android_kernel_motorola_sm6225/fs/binfmt_elf_fdpic.c

1840 lines
48 KiB
C
Raw Normal View History

/* binfmt_elf_fdpic.c: FDPIC ELF binary format
*
* Copyright (C) 2003, 2004, 2006 Red Hat, Inc. All Rights Reserved.
* Written by David Howells (dhowells@redhat.com)
* Derived from binfmt_elf.c
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version
* 2 of the License, or (at your option) any later version.
*/
#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/string.h>
#include <linux/file.h>
#include <linux/fcntl.h>
#include <linux/slab.h>
#include <linux/pagemap.h>
#include <linux/security.h>
#include <linux/highmem.h>
#include <linux/highuid.h>
#include <linux/personality.h>
#include <linux/ptrace.h>
#include <linux/init.h>
#include <linux/elf.h>
#include <linux/elf-fdpic.h>
#include <linux/elfcore.h>
#include <asm/uaccess.h>
#include <asm/param.h>
#include <asm/pgalloc.h>
typedef char *elf_caddr_t;
#if 0
#define kdebug(fmt, ...) printk("FDPIC "fmt"\n" ,##__VA_ARGS__ )
#else
#define kdebug(fmt, ...) do {} while(0)
#endif
#if 0
#define kdcore(fmt, ...) printk("FDPIC "fmt"\n" ,##__VA_ARGS__ )
#else
#define kdcore(fmt, ...) do {} while(0)
#endif
MODULE_LICENSE("GPL");
static int load_elf_fdpic_binary(struct linux_binprm *, struct pt_regs *);
static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *, struct file *);
static int elf_fdpic_map_file(struct elf_fdpic_params *, struct file *,
struct mm_struct *, const char *);
static int create_elf_fdpic_tables(struct linux_binprm *, struct mm_struct *,
struct elf_fdpic_params *,
struct elf_fdpic_params *);
#ifndef CONFIG_MMU
static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *,
unsigned long *);
static int elf_fdpic_map_file_constdisp_on_uclinux(struct elf_fdpic_params *,
struct file *,
struct mm_struct *);
#endif
static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *,
struct file *, struct mm_struct *);
#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE)
core_pattern: ignore RLIMIT_CORE if core_pattern is a pipe For some time /proc/sys/kernel/core_pattern has been able to set its output destination as a pipe, allowing a user space helper to receive and intellegently process a core. This infrastructure however has some shortcommings which can be enhanced. Specifically: 1) The coredump code in the kernel should ignore RLIMIT_CORE limitation when core_pattern is a pipe, since file system resources are not being consumed in this case, unless the user application wishes to save the core, at which point the app is restricted by usual file system limits and restrictions. 2) The core_pattern code should be able to parse and pass options to the user space helper as an argv array. The real core limit of the uid of the crashing proces should also be passable to the user space helper (since it is overridden to zero when called). 3) Some miscellaneous bugs need to be cleaned up (specifically the recognition of a recursive core dump, should the user mode helper itself crash. Also, the core dump code in the kernel should not wait for the user mode helper to exit, since the same context is responsible for writing to the pipe, and a read of the pipe by the user mode helper will result in a deadlock. This patch: Remove the check of RLIMIT_CORE if core_pattern is a pipe. In the event that core_pattern is a pipe, the entire core will be fed to the user mode helper. Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Cc: <martin.pitt@ubuntu.com> Cc: <wwoods@redhat.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 08:26:34 +02:00
static int elf_fdpic_core_dump(long, struct pt_regs *, struct file *, unsigned long limit);
#endif
static struct linux_binfmt elf_fdpic_format = {
.module = THIS_MODULE,
.load_binary = load_elf_fdpic_binary,
#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE)
.core_dump = elf_fdpic_core_dump,
#endif
.min_coredump = ELF_EXEC_PAGESIZE,
};
static int __init init_elf_fdpic_binfmt(void)
{
return register_binfmt(&elf_fdpic_format);
}
static void __exit exit_elf_fdpic_binfmt(void)
{
unregister_binfmt(&elf_fdpic_format);
}
core_initcall(init_elf_fdpic_binfmt);
module_exit(exit_elf_fdpic_binfmt);
static int is_elf_fdpic(struct elfhdr *hdr, struct file *file)
{
if (memcmp(hdr->e_ident, ELFMAG, SELFMAG) != 0)
return 0;
if (hdr->e_type != ET_EXEC && hdr->e_type != ET_DYN)
return 0;
if (!elf_check_arch(hdr) || !elf_check_fdpic(hdr))
return 0;
if (!file->f_op || !file->f_op->mmap)
return 0;
return 1;
}
/*****************************************************************************/
/*
* read the program headers table into memory
*/
static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params,
struct file *file)
{
struct elf32_phdr *phdr;
unsigned long size;
int retval, loop;
if (params->hdr.e_phentsize != sizeof(struct elf_phdr))
return -ENOMEM;
if (params->hdr.e_phnum > 65536U / sizeof(struct elf_phdr))
return -ENOMEM;
size = params->hdr.e_phnum * sizeof(struct elf_phdr);
params->phdrs = kmalloc(size, GFP_KERNEL);
if (!params->phdrs)
return -ENOMEM;
retval = kernel_read(file, params->hdr.e_phoff,
(char *) params->phdrs, size);
if (unlikely(retval != size))
return retval < 0 ? retval : -ENOEXEC;
/* determine stack size for this binary */
phdr = params->phdrs;
for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) {
if (phdr->p_type != PT_GNU_STACK)
continue;
if (phdr->p_flags & PF_X)
params->flags |= ELF_FDPIC_FLAG_EXEC_STACK;
else
params->flags |= ELF_FDPIC_FLAG_NOEXEC_STACK;
params->stack_size = phdr->p_memsz;
break;
}
return 0;
}
/*****************************************************************************/
/*
* load an fdpic binary into various bits of memory
*/
static int load_elf_fdpic_binary(struct linux_binprm *bprm,
struct pt_regs *regs)
{
struct elf_fdpic_params exec_params, interp_params;
struct elf_phdr *phdr;
unsigned long stack_size, entryaddr;
#ifdef ELF_FDPIC_PLAT_INIT
unsigned long dynaddr;
#endif
struct file *interpreter = NULL; /* to shut gcc up */
char *interpreter_name = NULL;
int executable_stack;
int retval, i;
[PATCH] FDPIC: fix the /proc/pid/stat representation of executable boundaries Fix the /proc/pid/stat representation of executable boundaries. It should show the bounds of the executable, but instead shows the bounds of the loader. Before the patch is applied, the bug can be seen by examining, say, inetd: # ps | grep inetd 610 root 0 S /usr/sbin/inetd -i # cat /proc/610/maps c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so c3180000-c31dede4 r-xs 00000000 00:0b 14582179 /lib/libuClibc-0.9.28.so c328c000-c328ea00 rw-p 00008000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so c3290000-c329b6c0 rw-p 00000000 00:00 0 c32a0000-c32c0000 rwxp 00000000 00:00 0 c32d4000-c32d8000 rw-p 00000000 00:00 0 c3394000-c3398000 rw-p 00000000 00:00 0 c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd c3470000-c34748f8 rw-p 00004000 00:0b 16384612 /usr/sbin/inetd c34cc000-c34d0000 rw-p 00000000 00:00 0 c34d4000-c34d8000 rw-p 00000000 00:00 0 c34d8000-c34dc000 rw-p 00000000 00:00 0 # cat /proc/610/stat 610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 8 0 0 19 0 1 0 94392000718 950272 0 4294967295 3233480704 3233523592 3274440352 3274439976 3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0 The code boundaries are 3233480704 to 3233523592, which are: (gdb) p/x 3233480704 $1 = 0xc0bb0000 (gdb) p/x 3233523592 $2 = 0xc0bba788 Which corresponds to this line in the maps file: c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so Which is wrong. After the patch is applied, the maps file is pretty much identical (there's some minor shuffling of the location of some of the anonymous VMAs), but the stat file is now: # cat /proc/610/stat 610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 7 0 0 18 0 1 0 94392000722 950272 0 4294967295 3276111872 3276141668 3274440352 3274439976 3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0 The code boundaries are then 3276111872 to 3276141668, which are: (gdb) p/x 3276111872 $1 = 0xc3458000 (gdb) p/x 3276141668 $2 = 0xc345f464 And these correspond to this line in the maps file instead: c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd Which is now correct. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-03-23 08:10:00 +01:00
kdebug("____ LOAD %d ____", current->pid);
memset(&exec_params, 0, sizeof(exec_params));
memset(&interp_params, 0, sizeof(interp_params));
exec_params.hdr = *(struct elfhdr *) bprm->buf;
exec_params.flags = ELF_FDPIC_FLAG_PRESENT | ELF_FDPIC_FLAG_EXECUTABLE;
/* check that this is a binary we know how to deal with */
retval = -ENOEXEC;
if (!is_elf_fdpic(&exec_params.hdr, bprm->file))
goto error;
/* read the program header table */
retval = elf_fdpic_fetch_phdrs(&exec_params, bprm->file);
if (retval < 0)
goto error;
/* scan for a program header that specifies an interpreter */
phdr = exec_params.phdrs;
for (i = 0; i < exec_params.hdr.e_phnum; i++, phdr++) {
switch (phdr->p_type) {
case PT_INTERP:
retval = -ENOMEM;
if (phdr->p_filesz > PATH_MAX)
goto error;
retval = -ENOENT;
if (phdr->p_filesz < 2)
goto error;
/* read the name of the interpreter into memory */
interpreter_name = kmalloc(phdr->p_filesz, GFP_KERNEL);
if (!interpreter_name)
goto error;
retval = kernel_read(bprm->file,
phdr->p_offset,
interpreter_name,
phdr->p_filesz);
if (unlikely(retval != phdr->p_filesz)) {
if (retval >= 0)
retval = -ENOEXEC;
goto error;
}
retval = -ENOENT;
if (interpreter_name[phdr->p_filesz - 1] != '\0')
goto error;
kdebug("Using ELF interpreter %s", interpreter_name);
/* replace the program with the interpreter */
interpreter = open_exec(interpreter_name);
retval = PTR_ERR(interpreter);
if (IS_ERR(interpreter)) {
interpreter = NULL;
goto error;
}
/*
* If the binary is not readable then enforce
* mm->dumpable = 0 regardless of the interpreter's
* permissions.
