android_kernel_motorola_sm6225/arch/s390/lib/uaccess_pt.c

440 lines
10 KiB
C
Raw Normal View History

/*
* arch/s390/lib/uaccess_pt.c
*
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
* User access functions based on page table walks for enhanced
* system layout without hardware support.
*
* Copyright IBM Corp. 2006
* Author(s): Gerald Schaefer (gerald.schaefer@de.ibm.com)
*/
#include <linux/errno.h>
#include <linux/hardirq.h>
#include <linux/mm.h>
#include <asm/uaccess.h>
#include <asm/futex.h>
#include "uaccess.h"
static inline pte_t *follow_table(struct mm_struct *mm, unsigned long addr)
{
pgd_t *pgd;
pud_t *pud;
pmd_t *pmd;
pgd = pgd_offset(mm, addr);
if (pgd_none(*pgd) || unlikely(pgd_bad(*pgd)))
return NULL;
pud = pud_offset(pgd, addr);
if (pud_none(*pud) || unlikely(pud_bad(*pud)))
return NULL;
pmd = pmd_offset(pud, addr);
if (pmd_none(*pmd) || unlikely(pmd_bad(*pmd)))
return NULL;
return pte_offset_map(pmd, addr);
}
static int __handle_fault(struct mm_struct *mm, unsigned long address,
int write_access)
{
struct vm_area_struct *vma;
int ret = -EFAULT;
mm: fault feedback #2 This patch completes Linus's wish that the fault return codes be made into bit flags, which I agree makes everything nicer. This requires requires all handle_mm_fault callers to be modified (possibly the modifications should go further and do things like fault accounting in handle_mm_fault -- however that would be for another patch). [akpm@linux-foundation.org: fix alpha build] [akpm@linux-foundation.org: fix s390 build] [akpm@linux-foundation.org: fix sparc build] [akpm@linux-foundation.org: fix sparc64 build] [akpm@linux-foundation.org: fix ia64 build] Signed-off-by: Nick Piggin <npiggin@suse.de> Cc: Richard Henderson <rth@twiddle.net> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ian Molton <spyro@f2s.com> Cc: Bryan Wu <bryan.wu@analog.com> Cc: Mikael Starvik <starvik@axis.com> Cc: David Howells <dhowells@redhat.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: "Luck, Tony" <tony.luck@intel.com> Cc: Hirokazu Takata <takata@linux-m32r.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Greg Ungerer <gerg@uclinux.org> Cc: Matthew Wilcox <willy@debian.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp> Cc: Richard Curnow <rc@rc0.org.uk> Cc: William Lee Irwin III <wli@holomorphy.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp> Cc: Chris Zankel <chris@zankel.net> Acked-by: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Haavard Skinnemoen <hskinnemoen@atmel.com> Acked-by: Ralf Baechle <ralf@linux-mips.org> Acked-by: Andi Kleen <ak@muc.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> [ Still apparently needs some ARM and PPC loving - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 10:47:05 +02:00
int fault;
if (in_atomic())
return ret;
down_read(&mm->mmap_sem);
vma = find_vma(mm, address);
if (unlikely(!vma))
goto out;
if (unlikely(vma->vm_start > address)) {
if (!(vma->vm_flags & VM_GROWSDOWN))
goto out;
if (expand_stack(vma, address))
goto out;
}
if (!write_access) {
/* page not present, check vm flags */
if (!(vma->vm_flags & (VM_READ | VM_EXEC | VM_WRITE)))
goto out;
} else {
if (!(vma->vm_flags & VM_WRITE))
goto out;
}
survive:
mm: fault feedback #2 This patch completes Linus's wish that the fault return codes be made into bit flags, which I agree makes everything nicer. This requires requires all handle_mm_fault callers to be modified (possibly the modifications should go further and do things like fault accounting in handle_mm_fault -- however that would be for another patch). [akpm@linux-foundation.org: fix alpha build] [akpm@linux-foundation.org: fix s390 build] [akpm@linux-foundation.org: fix sparc build] [akpm@linux-foundation.org: fix sparc64 build] [akpm@linux-foundation.org: fix ia64 build] Signed-off-by: Nick Piggin <npiggin@suse.de> Cc: Richard Henderson <rth@twiddle.net> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ian Molton <spyro@f2s.com> Cc: Bryan Wu <bryan.wu@analog.com> Cc: Mikael Starvik <starvik@axis.com> Cc: David Howells <dhowells@redhat.