* [uml-devel] [RFC PATCH 0/9] SKAS4
@ 2008-01-09 22:09 Jeff Dike
2008-01-10 2:48 ` Jeff Dike
0 siblings, 1 reply; 2+ messages in thread
From: Jeff Dike @ 2008-01-09 22:09 UTC (permalink / raw)
To: uml-devel
These patches implement host and guest SKAS4 support for both 32- and
64-bit x86.
I think the new interfaces here are much more acceptable, so I'm going
to push this to mainline.
What's new:
two new system calls -
new_mm - creates a new address space and returns a file
descriptor referencing it
switch_mm - moves the calling process to the address space
referenced by the file descriptor passed to it
/proc/<pid>/mm - opening this gives you a file descriptor
referencing the address space occupied by the given pid - this
descriptor can be given to switch_mm just as a new_mm descriptor can
siginfo_t extensions - siginfo_t always contained the faulting
address in the SIGSEGV case. However, for it to be useful to UML, it
needs the CPU trap number and error code. There already was
conditional support for the trap number, which x86 didn't implement.
This is enabled, and support for passing out the error code is added.
How this compares to skas3:
Creating a new address space:
skas3 - open /proc/mm
skas4 - new_mm()
Remapping pages within another address space
skas3 - write a structure desribing the change to a /proc/mm
descriptor
skas4 - switch_mm to the address space, run the necessary system
calls directly, switch_mm back to the UML kernel address space
Getting page fault information from a process
skas3 - PTRACE_FAULTINFO
skas4 - PTRACE_GETSIGINFO
This patchset contains 9 patches. Four of them contain the siginfo
extension, guest siginfo support, host new_mm/switch_mm support, and
guest new_mm/switch_mm support. The rest are preparation patches
which shouldn't change anything functionally, but which make the four
functional patches smaller and easier to read.
These are against 2.6.24-rc7. Build both host and guest from the same
tree. I will be tweaking the interfaces in incompatible ways, so if
you play with future versions of this patchset, throw this one out.
Definitely don't try to boot a guest built from one version on a host
built from another. It will work, but likely you'll get skas0, as
there will likely be some interface change which causes the host skas4
checks to fail.
This is very experimental at this point. Don't let it near anything
resembling a production system.
Jeff
--
Work email - jdike at linux dot intel dot com
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [uml-devel] [RFC PATCH 0/9] SKAS4
2008-01-09 22:09 [uml-devel] [RFC PATCH 0/9] SKAS4 Jeff Dike
@ 2008-01-10 2:48 ` Jeff Dike
0 siblings, 0 replies; 2+ messages in thread
From: Jeff Dike @ 2008-01-10 2:48 UTC (permalink / raw)
To: uml-devel
On Wed, Jan 09, 2008 at 05:09:49PM -0500, Jeff Dike wrote:
> These patches implement host and guest SKAS4 support for both 32- and
> 64-bit x86.
I forgot to include a single rolled-up patch, so it's below - again,
against 2.6.24-rc7.
Jeff
--
Work email - jdike at linux dot intel dot com
diff --git a/arch/um/include/as-layout.h b/arch/um/include/as-layout.h
index a5cdf95..90ee798 100644
--- a/arch/um/include/as-layout.h
+++ b/arch/um/include/as-layout.h
@@ -17,6 +17,7 @@
#define ASM_STUB_CODE (UML_CONFIG_TOP_ADDR - 2 * UM_KERN_PAGE_SIZE)
#define ASM_STUB_DATA (UML_CONFIG_TOP_ADDR - UM_KERN_PAGE_SIZE)
#define ASM_STUB_START ASM_STUB_CODE
+#define ASM_STUB_END UML_CONFIG_TOP_ADDR
/*
* This file is included by the assembly stubs, which just want the
@@ -27,6 +28,7 @@
#define STUB_CODE ((unsigned long) ASM_STUB_CODE)
#define STUB_DATA ((unsigned long) ASM_STUB_DATA)
#define STUB_START ((unsigned long) ASM_STUB_START)
+#define STUB_END ((unsigned long) ASM_STUB_END)
#include "sysdep/ptrace.h"
diff --git a/arch/um/include/os.h b/arch/um/include/os.h
index 6f0d1c7..f1c26f5 100644
--- a/arch/um/include/os.h
+++ b/arch/um/include/os.h
@@ -168,7 +168,7 @@ extern int os_fchange_dir(int fd);
/* start_up.c */
extern void os_early_checks(void);
-extern int can_do_skas(void);
+extern void can_do_skas(void);
extern void os_check_bugs(void);
extern void check_host_supports_tls(int *supports_tls, int *tls_min);
diff --git a/arch/um/include/skas/mm_id.h b/arch/um/include/skas/mm_id.h
index 48dd098..a2e7643 100644
--- a/arch/um/include/skas/mm_id.h
+++ b/arch/um/include/skas/mm_id.h
@@ -7,7 +7,7 @@
#define __MM_ID_H
struct mm_id {
- union {
+ struct {
int mm_fd;
int pid;
} u;
diff --git a/arch/um/include/skas/skas.h b/arch/um/include/skas/skas.h
index b073f8a..5953647 100644
--- a/arch/um/include/skas/skas.h
+++ b/arch/um/include/skas/skas.h
@@ -6,18 +6,61 @@
#ifndef __SKAS_H
#define __SKAS_H
+#ifndef __KERNEL__
+#include <unistd.h>
+#include <sys/syscall.h>
+#endif
+#include "uml-config.h"
+
+#ifdef UML_CONFIG_X86_32
+#define __NR_new_mm 325
+#define __NR_switch_mm 326
+#else
+#define __NR_new_mm 286
+#define __NR_switch_mm 287
+#endif
+
+#define MM_COPY 0
+
+#define MM_ALL_REGS 0
+#define MM_SP_IP 1
+#define MM_SAME 2
+
+#define PTRACE_SWITCH_MM 33
+
+#ifndef __ASSEMBLY__
+
#include "sysdep/ptrace.h"
extern int userspace_pid[];
extern int proc_mm, ptrace_faultinfo, ptrace_ldt;
extern int skas_needs_stub;
+extern int have_switch_mm;
+extern int self_mm_fd;
+extern int have_ptrace_switch_mm;
+
extern int user_thread(unsigned long stack, int flags);
extern void new_thread_handler(void);
extern void handle_syscall(struct uml_pt_regs *regs);
-extern int new_mm(unsigned long stack);
+extern int make_new_mm(unsigned long stack);
extern void get_skas_faultinfo(int pid, struct faultinfo * fi);
extern long execute_syscall_skas(void *r);
extern unsigned long current_stub_stack(void);
+#ifndef __KERNEL__
+static inline long new_mm(int flags)
+{
+ return syscall(__NR_new_mm, MM_COPY, 0, 0, 0, 0, 0, 0);
+}
+
+static inline long switch_mm(int mm_fd, unsigned long flags,
+ unsigned long *new_regs, unsigned long *save_regs)
+{
+ return syscall(__NR_switch_mm, mm_fd, flags, new_regs, save_regs, 0, 0);
+}
+#endif
+
+#endif
+
#endif
diff --git a/arch/um/include/skas_ptrace.h b/arch/um/include/skas_ptrace.h