*/
if (file_permission(interpreter, MAY_READ) < 0)
bprm->interp_flags |= BINPRM_FLAGS_ENFORCE_NONDUMP;
retval = kernel_read(interpreter, 0, bprm->buf,
BINPRM_BUF_SIZE);
if (unlikely(retval != BINPRM_BUF_SIZE)) {
if (retval >= 0)
retval = -ENOEXEC;
goto error;
}
interp_params.hdr = *((struct elfhdr *) bprm->buf);
break;
case PT_LOAD:
#ifdef CONFIG_MMU
if (exec_params.load_addr == 0)
exec_params.load_addr = phdr->p_vaddr;
#endif
break;
}
}
if (elf_check_const_displacement(&exec_params.hdr))
exec_params.flags |= ELF_FDPIC_FLAG_CONSTDISP;
/* perform insanity checks on the interpreter */
if (interpreter_name) {
retval = -ELIBBAD;
if (!is_elf_fdpic(&interp_params.hdr, interpreter))
goto error;
interp_params.flags = ELF_FDPIC_FLAG_PRESENT;
/* read the interpreter's program header table */
retval = elf_fdpic_fetch_phdrs(&interp_params, interpreter);
if (retval < 0)
goto error;
}
stack_size = exec_params.stack_size;
if (stack_size < interp_params.stack_size)
stack_size = interp_params.stack_size;
if (exec_params.flags & ELF_FDPIC_FLAG_EXEC_STACK)
executable_stack = EXSTACK_ENABLE_X;
else if (exec_params.flags & ELF_FDPIC_FLAG_NOEXEC_STACK)
executable_stack = EXSTACK_DISABLE_X;
else if (interp_params.flags & ELF_FDPIC_FLAG_EXEC_STACK)
executable_stack = EXSTACK_ENABLE_X;
else if (interp_params.flags & ELF_FDPIC_FLAG_NOEXEC_STACK)
executable_stack = EXSTACK_DISABLE_X;
else
executable_stack = EXSTACK_DEFAULT;
retval = -ENOEXEC;
if (stack_size == 0)
goto error;
if (elf_check_const_displacement(&interp_params.hdr))
interp_params.flags |= ELF_FDPIC_FLAG_CONSTDISP;
/* flush all traces of the currently running executable */
retval = flush_old_exec(bprm);
if (retval)
goto error;
/* there's now no turning back... the old userspace image is dead,
* defunct, deceased, etc. after this point we have to exit via
* error_kill */
set_personality(PER_LINUX_FDPIC);
set_binfmt(&elf_fdpic_format);
current->mm->start_code = 0;
current->mm->end_code = 0;
current->mm->start_stack = 0;
current->mm->start_data = 0;
current->mm->end_data = 0;
current->mm->context.exec_fdpic_loadmap = 0;
current->mm->context.interp_fdpic_loadmap = 0;
current->flags &= ~PF_FORKNOEXEC;
#ifdef CONFIG_MMU
elf_fdpic_arch_lay_out_mm(&exec_params,
&interp_params,
&current->mm->start_stack,
&current->mm->start_brk);
retval = setup_arg_pages(bprm, current->mm->start_stack,
executable_stack);
if (retval < 0) {
send_sig(SIGKILL, current, 0);
goto error_kill;
}
#endif
/* load the executable and interpreter into memory */
retval = elf_fdpic_map_file(&exec_params, bprm->file, current->mm,
"executable");
if (retval < 0)
goto error_kill;
if (interpreter_name) {
retval = elf_fdpic_map_file(&interp_params, interpreter,
current->mm, "interpreter");
if (retval < 0) {
printk(KERN_ERR "Unable to load interpreter\n");
goto error_kill;
}
allow_write_access(interpreter);
fput(interpreter);
interpreter = NULL;
}
#ifdef CONFIG_MMU
if (!current->mm->start_brk)
current->mm->start_brk = current->mm->end_data;
current->mm->brk = current->mm->start_brk =
PAGE_ALIGN(current->mm->start_brk);
#else
/* create a stack and brk area big enough for everyone
* - the brk heap starts at the bottom and works up
* - the stack starts at the top and works down
*/
stack_size = (stack_size + PAGE_SIZE - 1) & PAGE_MASK;
if (stack_size < PAGE_SIZE * 2)
stack_size = PAGE_SIZE * 2;
down_write(&current->mm->mmap_sem);
current->mm->start_brk = do_mmap(NULL, 0, stack_size,
PROT_READ | PROT_WRITE | PROT_EXEC,
MAP_PRIVATE | MAP_ANONYMOUS | MAP_GROWSDOWN,
0);
if (IS_ERR_VALUE(current->mm->start_brk)) {
up_write(&current->mm->mmap_sem);
retval = current->mm->start_brk;
current->mm->start_brk = 0;
goto error_kill;
}
up_write(&current->mm->mmap_sem);
current->mm->brk = current->mm->start_brk;
current->mm->context.end_brk = current->mm->start_brk;
current->mm->context.end_brk +=
(stack_size > PAGE_SIZE) ? (stack_size - PAGE_SIZE) : 0;
current->mm->start_stack = current->mm->start_brk + stack_size;
#endif
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 00:39:24 +01:00
install_exec_creds(bprm);
current->flags &= ~PF_FORKNOEXEC;
if (create_elf_fdpic_tables(bprm, current->mm,
&exec_params, &interp_params) < 0)
goto error_kill;
kdebug("- start_code %lx", current->mm->start_code);
kdebug("- end_code %lx", current->mm->end_code);
kdebug("- start_data %lx", current->mm->start_data);
kdebug("- end_data %lx", current->mm->end_data);
kdebug("- start_brk %lx", current->mm->start_brk);
kdebug("- brk %lx", current->mm->brk);
kdebug("- start_stack %lx", current->mm->start_stack);
#ifdef ELF_FDPIC_PLAT_INIT
/*
* The ABI may specify that certain registers be set up in special
* ways (on i386 %edx is the address of a DT_FINI function, for
* example. This macro performs whatever initialization to
* the regs structure is required.
*/
dynaddr = interp_params.dynamic_addr ?: exec_params.dynamic_addr;
ELF_FDPIC_PLAT_INIT(regs, exec_params.map_addr, interp_params.map_addr,
dynaddr);
#endif
/* everything is now ready... get the userspace context ready to roll */
entryaddr = interp_params.entry_addr ?: exec_params.entry_addr;
start_thread(regs, entryaddr, current->mm->start_stack);
retval = 0;
error:
if (interpreter) {
allow_write_access(interpreter);
fput(interpreter);
}
kfree(interpreter_name);
kfree(exec_params.phdrs);
kfree(exec_params.loadmap);
kfree(interp_params.phdrs);
kfree(interp_params.loadmap);
return retval;
/* unrecoverable error - kill the process */
error_kill:
send_sig(SIGSEGV, current, 0);
goto error;
}
/*****************************************************************************/
#ifndef ELF_BASE_PLATFORM
/*
* AT_BASE_PLATFORM indicates the "real" hardware/microarchitecture.
* If the arch defines ELF_BASE_PLATFORM (in asm/elf.h), the value
* will be copied to the user stack in the same manner as AT_PLATFORM.
*/
#define ELF_BASE_PLATFORM NULL
#endif
/*
* present useful information to the program by shovelling it onto the new
* process's stack
*/
static int create_elf_fdpic_tables(struct linux_binprm *bprm,
struct mm_struct *mm,
struct elf_fdpic_params *exec_params,
struct elf_fdpic_params *interp_params)
{
const struct cred *cred = current_cred();
unsigned long sp, csp, nitems;
elf_caddr_t __user *argv, *envp;
size_t platform_len = 0, len;
char *k_platform, *k_base_platform;
char __user *u_platform, *u_base_platform, *p;
long hwcap;
int loop;
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
int nr; /* reset for each csp adjustment */
#ifdef CONFIG_MMU
/* In some cases (e.g. Hyper-Threading), we want to avoid L1 evictions
* by the processes running on the same package. One thing we can do is
* to shuffle the initial stack for them, so we give the architecture
* an opportunity to do so here.
*/
sp = arch_align_stack(bprm->p);
#else
sp = mm->start_stack;
/* stack the program arguments and environment */
if (elf_fdpic_transfer_args_to_stack(bprm, &sp) < 0)
return -EFAULT;
#endif
hwcap = ELF_HWCAP;
/*
* If this architecture has a platform capability string, copy it
* to userspace. In some cases (Sparc), this info is impossible
* for userspace to get any other way, in others (i386) it is
* merely difficult.
*/
k_platform = ELF_PLATFORM;
u_platform = NULL;
if (k_platform) {
platform_len = strlen(k_platform) + 1;
sp -= platform_len;
u_platform = (char __user *) sp;
if (__copy_to_user(u_platform, k_platform, platform_len) != 0)
return -EFAULT;
}
/*
* If this architecture has a "base" platform capability
* string, copy it to userspace.