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: "Luck, Tony" <tony.luck@intel.com> Cc: Hirokazu Takata <takata@linux-m32r.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Greg Ungerer <gerg@uclinux.org> Cc: Matthew Wilcox <willy@debian.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp> Cc: Richard Curnow <rc@rc0.org.uk> Cc: William Lee Irwin III <wli@holomorphy.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp> Cc: Chris Zankel <chris@zankel.net> Acked-by: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Haavard Skinnemoen <hskinnemoen@atmel.com> Acked-by: Ralf Baechle <ralf@linux-mips.org> Acked-by: Andi Kleen <ak@muc.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> [ Still apparently needs some ARM and PPC loving - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 10:47:05 +02:00
fault = handle_mm_fault(mm, vma, address, write_access);
if (unlikely(fault & VM_FAULT_ERROR)) {
if (fault & VM_FAULT_OOM)
goto out_of_memory;
else if (fault & VM_FAULT_SIGBUS)
goto out_sigbus;
BUG();
}
mm: fault feedback #2 This patch completes Linus's wish that the fault return codes be made into bit flags, which I agree makes everything nicer. This requires requires all handle_mm_fault callers to be modified (possibly the modifications should go further and do things like fault accounting in handle_mm_fault -- however that would be for another patch). [akpm@linux-foundation.org: fix alpha build] [akpm@linux-foundation.org: fix s390 build] [akpm@linux-foundation.org: fix sparc build] [akpm@linux-foundation.org: fix sparc64 build] [akpm@linux-foundation.org: fix ia64 build] Signed-off-by: Nick Piggin <npiggin@suse.de> Cc: Richard Henderson <rth@twiddle.net> Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru> Cc: Russell King <rmk@arm.linux.org.uk> Cc: Ian Molton <spyro@f2s.com> Cc: Bryan Wu <bryan.wu@analog.com> Cc: Mikael Starvik <starvik@axis.com> Cc: David Howells <dhowells@redhat.com> Cc: Yoshinori Sato <ysato@users.sourceforge.jp> Cc: "Luck, Tony" <tony.luck@intel.com> Cc: Hirokazu Takata <takata@linux-m32r.org> Cc: Geert Uytterhoeven <geert@linux-m68k.org> Cc: Roman Zippel <zippel@linux-m68k.org> Cc: Greg Ungerer <gerg@uclinux.org> Cc: Matthew Wilcox <willy@debian.org> Cc: Paul Mackerras <paulus@samba.org> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Heiko Carstens <heiko.carstens@de.ibm.com> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com> Cc: Paul Mundt <lethal@linux-sh.org> Cc: Kazumoto Kojima <kkojima@rr.iij4u.or.jp> Cc: Richard Curnow <rc@rc0.org.uk> Cc: William Lee Irwin III <wli@holomorphy.com> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jeff Dike <jdike@addtoit.com> Cc: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it> Cc: Miles Bader <uclinux-v850@lsi.nec.co.jp> Cc: Chris Zankel <chris@zankel.net> Acked-by: Kyle McMartin <kyle@mcmartin.ca> Acked-by: Haavard Skinnemoen <hskinnemoen@atmel.com> Acked-by: Ralf Baechle <ralf@linux-mips.org> Acked-by: Andi Kleen <ak@muc.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> [ Still apparently needs some ARM and PPC loving - Linus ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-07-19 10:47:05 +02:00
if (fault & VM_FAULT_MAJOR)
current->maj_flt++;
else
current->min_flt++;
ret = 0;
out:
up_read(&mm->mmap_sem);
return ret;
out_of_memory:
up_read(&mm->mmap_sem);