index cd2327d..6b55c52 100644
--- a/arch/um/include/skas_ptrace.h
+++ b/arch/um/include/skas_ptrace.h
@@ -7,7 +7,9 @@
#define __SKAS_PTRACE_H
#define PTRACE_FAULTINFO 52
-#define PTRACE_SWITCH_MM 55
+#ifndef OLD_PTRACE_SWITCH_MM
+#define OLD_PTRACE_SWITCH_MM 55
+#endif
#include "sysdep/skas_ptrace.h"
diff --git a/arch/um/kernel/ptrace.c b/arch/um/kernel/ptrace.c
index 47b57b4..36f8ae5 100644
--- a/arch/um/kernel/ptrace.c
+++ b/arch/um/kernel/ptrace.c
@@ -192,7 +192,7 @@ long arch_ptrace(struct task_struct *child, long request, long addr, long data)
}
#endif
#ifdef CONFIG_PROC_MM
- case PTRACE_SWITCH_MM: {
+ case OLD_PTRACE_SWITCH_MM: {
struct mm_struct *old = child->mm;
struct mm_struct *new = proc_mm_get_mm(data);
diff --git a/arch/um/kernel/reboot.c b/arch/um/kernel/reboot.c
index 04cebcf..0a5468e 100644
--- a/arch/um/kernel/reboot.c
+++ b/arch/um/kernel/reboot.c
@@ -11,7 +11,7 @@ void (*pm_power_off)(void);
static void kill_off_processes(void)
{
- if(proc_mm)
+ if(proc_mm || have_switch_mm)
/*
* FIXME: need to loop over userspace_pids
*/
diff --git a/arch/um/kernel/skas/mmu.c b/arch/um/kernel/skas/mmu.c
index f859ec3..3155263 100644
--- a/arch/um/kernel/skas/mmu.c
+++ b/arch/um/kernel/skas/mmu.c
@@ -65,6 +65,9 @@ static int init_stub_pte(struct mm_struct *mm, unsigned long proc,
return -ENOMEM;
}
+extern int copy_context_skas4(struct mm_id *id);
+extern int get_new_mm(void);
+
int init_new_context(struct task_struct *task, struct mm_struct *mm)
{
struct mm_context *from_mm = NULL;
@@ -101,7 +104,7 @@ int init_new_context(struct task_struct *task, struct mm_struct *mm)
from_mm = ¤t->mm->context;
if (proc_mm) {
- ret = new_mm(stack);
+ ret = make_new_mm(stack);
if (ret < 0) {
printk(KERN_ERR "init_new_context_skas - "
"new_mm failed, errno = %d\n", ret);
@@ -109,6 +112,20 @@ int init_new_context(struct task_struct *task, struct mm_struct *mm)
}
to_mm->id.u.mm_fd = ret;
}
+ else if (have_switch_mm) {
+ to_mm->id.u.mm_fd = get_new_mm();
+ if(to_mm->id.u.mm_fd < 0) {
+ ret = to_mm->id.u.mm_fd;
+ goto out_free;
+ }
+
+ ret = copy_context_skas4(&to_mm->id);
+ if (ret < 0) {
+ os_close_file(to_mm->id.u.mm_fd);
+ to_mm->id.u.mm_fd = -1;
+ goto out_free;
+ }
+ }
else {
if (from_mm)
to_mm->id.u.pid = copy_context_skas0(stack,
@@ -136,11 +153,15 @@ void destroy_context(struct mm_struct *mm)
{
struct mm_context *mmu = &mm->context;
- if (proc_mm)
+ if (proc_mm || have_switch_mm)
os_close_file(mmu->id.u.mm_fd);
- else
+ else {
os_kill_ptraced_process(mmu->id.u.pid, 1);
+ if (have_switch_mm)
+ os_close_file(mmu->id.u.mm_fd);
+ }
+
if (!proc_mm || !ptrace_faultinfo) {
free_page(mmu->id.stack);
pte_lock_deinit(virt_to_page(mmu->last_page_table));
diff --git a/arch/um/kernel/skas/process.c b/arch/um/kernel/skas/process.c
index fce389c..e5e8613 100644
--- a/arch/um/kernel/skas/process.c
+++ b/arch/um/kernel/skas/process.c
@@ -9,7 +9,7 @@
#include "os.h"
#include "skas.h"
-int new_mm(unsigned long stack)
+int make_new_mm(unsigned long stack)
{
int fd;
@@ -49,7 +49,7 @@ int __init start_uml(void)
{
stack_protections((unsigned long) &cpu0_irqstack);
set_sigstack(cpu0_irqstack, THREAD_SIZE);
- if (proc_mm)
+ if (proc_mm || have_switch_mm)
userspace_pid[0] = start_userspace(0);
init_new_thread_signals();
diff --git a/arch/um/kernel/um_arch.c b/arch/um/kernel/um_arch.c
index f1c7139..d92108b 100644
--- a/arch/um/kernel/um_arch.c
+++ b/arch/um/kernel/um_arch.c
@@ -271,7 +271,9 @@ int __init linux_main(int argc, char **argv)
can_do_skas();
- if (proc_mm && ptrace_faultinfo)
+ if (have_switch_mm)
+ mode = "SKAS4";
+ else if (proc_mm && ptrace_faultinfo)
mode = "SKAS3";
else
mode = "SKAS0";
diff --git a/arch/um/os-Linux/skas/mem.c b/arch/um/os-Linux/skas/mem.c
index 484e68f..cc86c0b 100644
--- a/arch/um/os-Linux/skas/mem.c
+++ b/arch/um/os-Linux/skas/mem.c
@@ -6,6 +6,7 @@
#include <stddef.h>
#include <unistd.h>
#include <errno.h>
+#include <signal.h>
#include <string.h>
#include <sys/mman.h>
#include "init.h"
@@ -22,7 +23,7 @@
#include "sysdep/stub.h"
#include "uml-config.h"
-extern unsigned long batch_syscall_stub, __syscall_stub_start;
+extern unsigned long batch_syscall_stub, switch_mm_stub, __syscall_stub_start;
extern void wait_stub_done(int pid);
@@ -40,35 +41,72 @@ static unsigned long syscall_regs[MAX_REG_NR];
static int __init init_syscall_regs(void)
{
+ unsigned long *stub_entry;
+
get_safe_registers(syscall_regs);
+ if (have_switch_mm)
+ stub_entry = &switch_mm_stub;
+ else
+ stub_entry = &batch_syscall_stub;
+
syscall_regs[REGS_IP_INDEX] = STUB_CODE +
- ((unsigned long) &batch_syscall_stub -
+ ((unsigned long) stub_entry -
(unsigned long) &__syscall_stub_start);
return 0;
}
__initcall(init_syscall_regs);
-extern int proc_mm;
+static int syscall_stub_done(unsigned long stack)
+{
+ unsigned long *syscall, *data, offset;
+ int ret, n;
+
+ /*
+ * When the stub stops, we find the following values on the
+ * beginning of the stack:
+ * (long) return_value
+ * (long) offset to failed sycall data (0 if no error)
+ */
+ ret = *((unsigned long *) stack);
+ offset = *((unsigned long *) stack + 1);
+ if (offset == 0)
+ return 0;
+
+ data = (unsigned long *)(stack + offset - STUB_DATA);
+ printk(UM_KERN_ERR "syscall_stub_done : ret = %d, offset = %ld, "
+ "data = %p\n", ret, offset, data);
+ syscall = (unsigned long *)((unsigned long)data + data[0]);
+ printk(UM_KERN_ERR "syscall_stub_done : syscall %ld failed, "
+ "return value = 0x%x, expected return value = 0x%lx\n",
+ syscall[0], ret, syscall[7]);
+ printk(UM_KERN_ERR " syscall parameters: "
+ "0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx\n",
+ syscall[1], syscall[2], syscall[3],
+ syscall[4], syscall[5], syscall[6]);
+ for (n = 1; n < data[0]/sizeof(long); n++) {
+ if (n == 1)
+ printk(UM_KERN_ERR " additional syscall "
+ "data:");
+ if (n % 4 == 1)
+ printk("\n" UM_KERN_ERR " ");
+ printk(" 0x%lx", data[n]);
+ }
+ if (n > 1)
+ printk("\n");
-int single_count = 0;
-int multi_count = 0;
-int multi_op_count = 0;
+ return ret;
+}
-static inline long do_syscall_stub(struct mm_id * mm_idp, void **addr)
+static long do_syscall_stub(struct mm_id *mm_idp, void **addr)