*/
k_base_platform = ELF_BASE_PLATFORM;
u_base_platform = NULL;
if (k_base_platform) {
platform_len = strlen(k_base_platform) + 1;
sp -= platform_len;
u_base_platform = (char __user *) sp;
if (__copy_to_user(u_base_platform, k_base_platform, platform_len) != 0)
return -EFAULT;
}
sp &= ~7UL;
/* stack the load map(s) */
len = sizeof(struct elf32_fdpic_loadmap);
len += sizeof(struct elf32_fdpic_loadseg) * exec_params->loadmap->nsegs;
sp = (sp - len) & ~7UL;
exec_params->map_addr = sp;
if (copy_to_user((void __user *) sp, exec_params->loadmap, len) != 0)
return -EFAULT;
current->mm->context.exec_fdpic_loadmap = (unsigned long) sp;
if (interp_params->loadmap) {
len = sizeof(struct elf32_fdpic_loadmap);
len += sizeof(struct elf32_fdpic_loadseg) *
interp_params->loadmap->nsegs;
sp = (sp - len) & ~7UL;
interp_params->map_addr = sp;
if (copy_to_user((void __user *) sp, interp_params->loadmap,
len) != 0)
return -EFAULT;
current->mm->context.interp_fdpic_loadmap = (unsigned long) sp;
}
/* force 16 byte _final_ alignment here for generality */
#define DLINFO_ITEMS 15
nitems = 1 + DLINFO_ITEMS + (k_platform ? 1 : 0) +
(k_base_platform ? 1 : 0) + AT_VECTOR_SIZE_ARCH;
if (bprm->interp_flags & BINPRM_FLAGS_EXECFD)
nitems++;
csp = sp;
sp -= nitems * 2 * sizeof(unsigned long);
sp -= (bprm->envc + 1) * sizeof(char *); /* envv[] */
sp -= (bprm->argc + 1) * sizeof(char *); /* argv[] */
sp -= 1 * sizeof(unsigned long); /* argc */
csp -= sp & 15UL;
sp -= sp & 15UL;
/* put the ELF interpreter info on the stack */
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
#define NEW_AUX_ENT(id, val) \
do { \
struct { unsigned long _id, _val; } __user *ent; \
\
ent = (void __user *) csp; \
__put_user((id), &ent[nr]._id); \
__put_user((val), &ent[nr]._val); \
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
nr++; \
} while (0)
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
nr = 0;
csp -= 2 * sizeof(unsigned long);
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
NEW_AUX_ENT(AT_NULL, 0);
if (k_platform) {
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
nr = 0;
csp -= 2 * sizeof(unsigned long);
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
NEW_AUX_ENT(AT_PLATFORM,
(elf_addr_t) (unsigned long) u_platform);
}
if (k_base_platform) {
nr = 0;
csp -= 2 * sizeof(unsigned long);
NEW_AUX_ENT(AT_BASE_PLATFORM,
(elf_addr_t) (unsigned long) u_base_platform);
}
if (bprm->interp_flags & BINPRM_FLAGS_EXECFD) {
nr = 0;
csp -= 2 * sizeof(unsigned long);
NEW_AUX_ENT(AT_EXECFD, bprm->interp_data);
}
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
nr = 0;
csp -= DLINFO_ITEMS * 2 * sizeof(unsigned long);
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
NEW_AUX_ENT(AT_HWCAP, hwcap);
NEW_AUX_ENT(AT_PAGESZ, PAGE_SIZE);
NEW_AUX_ENT(AT_CLKTCK, CLOCKS_PER_SEC);
NEW_AUX_ENT(AT_PHDR, exec_params->ph_addr);
NEW_AUX_ENT(AT_PHENT, sizeof(struct elf_phdr));
NEW_AUX_ENT(AT_PHNUM, exec_params->hdr.e_phnum);
NEW_AUX_ENT(AT_BASE, interp_params->elfhdr_addr);
NEW_AUX_ENT(AT_FLAGS, 0);
NEW_AUX_ENT(AT_ENTRY, exec_params->entry_addr);
NEW_AUX_ENT(AT_UID, (elf_addr_t) cred->uid);
NEW_AUX_ENT(AT_EUID, (elf_addr_t) cred->euid);
NEW_AUX_ENT(AT_GID, (elf_addr_t) cred->gid);
NEW_AUX_ENT(AT_EGID, (elf_addr_t) cred->egid);
NEW_AUX_ENT(AT_SECURE, security_bprm_secureexec(bprm));
NEW_AUX_ENT(AT_EXECFN, bprm->exec);
#ifdef ARCH_DLINFO
binfmt_elf_fdpic: Magical stack pointer index, for NEW_AUX_ENT compat. While implementing binfmt_elf_fdpic on SH it quickly became apparent that SH was the first platform to support both binfmt_elf_fdpic and binfmt_elf, as well as the only of the FDPIC platforms to make use of the auxvt. Currently binfmt_elf_fdpic uses a special version of NEW_AUX_ENT() where the first argument is the entry displacement after csp has been adjusted, being reset after each adjustment. As we have no ability to sort this out through the platform's ARCH_DLINFO, this index needs to be managed entirely in create_elf_fdpic_tables(). Presently none of the platforms that set their own auxvt entries are able to do so through their respective ARCH_DLINFOs when using binfmt_elf_fdpic. In addition to this, binfmt_elf_fdpic has been looking at DLINFO_ARCH_ITEMS for the number of architecture-specific entries in the auxvt. This is legacy cruft, and is not defined by any platforms in-tree, even those that make heavy use of the auxvt. AT_VECTOR_SIZE_ARCH is always available, and contains the number that is of interest here, so we switch to using that unconditionally as well. As this has direct bearing on how much stack is used, platforms that have configurable (or dynamically adjustable) NEW_AUX_ENT calls need to either make AT_VECTOR_SIZE_ARCH more fine-grained, or leave it as a worst-case and live with some lost stack space if those entries aren't pushed (some platforms may also need to purposely sacrifice some space here for alignment considerations, as noted in the code -- although not an issue for any FDPIC-capable platform today). Signed-off-by: Paul Mundt <lethal@linux-sh.org> Acked-by: David Howells <dhowells@redhat.com>
2008-05-19 06:34:45 +02:00
nr = 0;
csp -= AT_VECTOR_SIZE_ARCH * 2 * sizeof(unsigned long);
/* ARCH_DLINFO must come last so platform specific code can enforce
* special alignment requirements on the AUXV if necessary (eg. PPC).
*/
ARCH_DLINFO;
#endif
#undef NEW_AUX_ENT
/* allocate room for argv[] and envv[] */
csp -= (bprm->envc + 1) * sizeof(elf_caddr_t);
envp = (elf_caddr_t __user *) csp;
csp -= (bprm->argc + 1) * sizeof(elf_caddr_t);
argv = (elf_caddr_t __user *) csp;
/* stack argc */
csp -= sizeof(unsigned long);
__put_user(bprm->argc, (unsigned long __user *) csp);
BUG_ON(csp != sp);
/* fill in the argv[] array */
#ifdef CONFIG_MMU
current->mm->arg_start = bprm->p;
#else
current->mm->arg_start = current->mm->start_stack -
(MAX_ARG_PAGES * PAGE_SIZE - bprm->p);
#endif
p = (char __user *) current->mm->arg_start;
for (loop = bprm->argc; loop > 0; loop--) {
__put_user((elf_caddr_t) p, argv++);
len = strnlen_user(p, MAX_ARG_STRLEN);
if (!len || len > MAX_ARG_STRLEN)
return -EINVAL;
p += len;
}
__put_user(NULL, argv);
current->mm->arg_end = (unsigned long) p;
/* fill in the envv[] array */
current->mm->env_start = (unsigned long) p;
for (loop = bprm->envc; loop > 0; loop--) {
__put_user((elf_caddr_t)(unsigned long) p, envp++);
len = strnlen_user(p, MAX_ARG_STRLEN);
if (!len || len > MAX_ARG_STRLEN)
return -EINVAL;
p += len;
}
__put_user(NULL, envp);
current->mm->env_end = (unsigned long) p;
mm->start_stack = (unsigned long) sp;
return 0;
}
/*****************************************************************************/
/*
* transfer the program arguments and environment from the holding pages onto
* the stack
*/
#ifndef CONFIG_MMU
static int elf_fdpic_transfer_args_to_stack(struct linux_binprm *bprm,
unsigned long *_sp)
{
unsigned long index, stop, sp;
char *src;
int ret = 0;
stop = bprm->p >> PAGE_SHIFT;
sp = *_sp;
for (index = MAX_ARG_PAGES - 1; index >= stop; index--) {
src = kmap(bprm->page[index]);
sp -= PAGE_SIZE;
if (copy_to_user((void *) sp, src, PAGE_SIZE) != 0)
ret = -EFAULT;
kunmap(bprm->page[index]);
if (ret < 0)
goto out;
}
*_sp = (*_sp - (MAX_ARG_PAGES * PAGE_SIZE - bprm->p)) & ~15;
out:
return ret;
}
#endif
/*****************************************************************************/
/*
* load the appropriate binary image (executable or interpreter) into memory
* - we assume no MMU is available
* - if no other PIC bits are set in params->hdr->e_flags
* - we assume that the LOADable segments in the binary are independently relocatable
* - we assume R/O executable segments are shareable
* - else
* - we assume the loadable parts of the image to require fixed displacement
* - the image is not shareable
*/
static int elf_fdpic_map_file(struct elf_fdpic_params *params,
struct file *file,
struct mm_struct *mm,
const char *what)
{
struct elf32_fdpic_loadmap *loadmap;
#ifdef CONFIG_MMU
struct elf32_fdpic_loadseg *mseg;
#endif
struct elf32_fdpic_loadseg *seg;
struct elf32_phdr *phdr;
unsigned long load_addr, stop;
unsigned nloads, tmp;
size_t size;
