@@ -268,19 +268,17 @@ extern long long os_persistent_clock_emulation(void);
extern long long os_nsecs(void);
/* skas/mem.c */
-extern long run_syscall_stub(struct mm_id * mm_idp,
- int syscall, unsigned long *args, long expected,
- void **addr, int done);
-extern long syscall_stub_data(struct mm_id * mm_idp,
- unsigned long *data, int data_count,
- void **addr, void **stub_addr);
-extern int map(struct mm_id * mm_idp, unsigned long virt,
- unsigned long len, int prot, int phys_fd,
- unsigned long long offset, int done, void **data);
-extern int unmap(struct mm_id * mm_idp, unsigned long addr, unsigned long len,
- int done, void **data);
-extern int protect(struct mm_id * mm_idp, unsigned long addr,
- unsigned long len, unsigned int prot, int done, void **data);
+int syscall_stub_flush(struct mm_id *mm_idp);
+struct stub_syscall *syscall_stub_alloc(struct mm_id *mm_idp,
+ unsigned long data_len,
+ unsigned long *data_addr);
+
+void map(struct mm_id *mm_idp, unsigned long virt,
+ unsigned long len, int prot, int phys_fd,
+ unsigned long long offset);
+void unmap(struct mm_id *mm_idp, unsigned long addr, unsigned long len);
+void protect(struct mm_id *mm_idp, unsigned long addr,
+ unsigned long len, unsigned int prot);
/* skas/process.c */
extern int is_skas_winch(int pid, int fd, void *data);
@@ -13,6 +13,7 @@ struct mm_id {
} u;
unsigned long stack;
int kill;
+ int syscall_data_len;
};
#endif
@@ -11,11 +11,23 @@
#include <linux/compiler_types.h>
#include <as-layout.h>
+#define STUB_NEXT_SYSCALL(s) \
+ ((struct stub_syscall *) (((unsigned long) s) + (s)->cmd_len))
+
+struct stub_syscall {
+ long syscall;
+ int cmd_len;
+ long expected_result;
+ long arg[6];
+ long data[];
+};
+
struct stub_data {
unsigned long offset;
int fd;
- long parent_err, child_err;
+ long err, child_err;
+ int syscall_data_len;
/* 128 leaves enough room for additional fields in the struct. */
unsigned char syscall_data[UM_KERN_PAGE_SIZE - MINSIGSTKSZ - 128]
__aligned(16);
@@ -40,11 +40,19 @@ extern void panic(const char *fmt, ...)
#ifdef UML_CONFIG_PRINTK
extern int printk(const char *fmt, ...)
__attribute__ ((format (printf, 1, 2)));
+extern void print_hex_dump(const char *level, const char *prefix_str,
+ int prefix_type, int rowsize, int groupsize,
+ const void *buf, size_t len, _Bool ascii);
#else
static inline int printk(const char *fmt, ...)