{
- int n, i;
- long ret, offset;
- unsigned long * data;
- unsigned long * syscall;
- int err, pid = mm_idp->u.pid;
+ long ret;
+ int n, i, err, pid = mm_idp->u.pid;
if (proc_mm)
/* FIXME: Need to look up userspace_pid by cpu */
pid = userspace_pid[0];
- multi_count++;
-
n = ptrace_setregs(pid, syscall_regs);
if (n < 0) {
printk(UM_KERN_ERR "Registers - \n");
@@ -85,52 +123,71 @@ static inline long do_syscall_stub(struct mm_id * mm_idp, void **addr)
wait_stub_done(pid);
- /*
- * When the stub stops, we find the following values on the
- * beginning of the stack:
- * (long )return_value
- * (long )offset to failed sycall-data (0, if no error)
- */
- ret = *((unsigned long *) mm_idp->stack);
- offset = *((unsigned long *) mm_idp->stack + 1);
- if (offset) {
- data = (unsigned long *)(mm_idp->stack + offset - STUB_DATA);
- printk(UM_KERN_ERR "do_syscall_stub : ret = %ld, offset = %ld, "
- "data = %p\n", ret, offset, data);
- syscall = (unsigned long *)((unsigned long)data + data[0]);
- printk(UM_KERN_ERR "do_syscall_stub: syscall %ld failed, "
- "return value = 0x%lx, expected return value = 0x%lx\n",
- syscall[0], ret, syscall[7]);
- printk(UM_KERN_ERR " syscall parameters: "
- "0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx\n",
- syscall[1], syscall[2], syscall[3],
- syscall[4], syscall[5], syscall[6]);
- for (n = 1; n < data[0]/sizeof(long); n++) {
- if (n == 1)
- printk(UM_KERN_ERR " additional syscall "
- "data:");
- if (n % 4 == 1)
- printk("\n" UM_KERN_ERR " ");
- printk(" 0x%lx", data[n]);
- }
- if (n > 1)
- printk("\n");
- }
- else ret = 0;
+ ret = syscall_stub_done(mm_idp->stack);
*addr = check_init_stack(mm_idp, NULL);
return ret;
}
-long run_syscall_stub(struct mm_id * mm_idp, int syscall,
+static long do_syscall_stub_skas4(struct mm_id *mm_idp, void **addr)
+{
+ long ret;
+ unsigned long *return_regs;
+ int err;
+ sigset_t sigs, old;
+
+ return_regs = (unsigned long *) (mm_idp->stack + UM_KERN_PAGE_SIZE) -
+ MAX_REG_NR;
+ *(return_regs - 1) = self_mm_fd;
+
+ return_regs = (unsigned long *) (STUB_DATA + UM_KERN_PAGE_SIZE) -
+ MAX_REG_NR;
+
+ sigfillset(&sigs);
+ sigprocmask(SIG_SETMASK, &sigs, &old);
+ err = switch_mm(mm_idp->u.mm_fd, MM_SP_IP, syscall_regs, return_regs);
+ sigprocmask(SIG_SETMASK, &old, NULL);
+
+ ret = syscall_stub_done(mm_idp->stack);
+
+ *addr = check_init_stack(mm_idp, NULL);
+
+ return ret;
+}
+
+static int flush_syscalls(struct mm_id *mm_idp, void **addr, int extra)
+{
+ unsigned long *stack = check_init_stack(mm_idp, *addr);
+ int current, end;
+
+ current = ((unsigned long) stack) & ~UM_KERN_PAGE_MASK;
+ end = UM_KERN_PAGE_SIZE;
+
+ if(have_switch_mm)
+ end -= (MAX_REG_NR + 1) * sizeof(long);
+
+ if (current + (10 + extra) * sizeof(long) < end)
+ return 0;
+
+ if (have_switch_mm)
+ return do_syscall_stub_skas4(mm_idp, addr);
+ else
+ return do_syscall_stub(mm_idp, addr);
+}
+
+long run_syscall_stub(struct mm_id *mm_idp, int syscall,
unsigned long *args, long expected, void **addr,
int done)
{
- unsigned long *stack = check_init_stack(mm_idp, *addr);
+ unsigned long *stack;
+ int ret;
+
+ ret = flush_syscalls(mm_idp, addr, 0);
+ if (ret)
+ return ret;
- if (done && *addr == NULL)
- single_count++;
+ stack = check_init_stack(mm_idp, *addr);
*stack += sizeof(long);
stack += *stack / sizeof(long);
@@ -144,45 +201,36 @@ long run_syscall_stub(struct mm_id * mm_idp, int syscall,
*stack++ = args[5];
*stack++ = expected;
*stack = 0;
- multi_op_count++;
- if (!done && ((((unsigned long) stack) & ~UM_KERN_PAGE_MASK) <
- UM_KERN_PAGE_SIZE - 10 * sizeof(long))) {
- *addr = stack;
- return 0;
+ if (done) {
+ if (have_switch_mm)
+ return do_syscall_stub_skas4(mm_idp, addr);
+ else
+ return do_syscall_stub(mm_idp, addr);
}
- return do_syscall_stub(mm_idp, addr);
+ *addr = stack;
+
+ return 0;
}
-long syscall_stub_data(struct mm_id * mm_idp,
- unsigned long *data, int data_count,
- void **addr, void **stub_addr)
+long syscall_stub_data(struct mm_id *mm_idp, unsigned long *data,
+ int data_count, void **addr, void **stub_addr)
{
unsigned long *stack;
- int ret = 0;
+ int ret;
- /*
- * If *addr still is uninitialized, it *must* contain NULL.
- * Thus in this case do_syscall_stub correctly won't be called.
- */
- if ((((unsigned long) *addr) & ~UM_KERN_PAGE_MASK) >=
- UM_KERN_PAGE_SIZE - (10 + data_count) * sizeof(long)) {
- ret = do_syscall_stub(mm_idp, addr);
- /* in case of error, don't overwrite data on stack */
- if (ret)
- return ret;
- }
+ ret = flush_syscalls(mm_idp, addr, data_count);
+ if (ret)
+ return ret;
stack = check_init_stack(mm_idp, *addr);
- *addr = stack;
-
- *stack = data_count * sizeof(long);
+ *stack++ = data_count * sizeof(long);
- memcpy(stack + 1, data, data_count * sizeof(long));
+ memcpy(stack, data, data_count * sizeof(long));
- *stub_addr = (void *)(((unsigned long)(stack + 1) &
- ~UM_KERN_PAGE_MASK) + STUB_DATA);
+ *stub_addr = (void *)(((unsigned long) stack & ~UM_KERN_PAGE_MASK) +
+ STUB_DATA);
return 0;
}
diff --git a/arch/um/os-Linux/skas/process.c b/arch/um/os-Linux/skas/process.c
index e8b7a97..725af32 100644
--- a/arch/um/os-Linux/skas/process.c
+++ b/arch/um/os-Linux/skas/process.c
@@ -3,6 +3,9 @@
* Licensed under the GPL
*/
+#define siginfo old_siginfo
+#define siginfo_t old_siginfo_t
+
#include <stdlib.h>
#include <unistd.h>
#include <sched.h>
@@ -26,6 +29,108 @@
#include "user.h"
#include "sysdep/stub.h"
+#undef siginfo
+#undef siginfo_t
+
+#define __ARCH_SI_TRAPNO
+#define __ARCH_SI_ERROR
+
+/*
+ * This is the size (including padding) of the part of the
+ * struct siginfo that is before the union.
+ */
+#ifndef __ARCH_SI_PREAMBLE_SIZE
+#define __ARCH_SI_PREAMBLE_SIZE (3 * sizeof(int))
+#endif
+
+#define SI_MAX_SIZE 128
+#ifndef SI_PAD_SIZE
+#define SI_PAD_SIZE ((SI_MAX_SIZE - __ARCH_SI_PREAMBLE_SIZE) / sizeof(int))
+#endif
+
+#ifndef __ARCH_SI_UID_T
+#define __ARCH_SI_UID_T uid_t
+#endif
+
+/*
+ * The default "si_band" type is "long", as specified by POSIX.
+ * However, some architectures want to override this to "int"
+ * for historical compatibility reasons, so we allow that.