int loop, ret;
/* allocate a load map table */
nloads = 0;
for (loop = 0; loop < params->hdr.e_phnum; loop++)
if (params->phdrs[loop].p_type == PT_LOAD)
nloads++;
if (nloads == 0)
return -ELIBBAD;
size = sizeof(*loadmap) + nloads * sizeof(*seg);
loadmap = kzalloc(size, GFP_KERNEL);
if (!loadmap)
return -ENOMEM;
params->loadmap = loadmap;
loadmap->version = ELF32_FDPIC_LOADMAP_VERSION;
loadmap->nsegs = nloads;
load_addr = params->load_addr;
seg = loadmap->segs;
/* map the requested LOADs into the memory space */
switch (params->flags & ELF_FDPIC_FLAG_ARRANGEMENT) {
case ELF_FDPIC_FLAG_CONSTDISP:
case ELF_FDPIC_FLAG_CONTIGUOUS:
#ifndef CONFIG_MMU
ret = elf_fdpic_map_file_constdisp_on_uclinux(params, file, mm);
if (ret < 0)
return ret;
break;
#endif
default:
ret = elf_fdpic_map_file_by_direct_mmap(params, file, mm);
if (ret < 0)
return ret;
break;
}
/* map the entry point */
if (params->hdr.e_entry) {
seg = loadmap->segs;
for (loop = loadmap->nsegs; loop > 0; loop--, seg++) {
if (params->hdr.e_entry >= seg->p_vaddr &&
params->hdr.e_entry < seg->p_vaddr + seg->p_memsz) {
params->entry_addr =
(params->hdr.e_entry - seg->p_vaddr) +
seg->addr;
break;
}
}
}
/* determine where the program header table has wound up if mapped */
stop = params->hdr.e_phoff;
stop += params->hdr.e_phnum * sizeof (struct elf_phdr);
phdr = params->phdrs;
for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) {
if (phdr->p_type != PT_LOAD)
continue;
if (phdr->p_offset > params->hdr.e_phoff ||
phdr->p_offset + phdr->p_filesz < stop)
continue;
seg = loadmap->segs;
for (loop = loadmap->nsegs; loop > 0; loop--, seg++) {
if (phdr->p_vaddr >= seg->p_vaddr &&
phdr->p_vaddr + phdr->p_filesz <=
seg->p_vaddr + seg->p_memsz) {
params->ph_addr =
(phdr->p_vaddr - seg->p_vaddr) +
seg->addr +
params->hdr.e_phoff - phdr->p_offset;
break;
}
}
break;
}
/* determine where the dynamic section has wound up if there is one */
phdr = params->phdrs;
for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) {
if (phdr->p_type != PT_DYNAMIC)
continue;
seg = loadmap->segs;
for (loop = loadmap->nsegs; loop > 0; loop--, seg++) {
if (phdr->p_vaddr >= seg->p_vaddr &&
phdr->p_vaddr + phdr->p_memsz <=
seg->p_vaddr + seg->p_memsz) {
params->dynamic_addr =
(phdr->p_vaddr - seg->p_vaddr) +
seg->addr;
/* check the dynamic section contains at least
* one item, and that the last item is a NULL
* entry */
if (phdr->p_memsz == 0 ||
phdr->p_memsz % sizeof(Elf32_Dyn) != 0)
goto dynamic_error;
tmp = phdr->p_memsz / sizeof(Elf32_Dyn);
if (((Elf32_Dyn *)
params->dynamic_addr)[tmp - 1].d_tag != 0)
goto dynamic_error;
break;
}
}
break;
}
/* now elide adjacent segments in the load map on MMU linux
* - on uClinux the holes between may actually be filled with system
* stuff or stuff from other processes
*/
#ifdef CONFIG_MMU
nloads = loadmap->nsegs;
mseg = loadmap->segs;
seg = mseg + 1;
for (loop = 1; loop < nloads; loop++) {
/* see if we have a candidate for merging */
if (seg->p_vaddr - mseg->p_vaddr == seg->addr - mseg->addr) {
load_addr = PAGE_ALIGN(mseg->addr + mseg->p_memsz);
if (load_addr == (seg->addr & PAGE_MASK)) {
mseg->p_memsz +=
load_addr -
(mseg->addr + mseg->p_memsz);
mseg->p_memsz += seg->addr & ~PAGE_MASK;
mseg->p_memsz += seg->p_memsz;
loadmap->nsegs--;
continue;
}
}
mseg++;
if (mseg != seg)
*mseg = *seg;
}
#endif
kdebug("Mapped Object [%s]:", what);
kdebug("- elfhdr : %lx", params->elfhdr_addr);
kdebug("- entry : %lx", params->entry_addr);
kdebug("- PHDR[] : %lx", params->ph_addr);
kdebug("- DYNAMIC[]: %lx", params->dynamic_addr);
seg = loadmap->segs;
for (loop = 0; loop < loadmap->nsegs; loop++, seg++)
kdebug("- LOAD[%d] : %08x-%08x [va=%x ms=%x]",
loop,
seg->addr, seg->addr + seg->p_memsz - 1,
seg->p_vaddr, seg->p_memsz);
return 0;
dynamic_error:
printk("ELF FDPIC %s with invalid DYNAMIC section (inode=%lu)\n",
what, file->f_path.dentry->d_inode->i_ino);
return -ELIBBAD;
}
/*****************************************************************************/
/*
* map a file with constant displacement under uClinux
*/
#ifndef CONFIG_MMU
static int elf_fdpic_map_file_constdisp_on_uclinux(
struct elf_fdpic_params *params,
struct file *file,
struct mm_struct *mm)
{
struct elf32_fdpic_loadseg *seg;
struct elf32_phdr *phdr;
unsigned long load_addr, base = ULONG_MAX, top = 0, maddr = 0, mflags;
loff_t fpos;
int loop, ret;
load_addr = params->load_addr;
seg = params->loadmap->segs;
/* determine the bounds of the contiguous overall allocation we must
* make */
phdr = params->phdrs;
for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) {
if (params->phdrs[loop].p_type != PT_LOAD)
continue;
if (base > phdr->p_vaddr)
base = phdr->p_vaddr;
if (top < phdr->p_vaddr + phdr->p_memsz)
top = phdr->p_vaddr + phdr->p_memsz;
}
/* allocate one big anon block for everything */
mflags = MAP_PRIVATE;
if (params->flags & ELF_FDPIC_FLAG_EXECUTABLE)
mflags |= MAP_EXECUTABLE;
down_write(&mm->mmap_sem);
maddr = do_mmap(NULL, load_addr, top - base,
PROT_READ | PROT_WRITE | PROT_EXEC, mflags, 0);
up_write(&mm->mmap_sem);
if (IS_ERR_VALUE(maddr))
return (int) maddr;
if (load_addr != 0)
load_addr += PAGE_ALIGN(top - base);
/* and then load the file segments into it */
phdr = params->phdrs;
for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) {
if (params->phdrs[loop].p_type != PT_LOAD)
continue;
fpos = phdr->p_offset;
seg->addr = maddr + (phdr->p_vaddr - base);
seg->p_vaddr = phdr->p_vaddr;
seg->p_memsz = phdr->p_memsz;
ret = file->f_op->read(file, (void *) seg->addr,
phdr->p_filesz, &fpos);
if (ret < 0)
return ret;
/* map the ELF header address if in this segment */
if (phdr->p_offset == 0)
params->elfhdr_addr = seg->addr;
/* clear any space allocated but not loaded */
if (phdr->p_filesz < phdr->p_memsz) {
ret = clear_user((void *) (seg->addr + phdr->p_filesz),
phdr->p_memsz - phdr->p_filesz);
if (ret)
return ret;
}
if (mm) {
if (phdr->p_flags & PF_X) {
[PATCH] FDPIC: fix the /proc/pid/stat representation of executable boundaries Fix the /proc/pid/stat representation of executable boundaries. It should show the bounds of the executable, but instead shows the bounds of the loader. Before the patch is applied, the bug can be seen by examining, say, inetd: # ps | grep inetd 610 root 0 S /usr/sbin/inetd -i # cat /proc/610/maps c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so c3180000-c31dede4 r-xs 00000000 00:0b 14582179 /lib/libuClibc-0.9.28.so c328c000-c328ea00 rw-p 00008000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so c3290000-c329b6c0 rw-p 00000000 00:00 0 c32a0000-c32c0000 rwxp 00000000 00:00 0 c32d4000-c32d8000 rw-p 00000000 00:00 0 c3394000-c3398000 rw-p 00000000 00:00 0 c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd c3470000-c34748f8 rw-p 00004000 00:0b 16384612 /usr/sbin/inetd c34cc000-c34d0000 rw-p 00000000 00:00 0 c34d4000-c34d8000 rw-p 00000000 00:00 0 c34d8000-c34dc000 rw-p 00000000 00:00 0 # cat /proc/610/stat 610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 8 0 0 19 0 1 0 94392000718 950272 0 4294967295 3233480704 3233523592 3274440352 3274439976 3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0 The code boundaries are 3233480704 to 3233523592, which are: (gdb) p/x 3233480704 $1 = 0xc0bb0000 (gdb) p/x 3233523592 $2 = 0xc0bba788 Which corresponds to this line in the maps file: c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so Which is wrong. After the patch is applied, the maps file is pretty much identical (there's some minor shuffling of the location of some of the anonymous VMAs), but the stat file is now: # cat /proc/610/stat 610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 7 0 0 18 0 1 0 94392000722 950272 0 4294967295 3276111872 3276141668 3274440352 3274439976 3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0 The code boundaries are then 3276111872 to 3276141668, which are: (gdb) p/x 3276111872 $1 = 0xc3458000 (gdb) p/x 3276141668 $2 = 0xc345f464 And these correspond to this line in the maps file instead: c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd Which is now correct. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-03-23 08:10:00 +01:00
