linux_dsm_epyc7002/fs/binfmt_aout.c
Wolfgang Wander 1363c3cd86 [PATCH] Avoiding mmap fragmentation
Ingo recently introduced a great speedup for allocating new mmaps using the
free_area_cache pointer which boosts the specweb SSL benchmark by 4-5% and
causes huge performance increases in thread creation.

The downside of this patch is that it does lead to fragmentation in the
mmap-ed areas (visible via /proc/self/maps), such that some applications
that work fine under 2.4 kernels quickly run out of memory on any 2.6
kernel.

The problem is twofold:

  1) the free_area_cache is used to continue a search for memory where
     the last search ended.  Before the change new areas were always
     searched from the base address on.

     So now new small areas are cluttering holes of all sizes
     throughout the whole mmap-able region whereas before small holes
     tended to close holes near the base leaving holes far from the base
     large and available for larger requests.

  2) the free_area_cache also is set to the location of the last
     munmap-ed area so in scenarios where we allocate e.g.  five regions of
     1K each, then free regions 4 2 3 in this order the next request for 1K
     will be placed in the position of the old region 3, whereas before we
     appended it to the still active region 1, placing it at the location
     of the old region 2.  Before we had 1 free region of 2K, now we only
     get two free regions of 1K -> fragmentation.

The patch addresses thes issues by introducing yet another cache descriptor
cached_hole_size that contains the largest known hole size below the
current free_area_cache.  If a new request comes in the size is compared
against the cached_hole_size and if the request can be filled with a hole
below free_area_cache the search is started from the base instead.

The results look promising: Whereas 2.6.12-rc4 fragments quickly and my
(earlier posted) leakme.c test program terminates after 50000+ iterations
with 96 distinct and fragmented maps in /proc/self/maps it performs nicely
(as expected) with thread creation, Ingo's test_str02 with 20000 threads
requires 0.7s system time.

Taking out Ingo's patch (un-patch available per request) by basically
deleting all mentions of free_area_cache from the kernel and starting the
search for new memory always at the respective bases we observe: leakme
terminates successfully with 11 distinctive hardly fragmented areas in
/proc/self/maps but thread creating is gringdingly slow: 30+s(!) system
time for Ingo's test_str02 with 20000 threads.

Now - drumroll ;-) the appended patch works fine with leakme: it ends with
only 7 distinct areas in /proc/self/maps and also thread creation seems
sufficiently fast with 0.71s for 20000 threads.

Signed-off-by: Wolfgang Wander <wwc@rentec.com>
Credit-to: "Richard Purdie" <rpurdie@rpsys.net>
Signed-off-by: Ken Chen <kenneth.w.chen@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu> (partly)
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-21 18:46:16 -07:00

