diff options
author | Ingo Molnar <mingo@kernel.org> | 2018-05-25 08:02:43 +0200 |
---|---|---|
committer | Ingo Molnar <mingo@kernel.org> | 2018-05-25 08:02:43 +0200 |
commit | bd9c67ad9693bacef086d65c1c6744645d4777e7 (patch) | |
tree | 1dcbac332819a79dbaa1a1a836a8485f866568ad /tools | |
parent | 861410270ab5b13f72fc9afc74fa60f3eb7c2b7e (diff) | |
parent | b50694381cfc22dce3a60a291cdae294a5e5777c (diff) | |
download | linux-stable-bd9c67ad9693bacef086d65c1c6744645d4777e7.tar.gz linux-stable-bd9c67ad9693bacef086d65c1c6744645d4777e7.tar.bz2 linux-stable-bd9c67ad9693bacef086d65c1c6744645d4777e7.zip |
Merge branch 'linus' into perf/core, to pick up fixes
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Diffstat (limited to 'tools')
22 files changed, 791 insertions, 192 deletions
diff --git a/tools/include/linux/spinlock.h b/tools/include/linux/spinlock.h index b21b586b9854..1738c0391da4 100644 --- a/tools/include/linux/spinlock.h +++ b/tools/include/linux/spinlock.h @@ -6,8 +6,9 @@ #include <stdbool.h> #define spinlock_t pthread_mutex_t -#define DEFINE_SPINLOCK(x) pthread_mutex_t x = PTHREAD_MUTEX_INITIALIZER; +#define DEFINE_SPINLOCK(x) pthread_mutex_t x = PTHREAD_MUTEX_INITIALIZER #define __SPIN_LOCK_UNLOCKED(x) (pthread_mutex_t)PTHREAD_MUTEX_INITIALIZER +#define spin_lock_init(x) pthread_mutex_init(x, NULL) #define spin_lock_irqsave(x, f) (void)f, pthread_mutex_lock(x) #define spin_unlock_irqrestore(x, f) (void)f, pthread_mutex_unlock(x) diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c index 5922443063f0..0f9f06df49bc 100644 --- a/tools/lib/bpf/libbpf.c +++ b/tools/lib/bpf/libbpf.c @@ -2035,7 +2035,7 @@ int bpf_prog_load_xattr(const struct bpf_prog_load_attr *attr, return -EINVAL; obj = bpf_object__open(attr->file); - if (IS_ERR(obj)) + if (IS_ERR_OR_NULL(obj)) return -ENOENT; bpf_object__for_each_program(prog, obj) { diff --git a/tools/objtool/arch/x86/include/asm/insn.h b/tools/objtool/arch/x86/include/asm/insn.h index b3e32b010ab1..c2c01f84df75 100644 --- a/tools/objtool/arch/x86/include/asm/insn.h +++ b/tools/objtool/arch/x86/include/asm/insn.h @@ -208,4 +208,22 @@ static inline int insn_offset_immediate(struct insn *insn) return insn_offset_displacement(insn) + insn->displacement.nbytes; } +#define POP_SS_OPCODE 0x1f +#define MOV_SREG_OPCODE 0x8e + +/* + * Intel SDM Vol.3A 6.8.3 states; + * "Any single-step trap that would be delivered following the MOV to SS + * instruction or POP to SS instruction (because EFLAGS.TF is 1) is + * suppressed." + * This function returns true if @insn is MOV SS or POP SS. On these + * instructions, single stepping is suppressed. + */ +static inline int insn_masking_exception(struct insn *insn) +{ + return insn->opcode.bytes[0] == POP_SS_OPCODE || + (insn->opcode.bytes[0] == MOV_SREG_OPCODE && + X86_MODRM_REG(insn->modrm.bytes[0]) == 2); +} + #endif /* _ASM_X86_INSN_H */ diff --git a/tools/objtool/check.c b/tools/objtool/check.c index 5409f6f6c48d..3a31b238f885 100644 --- a/tools/objtool/check.c +++ b/tools/objtool/check.c @@ -59,6 +59,31 @@ static struct instruction *next_insn_same_sec(struct objtool_file *file, return next; } +static struct instruction *next_insn_same_func(struct objtool_file *file, + struct instruction *insn) +{ + struct instruction *next = list_next_entry(insn, list); + struct symbol *func = insn->func; + + if (!func) + return NULL; + + if (&next->list != &file->insn_list && next->func == func) + return next; + + /* Check if we're already in the subfunction: */ + if (func == func->cfunc) + return NULL; + + /* Move to the subfunction: */ + return find_insn(file, func->cfunc->sec, func->cfunc->offset); +} + +#define func_for_each_insn_all(file, func, insn) \ + for (insn = find_insn(file, func->sec, func->offset); \ + insn; \ + insn = next_insn_same_func(file, insn)) + #define func_for_each_insn(file, func, insn) \ for (insn = find_insn(file, func->sec, func->offset); \ insn && &insn->list != &file->insn_list && \ @@ -149,10 +174,14 @@ static int __dead_end_function(struct objtool_file *file, struct symbol *func, if (!strcmp(func->name, global_noreturns[i])) return 1; - if (!func->sec) + if (!func->len) return 0; - func_for_each_insn(file, func, insn) { + insn = find_insn(file, func->sec, func->offset); + if (!insn->func) + return 0; + + func_for_each_insn_all(file, func, insn) { empty = false; if (insn->type == INSN_RETURN) @@ -167,35 +196,28 @@ static int __dead_end_function(struct objtool_file *file, struct symbol *func, * case, the function's dead-end status depends on whether the target * of the sibling call returns. */ - func_for_each_insn(file, func, insn) { - if (insn->sec != func->sec || - insn->offset >= func->offset + func->len) - break; - + func_for_each_insn_all(file, func, insn) { if (insn->type == INSN_JUMP_UNCONDITIONAL) { struct instruction *dest = insn->jump_dest; - struct symbol *dest_func; if (!dest) /* sibling call to another file */ return 0; - if (dest->sec != func->sec || - dest->offset < func->offset || - dest->offset >= func->offset + func->len) { - /* local sibling call */ - dest_func = find_symbol_by_offset(dest->sec, - dest->offset); - if (!dest_func) - continue; + if (dest->func && dest->func->pfunc != insn->func->pfunc) { + /* local sibling call */ if (recursion == 5) { - WARN_FUNC("infinite recursion (objtool bug!)", - dest->sec, dest->offset); - return -1; + /* + * Infinite recursion: two functions + * have sibling calls to each other. + * This is a very rare case. It means + * they aren't dead ends. + */ + return 0; } - return __dead_end_function(file, dest_func, + return __dead_end_function(file, dest->func, recursion + 1); } } @@ -422,7 +444,7 @@ static void add_ignores(struct objtool_file *file) if (!ignore_func(file, func)) continue; - func_for_each_insn(file, func, insn) + func_for_each_insn_all(file, func, insn) insn->ignore = true; } } @@ -782,30 +804,35 @@ out: return ret; } -static int add_switch_table(struct objtool_file *file, struct symbol *func, - struct instruction *insn, struct rela *table, - struct rela *next_table) +static int add_switch_table(struct objtool_file *file, struct instruction *insn, + struct rela *table, struct rela *next_table) { struct rela *rela = table; struct instruction *alt_insn; struct alternative *alt; + struct symbol *pfunc = insn->func->pfunc; + unsigned int prev_offset = 0; list_for_each_entry_from(rela, &file->rodata->rela->rela_list, list) { if (rela == next_table) break; - if (rela->sym->sec != insn->sec || - rela->addend <= func->offset || - rela->addend >= func->offset + func->len) + /* Make sure the switch table entries are consecutive: */ + if (prev_offset && rela->offset != prev_offset + 8) break; - alt_insn = find_insn(file, insn->sec, rela->addend); - if (!alt_insn) { - WARN("%s: can't find instruction at %s+0x%x", - file->rodata->rela->name, insn->sec->name, - rela->addend); - return -1; - } + /* Detect function pointers from contiguous objects: */ + if (rela->sym->sec == pfunc->sec && + rela->addend == pfunc->offset) + break; + + alt_insn = find_insn(file, rela->sym->sec, rela->addend); + if (!alt_insn) + break; + + /* Make sure the jmp dest is in the function or subfunction: */ + if (alt_insn->func->pfunc != pfunc) + break; alt = malloc(sizeof(*alt)); if (!alt) { @@ -815,6 +842,13 @@ static int add_switch_table(struct objtool_file *file, struct symbol *func, alt->insn = alt_insn; list_add_tail(&alt->list, &insn->alts); + prev_offset = rela->offset; + } + + if (!prev_offset) { + WARN_FUNC("can't find switch jump table", + insn->sec, insn->offset); + return -1; } return 0; @@ -869,40 +903,21 @@ static struct rela *find_switch_table(struct objtool_file *file, { struct rela *text_rela, *rodata_rela; struct instruction *orig_insn = insn; + unsigned long table_offset; - text_rela = find_rela_by_dest_range(insn->sec, insn->offset, insn->len); - if (text_rela && text_rela->sym == file->rodata->sym) { - /* case 1 */ - rodata_rela = find_rela_by_dest(file->rodata, - text_rela->addend); - if (rodata_rela) - return rodata_rela; - - /* case 2 */ - rodata_rela = find_rela_by_dest(file->rodata, - text_rela->addend + 4); - if (!rodata_rela) - return NULL; - - file->ignore_unreachables = true; - return rodata_rela; - } - - /* case 3 */ /* * Backward search using the @first_jump_src links, these help avoid * much of the 'in between' code. Which avoids us getting confused by * it. */ - for (insn = list_prev_entry(insn, list); - + for (; &insn->list != &file->insn_list && insn->sec == func->sec && insn->offset >= func->offset; insn = insn->first_jump_src ?: list_prev_entry(insn, list)) { - if (insn->type == INSN_JUMP_DYNAMIC) + if (insn != orig_insn && insn->type == INSN_JUMP_DYNAMIC) break; /* allow small jumps within the range */ @@ -918,18 +933,29 @@ static struct rela *find_switch_table(struct objtool_file *file, if (!text_rela || text_rela->sym != file->rodata->sym) continue; + table_offset = text_rela->addend; + if (text_rela->type == R_X86_64_PC32) + table_offset += 4; + /* * Make sure the .rodata address isn't associated with a * symbol. gcc jump tables are anonymous data. */ - if (find_symbol_containing(file->rodata, text_rela->addend)) + if (find_symbol_containing(file->rodata, table_offset)) continue; - rodata_rela = find_rela_by_dest(file->rodata, text_rela->addend); - if (!rodata_rela) - continue; + rodata_rela = find_rela_by_dest(file->rodata, table_offset); + if (rodata_rela) { + /* + * Use of RIP-relative switch jumps is quite rare, and + * indicates a rare GCC quirk/bug which can leave dead + * code behind. + */ + if (text_rela->type == R_X86_64_PC32) + file->ignore_unreachables = true; - return rodata_rela; + return rodata_rela; + } } return NULL; @@ -943,7 +969,7 @@ static int add_func_switch_tables(struct objtool_file *file, struct rela *rela, *prev_rela = NULL; int ret; - func_for_each_insn(file, func, insn) { + func_for_each_insn_all(file, func, insn) { if (!last) last = insn; @@ -974,8 +1000,7 @@ static int add_func_switch_tables(struct objtool_file *file, * the beginning of another switch table in the same function. */ if (prev_jump) { - ret = add_switch_table(file, func, prev_jump, prev_rela, - rela); + ret = add_switch_table(file, prev_jump, prev_rela, rela); if (ret) return ret; } @@ -985,7 +1010,7 @@ static int add_func_switch_tables(struct objtool_file *file, } if (prev_jump) { - ret = add_switch_table(file, func, prev_jump, prev_rela, NULL); + ret = add_switch_table(file, prev_jump, prev_rela, NULL); if (ret) return ret; } @@ -1749,15 +1774,13 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, while (1) { next_insn = next_insn_same_sec(file, insn); - - if (file->c_file && func && insn->func && func != insn->func) { + if (file->c_file && func && insn->func && func != insn->func->pfunc) { WARN("%s() falls through to next function %s()", func->name, insn->func->name); return 1; } - if (insn->func) - func = insn->func; + func = insn->func ? insn->func->pfunc : NULL; if (func && insn->ignore) { WARN_FUNC("BUG: why am I validating an ignored function?", @@ -1778,7 +1801,7 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, i = insn; save_insn = NULL; - func_for_each_insn_continue_reverse(file, func, i) { + func_for_each_insn_continue_reverse(file, insn->func, i) { if (i->save) { save_insn = i; break; @@ -1865,7 +1888,7 @@ static int validate_branch(struct objtool_file *file, struct instruction *first, case INSN_JUMP_UNCONDITIONAL: if (insn->jump_dest && (!func || !insn->jump_dest->func || - func == insn->jump_dest->func)) { + insn->jump_dest->func->pfunc == func)) { ret = validate_branch(file, insn->jump_dest, state); if (ret) @@ -2060,7 +2083,7 @@ static int validate_functions(struct objtool_file *file) for_each_sec(file, sec) { list_for_each_entry(func, &sec->symbol_list, list) { - if (func->type != STT_FUNC) + if (func->type != STT_FUNC || func->pfunc != func) continue; insn = find_insn(file, sec, func->offset); diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c index c1c338661699..4e60e105583e 100644 --- a/tools/objtool/elf.c +++ b/tools/objtool/elf.c @@ -79,6 +79,19 @@ struct symbol *find_symbol_by_offset(struct section *sec, unsigned long offset) return NULL; } +struct symbol *find_symbol_by_name(struct elf *elf, const char *name) +{ + struct section *sec; + struct symbol *sym; + + list_for_each_entry(sec, &elf->sections, list) + list_for_each_entry(sym, &sec->symbol_list, list) + if (!strcmp(sym->name, name)) + return sym; + + return NULL; +} + struct symbol *find_symbol_containing(struct section *sec, unsigned long offset) { struct symbol *sym; @@ -203,10 +216,11 @@ static int read_sections(struct elf *elf) static int read_symbols(struct elf *elf) { - struct section *symtab; - struct symbol *sym; + struct section *symtab, *sec; + struct symbol *sym, *pfunc; struct list_head *entry, *tmp; int symbols_nr, i; + char *coldstr; symtab = find_section_by_name(elf, ".symtab"); if (!symtab) { @@ -281,6 +295,30 @@ static int read_symbols(struct elf *elf) hash_add(sym->sec->symbol_hash, &sym->hash, sym->idx); } + /* Create parent/child links for any cold subfunctions */ + list_for_each_entry(sec, &elf->sections, list) { + list_for_each_entry(sym, &sec->symbol_list, list) { + if (sym->type != STT_FUNC) + continue; + sym->pfunc = sym->cfunc = sym; + coldstr = strstr(sym->name, ".cold."); + if (coldstr) { + coldstr[0] = '\0'; + pfunc = find_symbol_by_name(elf, sym->name); + coldstr[0] = '.'; + + if (!pfunc) { + WARN("%s(): can't find parent function", + sym->name); + goto err; + } + + sym->pfunc = pfunc; + pfunc->cfunc = sym; + } + } + } + return 0; err: diff --git a/tools/objtool/elf.h b/tools/objtool/elf.h index d86e2ff14466..de5cd2ddded9 100644 --- a/tools/objtool/elf.h +++ b/tools/objtool/elf.h @@ -61,6 +61,7 @@ struct symbol { unsigned char bind, type; unsigned long offset; unsigned int len; + struct symbol *pfunc, *cfunc; }; struct rela { @@ -86,6 +87,7 @@ struct elf { struct elf *elf_open(const char *name, int flags); struct section *find_section_by_name(struct elf *elf, const char *name); struct symbol *find_symbol_by_offset(struct section *sec, unsigned long offset); +struct symbol *find_symbol_by_name(struct elf *elf, const char *name); struct symbol *find_symbol_containing(struct section *sec, unsigned long offset); struct rela *find_rela_by_dest(struct section *sec, unsigned long offset); struct rela *find_rela_by_dest_range(struct section *sec, unsigned long offset, diff --git a/tools/testing/radix-tree/Makefile b/tools/testing/radix-tree/Makefile index fa7ee369b3c9..db66f8a0d4be 100644 --- a/tools/testing/radix-tree/Makefile +++ b/tools/testing/radix-tree/Makefile @@ -17,7 +17,7 @@ ifeq ($(BUILD), 32) LDFLAGS += -m32 endif -targets: mapshift $(TARGETS) +targets: generated/map-shift.h $(TARGETS) main: $(OFILES) @@ -42,9 +42,7 @@ radix-tree.c: ../../../lib/radix-tree.c idr.c: ../../../lib/idr.c sed -e 's/^static //' -e 's/__always_inline //' -e 's/inline //' < $< > $@ -.PHONY: mapshift - -mapshift: +generated/map-shift.h: @if ! grep -qws $(SHIFT) generated/map-shift.h; then \ echo "#define RADIX_TREE_MAP_SHIFT $(SHIFT)" > \ generated/map-shift.h; \ diff --git a/tools/testing/radix-tree/multiorder.c b/tools/testing/radix-tree/multiorder.c index 59245b3d587c..7bf405638b0b 100644 --- a/tools/testing/radix-tree/multiorder.c +++ b/tools/testing/radix-tree/multiorder.c @@ -16,6 +16,7 @@ #include <linux/radix-tree.h> #include <linux/slab.h> #include <linux/errno.h> +#include <pthread.h> #include "test.h" @@ -624,6 +625,67 @@ static void multiorder_account(void) item_kill_tree(&tree); } +bool stop_iteration = false; + +static void *creator_func(void *ptr) +{ + /* 'order' is set up to ensure we have sibling entries */ + unsigned int order = RADIX_TREE_MAP_SHIFT - 1; + struct radix_tree_root *tree = ptr; + int i; + + for (i = 0; i < 10000; i++) { + item_insert_order(tree, 0, order); + item_delete_rcu(tree, 0); + } + + stop_iteration = true; + return NULL; +} + +static void *iterator_func(void *ptr) +{ + struct radix_tree_root *tree = ptr; + struct radix_tree_iter iter; + struct item *item; + void **slot; + + while (!stop_iteration) { + rcu_read_lock(); + radix_tree_for_each_slot(slot, tree, &iter, 0) { + item = radix_tree_deref_slot(slot); + + if (!item) + continue; + if (radix_tree_deref_retry(item)) { + slot = radix_tree_iter_retry(&iter); + continue; + } + + item_sanity(item, iter.index); + } + rcu_read_unlock(); + } + return NULL; +} + +static void multiorder_iteration_race(void) +{ + const int num_threads = sysconf(_SC_NPROCESSORS_ONLN); + pthread_t worker_thread[num_threads]; + RADIX_TREE(tree, GFP_KERNEL); + int i; + + pthread_create(&worker_thread[0], NULL, &creator_func, &tree); + for (i = 1; i < num_threads; i++) + pthread_create(&worker_thread[i], NULL, &iterator_func, &tree); + + for (i = 0; i < num_threads; i++) + pthread_join(worker_thread[i], NULL); + + item_kill_tree(&tree); +} + void multiorder_checks(void) { int i; @@ -644,6 +706,7 @@ void multiorder_checks(void) multiorder_join(); multiorder_split(); multiorder_account(); + multiorder_iteration_race(); radix_tree_cpu_dead(0); } diff --git a/tools/testing/radix-tree/test.c b/tools/testing/radix-tree/test.c index 5978ab1f403d..def6015570b2 100644 --- a/tools/testing/radix-tree/test.c +++ b/tools/testing/radix-tree/test.c @@ -75,6 +75,25 @@ int item_delete(struct radix_tree_root *root, unsigned long index) return 0; } +static void item_free_rcu(struct rcu_head *head) +{ + struct item *item = container_of(head, struct item, rcu_head); + + free(item); +} + +int item_delete_rcu(struct radix_tree_root *root, unsigned long index) +{ + struct item *item = radix_tree_delete(root, index); + + if (item) { + item_sanity(item, index); + call_rcu(&item->rcu_head, item_free_rcu); + return 1; + } + return 0; +} + void item_check_present(struct radix_tree_root *root, unsigned long index) { struct item *item; diff --git a/tools/testing/radix-tree/test.h b/tools/testing/radix-tree/test.h index d9c031dbeb1a..31f1d9b6f506 100644 --- a/tools/testing/radix-tree/test.h +++ b/tools/testing/radix-tree/test.h @@ -5,6 +5,7 @@ #include <linux/rcupdate.h> struct item { + struct rcu_head rcu_head; unsigned long index; unsigned int order; }; @@ -12,9 +13,11 @@ struct item { struct item *item_create(unsigned long index, unsigned int order); int __item_insert(struct radix_tree_root *root, struct item *item); int item_insert(struct radix_tree_root *root, unsigned long index); +void item_sanity(struct item *item, unsigned long index); int item_insert_order(struct radix_tree_root *root, unsigned long index, unsigned order); int item_delete(struct radix_tree_root *root, unsigned long index); +int item_delete_rcu(struct radix_tree_root *root, unsigned long index); struct item *item_lookup(struct radix_tree_root *root, unsigned long index); void item_check_present(struct radix_tree_root *root, unsigned long index); diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c index 3e7718b1a9ae..fd7de7eb329e 100644 --- a/tools/testing/selftests/bpf/test_verifier.c +++ b/tools/testing/selftests/bpf/test_verifier.c @@ -11713,6 +11713,11 @@ static void get_unpriv_disabled() FILE *fd; fd = fopen("/proc/sys/"UNPRIV_SYSCTL, "r"); + if (!fd) { + perror("fopen /proc/sys/"UNPRIV_SYSCTL); + unpriv_disabled = true; + return; + } if (fgets(buf, 2, fd) == buf && atoi(buf)) unpriv_disabled = true; fclose(fd); diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile index 2ddcc96ae456..d9d00319b07c 100644 --- a/tools/testing/selftests/kvm/Makefile +++ b/tools/testing/selftests/kvm/Makefile @@ -15,7 +15,7 @@ LIBKVM += $(LIBKVM_$(UNAME_M)) INSTALL_HDR_PATH = $(top_srcdir)/usr LINUX_HDR_PATH = $(INSTALL_HDR_PATH)/include/ -CFLAGS += -O2 -g -std=gnu99 -I$(LINUX_HDR_PATH) -Iinclude -I$(<D) +CFLAGS += -O2 -g -std=gnu99 -I$(LINUX_HDR_PATH) -Iinclude -I$(<D) -I.. # After inclusion, $(OUTPUT) is defined and # $(TEST_GEN_PROGS) starts with $(OUTPUT)/ diff --git a/tools/testing/selftests/kvm/include/test_util.h b/tools/testing/selftests/kvm/include/test_util.h index 7ab98e41324f..ac53730b30aa 100644 --- a/tools/testing/selftests/kvm/include/test_util.h +++ b/tools/testing/selftests/kvm/include/test_util.h @@ -19,6 +19,7 @@ #include <errno.h> #include <unistd.h> #include <fcntl.h> +#include "kselftest.h" ssize_t test_write(int fd, const void *buf, size_t count); ssize_t test_read(int fd, void *buf, size_t count); diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c index 2cedfda181d4..37e2a787d2fc 100644 --- a/tools/testing/selftests/kvm/lib/kvm_util.c +++ b/tools/testing/selftests/kvm/lib/kvm_util.c @@ -50,8 +50,8 @@ int kvm_check_cap(long cap) int kvm_fd; kvm_fd = open(KVM_DEV_PATH, O_RDONLY); - TEST_ASSERT(kvm_fd >= 0, "open %s failed, rc: %i errno: %i", - KVM_DEV_PATH, kvm_fd, errno); + if (kvm_fd < 0) + exit(KSFT_SKIP); ret = ioctl(kvm_fd, KVM_CHECK_EXTENSION, cap); TEST_ASSERT(ret != -1, "KVM_CHECK_EXTENSION IOCTL failed,\n" @@ -91,8 +91,8 @@ struct kvm_vm *vm_create(enum vm_guest_mode mode, uint64_t phy_pages, int perm) vm->mode = mode; kvm_fd = open(KVM_DEV_PATH, perm); - TEST_ASSERT(kvm_fd >= 0, "open %s failed, rc: %i errno: %i", - KVM_DEV_PATH, kvm_fd, errno); + if (kvm_fd < 0) + exit(KSFT_SKIP); /* Create VM. */ vm->fd = ioctl(kvm_fd, KVM_CREATE_VM, NULL); @@ -418,8 +418,8 @@ struct kvm_cpuid2 *kvm_get_supported_cpuid(void) cpuid = allocate_kvm_cpuid2(); kvm_fd = open(KVM_DEV_PATH, O_RDONLY); - TEST_ASSERT(kvm_fd >= 0, "open %s failed, rc: %i errno: %i", - KVM_DEV_PATH, kvm_fd, errno); + if (kvm_fd < 0) + exit(KSFT_SKIP); ret = ioctl(kvm_fd, KVM_GET_SUPPORTED_CPUID, cpuid); TEST_ASSERT(ret == 0, "KVM_GET_SUPPORTED_CPUID failed %d %d\n", @@ -675,8 +675,8 @@ static int vcpu_mmap_sz(void) int dev_fd, ret; dev_fd = open(KVM_DEV_PATH, O_RDONLY); - TEST_ASSERT(dev_fd >= 0, "%s open %s failed, rc: %i errno: %i", - __func__, KVM_DEV_PATH, dev_fd, errno); + if (dev_fd < 0) + exit(KSFT_SKIP); ret = ioctl(dev_fd, KVM_GET_VCPU_MMAP_SIZE, NULL); TEST_ASSERT(ret >= sizeof(struct kvm_run), diff --git a/tools/testing/selftests/kvm/sync_regs_test.c b/tools/testing/selftests/kvm/sync_regs_test.c index 428e9473f5e2..eae1ece3c31b 100644 --- a/tools/testing/selftests/kvm/sync_regs_test.c +++ b/tools/testing/selftests/kvm/sync_regs_test.c @@ -85,6 +85,9 @@ static void compare_vcpu_events(struct kvm_vcpu_events *left, { } +#define TEST_SYNC_FIELDS (KVM_SYNC_X86_REGS|KVM_SYNC_X86_SREGS|KVM_SYNC_X86_EVENTS) +#define INVALID_SYNC_FIELD 0x80000000 + int main(int argc, char *argv[]) { struct kvm_vm *vm; @@ -98,9 +101,14 @@ int main(int argc, char *argv[]) setbuf(stdout, NULL); cap = kvm_check_cap(KVM_CAP_SYNC_REGS); - TEST_ASSERT((unsigned long)cap == KVM_SYNC_X86_VALID_FIELDS, - "KVM_CAP_SYNC_REGS (0x%x) != KVM_SYNC_X86_VALID_FIELDS (0x%lx)\n", - cap, KVM_SYNC_X86_VALID_FIELDS); + if ((cap & TEST_SYNC_FIELDS) != TEST_SYNC_FIELDS) { + fprintf(stderr, "KVM_CAP_SYNC_REGS not supported, skipping test\n"); + exit(KSFT_SKIP); + } + if ((cap & INVALID_SYNC_FIELD) != 0) { + fprintf(stderr, "The \"invalid\" field is not invalid, skipping test\n"); + exit(KSFT_SKIP); + } /* Create VM */ vm = vm_create_default(VCPU_ID, guest_code); @@ -108,7 +116,14 @@ int main(int argc, char *argv[]) run = vcpu_state(vm, VCPU_ID); /* Request reading invalid register set from VCPU. */ - run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS << 1; + run->kvm_valid_regs = INVALID_SYNC_FIELD; + rv = _vcpu_run(vm, VCPU_ID); + TEST_ASSERT(rv < 0 && errno == EINVAL, + "Invalid kvm_valid_regs did not cause expected KVM_RUN error: %d\n", + rv); + vcpu_state(vm, VCPU_ID)->kvm_valid_regs = 0; + + run->kvm_valid_regs = INVALID_SYNC_FIELD | TEST_SYNC_FIELDS; rv = _vcpu_run(vm, VCPU_ID); TEST_ASSERT(rv < 0 && errno == EINVAL, "Invalid kvm_valid_regs did not cause expected KVM_RUN error: %d\n", @@ -116,7 +131,14 @@ int main(int argc, char *argv[]) vcpu_state(vm, VCPU_ID)->kvm_valid_regs = 0; /* Request setting invalid register set into VCPU. */ - run->kvm_dirty_regs = KVM_SYNC_X86_VALID_FIELDS << 1; + run->kvm_dirty_regs = INVALID_SYNC_FIELD; + rv = _vcpu_run(vm, VCPU_ID); + TEST_ASSERT(rv < 0 && errno == EINVAL, + "Invalid kvm_dirty_regs did not cause expected KVM_RUN error: %d\n", + rv); + vcpu_state(vm, VCPU_ID)->kvm_dirty_regs = 0; + + run->kvm_dirty_regs = INVALID_SYNC_FIELD | TEST_SYNC_FIELDS; rv = _vcpu_run(vm, VCPU_ID); TEST_ASSERT(rv < 0 && errno == EINVAL, "Invalid kvm_dirty_regs did not cause expected KVM_RUN error: %d\n", @@ -125,7 +147,7 @@ int main(int argc, char *argv[]) /* Request and verify all valid register sets. */ /* TODO: BUILD TIME CHECK: TEST_ASSERT(KVM_SYNC_X86_NUM_FIELDS != 3); */ - run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS; + run->kvm_valid_regs = TEST_SYNC_FIELDS; rv = _vcpu_run(vm, VCPU_ID); TEST_ASSERT(run->exit_reason == KVM_EXIT_IO, "Unexpected exit reason: %u (%s),\n", @@ -146,7 +168,7 @@ int main(int argc, char *argv[]) run->s.regs.sregs.apic_base = 1 << 11; /* TODO run->s.regs.events.XYZ = ABC; */ - run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS; + run->kvm_valid_regs = TEST_SYNC_FIELDS; run->kvm_dirty_regs = KVM_SYNC_X86_REGS | KVM_SYNC_X86_SREGS; rv = _vcpu_run(vm, VCPU_ID); TEST_ASSERT(run->exit_reason == KVM_EXIT_IO, @@ -172,7 +194,7 @@ int main(int argc, char *argv[]) /* Clear kvm_dirty_regs bits, verify new s.regs values are * overwritten with existing guest values. */ - run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS; + run->kvm_valid_regs = TEST_SYNC_FIELDS; run->kvm_dirty_regs = 0; run->s.regs.regs.r11 = 0xDEADBEEF; rv = _vcpu_run(vm, VCPU_ID); @@ -211,7 +233,7 @@ int main(int argc, char *argv[]) * with kvm_sync_regs values. */ run->kvm_valid_regs = 0; - run->kvm_dirty_regs = KVM_SYNC_X86_VALID_FIELDS; + run->kvm_dirty_regs = TEST_SYNC_FIELDS; run->s.regs.regs.r11 = 0xBBBB; rv = _vcpu_run(vm, VCPU_ID); TEST_ASSERT(run->exit_reason == KVM_EXIT_IO, diff --git a/tools/testing/selftests/kvm/vmx_tsc_adjust_test.c b/tools/testing/selftests/kvm/vmx_tsc_adjust_test.c index 8f7f62093add..aaa633263b2c 100644 --- a/tools/testing/selftests/kvm/vmx_tsc_adjust_test.c +++ b/tools/testing/selftests/kvm/vmx_tsc_adjust_test.c @@ -189,8 +189,8 @@ int main(int argc, char *argv[]) struct kvm_cpuid_entry2 *entry = kvm_get_supported_cpuid_entry(1); if (!