[RFC PATCH v3 04/10] hw/arm/smmuv3: Add page table walk for stage-2

[Mostafa Saleh]

In preparation for adding stage-2 support, add Stage-2 PTW code.
Only Aarch64 format is supported as stage-1.

Nesting stage-1 and stage-2 is not supported right now.

HTTU is not supported, SW is expected to maintain the Access flag.
This is described in the SMMUv3 manual(IHI 0070.E)
"5.2. Stream Table Entry" in "[181] S2AFFD".
This flag determines the behavior on access of a stage-2 page whose
descriptor has AF == 0:
- 0b0: An Access flag fault occurs (stall not supported).
- 0b1: An Access flag fault never occurs.
An Access fault takes priority over a Permission fault.

There are 3 address size checks for stage-2 according to "IHI 0070.E"
in "3.4. Address sizes".
- As nesting is not supported, input address is passed directly to
stage-2, and is checked against IAS.
We use cfg->oas to hold the OAS when stage-1 is not used, this is set
in the next patch.
This check is done outside of smmu_ptw_64_s2 as it is not part of
stage-2(it throws stage-1 fault), and the stage-2 function shouldn't
change it's behavior when nesting is supported.
When nesting is supported and we figure out how to combine TLB for
stage-1 and stage-2 we can move this check into the stage-1 function
as described in ARM DDI0487I.a in pseudocode
aarch64/translation/vmsa_translation/AArch64.S1Translate
aarch64/translation/vmsa_translation/AArch64.S1DisabledOutput

- Input to stage-2 is checked against s2t0sz, and throws stage-2
transaltion fault if exceeds it.

- Output of stage-2 is checked against effective PA output range.

Signed-off-by: Mostafa Saleh <smostafa@google.com>
---
Changes in v3:
- Fix IPA address size check.
- s2cfg.oas renamed to s2cfg.eff_ps.
- s/iova/ipa
- s/ap/s2ap
- s/gran/granule_sz
- use level_shift instead of inline code.
- Add missing brackets in is_permission_fault_s2.
- Use new VMSA_* macros and functions instead of SMMU_*
- Rename pgd_idx to pgd_concat_idx.
- Move SMMU_MAX_S2_CONCAT to STE patch as it is not used here.
Changes in v2:
- Squash S2AFF PTW code.
- Use common functions between stage-1 and stage-2.
- Add checks for IPA and out PA.
---
 hw/arm/smmu-common.c   | 142 ++++++++++++++++++++++++++++++++++++++++-
 hw/arm/smmu-internal.h |  35 ++++++++++
 2 files changed, 176 insertions(+), 1 deletion(-)

diff --git a/hw/arm/smmu-common.c b/hw/arm/smmu-common.c
index 50391a8c94..a4ebd004c5 100644
--- a/hw/arm/smmu-common.c
+++ b/hw/arm/smmu-common.c
@@ -363,6 +363,127 @@ error:
     return -EINVAL;
 }
 
