kernel_optimize_test/fs/binfmt_som.c
David Howells a6f76f23d2 CRED: Make execve() take advantage of copy-on-write credentials
Make execve() take advantage of copy-on-write credentials, allowing it to set
up the credentials in advance, and then commit the whole lot after the point
of no return.

This patch and the preceding patches have been tested with the LTP SELinux
testsuite.

This patch makes several logical sets of alteration:

 (1) execve().

     The credential bits from struct linux_binprm are, for the most part,
     replaced with a single credentials pointer (bprm->cred).  This means that
     all the creds can be calculated in advance and then applied at the point
     of no return with no possibility of failure.

     I would like to replace bprm->cap_effective with:

	cap_isclear(bprm->cap_effective)

     but this seems impossible due to special behaviour for processes of pid 1
     (they always retain their parent's capability masks where normally they'd
     be changed - see cap_bprm_set_creds()).

     The following sequence of events now happens:

     (a) At the start of do_execve, the current task's cred_exec_mutex is
     	 locked to prevent PTRACE_ATTACH from obsoleting the calculation of
     	 creds that we make.

     (a) prepare_exec_creds() is then called to make a copy of the current
     	 task's credentials and prepare it.  This copy is then assigned to
     	 bprm->cred.

  	 This renders security_bprm_alloc() and security_bprm_free()
     	 unnecessary, and so they've been removed.

     (b) The determination of unsafe execution is now performed immediately
     	 after (a) rather than later on in the code.  The result is stored in
     	 bprm->unsafe for future reference.

     (c) prepare_binprm() is called, possibly multiple times.

     	 (i) This applies the result of set[ug]id binaries to the new creds
     	     attached to bprm->cred.  Personality bit clearance is recorded,
     	     but now deferred on the basis that the exec procedure may yet
     	     fail.

         (ii) This then calls the new security_bprm_set_creds().  This should
	     calculate the new LSM and capability credentials into *bprm->cred.

	     This folds together security_bprm_set() and parts of
	     security_bprm_apply_creds() (these two have been removed).
	     Anything that might fail must be done at this point.

         (iii) bprm->cred_prepared is set to 1.

	     bprm->cred_prepared is 0 on the first pass of the security
	     calculations, and 1 on all subsequent passes.  This allows SELinux
	     in (ii) to base its calculations only on the initial script and
	     not on the interpreter.

     (d) flush_old_exec() is called to commit the task to execution.  This
     	 performs the following steps with regard to credentials:

	 (i) Clear pdeath_signal and set dumpable on certain circumstances that
	     may not be covered by commit_creds().

         (ii) Clear any bits in current->personality that were deferred from
             (c.i).

     (e) install_exec_creds() [compute_creds() as was] is called to install the
     	 new credentials.  This performs the following steps with regard to
     	 credentials:

         (i) Calls security_bprm_committing_creds() to apply any security
             requirements, such as flushing unauthorised files in SELinux, that
             must be done before the credentials are changed.

	     This is made up of bits of security_bprm_apply_creds() and
	     security_bprm_post_apply_creds(), both of which have been removed.
	     This function is not allowed to fail; anything that might fail
	     must have been done in (c.ii).

         (ii) Calls commit_creds() to apply the new credentials in a single
             assignment (more or less).  Possibly pdeath_signal and dumpable
             should be part of struct creds.

	 (iii) Unlocks the task's cred_replace_mutex, thus allowing
	     PTRACE_ATTACH to take place.

         (iv) Clears The bprm->cred pointer as the credentials it was holding
             are now immutable.

         (v) Calls security_bprm_committed_creds() to apply any security
             alterations that must be done after the creds have been changed.
             SELinux uses this to flush signals and signal handlers.

     (f) If an error occurs before (d.i), bprm_free() will call abort_creds()
     	 to destroy the proposed new credentials and will then unlock
     	 cred_replace_mutex.  No changes to the credentials will have been
     	 made.

 (2) LSM interface.

     A number of functions have been changed, added or removed:

     (*) security_bprm_alloc(), ->bprm_alloc_security()
     (*) security_bprm_free(), ->bprm_free_security()

     	 Removed in favour of preparing new credentials and modifying those.

     (*) security_bprm_apply_creds(), ->bprm_apply_creds()
     (*) security_bprm_post_apply_creds(), ->bprm_post_apply_creds()

     	 Removed; split between security_bprm_set_creds(),
     	 security_bprm_committing_creds() and security_bprm_committed_creds().

     (*) security_bprm_set(), ->bprm_set_security()

     	 Removed; folded into security_bprm_set_creds().

     (*) security_bprm_set_creds(), ->bprm_set_creds()

     	 New.  The new credentials in bprm->creds should be checked and set up
     	 as appropriate.  bprm->cred_prepared is 0 on the first call, 1 on the
     	 second and subsequent calls.

     (*) security_bprm_committing_creds(), ->bprm_committing_creds()
     (*) security_bprm_committed_creds(), ->bprm_committed_creds()

     	 New.  Apply the security effects of the new credentials.  This
     	 includes closing unauthorised files in SELinux.  This function may not
     	 fail.  When the former is called, the creds haven't yet been applied
     	 to the process; when the latter is called, they have.