if (!mm->start_code) {
mm->start_code = seg->addr;
mm->end_code = seg->addr +
phdr->p_memsz;
}
} else if (!mm->start_data) {
mm->start_data = seg->addr;
#ifndef CONFIG_MMU
mm->end_data = seg->addr + phdr->p_memsz;
#endif
}
#ifdef CONFIG_MMU
if (seg->addr + phdr->p_memsz > mm->end_data)
mm->end_data = seg->addr + phdr->p_memsz;
#endif
}
seg++;
}
return 0;
}
#endif
/*****************************************************************************/
/*
* map a binary by direct mmap() of the individual PT_LOAD segments
*/
static int elf_fdpic_map_file_by_direct_mmap(struct elf_fdpic_params *params,
struct file *file,
struct mm_struct *mm)
{
struct elf32_fdpic_loadseg *seg;
struct elf32_phdr *phdr;
unsigned long load_addr, delta_vaddr;
int loop, dvset, ret;
load_addr = params->load_addr;
delta_vaddr = 0;
dvset = 0;
seg = params->loadmap->segs;
/* deal with each load segment separately */
phdr = params->phdrs;
for (loop = 0; loop < params->hdr.e_phnum; loop++, phdr++) {
unsigned long maddr, disp, excess, excess1;
int prot = 0, flags;
if (phdr->p_type != PT_LOAD)
continue;
kdebug("[LOAD] va=%lx of=%lx fs=%lx ms=%lx",
(unsigned long) phdr->p_vaddr,
(unsigned long) phdr->p_offset,
(unsigned long) phdr->p_filesz,
(unsigned long) phdr->p_memsz);
/* determine the mapping parameters */
if (phdr->p_flags & PF_R) prot |= PROT_READ;
if (phdr->p_flags & PF_W) prot |= PROT_WRITE;
if (phdr->p_flags & PF_X) prot |= PROT_EXEC;
flags = MAP_PRIVATE | MAP_DENYWRITE;
if (params->flags & ELF_FDPIC_FLAG_EXECUTABLE)
flags |= MAP_EXECUTABLE;
maddr = 0;
switch (params->flags & ELF_FDPIC_FLAG_ARRANGEMENT) {
case ELF_FDPIC_FLAG_INDEPENDENT:
/* PT_LOADs are independently locatable */
break;
case ELF_FDPIC_FLAG_HONOURVADDR:
/* the specified virtual address must be honoured */
maddr = phdr->p_vaddr;
flags |= MAP_FIXED;
break;
case ELF_FDPIC_FLAG_CONSTDISP:
/* constant displacement
* - can be mapped anywhere, but must be mapped as a
* unit
*/
if (!dvset) {
maddr = load_addr;
delta_vaddr = phdr->p_vaddr;
dvset = 1;
} else {
maddr = load_addr + phdr->p_vaddr - delta_vaddr;
flags |= MAP_FIXED;
}
break;
case ELF_FDPIC_FLAG_CONTIGUOUS:
/* contiguity handled later */
break;
default:
BUG();
}
maddr &= PAGE_MASK;
/* create the mapping */
disp = phdr->p_vaddr & ~PAGE_MASK;
down_write(&mm->mmap_sem);
maddr = do_mmap(file, maddr, phdr->p_memsz + disp, prot, flags,
phdr->p_offset - disp);
up_write(&mm->mmap_sem);
kdebug("mmap[%d] <file> sz=%lx pr=%x fl=%x of=%lx --> %08lx",
loop, phdr->p_memsz + disp, prot, flags,
phdr->p_offset - disp, maddr);
if (IS_ERR_VALUE(maddr))
return (int) maddr;
if ((params->flags & ELF_FDPIC_FLAG_ARRANGEMENT) ==
ELF_FDPIC_FLAG_CONTIGUOUS)
load_addr += PAGE_ALIGN(phdr->p_memsz + disp);
seg->addr = maddr + disp;
seg->p_vaddr = phdr->p_vaddr;
seg->p_memsz = phdr->p_memsz;
/* map the ELF header address if in this segment */
if (phdr->p_offset == 0)
params->elfhdr_addr = seg->addr;
/* clear the bit between beginning of mapping and beginning of
* PT_LOAD */
if (prot & PROT_WRITE && disp > 0) {
kdebug("clear[%d] ad=%lx sz=%lx", loop, maddr, disp);
ret = clear_user((void __user *) maddr, disp);
if (ret)
return ret;
maddr += disp;
}
/* clear any space allocated but not loaded
* - on uClinux we can just clear the lot
* - on MMU linux we'll get a SIGBUS beyond the last page
* extant in the file
*/
excess = phdr->p_memsz - phdr->p_filesz;
excess1 = PAGE_SIZE - ((maddr + phdr->p_filesz) & ~PAGE_MASK);
#ifdef CONFIG_MMU
if (excess > excess1) {
unsigned long xaddr = maddr + phdr->p_filesz + excess1;
unsigned long xmaddr;
flags |= MAP_FIXED | MAP_ANONYMOUS;
down_write(&mm->mmap_sem);
xmaddr = do_mmap(NULL, xaddr, excess - excess1,
prot, flags, 0);
up_write(&mm->mmap_sem);
kdebug("mmap[%d] <anon>"
" ad=%lx sz=%lx pr=%x fl=%x of=0 --> %08lx",
loop, xaddr, excess - excess1, prot, flags,
xmaddr);
if (xmaddr != xaddr)
return -ENOMEM;
}
if (prot & PROT_WRITE && excess1 > 0) {
kdebug("clear[%d] ad=%lx sz=%lx",
loop, maddr + phdr->p_filesz, excess1);
ret = clear_user((void __user *) maddr + phdr->p_filesz,
excess1);
if (ret)
return ret;
}
#else
if (excess > 0) {
kdebug("clear[%d] ad=%lx sz=%lx",
loop, maddr + phdr->p_filesz, excess);
ret = clear_user((void *) maddr + phdr->p_filesz, excess);
if (ret)
return ret;
}
#endif
if (mm) {
if (phdr->p_flags & PF_X) {
[PATCH] FDPIC: fix the /proc/pid/stat representation of executable boundaries Fix the /proc/pid/stat representation of executable boundaries. It should show the bounds of the executable, but instead shows the bounds of the loader. Before the patch is applied, the bug can be seen by examining, say, inetd: # ps | grep inetd 610 root 0 S /usr/sbin/inetd -i # cat /proc/610/maps c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so c3180000-c31dede4 r-xs 00000000 00:0b 14582179 /lib/libuClibc-0.9.28.so c328c000-c328ea00 rw-p 00008000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so c3290000-c329b6c0 rw-p 00000000 00:00 0 c32a0000-c32c0000 rwxp 00000000 00:00 0 c32d4000-c32d8000 rw-p 00000000 00:00 0 c3394000-c3398000 rw-p 00000000 00:00 0 c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd c3470000-c34748f8 rw-p 00004000 00:0b 16384612 /usr/sbin/inetd c34cc000-c34d0000 rw-p 00000000 00:00 0 c34d4000-c34d8000 rw-p 00000000 00:00 0 c34d8000-c34dc000 rw-p 00000000 00:00 0 # cat /proc/610/stat 610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 8 0 0 19 0 1 0 94392000718 950272 0 4294967295 3233480704 3233523592 3274440352 3274439976 3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0 The code boundaries are 3233480704 to 3233523592, which are: (gdb) p/x 3233480704 $1 = 0xc0bb0000 (gdb) p/x 3233523592 $2 = 0xc0bba788 Which corresponds to this line in the maps file: c0bb0000-c0bba788 r-xs 00000000 00:0b 14582157 /lib/ld-uClibc-0.9.28.so Which is wrong. After the patch is applied, the maps file is pretty much identical (there's some minor shuffling of the location of some of the anonymous VMAs), but the stat file is now: # cat /proc/610/stat 610 (inetd) S 1 610 610 0 -1 256 0 0 0 0 0 7 0 0 18 0 1 0 94392000722 950272 0 4294967295 3276111872 3276141668 3274440352 3274439976 3273467584 0 0 4096 90115 3221712796 0 0 17 0 0 0 0 The code boundaries are then 3276111872 to 3276141668, which are: (gdb) p/x 3276111872 $1 = 0xc3458000 (gdb) p/x 3276141668 $2 = 0xc345f464 And these correspond to this line in the maps file instead: c3458000-c345f464 r-xs 00000000 00:0b 16384612 /usr/sbin/inetd Which is now correct. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-03-23 08:10:00 +01:00
if (!mm->start_code) {
mm->start_code = maddr;
mm->end_code = maddr + phdr->p_memsz;
}
} else if (!mm->start_data) {
mm->start_data = maddr;
mm->end_data = maddr + phdr->p_memsz;
}
}
seg++;
}
return 0;
}
/*****************************************************************************/
/*
* ELF-FDPIC core dumper
*
* Modelled on fs/exec.c:aout_core_dump()
* Jeremy Fitzhardinge <jeremy@sw.oz.au>
*
* Modelled on fs/binfmt_elf.c core dumper
*/
#if defined(USE_ELF_CORE_DUMP) && defined(CONFIG_ELF_CORE)
/*
* These are the only things you should do on a core-file: use only these
* functions to write out all the necessary info.
*/
static int dump_write(struct file *file, const void *addr, int nr)
{
return file->f_op->write(file, addr, nr, &file->f_pos) == nr;
}
static int dump_seek(struct file *file, loff_t off)
{
if (file->f_op->llseek) {
if (file->f_op->llseek(file, off, SEEK_SET) != off)
return 0;
} else {
file->f_pos = off;
}
return 1;
}
/*
* Decide whether a segment is worth dumping; default is yes to be
* sure (missing info is worse than too much; etc).
* Personally I'd include everything, and use the coredump limit...
*
* I think we should skip something. But I am not sure how. H.J.
*/
static int maydump(struct vm_area_struct *vma, unsigned long mm_flags)
{
int dump_ok;
/* Do not dump I/O mapped devices or special mappings */
if (vma->vm_flags & (VM_IO | VM_RESERVED)) {
kdcore("%08lx: %08lx: no (IO)", vma->vm_start, vma->vm_flags);
return 0;
}
/* If we may not read the contents, don't allow us to dump
* them either. "dump_write()" can't handle it anyway.
