cregit-Linux how code gets into the kernel

Release 4.7 kernel/acct.c

Directory: kernel
/*
 *  linux/kernel/acct.c
 *
 *  BSD Process Accounting for Linux
 *
 *  Author: Marco van Wieringen <mvw@planets.elm.net>
 *
 *  Some code based on ideas and code from:
 *  Thomas K. Dyas <tdyas@eden.rutgers.edu>
 *
 *  This file implements BSD-style process accounting. Whenever any
 *  process exits, an accounting record of type "struct acct" is
 *  written to the file specified with the acct() system call. It is
 *  up to user-level programs to do useful things with the accounting
 *  log. The kernel just provides the raw accounting information.
 *
 * (C) Copyright 1995 - 1997 Marco van Wieringen - ELM Consultancy B.V.
 *
 *  Plugged two leaks. 1) It didn't return acct_file into the free_filps if
 *  the file happened to be read-only. 2) If the accounting was suspended
 *  due to the lack of space it happily allowed to reopen it and completely
 *  lost the old acct_file. 3/10/98, Al Viro.
 *
 *  Now we silently close acct_file on attempt to reopen. Cleaned sys_acct().
 *  XTerms and EMACS are manifestations of pure evil. 21/10/98, AV.
 *
 *  Fixed a nasty interaction with with sys_umount(). If the accointing
 *  was suspeneded we failed to stop it on umount(). Messy.
 *  Another one: remount to readonly didn't stop accounting.
 *      Question: what should we do if we have CAP_SYS_ADMIN but not
 *  CAP_SYS_PACCT? Current code does the following: umount returns -EBUSY
 *  unless we are messing with the root. In that case we are getting a
 *  real mess with do_remount_sb(). 9/11/98, AV.
 *
 *  Fixed a bunch of races (and pair of leaks). Probably not the best way,
 *  but this one obviously doesn't introduce deadlocks. Later. BTW, found
 *  one race (and leak) in BSD implementation.
 *  OK, that's better. ANOTHER race and leak in BSD variant. There always
 *  is one more bug... 10/11/98, AV.
 *
 *      Oh, fsck... Oopsable SMP race in do_process_acct() - we must hold
 * ->mmap_sem to walk the vma list of current->mm. Nasty, since it leaks
 * a struct file opened for write. Fixed. 2/6/2000, AV.
 */

#include <linux/mm.h>
#include <linux/slab.h>
#include <linux/acct.h>
#include <linux/capability.h>
#include <linux/file.h>
#include <linux/tty.h>
#include <linux/security.h>
#include <linux/vfs.h>
#include <linux/jiffies.h>
#include <linux/times.h>
#include <linux/syscalls.h>
#include <linux/mount.h>
#include <linux/uaccess.h>
#include <asm/div64.h>
#include <linux/blkdev.h> /* sector_div */
#include <linux/pid_namespace.h>
#include <linux/fs_pin.h>

/*
 * These constants control the amount of freespace that suspend and
 * resume the process accounting system, and the time delay between
 * each check.
 * Turned into sysctl-controllable parameters. AV, 12/11/98
 */


int acct_parm[3] = {4, 2, 30};

#define RESUME		(acct_parm[0])	
/* >foo% free space - resume */

#define SUSPEND		(acct_parm[1])	
/* <foo% free space - suspend */

#define ACCT_TIMEOUT	(acct_parm[2])	
/* foo second timeout between checks */

/*
 * External references and all of the globals.
 */


struct bsd_acct_struct {
	
struct fs_pin		pin;
	
atomic_long_t		count;
	
struct rcu_head		rcu;
	
struct mutex		lock;
	
int			active;
	
unsigned long		needcheck;
	
struct file		*file;
	
struct pid_namespace	*ns;
	
struct work_struct	work;
	
struct completion	done;
};

static void do_acct_process(struct bsd_acct_struct *acct);

