Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1763563AbXKTSdd (ORCPT ); Tue, 20 Nov 2007 13:33:33 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1761713AbXKTS1H (ORCPT ); Tue, 20 Nov 2007 13:27:07 -0500 Received: from [198.99.130.12] ([198.99.130.12]:53245 "EHLO saraswathi.solana.com" rhost-flags-FAIL-FAIL-OK-OK) by vger.kernel.org with ESMTP id S1761700AbXKTS1B (ORCPT ); Tue, 20 Nov 2007 13:27:01 -0500 Date: Tue, 20 Nov 2007 13:26:33 -0500 From: Jeff Dike To: Andrew Morton Cc: LKML , uml-devel Subject: [PATCH 3/4] UML - kill processes instead of panicing kernel Message-ID: <20071120182633.GA7694@c2.user-mode-linux.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.4.2.3i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 25455 Lines: 756 UML was panicing in the case of failures of libc calls which shouldn't happen. This is an overreaction since a failure from libc doesn't normally mean that kernel data structures are in an unknown state. Instead, the current process should just be killed if there is no way to recover. The case that prompted this was a failure of PTRACE_SETREGS restoring the same state that was read by PTRACE_GETREGS. It appears that when a process tries to load a bogus value into a segment register, it segfaults (as expected) and the value is actually loaded and is seen by PTRACE_GETREGS (not expected). This case is fixed by forcing a fatal SIGSEGV on the process so that it immediately dies. fatal_sigsegv was added for this purpose. It was declared as noreturn, so in order to pursuade gcc that it actually does not return, I added a call to os_dump_core (and declared it noreturn) so that I get a core file if somehow the process survives. All other calls in arch/um/os-Linux/skas/process.c got the same treatment, with failures causing the process to die instead of a kernel panic, with some exceptions. userspace_tramp exits with status 1 if anything goes wrong there. That will cause start_userspace to return an error. copy_context_skas0 and map_stub_pages also now return errors instead of panicing. Callers of thes functions were changed to check for errors and do something appropriate. Usually that's to return an error to their callers. check_skas3_ptrace_faultinfo just exits since that's too early to do anything else. save_registers, restore_registers, and init_registers now return status instead of panicing on failure, with their callers doing something appropriate. There were also duplicate declarations of save_registers and restore_registers in os.h - these are gone. I noticed and fixed up some whitespace damage. Signed-off-by: Jeff Dike --- arch/um/include/kern_util.h | 1 arch/um/include/os.h | 8 - arch/um/include/registers.h | 6 arch/um/kernel/skas/mmu.c | 5 arch/um/kernel/skas/process.c | 20 ++ arch/um/kernel/trap.c | 12 + arch/um/os-Linux/registers.c | 21 +-- arch/um/os-Linux/skas/process.c | 267 ++++++++++++++++++++++++++-------------- arch/um/os-Linux/start_up.c | 4 arch/um/sys-x86_64/syscalls.c | 8 - 10 files changed, 234 insertions(+), 118 deletions(-) Index: linux-2.6.22/arch/um/include/os.h =================================================================== --- linux-2.6.22.orig/arch/um/include/os.h 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/include/os.h 2007-11-20 13:05:54.000000000 -0500 @@ -238,7 +238,7 @@ extern int raw(int fd); extern void setup_machinename(char *machine_out); extern void setup_hostinfo(char *buf, int len); extern int