*/
if (!(vma->vm_flags & VM_READ)) {
kdcore("%08lx: %08lx: no (!read)", vma->vm_start, vma->vm_flags);
return 0;
}
/* By default, dump shared memory if mapped from an anonymous file. */
if (vma->vm_flags & VM_SHARED) {
if (vma->vm_file->f_path.dentry->d_inode->i_nlink == 0) {
dump_ok = test_bit(MMF_DUMP_ANON_SHARED, &mm_flags);
kdcore("%08lx: %08lx: %s (share)", vma->vm_start,
vma->vm_flags, dump_ok ? "yes" : "no");
return dump_ok;
}
dump_ok = test_bit(MMF_DUMP_MAPPED_SHARED, &mm_flags);
kdcore("%08lx: %08lx: %s (share)", vma->vm_start,
vma->vm_flags, dump_ok ? "yes" : "no");
return dump_ok;
}
#ifdef CONFIG_MMU
/* By default, if it hasn't been written to, don't write it out */
if (!vma->anon_vma) {
dump_ok = test_bit(MMF_DUMP_MAPPED_PRIVATE, &mm_flags);
kdcore("%08lx: %08lx: %s (!anon)", vma->vm_start,
vma->vm_flags, dump_ok ? "yes" : "no");
return dump_ok;
}
#endif
dump_ok = test_bit(MMF_DUMP_ANON_PRIVATE, &mm_flags);
kdcore("%08lx: %08lx: %s", vma->vm_start, vma->vm_flags,
dump_ok ? "yes" : "no");
return dump_ok;
}
/* An ELF note in memory */
struct memelfnote
{
const char *name;
int type;
unsigned int datasz;
void *data;
};
static int notesize(struct memelfnote *en)
{
int sz;
sz = sizeof(struct elf_note);
sz += roundup(strlen(en->name) + 1, 4);
sz += roundup(en->datasz, 4);
return sz;
}
/* #define DEBUG */
#define DUMP_WRITE(addr, nr) \
do { if (!dump_write(file, (addr), (nr))) return 0; } while(0)
#define DUMP_SEEK(off) \
do { if (!dump_seek(file, (off))) return 0; } while(0)
static int writenote(struct memelfnote *men, struct file *file)
{
struct elf_note en;
en.n_namesz = strlen(men->name) + 1;
en.n_descsz = men->datasz;
en.n_type = men->type;
DUMP_WRITE(&en, sizeof(en));
DUMP_WRITE(men->name, en.n_namesz);
/* XXX - cast from long long to long to avoid need for libgcc.a */
DUMP_SEEK(roundup((unsigned long)file->f_pos, 4)); /* XXX */
DUMP_WRITE(men->data, men->datasz);
DUMP_SEEK(roundup((unsigned long)file->f_pos, 4)); /* XXX */
return 1;
}
#undef DUMP_WRITE
#undef DUMP_SEEK
#define DUMP_WRITE(addr, nr) \
if ((size += (nr)) > limit || !dump_write(file, (addr), (nr))) \
goto end_coredump;
#define DUMP_SEEK(off) \
if (!dump_seek(file, (off))) \
goto end_coredump;
static inline void fill_elf_fdpic_header(struct elfhdr *elf, int segs)
{
memcpy(elf->e_ident, ELFMAG, SELFMAG);
elf->e_ident[EI_CLASS] = ELF_CLASS;
elf->e_ident[EI_DATA] = ELF_DATA;
elf->e_ident[EI_VERSION] = EV_CURRENT;
elf->e_ident[EI_OSABI] = ELF_OSABI;
memset(elf->e_ident+EI_PAD, 0, EI_NIDENT-EI_PAD);
elf->e_type = ET_CORE;
elf->e_machine = ELF_ARCH;
elf->e_version = EV_CURRENT;
elf->e_entry = 0;
elf->e_phoff = sizeof(struct elfhdr);
elf->e_shoff = 0;
elf->e_flags = ELF_FDPIC_CORE_EFLAGS;
elf->e_ehsize = sizeof(struct elfhdr);
elf->e_phentsize = sizeof(struct elf_phdr);
elf->e_phnum = segs;
elf->e_shentsize = 0;
elf->e_shnum = 0;
elf->e_shstrndx = 0;
return;
}
static inline void fill_elf_note_phdr(struct elf_phdr *phdr, int sz, loff_t offset)
{
phdr->p_type = PT_NOTE;
phdr->p_offset = offset;
phdr->p_vaddr = 0;
phdr->p_paddr = 0;
phdr->p_filesz = sz;
phdr->p_memsz = 0;
phdr->p_flags = 0;
phdr->p_align = 0;
return;
}
static inline void fill_note(struct memelfnote *note, const char *name, int type,
unsigned int sz, void *data)
{
note->name = name;
note->type = type;
note->datasz = sz;
note->data = data;
return;
}
/*
* fill up all the fields in prstatus from the given task struct, except
* registers which need to be filled up seperately.
*/
static void fill_prstatus(struct elf_prstatus *prstatus,
struct task_struct *p, long signr)
{
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
prstatus->pr_pid = task_pid_vnr(p);
prstatus->pr_ppid = task_pid_vnr(p->real_parent);
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
struct task_cputime cputime;
/*
* This is the record for the group leader. It shows the
* group-wide total, not its individual thread total.
*/
thread_group_cputime(p, &cputime);
cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
} else {
cputime_to_timeval(p->utime, &prstatus->pr_utime);
cputime_to_timeval(p->stime, &prstatus->pr_stime);
}
cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime);
cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime);
prstatus->pr_exec_fdpic_loadmap = p->mm->context.exec_fdpic_loadmap;
prstatus->pr_interp_fdpic_loadmap = p->mm->context.interp_fdpic_loadmap;
}
static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p,
struct mm_struct *mm)
{
const struct cred *cred;
unsigned int i, len;
/* first copy the parameters from user space */
memset(psinfo, 0, sizeof(struct elf_prpsinfo));
len = mm->arg_end - mm->arg_start;
if (len >= ELF_PRARGSZ)
len = ELF_PRARGSZ - 1;
if (copy_from_user(&psinfo->pr_psargs,
(const char __user *) mm->arg_start, len))
return -EFAULT;
for (i = 0; i < len; i++)
if (psinfo->pr_psargs[i] == 0)
psinfo->pr_psargs[i] = ' ';
psinfo->pr_psargs[len] = 0;
psinfo->pr_pid = task_pid_vnr(p);
psinfo->pr_ppid = task_pid_vnr(p->real_parent);
psinfo->pr_pgrp = task_pgrp_vnr(p);
psinfo->pr_sid = task_session_vnr(p);
i = p->state ? ffz(~p->state) + 1 : 0;
psinfo->pr_state = i;
psinfo->pr_sname = (i > 5) ? '.' : "RSDTZW"[i];
psinfo->pr_zomb = psinfo->pr_sname == 'Z';
psinfo->pr_nice = task_nice(p);
psinfo->pr_flag = p->flags;
rcu_read_lock();
cred = __task_cred(p);
SET_UID(psinfo->pr_uid, cred->uid);
SET_GID(psinfo->pr_gid, cred->gid);
rcu_read_unlock();
strncpy(psinfo->pr_fname, p->comm, sizeof(psinfo->pr_fname));
return 0;
}
/* Here is the structure in which status of each thread is captured. */
struct elf_thread_status
{
struct list_head list;
struct elf_prstatus prstatus; /* NT_PRSTATUS */
elf_fpregset_t fpu; /* NT_PRFPREG */
struct task_struct *thread;
#ifdef ELF_CORE_COPY_XFPREGS
elf_fpxregset_t xfpu; /* ELF_CORE_XFPREG_TYPE */
#endif
struct memelfnote notes[3];
int num_notes;
};
/*
* In order to add the specific thread information for the elf file format,
* we need to keep a linked list of every thread's pr_status and then create
* a single section for them in the final core file.