{
return 0;
}
+static inline void print_hex_dump(const char *level, const char *prefix_str,
+ int prefix_type, int rowsize, int groupsize,
+ const void *buf, size_t len, _Bool ascii)
+{
+}
#endif
extern int in_aton(char *str);
@@ -21,15 +21,11 @@
void flush_thread(void)
{
- void *data = NULL;
- int ret;
-
arch_flush_thread(¤t->thread.arch);
- ret = unmap(¤t->mm->context.id, 0, TASK_SIZE, 1, &data);
- if (ret) {
- printk(KERN_ERR "%s - clearing address space failed, err = %d\n",
- __func__, ret);
+ unmap(¤t->mm->context.id, 0, TASK_SIZE);
+ if (syscall_stub_flush(¤t->mm->context.id) < 0) {
+ printk(KERN_ERR "%s - clearing address space failed", __func__);
force_sig(SIGKILL);
}
get_safe_registers(current_pt_regs()->regs.gp,
@@ -3,14 +3,14 @@
# Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com)
#
-obj-y := clone.o mmu.o process.o syscall.o uaccess.o
+obj-y := clone.o stub.o mmu.o process.o syscall.o uaccess.o
# clone.o is in the stub, so it can't be built with profiling
# GCC hardened also auto-enables -fpic, but we need %ebx so it can't work ->
# disable it
CFLAGS_clone.o := $(CFLAGS_NO_HARDENING)
-UNPROFILE_OBJS := clone.o
+UNPROFILE_OBJS := clone.o stub.o
KCOV_INSTRUMENT := n
@@ -34,7 +34,7 @@ stub_clone_handler(void)
sizeof(data->syscall_data) -
sizeof(void *));
if (err) {
- data->parent_err = err;
+ data->err = err;
goto done;
}
new file mode 100644
@@ -0,0 +1,52 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2021 Benjamin Berg <benjamin@sipsolutions.net>
+ */
+
+#include <sysdep/stub.h>
+
+static __always_inline int
+syscall_handler(struct stub_data *d)
+{
+ struct stub_syscall *sc;
+ long ret;
+
+ sc = (void *) &d->syscall_data;
+ while ((unsigned long) sc - (unsigned long) d->syscall_data < d->syscall_data_len) {
+ ret = stub_syscall6(sc->syscall,
+ sc->arg[0], sc->arg[1], sc->arg[2],
+ sc->arg[3], sc->arg[4], sc->arg[5]);
+
+ /*
+ * If there was an error, then set d->err and set
+ * d->syscall_data_len to point to the failed syscall.
+ */
+ if (ret != sc->expected_result) {
+ d->err = ret;
+ d->syscall_data_len = (unsigned long) sc - (unsigned long) d->syscall_data;
+
+ return -1;
+ }
+
+ sc = STUB_NEXT_SYSCALL(sc);
+ }
+ d->err = 0;
+ d->syscall_data_len = 0;
+
+ return 0;
+}
+
+void __section(".__syscall_stub")
+stub_syscall_handler(void)
+{
+ /*
+ * NOTE: Putting this inside the inlined function will result in
+ * incorrect optimizations with GCC 10.2.1.
+ */
+ int stack;
+ struct stub_data *d = (void *) ((unsigned long)&stack & ~(UM_KERN_PAGE_SIZE - 1));
+
+ syscall_handler(d);
+
+ trap_myself();
+}
@@ -70,21 +70,19 @@ static int do_ops(struct host_vm_change *hvc, int end,
switch (op->type) {
case MMAP:
if (hvc->userspace)
- ret = map(&hvc->mm->context.id, op->u.mmap.addr,
- op->u.mmap.len, op->u.mmap.prot,
- op->u.mmap.fd,
- op->u.mmap.offset, finished,
- &hvc->data);
+ map(&hvc->mm->context.id, op->u.mmap.addr,
+ op->u.mmap.len, op->u.mmap.prot,
+ op->u.mmap.fd,
+ op->u.mmap.offset);
else
map_memory(op->u.mmap.addr, op->u.mmap.offset,
op->u.mmap.len, 1, 1, 1);
break;
case MUNMAP:
if (hvc->userspace)
- ret = unmap(&hvc->mm->context.id,
- op->u.munmap.addr,
- op->u.munmap.len, finished,
- &hvc->data);
+ unmap(&hvc->mm->context.id,
+ op->u.munmap.addr,
+ op->u.munmap.len);
else
ret = os_unmap_memory(
(void *) op->u.munmap.addr,
@@ -93,11 +91,10 @@ static int do_ops(struct host_vm_change *hvc, int end,
break;
case MPROTECT:
if (hvc->userspace)
- ret = protect(&hvc->mm->context.id,
- op->u.mprotect.addr,
- op->u.mprotect.len,
- op->u.mprotect.prot,
- finished, &hvc->data);
+ protect(&hvc->mm->context.id,
+ op->u.mprotect.addr,
+ op->u.mprotect.len,
+ op->u.mprotect.prot);
else
ret = os_protect_memory(
(void *) op->u.mprotect.addr,
@@ -112,6 +109,9 @@ static int do_ops(struct host_vm_change *hvc, int end,
}
}
+ if (hvc->userspace && finished)
+ ret = syscall_stub_flush(&hvc->mm->context.id);
+
if (ret == -ENOMEM)
report_enomem();
@@ -460,7 +460,6 @@ void flush_tlb_page(struct vm_area_struct *vma, unsigned long address)
pmd_t *pmd;
pte_t *pte;
struct mm_struct *mm = vma->vm_mm;
- void *flush = NULL;
int r, w, x, prot, err = 0;
struct mm_id *mm_id;
@@ -503,14 +502,13 @@ void flush_tlb_page(struct vm_area_struct *vma, unsigned long address)
int fd;
fd = phys_mapping(pte_val(*pte) & PAGE_MASK, &offset);
- err = map(mm_id, address, PAGE_SIZE, prot, fd, offset,
- 1, &flush);
- }
- else err = unmap(mm_id, address, PAGE_SIZE, 1, &flush);
- }
- else if (pte_newprot(*pte))
- err = protect(mm_id, address, PAGE_SIZE, prot, 1, &flush);
+ map(mm_id, address, PAGE_SIZE, prot, fd, offset);
+ } else
+ unmap(mm_id, address, PAGE_SIZE);
+ } else if (pte_newprot(*pte))
+ protect(mm_id, address, PAGE_SIZE, prot);
+ err = syscall_stub_flush(mm_id);
if (err) {
if (err == -ENOMEM)
report_enomem();
@@ -1,5 +1,6 @@
// SPDX-License-Identifier: GPL-2.0
/*
+ * Copyright (C) 2021 Benjamin Berg <benjamin@sipsolutions.net>
* Copyright (C) 2002 - 2007 Jeff Dike (jdike@{addtoit,linux.intel}.com)
*/
@@ -18,11 +19,11 @@
#include <sysdep/ptrace.h>
#include <sysdep/stub.h>
-extern char batch_syscall_stub[], __syscall_stub_start[];
+extern char __syscall_stub_start[];
extern void wait_stub_done(int pid);
-static inline unsigned long *check_init_stack(struct mm_id * mm_idp,
+static inline unsigned long *check_init_stack(struct mm_id *mm_idp,
unsigned long *stack)
{
if (stack == NULL) {
@@ -37,22 +38,22 @@ static unsigned long syscall_regs[MAX_REG_NR];
static int __init init_syscall_regs(void)
{
get_safe_registers(syscall_regs, NULL);
+
syscall_regs[REGS_IP_INDEX] = STUB_CODE +
- ((unsigned long) batch_syscall_stub -
+ ((unsigned long) stub_syscall_handler -
(unsigned long) __syscall_stub_start);
- syscall_regs[REGS_SP_INDEX] = STUB_DATA;
+ syscall_regs[REGS_SP_INDEX] = STUB_DATA + UM_KERN_PAGE_SIZE -
+ sizeof(void *);
return 0;
}
__initcall(init_syscall_regs);
-static inline long do_syscall_stub(struct mm_id * mm_idp, void **addr)
+static inline long do_syscall_stub(struct mm_id *mm_idp)
{
+ struct stub_data *proc_data = (void *)mm_idp->stack;
int n, i;
- long ret, offset;
- unsigned long * data;
- unsigned long * syscall;
int err, pid = mm_idp->u.pid;
n = ptrace_setregs(pid, syscall_regs);
@@ -64,6 +65,9 @@ static inline long do_syscall_stub(struct mm_id * mm_idp, void **addr)
__func__, -n);
}
+ /* Inform process how much we have filled in. */
+ proc_data->syscall_data_len = mm_idp->syscall_data_len;
+
err = ptrace(PTRACE_CONT, pid, 0, 0);
if (err)
panic("Failed to continue stub, pid = %d, errno = %d\n", pid,
@@ -72,135 +76,148 @@ 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)
+ * proc_data->err will be non-zero if there was an (unexpected) error.
+ * In that case, syscall_data_len points to the last executed syscall,
+ * otherwise it will be zero (but we do not need to rely on that).