 	 The former may access bprm->cred, the latter may not.

 (3) SELinux.

     SELinux has a number of changes, in addition to those to support the LSM
     interface changes mentioned above:

     (a) The bprm_security_struct struct has been removed in favour of using
     	 the credentials-under-construction approach.

     (c) flush_unauthorized_files() now takes a cred pointer and passes it on
     	 to inode_has_perm(), file_has_perm() and dentry_open().

Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: James Morris <jmorris@namei.org>
Acked-by: Serge Hallyn <serue@us.ibm.com>
Signed-off-by: James Morris <jmorris@namei.org>
2008-11-14 10:39:24 +11:00

311 lines
7.5 KiB
C

/*
* linux/fs/binfmt_som.c
*
* These are the functions used to load SOM format executables as used
* by HP-UX.
*
* Copyright 1999 Matthew Wilcox <willy@bofh.ai>
* based on binfmt_elf which is
* Copyright 1993, 1994: Eric Youngdale (ericy@cais.com).
*/
#include <linux/module.h>
#include <linux/fs.h>
#include <linux/stat.h>
#include <linux/sched.h>
#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/errno.h>
#include <linux/signal.h>
#include <linux/binfmts.h>
#include <linux/som.h>
#include <linux/string.h>
#include <linux/file.h>
#include <linux/fcntl.h>
#include <linux/ptrace.h>
#include <linux/slab.h>
#include <linux/shm.h>
#include <linux/personality.h>
#include <linux/init.h>
#include <asm/uaccess.h>
#include <asm/pgtable.h>
#include <linux/elf.h>
static int load_som_binary(struct linux_binprm * bprm, struct pt_regs * regs);
static int load_som_library(struct file *);
/*
* If we don't support core dumping, then supply a NULL so we
* don't even try.
*/
#if 0
static int som_core_dump(long signr, struct pt_regs *regs, unsigned long limit);
#else
#define som_core_dump NULL
#endif
#define SOM_PAGESTART(_v) ((_v) & ~(unsigned long)(SOM_PAGESIZE-1))
#define SOM_PAGEOFFSET(_v) ((_v) & (SOM_PAGESIZE-1))
#define SOM_PAGEALIGN(_v) (((_v) + SOM_PAGESIZE - 1) & ~(SOM_PAGESIZE - 1))
static struct linux_binfmt som_format = {
.module = THIS_MODULE,
.load_binary = load_som_binary,
.load_shlib = load_som_library,
.core_dump = som_core_dump,
.min_coredump = SOM_PAGESIZE
};
/*
* create_som_tables() parses the env- and arg-strings in new user
* memory and creates the pointer tables from them, and puts their
* addresses on the "stack", returning the new stack pointer value.
*/
static void create_som_tables(struct linux_binprm *bprm)
{
char **argv, **envp;
int argc = bprm->argc;
int envc = bprm->envc;
unsigned long p;
unsigned long *sp;
/* Word-align the stack pointer */
sp = (unsigned long *)((bprm->p + 3) & ~3);
envp = (char **) sp;
sp += envc + 1;
argv = (char **) sp;
sp += argc + 1;
__put_user((unsigned long) envp,++sp);
__put_user((unsigned long) argv,++sp);
__put_user(argc, ++sp);
bprm->p = (unsigned long) sp;
p = current->mm->arg_start;
while (argc-- > 0) {
__put_user((char *)p,argv++);
p += strlen_user((char *)p);
}
__put_user(NULL, argv);
current->mm->arg_end = current->mm->env_start = p;
while (envc-- > 0) {
__put_user((char *)p,envp++);
p += strlen_user((char *)p);
}
__put_user(NULL, envp);
current->mm->env_end = p;
}
static int check_som_header(struct som_hdr *som_ex)
{
int *buf = (int *)som_ex;
int i, ck;
if (som_ex->system_id != SOM_SID_PARISC_1_0 &&
som_ex->system_id != SOM_SID_PARISC_1_1 &&
som_ex->system_id != SOM_SID_PARISC_2_0)
return -ENOEXEC;
if (som_ex->a_magic != SOM_EXEC_NONSHARE &&
som_ex->a_magic != SOM_EXEC_SHARE &&
som_ex->a_magic != SOM_EXEC_DEMAND)
return -ENOEXEC;
if (som_ex->version_id != SOM_ID_OLD &&
som_ex->version_id != SOM_ID_NEW)
return -ENOEXEC;
ck = 0;
for (i=0; i<32; i++)
ck ^= buf[i];
if (ck != 0)
return -ENOEXEC;
return 0;
}
static int map_som_binary(struct file *file,
const struct som_exec_auxhdr *hpuxhdr)
{
unsigned long code_start, code_size, data_start, data_size;
unsigned long bss_start, som_brk;
int retval;
int prot = PROT_READ | PROT_EXEC;
int flags = MAP_FIXED|MAP_PRIVATE|MAP_DENYWRITE|MAP_EXECUTABLE;
mm_segment_t old_fs = get_fs();
set_fs(get_ds());