/*
 * Check the amount of free space and suspend/resume accordingly.
 */

static int check_free_space(struct bsd_acct_struct *acct) { struct kstatfs sbuf; if (time_is_before_jiffies(acct->needcheck)) goto out; /* May block */ if (vfs_statfs(&acct->file->f_path, &sbuf)) goto out; if (acct->active) { u64 suspend = sbuf.f_blocks * SUSPEND; do_div(suspend, 100); if (sbuf.f_bavail <= suspend) { acct->active = 0; pr_info("Process accounting paused\n"); } } else { u64 resume = sbuf.f_blocks * RESUME; do_div(resume, 100); if (sbuf.f_bavail >= resume) { acct->active = 1; pr_info("Process accounting resumed\n"); } } acct->needcheck = jiffies + ACCT_TIMEOUT*HZ; out: return acct->active; }

Contributors

PersonTokensPropCommitsCommitProp
pre-gitpre-git7047.30%538.46%
al viroal viro4329.05%215.38%
andrew mortonandrew morton1610.81%17.69%
pavel emelianovpavel emelianov117.43%17.69%
bob millerbob miller42.70%17.69%
ionut alexaionut alexa21.35%17.69%
christoph hellwigchristoph hellwig10.68%17.69%
josef 'jeff' sipekjosef 'jeff' sipek10.68%17.69%
Total148100.00%13100.00%


static void acct_put(struct bsd_acct_struct *p) { if (atomic_long_dec_and_test(&p->count)) kfree_rcu(p, rcu); }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro28100.00%1100.00%
Total28100.00%1100.00%


static inline struct bsd_acct_struct *to_acct(struct fs_pin *p) { return p ? container_of(p, struct bsd_acct_struct, pin) : NULL; }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro29100.00%1100.00%
Total29100.00%1100.00%


static struct bsd_acct_struct *acct_get(struct pid_namespace *ns) { struct bsd_acct_struct *res; again: smp_rmb(); rcu_read_lock(); res = to_acct(ACCESS_ONCE(ns->bacct)); if (!res) { rcu_read_unlock(); return NULL; } if (!atomic_long_inc_not_zero(&res->count)) { rcu_read_unlock(); cpu_relax(); goto again; } rcu_read_unlock(); mutex_lock(&res->lock); if (res != to_acct(ACCESS_ONCE(ns->bacct))) { mutex_unlock(&res->lock); acct_put(res); goto again; } return res; }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro10588.24%654.55%
pavel emelianovpavel emelianov75.88%327.27%
bob millerbob miller65.04%19.09%
adrian bunkadrian bunk10.84%19.09%
Total119100.00%11100.00%


static void acct_pin_kill(struct fs_pin *pin) { struct bsd_acct_struct *acct = to_acct(pin); mutex_lock(&acct->lock); do_acct_process(acct); schedule_work(&acct->work); wait_for_completion(&acct->done); cmpxchg(&acct->ns->bacct, pin, NULL); mutex_unlock(&acct->lock); pin_remove(pin); acct_put(acct); }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro7793.90%880.00%
pavel emelianovpavel emelianov44.88%110.00%
bob millerbob miller11.22%110.00%
Total82100.00%10100.00%


static void close_work(struct work_struct *work) { struct bsd_acct_struct *acct = container_of(work, struct bsd_acct_struct, work); struct file *file = acct->file; if (file->f_op->flush) file->f_op->flush(file, NULL); __fput_sync(file); complete(&acct->done); }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro67100.00%5100.00%
Total67100.00%5100.00%