(entry->ecx & CPUID_VMX)) { - printf("nested VMX not enabled, skipping test"); - return 0; + fprintf(stderr, "nested VMX not enabled, skipping test\n"); + exit(KSFT_SKIP); } vm = vm_create_default_vmx(VCPU_ID, (void *) l1_guest_code); diff --git a/tools/testing/selftests/seccomp/seccomp_bpf.c b/tools/testing/selftests/seccomp/seccomp_bpf.c index 168c66d74fc5..e1473234968d 100644 --- a/tools/testing/selftests/seccomp/seccomp_bpf.c +++ b/tools/testing/selftests/seccomp/seccomp_bpf.c @@ -134,11 +134,15 @@ struct seccomp_data { #endif #ifndef SECCOMP_FILTER_FLAG_TSYNC -#define SECCOMP_FILTER_FLAG_TSYNC 1 +#define SECCOMP_FILTER_FLAG_TSYNC (1UL << 0) #endif #ifndef SECCOMP_FILTER_FLAG_LOG -#define SECCOMP_FILTER_FLAG_LOG 2 +#define SECCOMP_FILTER_FLAG_LOG (1UL << 1) +#endif + +#ifndef SECCOMP_FILTER_FLAG_SPEC_ALLOW +#define SECCOMP_FILTER_FLAG_SPEC_ALLOW (1UL << 2) #endif #ifndef PTRACE_SECCOMP_GET_METADATA @@ -2072,14 +2076,26 @@ TEST(seccomp_syscall_mode_lock) TEST(detect_seccomp_filter_flags) { unsigned int flags[] = { SECCOMP_FILTER_FLAG_TSYNC, - SECCOMP_FILTER_FLAG_LOG }; + SECCOMP_FILTER_FLAG_LOG, + SECCOMP_FILTER_FLAG_SPEC_ALLOW }; unsigned int flag, all_flags; int i; long ret; /* Test detection of known-good filter flags */ for (i = 0, all_flags = 0; i < ARRAY_SIZE(flags); i++) { + int bits = 0; + flag = flags[i]; + /* Make sure the flag is a single bit! */ + while (flag) { + if (flag & 0x1) + bits ++; + flag >>= 1; + } + ASSERT_EQ(1, bits); + flag = flags[i]; + ret = seccomp(SECCOMP_SET_MODE_FILTER, flag, NULL); ASSERT_NE(ENOSYS, errno) { TH_LOG("Kernel does not support seccomp syscall!"); diff --git a/tools/testing/selftests/x86/Makefile b/tools/testing/selftests/x86/Makefile index d744991c0f4f..39f66bc29b82 100644 --- a/tools/testing/selftests/x86/Makefile +++ b/tools/testing/selftests/x86/Makefile @@ -11,7 +11,7 @@ CAN_BUILD_X86_64 := $(shell ./check_cc.sh $(CC) trivial_64bit_program.c) TARGETS_C_BOTHBITS := single_step_syscall sysret_ss_attrs syscall_nt test_mremap_vdso \ check_initial_reg_state sigreturn iopl mpx-mini-test ioperm \ - protection_keys test_vdso test_vsyscall + protection_keys test_vdso test_vsyscall mov_ss_trap TARGETS_C_32BIT_ONLY := entry_from_vm86 syscall_arg_fault test_syscall_vdso unwind_vdso \ test_FCMOV test_FCOMI test_FISTTP \ vdso_restorer diff --git a/tools/testing/selftests/x86/mov_ss_trap.c b/tools/testing/selftests/x86/mov_ss_trap.c new file mode 100644 index 000000000000..3c3a022654f3 --- /dev/null +++ b/tools/testing/selftests/x86/mov_ss_trap.c @@ -0,0 +1,285 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * mov_ss_trap.c: Exercise the bizarre side effects of a watchpoint on MOV SS + * + * This does MOV SS from a watchpointed address followed by various + * types of kernel entries. A MOV SS that hits a watchpoint will queue + * up a #DB trap but will not actually deliver that trap. The trap + * will be delivered after the next instruction instead. The CPU's logic + * seems to be: + * + * - Any fault: drop the pending #DB trap. + * - INT $N, INT3, INTO, SYSCALL, SYSENTER: enter the kernel and then + * deliver #DB. + * - ICEBP: enter the kernel but do not deliver the watchpoint trap + * - breakpoint: only one #DB is delivered (phew!) + * + * There are plenty of ways for a kernel to handle this incorrectly. This + * test tries to exercise all the cases. + * + * This should mostly cover CVE-2018-1087 and CVE-2018-8897. + */ +#define _GNU_SOURCE + +#include <stdlib.h> +#include <sys/ptrace.h> +#include <sys/types.h> +#include <sys/wait.h> +#include <sys/user.h> +#include <sys/syscall.h> +#include <unistd.h> +#include <errno.h> +#include <stddef.h> +#include <stdio.h> +#include <err.h> +#include <string.h> +#include <setjmp.h> +#include <sys/prctl.h> + +#define X86_EFLAGS_RF (1UL << 16) + +#if __x86_64__ +# define REG_IP REG_RIP +#else +# define REG_IP REG_EIP +#endif + +unsigned short ss; +extern unsigned char breakpoint_insn[]; +sigjmp_buf jmpbuf; +static unsigned char altstack_data[SIGSTKSZ]; + +static void enable_watchpoint(void) +{ + pid_t parent = getpid(); + int status; + + pid_t child = fork(); + if (child < 0) + err(1, "fork"); + + if (child) { + if (waitpid(child, &status, 0) != child) + err(1, "waitpid for child"); + } else { + unsigned long dr0, dr1, dr7; + + dr0 = (unsigned long)&ss; + dr1 = (unsigned long)breakpoint_insn; + dr7 = ((1UL << 1) | /* G0 */ + (3UL << 16) | /* RW0 = read or write */ + (1UL << 18) | /* LEN0 = 2 bytes */ + (1UL << 3)); /* G1, RW1 = insn */ + + if (ptrace(PTRACE_ATTACH, parent, NULL, NULL) != 0) + err(1, "PTRACE_ATTACH"); + + if (waitpid(parent, &status, 0) != parent) + err(1, "waitpid for child"); + + if (ptrace(PTRACE_POKEUSER, parent, (void *)offsetof(struct user, u_debugreg[0]), dr0) != 0) + err(1, "PTRACE_POKEUSER DR0"); + + if (ptrace(PTRACE_POKEUSER, parent, (void *)offsetof(struct user, u_debugreg[1]), dr1) != 0) + err(1, "PTRACE_POKEUSER DR1"); + + if (ptrace(PTRACE_POKEUSER, parent, (void *)offsetof(struct user, u_debugreg[7]), dr7) != 0) + err(1, "PTRACE_POKEUSER DR7"); + + printf("\tDR0 = %lx, DR1 = %lx, DR7 = %lx\n", dr0, dr1, dr7); + + if (ptrace(PTRACE_DETACH, parent, NULL, NULL) != 0) + err(1, "PTRACE_DETACH"); + + exit(0); + } +} + +static void sethandler(int sig, void (*handler)(int, siginfo_t *, void *), + int flags) +{ + struct sigaction sa; + memset(&sa, 0, sizeof(sa)); + sa.sa_sigaction = handler; + sa.sa_flags = SA_SIGINFO | flags; + sigemptyset(&sa.sa_mask); + if (sigaction(sig, &sa, 0)) + err(1, "sigaction"); +} + +static char const * const signames[] = { + [SIGSEGV] = "SIGSEGV", + [SIGBUS] = "SIBGUS", + [SIGTRAP] = "SIGTRAP", + [SIGILL] = "SIGILL", +}; + +static void sigtrap(int sig, siginfo_t *si, void *ctx_void) +{ + ucontext_t *ctx = ctx_void; + + printf("\tGot SIGTRAP with RIP=%lx, EFLAGS.RF=%d\n", + (unsigned long)ctx->uc_mcontext.gregs[REG_IP], + !!(ctx->uc_mcontext.gregs[REG_EFL] & X86_EFLAGS_RF)); +} + +static void handle_and_return(int sig, siginfo_t *si, void *ctx_void) +{ + ucontext_t *ctx = ctx_void; + + printf("\tGot %s with RIP=%lx\n", signames[sig], + (unsigned long)ctx->uc_mcontext.gregs[REG_IP]); +} + +static void handle_and_longjmp(int sig, siginfo_t *si, void *ctx_void) +{ + ucontext_t *ctx = ctx_void; + + printf("\tGot %s with RIP=%lx\n", signames[sig], + (unsigned long)ctx->uc_mcontext.gregs[REG_IP]); + + siglongjmp(jmpbuf, 1); +} + +int main() +{ + unsigned long nr; + + asm volatile ("mov %%ss, %[ss]" : [ss] "=m" (ss)); + printf("\tSS = 0x%hx, &SS = 0x%p\n", ss, &ss); + + if (prctl(PR_SET_PTRACER, PR_SET_PTRACER_ANY, 0, 0, 0) == 0) + printf("\tPR_SET_PTRACER_ANY succeeded\n"); + + printf("\tSet up a watchpoint\n"); + sethandler(SIGTRAP, sigtrap, 0); + enable_watchpoint(); + + printf("[RUN]\tRead from watched memory (should get SIGTRAP)\n"); + asm volatile ("mov %[ss], %[tmp]" : [tmp] "=r" (nr) : [ss] "m" (ss)); + + printf("[RUN]\tMOV SS; INT3\n"); + asm volatile ("mov %[ss], %%ss; int3" :: [ss] "m" (ss)); + + printf("[RUN]\tMOV SS; INT 3\n"); + asm volatile ("mov %[ss], %%ss; .byte 0xcd, 0x3" :: [ss] "m" (ss)); + + printf("[RUN]\tMOV SS; CS CS INT3\n"); + asm volatile ("mov %[ss], %%ss; .byte 0x2e, 0x2e; int3" :: [ss] "m" (ss)); + + printf("[RUN]\tMOV SS; CSx14 INT3\n"); + asm volatile ("mov %[ss], %%ss; .fill 14,1,0x2e; int3" :: [ss] "m" (ss)); + + printf("[RUN]\tMOV