*/
static int elf_dump_thread_status(long signr, struct elf_thread_status *t)
{
struct task_struct *p = t->thread;
int sz = 0;
t->num_notes = 0;
fill_prstatus(&t->prstatus, p, signr);
elf_core_copy_task_regs(p, &t->prstatus.pr_reg);
fill_note(&t->notes[0], "CORE", NT_PRSTATUS, sizeof(t->prstatus),
&t->prstatus);
t->num_notes++;
sz += notesize(&t->notes[0]);
t->prstatus.pr_fpvalid = elf_core_copy_task_fpregs(p, NULL, &t->fpu);
if (t->prstatus.pr_fpvalid) {
fill_note(&t->notes[1], "CORE", NT_PRFPREG, sizeof(t->fpu),
&t->fpu);
t->num_notes++;
sz += notesize(&t->notes[1]);
}
#ifdef ELF_CORE_COPY_XFPREGS
if (elf_core_copy_task_xfpregs(p, &t->xfpu)) {
fill_note(&t->notes[2], "LINUX", ELF_CORE_XFPREG_TYPE,
sizeof(t->xfpu), &t->xfpu);
t->num_notes++;
sz += notesize(&t->notes[2]);
}
#endif
return sz;
}
/*
* dump the segments for an MMU process
*/
#ifdef CONFIG_MMU
static int elf_fdpic_dump_segments(struct file *file, size_t *size,
unsigned long *limit, unsigned long mm_flags)
{
struct vm_area_struct *vma;
for (vma = current->mm->mmap; vma; vma = vma->vm_next) {
unsigned long addr;
if (!maydump(vma, mm_flags))
continue;
for (addr = vma->vm_start;
addr < vma->vm_end;
addr += PAGE_SIZE
) {
struct vm_area_struct *vma;
struct page *page;
if (get_user_pages(current, current->mm, addr, 1, 0, 1,
&page, &vma) <= 0) {
DUMP_SEEK(file->f_pos + PAGE_SIZE);
}
remove ZERO_PAGE The commit b5810039a54e5babf428e9a1e89fc1940fabff11 contains the note A last caveat: the ZERO_PAGE is now refcounted and managed with rmap (and thus mapcounted and count towards shared rss). These writes to the struct page could cause excessive cacheline bouncing on big systems. There are a number of ways this could be addressed if it is an issue. And indeed this cacheline bouncing has shown up on large SGI systems. There was a situation where an Altix system was essentially livelocked tearing down ZERO_PAGE pagetables when an HPC app aborted during startup. This situation can be avoided in userspace, but it does highlight the potential scalability problem with refcounting ZERO_PAGE, and corner cases where it can really hurt (we don't want the system to livelock!). There are several broad ways to fix this problem: 1. add back some special casing to avoid refcounting ZERO_PAGE 2. per-node or per-cpu ZERO_PAGES 3. remove the ZERO_PAGE completely I will argue for 3. The others should also fix the problem, but they result in more complex code than does 3, with little or no real benefit that I can see. Why? Inserting a ZERO_PAGE for anonymous read faults appears to be a false optimisation: if an application is performance critical, it would not be doing many read faults of new memory, or at least it could be expected to write to that memory soon afterwards. If cache or memory use is critical, it should not be working with a significant number of ZERO_PAGEs anyway (a more compact representation of zeroes should be used). As a sanity check -- mesuring on my desktop system, there are never many mappings to the ZERO_PAGE (eg. 2 or 3), thus memory usage here should not increase much without it. When running a make -j4 kernel compile on my dual core system, there are about 1,000 mappings to the ZERO_PAGE created per second, but about 1,000 ZERO_PAGE COW faults per second (less than 1 ZERO_PAGE mapping per second is torn down without being COWed). So removing ZERO_PAGE will save 1,000 page faults per second when running kbuild, while keeping it only saves less than 1 page clearing operation per second. 1 page clear is cheaper than a thousand faults, presumably, so there isn't an obvious loss. Neither the logical argument nor these basic tests give a guarantee of no regressions. However, this is a reasonable opportunity to try to remove the ZERO_PAGE from the pagefault path. If it is found to cause regressions, we can reintroduce it and just avoid refcounting it. The /dev/zero ZERO_PAGE usage and TLB tricks also get nuked. I don't see much use to them except on benchmarks. All other users of ZERO_PAGE are converted just to use ZERO_PAGE(0) for simplicity. We can look at replacing them all and maybe ripping out ZERO_PAGE completely when we are more satisfied with this solution. Signed-off-by: Nick Piggin <npiggin@suse.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus "snif" Torvalds <torvalds@linux-foundation.org>
2007-10-16 10:24:40 +02:00
else if (page == ZERO_PAGE(0)) {
page_cache_release(page);
DUMP_SEEK(file->f_pos + PAGE_SIZE);
}
else {
void *kaddr;
flush_cache_page(vma, addr, page_to_pfn(page));
kaddr = kmap(page);
if ((*size += PAGE_SIZE) > *limit ||
!dump_write(file, kaddr, PAGE_SIZE)
) {
kunmap(page);
page_cache_release(page);
return -EIO;
}
kunmap(page);
page_cache_release(page);
}
}
}
return 0;
end_coredump:
return -EFBIG;
}
#endif
/*
* dump the segments for a NOMMU process
*/
#ifndef CONFIG_MMU
static int elf_fdpic_dump_segments(struct file *file, size_t *size,
unsigned long *limit, unsigned long mm_flags)
{
NOMMU: Make VMAs per MM as for MMU-mode linux Make VMAs per mm_struct as for MMU-mode linux. This solves two problems: (1) In SYSV SHM where nattch for a segment does not reflect the number of shmat's (and forks) done. (2) In mmap() where the VMA's vm_mm is set to point to the parent mm by an exec'ing process when VM_EXECUTABLE is specified, regardless of the fact that a VMA might be shared and already have its vm_mm assigned to another process or a dead process. A new struct (vm_region) is introduced to track a mapped region and to remember the circumstances under which it may be shared and the vm_list_struct structure is discarded as it's no longer required. This patch makes the following additional changes: (1) Regions are now allocated with alloc_pages() rather than kmalloc() and with no recourse to __GFP_COMP, so the pages are not composite. Instead, each page has a reference on it held by the region. Anything else that is interested in such a page will have to get a reference on it to retain it. When the pages are released due to unmapping, each page is passed to put_page() and will be freed when the page usage count reaches zero. (2) Excess pages are trimmed after an allocation as the allocation must be made as a power-of-2 quantity of pages. (3) VMAs are added to the parent MM's R/B tree and mmap lists. As an MM may end up with overlapping VMAs within the tree, the VMA struct address is appended to the sort key. (4) Non-anonymous VMAs are now added to the backing inode's prio list. (5) Holes may be punched in anonymous VMAs with munmap(), releasing parts of the backing region. The VMA and region structs will be split if necessary. (6) sys_shmdt() only releases one attachment to a SYSV IPC shared memory segment instead of all the attachments at that addresss. Multiple shmat()'s return the same address under NOMMU-mode instead of different virtual addresses as under MMU-mode. (7) Core dumping for ELF-FDPIC requires fewer exceptions for NOMMU-mode. (8) /proc/maps is now the global list of mapped regions, and may list bits that aren't actually mapped anywhere. (9) /proc/meminfo gains a line (tagged "MmapCopy") that indicates the amount of RAM currently allocated by mmap to hold mappable regions that can't be mapped directly. These are copies of the backing device or file if not anonymous. These changes make NOMMU mode more similar to MMU mode. The downside is that NOMMU mode requires some extra memory to track things over NOMMU without this patch (VMAs are no longer shared, and there are now region structs). Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Mike Frysinger <vapier.adi@gmail.com> Acked-by: Paul Mundt <lethal@linux-sh.org>
2009-01-08 13:04:47 +01:00
struct vm_area_struct *vma;
NOMMU: Make VMAs per MM as for MMU-mode linux Make VMAs per mm_struct as for MMU-mode linux. This solves two problems: (1) In SYSV SHM where nattch for a segment does not reflect the number of shmat's (and forks) done. (2) In mmap() where the VMA's vm_mm is set to point to the parent mm by an exec'ing process when VM_EXECUTABLE is specified, regardless of the fact that a VMA might be shared and already have its vm_mm assigned to another process or a dead process. A new struct (vm_region) is introduced to track a mapped region and to remember the circumstances under which it may be shared and the vm_list_struct structure is discarded as it's no longer required. This patch makes the following additional changes: (1) Regions are now allocated with alloc_pages() rather than kmalloc() and with no recourse to __GFP_COMP, so the pages are not composite. Instead, each page has a reference on it held by the region. Anything else that is interested in such a page will have to get a reference on it to retain it. When the pages are released due to unmapping, each page is passed to put_page() and will be freed when the page usage count reaches zero. (2) Excess pages are trimmed after an allocation as the allocation must be made as a power-of-2 quantity of pages. (3) VMAs are added to the parent MM's R/B tree and mmap lists. As an MM may end up with overlapping VMAs within the tree, the VMA struct address is appended to the sort key. (4) Non-anonymous VMAs are now added to the backing inode's prio list. (5) Holes may be punched in anonymous VMAs with munmap(), releasing parts of the backing region. The VMA and region structs will be split if necessary. (6) sys_shmdt() only releases one attachment to a SYSV IPC shared memory segment instead of all the attachments at that addresss. Multiple shmat()'s return the same address under NOMMU-mode instead of different virtual addresses as under MMU-mode. (7) Core dumping for ELF-FDPIC requires fewer exceptions for NOMMU-mode. (8) /proc/maps is now the global list of mapped regions, and may list bits that aren't actually mapped anywhere. (9) /proc/meminfo gains a line (tagged "MmapCopy") that indicates the amount of RAM currently allocated by mmap to hold mappable regions that can't be mapped directly. These are copies of the backing device or file if not anonymous. These changes make NOMMU mode more similar to MMU mode. The downside is that NOMMU mode requires some extra memory to track things over NOMMU without this patch (VMAs are no longer shared, and there are now region structs). Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Mike Frysinger <vapier.adi@gmail.com> Acked-by: Paul Mundt <lethal@linux-sh.org>
2009-01-08 13:04:47 +01:00
for (vma = current->mm->mmap; vma; vma = vma->vm_next) {
if (!maydump(vma, mm_flags))
continue;
if ((*size += PAGE_SIZE) > *limit)
return -EFBIG;
if (!dump_write(file, (void *) vma->vm_start,
vma->vm_end - vma->vm_start))
return -EIO;
}
return 0;
}
#endif
/*
* Actual dumper
*
* This is a two-pass process; first we find the offsets of the bits,
* and then they are actually written out. If we run out of core limit
* we just truncate.
*/
static int elf_fdpic_core_dump(long signr, struct pt_regs *regs,
core_pattern: ignore RLIMIT_CORE if core_pattern is a pipe For some time /proc/sys/kernel/core_pattern has been able to set its output destination as a pipe, allowing a user space helper to receive and intellegently process a core. This infrastructure however has some shortcommings which can be enhanced. Specifically: 1) The coredump code in the kernel should ignore RLIMIT_CORE limitation when core_pattern is a pipe, since file system resources are not being consumed in this case, unless the user application wishes to save the core, at which point the app is restricted by usual file system limits and restrictions. 2) The core_pattern code should be able to parse and pass options to the user space helper as an argv array. The real core limit of the uid of the crashing proces should also be passable to the user space helper (since it is overridden to zero when called). 3) Some miscellaneous bugs need to be cleaned up (specifically the recognition of a recursive core dump, should the user mode helper itself crash. Also, the core dump code in the kernel should not wait for the user mode helper to exit, since the same context is responsible for writing to the pipe, and a read of the pipe by the user mode helper will result in a deadlock. This patch: Remove the check of RLIMIT_CORE if core_pattern is a pipe. In the event that core_pattern is a pipe, the entire core will be fed to the user mode helper. Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Cc: <martin.pitt@ubuntu.com> Cc: <wwoods@redhat.com> Cc: Jeremy Fitzhardinge <jeremy@goop.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-17 08:26:34 +02:00
struct file *file, unsigned long limit)
{
#define NUM_NOTES 6
int has_dumped = 0;
mm_segment_t fs;
int segs;
size_t size = 0;
int i;
struct vm_area_struct *vma;
struct elfhdr *elf = NULL;
loff_t offset = 0, dataoff;
int numnote;
struct memelfnote *notes = NULL;
struct elf_prstatus *prstatus = NULL; /* NT_PRSTATUS */
struct elf_prpsinfo *psinfo = NULL; /* NT_PRPSINFO */
LIST_HEAD(thread_list);
struct list_head *t;
elf_fpregset_t *fpu = NULL;
#ifdef ELF_CORE_COPY_XFPREGS
elf_fpxregset_t *xfpu = NULL;
#endif
int thread_status_size = 0;
elf_addr_t *auxv;
unsigned long mm_flags;
/*
* We no longer stop all VM operations.