*/
- 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 "%s : ret = %ld, offset = %ld, data = %p\n",
- __func__, ret, offset, data);
- syscall = (unsigned long *)((unsigned long)data + data[0]);
- printk(UM_KERN_ERR "%s: syscall %ld failed, return value = 0x%lx, expected return value = 0x%lx\n",
- __func__, syscall[0], ret, syscall[7]);
+ if (proc_data->err) {
+ struct stub_syscall *sc;
+
+ if (proc_data->syscall_data_len < 0 ||
+ proc_data->syscall_data_len > (long) mm_idp->syscall_data_len - sizeof(*sc))
+ panic("Syscall data was corrupted by stub (len is: %d, expected maximum: %d)!",
+ proc_data->syscall_data_len,
+ mm_idp->syscall_data_len);
+
+ sc = (void *) (((unsigned long) &proc_data->syscall_data) +
+ proc_data->syscall_data_len);
+
+ printk(UM_KERN_ERR "%s : length = %d, last offset = %d",
+ __func__, mm_idp->syscall_data_len,
+ proc_data->syscall_data_len);
+ printk(UM_KERN_ERR "%s : syscall %ld failed, return value = 0x%lx, expected return value = 0x%lx\n",
+ __func__, sc->syscall, proc_data->err,
+ sc->expected_result);
+
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]);
+ sc->arg[0], sc->arg[1], sc->arg[2],
+ sc->arg[3], sc->arg[4], sc->arg[5]);
+
+ n = sc->cmd_len - sizeof(*sc);
+ if (n > 0) {
+ printk(UM_KERN_ERR " syscall data 0x%lx + %d",
+ STUB_DATA + ((unsigned long) (&sc->data) &
+ (UM_KERN_PAGE_SIZE - 1)),
+ n);
+ print_hex_dump(UM_KERN_ERR,
+ " syscall data: ", 0,
+ 16, 4, sc->data, n, 0);
}
- if (n > 1)
- printk("\n");
- }
- else ret = 0;
- *addr = check_init_stack(mm_idp, NULL);
+ /* Store error code in case someone tries to add more syscalls */
+ mm_idp->syscall_data_len = proc_data->err;
+ } else {
+ mm_idp->syscall_data_len = 0;
+ }
- return ret;
+ return mm_idp->syscall_data_len;
}
-long run_syscall_stub(struct mm_id * mm_idp, int syscall,
- unsigned long *args, long expected, void **addr,
- int done)
+int syscall_stub_flush(struct mm_id *mm_idp)
{
- unsigned long *stack = check_init_stack(mm_idp, *addr);
-
- *stack += sizeof(long);
- stack += *stack / sizeof(long);
-
- *stack++ = syscall;
- *stack++ = args[0];
- *stack++ = args[1];
- *stack++ = args[2];
- *stack++ = args[3];
- *stack++ = args[4];
- *stack++ = args[5];
- *stack++ = expected;
- *stack = 0;
-
- if (!done && ((((unsigned long) stack) & ~UM_KERN_PAGE_MASK) <
- UM_KERN_PAGE_SIZE - 10 * sizeof(long))) {
- *addr = stack;
+ int res;
+
+ if (mm_idp->syscall_data_len == 0)
return 0;
+
+ /* If an error happened already, report it and reset the state. */
+ if (mm_idp->syscall_data_len < 0) {
+ res = mm_idp->syscall_data_len;
+ mm_idp->syscall_data_len = 0;
+ return res;
}
- return do_syscall_stub(mm_idp, addr);
+ res = do_syscall_stub(mm_idp);
+ mm_idp->syscall_data_len = 0;
+
+ return res;
}
-long syscall_stub_data(struct mm_id * mm_idp,
- unsigned long *data, int data_count,
- void **addr, void **stub_addr)
+struct stub_syscall *syscall_stub_alloc(struct mm_id *mm_idp,
+ unsigned long data_len,
+ unsigned long *data_addr)
{
- unsigned long *stack;
- int ret = 0;
-
- /*
- * 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;
+ struct stub_syscall *sc;
+ struct stub_data *proc_data = (struct stub_data *) mm_idp->stack;
+ int len;
+
+ /* Align to sizeof(long) */
+ data_len = (data_len + sizeof(long) - 1) & ~(sizeof(long) - 1);
+ len = sizeof(struct stub_syscall) + data_len;
+
+ if (len > sizeof(proc_data->syscall_data))
+ panic("Syscall data too large to marshal!");
+
+ if (mm_idp->syscall_data_len > 0 &&
+ mm_idp->syscall_data_len + len > sizeof(proc_data->syscall_data))
+ do_syscall_stub(mm_idp);
+
+ if (mm_idp->syscall_data_len < 0) {
+ /* Return dummy without changing the syscall_next_offset to
+ * retain error state.