+ */
+#ifndef __ARCH_SI_BAND_T
+#define __ARCH_SI_BAND_T long
+#endif
+
+#define __user
+
+typedef struct siginfo {
+ int si_signo;
+ int si_errno;
+ int si_code;
+
+ union {
+ int _pad[SI_PAD_SIZE];
+
+ /* kill() */
+ struct {
+ pid_t _pid; /* sender's pid */
+ __ARCH_SI_UID_T _uid; /* sender's uid */
+ } _kill;
+
+ /* POSIX.1b timers */
+ struct {
+ timer_t _tid; /* timer id */
+ int _overrun; /* overrun count */
+ char _pad[sizeof( __ARCH_SI_UID_T) - sizeof(int)];
+ sigval_t _sigval; /* same as below */
+ int _sys_private; /* not to be passed to user */
+ } _timer;
+
+ /* POSIX.1b signals */
+ struct {
+ pid_t _pid; /* sender's pid */
+ __ARCH_SI_UID_T _uid; /* sender's uid */
+ sigval_t _sigval;
+ } _rt;
+
+ /* SIGCHLD */
+ struct {
+ pid_t _pid; /* which child */
+ __ARCH_SI_UID_T _uid; /* sender's uid */
+ int _status; /* exit code */
+ clock_t _utime;
+ clock_t _stime;
+ } _sigchld;
+
+ /* SIGILL, SIGFPE, SIGSEGV, SIGBUS */
+ struct {
+ void __user *_addr; /* faulting insn/memory ref. */
+#ifdef __ARCH_SI_TRAPNO
+ int _trapno; /* TRAP # which caused the signal */
+#endif
+#ifdef __ARCH_SI_ERROR
+ int _error; /* CPU error code */
+#endif
+ } _sigfault;
+
+ /* SIGPOLL */
+ struct {
+ __ARCH_SI_BAND_T _band; /* POLL_IN, POLL_OUT, POLL_MSG */
+ int _fd;
+ } _sigpoll;
+ } _sifields;
+} siginfo_t;
+
+#ifdef __ARCH_SI_TRAPNO
+#define si_trapno _sifields._sigfault._trapno
+#endif
+#ifdef __ARCH_SI_ERROR
+#define si_error _sifields._sigfault._error
+#endif
+
+#undef si_addr
+#define si_addr _sifields._sigfault._addr
+
int is_skas_winch(int pid, int fd, void *data)
{
if (pid != getpgrp())
@@ -91,11 +196,25 @@ bad_wait:
extern unsigned long current_stub_stack(void);
+#ifndef PTRACE_GETSIGINFO
+#define PTRACE_GETSIGINFO 0x4202
+#endif
+
void get_skas_faultinfo(int pid, struct faultinfo * fi)
{
+ siginfo_t si;
int err;
- if (ptrace_faultinfo) {
+ if(1){
+ err = ptrace(PTRACE_GETSIGINFO, pid, 0, &si);
+ if(err)
+ printk("PTRACE_GETSIGINFO failed, err = %d\n", errno);
+
+ fi->cr2 = (unsigned long) si.si_addr;
+ fi->error_code = si.si_error;
+ fi->trap_no = si.si_trapno;
+ }
+ else if (ptrace_faultinfo) {
err = ptrace(PTRACE_FAULTINFO, pid, 0, fi);
if (err)
panic("get_skas_faultinfo - PTRACE_FAULTINFO failed, "
@@ -212,7 +331,7 @@ static int userspace_tramp(void *stack)
}
}
}
- if (!ptrace_faultinfo && (stack != NULL)) {
+ if (!ptrace_faultinfo) {
struct sigaction sa;
unsigned long v = STUB_CODE +
@@ -256,7 +375,7 @@ int start_userspace(unsigned long stub_stack)
sp = (unsigned long) stack + UM_KERN_PAGE_SIZE - sizeof(void *);
flags = CLONE_FILES;
- if (proc_mm)
+ if (proc_mm || have_switch_mm)
flags |= CLONE_VM;
else
flags |= SIGCHLD;
@@ -369,8 +488,14 @@ void userspace(struct uml_pt_regs *regs)
printk(UM_KERN_ERR "userspace - child stopped "
"with signal %d\n", sig);
}
- pid = userspace_pid[0];
+
+ /*
+ * userspace_pid can change in in_interrupt since
+ * PTRACE_SWITCH_MM can cause a process to change
+ * address spaces
+ */
interrupt_end();
+ pid = userspace_pid[0];
/* Avoid -ERESTARTSYS handling in host */
if (PT_SYSCALL_NR_OFFSET != PT_SYSCALL_RET_OFFSET)
@@ -458,6 +583,69 @@ int copy_context_skas0(unsigned long new_stack, int pid)
return pid;
}
+extern unsigned long switch_mm_stub;
+extern long task_size;
+
+static void unmap_new_as(void)
+{
+ void (*p)(void);
+ void *addr;
+ unsigned long stack = (unsigned long) &stack & ~(UM_KERN_PAGE_SIZE - 1);
+ unsigned long long data_offset, code_offset;
+ int fd = phys_mapping(to_phys((void *) stack), &data_offset);
+
+ addr = mmap((void *) STUB_DATA, UM_KERN_PAGE_SIZE,
+ PROT_READ | PROT_WRITE, MAP_SHARED | MAP_FIXED, fd,
+ data_offset);
+ if (addr == MAP_FAILED)
+ panic("Failed to remap stack");
+
+ fd = phys_mapping(to_phys(&__syscall_stub_start), &code_offset);
+ addr = mmap((void *) STUB_CODE, UM_KERN_PAGE_SIZE,
+ PROT_READ | PROT_WRITE, MAP_SHARED | MAP_FIXED, fd,
+ code_offset);
+ if (addr == MAP_FAILED)
+ panic("Failed to remap code");
+
+ p = (void (*)(void)) (STUB_CODE +
+ ((unsigned long) &switch_mm_stub -
+ (unsigned long) &__syscall_stub_start));
+ (*p)();
+}
+
+int copy_context_skas4(struct mm_id *id)
+{
+ void *data = NULL;
+ unsigned long *return_regs, *fd_ptr, regs[MAX_REG_NR];
+ sigset_t sigs, old;
+ int err;
+
+ err = unmap(id, 0, STUB_START, 0, &data);
+ if (err)
+ return err;
+
+ err = unmap(id, STUB_END, task_size - STUB_END, 0, &data);
+ if (err)
+ return err;
+
+ return_regs = (unsigned long *) (id->stack + UM_KERN_PAGE_SIZE -
+ MAX_REG_NR * sizeof(long));
+ fd_ptr = return_regs - 1;
+ *fd_ptr = self_mm_fd;
+
+ regs[REGS_IP_INDEX] = (unsigned long) unmap_new_as;
+ regs[REGS_SP_INDEX] = id->stack + UM_KERN_PAGE_SIZE / 2;
+
+ sigfillset(&sigs);
+ sigprocmask(SIG_SETMASK, &sigs, &old);
+
+ err = switch_mm(id->u.mm_fd, MM_SP_IP, regs, return_regs);
+
+ sigprocmask(SIG_SETMASK, &old, NULL);
+
+ return err;
+}
+
/*
* This is used only, if stub pages are needed, while proc_mm is
* available. Opening /proc/mm creates a new mm_context, which lacks
@@ -612,11 +800,18 @@ void __switch_mm(struct mm_id *mm_idp)
/* FIXME: need cpu pid in __switch_mm */
if (proc_mm) {
- err = ptrace(PTRACE_SWITCH_MM, userspace_pid[0], 0,
+ err = ptrace(OLD_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);
}
+ else if (have_ptrace_switch_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);
+ }
else userspace_pid[0] = mm_idp->u.pid;
}
diff --git a/arch/um/os-Linux/start_up.c b/arch/um/os-Linux/start_up.c
index 7b81f6c..20a3c0c 100644
--- a/arch/um/os-Linux/start_up.c
+++ b/arch/um/os-Linux/start_up.c
@@ -23,6 +23,7 @@
#include "mem_user.h"
#include "ptrace_user.h"
#include "registers.h"
+#include "skas.h"
#include "skas_ptrace.h"
static int ptrace_child(void)
@@ -141,14 +142,27 @@ static int stop_ptraced_child(int pid, int exitcode, int mustexit)
}
/* Changed only during early boot */
-int ptrace_faultinfo = 1;
-int ptrace_ldt = 1;
-int proc_mm = 1;
-int skas_needs_stub = 0;
+int ptrace_faultinfo;
+static int disable_ptrace_faultinfo;
+
+int ptrace_ldt;
+static int disable_ptrace_ldt;
+
+int proc_mm;
+static int disable_proc_mm;
+
+int have_switch_mm;
+static int disable_switch_mm;
+
+int skas_needs_stub;
static int __init skas0_cmd_param(char *str, int* add)
{
- ptrace_faultinfo = proc_mm = 0;
+ disable_ptrace_faultinfo = 1;
+ disable_ptrace_ldt = 1;
+ disable_proc_mm = 1;
+ disable_switch_mm = 1;
+
return 0;
}
@@ -158,15 +172,12 @@ static int __init mode_skas0_cmd_param(char *str, int* add)
__attribute__((alias("skas0_cmd_param")));
__uml_setup("skas0", skas0_cmd_param,
- "skas0\n"
- " Disables SKAS3 usage, so that SKAS0 is used, unless \n"
- " you specify mode=tt.\n\n");
+"skas0\n"
+" Disables SKAS3 and SKAS4 usage, so that SKAS0 is used\n\n");
__uml_setup("mode=skas0", mode_skas0_cmd_param,
- "mode=skas0\n"
- " Disables SKAS3 usage, so that SKAS0 is used, unless you \n"
- " specify mode=tt. Note that this was recently added - on \n"
- " older kernels you must use simply \"skas0\".\n\n");
+"mode=skas0\n"
+" Disables SKAS3 and SKAS4 usage, so that SKAS0 is used.\n\n");
/* Changed only during early boot */
static int force_sysemu_disabled = 0;
@@ -341,6 +352,8 @@ static void __init check_coredump_limit(void)
void __init os_early_checks(void)
{
+ int pid;
+
/* Print out the core dump limits early */
check_coredump_limit();
@@ -350,11 +363,15 @@ void __init os_early_checks(void)
* kernel is running.