SS; INT 4\n"); + sethandler(SIGSEGV, handle_and_return, SA_RESETHAND); + asm volatile ("mov %[ss], %%ss; int $4" :: [ss] "m" (ss)); + +#ifdef __i386__ + printf("[RUN]\tMOV SS; INTO\n"); + sethandler(SIGSEGV, handle_and_return, SA_RESETHAND); + nr = -1; + asm volatile ("add $1, %[tmp]; mov %[ss], %%ss; into" + : [tmp] "+r" (nr) : [ss] "m" (ss)); +#endif + + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; ICEBP\n"); + + /* Some emulators (e.g. QEMU TCG) don't emulate ICEBP. */ + sethandler(SIGILL, handle_and_longjmp, SA_RESETHAND); + + asm volatile ("mov %[ss], %%ss; .byte 0xf1" :: [ss] "m" (ss)); + } + + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; CLI\n"); + sethandler(SIGSEGV, handle_and_longjmp, SA_RESETHAND); + asm volatile ("mov %[ss], %%ss; cli" :: [ss] "m" (ss)); + } + + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; #PF\n"); + sethandler(SIGSEGV, handle_and_longjmp, SA_RESETHAND); + asm volatile ("mov %[ss], %%ss; mov (-1), %[tmp]" + : [tmp] "=r" (nr) : [ss] "m" (ss)); + } + + /* + * INT $1: if #DB has DPL=3 and there isn't special handling, + * then the kernel will die. + */ + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; INT 1\n"); + sethandler(SIGSEGV, handle_and_longjmp, SA_RESETHAND); + asm volatile ("mov %[ss], %%ss; int $1" :: [ss] "m" (ss)); + } + +#ifdef __x86_64__ + /* + * In principle, we should test 32-bit SYSCALL as well, but + * the calling convention is so unpredictable that it's + * not obviously worth the effort. + */ + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; SYSCALL\n"); + sethandler(SIGILL, handle_and_longjmp, SA_RESETHAND); + nr = SYS_getpid; + /* + * Toggle the high bit of RSP to make it noncanonical to + * strengthen this test on non-SMAP systems. + */ + asm volatile ("btc $63, %%rsp\n\t" + "mov %[ss], %%ss; syscall\n\t" + "btc $63, %%rsp" + : "+a" (nr) : [ss] "m" (ss) + : "rcx" +#ifdef __x86_64__ + , "r11" +#endif + ); + } +#endif + + printf("[RUN]\tMOV SS; breakpointed NOP\n"); + asm volatile ("mov %[ss], %%ss; breakpoint_insn: nop" :: [ss] "m" (ss)); + + /* + * Invoking SYSENTER directly breaks all the rules. Just handle + * the SIGSEGV. + */ + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; SYSENTER\n"); + stack_t stack = { + .ss_sp = altstack_data, + .ss_size = SIGSTKSZ, + }; + if (sigaltstack(&stack, NULL) != 0) + err(1, "sigaltstack"); + sethandler(SIGSEGV, handle_and_longjmp, SA_RESETHAND | SA_ONSTACK); + nr = SYS_getpid; + asm volatile ("mov %[ss], %%ss; SYSENTER" : "+a" (nr) + : [ss] "m" (ss) : "flags", "rcx" +#ifdef __x86_64__ + , "r11" +#endif + ); + + /* We're unreachable here. SYSENTER forgets RIP. */ + } + + if (sigsetjmp(jmpbuf, 1) == 0) { + printf("[RUN]\tMOV SS; INT $0x80\n"); + sethandler(SIGSEGV, handle_and_longjmp, SA_RESETHAND); + nr = 20; /* compat getpid */ + asm volatile ("mov %[ss], %%ss; int $0x80" + : "+a" (nr) : [ss] "m" (ss) + : "flags" +#ifdef __x86_64__ + , "r8", "r9", "r10", "r11" +#endif + ); + } + + printf("[OK]\tI aten't dead\n"); + return 0; +} diff --git a/tools/testing/selftests/x86/mpx-mini-test.c b/tools/testing/selftests/x86/mpx-mini-test.c index 9c0325e1ea68..50f7e9272481 100644 --- a/tools/testing/selftests/x86/mpx-mini-test.c +++ b/tools/testing/selftests/x86/mpx-mini-test.c @@ -368,6 +368,11 @@ static int expected_bnd_index = -1; uint64_t shadow_plb[NR_MPX_BOUNDS_REGISTERS][2]; /* shadow MPX bound registers */ unsigned long shadow_map[NR_MPX_BOUNDS_REGISTERS]; +/* Failed address bound checks: */ +#ifndef SEGV_BNDERR +# define SEGV_BNDERR 3 +#endif + /* * The kernel is supposed to provide some information about the bounds * exception in the siginfo. It should match what we have in the bounds @@ -419,8 +424,6 @@ void handler(int signum, siginfo_t *si, void *vucontext) br_count++; dprintf1("#BR 0x%jx (total seen: %d)\n", status, br_count); -#define SEGV_BNDERR 3 /* failed address bound checks */ - dprintf2("Saw a #BR! status 0x%jx at %016lx br_reason: %jx\n", status, ip, br_reason); dprintf2("si_signo: %d\n", si->si_signo); diff --git a/tools/testing/selftests/x86/pkey-helpers.h b/tools/testing/selftests/x86/pkey-helpers.h index b3cb7670e026..254e5436bdd9 100644 --- a/tools/testing/selftests/x86/pkey-helpers.h +++ b/tools/testing/selftests/x86/pkey-helpers.h @@ -26,30 +26,26 @@ static inline void sigsafe_printf(const char *format, ...) { va_list ap; - va_start(ap, format); if (!dprint_in_signal) { + va_start(ap, format); vprintf(format, ap); + va_end(ap); } else { int ret; - int len = vsnprintf(dprint_in_signal_buffer, - DPRINT_IN_SIGNAL_BUF_SIZE, - format, ap); /* - * len is amount that would have been printed, - * but actual write is truncated at BUF_SIZE. + * No printf() functions are signal-safe. + * They deadlock easily. Write the format + * string to get some output, even if + * incomplete. */ - if (len > DPRINT_IN_SIGNAL_BUF_SIZE) - len = DPRINT_IN_SIGNAL_BUF_SIZE; - ret = write(1, dprint_in_signal_buffer, len); + ret = write(1, format, strlen(format)); if (ret < 0) - abort(); + exit(1); } - va_end(ap); } #define dprintf_level(level, args...) do { \ if (level <= DEBUG_LEVEL) \ sigsafe_printf(args); \ - fflush(NULL); \ } while (0) #define dprintf0(args...) dprintf_level(0, args) #define dprintf1(args...) dprintf_level(1, args) diff --git a/tools/testing/selftests/x86/protection_keys.c b/tools/testing/selftests/x86/protection_keys.c index f15aa5a76fe3..460b4bdf4c1e 100644 --- a/tools/testing/selftests/x86/protection_keys.c +++ b/tools/testing/selftests/x86/protection_keys.c @@ -72,10 +72,9 @@ extern void abort_hooks(void); test_nr, iteration_nr); \ dprintf0("errno at assert: %d", errno); \ abort_hooks(); \ - assert(condition); \ + exit(__LINE__); \ } \ } while (0) -#define raw_assert(cond) assert(cond) void cat_into_file(char *str, char *file) { @@ -87,12 +86,17 @@ void cat_into_file(char *str, char *file) * these need to be raw because they are called under * pkey_assert() */ - raw_assert(fd >= 0); + if (fd < 0) { + fprintf(stderr, "error opening '%s'\n", str); + perror("error: "); + exit(__LINE__); + } + ret = write(fd, str, strlen(str)); if (ret != strlen(str)) { perror("write to file failed"); fprintf(stderr, "filename: '%s' str: '%s'\n", file, str); - raw_assert(0); + exit(__LINE__); } close(fd); } @@ -191,26 +195,30 @@ void lots_o_noops_around_write(int *write_to_me) #ifdef __i386__ #ifndef SYS_mprotect_key -# define SYS_mprotect_key 380 +# define SYS_mprotect_key 380 #endif + #ifndef SYS_pkey_alloc -# define SYS_pkey_alloc 381 -# define SYS_pkey_free 382 +# define SYS_pkey_alloc 381 +# define SYS_pkey_free 382 #endif -#define REG_IP_IDX REG_EIP -#define si_pkey_offset 0x14 + +#define REG_IP_IDX REG_EIP +#define si_pkey_offset 0x14 #else #ifndef