pid namespaces: define is_global_init() and is_container_init() is_init() is an ambiguous name for the pid==1 check. Split it into is_global_init() and is_container_init(). A cgroup init has it's tsk->pid == 1. A global init also has it's tsk->pid == 1 and it's active pid namespace is the init_pid_ns. But rather than check the active pid namespace, compare the task structure with 'init_pid_ns.child_reaper', which is initialized during boot to the /sbin/init process and never changes. Changelog: 2.6.22-rc4-mm2-pidns1: - Use 'init_pid_ns.child_reaper' to determine if a given task is the global init (/sbin/init) process. This would improve performance and remove dependence on the task_pid(). 2.6.21-mm2-pidns2: - [Sukadev Bhattiprolu] Changed is_container_init() calls in {powerpc, ppc,avr32}/traps.c for the _exception() call to is_global_init(). This way, we kill only the cgroup if the cgroup's init has a bug rather than force a kernel panic. [akpm@linux-foundation.org: fix comment] [sukadev@us.ibm.com: Use is_global_init() in arch/m32r/mm/fault.c] [bunk@stusta.de: kernel/pid.c: remove unused exports] [sukadev@us.ibm.com: Fix capability.c to work with threaded init] Signed-off-by: Serge E. Hallyn <serue@us.ibm.com> Signed-off-by: Sukadev Bhattiprolu <sukadev@us.ibm.com> Acked-by: Pavel Emelianov <xemul@openvz.org> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: Cedric Le Goater <clg@fr.ibm.com> Cc: Dave Hansen <haveblue@us.ibm.com> Cc: Herbert Poetzel <herbert@13thfloor.at> Cc: Kirill Korotaev <dev@sw.ru> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 08:39:52 +02:00
if (is_global_init(current)) {
yield();
down_read(&mm->mmap_sem);
goto survive;
}
printk("VM: killing process %s\n", current->comm);
return ret;
out_sigbus:
up_read(&mm->mmap_sem);
current->thread.prot_addr = address;
current->thread.trap_no = 0x11;
force_sig(SIGBUS, current);
return ret;
}
static size_t __user_copy_pt(unsigned long uaddr, void *kptr,
size_t n, int write_user)
{
struct mm_struct *mm = current->mm;
unsigned long offset, pfn, done, size;
pte_t *pte;
void *from, *to;
done = 0;
retry:
spin_lock(&mm->page_table_lock);
do {
pte = follow_table(mm, uaddr);
if (!pte || !pte_present(*pte) ||
(write_user && !pte_write(*pte)))
goto fault;
pfn = pte_pfn(*pte);
if (!pfn_valid(pfn))
goto out;
offset = uaddr & (PAGE_SIZE - 1);
size = min(n - done, PAGE_SIZE - offset);
if (write_user) {
to = (void *)((pfn << PAGE_SHIFT) + offset);
from = kptr + done;
} else {
from = (void *)((pfn << PAGE_SHIFT) + offset);
to = kptr + done;
}
memcpy(to, from, size);
done += size;
uaddr += size;
} while (done < n);
out:
spin_unlock(&mm->page_table_lock);
return n - done;
fault:
spin_unlock(&mm->page_table_lock);
if (__handle_fault(mm, uaddr, write_user))
return n - done;
goto retry;
}
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
/*
* Do DAT for user address by page table walk, return kernel address.
* This function needs to be called with current->mm->page_table_lock held.
*/
static unsigned long __dat_user_addr(unsigned long uaddr)
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
{
struct mm_struct *mm = current->mm;
unsigned long pfn, ret;
pte_t *pte;
int rc;
ret = 0;
retry:
pte = follow_table(mm, uaddr);
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
if (!pte || !pte_present(*pte))
goto fault;
pfn = pte_pfn(*pte);
if (!pfn_valid(pfn))
goto out;
ret = (pfn << PAGE_SHIFT) + (uaddr & (PAGE_SIZE - 1));
out:
return ret;
fault:
spin_unlock(&mm->page_table_lock);
rc = __handle_fault(mm, uaddr, 0);
spin_lock(&mm->page_table_lock);
if (rc)
goto out;
goto retry;
}
size_t copy_from_user_pt(size_t n, const void __user *from, void *to)
{
size_t rc;
if (segment_eq(get_fs(), KERNEL_DS)) {
memcpy(to, (void __kernel __force *) from, n);
return 0;
}
rc = __user_copy_pt((unsigned long) from, to, n, 0);
if (unlikely(rc))
memset(to + n - rc, 0, rc);
return rc;
}
size_t copy_to_user_pt(size_t n, void __user *to, const void *from)
{
if (segment_eq(get_fs(), KERNEL_DS)) {
memcpy((void __kernel __force *) to, from, n);
return 0;
}
return __user_copy_pt((unsigned long) to, (void *) from, n, 1);
}
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
static size_t clear_user_pt(size_t n, void __user *to)
