RISC-V Fixes for 6.7-rc5

* A pair of fixes to the new module load-time relocation code.
 * A fix for hwprobe overflowing on rv32.
 * A fix for to correctly decode C.SWSP and C.SDSP, which manifests in
   misaligned access handling.
 * A fix for a boot-time shadow call stack initialization ordering issue.
 * A fix for Andes' errata probing, which was calling
   riscv_noncoherent_supported() too late in the boot process and
   triggering an oops.
 -----BEGIN PGP SIGNATURE-----
 
 iQJHBAABCAAxFiEEKzw3R0RoQ7JKlDp6LhMZ81+7GIkFAmVzPZ8THHBhbG1lckBk
 YWJiZWx0LmNvbQAKCRAuExnzX7sYiV0LD/0S5ehBNExoF5B/uByAV+5Bgjvee2gc
 UyYcPoFqhHrd3EUBRNO8bCWXphMpFe19HvR4DDjQwOxmSWKFN58TEjkLZQTmfNGX
 z/iRdwkwCKWjnjnGdRupJT7nlPWnXDVbwosxcnRhkxBBZRPA29kjf+Hqg+vF0pIQ
 JMXhUwcV6FrZrKszV/erjNp4L/B1N9uHb5CBMGaYBTaQzMzmcNsCGDZpJrZxlEV/
 MESSwgZRc5fUEhCVyafXoq1VVnjC2yM1fgrigpNmilDDLXYZ6iHk8h19cX2Wk2Ns
 AJ7PX9PR/xuDyaxPD8ANCj/isxqf06T8r3rdSn6aNjP+nzrAoSUAgKQDZNyZclFj
 nymGD/ZvE3gCqoZ6eYJmA0a3qqxQYpL154F1jJiyB+lIUPiqKIQgc5ksLRNCdvF4
 60wc4qKkBKc2PpxU+tD2rn9nl9pAOXiuDGyoT452I3W5SlXb1qDdd5g4hrOlYg4y
 tSmnzyoI2L6qQ+Gux7TCAzTMTrBsSpMkH8uuPRFvk9DmV6bLsUHIlh/hGmqTQBdr
 eZM4vO9Xi1haF8THeiuf4T1PgTfG7qFGpzs3nZwz16WQKd4tm5Qe6Ms3oDa1RM9D
 ZX7feTdw859Pz6ULsm8QMTYn3mBKraUz1EdiNClsKFVBTw722pCk8VJhJbjWFQpZ
 h9IAyKdQjNeTEA==
 =Hb40
 -----END PGP SIGNATURE-----

Merge tag 'riscv-for-linus-6.7-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux

Pull RISC-V fixes from Palmer Dabbelt:

 - A pair of fixes to the new module load-time relocation code

 - A fix for hwprobe overflowing on rv32

 - A fix for to correctly decode C.SWSP and C.SDSP, which manifests in
   misaligned access handling

 - A fix for a boot-time shadow call stack initialization ordering issue

 - A fix for Andes' errata probing, which was calling
   riscv_noncoherent_supported() too late in the boot process and
   triggering an oops

* tag 'riscv-for-linus-6.7-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
  riscv: errata: andes: Probe for IOCP only once in boot stage
  riscv: Fix SMP when shadow call stacks are enabled
  dt-bindings: perf: riscv,pmu: drop unneeded quotes
  riscv: fix misaligned access handling of C.SWSP and C.SDSP
  RISC-V: hwprobe: Always use u64 for extension bits
  Support rv32 ULEB128 test
  riscv: Correct type casting in module loading
  riscv: Safely remove entries from relocation list

Documentation/devicetree/bindings/perf/riscv,pmu.yaml

@@ -90,7 +90,7 @@ properties:
             bitmap of all MHPMCOUNTERx that can monitor the range of events
 
 dependencies:
-  "riscv,event-to-mhpmevent": [ "riscv,event-to-mhpmcounters" ]
+  riscv,event-to-mhpmevent: [ "riscv,event-to-mhpmcounters" ]
 
 required:
   - compatible

arch/riscv/errata/andes/errata.c

@@ -38,28 +38,34 @@ static long ax45mp_iocp_sw_workaround(void)
 	return ret.error ? 0 : ret.value;
 }
 