setjmp_wrapper(void (*proc)(void *, void *), ...); -extern void os_dump_core(void); +extern void os_dump_core(void) __attribute__ ((noreturn)); extern void os_log_info(char *fmt, va_list ap); /* time.c */ @@ -268,11 +268,9 @@ extern int protect(struct mm_id * mm_idp extern int is_skas_winch(int pid, int fd, void *data); extern int start_userspace(unsigned long stub_stack); extern int copy_context_skas0(unsigned long stack, int pid); -extern void save_registers(int pid, struct uml_pt_regs *regs); -extern void restore_registers(int pid, struct uml_pt_regs *regs); extern void userspace(struct uml_pt_regs *regs); -extern void map_stub_pages(int fd, unsigned long code, - unsigned long data, unsigned long stack); +extern int map_stub_pages(int fd, unsigned long code, unsigned long data, + unsigned long stack); extern void new_thread(void *stack, jmp_buf *buf, void (*handler)(void)); extern void switch_threads(jmp_buf *me, jmp_buf *you); extern int start_idle_thread(void *stack, jmp_buf *switch_buf); Index: linux-2.6.22/arch/um/include/registers.h =================================================================== --- linux-2.6.22.orig/arch/um/include/registers.h 2007-11-20 13:05:24.000000000 -0500 +++ linux-2.6.22/arch/um/include/registers.h 2007-11-20 13:05:54.000000000 -0500 @@ -13,9 +13,9 @@ extern int save_fp_registers(int pid, un extern int restore_fp_registers(int pid, unsigned long *fp_regs); extern int save_fpx_registers(int pid, unsigned long *fp_regs); extern int restore_fpx_registers(int pid, unsigned long *fp_regs); -extern void save_registers(int pid, struct uml_pt_regs *regs); -extern void restore_registers(int pid, struct uml_pt_regs *regs); -extern void init_registers(int pid); +extern int save_registers(int pid, struct uml_pt_regs *regs); +extern int restore_registers(int pid, struct uml_pt_regs *regs); +extern int init_registers(int pid); extern void get_safe_registers(unsigned long *regs); extern unsigned long get_thread_reg(int reg, jmp_buf *buf); Index: linux-2.6.22/arch/um/kernel/trap.c =================================================================== --- linux-2.6.22.orig/arch/um/kernel/trap.c 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/kernel/trap.c 2007-11-20 13:05:54.000000000 -0500 @@ -129,6 +129,18 @@ static void bad_segv(struct faultinfo fi force_sig_info(SIGSEGV, &si, current); } +void fatal_sigsegv(void) +{ + force_sigsegv(SIGSEGV, current); + do_signal(); + /* + * This is to tell gcc that we're not returning - do_signal + * can, in general, return, but in this case, it's not, since + * we just got a fatal SIGSEGV queued. + */ + os_dump_core(); +} + void segv_handler(int sig, struct uml_pt_regs *regs) { struct faultinfo * fi = UPT_FAULTINFO(regs); Index: linux-2.6.22/arch/um/os-Linux/registers.c =================================================================== --- linux-2.6.22.orig/arch/um/os-Linux/registers.c 2007-11-20 13:05:24.000000000 -0500 +++ linux-2.6.22/arch/um/os-Linux/registers.c 2007-11-20 13:05:54.000000000 -0500 @@ -8,42 +8,41 @@ #include #include #include "sysdep/ptrace.h" -#include "user.h" -void save_registers(int pid, struct uml_pt_regs *regs) +int save_registers(int pid, struct uml_pt_regs *regs) { int err; err = ptrace(PTRACE_GETREGS, pid, 0, regs->gp); if (err < 0) - panic("save_registers - saving registers failed, errno = %d\n", - errno); + return -errno; + return 0; } -void restore_registers(int pid, struct uml_pt_regs *regs) +int restore_registers(int