552 lines
15 KiB
C

/*
* linux/fs/binfmt_aout.c
*
* Copyright (C) 1991, 1992, 1996 Linus Torvalds
*/
#include <linux/module.h>
#include <linux/time.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/mman.h>
#include <linux/a.out.h>
#include <linux/errno.h>
#include <linux/signal.h>
#include <linux/string.h>
#include <linux/fs.h>
#include <linux/file.h>
#include <linux/stat.h>
#include <linux/fcntl.h>
#include <linux/ptrace.h>
#include <linux/user.h>
#include <linux/slab.h>
#include <linux/binfmts.h>
#include <linux/personality.h>
#include <linux/init.h>
#include <asm/system.h>
#include <asm/uaccess.h>
#include <asm/cacheflush.h>
static int load_aout_binary(struct linux_binprm *, struct pt_regs * regs);
static int load_aout_library(struct file*);
static int aout_core_dump(long signr, struct pt_regs * regs, struct file *file);
extern void dump_thread(struct pt_regs *, struct user *);
static struct linux_binfmt aout_format = {
.module = THIS_MODULE,
.load_binary = load_aout_binary,
.load_shlib = load_aout_library,
.core_dump = aout_core_dump,
.min_coredump = PAGE_SIZE
};
#define BAD_ADDR(x) ((unsigned long)(x) >= TASK_SIZE)
static int set_brk(unsigned long start, unsigned long end)
{
start = PAGE_ALIGN(start);
end = PAGE_ALIGN(end);
if (end > start) {
unsigned long addr;
down_write(&current->mm->mmap_sem);
addr = do_brk(start, end - start);
up_write(&current->mm->mmap_sem);
if (BAD_ADDR(addr))
return addr;
}
return 0;
}
/*
* These are the only things you should do on a core-file: use only these
* macros 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;
}
#define DUMP_WRITE(addr, nr) \
if (!dump_write(file, (void *)(addr), (nr))) \
goto end_coredump;
#define DUMP_SEEK(offset) \
if (file->f_op->llseek) { \
if (file->f_op->llseek(file,(offset),0) != (offset)) \
goto end_coredump; \
} else file->f_pos = (offset)
/*
* Routine writes a core dump image in the current directory.
* Currently only a stub-function.
*
* Note that setuid/setgid files won't make a core-dump if the uid/gid
* changed due to the set[u|g]id. It's enforced by the "current->mm->dumpable"
* field, which also makes sure the core-dumps won't be recursive if the
* dumping of the process results in another error..
*/
static int aout_core_dump(long signr, struct pt_regs * regs, struct file *file)
{
mm_segment_t fs;
int has_dumped = 0;
unsigned long dump_start, dump_size;
struct user dump;
#if defined(__alpha__)
# define START_DATA(u) (u.start_data)
#elif defined(__arm__)
# define START_DATA(u) ((u.u_tsize << PAGE_SHIFT) + u.start_code)
#elif defined(__sparc__)
# define START_DATA(u) (u.u_tsize)
#elif defined(__i386__) || defined(__mc68000__) || defined(__arch_um__)
# define START_DATA(u) (u.u_tsize << PAGE_SHIFT)
#endif
#ifdef __sparc__
# define START_STACK(u) ((regs->u_regs[UREG_FP]) & ~(PAGE_SIZE - 1))
#else
# define START_STACK(u) (u.start_stack)
#endif
fs = get_fs();
set_fs(KERNEL_DS);
has_dumped = 1;
current->flags |= PF_DUMPCORE;
strncpy(dump.u_comm, current->comm, sizeof(dump.u_comm));
#ifndef __sparc__
dump.u_ar0 = (void *)(((unsigned long)(&dump.regs)) - ((unsigned long)(&dump)));
#endif
dump.signal = signr;
dump_thread(regs, &dump);
/* If the size of the dump file exceeds the rlimit, then see what would happen
if we wrote the stack, but not the data area. */
#ifdef __sparc__
if ((dump.u_dsize+dump.u_ssize) >
current->signal->rlim[RLIMIT_CORE].rlim_cur)
dump.u_dsize = 0;
#else
if ((dump.u_dsize+dump.u_ssize+1) * PAGE_SIZE >
current->signal->rlim[RLIMIT_CORE].rlim_cur)
dump.u_dsize = 0;
#endif
