[RFC PATCH v2 6/9] um: Add SECCOMP support detection and initialization

[Benjamin Berg <benjamin@sipsolutions.net>]

This detects seccomp support, sets the global using_seccomp variable and
initilizes the exec registers. For now, the implementation simply falls
through to the ptrace startup code, meaning that it is unused.

Signed-off-by: Benjamin Berg <benjamin@sipsolutions.net>
Signed-off-by: Benjamin Berg <benjamin.berg@intel.com>
---
 arch/um/include/shared/skas/skas.h |   6 ++
 arch/um/os-Linux/registers.c       |   4 +-
 arch/um/os-Linux/skas/process.c    |   3 +
 arch/um/os-Linux/start_up.c        | 151 ++++++++++++++++++++++++++++-
 4 files changed, 160 insertions(+), 4 deletions(-)

diff --git a/arch/um/include/shared/skas/skas.h b/arch/um/include/shared/skas/skas.h
index 85c50122ab98..2ff01c773483 100644
--- a/arch/um/include/shared/skas/skas.h
+++ b/arch/um/include/shared/skas/skas.h
@@ -6,8 +6,14 @@
 #ifndef __SKAS_H
 #define __SKAS_H
 
+#include <linux/kconfig.h>
 #include <sysdep/ptrace.h>
 
+#ifdef CONFIG_UML_SECCOMP
+extern int using_seccomp;
+#else
+#define using_seccomp 0
+#endif
 extern int userspace_pid[];
 
 extern void new_thread_handler(void);
diff --git a/arch/um/os-Linux/registers.c b/arch/um/os-Linux/registers.c
index d7ca148807b2..bfba2cbc9478 100644
--- a/arch/um/os-Linux/registers.c
+++ b/arch/um/os-Linux/registers.c
@@ -14,8 +14,8 @@
 
 /* This is set once at boot time and not changed thereafter */
 
-static unsigned long exec_regs[MAX_REG_NR];
-static unsigned long *exec_fp_regs;
+unsigned long exec_regs[MAX_REG_NR];
+unsigned long *exec_fp_regs;
 
 int init_pid_registers(int pid)
 {
diff --git a/arch/um/os-Linux/skas/process.c b/arch/um/os-Linux/skas/process.c
index e99fdb100715..5eb0155ff7a7 100644
--- a/arch/um/os-Linux/skas/process.c
+++ b/arch/um/os-Linux/skas/process.c
@@ -299,6 +299,9 @@ static int __init init_stub_exe_fd(void)
 }
 __initcall(init_stub_exe_fd);
 
+#ifdef CONFIG_UML_SECCOMP
+int using_seccomp;
+#endif
 int userspace_pid[NR_CPUS];
 
 /**
diff --git a/arch/um/os-Linux/start_up.c b/arch/um/os-Linux/start_up.c
index 93fc82c01aba..f1064817e719 100644
--- a/arch/um/os-Linux/start_up.c
+++ b/arch/um/os-Linux/start_up.c
@@ -1,8 +1,10 @@
 // SPDX-License-Identifier: GPL-2.0
 /*
+ * Copyright (C) 2021 Benjamin Berg <benjamin@sipsolutions.net>
  * Copyright (C) 2000 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com)
  */
 
+#include <linux/kconfig.h>
 #include <stdio.h>
 #include <stdlib.h>
 #include <stdarg.h>
@@ -24,6 +26,15 @@
 #include <kern_util.h>
 #include <mem_user.h>
 #include <ptrace_user.h>
+#ifdef CONFIG_UML_SECCOMP
+#include <stdbool.h>
+#include <stub-data.h>
+#include <sys/prctl.h>
+#include <linux/seccomp.h>
+#include <linux/filter.h>
+#include <sysdep/mcontext.h>
+#include <sysdep/stub.h>
+#endif
 #include <registers.h>
 #include <skas.h>
 #include "internal.h"
@@ -224,6 +235,129 @@ static void __init check_ptrace(void)
 	check_sysemu();
 }
 