+ */
+ sc = (void *) &proc_data->syscall_data;
+ } else {
+ sc = (void *) (((unsigned long) &proc_data->syscall_data) +
+ mm_idp->syscall_data_len);
+ mm_idp->syscall_data_len += len;
}
+ memset(sc, 0, len);
+ sc->cmd_len = len;
- stack = check_init_stack(mm_idp, *addr);
- *addr = stack;
-
- *stack = data_count * sizeof(long);
+ if (data_addr)
+ *data_addr = STUB_DATA +
+ ((unsigned long) (&sc->data) &
+ (UM_KERN_PAGE_SIZE - 1));
- memcpy(stack + 1, data, data_count * sizeof(long));
-
- *stub_addr = (void *)(((unsigned long)(stack + 1) &
- ~UM_KERN_PAGE_MASK) + STUB_DATA);
-
- return 0;
+ return sc;
}
-int map(struct mm_id * mm_idp, unsigned long virt, unsigned long len, int prot,
- int phys_fd, unsigned long long offset, int done, void **data)
-{
- int ret;
- unsigned long args[] = { virt, len, prot,
- MAP_SHARED | MAP_FIXED, phys_fd,
- MMAP_OFFSET(offset) };
-
- ret = run_syscall_stub(mm_idp, STUB_MMAP_NR, args, virt,
- data, done);
- return ret;
+void map(struct mm_id *mm_idp, unsigned long virt, unsigned long len, int prot,
+ int phys_fd, unsigned long long offset)
+{
+ struct stub_syscall *sc;
+
+ sc = syscall_stub_alloc(mm_idp, 0, NULL);
+ sc->syscall = STUB_MMAP_NR;
+ sc->expected_result = virt;
+ sc->arg[0] = virt;
+ sc->arg[1] = len;
+ sc->arg[2] = prot;
+ sc->arg[3] = MAP_SHARED | MAP_FIXED;
+ sc->arg[4] = phys_fd;
+ sc->arg[5] = MMAP_OFFSET(offset);
}
-int unmap(struct mm_id * mm_idp, unsigned long addr, unsigned long len,
- int done, void **data)
+void unmap(struct mm_id *mm_idp, unsigned long addr, unsigned long len)
{
- int ret;
- unsigned long args[] = { (unsigned long) addr, len, 0, 0, 0,
- 0 };
+ struct stub_syscall *sc;
- ret = run_syscall_stub(mm_idp, __NR_munmap, args, 0,
- data, done);
-
- return ret;
+ sc = syscall_stub_alloc(mm_idp, 0, NULL);
+ sc->syscall = __NR_munmap;
+ sc->arg[0] = addr;
+ sc->arg[1] = len;
}
-int protect(struct mm_id * mm_idp, unsigned long addr, unsigned long len,
- unsigned int prot, int done, void **data)
+void protect(struct mm_id *mm_idp, unsigned long addr, unsigned long len,
+ unsigned int prot)
{
- int ret;
- unsigned long args[] = { addr, len, prot, 0, 0, 0 };
-
- ret = run_syscall_stub(mm_idp, __NR_mprotect, args, 0,
- data, done);
+ struct stub_syscall *sc;
- return ret;
+ sc = syscall_stub_alloc(mm_idp, 0, NULL);
+ sc->syscall = __NR_mprotect;
+ sc->arg[0] = addr;
+ sc->arg[1] = len;
+ sc->arg[2] = prot;
}
@@ -502,7 +502,7 @@ int copy_context_skas0(unsigned long new_stack, int pid)
*data = ((struct stub_data) {
.offset = MMAP_OFFSET(new_offset),
.fd = new_fd,
- .parent_err = -ESRCH,
+ .err = -ESRCH,
.child_err = 0,
});
@@ -539,7 +539,7 @@ int copy_context_skas0(unsigned long new_stack, int pid)
wait_stub_done(pid);
- pid = data->parent_err;
+ pid = data->err;
if (pid < 0) {
printk(UM_KERN_ERR "%s - stub-parent reports error %d\n",
