Merge tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux

Pull arm64 updates from Will Deacon:
 "A sizeable pile of arm64 updates for 5.8.

  Summary below, but the big two features are support for Branch Target
  Identification and Clang's Shadow Call stack. The latter is currently
  arm64-only, but the high-level parts are all in core code so it could
  easily be adopted by other architectures pending toolchain support

  Branch Target Identification (BTI):

   - Support for ARMv8.5-BTI in both user- and kernel-space. This allows
     branch targets to limit the types of branch from which they can be
     called and additionally prevents branching to arbitrary code,
     although kernel support requires a very recent toolchain.

   - Function annotation via SYM_FUNC_START() so that assembly functions
     are wrapped with the relevant "landing pad" instructions.

   - BPF and vDSO updates to use the new instructions.

   - Addition of a new HWCAP and exposure of BTI capability to userspace
     via ID register emulation, along with ELF loader support for the
     BTI feature in .note.gnu.property.

   - Non-critical fixes to CFI unwind annotations in the sigreturn
     trampoline.

  Shadow Call Stack (SCS):

   - Support for Clang's Shadow Call Stack feature, which reserves
     platform register x18 to point at a separate stack for each task
     that holds only return addresses. This protects function return
     control flow from buffer overruns on the main stack.

   - Save/restore of x18 across problematic boundaries (user-mode,
     hypervisor, EFI, suspend, etc).

   - Core support for SCS, should other architectures want to use it
     too.

   - SCS overflow checking on context-switch as part of the existing
     stack limit check if CONFIG_SCHED_STACK_END_CHECK=y.

  CPU feature detection:

   - Removed numerous "SANITY CHECK" errors when running on a system
     with mismatched AArch32 support at EL1. This is primarily a concern
     for KVM, which disabled support for 32-bit guests on such a system.

   - Addition of new ID registers and fields as the architecture has
     been extended.

  Perf and PMU drivers:

   - Minor fixes and cleanups to system PMU drivers.

  Hardware errata:

   - Unify KVM workarounds for VHE and nVHE configurations.

   - Sort vendor errata entries in Kconfig.

  Secure Monitor Call Calling Convention (SMCCC):

   - Update to the latest specification from Arm (v1.2).

   - Allow PSCI code to query the SMCCC version.

  Software Delegated Exception Interface (SDEI):

   - Unexport a bunch of unused symbols.

   - Minor fixes to handling of firmware data.

  Pointer authentication:

   - Add support for dumping the kernel PAC mask in vmcoreinfo so that
     the stack can be unwound by tools such as kdump.

   - Simplification of key initialisation during CPU bringup.

  BPF backend:

   - Improve immediate generation for logical and add/sub instructions.

  vDSO:

   - Minor fixes to the linker flags for consistency with other
     architectures and support for LLVM's unwinder.

   - Clean up logic to initialise and map the vDSO into userspace.

  ACPI:

   - Work around for an ambiguity in the IORT specification relating to
     the "num_ids" field.

   - Support _DMA method for all named components rather than only PCIe
     root complexes.

   - Minor other IORT-related fixes.

  Miscellaneous:

   - Initialise debug traps early for KGDB and fix KDB cacheflushing
     deadlock.

   - Minor tweaks to early boot state (documentation update, set
     TEXT_OFFSET to 0x0, increase alignment of PE/COFF sections).

   - Refactoring and cleanup"

* tag 'arm64-upstream' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux: (148 commits)
  KVM: arm64: Move __load_guest_stage2 to kvm_mmu.h
  KVM: arm64: Check advertised Stage-2 page size capability
  arm64/cpufeature: Add get_arm64_ftr_reg_nowarn()
  ACPI/IORT: Remove the unused __get_pci_rid()
  arm64/cpuinfo: Add ID_MMFR4_EL1 into the cpuinfo_arm64 context
  arm64/cpufeature: Add remaining feature bits in ID_AA64PFR1 register
  arm64/cpufeature: Add remaining feature bits in ID_AA64PFR0 register
  arm64/cpufeature: Add remaining feature bits in ID_AA64ISAR0 register
  arm64/cpufeature: Add remaining feature bits in ID_MMFR4 register
  arm64/cpufeature: Add remaining feature bits in ID_PFR0 register
  arm64/cpufeature: Introduce ID_MMFR5 CPU register
  arm64/cpufeature: Introduce ID_DFR1 CPU register
  arm64/cpufeature: Introduce ID_PFR2 CPU register
  arm64/cpufeature: Make doublelock a signed feature in ID_AA64DFR0
  arm64/cpufeature: Drop TraceFilt feature exposure from ID_DFR0 register
  arm64/cpufeature: Add explicit ftr_id_isar0[] for ID_ISAR0 register
  arm64: mm: Add asid_gen_match() helper
  firmware: smccc: Fix missing prototype warning for arm_smccc_version_init
  arm64: vdso: Fix CFI directives in sigreturn trampoline
  arm64: vdso: Don't prefix sigreturn trampoline with a BTI C instruction
  ...