pid, struct uml_pt_regs *regs) { int err; err = ptrace(PTRACE_SETREGS, pid, 0, regs->gp); if (err < 0) - panic("restore_registers - saving registers failed, " - "errno = %d\n", errno); + return -errno; + return 0; } /* This is set once at boot time and not changed thereafter */ static unsigned long exec_regs[MAX_REG_NR]; -void init_registers(int pid) +int init_registers(int pid) { int err; err = ptrace(PTRACE_GETREGS, pid, 0, exec_regs); - if (err) - panic("check_ptrace : PTRACE_GETREGS failed, errno = %d", - errno); + if (err < 0) + return -errno; arch_init_registers(pid); + return 0; } void get_safe_registers(unsigned long *regs) Index: linux-2.6.22/arch/um/os-Linux/skas/process.c =================================================================== --- linux-2.6.22.orig/arch/um/os-Linux/skas/process.c 2007-11-20 13:05:45.000000000 -0500 +++ linux-2.6.22/arch/um/os-Linux/skas/process.c 2007-11-20 13:09:27.000000000 -0500 @@ -38,17 +38,17 @@ int is_skas_winch(int pid, int fd, void static int ptrace_dump_regs(int pid) { - unsigned long regs[MAX_REG_NR]; - int i; + unsigned long regs[MAX_REG_NR]; + int i; - if (ptrace(PTRACE_GETREGS, pid, 0, regs) < 0) - return -errno; + if (ptrace(PTRACE_GETREGS, pid, 0, regs) < 0) + return -errno; printk(UM_KERN_ERR "Stub registers -\n"); for (i = 0; i < ARRAY_SIZE(regs); i++) printk(UM_KERN_ERR "\t%d - %lx\n", i, regs[i]); - return 0; + return 0; } /* @@ -73,9 +73,11 @@ void wait_stub_done(int pid) break; err = ptrace(PTRACE_CONT, pid, 0, 0); - if (err) - panic("wait_stub_done : continue failed, errno = %d\n", - errno); + if (err) { + printk(UM_KERN_ERR "wait_stub_done : continue failed, " + "errno = %d\n", errno); + fatal_sigsegv(); + } } if (((1 << WSTOPSIG(status)) & STUB_DONE_MASK) != 0) @@ -86,8 +88,10 @@ bad_wait: if (err) printk(UM_KERN_ERR "Failed to get registers from stub, " "errno = %d\n", -err); - panic("wait_stub_done : failed to wait for SIGUSR1/SIGTRAP, pid = %d, " - "n = %d, errno = %d, status = 0x%x\n", pid, n, errno, status); + printk(UM_KERN_ERR "wait_stub_done : failed to wait for SIGTRAP, " + "pid = %d, n = %d, errno = %d, status = 0x%x\n", pid, n, errno, + status); + fatal_sigsegv(); } extern unsigned long current_stub_stack(void); @@ -98,9 +102,11 @@ void get_skas_faultinfo(int pid, struct if (ptrace_faultinfo) { err = ptrace(PTRACE_FAULTINFO, pid, 0, fi); - if (err) - panic("get_skas_faultinfo - PTRACE_FAULTINFO failed, " - "errno = %d\n", errno); + if (err) { + printk(UM_KERN_ERR "get_skas_faultinfo - " + "PTRACE_FAULTINFO failed, errno = %d\n", errno); + fatal_sigsegv(); + } /* Special handling for i386, which has different structs */ if (sizeof(struct ptrace_faultinfo) < sizeof(struct faultinfo)) @@ -110,9 +116,11 @@ void get_skas_faultinfo(int pid, struct } else { err = ptrace(PTRACE_CONT, pid, 0, SIGSEGV); - if (err) - panic("Failed to continue stub, pid = %d, errno = %d\n", - pid, errno); + if (err) { + printk(UM_KERN_ERR "Failed to continue stub, pid = %d, " + "errno = %d\n", pid, errno); + fatal_sigsegv(); + } wait_stub_done(pid); /* @@ -145,25 +153,31 @@ static void handle_trap(int pid, struct { err = ptrace(PTRACE_POKEUSR, pid, PT_SYSCALL_NR_OFFSET, __NR_getpid); - if (err < 0) - panic("handle_trap - nullifying syscall failed, " - "errno = %d\n", errno); + if (err < 0) { + printk(UM_KERN_ERR "handle_trap - nullifying syscall " + "failed, errno = %d\n", errno); + fatal_sigsegv(); + } err = ptrace(PTRACE_SYSCALL, pid, 0, 0); - if (err < 0) - panic("handle_trap - continuing to end of syscall " - "failed, errno = %d\n", errno); + if (err < 0) { + printk(UM_KERN_ERR "handle_trap - continuing to end of " + "syscall failed, errno = %d\n", errno); + fatal_sigsegv(); + } CATCH_EINTR(err = waitpid(pid, &status, WUNTRACED)); if ((err < 0) || !WIFSTOPPED(status) || - (WSTOPSIG(status) != SIGTRAP + 0x80)) { - err = ptrace_dump_regs(pid); - if (err) - printk(UM_KERN_ERR "Failed to get registers " + (WSTOPSIG(status) != SIGTRAP + 0x80)) { + err = ptrace_dump_regs(pid); + if (err) + printk(UM_KERN_ERR "Failed to get registers " "from process, errno = %d\n", -err); - panic("handle_trap - failed to wait at end of syscall, " - "errno = %d, status = %d\n", errno, status); - } + printk(UM_KERN_ERR "handle_trap - failed to wait at " + "end of syscall, errno = %d, status = %d\n", + errno, status); + fatal_sigsegv(); + } } handle_syscall(regs); @@ -181,9 +195,11 @@ static int userspace_tramp(void *stack) signal(SIGTERM, SIG_DFL); signal(SIGWINCH, SIG_IGN); err = set_interval(); - if (err) - panic("userspace_tramp - setting timer failed, errno = %d\n", - err); + if (err) { + printk(UM_KERN_ERR "userspace_tramp - setting timer failed, " + "errno = %d\n", err); + exit(1); + } if (!proc_mm) { /* @@ -226,9 +242,11 @@ static int userspace_tramp(void *stack) sa.sa_flags = SA_ONSTACK | SA_NODEFER; sa.sa_handler = (void *) v; sa.sa_restorer = NULL; - if (sigaction(SIGSEGV, &sa, NULL) < 0) - panic("userspace_tramp - setting SIGSEGV handler " - "failed - errno = %d\n", errno); + if (sigaction(SIGSEGV, &sa, NULL) < 0) { + printk(UM_KERN_ERR "userspace_tramp - setting SIGSEGV " + "handler failed - errno = %d\n", errno); + exit(1); + } } kill(os_getpid(), SIGSTOP); @@ -244,13 +262,18 @@ int start_userspace(unsigned long stub_s { void *stack; unsigned long sp; - int pid, status, n, flags; + int pid, status, n, flags, err; stack = mmap(NULL, UM_KERN_PAGE_SIZE, PROT_READ | PROT_WRITE | PROT_EXEC, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); - if (stack == MAP_FAILED) - panic("start_userspace : mmap failed, errno = %d", errno); + if (stack == MAP_FAILED) { + err = -errno; + printk(UM_KERN_ERR "start_userspace : mmap failed, " + "errno = %d", errno); + return err; + } + sp = (unsigned long) stack + UM_KERN_PAGE_SIZE - sizeof(void *); flags = CLONE_FILES | SIGCHLD; @@ -258,29 +281,50 @@ int start_userspace(unsigned long stub_s flags |= CLONE_VM; pid = clone(userspace_tramp, (void *) sp, flags, (void *) stub_stack); - if (pid < 0) - panic("start_userspace : clone failed, errno = %d", errno); + if (pid < 0) { + err = -errno; + printk(UM_KERN_ERR "start_userspace : clone failed, " + "errno = %d", errno); + return err; + } do { CATCH_EINTR(n = waitpid(pid, &status, WUNTRACED)); - if (n < 0) - panic("start_userspace : wait failed, errno = %d", - errno); + if (n < 0) { + err = -errno; + printk(UM_KERN_ERR "start_userspace : wait failed, " + "errno = %d", errno); + goto out_kill; + } } while (WIFSTOPPED(status) && (WSTOPSIG(status) == SIGVTALRM)); - if (!WIFSTOPPED(status) || (WSTOPSIG(status) != SIGSTOP)) - panic("start_userspace : expected SIGSTOP, got status = %d", - status); + if (!WIFSTOPPED(status) || (WSTOPSIG(status) != SIGSTOP)) { + err = -EINVAL; + printk(UM_KERN_ERR "start_userspace : expected SIGSTOP, got " + "status = %d", status); + goto out_kill; + } if (ptrace(PTRACE_OLDSETOPTIONS, pid, NULL, - (void *) PTRACE_O_TRACESYSGOOD) < 0) - panic("start_userspace : PTRACE_OLDSETOPTIONS failed, " - "errno = %d\n", errno); + (void *) PTRACE_O_TRACESYSGOOD) < 0) { + err = -errno; + printk(UM_KERN_ERR "start_userspace : PTRACE_OLDSETOPTIONS " + "failed, errno = %d\n", errno); + goto out_kill; + } - if (munmap(stack, UM_KERN_PAGE_SIZE) < 0) - panic("start_userspace : munmap failed, errno = %d\n", errno); + if (munmap(stack, UM_KERN_PAGE_SIZE) < 0) { + err = -errno; + printk(UM_KERN_ERR "start_userspace : munmap failed, " + "errno = %d\n", errno); + goto out_kill; + } return pid; + + out_kill: + os_kill_ptraced_process(pid, 1); + return err; } void userspace(struct uml_pt_regs *regs) @@ -298,9 +342,16 @@ void userspace(struct uml_pt_regs *regs) nsecs += os_nsecs(); while (1) { + /* + * This can legitimately fail if the process loads a + * bogus value into a segment register. It will + * segfault and PTRACE_GETREGS will read that value + * out of the process. However, PTRACE_SETREGS will + * fail. In this case, there is nothing to do but + * just kill the process. + */ if (ptrace(PTRACE_SETREGS, pid, 0, regs->gp)) - panic("userspace - PTRACE_SETREGS failed, " - "errno = %d\n", errno); + fatal_sigsegv(); /* Now we set local_using_sysemu to be used for one loop */ local_using_sysemu = get_using_sysemu(); @@ -308,21 +359,25 @@ void userspace(struct uml_pt_regs *regs) op = SELECT_PTRACE_OPERATION(local_using_sysemu, singlestepping(NULL)); - err = ptrace(op, pid, 0, 0); - if (err) - panic("userspace - could not resume userspace process, " - "pid=%d, ptrace operation = %d, errno = %d\n", - pid, op, errno); + if (ptrace(op, pid, 0, 0)) { + printk(UM_KERN_ERR "userspace - ptrace continue " + "failed, op = %d, errno = %d\n", op, errno); + fatal_sigsegv(); + } CATCH_EINTR(err = waitpid(pid, &status, WUNTRACED)); - if (err < 0) - panic("userspace - waitpid failed, errno = %d\n", - errno); + if (err < 0) { + printk(UM_KERN_ERR "userspace - wait failed, " + "errno = %d\n", errno); + fatal_sigsegv(); + } regs->is_user = 1; - if (ptrace(PTRACE_GETREGS, pid, 0, regs->gp)) - panic("userspace - saving registers failed, " - "errno = %d\n", errno); + if (ptrace(PTRACE_GETREGS, pid, 0, regs->gp)) { + printk(UM_KERN_ERR "userspace - PTRACE_GETREGS failed, " + "errno = %d\n", errno); + fatal_sigsegv(); + } UPT_SYSCALL_NR(regs) = -1; /* Assume: It's not a syscall */ @@ -369,6 +424,7 @@ void userspace(struct uml_pt_regs *regs) default: printk(UM_KERN_ERR "userspace - child stopped " "with signal %d\n", sig); + fatal_sigsegv(); } pid = userspace_pid[0]; interrupt_end(); @@ -420,9 +476,12 @@ int copy_context_skas0(unsigned long new .it_interval = tv }) }); err = ptrace_setregs(pid, thread_regs); - if (err < 0) - panic("copy_context_skas0 : PTRACE_SETREGS failed, " - "pid = %d, errno = %d\n", pid, -err); + if (err < 0) { + err = -errno; + printk(UM_KERN_ERR "copy_context_skas0 : PTRACE_SETREGS " + "failed, pid = %d, errno = %d\n", pid, -err); + return err; + } /* set a well known return code for detection of child write failure */ child_data->err = 12345678; @@ -432,31 +491,47 @@ int copy_context_skas0(unsigned long new * parent's stack, and check, if bad result. */ err = ptrace(PTRACE_CONT, pid, 0, 0); - if (err) - panic("Failed to continue new process, pid = %d, " - "errno = %d\n", pid, errno); + if (err) { + err = -errno; + printk(UM_KERN_ERR "Failed to continue new process, pid = %d, " + "errno = %d\n", pid, errno); + return err; + } + wait_stub_done(pid); pid = data->err; - if (pid < 