+/**
+ * smmu_ptw_64_s2 - VMSAv8-64 Walk of the page tables for a given ipa
+ * for stage-2.
+ * @cfg: translation config
+ * @ipa: ipa to translate
+ * @perm: access type
+ * @tlbe: SMMUTLBEntry (out)
+ * @info: handle to an error info
+ *
+ * Return 0 on success, < 0 on error. In case of error, @info is filled
+ * and tlbe->perm is set to IOMMU_NONE.
+ * Upon success, @tlbe is filled with translated_addr and entry
+ * permission rights.
+ */
+static int smmu_ptw_64_s2(SMMUTransCfg *cfg,
+                          dma_addr_t ipa, IOMMUAccessFlags perm,
+                          SMMUTLBEntry *tlbe, SMMUPTWEventInfo *info)
+{
+    const int stage = 2;
+    int granule_sz = cfg->s2cfg.granule_sz;
+    /* ARM DDI0487I.a: Table D8-7. */
+    int inputsize = 64 - cfg->s2cfg.tsz;
+    int level = get_start_level(cfg->s2cfg.sl0, granule_sz);
+    int stride = VMSA_STRIDE(granule_sz);
+    int idx = pgd_concat_idx(level, granule_sz, ipa);
+    /*
+     * Get the ttb from concatenated structure.
+     * The offset is the idx * size of each ttb(number of ptes * (sizeof(pte))
+     */
+    uint64_t baseaddr = extract64(cfg->s2cfg.vttb, 0, 48) + (1 << stride) *
+                                  idx * sizeof(uint64_t);
+    dma_addr_t indexmask = VMSA_IDXMSK(inputsize, stride, level);
+
+    baseaddr &= ~indexmask;
+
+    /*
+     * On input, a stage 2 Translation fault occurs if the IPA is outside the
+     * range configured by the relevant S2T0SZ field of the STE.
+     */
+    if (ipa >= (1ULL << inputsize)) {
+        info->type = SMMU_PTW_ERR_TRANSLATION;
+        goto error;
+    }
+
+    while (level < VMSA_LEVELS) {
+        uint64_t subpage_size = 1ULL << level_shift(level, granule_sz);
+        uint64_t mask = subpage_size - 1;
+        uint32_t offset = iova_level_offset(ipa, inputsize, level, granule_sz);
+        uint64_t pte, gpa;
+        dma_addr_t pte_addr = baseaddr + offset * sizeof(pte);
+        uint8_t s2ap;
+
+        if (get_pte(baseaddr, offset, &pte, info)) {
+                goto error;
+        }
+        trace_smmu_ptw_level(stage, level, ipa, subpage_size,
+                             baseaddr, offset, pte);
+        if (is_invalid_pte(pte) || is_reserved_pte(pte, level)) {
+            trace_smmu_ptw_invalid_pte(stage, level, baseaddr,
+                                       pte_addr, offset, pte);
+            break;
+        }
+
+        if (is_table_pte(pte, level)) {
+            baseaddr = get_table_pte_address(pte, granule_sz);
+            level++;
+            continue;
+        } else if (is_page_pte(pte, level)) {
+            gpa = get_page_pte_address(pte, granule_sz);
+            trace_smmu_ptw_page_pte(stage, level, ipa,
+                                    baseaddr, pte_addr, pte, gpa);
+        } else {
+            uint64_t block_size;
+
+            gpa = get_block_pte_address(pte, level, granule_sz,
+                                        &block_size);
+            trace_smmu_ptw_block_pte(stage, level, baseaddr,
+                                     pte_addr, pte, ipa, gpa,
+                                     block_size >> 20);
+        }
+
+        /*
+         * If S2AFFD and PTE.AF are 0 => fault. (5.2. Stream Table Entry)
+         * An Access fault takes priority over a Permission fault.
+         */
+        if (!PTE_AF(pte) && !cfg->s2cfg.affd) {
+            info->type = SMMU_PTW_ERR_ACCESS;
+            goto error;
+        }
+
+        s2ap = PTE_AP(pte);
+        if (is_permission_fault_s2(s2ap, perm)) {
+            info->type = SMMU_PTW_ERR_PERMISSION;
+            goto error;
+        }
+
+        /*
+         * The address output from the translation causes a stage 2 Address
+         * Size fault if it exceeds the effective PA output range.
+         */
+        if (gpa >= (1ULL << cfg->s2cfg.eff_ps)) {
+            info->type = SMMU_PTW_ERR_ADDR_SIZE;
+            goto error;
+        }
+
+        tlbe->entry.translated_addr = gpa;
+        tlbe->entry.iova = ipa & ~mask;
+        tlbe->entry.addr_mask = mask;
+        tlbe->entry.perm = s2ap;
+        tlbe->level = level;
+        tlbe->granule = granule_sz;
+        return 0;
+    }
+    info->type = SMMU_PTW_ERR_TRANSLATION;
+
+error:
+    info->stage = 2;
+    tlbe->entry.perm = IOMMU_NONE;
+    return -EINVAL;
+}
+
 /**
  * smmu_ptw - Walk the page tables for an IOVA, according to @cfg
  *
@@ -377,7 +498,26 @@ error:
 int smmu_ptw(SMMUTransCfg *cfg, dma_addr_t iova, IOMMUAccessFlags perm,
              SMMUTLBEntry *tlbe, SMMUPTWEventInfo *info)
 {
-    return smmu_ptw_64_s1(cfg, iova, perm, tlbe, info);
+    if (cfg->stage == 1) {
+        return smmu_ptw_64_s1(cfg, iova, perm, tlbe, info);
+    } else if (cfg->stage == 2) {
+        /*
+         * If bypassing stage 1(or unimplemented), the input address is passed
+         * directly to stage 2 as IPA. If the input address of a transaction
+         * exceeds the size of the IAS, a stage 1 Address Size fault occurs.
+         * For AA64, IAS = OAS according to (IHI 0070.E) "3.4 Address sizes"
+         */
+        if (iova >= (1ULL << cfg->oas)) {
+            info->type = SMMU_PTW_ERR_ADDR_SIZE;
+            info->stage = 1;
+            tlbe->entry.perm = IOMMU_NONE;
+            return -EINVAL;
+        }
+
+        return smmu_ptw_64_s2(cfg, iova, perm, tlbe, info);
+    }
+
+    g_assert_not_reached();
 }
 