static int acct_on(struct filename *pathname) { struct file *file; struct vfsmount *mnt, *internal; struct pid_namespace *ns = task_active_pid_ns(current); struct bsd_acct_struct *acct; struct fs_pin *old; int err; acct = kzalloc(sizeof(struct bsd_acct_struct), GFP_KERNEL); if (!acct) return -ENOMEM; /* Difference from BSD - they don't do O_APPEND */ file = file_open_name(pathname, O_WRONLY|O_APPEND|O_LARGEFILE, 0); if (IS_ERR(file)) { kfree(acct); return PTR_ERR(file); } if (!S_ISREG(file_inode(file)->i_mode)) { kfree(acct); filp_close(file, NULL); return -EACCES; } if (!(file->f_mode & FMODE_CAN_WRITE)) { kfree(acct); filp_close(file, NULL); return -EIO; } internal = mnt_clone_internal(&file->f_path); if (IS_ERR(internal)) { kfree(acct); filp_close(file, NULL); return PTR_ERR(internal); } err = mnt_want_write(internal); if (err) { mntput(internal); kfree(acct); filp_close(file, NULL); return err; } mnt = file->f_path.mnt; file->f_path.mnt = internal; atomic_long_set(&acct->count, 1); init_fs_pin(&acct->pin, acct_pin_kill); acct->file = file; acct->needcheck = jiffies; acct->ns = ns; mutex_init(&acct->lock); INIT_WORK(&acct->work, close_work); init_completion(&acct->done); mutex_lock_nested(&acct->lock, 1); /* nobody has seen it yet */ pin_insert(&acct->pin, mnt); rcu_read_lock(); old = xchg(&ns->bacct, &acct->pin); mutex_unlock(&acct->lock); pin_kill(old); mnt_drop_write(mnt); mntput(mnt); return 0; }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro28976.05%952.94%
pre-gitpre-git4511.84%211.76%
bob millerbob miller246.32%211.76%
pavel emelianovpavel emelianov174.47%15.88%
jeff laytonjeff layton20.53%15.88%
peter staubachpeter staubach20.53%15.88%
josef 'jeff' sipekjosef 'jeff' sipek10.26%15.88%
Total380100.00%17100.00%

static DEFINE_MUTEX(acct_on_mutex); /** * sys_acct - enable/disable process accounting * @name: file name for accounting records or NULL to shutdown accounting * * Returns 0 for success or negative errno values for failure. * * sys_acct() is the only system call needed to implement process * accounting. It takes the name of the file where accounting records * should be written. If the filename is NULL, accounting will be * shutdown. */ SYSCALL_DEFINE1(acct, const char __user *, name) { int error = 0; if (!capable(CAP_SYS_PACCT)) return -EPERM; if (name) { struct filename *tmp = getname(name); if (IS_ERR(tmp)) return PTR_ERR(tmp); mutex_lock(&acct_on_mutex); error = acct_on(tmp); mutex_unlock(&acct_on_mutex); putname(tmp); } else { rcu_read_lock(); pin_kill(task_active_pid_ns(current)->bacct); } return error; }
void acct_exit_ns(struct pid_namespace *ns) { rcu_read_lock(); pin_kill(ns->bacct); }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro1785.00%250.00%
vitaliy gusevvitaliy gusev210.00%125.00%
pre-gitpre-git15.00%125.00%
Total20100.00%4100.00%

/* * encode an unsigned long into a comp_t * * This routine has been adopted from the encode_comp_t() function in * the kern_acct.c file of the FreeBSD operating system. The encoding * is a 13-bit fraction with a 3-bit (base 8) exponent. */ #define MANTSIZE 13 /* 13 bit mantissa. */ #define EXPSIZE 3 /* Base 8 (3 bit) exponent. */ #define MAXFRACT ((1 << MANTSIZE) - 1) /* Maximum fractional value. */
static comp_t encode_comp_t(unsigned long value) { int exp, rnd; exp = rnd = 0; while (value > MAXFRACT) { rnd = value & (1 << (EXPSIZE - 1)); /* Round up? */ value >>= EXPSIZE; /* Base 8 exponent == 3 bit shift. */ exp++; } /* * If we need to round up, do it (and handle overflow correctly). */ if (rnd && (++value > MAXFRACT)) { value >>= EXPSIZE; exp++; } /* * Clean it up and polish it off. */ exp <<= MANTSIZE; /* Shift the exponent into place */ exp += value; /* and add on the mantissa. */ return exp; }

Contributors

PersonTokensPropCommitsCommitProp
pre-gitpre-git8597.70%150.00%
daniel walkerdaniel walker22.30%150.00%
Total87100.00%2100.00%