0) - panic("copy_context_skas0 - stub-parent reports error %d\n", - -pid); + if (pid < 0) { + printk(UM_KERN_ERR "copy_context_skas0 - stub-parent reports " + "error %d\n", -pid); + return pid; + } /* * Wait, until child has finished too: read child's result from * child's stack and check it. */ wait_stub_done(pid); - if (child_data->err != STUB_DATA) - panic("copy_context_skas0 - stub-child reports error %ld\n", - child_data->err); + if (child_data->err != STUB_DATA) { + printk(UM_KERN_ERR "copy_context_skas0 - stub-child reports " + "error %ld\n", child_data->err); + err = child_data->err; + goto out_kill; + } if (ptrace(PTRACE_OLDSETOPTIONS, pid, NULL, - (void *)PTRACE_O_TRACESYSGOOD) < 0) - panic("copy_context_skas0 : PTRACE_OLDSETOPTIONS failed, " - "errno = %d\n", errno); + (void *)PTRACE_O_TRACESYSGOOD) < 0) { + err = -errno; + printk(UM_KERN_ERR "copy_context_skas0 : PTRACE_OLDSETOPTIONS " + "failed, errno = %d\n", errno); + goto out_kill; + } return pid; + + out_kill: + os_kill_ptraced_process(pid, 1); + return err; } /* @@ -464,8 +539,8 @@ int copy_context_skas0(unsigned long new * available. Opening /proc/mm creates a new mm_context, which lacks * the stub-pages. Thus, we map them using /proc/mm-fd */ -void map_stub_pages(int fd, unsigned long code, - unsigned long data, unsigned long stack) +int map_stub_pages(int fd, unsigned long code, unsigned long data, + unsigned long stack) { struct proc_mm_op mmop; int n; @@ -489,8 +564,9 @@ void map_stub_pages(int fd, unsigned lon printk(UM_KERN_ERR "mmap args - addr = 0x%lx, fd = %d, " "offset = %llx\n", code, code_fd, (unsigned long long) code_offset); - panic("map_stub_pages : /proc/mm map for code failed, " - "err = %d\n", n); + printk(UM_KERN_ERR "map_stub_pages : /proc/mm map for code " + "failed, err = %d\n", n); + return -n; } if (stack) { @@ -508,10 +584,15 @@ void map_stub_pages(int fd, unsigned lon .offset = map_offset } } }); CATCH_EINTR(n = write(fd, &mmop, sizeof(mmop))); - if (n != sizeof(mmop)) - panic("map_stub_pages : /proc/mm map for data failed, " - "err = %d\n", errno); + if (n != sizeof(mmop)) { + n = errno; + printk(UM_KERN_ERR "map_stub_pages : /proc/mm map for " + "data failed, err = %d\n", n); + return -n; + } } + + return 0; } void new_thread(void *stack, jmp_buf *buf, void (*handler)(void)) @@ -572,7 +653,9 @@ int start_idle_thread(void *stack, jmp_b kmalloc_ok = 0; return 1; default: - panic("Bad sigsetjmp return in start_idle_thread - %d\n", n); + printk(UM_KERN_ERR "Bad sigsetjmp return in " + "start_idle_thread - %d\n", n); + fatal_sigsegv(); } longjmp(*switch_buf, 1); } @@ -615,9 +698,11 @@ void __switch_mm(struct mm_id *mm_idp) if (proc_mm) { err = ptrace(PTRACE_SWITCH_MM, userspace_pid[0], 0, mm_idp->u.mm_fd); - if (err) - panic("__switch_mm - PTRACE_SWITCH_MM failed, " - "errno = %d\n", errno); + if (err) { + printk(UM_KERN_ERR "__switch_mm - PTRACE_SWITCH_MM " + "failed, errno = %d\n", errno); + fatal_sigsegv(); + } } else userspace_pid[0] = mm_idp->u.pid; } Index: linux-2.6.22/arch/um/include/kern_util.h =================================================================== --- linux-2.6.22.orig/arch/um/include/kern_util.h 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/include/kern_util.h 2007-11-20 13:05:54.000000000 -0500 @@ -62,6 +62,7 @@ extern int singlestepping(void *t); extern void segv_handler(int sig, struct uml_pt_regs *regs); extern void bus_handler(int sig, struct uml_pt_regs *regs); extern void winch(int