SYS_mprotect_key -# define SYS_mprotect_key 329 +# define SYS_mprotect_key 329 #endif + #ifndef SYS_pkey_alloc -# define SYS_pkey_alloc 330 -# define SYS_pkey_free 331 +# define SYS_pkey_alloc 330 +# define SYS_pkey_free 331 #endif -#define REG_IP_IDX REG_RIP -#define si_pkey_offset 0x20 + +#define REG_IP_IDX REG_RIP +#define si_pkey_offset 0x20 #endif @@ -225,8 +233,14 @@ void dump_mem(void *dumpme, int len_bytes) } } -#define SEGV_BNDERR 3 /* failed address bound checks */ -#define SEGV_PKUERR 4 +/* Failed address bound checks: */ +#ifndef SEGV_BNDERR +# define SEGV_BNDERR 3 +#endif + +#ifndef SEGV_PKUERR +# define SEGV_PKUERR 4 +#endif static char *si_code_str(int si_code) { @@ -289,13 +303,6 @@ void signal_handler(int signum, siginfo_t *si, void *vucontext) dump_mem(pkru_ptr - 128, 256); pkey_assert(*pkru_ptr); - si_pkey_ptr = (u32 *)(((u8 *)si) + si_pkey_offset); - dprintf1("si_pkey_ptr: %p\n", si_pkey_ptr); - dump_mem(si_pkey_ptr - 8, 24); - siginfo_pkey = *si_pkey_ptr; - pkey_assert(siginfo_pkey < NR_PKEYS); - last_si_pkey = siginfo_pkey; - if ((si->si_code == SEGV_MAPERR) || (si->si_code == SEGV_ACCERR) || (si->si_code == SEGV_BNDERR)) { @@ -303,6 +310,13 @@ void signal_handler(int signum, siginfo_t *si, void *vucontext) exit(4); } + si_pkey_ptr = (u32 *)(((u8 *)si) + si_pkey_offset); + dprintf1("si_pkey_ptr: %p\n", si_pkey_ptr); + dump_mem((u8 *)si_pkey_ptr - 8, 24); + siginfo_pkey = *si_pkey_ptr; + pkey_assert(siginfo_pkey < NR_PKEYS); + last_si_pkey = siginfo_pkey; + dprintf1("signal pkru from xsave: %08x\n", *pkru_ptr); /* need __rdpkru() version so we do not do shadow_pkru checking */ dprintf1("signal pkru from pkru: %08x\n", __rdpkru()); @@ -311,22 +325,6 @@ void signal_handler(int signum, siginfo_t *si, void *vucontext) dprintf1("WARNING: set PRKU=0 to allow faulting instruction to continue\n"); pkru_faults++; dprintf1("<<<<==================================================\n"); - return; - if (trapno == 14) { - fprintf(stderr, - "ERROR: In signal handler, page fault, trapno = %d, ip = %016lx\n", - trapno, ip); - fprintf(stderr, "si_addr %p\n", si->si_addr); - fprintf(stderr, "REG_ERR: %lx\n", - (unsigned long)uctxt->uc_mcontext.gregs[REG_ERR]); - exit(1); - } else { - fprintf(stderr, "unexpected trap %d! at 0x%lx\n", trapno, ip); - fprintf(stderr, "si_addr %p\n", si->si_addr); - fprintf(stderr, "REG_ERR: %lx\n", - (unsigned long)uctxt->uc_mcontext.gregs[REG_ERR]); - exit(2); - } dprint_in_signal = 0; } @@ -393,10 +391,15 @@ pid_t fork_lazy_child(void) return forkret; } -#define PKEY_DISABLE_ACCESS 0x1 -#define PKEY_DISABLE_WRITE 0x2 +#ifndef PKEY_DISABLE_ACCESS +# define PKEY_DISABLE_ACCESS 0x1 +#endif + +#ifndef PKEY_DISABLE_WRITE +# define PKEY_DISABLE_WRITE 0x2 +#endif -u32 pkey_get(int pkey, unsigned long flags) +static u32 hw_pkey_get(int pkey, unsigned long flags) { u32 mask = (PKEY_DISABLE_ACCESS|PKEY_DISABLE_WRITE); u32 pkru = __rdpkru(); @@ -418,7 +421,7 @@ u32 pkey_get(int pkey, unsigned long flags) return masked_pkru; } -int pkey_set(int pkey, unsigned long rights, unsigned long flags) +static int hw_pkey_set(int pkey, unsigned long rights, unsigned long flags) { u32 mask = (PKEY_DISABLE_ACCESS|PKEY_DISABLE_WRITE); u32 old_pkru = __rdpkru(); @@ -452,15 +455,15 @@ void pkey_disable_set(int pkey, int flags) pkey, flags); pkey_assert(flags & (PKEY_DISABLE_ACCESS | PKEY_DISABLE_WRITE)); - pkey_rights = pkey_get(pkey, syscall_flags); + pkey_rights = hw_pkey_get(pkey, syscall_flags); - dprintf1("%s(%d) pkey_get(%d): %x\n", __func__, + dprintf1("%s(%d) hw_pkey_get(%d): %x\n", __func__, pkey, pkey, pkey_rights); pkey_assert(pkey_rights >= 0); pkey_rights |= flags; - ret = pkey_set(pkey, pkey_rights, syscall_flags); + ret = hw_pkey_set(pkey, pkey_rights, syscall_flags); assert(!ret); /*pkru and flags have the same format */ shadow_pkru |= flags << (pkey * 2); @@ -468,8 +471,8 @@ void pkey_disable_set(int pkey, int flags) pkey_assert(ret >= 0); - pkey_rights = pkey_get(pkey, syscall_flags); - dprintf1("%s(%d) pkey_get(%d): %x\n", __func__, + pkey_rights = hw_pkey_get(pkey, syscall_flags); + dprintf1("%s(%d) hw_pkey_get(%d): %x\n", __func__, pkey, pkey, pkey_rights); dprintf1("%s(%d) pkru: 0x%x\n", __func__, pkey, rdpkru()); @@ -483,24 +486,24 @@ void pkey_disable_clear(int pkey, int flags) { unsigned long syscall_flags = 0; int ret; - int pkey_rights = pkey_get(pkey, syscall_flags); + int pkey_rights = hw_pkey_get(pkey, syscall_flags); u32 orig_pkru = rdpkru(); pkey_assert(flags & (PKEY_DISABLE_ACCESS | PKEY_DISABLE_WRITE)); - dprintf1("%s(%d) pkey_get(%d): %x\n", __func__, + dprintf1("%s(%d) hw_pkey_get(%d): %x\n", __func__, pkey, pkey, pkey_rights); pkey_assert(pkey_rights >= 0); pkey_rights |= flags; - ret = pkey_set(pkey, pkey_rights, 0); + ret = hw_pkey_set(pkey, pkey_rights, 0); /* pkru and flags have the same format */ shadow_pkru &= ~(flags << (pkey * 2)); pkey_assert(ret >= 0); - pkey_rights = pkey_get(pkey, syscall_flags); - dprintf1("%s(%d) pkey_get(%d): %x\n", __func__, + pkey_rights = hw_pkey_get(pkey, syscall_flags); + dprintf1("%s(%d) hw_pkey_get(%d): %x\n", __func__, pkey, pkey, pkey_rights); dprintf1("%s(%d) pkru: 0x%x\n", __func__, pkey, rdpkru()); @@ -674,10 +677,12 @@ int mprotect_pkey(void *ptr, size_t size, unsigned long orig_prot, struct pkey_malloc_record { void *ptr; long size; + int prot; }; struct pkey_malloc_record *pkey_malloc_records; +struct pkey_malloc_record *pkey_last_malloc_record; long nr_pkey_malloc_records; -void record_pkey_malloc(void *ptr, long size) +void record_pkey_malloc(void *ptr, long size, int prot) { long i; struct pkey_malloc_record *rec = NULL; @@ -709,6 +714,8 @@ void record_pkey_malloc(void *ptr, long size) (int)(rec - pkey_malloc_records), rec, ptr, size); rec->ptr = ptr; rec->size = size; + rec->prot = prot; + pkey_last_malloc_record = rec; nr_pkey_malloc_records++; } @@ -753,7 +760,7 @@ void *malloc_pkey_with_mprotect(long size, int prot, u16 pkey) pkey_assert(ptr != (void *)-1); ret = mprotect_pkey((void *)ptr, PAGE_SIZE, prot, pkey); pkey_assert(!ret); - record_pkey_malloc(ptr, size); + record_pkey_malloc(ptr, size, prot); rdpkru(); dprintf1("%s() for pkey %d @ %p\n", __func__, pkey, ptr); @@ -774,7 +781,7 @@ void *malloc_pkey_anon_huge(long size, int prot, u16 pkey) size = ALIGN_UP(size, HPAGE_SIZE * 2); ptr = mmap(NULL, size, PROT_NONE, MAP_ANONYMOUS|MAP_PRIVATE, -1, 0); pkey_assert(ptr != (void *)-1); - record_pkey_malloc(ptr, size); + record_pkey_malloc(ptr, size, prot); mprotect_pkey(ptr, size, prot, pkey); dprintf1("unaligned ptr: %p\n", ptr); @@ -847,7 +854,7 @@ void *malloc_pkey_hugetlb(long size, int prot, u16 pkey) pkey_assert(ptr != (void *)-1); mprotect_pkey(ptr, size, prot, pkey); - record_pkey_malloc(ptr, size); + record_pkey_malloc(ptr, size, prot); dprintf1("mmap()'d hugetlbfs for pkey %d @ %p\n", pkey, ptr); return ptr; @@ -869,7 +876,7 @@ void *malloc_pkey_mmap_dax(long size, int prot, u16 pkey) mprotect_pkey(ptr, size, prot, pkey); - record_pkey_malloc(ptr, size); + record_pkey_malloc(ptr, size, prot); dprintf1("mmap()'d for pkey %d @ %p\n", pkey, ptr); close(fd); @@ -918,13 +925,21 @@ void *malloc_pkey(long size, int prot, u16 pkey) } int last_pkru_faults; +#define UNKNOWN_PKEY -2 void expected_pk_fault(int pkey) { dprintf2("%s(): last_pkru_faults: %d pkru_faults: %d\n", __func__, last_pkru_faults, pkru_faults); dprintf2("%s(%d): last_si_pkey: %d\n", __func__, pkey, last_si_pkey); pkey_assert(last_pkru_faults + 1 == pkru_faults); - pkey_assert(last_si_pkey == pkey); + + /* + * For exec-only memory, we do not know the pkey in + * advance, so skip this check. + */ + if (pkey != UNKNOWN_PKEY) + pkey_assert(last_si_pkey == pkey); + /* * The signal handler shold have cleared out PKRU to let the * test program continue. We now have to restore it. @@ -939,10 +954,11 @@ void expected_pk_fault(int pkey) last_si_pkey = -1; } -void do_not_expect_pk_fault(void) -{ - pkey_assert(last_pkru_faults == pkru_faults); -} +#define do_not_expect_pk_fault(msg) do { \ + if (last_pkru_faults != pkru_faults) \ + dprintf0("unexpected PK fault: %s\n", msg); \ + pkey_assert(last_pkru_faults == pkru_faults); \ +} while (0) int test_fds[10] = { -1 }; int nr_test_fds; @@ -1151,12 +1167,15 @@ void test_pkey_alloc_exhaust(int *ptr, u16 pkey) pkey_assert(i < NR_PKEYS*2); /* - * There are 16 pkeys supported in hardware. One is taken - * up for the default (0) and another can be taken up by - * an execute-only mapping. Ensure that we can allocate - * at least 14 (16-2). + * There are 16 pkeys supported in hardware. Three are + * allocated by the time we get here: + * 1. The default key (0) + * 2. One possibly consumed by an execute-only mapping. + * 3. One allocated by the test code and passed in via + * 'pkey' to this function. + * Ensure that we can allocate at least another 13 (16-3). */ - pkey_assert(i >= NR_PKEYS-2); + pkey_assert(i >= NR_PKEYS-3); for (i = 0; i < nr_allocated_pkeys; i++) { err = sys_pkey_free(allocated_pkeys[i]); @@ -1165,6 +1184,35 @@ void test_pkey_alloc_exhaust(int *ptr, u16 pkey) } } +/* + * pkey 0 is special. It is allocated by default, so you do not + * have to call pkey_alloc() to use it first. Make sure that it + * is usable. + */ +void test_mprotect_with_pkey_0(int *ptr, u16 pkey) +{ + long size; + int prot; + + assert(pkey_last_malloc_record); + size = pkey_last_malloc_record->size; + /* + * This is a bit of a hack. But mprotect() requires + * huge-page-aligned sizes when operating on hugetlbfs. + * So, make sure that we use something that's a multiple + * of a huge page when we can. + */ + if (size >= HPAGE_SIZE) + size = HPAGE_SIZE; + prot = pkey_last_malloc_record->prot; + + /* Use pkey 0 */ + mprotect_pkey(ptr, size, prot, 0); + + /* Make sure that we can set it back to the original pkey. */ + mprotect_pkey(ptr, size, prot, pkey); +} + void test_ptrace_of_child(int *ptr, u16 pkey) { __attribute__((__unused__)) int peek_result; @@ -1228,7 +1276,7 @@ void test_ptrace_of_child(int *ptr, u16 pkey) pkey_assert(ret != -1); /* Now access from the current task, and expect NO exception: */ peek_result = read_ptr(plain_ptr); - do_not_expect_pk_fault(); + do_not_expect_pk_fault("read plain pointer after ptrace"); ret = ptrace(PTRACE_DETACH, child_pid, ignored, 0); pkey_assert(ret != -1); @@ -1241,12 +1289,9 @@ void test_ptrace_of_child(int *ptr, u16 pkey) free(plain_ptr_unaligned); } -void test_executing_on_unreadable_memory(int *ptr, u16 pkey) +void *get_pointer_to_instructions(void) { void *p1; - int scratch; - int ptr_contents; - int ret; p1 = ALIGN_PTR_UP(&lots_o_noops_around_write, PAGE_SIZE); dprintf3("&lots_o_noops: %p\n", &lots_o_noops_around_write); @@ -1256,7 +1301,23 @@ void test_executing_on_unreadable_memory(int *ptr, u16 pkey) /* Point 'p1' at the *second* page of the function: */ p1 += PAGE_SIZE; + /* + * Try to ensure we fault this in on next touch to ensure + * we get an instruction fault as opposed to a data one + */ madvise(p1, PAGE_SIZE, MADV_DONTNEED); + + return p1; +} + +void test_executing_on_unreadable_memory(int *ptr, u16 pkey) +{ + void *p1; + int scratch; + int ptr_contents; + int ret; + + p1 = get_pointer_to_instructions(); lots_o_noops_around_write(&scratch); ptr_contents = read_ptr(p1); dprintf2("ptr (%p) contents@%d: %x\n", p1, __LINE__, ptr_contents); @@ -1272,12 +1333,55 @@ void test_executing_on_unreadable_memory(int *ptr, u16 pkey) */ madvise(p1, PAGE_SIZE, MADV_DONTNEED); lots_o_noops_around_write(&scratch); - do_not_expect_pk_fault(); + do_not_expect_pk_fault("executing on PROT_EXEC memory"); ptr_contents = read_ptr(p1); dprintf2("ptr (%p) contents@%d: %x\n", p1, __LINE__, ptr_contents); expected_pk_fault(pkey); } +void test_implicit_mprotect_exec_only_memory(int *ptr, u16 pkey) +{ + void *p1; + int scratch; + int ptr_contents; + int ret; + + dprintf1("%s() start\n", __func__); + + p1 = get_pointer_to_instructions(); + lots_o_noops_around_write(&scratch); + ptr_contents = read_ptr(p1); + dprintf2("ptr (%p) contents@%d: %x\n", p1, __LINE__, ptr_contents); + + /* Use a *normal* mprotect(), not mprotect_pkey(): */ + ret = mprotect(p1, PAGE_SIZE, PROT_EXEC); + pkey_assert(!ret); + + dprintf2("pkru: %x\n", rdpkru()); + + /* Make sure this is an *instruction* fault */ + madvise(p1, PAGE_SIZE, MADV_DONTNEED); + lots_o_noops_around_write(&scratch); + do_not_expect_pk_fault("executing on PROT_EXEC memory"); + ptr_contents = read_ptr(p1); + dprintf2("ptr (%p) contents@%d: %x\n", p1, __LINE__, ptr_contents); + expected_pk_fault(UNKNOWN_PKEY); + + /* + * Put the memory back to non-PROT_EXEC. Should clear the + * exec-only pkey off the VMA and allow it to be readable + * again. Go to PROT_NONE first to check for a kernel bug + * that did not clear the pkey when doing PROT_NONE. + */ + ret = mprotect(p1, PAGE_SIZE, PROT_NONE); + pkey_assert(!ret); + + ret = mprotect(p1, PAGE_SIZE, PROT_READ|PROT_EXEC); + pkey_assert(!ret); + ptr_contents = read_ptr(p1); + do_not_expect_pk_fault("plain read on recently PROT_EXEC area"); +} + void test_mprotect_pkey_on_unsupported_cpu(int *ptr, u16 pkey) { int size = PAGE_SIZE; @@ -1302,6 +1406,8 @@ void (*pkey_tests[])(int *ptr, u16 pkey) = { test_kernel_gup_of_access_disabled_region, test_kernel_gup_write_to_write_disabled_region, test_executing_on_unreadable_memory, + test_implicit_mprotect_exec_only_memory, + test_mprotect_with_pkey_0, test_ptrace_of_child, test_pkey_syscalls_on_non_allocated_pkey, test_pkey_syscalls_bad_args, |