#if ACCT_VERSION == 1 || ACCT_VERSION == 2 /* * encode an u64 into a comp2_t (24 bits) * * Format: 5 bit base 2 exponent, 20 bits mantissa. * The leading bit of the mantissa is not stored, but implied for * non-zero exponents. * Largest encodable value is 50 bits. */ #define MANTSIZE2 20 /* 20 bit mantissa. */ #define EXPSIZE2 5 /* 5 bit base 2 exponent. */ #define MAXFRACT2 ((1ul << MANTSIZE2) - 1) /* Maximum fractional value. */ #define MAXEXP2 ((1 << EXPSIZE2) - 1) /* Maximum exponent. */
static comp2_t encode_comp2_t(u64 value) { int exp, rnd; exp = (value > (MAXFRACT2>>1)); rnd = 0; while (value > MAXFRACT2) { rnd = value & 1; value >>= 1; exp++; } /* * If we need to round up, do it (and handle overflow correctly). */ if (rnd && (++value > MAXFRACT2)) { value >>= 1; exp++; } if (exp > MAXEXP2) { /* Overflow. Return largest representable number instead. */ return (1ul << (MANTSIZE2+EXPSIZE2-1)) - 1; } else { return (value & (MAXFRACT2>>1)) | (exp << (MANTSIZE2-1)); } }

Contributors

PersonTokensPropCommitsCommitProp
tim schmielautim schmielau11999.17%150.00%
daniel walkerdaniel walker10.83%150.00%
Total120100.00%2100.00%

#endif #if ACCT_VERSION == 3 /* * encode an u64 into a 32 bit IEEE float */
static u32 encode_float(u64 value) { unsigned exp = 190; unsigned u; if (value == 0) return 0; while ((s64)value > 0) { value <<= 1; exp--; } u = (u32)(value >> 40) & 0x7fffffu; return u | (exp << 23); }

Contributors

PersonTokensPropCommitsCommitProp
tim schmielautim schmielau66100.00%1100.00%
Total66100.00%1100.00%

#endif /* * Write an accounting entry for an exiting process * * The acct_process() call is the workhorse of the process * accounting system. The struct acct is built here and then written * into the accounting file. This function should only be called from * do_exit() or when switching to a different output file. */
static void fill_ac(acct_t *ac) { struct pacct_struct *pacct = &current->signal->pacct; u64 elapsed, run_time; struct tty_struct *tty; /* * Fill the accounting struct with the needed info as recorded * by the different kernel functions. */ memset(ac, 0, sizeof(acct_t)); ac->ac_version = ACCT_VERSION | ACCT_BYTEORDER; strlcpy(ac->ac_comm, current->comm, sizeof(ac->ac_comm)); /* calculate run_time in nsec*/ run_time = ktime_get_ns(); run_time -= current->group_leader->start_time; /* convert nsec -> AHZ */ elapsed = nsec_to_AHZ(run_time); #if ACCT_VERSION == 3 ac->ac_etime = encode_float(elapsed); #else ac->ac_etime = encode_comp_t(elapsed < (unsigned long) -1l ? (unsigned long) elapsed : (unsigned long) -1l); #endif #if ACCT_VERSION == 1 || ACCT_VERSION == 2 { /* new enlarged etime field */ comp2_t etime = encode_comp2_t(elapsed); ac->ac_etime_hi = etime >> 16; ac->ac_etime_lo = (u16) etime; } #endif do_div(elapsed, AHZ); ac->ac_btime = get_seconds() - elapsed; #if ACCT_VERSION==2 ac->ac_ahz = AHZ; #endif spin_lock_irq(&current->sighand->siglock); tty = current->signal->tty; /* Safe as we hold the siglock */ ac->ac_tty = tty ? old_encode_dev(tty_devnum(tty)) : 0; ac->ac_utime = encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_utime))); ac->ac_stime = encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_stime))); ac->ac_flag = pacct->ac_flag; ac->ac_mem = encode_comp_t(pacct->ac_mem); ac->ac_minflt = encode_comp_t(pacct->ac_minflt); ac->ac_majflt = encode_comp_t(pacct->ac_majflt); ac->ac_exitcode = pacct->ac_exitcode; spin_unlock_irq(&current->sighand->siglock); }