/* Make sure we have enough room to write the stack and data areas. */
#ifdef __sparc__
if ((dump.u_ssize) >
current->signal->rlim[RLIMIT_CORE].rlim_cur)
dump.u_ssize = 0;
#else
if ((dump.u_ssize+1) * PAGE_SIZE >
current->signal->rlim[RLIMIT_CORE].rlim_cur)
dump.u_ssize = 0;
#endif
/* make sure we actually have a data and stack area to dump */
set_fs(USER_DS);
#ifdef __sparc__
if (!access_ok(VERIFY_READ, (void __user *)START_DATA(dump), dump.u_dsize))
dump.u_dsize = 0;
if (!access_ok(VERIFY_READ, (void __user *)START_STACK(dump), dump.u_ssize))
dump.u_ssize = 0;
#else
if (!access_ok(VERIFY_READ, (void __user *)START_DATA(dump), dump.u_dsize << PAGE_SHIFT))
dump.u_dsize = 0;
if (!access_ok(VERIFY_READ, (void __user *)START_STACK(dump), dump.u_ssize << PAGE_SHIFT))
dump.u_ssize = 0;
#endif
set_fs(KERNEL_DS);
/* struct user */
DUMP_WRITE(&dump,sizeof(dump));
/* Now dump all of the user data. Include malloced stuff as well */
#ifndef __sparc__
DUMP_SEEK(PAGE_SIZE);
#endif
/* now we start writing out the user space info */
set_fs(USER_DS);
/* Dump the data area */
if (dump.u_dsize != 0) {
dump_start = START_DATA(dump);
#ifdef __sparc__
dump_size = dump.u_dsize;
#else
dump_size = dump.u_dsize << PAGE_SHIFT;
#endif
DUMP_WRITE(dump_start,dump_size);
}
/* Now prepare to dump the stack area */
if (dump.u_ssize != 0) {
dump_start = START_STACK(dump);
#ifdef __sparc__
dump_size = dump.u_ssize;
#else
dump_size = dump.u_ssize << PAGE_SHIFT;
#endif
DUMP_WRITE(dump_start,dump_size);
}
/* Finally dump the task struct. Not be used by gdb, but could be useful */
set_fs(KERNEL_DS);
DUMP_WRITE(current,sizeof(*current));
end_coredump:
set_fs(fs);
return has_dumped;
}
/*
* create_aout_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 unsigned long __user *create_aout_tables(char __user *p, struct linux_binprm * bprm)
{
char __user * __user *argv;
char __user * __user *envp;
unsigned long __user *sp;
int argc = bprm->argc;
int envc = bprm->envc;
sp = (void __user *)((-(unsigned long)sizeof(char *)) & (unsigned long) p);
#ifdef __sparc__
/* This imposes the proper stack alignment for a new process. */
sp = (void __user *) (((unsigned long) sp) & ~7);
if ((envc+argc+3)&1) --sp;
#endif
#ifdef __alpha__
/* whee.. test-programs are so much fun. */
put_user(0, --sp);
put_user(0, --sp);
if (bprm->loader) {
put_user(0, --sp);
put_user(0x3eb, --sp);
put_user(bprm->loader, --sp);
put_user(0x3ea, --sp);
}
put_user(bprm->exec, --sp);
put_user(0x3e9, --sp);
#endif
sp -= envc+1;
envp = (char __user * __user *) sp;
sp -= argc+1;
argv = (char __user * __user *) sp;
#if defined(__i386__) || defined(__mc68000__) || defined(__arm__) || defined(__arch_um__)
put_user((unsigned long) envp,--sp);
put_user((unsigned long) argv,--sp);
#endif
put_user(argc,--sp);
current->mm->arg_start = (unsigned long) p;
while (argc-->0) {
char c;
put_user(p,argv++);
do {
get_user(c,p++);
} while (c);
}
put_user(NULL,argv);
current->mm->arg_end = current->mm->env_start = (unsigned long) p;
while (envc-->0) {
char c;
put_user(p,envp++);
do {
get_user(c,p++);
} while (c);
}
put_user(NULL,envp);
current->mm->env_end = (unsigned long) p;
return sp;
}
/*
* These are the functions used to load a.out style executables and shared
* libraries. There is no binary dependent code anywhere else.
*/
static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs)
{
struct exec ex;
unsigned long error;
unsigned long fd_offset;
unsigned long rlim;
int retval;
ex = *((struct exec *) bprm->buf); /* exec-header */
if ((N_MAGIC(ex) != ZMAGIC && N_MAGIC(ex) != OMAGIC &&