-static bool errata_probe_iocp(unsigned int stage, unsigned long arch_id, unsigned long impid)
+static void errata_probe_iocp(unsigned int stage, unsigned long arch_id, unsigned long impid)
 {
+	static bool done;
+
 	if (!IS_ENABLED(CONFIG_ERRATA_ANDES_CMO))
-		return false;
+		return;
+
+	if (done)
+		return;
+
+	done = true;
 
 	if (arch_id != ANDESTECH_AX45MP_MARCHID || impid != ANDESTECH_AX45MP_MIMPID)
-		return false;
+		return;
 
 	if (!ax45mp_iocp_sw_workaround())
-		return false;
+		return;
 
 	/* Set this just to make core cbo code happy */
 	riscv_cbom_block_size = 1;
 	riscv_noncoherent_supported();
-
-	return true;
 }
 
 void __init_or_module andes_errata_patch_func(struct alt_entry *begin, struct alt_entry *end,
 					      unsigned long archid, unsigned long impid,
 					      unsigned int stage)
 {
+	if (stage == RISCV_ALTERNATIVES_BOOT)
 		errata_probe_iocp(stage, archid, impid);
 
 	/* we have nothing to patch here ATM so just return back */

arch/riscv/kernel/head.S

@@ -154,7 +154,6 @@ secondary_start_sbi:
 	XIP_FIXUP_OFFSET a3
 	add a3, a3, a1
 	REG_L sp, (a3)
-	scs_load_current
 
 .Lsecondary_start_common:
 
@@ -165,6 +164,7 @@ secondary_start_sbi:
 	call relocate_enable_mmu
 #endif
 	call .Lsetup_trap_vector
+	scs_load_current
 	tail smp_callin
 #endif /* CONFIG_SMP */
 

arch/riscv/kernel/module.c

@@ -40,15 +40,6 @@ struct relocation_handlers {
 				  long buffer);
 };
 
-unsigned int initialize_relocation_hashtable(unsigned int num_relocations);
-void process_accumulated_relocations(struct module *me);
-int add_relocation_to_accumulate(struct module *me, int type, void *location,
-				 unsigned int hashtable_bits, Elf_Addr v);
-
-struct hlist_head *relocation_hashtable;
-
-struct list_head used_buckets_list;
-
 /*
  * The auipc+jalr instruction pair can reach any PC-relative offset
  * in the range [-2^31 - 2^11, 2^31 - 2^11)
@@ -64,7 +55,7 @@ static bool riscv_insn_valid_32bit_offset(ptrdiff_t val)
 
 static int riscv_insn_rmw(void *location, u32 keep, u32 set)
 {
-	u16 *parcel = location;
+	__le16 *parcel = location;
 	u32 insn = (u32)le16_to_cpu(parcel[0]) | (u32)le16_to_cpu(parcel[1]) << 16;
 
 	insn &= keep;
@@ -77,7 +68,7 @@ static int riscv_insn_rmw(void *location, u32 keep, u32 set)
 
 static int riscv_insn_rvc_rmw(void *location, u16 keep, u16 set)
 {
-	u16 *parcel = location;
+	__le16 *parcel = location;
 	u16 insn = le16_to_cpu(*parcel);
 
 	insn &= keep;
@@ -604,7 +595,10 @@ static const struct relocation_handlers reloc_handlers[] = {
 	/* 192-255 nonstandard ABI extensions  */
 };
 
-void process_accumulated_relocations(struct module *me)
+static void
+process_accumulated_relocations(struct module *me,
+				struct hlist_head **relocation_hashtable,
+				struct list_head *used_buckets_list)
 {
 	/*
 	 * Only ADD/SUB/SET/ULEB128 should end up here.
@@ -624,18 +618,25 @@ void process_accumulated_relocations(struct module *me)
 	 *	- Each relocation entry for a location address
 	 */
 	struct used_bucket *bucket_iter;
+	struct used_bucket *bucket_iter_tmp;
 	struct relocation_head *rel_head_iter;
+	struct hlist_node *rel_head_iter_tmp;
 	struct relocation_entry *rel_entry_iter;
+	struct relocation_entry *rel_entry_iter_tmp;
 	int curr_type;
 	void *location;
 	long buffer;
 