Contributors

PersonTokensPropCommitsCommitProp
tim schmielautim schmielau12537.20%414.81%
kaigai koheikaigai kohei8926.49%414.81%
pre-gitpre-git5817.26%518.52%
al viroal viro278.04%311.11%
oleg nesterovoleg nesterov113.27%13.70%
peter zijlstrapeter zijlstra61.79%13.70%
ben collinsben collins51.49%13.70%
alan coxalan cox51.49%27.41%
thomas gleixnerthomas gleixner30.89%13.70%
linus torvaldslinus torvalds20.60%13.70%
andrew mortonandrew morton20.60%27.41%
john stultzjohn stultz20.60%13.70%
hirofumi ogawahirofumi ogawa10.30%13.70%
Total336100.00%27100.00%

/* * do_acct_process does all actual work. Caller holds the reference to file. */
static void do_acct_process(struct bsd_acct_struct *acct) { acct_t ac; unsigned long flim; const struct cred *orig_cred; struct file *file = acct->file; /* * Accounting records are not subject to resource limits. */ flim = current->signal->rlim[RLIMIT_FSIZE].rlim_cur; current->signal->rlim[RLIMIT_FSIZE].rlim_cur = RLIM_INFINITY; /* Perform file operations on behalf of whoever enabled accounting */ orig_cred = override_creds(file->f_cred); /* * First check to see if there is enough free_space to continue * the process accounting system. */ if (!check_free_space(acct)) goto out; fill_ac(&ac); /* we really need to bite the bullet and change layout */ ac.ac_uid = from_kuid_munged(file->f_cred->user_ns, orig_cred->uid); ac.ac_gid = from_kgid_munged(file->f_cred->user_ns, orig_cred->gid); #if ACCT_VERSION == 1 || ACCT_VERSION == 2 /* backward-compatible 16 bit fields */ ac.ac_uid16 = ac.ac_uid; ac.ac_gid16 = ac.ac_gid; #endif #if ACCT_VERSION == 3 { struct pid_namespace *ns = acct->ns; ac.ac_pid = task_tgid_nr_ns(current, ns); rcu_read_lock(); ac.ac_ppid = task_tgid_nr_ns(rcu_dereference(current->real_parent), ns); rcu_read_unlock(); } #endif /* * Get freeze protection. If the fs is frozen, just skip the write * as we could deadlock the system otherwise. */ if (file_start_write_trylock(file)) { /* it's been opened O_APPEND, so position is irrelevant */ loff_t pos = 0; __kernel_write(file, (char *)&ac, sizeof(acct_t), &pos); file_end_write(file); } out: current->signal->rlim[RLIMIT_FSIZE].rlim_cur = flim; revert_creds(orig_cred); }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro20979.47%430.77%
pre-gitpre-git186.84%215.38%
ying xueying xue114.18%17.69%
alan coxalan cox114.18%17.69%
michal schmidtmichal schmidt51.90%17.69%
jan karajan kara51.90%17.69%
roland mcgrathroland mcgrath20.76%17.69%
tim schmielautim schmielau10.38%17.69%
kaigai koheikaigai kohei10.38%17.69%
Total263100.00%13100.00%

/** * acct_collect - collect accounting information into pacct_struct * @exitcode: task exit code * @group_dead: not 0, if this thread is the last one in the process. */
void acct_collect(long exitcode, int group_dead) { struct pacct_struct *pacct = &current->signal->pacct; cputime_t utime, stime; unsigned long vsize = 0; if (group_dead && current->mm) { struct vm_area_struct *vma; down_read(&current->mm->mmap_sem); vma = current->mm->mmap; while (vma) { vsize += vma->vm_end - vma->vm_start; vma = vma->vm_next; } up_read(&current->mm->mmap_sem); } spin_lock_irq(&current->sighand->siglock); if (group_dead) pacct->ac_mem = vsize / 1024; if (thread_group_leader(current)) { pacct->ac_exitcode = exitcode; if (current->flags & PF_FORKNOEXEC) pacct->ac_flag |= AFORK; } if (current->flags & PF_SUPERPRIV) pacct->ac_flag |= ASU; if (current->flags & PF_DUMPCORE) pacct->ac_flag |= ACORE; if (current->flags & PF_SIGNALED) pacct->ac_flag |= AXSIG; task_cputime(current, &utime, &stime); pacct->ac_utime += utime; pacct->ac_stime += stime; pacct->ac_minflt += current->min_flt; pacct->ac_majflt += current->maj_flt; spin_unlock_irq(&current->sighand->siglock); }