*/
check_tmpexec();
+
+ pid = start_ptraced_child();
+ init_registers(pid);
+ stop_ptraced_child(pid, 1, 1);
}
static int __init noprocmm_cmd_param(char *str, int* add)
{
- proc_mm = 0;
+ disable_proc_mm = 1;
return 0;
}
@@ -366,7 +383,7 @@ __uml_setup("noprocmm", noprocmm_cmd_param,
static int __init noptracefaultinfo_cmd_param(char *str, int* add)
{
- ptrace_faultinfo = 0;
+ disable_ptrace_faultinfo = 1;
return 0;
}
@@ -378,7 +395,7 @@ __uml_setup("noptracefaultinfo", noptracefaultinfo_cmd_param,
static int __init noptraceldt_cmd_param(char *str, int* add)
{
- ptrace_ldt = 0;
+ disable_ptrace_ldt = 1;
return 0;
}
@@ -398,20 +415,18 @@ static inline void check_skas3_ptrace_faultinfo(void)
n = ptrace(PTRACE_FAULTINFO, pid, 0, &fi);
if (n < 0) {
- ptrace_faultinfo = 0;
if (errno == EIO)
non_fatal("not found\n");
else
perror("not found");
}
+ else if (disable_ptrace_faultinfo)
+ non_fatal("found but disabled on command line\n");
else {
- if (!ptrace_faultinfo)
- non_fatal("found but disabled on command line\n");
- else
- non_fatal("found\n");
+ ptrace_faultinfo = 1;
+ non_fatal("found\n");
}
- init_registers(pid);
stop_ptraced_child(pid, 1, 1);
}
@@ -435,49 +450,236 @@ static inline void check_skas3_ptrace_ldt(void)
else {
perror("not found");
}
- ptrace_ldt = 0;
}
+ else if (disable_ptrace_ldt)
+ non_fatal("found, but use is disabled\n");
else {
- if (ptrace_ldt)
- non_fatal("found\n");
- else
- non_fatal("found, but use is disabled\n");
+ ptrace_ldt = 1;
+ non_fatal("found\n");
}
stop_ptraced_child(pid, 1, 1);
-#else
- /* PTRACE_LDT might be disabled via cmdline option.
- * We want to override this, else we might use the stub
- * without real need
- */
- ptrace_ldt = 1;
#endif
}
static inline void check_skas3_proc_mm(void)
{
non_fatal(" - /proc/mm...");
- if (access("/proc/mm", W_OK) < 0) {
- proc_mm = 0;
+ if (access("/proc/mm", W_OK) < 0)
perror("not found");
- }
- else if (!proc_mm)
+ else if (disable_proc_mm)
non_fatal("found but disabled on command line\n");
- else non_fatal("found\n");
+ else {
+ proc_mm = 1;
+ non_fatal("found\n");
+ }
}
-int can_do_skas(void)
+static void can_do_skas3(void)
{
non_fatal("Checking for the skas3 patch in the host:\n");
check_skas3_proc_mm();
check_skas3_ptrace_faultinfo();
check_skas3_ptrace_ldt();
+}
- if (!proc_mm || !ptrace_faultinfo || !ptrace_ldt)
- skas_needs_stub = 1;
+int get_new_mm(void)
+{
+ int err;
+
+ err = new_mm(MM_COPY);
+ if (err < 0)
+ err = -errno;
+
+ return err;
+}
+
+static char *mm_stack;
+static unsigned long return_regs[MAX_REG_NR];
+int self_mm_fd;
+
+static int switch_mm_works;
+
+static void after_switch(void)
+{
+ switch_mm_works = 0;
+
+ switch_mm(self_mm_fd, MM_ALL_REGS, return_regs, NULL);
+}
+
+static int check_switch_mm(void)
+{
+ unsigned long regs[MAX_REG_NR];
+ int err, there = -1;
+
+ non_fatal("\t/proc/self/mm ... ");
+ self_mm_fd = open("/proc/self/mm", O_RDONLY);
+ if (self_mm_fd < 0)
+ goto bad;
+ non_fatal("OK\n");
+
+ mm_stack = mmap(NULL, UM_KERN_PAGE_SIZE, PROT_READ | PROT_WRITE,
+ MAP_PRIVATE | MAP_ANONYMOUS, -1, 0);
+ if(mm_stack == MAP_FAILED)
+ goto bad;
+
+ non_fatal("\tnew_mm ... ");
+ there = new_mm(MM_COPY);
+ if(there < 0)
+ goto bad;
+ non_fatal("OK\n");
+
+ regs[REGS_IP_INDEX] = (unsigned long) after_switch;
+ regs[REGS_SP_INDEX] = ((unsigned long) &mm_stack[UM_KERN_PAGE_SIZE]) -
+ sizeof(void *);
+
+ non_fatal("\tswitching over ... ");
+ err = switch_mm(there, MM_SP_IP, regs, return_regs);
+ if (err < 0)
+ goto bad;
+ non_fatal("switched back ... ");
+ switch_mm_works = 1;
+ if(!switch_mm_works)
+ goto bad;
+ else
+ non_fatal("OK\n");
+
+ munmap(mm_stack, UM_KERN_PAGE_SIZE);
+ close(there);
+
+ if (disable_switch_mm)
+ non_fatal("SKAS4 support disabled on command line\n");
+ else
+ have_switch_mm = 1;
+
+ return 1;
+ bad:
+ if (there > 0)
+ close(there);
+ munmap(mm_stack, UM_KERN_PAGE_SIZE);
+ non_fatal("Failed - \n");
+ perror("");
+ return 0;
+}
+
+int have_ptrace_switch_mm;
+static int disable_ptrace_switch_mm;
+
+static int ptrace_switch_mm_works;
+
+static int after_ptrace_switch(void)
+{
+ ptrace_switch_mm_works = 1;
+ exit(0);
+}
+
+static int check_ptrace_switch_mm(void)
+{
+ void *stack;
+ unsigned long regs[MAX_REG_NR];
+ int pid, here, err, status;
+
+ non_fatal("\tPTRACE_SWITCH_MM ... ");
+ pid = fork();
+ if(pid == 0){
+ ptrace(PTRACE_TRACEME, 0, 0, 0);
+ kill(getpid(), SIGSTOP);
+
+ exit(0);
+ }
+ else if(pid < 0)
+ goto bad;
+
+ stack = mmap(NULL, UM_KERN_PAGE_SIZE, PROT_READ | PROT_WRITE,
+ MAP_SHARED | MAP_ANONYMOUS, -1, 0);
+ if(stack == MAP_FAILED)
+ goto bad;
+
+ here = open("/proc/self/mm", O_RDONLY);
+ if(here < 0)
+ goto bad_unmap;
+
+ err = waitpid(pid, &status, WUNTRACED);
+ if (err < 0)
+ goto bad_close;
+ else if (err != pid) {
+ non_fatal("waitpid returned %d, expected %d\n", err, pid);
+ goto bad_close;
+ }
+ else if (!WIFSTOPPED(status) || (WSTOPSIG(status) != SIGSTOP)) {
+ non_fatal("waitpid returned status 0x%d\n", status);
+ goto bad_close;
+ }
+
+ err = ptrace(PTRACE_GETREGS, pid, 0, regs);
+ if (err < 0)
+ goto bad_close;
+
+ regs[REGS_IP_INDEX] = (unsigned long) after_ptrace_switch;
+ regs[REGS_SP_INDEX] = (unsigned long) stack + UM_KERN_PAGE_SIZE -
+ sizeof(void *);
+
+ if (ptrace(PTRACE_SETREGS, pid, 0, regs) < 0)
+ goto bad_close;
+
+ if (ptrace(PTRACE_SWITCH_MM, pid, NULL, here) < 0)
+ goto bad_close;
+
+ if (ptrace(PTRACE_CONT, pid, NULL, 0) < 0)
+ goto bad_close;
+
+ err = waitpid(pid, &status, WUNTRACED);
+ if (err < 0)
+ goto bad_close;
+ else if(err != pid) {
+ non_fatal("waitpid returned %d, expected %d\n", err, pid);
+ goto bad_close;
+ }
+ else if (!WIFEXITED(status) || (WEXITSTATUS(status) != 0)) {
+ non_fatal("waitpid returned status 0x%d\n", status);
+ goto bad_close;
+ }
+
+ if (!ptrace_switch_mm_works)
+ goto bad_close;
+ else non_fatal("OK\n");
+
+ if (disable_ptrace_switch_mm)
+ non_fatal("PTRACE_SWITCH_MM support disabled on command "
+ "line\n");
+ else
+ have_ptrace_switch_mm = 1;
+
+ close(here);
+ munmap(stack, UM_KERN_PAGE_SIZE);
return 1;
+
+ bad_close:
+ close(here);
+ bad_unmap:
+ munmap(stack, UM_KERN_PAGE_SIZE);
+ bad:
+ non_fatal("Failed - \n");
+ perror("");
+ return 0;
+}
+
+static int can_do_skas4(void)
+{
+ non_fatal("Checking for new_mm and switch_mm support in the host:\n");
+
+ return check_switch_mm() && check_ptrace_switch_mm();
+}
+
+void can_do_skas(void)
+{
+ if(!can_do_skas4())
+ can_do_skas3();
+
+ if (!proc_mm || !ptrace_faultinfo || !ptrace_ldt)
+ skas_needs_stub = 1;