-	list_for_each_entry(bucket_iter, &used_buckets_list, head) {
-		hlist_for_each_entry(rel_head_iter, bucket_iter->bucket, node) {
+	list_for_each_entry_safe(bucket_iter, bucket_iter_tmp,
+				 used_buckets_list, head) {
+		hlist_for_each_entry_safe(rel_head_iter, rel_head_iter_tmp,
+					  bucket_iter->bucket, node) {
 			buffer = 0;
 			location = rel_head_iter->location;
-			list_for_each_entry(rel_entry_iter,
-					    rel_head_iter->rel_entry, head) {
+			list_for_each_entry_safe(rel_entry_iter,
+						 rel_entry_iter_tmp,
+						 rel_head_iter->rel_entry,
+						 head) {
 				curr_type = rel_entry_iter->type;
 				reloc_handlers[curr_type].reloc_handler(
 					me, &buffer, rel_entry_iter->value);
@@ -648,11 +649,14 @@ void process_accumulated_relocations(struct module *me)
 		kfree(bucket_iter);
 	}
 
-	kfree(relocation_hashtable);
+	kfree(*relocation_hashtable);
 }
 
-int add_relocation_to_accumulate(struct module *me, int type, void *location,
-				 unsigned int hashtable_bits, Elf_Addr v)
+static int add_relocation_to_accumulate(struct module *me, int type,
+					void *location,
+					unsigned int hashtable_bits, Elf_Addr v,
+					struct hlist_head *relocation_hashtable,
+					struct list_head *used_buckets_list)
 {
 	struct relocation_entry *entry;
 	struct relocation_head *rel_head;
@@ -661,6 +665,10 @@ int add_relocation_to_accumulate(struct module *me, int type, void *location,
 	unsigned long hash;
 
 	entry = kmalloc(sizeof(*entry), GFP_KERNEL);
+
+	if (!entry)
+		return -ENOMEM;
+
 	INIT_LIST_HEAD(&entry->head);
 	entry->type = type;
 	entry->value = v;
@@ -669,7 +677,10 @@ int add_relocation_to_accumulate(struct module *me, int type, void *location,
 
 	current_head = &relocation_hashtable[hash];
 
-	/* Find matching location (if any) */
+	/*
+	 * Search for the relocation_head for the relocations that happen at the
+	 * provided location
+	 */
 	bool found = false;
 	struct relocation_head *rel_head_iter;
 
@@ -681,19 +692,45 @@ int add_relocation_to_accumulate(struct module *me, int type, void *location,
 		}
 	}
 
+	/*
+	 * If there has not yet been any relocations at the provided location,
+	 * create a relocation_head for that location and populate it with this
+	 * relocation_entry.
+	 */
 	if (!found) {
 		rel_head = kmalloc(sizeof(*rel_head), GFP_KERNEL);
+
+		if (!rel_head) {
+			kfree(entry);
+			return -ENOMEM;
+		}
+
 		rel_head->rel_entry =
 			kmalloc(sizeof(struct list_head), GFP_KERNEL);
+
+		if (!rel_head->rel_entry) {
+			kfree(entry);
+			kfree(rel_head);
+			return -ENOMEM;
+		}
+
 		INIT_LIST_HEAD(rel_head->rel_entry);
 		rel_head->location = location;
 		INIT_HLIST_NODE(&rel_head->node);
 		if (!current_head->first) {
 			bucket =
 				kmalloc(sizeof(struct used_bucket), GFP_KERNEL);
+
+			if (!bucket) {
+				kfree(entry);
+				kfree(rel_head);
+				kfree(rel_head->rel_entry);
+				return -ENOMEM;
+			}
+
 			INIT_LIST_HEAD(&bucket->head);
 			bucket->bucket = current_head;
-			list_add(&bucket->head, &used_buckets_list);
+			list_add(&bucket->head, used_buckets_list);
 		}
 		hlist_add_head(&rel_head->node, current_head);
 	}
@@ -704,7 +741,9 @@ int add_relocation_to_accumulate(struct module *me, int type, void *location,
 	return 0;
 }
 
-unsigned int initialize_relocation_hashtable(unsigned int num_relocations)
+static unsigned int
+initialize_relocation_hashtable(unsigned int num_relocations,
+				struct hlist_head **relocation_hashtable)
 {
 	/* Can safely assume that bits is not greater than sizeof(long) */
 	unsigned long hashtable_size = roundup_pow_of_two(num_relocations);
@@ -720,12 +759,13 @@ unsigned int initialize_relocation_hashtable(unsigned int num_relocations)
 
 	hashtable_size <<= should_double_size;
 
-	relocation_hashtable = kmalloc_array(hashtable_size,
+	*relocation_hashtable = kmalloc_array(hashtable_size,
 					      sizeof(*relocation_hashtable),
 					      GFP_KERNEL);
-	__hash_init(relocation_hashtable, hashtable_size);
+	if (!*relocation_hashtable)
+		return -ENOMEM;
 