drivers/acpi/arm64/gtdt.c

@@ -394,7 +394,7 @@ static int __init gtdt_sbsa_gwdt_init(void)
 	 */
 	ret = acpi_gtdt_init(table, &timer_count);
 	if (ret || !timer_count)
-		return ret;
+		goto out_put_gtdt;
 
 	for_each_platform_timer(platform_timer) {
 		if (is_non_secure_watchdog(platform_timer)) {
@@ -408,6 +408,8 @@ static int __init gtdt_sbsa_gwdt_init(void)
 	if (gwdt_count)
 		pr_info("found %d SBSA generic Watchdog(s).\n", gwdt_count);
 
+out_put_gtdt:
+	acpi_put_table(table);
 	return ret;
 }
 

drivers/acpi/arm64/iort.c

@@ -299,61 +299,8 @@ out:
 	return status;
 }
 
-struct iort_workaround_oem_info {
-	char oem_id[ACPI_OEM_ID_SIZE + 1];
-	char oem_table_id[ACPI_OEM_TABLE_ID_SIZE + 1];
-	u32 oem_revision;
-};
-
-static bool apply_id_count_workaround;
-
-static struct iort_workaround_oem_info wa_info[] __initdata = {
-	{
-		.oem_id		= "HISI  ",
-		.oem_table_id	= "HIP07   ",
-		.oem_revision	= 0,
-	}, {
-		.oem_id		= "HISI  ",
-		.oem_table_id	= "HIP08   ",
-		.oem_revision	= 0,
-	}
-};
-
-static void __init
-iort_check_id_count_workaround(struct acpi_table_header *tbl)
-{
-	int i;
-
-	for (i = 0; i < ARRAY_SIZE(wa_info); i++) {
-		if (!memcmp(wa_info[i].oem_id, tbl->oem_id, ACPI_OEM_ID_SIZE) &&
-		    !memcmp(wa_info[i].oem_table_id, tbl->oem_table_id, ACPI_OEM_TABLE_ID_SIZE) &&
-		    wa_info[i].oem_revision == tbl->oem_revision) {
-			apply_id_count_workaround = true;
-			pr_warn(FW_BUG "ID count for ID mapping entry is wrong, applying workaround\n");
-			break;
-		}
-	}
-}
-
-static inline u32 iort_get_map_max(struct acpi_iort_id_mapping *map)
-{
-	u32 map_max = map->input_base + map->id_count;
-
-	/*
-	 * The IORT specification revision D (Section 3, table 4, page 9) says
-	 * Number of IDs = The number of IDs in the range minus one, but the
-	 * IORT code ignored the "minus one", and some firmware did that too,
-	 * so apply a workaround here to keep compatible with both the spec
-	 * compliant and non-spec compliant firmwares.
-	 */
-	if (apply_id_count_workaround)
-		map_max--;
-
-	return map_max;
-}
-
 static int iort_id_map(struct acpi_iort_id_mapping *map, u8 type, u32 rid_in,
-		       u32 *rid_out)
+		       u32 *rid_out, bool check_overlap)
 {
 	/* Single mapping does not care for input id */
 	if (map->flags & ACPI_IORT_ID_SINGLE_MAPPING) {
@@ -368,10 +315,37 @@ static int iort_id_map(struct acpi_iort_id_mapping *map, u8 type, u32 rid_in,
 		return -ENXIO;
 	}
 
-	if (rid_in < map->input_base || rid_in > iort_get_map_max(map))
+	if (rid_in < map->input_base ||
+	    (rid_in > map->input_base + map->id_count))
 		return -ENXIO;
 
+	if (check_overlap) {
+		/*
+		 * We already found a mapping for this input ID at the end of
+		 * another region. If it coincides with the start of this
+		 * region, we assume the prior match was due to the off-by-1
+		 * issue mentioned below, and allow it to be superseded.
+		 * Otherwise, things are *really* broken, and we just disregard
+		 * duplicate matches entirely to retain compatibility.
+		 */
+		pr_err(FW_BUG "[map %p] conflicting mapping for input ID 0x%x\n",
+		       map, rid_in);
+		if (rid_in != map->input_base)
+			return -ENXIO;
+
+		pr_err(FW_BUG "applying workaround.\n");
+	}
+
 	*rid_out = map->output_base + (rid_in - map->input_base);
+
+	/*
+	 * Due to confusion regarding the meaning of the id_count field (which
+	 * carries the number of IDs *minus 1*), we may have to disregard this
+	 * match if it is at the end of the range, and overlaps with the start
+	 * of another one.
+	 */
+	if (map->id_count > 0 && rid_in == map->input_base + map->id_count)
+		return -EAGAIN;
 	return 0;
 }
 