N_MAGIC(ex) != QMAGIC && N_MAGIC(ex) != NMAGIC) ||
N_TRSIZE(ex) || N_DRSIZE(ex) ||
i_size_read(bprm->file->f_dentry->d_inode) < ex.a_text+ex.a_data+N_SYMSIZE(ex)+N_TXTOFF(ex)) {
return -ENOEXEC;
}
fd_offset = N_TXTOFF(ex);
/* Check initial limits. This avoids letting people circumvent
* size limits imposed on them by creating programs with large
* arrays in the data or bss.
*/
rlim = current->signal->rlim[RLIMIT_DATA].rlim_cur;
if (rlim >= RLIM_INFINITY)
rlim = ~0;
if (ex.a_data + ex.a_bss > rlim)
return -ENOMEM;
/* Flush all traces of the currently running executable */
retval = flush_old_exec(bprm);
if (retval)
return retval;
/* OK, This is the point of no return */
#if defined(__alpha__)
SET_AOUT_PERSONALITY(bprm, ex);
#elif defined(__sparc__)
set_personality(PER_SUNOS);
#if !defined(__sparc_v9__)
memcpy(&current->thread.core_exec, &ex, sizeof(struct exec));
#endif
#else
set_personality(PER_LINUX);
#endif
current->mm->end_code = ex.a_text +
(current->mm->start_code = N_TXTADDR(ex));
current->mm->end_data = ex.a_data +
(current->mm->start_data = N_DATADDR(ex));
current->mm->brk = ex.a_bss +
(current->mm->start_brk = N_BSSADDR(ex));
current->mm->free_area_cache = current->mm->mmap_base;
current->mm->cached_hole_size = 0;
set_mm_counter(current->mm, rss, 0);
current->mm->mmap = NULL;
compute_creds(bprm);
current->flags &= ~PF_FORKNOEXEC;
#ifdef __sparc__
if (N_MAGIC(ex) == NMAGIC) {
loff_t pos = fd_offset;
/* Fuck me plenty... */
/* <AOL></AOL> */
down_write(&current->mm->mmap_sem);
error = do_brk(N_TXTADDR(ex), ex.a_text);
up_write(&current->mm->mmap_sem);
bprm->file->f_op->read(bprm->file, (char *) N_TXTADDR(ex),
ex.a_text, &pos);
down_write(&current->mm->mmap_sem);
error = do_brk(N_DATADDR(ex), ex.a_data);
up_write(&current->mm->mmap_sem);
bprm->file->f_op->read(bprm->file, (char *) N_DATADDR(ex),
ex.a_data, &pos);
goto beyond_if;
}
#endif
if (N_MAGIC(ex) == OMAGIC) {
unsigned long text_addr, map_size;
loff_t pos;
text_addr = N_TXTADDR(ex);
#if defined(__alpha__) || defined(__sparc__)
pos = fd_offset;
map_size = ex.a_text+ex.a_data + PAGE_SIZE - 1;
#else
pos = 32;
map_size = ex.a_text+ex.a_data;
#endif
down_write(&current->mm->mmap_sem);
error = do_brk(text_addr & PAGE_MASK, map_size);
up_write(&current->mm->mmap_sem);
if (error != (text_addr & PAGE_MASK)) {
send_sig(SIGKILL, current, 0);
return error;
}
error = bprm->file->f_op->read(bprm->file,
(char __user *)text_addr,
ex.a_text+ex.a_data, &pos);
if ((signed long)error < 0) {
send_sig(SIGKILL, current, 0);
return error;
}
flush_icache_range(text_addr, text_addr+ex.a_text+ex.a_data);
} else {
static unsigned long error_time, error_time2;
if ((ex.a_text & 0xfff || ex.a_data & 0xfff) &&
(N_MAGIC(ex) != NMAGIC) && (jiffies-error_time2) > 5*HZ)
{
printk(KERN_NOTICE "executable not page aligned\n");
error_time2 = jiffies;
}
if ((fd_offset & ~PAGE_MASK) != 0 &&
(jiffies-error_time) > 5*HZ)
{
printk(KERN_WARNING
"fd_offset is not page aligned. Please convert program: %s\n",
bprm->file->f_dentry->d_name.name);
error_time = jiffies;
}
if (!bprm->file->f_op->mmap||((fd_offset & ~PAGE_MASK) != 0)) {
loff_t pos = fd_offset;
down_write(&current->mm->mmap_sem);
do_brk(N_TXTADDR(ex), ex.a_text+ex.a_data);
up_write(&current->mm->mmap_sem);
bprm->file->f_op->read(bprm->file,
(char __user *)N_TXTADDR(ex),
ex.a_text+ex.a_data, &pos);
flush_icache_range((unsigned long) N_TXTADDR(ex),
(unsigned long) N_TXTADDR(ex) +
ex.a_text+ex.a_data);
goto beyond_if;
}
down_write(&current->mm->mmap_sem);
error = do_mmap(bprm->file, N_TXTADDR(ex), ex.a_text,
