x86/alternative: Use .ibt_endbr_seal to seal indirect calls

Objtool's --ibt option generates .ibt_endbr_seal which lists
superfluous ENDBR instructions. That is those instructions for which
the function is never indirectly called.

Overwrite these ENDBR instructions with a NOP4 such that these
function can never be indirect called, reducing the number of viable
ENDBR targets in the kernel.

Signed-off-by: Peter Zijlstra (Intel) <[e-mail]>
Acked-by: Josh Poimboeuf <[e-mail]>
Link: https://lore.kernel.org/r/[e-mail]

arch/um/kernel/um_arch.c

···
 	os_check_bugs();
 }
 
+void apply_ibt_endbr(s32 *start, s32 *end)
+{
+}
+
 void apply_retpolines(s32 *start, s32 *end)
 {
 }

arch/x86/Kconfig

···
 config X86_KERNEL_IBT
 	prompt "Indirect Branch Tracking"
 	bool
-	depends on X86_64 && CC_HAS_IBT
+	depends on X86_64 && CC_HAS_IBT && STACK_VALIDATION
 	help
 	  Build the kernel with support for Indirect Branch Tracking, a
 	  hardware support course-grain forward-edge Control Flow Integrity
 	  protection. It enforces that all indirect calls must land on
 	  an ENDBR instruction, as such, the compiler will instrument the
 	  code with them to make this happen.
+
+	  In addition to building the kernel with IBT, seal all functions that
+	  are not indirect call targets, avoiding them ever becomming one.
+
+	  This requires LTO like objtool runs and will slow down the build. It
+	  does significantly reduce the number of ENDBR instructions in the
+	  kernel image.
 
 config X86_INTEL_MEMORY_PROTECTION_KEYS
 	prompt "Memory Protection Keys"

arch/x86/include/asm/alternative.h

···
 extern void alternative_instructions(void);
 extern void apply_alternatives(struct alt_instr *start, struct alt_instr *end);
 extern void apply_retpolines(s32 *start, s32 *end);
+extern void apply_ibt_endbr(s32 *start, s32 *end);
 
 struct module;
 

arch/x86/include/asm/ibt.h

···
 	return endbr;
 }
 
+static inline __attribute_const__ u32 gen_endbr_poison(void)
+{
+	/*
+	 * 4 byte NOP that isn't NOP4 (in fact it is OSP NOP3), such that it
+	 * will be unique to (former) ENDBR sites.
+	 */
+	return 0x001f0f66; /* osp nopl (%rax) */
+}
+
 static inline bool is_endbr(u32 val)
 {
+	if (val == gen_endbr_poison())
+		return true;
+
 	val &= ~0x01000000U; /* ENDBR32 -> ENDBR64 */
 	return val == gen_endbr();
 }

arch/x86/kernel/alternative.c

···
 }
 
 extern s32 __retpoline_sites[], __retpoline_sites_end[];
+extern s32 __ibt_endbr_seal[], __ibt_endbr_seal_end[];
 extern struct alt_instr __alt_instructions[], __alt_instructions_end[];
 extern s32 __smp_locks[], __smp_locks_end[];
 void text_poke_early(void *addr, const void *opcode, size_t len);
···
 