@@ -414,6 +388,7 @@ static struct acpi_iort_node *iort_node_get_id(struct acpi_iort_node *node,
 static int iort_get_id_mapping_index(struct acpi_iort_node *node)
 {
 	struct acpi_iort_smmu_v3 *smmu;
+	struct acpi_iort_pmcg *pmcg;
 
 	switch (node->type) {
 	case ACPI_IORT_NODE_SMMU_V3:
@@ -441,6 +416,10 @@ static int iort_get_id_mapping_index(struct acpi_iort_node *node)
 
 		return smmu->id_mapping_index;
 	case ACPI_IORT_NODE_PMCG:
+		pmcg = (struct acpi_iort_pmcg *)node->node_data;
+		if (pmcg->overflow_gsiv || node->mapping_count == 0)
+			return -EINVAL;
+
 		return 0;
 	default:
 		return -EINVAL;
@@ -456,7 +435,8 @@ static struct acpi_iort_node *iort_node_map_id(struct acpi_iort_node *node,
 	/* Parse the ID mapping tree to find specified node type */
 	while (node) {
 		struct acpi_iort_id_mapping *map;
-		int i, index;
+		int i, index, rc = 0;
+		u32 out_ref = 0, map_id = id;
 
 		if (IORT_TYPE_MASK(node->type) & type_mask) {
 			if (id_out)
@@ -490,15 +470,18 @@ static struct acpi_iort_node *iort_node_map_id(struct acpi_iort_node *node,
 			if (i == index)
 				continue;
 
-			if (!iort_id_map(map, node->type, id, &id))
+			rc = iort_id_map(map, node->type, map_id, &id, out_ref);
+			if (!rc)
 				break;
+			if (rc == -EAGAIN)
+				out_ref = map->output_reference;
 		}
 
-		if (i == node->mapping_count)
+		if (i == node->mapping_count && !out_ref)
 			goto fail_map;
 
 		node = ACPI_ADD_PTR(struct acpi_iort_node, iort_table,
-				    map->output_reference);
+				    rc ? out_ref : map->output_reference);
 	}
 
 fail_map:
@@ -789,15 +772,6 @@ void acpi_configure_pmsi_domain(struct device *dev)
 		dev_set_msi_domain(dev, msi_domain);
 }
 
-static int __maybe_unused __get_pci_rid(struct pci_dev *pdev, u16 alias,
-					void *data)
-{
-	u32 *rid = data;
-
-	*rid = alias;
-	return 0;
-}
-
 #ifdef CONFIG_IOMMU_API
 static struct acpi_iort_node *iort_get_msi_resv_iommu(struct device *dev)
 {
@@ -1148,13 +1122,10 @@ void iort_dma_setup(struct device *dev, u64 *dma_addr, u64 *dma_size)
 	else
 		size = 1ULL << 32;
 
-	if (dev_is_pci(dev)) {
-		ret = acpi_dma_get_range(dev, &dmaaddr, &offset, &size);
-		if (ret == -ENODEV)
-			ret = rc_dma_get_range(dev, &size);
-	} else {
-		ret = nc_dma_get_range(dev, &size);
-	}
+	ret = acpi_dma_get_range(dev, &dmaaddr, &offset, &size);
+	if (ret == -ENODEV)
+		ret = dev_is_pci(dev) ? rc_dma_get_range(dev, &size)
+				      : nc_dma_get_range(dev, &size);
 
 	if (!ret) {
 		/*
@@ -1692,6 +1663,10 @@ void __init acpi_iort_init(void)
 {
 	acpi_status status;
 
+	/* iort_table will be used at runtime after the iort init,
+	 * so we don't need to call acpi_put_table() to release
+	 * the IORT table mapping.
+	 */
 	status = acpi_get_table(ACPI_SIG_IORT, 0, &iort_table);
 	if (ACPI_FAILURE(status)) {
 		if (status != AE_NOT_FOUND) {
@@ -1703,6 +1678,5 @@ void __init acpi_iort_init(void)
 		return;
 	}
 
-	iort_check_id_count_workaround(iort_table);
 	iort_init_platform_devices();
 }