Contributors

PersonTokensPropCommitsCommitProp
kaigai koheikaigai kohei22292.12%350.00%
frederic weisbeckerfrederic weisbecker166.64%116.67%
martin schwidefskymartin schwidefsky20.83%116.67%
pre-gitpre-git10.41%116.67%
Total241100.00%6100.00%


static void slow_acct_process(struct pid_namespace *ns) { for ( ; ns; ns = ns->parent) { struct bsd_acct_struct *acct = acct_get(ns); if (acct) { do_acct_process(acct); mutex_unlock(&acct->lock); acct_put(acct); } } }

Contributors

PersonTokensPropCommitsCommitProp
al viroal viro2746.55%327.27%
pre-gitpre-git1118.97%218.18%
pavel emelianovpavel emelianov1017.24%218.18%
andrew mortonandrew morton46.90%19.09%
bob millerbob miller35.17%19.09%
randy dunlaprandy dunlap23.45%19.09%
ingo molnaringo molnar11.72%19.09%
Total58100.00%11100.00%

/** * acct_process * * handles process accounting for an exiting task */
void acct_process(void) { struct pid_namespace *ns; /* * This loop is safe lockless, since current is still * alive and holds its namespace, which in turn holds * its parent. */ for (ns = task_active_pid_ns(current); ns != NULL; ns = ns->parent) { if (ns->bacct) break; } if (unlikely(ns)) slow_acct_process(ns); }

Contributors

PersonTokensPropCommitsCommitProp
pavel emelianovpavel emelianov3667.92%250.00%
al viroal viro1732.08%250.00%
Total53100.00%4100.00%


Overall Contributors

PersonTokensPropCommitsCommitProp
al viroal viro107044.53%2426.97%
pre-gitpre-git35914.94%910.11%
tim schmielautim schmielau35614.81%44.49%
kaigai koheikaigai kohei31313.03%44.49%
pavel emelianovpavel emelianov974.04%66.74%
bob millerbob miller441.83%22.25%
andrew mortonandrew morton291.21%44.49%
frederic weisbeckerfrederic weisbecker160.67%11.12%
alan coxalan cox160.67%33.37%
ying xueying xue110.46%11.12%
oleg nesterovoleg nesterov110.46%11.12%
heiko carstensheiko carstens80.33%11.12%
peter zijlstrapeter zijlstra60.25%11.12%
jan karajan kara50.21%11.12%
randy dunlaprandy dunlap50.21%22.25%
ben collinsben collins50.21%11.12%
linus torvaldslinus torvalds50.21%22.25%
michal schmidtmichal schmidt50.21%11.12%
jeff laytonjeff layton40.17%22.25%
christoph hellwigchristoph hellwig40.17%22.25%
greg kroah-hartmangreg kroah-hartman30.12%11.12%
daniel walkerdaniel walker30.12%11.12%
ionut alexaionut alexa30.12%11.12%
arnd bergmannarnd bergmann30.12%11.12%
thomas gleixnerthomas gleixner30.12%11.12%
josef 'jeff' sipekjosef 'jeff' sipek20.08%11.12%
vitaliy gusevvitaliy gusev20.08%11.12%
peter staubachpeter staubach20.08%11.12%
eric pariseric paris20.08%11.12%
martin schwidefskymartin schwidefsky20.08%11.12%
john stultzjohn stultz20.08%11.12%
roland mcgrathroland mcgrath20.08%11.12%
ingo molnaringo molnar20.08%22.25%
hirofumi ogawahirofumi ogawa10.04%11.12%
paul mcquadepaul mcquade10.04%11.12%
adrian bunkadrian bunk10.04%11.12%
Total2403100.00%89100.00%
Directory: kernel
Information contained on this website is for historical information purposes only and does not indicate or represent copyright ownership.
{% endraw %}