 #endif /* CONFIG_RETPOLINE && CONFIG_STACK_VALIDATION */
 
+#ifdef CONFIG_X86_KERNEL_IBT
+
+/*
+ * Generated by: objtool --ibt
+ */
+void __init_or_module noinline apply_ibt_endbr(s32 *start, s32 *end)
+{
+	s32 *s;
+
+	for (s = start; s < end; s++) {
+		u32 endbr, poison = gen_endbr_poison();
+		void *addr = (void *)s + *s;
+
+		if (WARN_ON_ONCE(get_kernel_nofault(endbr, addr)))
+			continue;
+
+		if (WARN_ON_ONCE(!is_endbr(endbr)))
+			continue;
+
+		DPRINTK("ENDBR at: %pS (%px)", addr, addr);
+
+		/*
+		 * When we have IBT, the lack of ENDBR will trigger #CP
+		 */
+		DUMP_BYTES(((u8*)addr), 4, "%px: orig: ", addr);
+		DUMP_BYTES(((u8*)&poison), 4, "%px: repl: ", addr);
+		text_poke_early(addr, &poison, 4);
+	}
+}
+
+#else
+
+void __init_or_module noinline apply_ibt_endbr(s32 *start, s32 *end) { }
+
+#endif /* CONFIG_X86_KERNEL_IBT */
+
 #ifdef CONFIG_SMP
 static void alternatives_smp_lock(const s32 *start, const s32 *end,
 				  u8 *text, u8 *text_end)
···
 	 * alternatives can be overwritten by their immediate fragments.
 	 */
 	apply_alternatives(__alt_instructions, __alt_instructions_end);
+
+	apply_ibt_endbr(__ibt_endbr_seal, __ibt_endbr_seal_end);
 
 #ifdef CONFIG_SMP
 	/* Patch to UP if other cpus not imminent. */

arch/x86/kernel/module.c

···
 {
 	const Elf_Shdr *s, *text = NULL, *alt = NULL, *locks = NULL,
 		*para = NULL, *orc = NULL, *orc_ip = NULL,
-		*retpolines = NULL;
+		*retpolines = NULL, *ibt_endbr = NULL;
 	char *secstrings = (void *)hdr + sechdrs[hdr->e_shstrndx].sh_offset;
 
 	for (s = sechdrs; s < sechdrs + hdr->e_shnum; s++) {
···
 			orc_ip = s;
 		if (!strcmp(".retpoline_sites", secstrings + s->sh_name))
 			retpolines = s;
+		if (!strcmp(".ibt_endbr_seal", secstrings + s->sh_name))
+			ibt_endbr = s;
 	}
 
 	/*
···
 		/* patch .altinstructions */
 		void *aseg = (void *)alt->sh_addr;
 		apply_alternatives(aseg, aseg + alt->sh_size);
+	}
+	if (ibt_endbr) {
+		void *iseg = (void *)ibt_endbr->sh_addr;
+		apply_ibt_endbr(iseg, iseg + ibt_endbr->sh_size);
 	}
 	if (locks && text) {
 		void *lseg = (void *)locks->sh_addr;

scripts/Makefile.build

···
 targets-for-builtin += $(obj)/built-in.a
 endif
 
-targets-for-modules := $(patsubst %.o, %.mod, $(filter %.o, $(obj-m)))
+targets-for-modules :=
 
 ifdef CONFIG_LTO_CLANG
 targets-for-modules += $(patsubst %.o, %.lto.o, $(filter %.o, $(obj-m)))
 endif
+
+ifdef CONFIG_X86_KERNEL_IBT
+targets-for-modules += $(patsubst %.o, %.objtool, $(filter %.o, $(obj-m)))
+endif
+
+targets-for-modules += $(patsubst %.o, %.mod, $(filter %.o, $(obj-m)))
 
 ifdef need-modorder
 targets-for-modules += $(obj)/modules.order
···
 objtool_args =								\
 	$(if $(CONFIG_UNWINDER_ORC),orc generate,check)			\
 	$(if $(part-of-module), --module)				\
+	$(if $(CONFIG_X86_KERNEL_IBT), --lto --ibt)			\
 	$(if $(CONFIG_FRAME_POINTER),, --no-fp)				\
 	$(if $(CONFIG_GCOV_KERNEL)$(CONFIG_LTO_CLANG), --no-unreachable)\
 	$(if $(CONFIG_RETPOLINE), --retpoline)				\
···
 	$(if $(CONFIG_FTRACE_MCOUNT_USE_OBJTOOL), --mcount)		\
 	$(if $(CONFIG_SLS), --sls)
 