sig, struct uml_pt_regs *regs); +extern void fatal_sigsegv(void) __attribute__ ((noreturn)); extern void log_info(char *fmt, ...) __attribute__ ((format (printf, 1, 2))); extern void log_info_nolock(char *fmt, ...) Index: linux-2.6.22/arch/um/os-Linux/start_up.c =================================================================== --- linux-2.6.22.orig/arch/um/os-Linux/start_up.c 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/os-Linux/start_up.c 2007-11-20 13:05:54.000000000 -0500 @@ -411,7 +411,9 @@ static inline void check_skas3_ptrace_fa non_fatal("found\n"); } - init_registers(pid); + if (init_registers(pid)) + fatal("Failed to initialize default registers"); + stop_ptraced_child(pid, 1, 1); } Index: linux-2.6.22/arch/um/sys-x86_64/syscalls.c =================================================================== --- linux-2.6.22.orig/arch/um/sys-x86_64/syscalls.c 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/sys-x86_64/syscalls.c 2007-11-20 13:05:54.000000000 -0500 @@ -48,7 +48,9 @@ long arch_prctl(struct task_struct *task switch (code) { case ARCH_SET_FS: case ARCH_SET_GS: - restore_registers(pid, ¤t->thread.regs.regs); + ret = restore_registers(pid, ¤t->thread.regs.regs); + if (ret) + return ret; break; case ARCH_GET_FS: case ARCH_GET_GS: @@ -70,10 +72,10 @@ long arch_prctl(struct task_struct *task switch (code) { case ARCH_SET_FS: current->thread.arch.fs = (unsigned long) ptr; - save_registers(pid, ¤t->thread.regs.regs); + ret = save_registers(pid, ¤t->thread.regs.regs); break; case ARCH_SET_GS: - save_registers(pid, ¤t->thread.regs.regs); + ret = save_registers(pid, ¤t->thread.regs.regs); break; case ARCH_GET_FS: ret = put_user(tmp, addr); Index: linux-2.6.22/arch/um/kernel/skas/mmu.c =================================================================== --- linux-2.6.22.orig/arch/um/kernel/skas/mmu.c 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/kernel/skas/mmu.c 2007-11-20 13:05:54.000000000 -0500 @@ -86,6 +86,11 @@ int init_new_context(struct task_struct to_mm->id.u.pid = copy_context_skas0(stack, from_mm->id.u.pid); else to_mm->id.u.pid = start_userspace(stack); + + if (to_mm->id.u.pid < 0) { + ret = to_mm->id.u.pid; + goto out_free; + } } ret = init_new_ldt(to_mm, from_mm); Index: linux-2.6.22/arch/um/kernel/skas/process.c =================================================================== --- linux-2.6.22.orig/arch/um/kernel/skas/process.c 2007-11-20 13:05:19.000000000 -0500 +++ linux-2.6.22/arch/um/kernel/skas/process.c 2007-11-20 13:05:54.000000000 -0500 @@ -6,19 +6,25 @@ #include "linux/init.h" #include "linux/sched.h" #include "as-layout.h" +#include "kern.h" #include "os.h" #include "skas.h" int new_mm(unsigned long stack) { - int fd; + int fd, err; fd = os_open_file("/proc/mm", of_cloexec(of_write(OPENFLAGS())), 0); if (fd < 0) return fd; - if (skas_needs_stub) - map_stub_pages(fd, STUB_CODE, STUB_DATA, stack); + if (skas_needs_stub) { + err = map_stub_pages(fd, STUB_CODE, STUB_DATA, stack); + if (err) { + os_close_file(fd); + return err; + } + } return fd; } @@ -49,8 +55,14 @@ int __init start_uml(void) { stack_protections((unsigned long) &cpu0_irqstack); set_sigstack(cpu0_irqstack, THREAD_SIZE); - if (proc_mm) + if (proc_mm) { userspace_pid[0] = start_userspace(0); + if (userspace_pid[0] < 0) { + printf("start_uml - start_userspace returned %d\n", + userspace_pid[0]); + exit(1); + } + } init_new_thread_signals(); - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/