}
int __init parse_iomem(char *str, int *add)
diff --git a/arch/um/sys-i386/ldt.c b/arch/um/sys-i386/ldt.c
index 67c0958..c06c66c 100644
--- a/arch/um/sys-i386/ldt.c
+++ b/arch/um/sys-i386/ldt.c
@@ -436,7 +436,7 @@ long init_new_ldt(struct mm_context *new_mm, struct mm_context *from_mm)
/*
* We have a valid from_mm, so we now have to copy the LDT of
* from_mm to new_mm, because using proc_mm an new mm with
- * an empty/default LDT was created in new_mm()
+ * an empty/default LDT was created in make_new_mm()
*/
copy = ((struct proc_mm_op) { .op = MM_COPY_SEGMENTS,
.u =
diff --git a/arch/um/sys-i386/stub.S b/arch/um/sys-i386/stub.S
index e730772..3cfb6e8 100644
--- a/arch/um/sys-i386/stub.S
+++ b/arch/um/sys-i386/stub.S
@@ -1,52 +1,61 @@
#include "uml-config.h"
#include "as-layout.h"
+#include "skas/skas.h"
+
+#define MAX_REG_NR 17
+
+#define PROCESS_LIST \
+ /* load pointer to first operation */ \
+ mov $(ASM_STUB_DATA + 8), %esp; \
+1: \
+ /* load length of additional data */ \
+ mov 0x0(%esp), %eax; \
+ /* if(length == 0) : end of list */ \
+ /* write possible 0 to header */ \
+ mov %eax, ASM_STUB_DATA + 4; \
+ cmpl $0, %eax; \
+ jz 2f; \
+ /* save current pointer */ \
+ mov %esp, ASM_STUB_DATA + 4; \
+ /* skip additional data */ \
+ add %eax, %esp; \
+ /* load syscall-# */ \
+ pop %eax; \
+ /* load syscall params */ \
+ pop %ebx; \
+ pop %ecx; \
+ pop %edx; \
+ pop %esi; \
+ pop %edi; \
+ pop %ebp; \
+ /* execute syscall */ \
+ int $0x80; \
+ /* check return value */ \
+ pop %ebx; \
+ cmp %ebx, %eax; \
+ je 1b; \
+2: \
+ /* save return value */ \
+ mov %eax, ASM_STUB_DATA;
.globl syscall_stub
.section .__syscall_stub, "x"
.globl batch_syscall_stub
batch_syscall_stub:
- /* load pointer to first operation */
- mov $(ASM_STUB_DATA+8), %esp
-
-again:
- /* load length of additional data */
- mov 0x0(%esp), %eax
-
- /* if(length == 0) : end of list */
- /* write possible 0 to header */
- mov %eax, ASM_STUB_DATA+4
- cmpl $0, %eax
- jz done
-
- /* save current pointer */
- mov %esp, ASM_STUB_DATA+4
-
- /* skip additional data */
- add %eax, %esp
-
- /* load syscall-# */
- pop %eax
+ PROCESS_LIST
+ /* stop */
+ int3
- /* load syscall params */
- pop %ebx
- pop %ecx
- pop %edx
- pop %esi
- pop %edi
- pop %ebp
+ .globl switch_mm_stub
+switch_mm_stub:
+ PROCESS_LIST
- /* execute syscall */
+ mov $__NR_switch_mm, %eax
+ mov ASM_STUB_DATA + UM_KERN_PAGE_SIZE - MAX_REG_NR * 4 - 4, %ebx
+ mov $MM_ALL_REGS, %ecx
+ mov $(ASM_STUB_DATA + UM_KERN_PAGE_SIZE - MAX_REG_NR * 4), %edx
+ xor %esi, %esi
int $0x80
- /* check return value */
- pop %ebx
- cmp %ebx, %eax
- je again
-
-done:
- /* save return value */
- mov %eax, ASM_STUB_DATA
-
- /* stop */
int3
diff --git a/arch/um/sys-i386/syscalls.c b/arch/um/sys-i386/syscalls.c
index e2d1426..85621a2 100644
--- a/arch/um/sys-i386/syscalls.c
+++ b/arch/um/sys-i386/syscalls.c
@@ -200,3 +200,11 @@ long sys_sigaction(int sig, const struct old_sigaction __user *act,
return ret;
}
+
+extern long do_switch_mm(int fd, int flags, long __user *new,
+ long __user *save, struct pt_regs *regs);
+
+long sys_switch_mm(int fd, int flags, long __user *new, long __user *save)
+{
+ return do_switch_mm(fd, flags, new, save, ¤t->thread.regs);
+}
diff --git a/arch/um/sys-x86_64/stub.S b/arch/um/sys-x86_64/stub.S
index 4afe204..fb10353 100644
--- a/arch/um/sys-x86_64/stub.S
+++ b/arch/um/sys-x86_64/stub.S
@@ -1,67 +1,69 @@
#include "uml-config.h"
#include "as-layout.h"
+#include "skas/skas.h"
- .globl syscall_stub
-.section .__syscall_stub, "x"
-syscall_stub:
- syscall
- /* We don't have 64-bit constants, so this constructs the address
- * we need.
- */
- movq $(ASM_STUB_DATA >> 32), %rbx
- salq $32, %rbx
- movq $(ASM_STUB_DATA & 0xffffffff), %rcx
- or %rcx, %rbx
- movq %rax, (%rbx)
- int3
+#define MAX_REG_NR 27
+
+#define PROCESS_LIST \
+ mov $(ASM_STUB_DATA >> 32), %rbx; \
+ sal $32, %rbx; \
+ mov $(ASM_STUB_DATA & 0xffffffff), %rax; \
+ or %rax, %rbx; \
+ /* load pointer to first operation */ \
+ mov %rbx, %rsp; \
+ add $0x10, %rsp; \
+1: \
+ /* load length of additional data */ \
+ mov 0x0(%rsp), %rax; \
+ /* if(length == 0) : end of list */ \
+ /* write possible 0 to header */ \
+ mov %rax, 8(%rbx); \
+ cmp $0, %rax; \
+ jz 2f; \
+ /* save current pointer */ \
+ mov %rsp, 8(%rbx); \
+ /* skip additional data */ \
+ add %rax, %rsp; \
+ /* load syscall-# */ \
+ pop %rax; \
+ /* load syscall params */ \
+ pop %rdi; \
+ pop %rsi; \
+ pop %rdx; \
+ pop %r10; \
+ pop %r8; \
+ pop %r9; \
+ /* execute syscall */ \
+ syscall; \
+ /* check return value */ \
+ pop %rcx; \
+ cmp %rcx, %rax; \
+ je 1b; \
+2: \
+ /* save return value */ \
+ mov %rax, (%rbx); \
+.section .__syscall_stub, "x"
.globl batch_syscall_stub
batch_syscall_stub:
- mov $(ASM_STUB_DATA >> 32), %rbx
- sal $32, %rbx
- mov $(ASM_STUB_DATA & 0xffffffff), %rax
- or %rax, %rbx
- /* load pointer to first operation */
- mov %rbx, %rsp
- add $0x10, %rsp
-again:
- /* load length of additional data */
- mov 0x0(%rsp), %rax
-
- /* if(length == 0) : end of list */
- /* write possible 0 to header */
- mov %rax, 8(%rbx)
- cmp $0, %rax
- jz done
-
- /* save current pointer */
- mov %rsp, 8(%rbx)
-
- /* skip additional data */
- add %rax, %rsp
-
- /* load syscall-# */
- pop %rax
+ PROCESS_LIST
+ /* stop */
+ int3
- /* load syscall params */
- pop %rdi
- pop %rsi
- pop %rdx
- pop %r10
- pop %r8
- pop %r9
+ .globl switch_mm_stub
+switch_mm_stub:
+ PROCESS_LIST
- /* execute syscall */
+ mov $__NR_switch_mm, %rax
+ mov $(ASM_STUB_DATA >> 32), %rdi
+ sal $32, %rdi
+ mov $(ASM_STUB_DATA & 0xffffffff), %rsi
+ add %rsi, %rdi
+ add $(UM_KERN_PAGE_SIZE - MAX_REG_NR * 8 - 8), %rdi
+ mov (%rdi), %rdi
+ mov $MM_ALL_REGS, %rsi
+ mov $(ASM_STUB_DATA + UM_KERN_PAGE_SIZE - MAX_REG_NR * 8), %rdx
+ xor %r10, %r10
syscall
- /* check return value */
- pop %rcx
- cmp %rcx, %rax
- je again
-
-done:
- /* save return value */
- mov %rax, (%rbx)
-
- /* stop */
int3
diff --git a/arch/um/sys-x86_64/syscalls.c b/arch/um/sys-x86_64/syscalls.c
index 86f6b18..00131f9 100644
--- a/arch/um/sys-x86_64/syscalls.c
+++ b/arch/um/sys-x86_64/syscalls.c
@@ -30,7 +30,7 @@ long arch_prctl(struct task_struct *task, int code, unsigned long __user *addr)
{
unsigned long *ptr = addr, tmp;
long ret;
- int pid = task->mm->context.id.u.pid;
+ int pid = userspace_pid[0];
/*
* With ARCH_SET_FS (and ARCH_SET_GS is treated similarly to
@@ -112,3 +112,11 @@ void arch_switch_to(struct task_struct *from, struct task_struct *to)