-cmd_objtool = $(if $(objtool-enabled), ; $(objtool) $(objtool_args) $@)
-cmd_gen_objtooldep = $(if $(objtool-enabled), { echo ; echo '$@: $$(wildcard $(objtool))' ; } >> $(dot-target).cmd)
+cmd_objtool = $(if $(objtool-enabled), ; $(objtool) $(objtool_args) $(@:.objtool=.o))
+cmd_gen_objtooldep = $(if $(objtool-enabled), { echo ; echo '$(@:.objtool=.o): $$(wildcard $(objtool))' ; } >> $(dot-target).cmd)
 
 endif # CONFIG_STACK_VALIDATION
 
···
 
 # Skip objtool for LLVM bitcode
 $(obj)/%.o: objtool-enabled :=
+
+# objtool was skipped for LLVM bitcode, run it now that we have compiled
+# modules into native code
+$(obj)/%.lto.o: objtool-enabled = y
+$(obj)/%.lto.o: part-of-module := y
+
+else ifdef CONFIG_X86_KERNEL_IBT
+
+# Skip objtool on individual files
+$(obj)/%.o: objtool-enabled :=
+
+# instead run objtool on the module as a whole, right before
+# the final link pass with the linker script.
+$(obj)/%.objtool: objtool-enabled = y
+$(obj)/%.objtool: part-of-module := y
 
 else
 
···
 # Module .o files may contain LLVM bitcode, compile them into native code
 # before ELF processing
 quiet_cmd_cc_lto_link_modules = LTO [M] $@
-cmd_cc_lto_link_modules =						\
+      cmd_cc_lto_link_modules =						\
 	$(LD) $(ld_flags) -r -o $@					\
 		$(shell [ -s $(@:.lto.o=.o.symversions) ] &&		\
 			echo -T $(@:.lto.o=.o.symversions))		\
 		--whole-archive $(filter-out FORCE,$^)			\
 		$(cmd_objtool)
-
-# objtool was skipped for LLVM bitcode, run it now that we have compiled
-# modules into native code
-$(obj)/%.lto.o: objtool-enabled = y
-$(obj)/%.lto.o: part-of-module := y
 
 $(obj)/%.lto.o: $(obj)/%.o FORCE
 	$(call if_changed,cc_lto_link_modules)
···
 
 $(obj)/%.mod: $(obj)/%$(mod-prelink-ext).o FORCE
 	$(call if_changed,mod)
+
+#
+# Since objtool will re-write the file it will change the timestamps, therefore
+# it is critical that the %.objtool file gets a timestamp *after* objtool runs.
+#
+# Additionally, care must be had with ordering this rule against the other rules
+# that take %.o as a dependency.
+#
+cmd_objtool_mod = true $(cmd_objtool) ; touch $@
+
+$(obj)/%.objtool: $(obj)/%$(mod-prelink-ext).o FORCE
+	$(call if_changed,objtool_mod)
 
 quiet_cmd_cc_lst_c = MKLST   $@
       cmd_cc_lst_c = $(CC) $(c_flags) -g -c -o $*.o $< && \

scripts/link-vmlinux.sh

···
 	local objtoolcmd;
 	local objtoolopt;
 
-	if is_enabled CONFIG_LTO_CLANG && is_enabled CONFIG_STACK_VALIDATION; then
+	if is_enabled CONFIG_STACK_VALIDATION && \
+	   ( is_enabled CONFIG_LTO_CLANG || is_enabled CONFIG_X86_KERNEL_IBT ); then
+
 		# Don't perform vmlinux validation unless explicitly requested,
 		# but run objtool on vmlinux.o now that we have an object file.
 		if is_enabled CONFIG_UNWINDER_ORC; then
···
 		fi
 
 		objtoolopt="${objtoolopt} --lto"
+
+		if is_enabled CONFIG_X86_KERNEL_IBT; then
+			objtoolopt="${objtoolopt} --ibt"
+		fi
 
 		if is_enabled CONFIG_FTRACE_MCOUNT_USE_OBJTOOL; then
 			objtoolopt="${objtoolopt} --mcount"
···
 	# skip output file argument
 	shift
 
-	if is_enabled CONFIG_LTO_CLANG; then
+	if is_enabled CONFIG_LTO_CLANG || is_enabled CONFIG_X86_KERNEL_IBT; then
 		# Use vmlinux.o instead of performing the slow LTO link again.
 		objs=vmlinux.o
 		libs=