{
long done, size, ret;
if (segment_eq(get_fs(), KERNEL_DS)) {
memset((void __kernel __force *) to, 0, n);
return 0;
}
done = 0;
do {
if (n - done > PAGE_SIZE)
size = PAGE_SIZE;
else
size = n - done;
ret = __user_copy_pt((unsigned long) to + done,
&empty_zero_page, size, 1);
done += size;
if (ret)
return ret + n - done;
} while (done < n);
return 0;
}
static size_t strnlen_user_pt(size_t count, const char __user *src)
{
char *addr;
unsigned long uaddr = (unsigned long) src;
struct mm_struct *mm = current->mm;
unsigned long offset, pfn, done, len;
pte_t *pte;
size_t len_str;
if (segment_eq(get_fs(), KERNEL_DS))
return strnlen((const char __kernel __force *) src, count) + 1;
done = 0;
retry:
spin_lock(&mm->page_table_lock);
do {
pte = follow_table(mm, uaddr);
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
if (!pte || !pte_present(*pte))
goto fault;
pfn = pte_pfn(*pte);
if (!pfn_valid(pfn)) {
done = -1;
goto out;
}
offset = uaddr & (PAGE_SIZE-1);
addr = (char *)(pfn << PAGE_SHIFT) + offset;
len = min(count - done, PAGE_SIZE - offset);
len_str = strnlen(addr, len);
done += len_str;
uaddr += len_str;
} while ((len_str == len) && (done < count));
out:
spin_unlock(&mm->page_table_lock);
return done + 1;
fault:
spin_unlock(&mm->page_table_lock);
if (__handle_fault(mm, uaddr, 0)) {
return 0;
}
goto retry;
}
static size_t strncpy_from_user_pt(size_t count, const char __user *src,
char *dst)
{
size_t n = strnlen_user_pt(count, src);
if (!n)
return -EFAULT;
if (n > count)
n = count;
if (segment_eq(get_fs(), KERNEL_DS)) {
memcpy(dst, (const char __kernel __force *) src, n);
if (dst[n-1] == '\0')
return n-1;
else
return n;
}
if (__user_copy_pt((unsigned long) src, dst, n, 0))
return -EFAULT;
if (dst[n-1] == '\0')
return n-1;
else
return n;
}
static size_t copy_in_user_pt(size_t n, void __user *to,
const void __user *from)
{
struct mm_struct *mm = current->mm;
unsigned long offset_from, offset_to, offset_max, pfn_from, pfn_to,
uaddr, done, size;
unsigned long uaddr_from = (unsigned long) from;
unsigned long uaddr_to = (unsigned long) to;
pte_t *pte_from, *pte_to;
int write_user;
done = 0;
retry:
spin_lock(&mm->page_table_lock);
do {
pte_from = follow_table(mm, uaddr_from);
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
if (!pte_from || !pte_present(*pte_from)) {
uaddr = uaddr_from;
write_user = 0;
goto fault;
}
pte_to = follow_table(mm, uaddr_to);
[S390] noexec protection This provides a noexec protection on s390 hardware. Our hardware does not have any bits left in the pte for a hw noexec bit, so this is a different approach using shadow page tables and a special addressing mode that allows separate address spaces for code and data. As a special feature of our "secondary-space" addressing mode, separate page tables can be specified for the translation of data addresses (storage operands) and instruction addresses. The shadow page table is used for the instruction addresses and the standard page table for the data addresses. The shadow page table is linked to the standard page table by a pointer in page->lru.next of the struct page corresponding to the page that contains the standard page table (since page->private is not really private with the pte_lock and the page table pages are not in the LRU list). Depending on the software bits of a pte, it is either inserted into both page tables or just into the standard (data) page table. Pages of a vma that does not have the VM_EXEC bit set get mapped only in the data address space. Any try to execute code on such a page will cause a page translation exception. The standard reaction to this is a SIGSEGV with two exceptions: the two system call opcodes 0x0a77 (sys_sigreturn) and 0x0aad (sys_rt_sigreturn) are allowed. They are stored by the kernel to the signal stack frame. Unfortunately, the signal return mechanism cannot be modified to