arch_prctl(to, ARCH_SET_FS, (void __user *) to->thread.arch.fs);
}
+
+extern long do_switch_mm(int fd, int flags, long __user *new,
+ long __user *save, struct pt_regs *regs);
+
+long stub_switch_mm(int fd, int flags, long __user *new, long __user *save)
+{
+ return do_switch_mm(fd, flags, new, save, ¤t->thread.regs);
+}
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 3a058bb..b130f88 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -412,6 +412,7 @@ END(\label)
PTREGSCALL stub_rt_sigsuspend, sys_rt_sigsuspend, %rdx
PTREGSCALL stub_sigaltstack, sys_sigaltstack, %rdx
PTREGSCALL stub_iopl, sys_iopl, %rsi
+ PTREGSCALL stub_switch_mm, sys_switch_mm, %r8
ENTRY(ptregscall_common)
popq %r11
diff --git a/arch/x86/kernel/ptrace_32.c b/arch/x86/kernel/ptrace_32.c
index ff5431c..c3bd8cd 100644
--- a/arch/x86/kernel/ptrace_32.c
+++ b/arch/x86/kernel/ptrace_32.c
@@ -83,8 +83,8 @@ static inline int put_stack_long(struct task_struct *task, int offset,
return 0;
}
-static int putreg(struct task_struct *child,
- unsigned long regno, unsigned long value)
+int putreg(struct task_struct *child,
+ unsigned long regno, unsigned long value)
{
switch (regno >> 2) {
case GS:
@@ -116,7 +116,7 @@ static int putreg(struct task_struct *child,
return 0;
}
-static unsigned long getreg(struct task_struct *child,
+unsigned long getreg(struct task_struct *child,
unsigned long regno)
{
unsigned long retval = ~0UL;
diff --git a/arch/x86/kernel/ptrace_64.c b/arch/x86/kernel/ptrace_64.c
index 607085f..a568429 100644
--- a/arch/x86/kernel/ptrace_64.c
+++ b/arch/x86/kernel/ptrace_64.c
@@ -226,7 +226,7 @@ void ptrace_disable(struct task_struct *child)
clear_singlestep(child);
}
-static int putreg(struct task_struct *child,
+int putreg(struct task_struct *child,
unsigned long regno, unsigned long value)
{
unsigned long tmp;
@@ -283,7 +283,7 @@ static int putreg(struct task_struct *child,
return 0;
}
-static unsigned long getreg(struct task_struct *child, unsigned long regno)
+unsigned long getreg(struct task_struct *child, unsigned long regno)
{
unsigned long val;
switch (regno) {
diff --git a/arch/x86/kernel/sys_i386_32.c b/arch/x86/kernel/sys_i386_32.c
index a86d26f..7b9d43b 100644
--- a/arch/x86/kernel/sys_i386_32.c
+++ b/arch/x86/kernel/sys_i386_32.c
@@ -261,3 +261,12 @@ int kernel_execve(const char *filename, char *const argv[], char *const envp[])
: "0" (__NR_execve),"ri" (filename),"c" (argv), "d" (envp) : "memory");
return __res;
}
+
+extern long do_switch_mm(int fd, int flags, long __user *new, long __user *save,
+ struct pt_regs *regs);
+
+asmlinkage long sys_switch_mm(struct pt_regs regs)
+{
+ return do_switch_mm(regs.ebx, regs.ecx, (long __user *) regs.edx,
+ (long __user *) regs.esi, ®s);
+}
diff --git a/arch/x86/kernel/sys_x86_64.c b/arch/x86/kernel/sys_x86_64.c
index 907942e..ddc1c98 100644
--- a/arch/x86/kernel/sys_x86_64.c
+++ b/arch/x86/kernel/sys_x86_64.c
@@ -153,3 +153,12 @@ asmlinkage long sys_uname(struct new_utsname __user * name)
err |= copy_to_user(&name->machine, "i686", 5);
return err ? -EFAULT : 0;
}
+
+extern long do_switch_mm(int fd, int flags, long __user *new,
+ long __user *save, struct pt_regs *regs);
+
+asmlinkage long sys_switch_mm(int fd, int flags, long __user *new,
+ long __user *save, struct pt_regs *regs)
+{
+ return do_switch_mm(fd, flags, new, save, regs);
+}
diff --git a/arch/x86/kernel/syscall_table_32.S b/arch/x86/kernel/syscall_table_32.S
index 8344c70..3346997 100644
--- a/arch/x86/kernel/syscall_table_32.S
+++ b/arch/x86/kernel/syscall_table_32.S
@@ -324,3 +324,5 @@ ENTRY(sys_call_table)
.long sys_timerfd
.long sys_eventfd
.long sys_fallocate
+ .long sys_new_mm
+ .long sys_switch_mm
diff --git a/arch/x86/mm/fault_32.c b/arch/x86/mm/fault_32.c
index a2273d4..0e7940d 100644
--- a/arch/x86/mm/fault_32.c
+++ b/arch/x86/mm/fault_32.c
@@ -211,6 +211,8 @@ static noinline void force_sig_info_fault(int si_signo, int si_code,
info.si_errno = 0;
info.si_code = si_code;
info.si_addr = (void __user *)address;
+ info.si_trapno = tsk->thread.trap_no;
+ info.si_error = tsk->thread.error_code;
force_sig_info(si_signo, &info, tsk);
}
diff --git a/arch/x86/mm/fault_64.c b/arch/x86/mm/fault_64.c
index 0e26230..6365ba8 100644
--- a/arch/x86/mm/fault_64.c
+++ b/arch/x86/mm/fault_64.c
@@ -502,6 +502,8 @@ bad_area_nosemaphore:
info.si_errno = 0;
/* info.si_code has been set above */
info.si_addr = (void __user *)address;
+ info.si_trapno = tsk->thread.trap_no;
+ info.si_error = tsk->thread.error_code;
force_sig_info(SIGSEGV, &info, tsk);
return;
}
@@ -577,6 +579,8 @@ do_sigbus:
info.si_errno = 0;
info.si_code = BUS_ADRERR;
info.si_addr = (void __user *)address;
+ info.si_trapno = tsk->thread.trap_no;
+ info.si_error = tsk->thread.error_code;
force_sig_info(SIGBUS, &info, tsk);
return;
}
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 7411bfb..6dd8e34 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -2187,6 +2187,34 @@ static int proc_pid_io_accounting(struct task_struct *task, char *buffer)
}
#endif
+static int proc_pid_mm_open(struct inode *inode, struct file *file)
+{
+ struct task_struct *task = pid_task(proc_pid(inode), PIDTYPE_PID);
+
+ if (task == NULL)
+ return -ENOENT;
+
+ if(task->mm != NULL)
+ atomic_inc(&task->mm->mm_users);
+ file->private_data = task->mm;
+ return 0;
+}
+
+static int proc_pid_mm_release(struct inode *inode, struct file *file)
+{
+ struct mm_struct *mm = file->private_data;
+
+ if(mm != NULL)
+ mmput(mm);
+
+ return 0;
+}
+
+const struct file_operations proc_pid_mm_operations = {
+ .open = proc_pid_mm_open,
+ .release = proc_pid_mm_release,
+};
+
/*
* Thread groups
*/
@@ -2250,6 +2278,7 @@ static const struct pid_entry tgid_base_stuff[] = {
#ifdef CONFIG_TASK_IO_ACCOUNTING
INF("io", S_IRUGO, pid_io_accounting),
#endif
+ REG("mm", S_IRUSR | S_IWUSR, pid_mm),
};
static int proc_tgid_base_readdir(struct file * filp,
diff --git a/include/asm-generic/siginfo.h b/include/asm-generic/siginfo.h
index 8786e01..ffe5e49 100644
--- a/include/asm-generic/siginfo.h
+++ b/include/asm-generic/siginfo.h
@@ -82,6 +82,9 @@ typedef struct siginfo {
#ifdef __ARCH_SI_TRAPNO
int _trapno; /* TRAP # which caused the signal */
#endif
+#ifdef __ARCH_SI_ERROR
+ int _error; /* CPU error code */
+#endif
} _sigfault;
/* SIGPOLL */
@@ -112,6 +115,12 @@ typedef struct siginfo {
#ifdef __ARCH_SI_TRAPNO
#define si_trapno _sifields._sigfault._trapno
#endif
+#ifdef __ARCH_SI_ERROR
+#define si_error _sifields._sigfault._error
+#endif