 /**
diff --git a/hw/arm/smmu-internal.h b/hw/arm/smmu-internal.h
index 2d75b31953..a9454f914e 100644
--- a/hw/arm/smmu-internal.h
+++ b/hw/arm/smmu-internal.h
@@ -66,6 +66,8 @@
 #define PTE_APTABLE(pte) \
     (extract64(pte, 61, 2))
 
+#define PTE_AF(pte) \
+    (extract64(pte, 10, 1))
 /*
  * TODO: At the moment all transactions are considered as privileged (EL1)
  * as IOMMU translation callback does not pass user/priv attributes.
@@ -73,6 +75,9 @@
 #define is_permission_fault(ap, perm) \
     (((perm) & IOMMU_WO) && ((ap) & 0x2))
 
+#define is_permission_fault_s2(s2ap, perm) \
+    (!(((s2ap) & (perm)) == (perm)))
+
 #define PTE_AP_TO_PERM(ap) \
     (IOMMU_ACCESS_FLAG(true, !((ap) & 0x2)))
 
@@ -96,6 +101,36 @@ uint64_t iova_level_offset(uint64_t iova, int inputsize,
             MAKE_64BIT_MASK(0, gsz - 3);
 }
 
+/* FEAT_LPA2 and FEAT_TTST are not implemented. */
+static inline int get_start_level(int sl0 , int granule_sz)
+{
+    /* ARM DDI0487I.a: Table D8-12. */
+    if (granule_sz == 12) {
+        return 2 - sl0;
+    }
+    /* ARM DDI0487I.a: Table D8-22 and Table D8-31. */
+    return 3 - sl0;
+}
+
+/*
+ * Index in a concatenated first level stage-2 page table.
+ * ARM DDI0487I.a: D8.2.2 Concatenated translation tables.
+ */
+static inline int pgd_concat_idx(int start_level, int granule_sz,
+                                 dma_addr_t ipa)
+{
+    uint64_t ret;
+    /*
+     * Get the number of bits handled by next levels, then any extra bits in
+     * the address should index the concatenated tables. This relation can be
+     * deduced from tables in ARM DDI0487I.a: D8.2.7-9
+     */
+    int shift =  level_shift(start_level - 1, granule_sz);
+
+    ret = ipa >> shift;
+    return ret;
+}
+
 #define SMMU_IOTLB_ASID(key) ((key).asid)
 
 typedef struct SMMUIOTLBPageInvInfo {
-- 
2.40.0.348.gf938b09366-goog