PROT_READ | PROT_EXEC,
MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE | MAP_EXECUTABLE,
fd_offset);
up_write(&current->mm->mmap_sem);
if (error != N_TXTADDR(ex)) {
send_sig(SIGKILL, current, 0);
return error;
}
down_write(&current->mm->mmap_sem);
error = do_mmap(bprm->file, N_DATADDR(ex), ex.a_data,
PROT_READ | PROT_WRITE | PROT_EXEC,
MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE | MAP_EXECUTABLE,
fd_offset + ex.a_text);
up_write(&current->mm->mmap_sem);
if (error != N_DATADDR(ex)) {
send_sig(SIGKILL, current, 0);
return error;
}
}
beyond_if:
set_binfmt(&aout_format);
retval = set_brk(current->mm->start_brk, current->mm->brk);
if (retval < 0) {
send_sig(SIGKILL, current, 0);
return retval;
}
retval = setup_arg_pages(bprm, STACK_TOP, EXSTACK_DEFAULT);
if (retval < 0) {
/* Someone check-me: is this error path enough? */
send_sig(SIGKILL, current, 0);
return retval;
}
current->mm->start_stack =
(unsigned long) create_aout_tables((char __user *) bprm->p, bprm);
#ifdef __alpha__
regs->gp = ex.a_gpvalue;
#endif
start_thread(regs, ex.a_entry, current->mm->start_stack);
if (unlikely(current->ptrace & PT_PTRACED)) {
if (current->ptrace & PT_TRACE_EXEC)
ptrace_notify ((PTRACE_EVENT_EXEC << 8) | SIGTRAP);
else
send_sig(SIGTRAP, current, 0);
}
return 0;
}
static int load_aout_library(struct file *file)
{
struct inode * inode;
unsigned long bss, start_addr, len;
unsigned long error;
int retval;
struct exec ex;
inode = file->f_dentry->d_inode;
retval = -ENOEXEC;
error = kernel_read(file, 0, (char *) &ex, sizeof(ex));
if (error != sizeof(ex))
goto out;
/* We come in here for the regular a.out style of shared libraries */
if ((N_MAGIC(ex) != ZMAGIC && N_MAGIC(ex) != QMAGIC) || N_TRSIZE(ex) ||
N_DRSIZE(ex) || ((ex.a_entry & 0xfff) && N_MAGIC(ex) == ZMAGIC) ||
i_size_read(inode) < ex.a_text+ex.a_data+N_SYMSIZE(ex)+N_TXTOFF(ex)) {
goto out;
}
if (N_FLAGS(ex))
goto out;
/* For QMAGIC, the starting address is 0x20 into the page. We mask
this off to get the starting address for the page */
start_addr = ex.a_entry & 0xfffff000;
if ((N_TXTOFF(ex) & ~PAGE_MASK) != 0) {
static unsigned long error_time;
loff_t pos = N_TXTOFF(ex);
if ((jiffies-error_time) > 5*HZ)
{
printk(KERN_WARNING
"N_TXTOFF is not page aligned. Please convert library: %s\n",
file->f_dentry->d_name.name);
error_time = jiffies;
}
down_write(&current->mm->mmap_sem);
do_brk(start_addr, ex.a_text + ex.a_data + ex.a_bss);
up_write(&current->mm->mmap_sem);
file->f_op->read(file, (char __user *)start_addr,
ex.a_text + ex.a_data, &pos);
flush_icache_range((unsigned long) start_addr,
(unsigned long) start_addr + ex.a_text + ex.a_data);
retval = 0;
goto out;
}
/* Now use mmap to map the library into memory. */
down_write(&current->mm->mmap_sem);
error = do_mmap(file, start_addr, ex.a_text + ex.a_data,
PROT_READ | PROT_WRITE | PROT_EXEC,
MAP_FIXED | MAP_PRIVATE | MAP_DENYWRITE,
N_TXTOFF(ex));
up_write(&current->mm->mmap_sem);
retval = error;
if (error != start_addr)
goto out;
len = PAGE_ALIGN(ex.a_text + ex.a_data);
bss = ex.a_text + ex.a_data + ex.a_bss;
if (bss > len) {
down_write(&current->mm->mmap_sem);
error = do_brk(start_addr + len, bss - len);
up_write(&current->mm->mmap_sem);
retval = error;
if (error != start_addr + len)
goto out;
}
retval = 0;
out:
return retval;
}
static int __init init_aout_binfmt(void)
{
return register_binfmt(&aout_format);
}
static void __exit exit_aout_binfmt(void)
{
unregister_binfmt(&aout_format);
}
core_initcall(init_aout_binfmt);
module_exit(exit_aout_binfmt);
MODULE_LICENSE("GPL");