use an SA_RESTORER because the exception unwinding code depends on the system call opcode stored behind the signal stack frame. This feature requires that user space is executed in secondary-space mode and the kernel in home-space mode, which means that the addressing modes need to be switched and that the noexec protection only works for user space. After switching the addressing modes, we cannot use the mvcp/mvcs instructions anymore to copy between kernel and user space. A new mvcos instruction has been added to the z9 EC/BC hardware which allows to copy between arbitrary address spaces, but on older hardware the page tables need to be walked manually. Signed-off-by: Gerald Schaefer <geraldsc@de.ibm.com> Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
2007-02-05 21:18:17 +01:00
if (!pte_to || !pte_present(*pte_to) || !pte_write(*pte_to)) {
uaddr = uaddr_to;
write_user = 1;
goto fault;
}
pfn_from = pte_pfn(*pte_from);
if (!pfn_valid(pfn_from))
goto out;
pfn_to = pte_pfn(*pte_to);
if (!pfn_valid(pfn_to))
goto out;
offset_from = uaddr_from & (PAGE_SIZE-1);
offset_to = uaddr_from & (PAGE_SIZE-1);
offset_max = max(offset_from, offset_to);
size = min(n - done, PAGE_SIZE - offset_max);
memcpy((void *)(pfn_to << PAGE_SHIFT) + offset_to,
(void *)(pfn_from << PAGE_SHIFT) + offset_from, size);
done += size;
uaddr_from += size;
uaddr_to += size;
} while (done < n);
out:
spin_unlock(&mm->page_table_lock);
return n - done;
fault:
spin_unlock(&mm->page_table_lock);
if (__handle_fault(mm, uaddr, write_user))
return n - done;
goto retry;
}
#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg) \
asm volatile("0: l %1,0(%6)\n" \
"1: " insn \
"2: cs %1,%2,0(%6)\n" \
"3: jl 1b\n" \
" lhi %0,0\n" \
"4:\n" \
EX_TABLE(0b,4b) EX_TABLE(2b,4b) EX_TABLE(3b,4b) \
: "=d" (ret), "=&d" (oldval), "=&d" (newval), \
"=m" (*uaddr) \
: "0" (-EFAULT), "d" (oparg), "a" (uaddr), \
"m" (*uaddr) : "cc" );
int futex_atomic_op_pt(int op, int __user *uaddr, int oparg, int *old)
{
int oldval = 0, newval, ret;
spin_lock(&current->mm->page_table_lock);
uaddr = (int __user *) __dat_user_addr((unsigned long) uaddr);
if (!uaddr) {
spin_unlock(&current->mm->page_table_lock);
return -EFAULT;
}
get_page(virt_to_page(uaddr));
spin_unlock(&current->mm->page_table_lock);
switch (op) {
case FUTEX_OP_SET:
__futex_atomic_op("lr %2,%5\n",
ret, oldval, newval, uaddr, oparg);
break;
case FUTEX_OP_ADD:
__futex_atomic_op("lr %2,%1\nar %2,%5\n",
ret, oldval, newval, uaddr, oparg);
break;
case FUTEX_OP_OR:
__futex_atomic_op("lr %2,%1\nor %2,%5\n",
ret, oldval, newval, uaddr, oparg);
break;
case FUTEX_OP_ANDN:
__futex_atomic_op("lr %2,%1\nnr %2,%5\n",
ret, oldval, newval, uaddr, oparg);
break;
case FUTEX_OP_XOR:
__futex_atomic_op("lr %2,%1\nxr %2,%5\n",
ret, oldval, newval, uaddr, oparg);
break;
default:
ret = -ENOSYS;
}
put_page(virt_to_page(uaddr));
*old = oldval;
return ret;
}
int futex_atomic_cmpxchg_pt(int __user *uaddr, int oldval, int newval)
{
int ret;
spin_lock(&current->mm->page_table_lock);
uaddr = (int __user *) __dat_user_addr((unsigned long) uaddr);
if (!uaddr) {
spin_unlock(&current->mm->page_table_lock);
return -EFAULT;
}
get_page(virt_to_page(uaddr));
spin_unlock(&current->mm->page_table_lock);
asm volatile(" cs %1,%4,0(%5)\n"
"0: lr %0,%1\n"
"1:\n"
EX_TABLE(0b,1b)
: "=d" (ret), "+d" (oldval), "=m" (*uaddr)
: "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr)
: "cc", "memory" );
put_page(virt_to_page(uaddr));
return ret;
}
struct uaccess_ops uaccess_pt = {
.copy_from_user = copy_from_user_pt,
.copy_from_user_small = copy_from_user_pt,
.copy_to_user = copy_to_user_pt,
.copy_to_user_small = copy_to_user_pt,
.copy_in_user = copy_in_user_pt,
.clear_user = clear_user_pt,
.strnlen_user = strnlen_user_pt,
.strncpy_from_user = strncpy_from_user_pt,
.futex_atomic_op = futex_atomic_op_pt,
.futex_atomic_cmpxchg = futex_atomic_cmpxchg_pt,
};