__func__, -pid);
@@ -11,7 +11,7 @@ endif
obj-y = bugs_$(BITS).o delay.o fault.o ldt.o \
ptrace_$(BITS).o ptrace_user.o setjmp_$(BITS).o signal.o \
- stub_$(BITS).o stub_segv.o \
+ stub_segv.o \
sys_call_table_$(BITS).o sysrq_$(BITS).o tls_$(BITS).o \
mem_$(BITS).o subarch.o os-$(OS)/
@@ -12,31 +12,26 @@
#include <os.h>
#include <skas.h>
#include <sysdep/tls.h>
+#include <stub-data.h>
static inline int modify_ldt (int func, void *ptr, unsigned long bytecount)
{
return syscall(__NR_modify_ldt, func, ptr, bytecount);
}
-static long write_ldt_entry(struct mm_id *mm_idp, int func,
- struct user_desc *desc, void **addr, int done)
+static void write_ldt_entry(struct mm_id *mm_idp, int func,
+ struct user_desc *desc)
{
- long res;
- void *stub_addr;
- res = syscall_stub_data(mm_idp, (unsigned long *)desc,
- (sizeof(*desc) + sizeof(long) - 1) &
- ~(sizeof(long) - 1),
- addr, &stub_addr);
- if (!res) {
- unsigned long args[] = { func,
- (unsigned long)stub_addr,
- sizeof(*desc),
- 0, 0, 0 };
- res = run_syscall_stub(mm_idp, __NR_modify_ldt, args,
- 0, addr, done);
- }
-
- return res;
+ struct stub_syscall *sc;
+ unsigned long data_addr;
+
+ sc = syscall_stub_alloc(mm_idp, sizeof(*desc), &data_addr);
+ memcpy(sc->data, desc, sizeof(*desc));
+ sc->expected_result = 0;
+ sc->syscall = __NR_modify_ldt;
+ sc->arg[0] = func;
+ sc->arg[1] = data_addr;
+ sc->arg[2] = sizeof(*desc);
}
/*
@@ -125,7 +120,6 @@ static int write_ldt(void __user * ptr, unsigned long bytecount, int func)
int i, err;
struct user_desc ldt_info;
struct ldt_entry entry0, *ldt_p;
- void *addr = NULL;
err = -EINVAL;
if (bytecount != sizeof(ldt_info))
@@ -146,7 +140,8 @@ static int write_ldt(void __user * ptr, unsigned long bytecount, int func)
mutex_lock(&ldt->lock);
- err = write_ldt_entry(mm_idp, func, &ldt_info, &addr, 1);
+ write_ldt_entry(mm_idp, func, &ldt_info);
+ err = syscall_stub_flush(mm_idp);
if (err)
goto out_unlock;
@@ -164,7 +159,8 @@ static int write_ldt(void __user * ptr, unsigned long bytecount, int func)
err = -ENOMEM;
/* Undo the change in host */
memset(&ldt_info, 0, sizeof(ldt_info));
- write_ldt_entry(mm_idp, 1, &ldt_info, &addr, 1);
+ write_ldt_entry(mm_idp, 1, &ldt_info);
+ err = syscall_stub_flush(mm_idp);
goto out_unlock;
}
if (i == 0) {
@@ -301,7 +297,6 @@ long init_new_ldt(struct mm_context *new_mm, struct mm_context *from_mm)
short * num_p;
int i;
long page, err=0;
- void *addr = NULL;
mutex_init(&new_mm->arch.ldt.lock);
@@ -316,11 +311,9 @@ long init_new_ldt(struct mm_context *new_mm, struct mm_context *from_mm)
ldt_get_host_info();
for (num_p=host_ldt_entries; *num_p != -1; num_p++) {
desc.entry_number = *num_p;
- err = write_ldt_entry(&new_mm->id, 1, &desc,
- &addr, *(num_p + 1) == -1);
- if (err)
- break;
+ write_ldt_entry(&new_mm->id, 1, &desc);
}
+ err = syscall_stub_flush(&new_mm->id);