*
* This is because those proceses that could possibly change map_count
* or the mmap / vma pages are now blocked in do_exit on current
* finishing this core dump.
*
* Only ptrace can touch these memory addresses, but it doesn't change
* the map_count or the pages allocated. So no possibility of crashing
* exists while dumping the mm->vm_next areas to the core file.
*/
/* alloc memory for large data structures: too large to be on stack */
elf = kmalloc(sizeof(*elf), GFP_KERNEL);
if (!elf)
goto cleanup;
prstatus = kzalloc(sizeof(*prstatus), GFP_KERNEL);
if (!prstatus)
goto cleanup;
psinfo = kmalloc(sizeof(*psinfo), GFP_KERNEL);
if (!psinfo)
goto cleanup;
notes = kmalloc(NUM_NOTES * sizeof(struct memelfnote), GFP_KERNEL);
if (!notes)
goto cleanup;
fpu = kmalloc(sizeof(*fpu), GFP_KERNEL);
if (!fpu)
goto cleanup;
#ifdef ELF_CORE_COPY_XFPREGS
xfpu = kmalloc(sizeof(*xfpu), GFP_KERNEL);
if (!xfpu)
goto cleanup;
#endif
if (signr) {
struct core_thread *ct;
struct elf_thread_status *tmp;
for (ct = current->mm->core_state->dumper.next;
ct; ct = ct->next) {
tmp = kzalloc(sizeof(*tmp), GFP_KERNEL);
if (!tmp)
goto cleanup;
tmp->thread = ct->task;
list_add(&tmp->list, &thread_list);
}
list_for_each(t, &thread_list) {
struct elf_thread_status *tmp;
int sz;
tmp = list_entry(t, struct elf_thread_status, list);
sz = elf_dump_thread_status(signr, tmp);
thread_status_size += sz;
}
}
/* now collect the dump for the current */
fill_prstatus(prstatus, current, signr);
elf_core_copy_regs(&prstatus->pr_reg, regs);
segs = current->mm->map_count;
#ifdef ELF_CORE_EXTRA_PHDRS
segs += ELF_CORE_EXTRA_PHDRS;
#endif
/* Set up header */
fill_elf_fdpic_header(elf, segs + 1); /* including notes section */
has_dumped = 1;
current->flags |= PF_DUMPCORE;
/*
* Set up the notes in similar form to SVR4 core dumps made
* with info from their /proc.
*/
fill_note(notes + 0, "CORE", NT_PRSTATUS, sizeof(*prstatus), prstatus);
fill_psinfo(psinfo, current->group_leader, current->mm);
fill_note(notes + 1, "CORE", NT_PRPSINFO, sizeof(*psinfo), psinfo);
numnote = 2;
auxv = (elf_addr_t *) current->mm->saved_auxv;
i = 0;
do
i += 2;
while (auxv[i - 2] != AT_NULL);
fill_note(&notes[numnote++], "CORE", NT_AUXV,
i * sizeof(elf_addr_t), auxv);
/* Try to dump the FPU. */
if ((prstatus->pr_fpvalid =
elf_core_copy_task_fpregs(current, regs, fpu)))
fill_note(notes + numnote++,
"CORE", NT_PRFPREG, sizeof(*fpu), fpu);
#ifdef ELF_CORE_COPY_XFPREGS
if (elf_core_copy_task_xfpregs(current, xfpu))
fill_note(notes + numnote++,
"LINUX", ELF_CORE_XFPREG_TYPE, sizeof(*xfpu), xfpu);
#endif
fs = get_fs();
set_fs(KERNEL_DS);
DUMP_WRITE(elf, sizeof(*elf));
offset += sizeof(*elf); /* Elf header */
offset += (segs+1) * sizeof(struct elf_phdr); /* Program headers */
/* Write notes phdr entry */
{
struct elf_phdr phdr;
int sz = 0;
for (i = 0; i < numnote; i++)
sz += notesize(notes + i);
sz += thread_status_size;
fill_elf_note_phdr(&phdr, sz, offset);
offset += sz;
DUMP_WRITE(&phdr, sizeof(phdr));
}
/* Page-align dumped data */
dataoff = offset = roundup(offset, ELF_EXEC_PAGESIZE);
/*
* We must use the same mm->flags while dumping core to avoid
* inconsistency between the program headers and bodies, otherwise an
* unusable core file can be generated.
*/
mm_flags = current->mm->flags;
/* write program headers for segments dump */
NOMMU: Make VMAs per MM as for MMU-mode linux Make VMAs per mm_struct as for MMU-mode linux. This solves two problems: (1) In SYSV SHM where nattch for a segment does not reflect the number of shmat's (and forks) done. (2) In mmap() where the VMA's vm_mm is set to point to the parent mm by an exec'ing process when VM_EXECUTABLE is specified, regardless of the fact that a VMA might be shared and already have its vm_mm assigned to another process or a dead process. A new struct (vm_region) is introduced to track a mapped region and to remember the circumstances under which it may be shared and the vm_list_struct structure is discarded as it's no longer required. This patch makes the following additional changes: (1) Regions are now allocated with alloc_pages() rather than kmalloc() and with no recourse to __GFP_COMP, so the pages are not composite. Instead, each page has a reference on it held by the region. Anything else that is interested in such a page will have to get a reference on it to retain it. When the pages are released due to unmapping, each page is passed to put_page() and will be freed when the page usage count reaches zero. (2) Excess pages are trimmed after an allocation as the allocation must be made as a power-of-2 quantity of pages. (3) VMAs are added to the parent MM's R/B tree and mmap lists. As an MM may end up with overlapping VMAs within the tree, the VMA struct address is appended to the sort key. (4) Non-anonymous VMAs are now added to the backing inode's prio list. (5) Holes may be punched in anonymous VMAs with munmap(), releasing parts of the backing region. The VMA and region structs will be split if necessary. (6) sys_shmdt() only releases one attachment to a SYSV IPC shared memory segment instead of all the attachments at that addresss. Multiple shmat()'s return the same address under NOMMU-mode instead of different virtual addresses as under MMU-mode. (7) Core dumping for ELF-FDPIC requires fewer exceptions for NOMMU-mode. (8) /proc/maps is now the global list of mapped regions, and may list bits that aren't actually mapped anywhere. (9) /proc/meminfo gains a line (tagged "MmapCopy") that indicates the amount of RAM currently allocated by mmap to hold mappable regions that can't be mapped directly. These are copies of the backing device or file if not anonymous. These changes make NOMMU mode more similar to MMU mode. The downside is that NOMMU mode requires some extra memory to track things over NOMMU without this patch (VMAs are no longer shared, and there are now region structs). Signed-off-by: David Howells <dhowells@redhat.com> Tested-by: Mike Frysinger <vapier.adi@gmail.com> Acked-by: Paul Mundt <lethal@linux-sh.org>
2009-01-08 13:04:47 +01:00
for (vma = current->mm->mmap; vma; vma = vma->vm_next) {
struct elf_phdr phdr;
size_t sz;
sz = vma->vm_end - vma->vm_start;
phdr.p_type = PT_LOAD;
phdr.p_offset = offset;
phdr.p_vaddr = vma->vm_start;
phdr.p_paddr = 0;
phdr.p_filesz = maydump(vma, mm_flags) ? sz : 0;
phdr.p_memsz = sz;
offset += phdr.p_filesz;
phdr.p_flags = vma->vm_flags & VM_READ ? PF_R : 0;
if (vma->vm_flags & VM_WRITE)
phdr.p_flags |= PF_W;
if (vma->vm_flags & VM_EXEC)
phdr.p_flags |= PF_X;
phdr.p_align = ELF_EXEC_PAGESIZE;
DUMP_WRITE(&phdr, sizeof(phdr));
}
#ifdef ELF_CORE_WRITE_EXTRA_PHDRS
ELF_CORE_WRITE_EXTRA_PHDRS;
#endif
/* write out the notes section */
for (i = 0; i < numnote; i++)
if (!writenote(notes + i, file))
goto end_coredump;
/* write out the thread status notes section */
list_for_each(t, &thread_list) {
struct elf_thread_status *tmp =
list_entry(t, struct elf_thread_status, list);
for (i = 0; i < tmp->num_notes; i++)
if (!writenote(&tmp->notes[i], file))
goto end_coredump;
}
DUMP_SEEK(dataoff);
if (elf_fdpic_dump_segments(file, &size, &limit, mm_flags) < 0)
goto end_coredump;
#ifdef ELF_CORE_WRITE_EXTRA_DATA
ELF_CORE_WRITE_EXTRA_DATA;
#endif
if (file->f_pos != offset) {
/* Sanity check */
printk(KERN_WARNING
"elf_core_dump: file->f_pos (%lld) != offset (%lld)\n",
file->f_pos, offset);
}
end_coredump:
set_fs(fs);
cleanup:
while (!list_empty(&thread_list)) {
struct list_head *tmp = thread_list.next;
list_del(tmp);
kfree(list_entry(tmp, struct elf_thread_status, list));
}
kfree(elf);
kfree(prstatus);
kfree(psinfo);
kfree(notes);
kfree(fpu);
#ifdef ELF_CORE_COPY_XFPREGS
kfree(xfpu);
#endif
return has_dumped;
#undef NUM_NOTES
}
#endif /* USE_ELF_CORE_DUMP */