
> hw/arm/smmuv3.c | 45 ++++++++++++++++++++++++++++++++++-----------
> 1 file changed, 34 insertions(+), 11 deletions(-)
>
> diff --git a/hw/arm/smmuv3.c b/hw/arm/smmuv3.c
> index dc74a5442d..ce193e9598 100644
> --- a/hw/arm/smmuv3.c
> +++ b/hw/arm/smmuv3.c
> @@ -697,6 +697,9 @@ static int smmuv3_decode_config(IOMMUMemoryRegion *mr, SMMUTransCfg *cfg,
> STE ste;
> CD cd;
>
> + /* ASID defaults to -1 (if s1 is not supported). */
> + cfg->asid = -1;
> +
> ret = smmu_find_ste(s, sid, &ste, event);
> if (ret) {
> return ret;
> @@ -787,6 +790,7 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
> SMMUTLBEntry *cached_entry = NULL;
> SMMUTransTableInfo *tt;
> SMMUTransCfg *cfg = NULL;
> + uint8_t granule_sz, tsz;
> IOMMUTLBEntry entry = {
> .target_as = &address_space_memory,
> .iova = addr,
> @@ -822,21 +826,40 @@ static IOMMUTLBEntry smmuv3_translate(IOMMUMemoryRegion *mr, hwaddr addr,
> goto epilogue;
> }
>
> - tt = select_tt(cfg, addr);
> - if (!tt) {
> - if (cfg->record_faults) {
> - event.type = SMMU_EVT_F_TRANSLATION;
> - event.u.f_translation.addr = addr;
> - event.u.f_translation.rnw = flag & 0x1;
> + if (cfg->stage == 1) {
> + /* Select stage1 translation table. */
> + tt = select_tt(cfg, addr);
> + if (!tt) {
> + if (cfg->record_faults) {
> + event.type = SMMU_EVT_F_TRANSLATION;
> + event.u.f_translation.addr = addr;
> + event.u.f_translation.rnw = flag & 0x1;
> + }
> + status = SMMU_TRANS_ERROR;
> + goto epilogue;
> }
> - status = SMMU_TRANS_ERROR;
> - goto epilogue;
> - }
> + granule_sz = tt->granule_sz;
> + tsz = tt->tsz;
>
> - page_mask = (1ULL << (tt->granule_sz)) - 1;
> + } else {
> + /* Stage2. */
> + granule_sz = cfg->s2cfg.granule_sz;
> + tsz = cfg->s2cfg.tsz;
> + }
> + /*
> + * TLB lookup looks for granule and input size for a translation stage,
> + * as only one stage is supported right now, choose the right values
> + * from the configuration.
> + */
> + page_mask = (1ULL << granule_sz) - 1;
> aligned_addr = addr & ~page_mask;
>
> - cached_entry = smmu_iotlb_lookup(bs, cfg, tt, aligned_addr);
> + SMMUTransTableInfo temp = {
Move the declaration at the top. Also rename temp into tt to be more
explicit about what it is?
> + .granule_sz = granule_sz,
> + .tsz = tsz,
> + };
> +
> + cached_entry = smmu_iotlb_lookup(bs, cfg, &temp, aligned_addr);
> if (cached_entry) {
> if ((flag & IOMMU_WO) && !(cached_entry->entry.perm & IOMMU_WO)) {
> status = SMMU_TRANS_ERROR;
Besides, looks good to me
Reviewed-by: Eric Auger <eric.auger@redhat.com>
Thanks
Eric
next prev parent reply other threads:[~2023-03-20 16:06 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-26 22:06 [RFC PATCH v2 00/11] Add stage-2 translation for SMMUv3 Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 01/11] hw/arm/smmuv3: Add missing fields for IDR0 Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 02/11] hw/arm/smmuv3: Update translation config to hold stage-2 Mostafa Saleh
2023-03-17 11:37 ` Eric Auger
2023-03-17 14:43 ` Mostafa Saleh
2023-03-17 17:36 ` Eric Auger
2023-02-26 22:06 ` [RFC PATCH v2 03/11] hw/arm/smmuv3: Refactor stage-1 PTW Mostafa Saleh
2023-03-17 18:31 ` Eric Auger
2023-03-19 8:38 ` Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 04/11] hw/arm/smmuv3: Add page table walk for stage-2 Mostafa Saleh
2023-03-20 14:56 ` Eric Auger
2023-03-20 18:52 ` Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 05/11] hw/arm/smmuv3: Parse STE config " Mostafa Saleh
2023-03-20 15:14 ` Eric Auger
2023-03-20 19:11 ` Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 06/11] hw/arm/smmuv3: Make TLB lookup work " Mostafa Saleh
2023-03-20 16:05 ` Eric Auger [this message]
2023-03-20 19:14 ` Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 07/11] hw/arm/smmuv3: Add VMID to tlb tagging Mostafa Saleh
2023-03-20 16:16 ` Eric Auger
2023-02-26 22:06 ` [RFC PATCH v2 08/11] hw/arm/smmuv3: Add CMDs related to stage-2 Mostafa Saleh
2023-03-20 16:51 ` Eric Auger
2023-03-20 19:29 ` Mostafa Saleh
2023-02-26 22:06 ` [RFC PATCH v2 09/11] hw/arm/smmuv3: Add stage-2 support in iova notifier Mostafa Saleh
2023-03-20 16:57 ` Eric Auger
2023-02-26 22:06 ` [RFC PATCH v2 10/11] hw/arm/smmuv3: Populate OAS based on CPU PARANGE Mostafa Saleh
2023-03-20 17:12 ` Eric Auger
2023-03-21 13:06 ` Mostafa Saleh
2023-03-21 13:23 ` Eric Auger
2023-03-21 13:29 ` Mostafa Saleh
2023-03-21 13:34 ` Eric Auger
2023-03-21 13:34 ` Peter Maydell
2023-03-21 13:42 ` Mostafa Saleh
2023-03-21 13:45 ` Eric Auger
2023-03-21 13:54 ` Mostafa Saleh
2023-03-21 14:08 ` Peter Maydell
2023-02-26 22:06 ` [RFC PATCH v2 11/11] hw/arm/smmuv3: Add knob to choose translation stage and enable stage-2 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=45961bdc-cd3a-c35a-8f10-bb9a52dd2e33@redhat.com \
--to=eric.auger@redhat.com \
--cc=jean-philippe@linaro.org \
--cc=peter.maydell@linaro.org \
--cc=qemu-arm@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.org \
--cc=smostafa@google.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).