new_mm->arch.ldt.entry_count = 0;
goto out;
@@ -12,4 +12,5 @@
#endif
extern void stub_segv_handler(int, siginfo_t *, void *);
+extern void stub_syscall_handler(void);
extern void stub_clone_handler(void);
deleted file mode 100644
@@ -1,56 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#include <as-layout.h>
-
-.section .__syscall_stub, "ax"
-
- .globl batch_syscall_stub
-batch_syscall_stub:
- /* %esp comes in as "top of page" */
- mov %esp, %ecx
- /* %esp has pointer to first operation */
- add $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, 0x4(%ecx)
- cmpl $0, %eax
- jz done
-
- /* save current pointer */
- mov %esp, 0x4(%ecx)
-
- /* 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
-
- /* restore top of page pointer in %ecx */
- mov %esp, %ecx
- andl $(~UM_KERN_PAGE_SIZE) + 1, %ecx
-
- /* check return value */
- pop %ebx
- cmp %ebx, %eax
- je again
-
-done:
- /* save return value */
- mov %eax, (%ecx)
-
- /* stop */
- int3
deleted file mode 100644
@@ -1,50 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-#include <as-layout.h>
-
-.section .__syscall_stub, "ax"
- .globl batch_syscall_stub
-batch_syscall_stub:
- /* %rsp has the pointer to first operation */
- mov %rsp, %rbx
- 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
-
- /* 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 again
-
-done:
- /* save return value */
- mov %rax, (%rbx)
-
- /* stop */
- int3
Rework syscall handling to be platform independent. Also create a clean split between queueing of syscalls and flushing them out, removing the need to keep state in the code that triggers the syscalls. The code adds syscall_data_len to the global mm_id structure. This will be used later to allow surrounding code to track whether syscalls still need to run and if errors occurred. The patch decreases the amount of memory available to queue syscalls, as it prevents overlap between the signal stack and syscall data. This is intentional, as such an overlap must not happen when using seccomp. Signed-off-by: Benjamin Berg <benjamin@sipsolutions.net> --- arch/um/include/shared/os.h | 24 ++- arch/um/include/shared/skas/mm_id.h | 1 + arch/um/include/shared/skas/stub-data.h | 14 +- arch/um/include/shared/user.h | 8 + arch/um/kernel/exec.c | 10 +- arch/um/kernel/skas/Makefile | 4 +- arch/um/kernel/skas/clone.c | 2 +- arch/um/kernel/skas/stub.c | 52 ++++++ arch/um/kernel/tlb.c | 42 ++--- arch/um/os-Linux/skas/mem.c | 239 +++++++++++++----------- arch/um/os-Linux/skas/process.c | 4 +- arch/x86/um/Makefile | 2 +- arch/x86/um/ldt.c | 45 ++--- arch/x86/um/shared/sysdep/stub.h | 1 + arch/x86/um/stub_32.S | 56 ------ arch/x86/um/stub_64.S | 50 ----- 16 files changed, 262 insertions(+), 292 deletions(-) create mode 100644 arch/um/kernel/skas/stub.c delete mode 100644 arch/x86/um/stub_32.S delete mode 100644 arch/x86/um/stub_64.S