+#ifdef CONFIG_UML_SECCOMP
+extern unsigned long host_fp_size;
+extern unsigned long exec_regs[MAX_REG_NR];
+extern unsigned long *exec_fp_regs;
+
+__initdata static struct stub_data *seccomp_test_stub_data;
+
+static void __init sigsys_handler(int sig, siginfo_t *info, void *p)
+{
+	ucontext_t *uc = p;
+
+	/* Stow away the location of the mcontext in the stack */
+	seccomp_test_stub_data->mctx_offset = (unsigned long)&uc->uc_mcontext -
+					      (unsigned long)&seccomp_test_stub_data->sigstack[0];
+	exit(0);
+}
+
+static bool __init init_seccomp(void)
+{
+	int pid;
+	int status;
+	int n;
+
+	/* We check that we can install a seccomp filter and then exit(0)
+	 * from a trapped syscall.
+	 *
+	 * Note that we cannot verify that no seccomp filter already exists
+	 * for a syscall that results in the process/thread to be killed.
+	 */
+
+	os_info("Checking that seccomp filters can be installed...");
+
+	/* data needs to be page aligned, so allocate twice the amount */
+	seccomp_test_stub_data = mmap(0, sizeof(*seccomp_test_stub_data),
+				      PROT_READ | PROT_WRITE,
+				      MAP_SHARED | MAP_ANON, 0, 0);
+
+	pid = fork();
+	if (pid == 0) {
+		static struct sock_filter filter[] = {
+			BPF_STMT(BPF_LD | BPF_W | BPF_ABS,
+				offsetof(struct seccomp_data, nr)),
+			BPF_JUMP(BPF_JMP | BPF_JEQ | BPF_K, __NR_clock_nanosleep, 1, 0),
+			BPF_STMT(BPF_RET | BPF_K, SECCOMP_RET_ALLOW),
+			BPF_STMT(BPF_RET | BPF_K, SECCOMP_RET_TRAP),
+		};
+		static struct sock_fprog prog = {
+			.len = ARRAY_SIZE(filter),
+			.filter = filter,
+		};
+		struct sigaction sa;
+
+		set_sigstack(seccomp_test_stub_data->sigstack,
+			     sizeof(seccomp_test_stub_data->sigstack));
+
+		sa.sa_flags = SA_ONSTACK | SA_NODEFER | SA_SIGINFO;
+		sa.sa_sigaction = (void *) sigsys_handler;
+		sa.sa_restorer = NULL;
+		if (sigaction(SIGSYS, &sa, NULL) < 0)
+			exit(1);
+
+		prctl(PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0);
+		if (syscall(__NR_seccomp, SECCOMP_SET_MODE_FILTER,
+			    SECCOMP_FILTER_FLAG_TSYNC, &prog) != 0)
+			exit(2);
+
+		sleep(0);
+
+		/* Never reached. */
+		exit(3);
+	}
+
+	if (pid < 0)
+		fatal_perror("check_seccomp : fork failed");
+
+	CATCH_EINTR(n = waitpid(pid, &status, 0));
+	if (n < 0)
+		fatal_perror("check_seccomp : waitpid failed");
+
+	if (WIFEXITED(status) && WEXITSTATUS(status) == 0) {
+		struct uml_pt_regs *regs;
+		unsigned long fp_size;
+		int r;
+
+		/* Fill in the host_fp_size from the mcontext. */
+		regs = calloc(1, sizeof(struct uml_pt_regs));
+		get_stub_state(regs, seccomp_test_stub_data, &fp_size);
+		host_fp_size = fp_size;
+		free(regs);
+
+		/* Repeat with the correct size */
+		regs = calloc(1, sizeof(struct uml_pt_regs) + host_fp_size);
+		r = get_stub_state(regs, seccomp_test_stub_data, NULL);
+
+		/* Store as the default startup registers */
+		exec_fp_regs = malloc(host_fp_size);
+		memcpy(exec_regs, regs->gp, sizeof(exec_regs));
+		memcpy(exec_fp_regs, regs->fp, host_fp_size);
+
+		munmap(seccomp_test_stub_data, sizeof(*seccomp_test_stub_data));
+
+		free(regs);
+
+		if (r) {
+			os_info("failed to fetch registers: %d\n", r);
+			return false;
+		}
+
+		os_info("OK\n");
+		return true;
+	}
+
+	if (WIFEXITED(status) && WEXITSTATUS(status) == 2)
+		os_info("missing\n");
+	else
+		os_info("error\n");
+
+	munmap(seccomp_test_stub_data, sizeof(*seccomp_test_stub_data));
+	return false;
+}
+#endif
+
+
 static void __init check_coredump_limit(void)
 {
 	struct rlimit lim;
@@ -286,13 +420,26 @@ void __init os_early_checks(void)
 	/* Print out the core dump limits early */
 	check_coredump_limit();
 
-	check_ptrace();
-
 	/* Need to check this early because mmapping happens before the
 	 * kernel is running.
 	 */
 	check_tmpexec();
 
+#ifdef CONFIG_UML_SECCOMP
+	using_seccomp = 0;
+
+	if (init_seccomp()) {
+		/* Not yet fully implemented */
+#if 0
+		using_seccomp = 1;
+
+		return;
+#endif
+	}
+#endif
+
+	check_ptrace();
+
 	pid = start_ptraced_child();
 	if (init_pid_registers(pid))
 		fatal("Failed to initialize default registers");
-- 
2.47.0