code_start = SOM_PAGESTART(hpuxhdr->exec_tmem);
code_size = SOM_PAGEALIGN(hpuxhdr->exec_tsize);
current->mm->start_code = code_start;
current->mm->end_code = code_start + code_size;
down_write(&current->mm->mmap_sem);
retval = do_mmap(file, code_start, code_size, prot,
flags, SOM_PAGESTART(hpuxhdr->exec_tfile));
up_write(&current->mm->mmap_sem);
if (retval < 0 && retval > -1024)
goto out;
data_start = SOM_PAGESTART(hpuxhdr->exec_dmem);
data_size = SOM_PAGEALIGN(hpuxhdr->exec_dsize);
current->mm->start_data = data_start;
current->mm->end_data = bss_start = data_start + data_size;
down_write(&current->mm->mmap_sem);
retval = do_mmap(file, data_start, data_size,
prot | PROT_WRITE, flags,
SOM_PAGESTART(hpuxhdr->exec_dfile));
up_write(&current->mm->mmap_sem);
if (retval < 0 && retval > -1024)
goto out;
som_brk = bss_start + SOM_PAGEALIGN(hpuxhdr->exec_bsize);
current->mm->start_brk = current->mm->brk = som_brk;
down_write(&current->mm->mmap_sem);
retval = do_mmap(NULL, bss_start, som_brk - bss_start,
prot | PROT_WRITE, MAP_FIXED | MAP_PRIVATE, 0);
up_write(&current->mm->mmap_sem);
if (retval > 0 || retval < -1024)
retval = 0;
out:
set_fs(old_fs);
return retval;
}
/*
* These are the functions used to load SOM executables and shared
* libraries. There is no binary dependent code anywhere else.
*/
static int
load_som_binary(struct linux_binprm * bprm, struct pt_regs * regs)
{
int som_exec_fileno;
int retval;
unsigned int size;
unsigned long som_entry;
struct som_hdr *som_ex;
struct som_exec_auxhdr *hpuxhdr;
/* Get the exec-header */
som_ex = (struct som_hdr *) bprm->buf;
retval = check_som_header(som_ex);
if (retval != 0)
goto out;
/* Now read in the auxiliary header information */
retval = -ENOMEM;
size = som_ex->aux_header_size;
if (size > SOM_PAGESIZE)
goto out;
hpuxhdr = kmalloc(size, GFP_KERNEL);
if (!hpuxhdr)
goto out;
retval = kernel_read(bprm->file, som_ex->aux_header_location,
(char *) hpuxhdr, size);
if (retval != size) {
if (retval >= 0)
retval = -EIO;
goto out_free;
}
retval = get_unused_fd();
if (retval < 0)
goto out_free;
get_file(bprm->file);
fd_install(som_exec_fileno = retval, bprm->file);
/* Flush all traces of the currently running executable */
retval = flush_old_exec(bprm);
if (retval)
goto out_free;
/* OK, This is the point of no return */
current->flags &= ~PF_FORKNOEXEC;
current->personality = PER_HPUX;
/* Set the task size for HP-UX processes such that
* the gateway page is outside the address space.
* This can be fixed later, but for now, this is much
* easier.
*/
current->thread.task_size = 0xc0000000;
/* Set map base to allow enough room for hp-ux heap growth */
current->thread.map_base = 0x80000000;
retval = map_som_binary(bprm->file, hpuxhdr);
if (retval < 0)
goto out_free;
som_entry = hpuxhdr->exec_entry;
kfree(hpuxhdr);
set_binfmt(&som_format);
install_exec_creds(bprm);
setup_arg_pages(bprm, STACK_TOP, EXSTACK_DEFAULT);
create_som_tables(bprm);
current->mm->start_stack = bprm->p;
#if 0
printk("(start_brk) %08lx\n" , (unsigned long) current->mm->start_brk);
printk("(end_code) %08lx\n" , (unsigned long) current->mm->end_code);
printk("(start_code) %08lx\n" , (unsigned long) current->mm->start_code);
printk("(end_data) %08lx\n" , (unsigned long) current->mm->end_data);
printk("(start_stack) %08lx\n" , (unsigned long) current->mm->start_stack);
printk("(brk) %08lx\n" , (unsigned long) current->mm->brk);
#endif
map_hpux_gateway_page(current,current->mm);
start_thread_som(regs, som_entry, bprm->p);
return 0;
/* error cleanup */
out_free:
kfree(hpuxhdr);
out:
return retval;
}
static int load_som_library(struct file *f)
{
/* No lib support in SOM yet. gizza chance.. */
return -ENOEXEC;
}
/* Install the SOM loader.
* N.B. We *rely* on the table being the right size with the
* right number of free slots...
*/
static int __init init_som_binfmt(void)
{
return register_binfmt(&som_format);
}
static void __exit exit_som_binfmt(void)
{
/* Remove the SOM loader. */
unregister_binfmt(&som_format);
}
core_initcall(init_som_binfmt);
module_exit(exit_som_binfmt);
MODULE_LICENSE("GPL");