-	INIT_LIST_HEAD(&used_buckets_list);
+	__hash_init(*relocation_hashtable, hashtable_size);
 
 	return hashtable_bits;
 }
@@ -742,7 +782,17 @@ int apply_relocate_add(Elf_Shdr *sechdrs, const char *strtab,
 	Elf_Addr v;
 	int res;
 	unsigned int num_relocations = sechdrs[relsec].sh_size / sizeof(*rel);
-	unsigned int hashtable_bits = initialize_relocation_hashtable(num_relocations);
+	struct hlist_head *relocation_hashtable;
+	struct list_head used_buckets_list;
+	unsigned int hashtable_bits;
+
+	hashtable_bits = initialize_relocation_hashtable(num_relocations,
+							 &relocation_hashtable);
+
+	if (hashtable_bits < 0)
+		return hashtable_bits;
+
+	INIT_LIST_HEAD(&used_buckets_list);
 
 	pr_debug("Applying relocate section %u to %u\n", relsec,
 	       sechdrs[relsec].sh_info);
@@ -823,14 +873,18 @@ int apply_relocate_add(Elf_Shdr *sechdrs, const char *strtab,
 		}
 
 		if (reloc_handlers[type].accumulate_handler)
-			res = add_relocation_to_accumulate(me, type, location, hashtable_bits, v);
+			res = add_relocation_to_accumulate(me, type, location,
+							   hashtable_bits, v,
+							   relocation_hashtable,
+							   &used_buckets_list);
 		else
 			res = handler(me, location, v);
 		if (res)
 			return res;
 	}
 
-	process_accumulated_relocations(me);
+	process_accumulated_relocations(me, &relocation_hashtable,
+					&used_buckets_list);
 
 	return 0;
 }

arch/riscv/kernel/sys_riscv.c

@@ -169,7 +169,7 @@ static void hwprobe_isa_ext0(struct riscv_hwprobe *pair,
 	pair->value &= ~missing;
 }
 
-static bool hwprobe_ext0_has(const struct cpumask *cpus, unsigned long ext)
+static bool hwprobe_ext0_has(const struct cpumask *cpus, u64 ext)
 {
 	struct riscv_hwprobe pair;
 

arch/riscv/kernel/tests/module_test/test_uleb128.S

@@ -6,13 +6,13 @@
 .text
 .global test_uleb_basic
 test_uleb_basic:
-	ld	a0, second
+	lw	a0, second
 	addi	a0, a0, -127
 	ret
 
 .global test_uleb_large
 test_uleb_large:
-	ld	a0, fourth
+	lw	a0, fourth
 	addi	a0, a0, -0x07e8
 	ret
 
@@ -22,10 +22,10 @@ first:
 second:
 	.reloc second, R_RISCV_SET_ULEB128, second
 	.reloc second, R_RISCV_SUB_ULEB128, first
-	.dword 0
+	.word 0
 third:
 	.space 1000
 fourth:
 	.reloc fourth, R_RISCV_SET_ULEB128, fourth
 	.reloc fourth, R_RISCV_SUB_ULEB128, third
-	.dword 0
+	.word 0

arch/riscv/kernel/traps_misaligned.c

@@ -550,16 +550,14 @@ int handle_misaligned_store(struct pt_regs *regs)
 	} else if ((insn & INSN_MASK_C_SD) == INSN_MATCH_C_SD) {
 		len = 8;
 		val.data_ulong = GET_RS2S(insn, regs);
-	} else if ((insn & INSN_MASK_C_SDSP) == INSN_MATCH_C_SDSP &&
-		   ((insn >> SH_RD) & 0x1f)) {
+	} else if ((insn & INSN_MASK_C_SDSP) == INSN_MATCH_C_SDSP) {
 		len = 8;
 		val.data_ulong = GET_RS2C(insn, regs);
 #endif
 	} else if ((insn & INSN_MASK_C_SW) == INSN_MATCH_C_SW) {
 		len = 4;
 		val.data_ulong = GET_RS2S(insn, regs);
-	} else if ((insn & INSN_MASK_C_SWSP) == INSN_MATCH_C_SWSP &&
-		   ((insn >> SH_RD) & 0x1f)) {
+	} else if ((insn & INSN_MASK_C_SWSP) == INSN_MATCH_C_SWSP) {
 		len = 4;
 		val.data_ulong = GET_RS2C(insn, regs);
 	} else if ((insn & INSN_MASK_C_FSD) == INSN_MATCH_C_FSD) {