+#ifdef __ARCH_SI_ERROR
+#define si_error _sifields._sigfault._error
+#endif
#define si_band _sifields._sigpoll._band
#define si_fd _sifields._sigpoll._fd
diff --git a/include/asm-um/processor-x86_64.h b/include/asm-um/processor-x86_64.h
index d946bf2..fd026c3 100644
--- a/include/asm-um/processor-x86_64.h
+++ b/include/asm-um/processor-x86_64.h
@@ -26,7 +26,7 @@ static inline void rep_nop(void)
#define cpu_relax() rep_nop()
#define INIT_ARCH_THREAD { .debugregs = { [ 0 ... 7 ] = 0 }, \
- .debugregs_seq = 0, \
+ .debugregs_seq = 0, \
.fs = 0, \
.faultinfo = { 0, 0, 0 } }
@@ -37,6 +37,7 @@ static inline void arch_flush_thread(struct arch_thread *thread)
static inline void arch_copy_thread(struct arch_thread *from,
struct arch_thread *to)
{
+ to->fs = from->fs;
}
#include "asm/arch/user.h"
diff --git a/include/asm-x86/siginfo.h b/include/asm-x86/siginfo.h
index a477bea..796ac81 100644
--- a/include/asm-x86/siginfo.h
+++ b/include/asm-x86/siginfo.h
@@ -5,6 +5,12 @@
# define __ARCH_SI_PREAMBLE_SIZE (4 * sizeof(int))
#endif
+#define __ARCH_SI_TRAPNO
+#define __ARCH_SI_ERROR
+
+#define __ARCH_SI_TRAPNO
+#define __ARCH_SI_ERROR
+
#include <asm-generic/siginfo.h>
#endif
diff --git a/include/asm-x86/unistd_32.h b/include/asm-x86/unistd_32.h
index 9b15545..3477555 100644
--- a/include/asm-x86/unistd_32.h
+++ b/include/asm-x86/unistd_32.h
@@ -330,10 +330,12 @@
#define __NR_timerfd 322
#define __NR_eventfd 323
#define __NR_fallocate 324
+#define __NR_new_mm 325
+#define __NR_switch_mm 326
#ifdef __KERNEL__
-#define NR_syscalls 325
+#define NR_syscalls 327
#define __ARCH_WANT_IPC_PARSE_VERSION
#define __ARCH_WANT_OLD_READDIR
diff --git a/include/asm-x86/unistd_64.h b/include/asm-x86/unistd_64.h
index 5ff4d3e..baf4c0c 100644
--- a/include/asm-x86/unistd_64.h
+++ b/include/asm-x86/unistd_64.h
@@ -635,6 +635,10 @@ __SYSCALL(__NR_timerfd, sys_timerfd)
__SYSCALL(__NR_eventfd, sys_eventfd)
#define __NR_fallocate 285
__SYSCALL(__NR_fallocate, sys_fallocate)
+#define __NR_new_mm 286
+__SYSCALL(__NR_new_mm, sys_new_mm)
+#define __NR_switch_mm 287
+__SYSCALL(__NR_switch_mm, stub_switch_mm)
#ifndef __NO_STUBS
#define __ARCH_WANT_OLD_READDIR
diff --git a/include/linux/ptrace.h b/include/linux/ptrace.h
index 3ea5750..6758e86 100644
--- a/include/linux/ptrace.h
+++ b/include/linux/ptrace.h
@@ -21,6 +21,8 @@
#define PTRACE_SYSCALL 24
+#define PTRACE_SWITCH_MM 33
+
/* 0x4200-0x4300 are reserved for architecture-independent additions. */
#define PTRACE_SETOPTIONS 0x4200
#define PTRACE_GETEVENTMSG 0x4201
diff --git a/include/linux/sched.h b/include/linux/sched.h
index ac3d496..7707a43 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1665,6 +1665,7 @@ static inline int sas_ss_flags(unsigned long sp)
* Routines for handling mm_structs
*/
extern struct mm_struct * mm_alloc(void);
+extern struct mm_struct *dup_mm(struct task_struct *tsk);
/* mmdrop drops the mm and the page tables */
extern void FASTCALL(__mmdrop(struct mm_struct *));
diff --git a/include/linux/signalfd.h b/include/linux/signalfd.h
index 86f9b1e..71e3c45 100644
--- a/include/linux/signalfd.h
+++ b/include/linux/signalfd.h
@@ -26,6 +26,8 @@ struct signalfd_siginfo {
__u64 ssi_utime;
__u64 ssi_stime;
__u64 ssi_addr;
+ __u32 ssi_trap_no;
+ __u32 ssi_error_code;
/*
* Pad strcture to 128 bytes. Remember to update the
@@ -36,7 +38,7 @@ struct signalfd_siginfo {
* comes out of a read(2) and we really don't want to have
* a compat on read(2).
*/
- __u8 __pad[48];
+ __u8 __pad[40];
};
diff --git a/kernel/Makefile b/kernel/Makefile
index dfa9695..ecaf05e 100644
--- a/kernel/Makefile
+++ b/kernel/Makefile
@@ -3,7 +3,7 @@
#
obj-y = sched.o fork.o exec_domain.o panic.o printk.o profile.o \
- exit.o itimer.o time.o softirq.o resource.o \
+ exit.o itimer.o time.o softirq.o resource.o mmfs.o \
sysctl.o capability.o ptrace.o timer.o user.o user_namespace.o \
signal.o sys.o kmod.o workqueue.o pid.o \
rcupdate.o extable.o params.o posix-timers.o \
diff --git a/kernel/fork.c b/kernel/fork.c
index 8dd8ff2..bd9afde 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -491,7 +491,7 @@ void mm_release(struct task_struct *tsk, struct mm_struct *mm)
* Allocate a new mm structure and copy contents from the
* mm structure of the passed in task structure.
*/
-static struct mm_struct *dup_mm(struct task_struct *tsk)
+struct mm_struct *dup_mm(struct task_struct *tsk)
{
struct mm_struct *mm, *oldmm = current->mm;
int err;
diff --git a/kernel/ptrace.c b/kernel/ptrace.c
index c25db86..317e888 100644
--- a/kernel/ptrace.c
+++ b/kernel/ptrace.c
@@ -366,6 +366,29 @@ static int ptrace_setsiginfo(struct task_struct *child, siginfo_t __user * data)
return error;
}
+extern struct mm_struct *fd_to_mm(int fd);
+
+static int ptrace_switch_mm(struct task_struct *child, int mm_fd)
+{
+ struct mm_struct *old = child->mm;
+ struct mm_struct *new = fd_to_mm(mm_fd);
+
+ if (IS_ERR(new))
+ return PTR_ERR(new);
+
+ task_lock(child);
+
+ atomic_inc(&new->mm_users);
+
+ child->mm = new;
+ child->active_mm = new;
+
+ task_unlock(child);
+ mmput(old);
+
+ return 0;
+}
+
int ptrace_request(struct task_struct *child, long request,
long addr, long data)
{
@@ -390,6 +413,9 @@ int ptrace_request(struct task_struct *child, long request,
case PTRACE_DETACH: /* detach a process that was attached. */
ret = ptrace_detach(child, data);
break;
+ case PTRACE_SWITCH_MM:
+ ret = ptrace_switch_mm(child, data);
+ break;
default:
break;
}
diff --git a/kernel/signal.c b/kernel/signal.c
index afa4f78..1e067a1 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -2108,6 +2108,9 @@ int copy_siginfo_to_user(siginfo_t __user *to, siginfo_t *from)
#ifdef __ARCH_SI_TRAPNO
err |= __put_user(from->si_trapno, &to->si_trapno);
#endif
+#ifdef __ARCH_SI_ERROR
+ err |= __put_user(from->si_error, &to->si_error);
+#endif
break;
case __SI_CHLD:
err |= __put_user(from->si_pid, &to->si_pid);
-------------------------------------------------------------------------
Check out the new SourceForge.net Marketplace.
It's the best place to buy or sell services for
just about anything Open Source.
http://ad.doubleclick.net/clk;164216239;13503038;w?http://sf.net/marketplace
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel
^ permalink raw reply related [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-01-10 2:49 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-01-09 22:09 [uml-devel] [RFC PATCH 0/9] SKAS4 Jeff Dike
2008-01-10 2:48 ` Jeff Dike
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.