proc: revert /proc/<pid>/maps [stack:TID] annotation

Commit b76437579d ("procfs: mark thread stack correctly in
proc/<pid>/maps") added [stack:TID] annotation to /proc/<pid>/maps.

Finding the task of a stack VMA requires walking the entire thread list,
turning this into quadratic behavior: a thousand threads means a
thousand stacks, so the rendering of /proc/<pid>/maps needs to look at a
million combinations.

The cost is not in proportion to the usefulness as described in the
patch.

Drop the [stack:TID] annotation to make /proc/<pid>/maps (and
/proc/<pid>/numa_maps) usable again for higher thread counts.

The [stack] annotation inside /proc/<pid>/task/<tid>/maps is retained, as
identifying the stack VMA there is an O(1) operation.

Siddesh said:
 "The end users needed a way to identify thread stacks programmatically and
  there wasn't a way to do that.  I'm afraid I no longer remember (or have
  access to the resources that would aid my memory since I changed
  employers) the details of their requirement.  However, I did do this on my
  own time because I thought it was an interesting project for me and nobody
  really gave any feedback then as to its utility, so as far as I am
  concerned you could roll back the main thread maps information since the
  information is available in the thread-specific files"

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>
Cc: Siddhesh Poyarekar <siddhesh.poyarekar@gmail.com>
Cc: Shaohua Li <shli@fb.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This commit is contained in:
Johannes Weiner 2016-02-02 16:57:29 -08:00 committed by Linus Torvalds
parent 5c2ff95e41
commit 65376df582
5 changed files with 47 additions and 105 deletions

View file

@ -356,7 +356,7 @@ address perms offset dev inode pathname
a7cb1000-a7cb2000 ---p 00000000 00:00 0 a7cb1000-a7cb2000 ---p 00000000 00:00 0
a7cb2000-a7eb2000 rw-p 00000000 00:00 0 a7cb2000-a7eb2000 rw-p 00000000 00:00 0
a7eb2000-a7eb3000 ---p 00000000 00:00 0 a7eb2000-a7eb3000 ---p 00000000 00:00 0
a7eb3000-a7ed5000 rw-p 00000000 00:00 0 [stack:1001] a7eb3000-a7ed5000 rw-p 00000000 00:00 0
a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6 a7ed5000-a8008000 r-xp 00000000 03:00 4222 /lib/libc.so.6
a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6 a8008000-a800a000 r--p 00133000 03:00 4222 /lib/libc.so.6
a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6 a800a000-a800b000 rw-p 00135000 03:00 4222 /lib/libc.so.6
@ -388,7 +388,6 @@ is not associated with a file:
[heap] = the heap of the program [heap] = the heap of the program
[stack] = the stack of the main process [stack] = the stack of the main process
[stack:1001] = the stack of the thread with tid 1001
[vdso] = the "virtual dynamic shared object", [vdso] = the "virtual dynamic shared object",
the kernel system call handler the kernel system call handler
@ -396,10 +395,8 @@ is not associated with a file:
The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint The /proc/PID/task/TID/maps is a view of the virtual memory from the viewpoint
of the individual tasks of a process. In this file you will see a mapping marked of the individual tasks of a process. In this file you will see a mapping marked
as [stack] if that task sees it as a stack. This is a key difference from the as [stack] if that task sees it as a stack. Hence, for the example above, the
content of /proc/PID/maps, where you will see all mappings that are being used task-level map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
as stack by all of those tasks. Hence, for the example above, the task-level
map, i.e. /proc/PID/task/TID/maps for thread 1001 will look like this:
08048000-08049000 r-xp 00000000 03:00 8312 /opt/test 08048000-08049000 r-xp 00000000 03:00 8312 /opt/test
08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test 08049000-0804a000 rw-p 00001000 03:00 8312 /opt/test

View file

@ -259,23 +259,29 @@ static int do_maps_open(struct inode *inode, struct file *file,
sizeof(struct proc_maps_private)); sizeof(struct proc_maps_private));
} }
static pid_t pid_of_stack(struct proc_maps_private *priv, /*
struct vm_area_struct *vma, bool is_pid) * Indicate if the VMA is a stack for the given task; for
* /proc/PID/maps that is the stack of the main task.
*/
static int is_stack(struct proc_maps_private *priv,
struct vm_area_struct *vma, int is_pid)
{ {
int stack = 0;
if (is_pid) {
stack = vma->vm_start <= vma->vm_mm->start_stack &&
vma->vm_end >= vma->vm_mm->start_stack;
} else {
struct inode *inode = priv->inode; struct inode *inode = priv->inode;
struct task_struct *task; struct task_struct *task;
pid_t ret = 0;
rcu_read_lock(); rcu_read_lock();
task = pid_task(proc_pid(inode), PIDTYPE_PID); task = pid_task(proc_pid(inode), PIDTYPE_PID);
if (task) {
task = task_of_stack(task, vma, is_pid);
if (task) if (task)
ret = task_pid_nr_ns(task, inode->i_sb->s_fs_info); stack = vma_is_stack_for_task(vma, task);
}
rcu_read_unlock(); rcu_read_unlock();
}
return ret; return stack;
} }
static void static void
@ -335,8 +341,6 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma, int is_pid)
name = arch_vma_name(vma); name = arch_vma_name(vma);
if (!name) { if (!name) {
pid_t tid;
if (!mm) { if (!mm) {
name = "[vdso]"; name = "[vdso]";
goto done; goto done;
@ -348,21 +352,8 @@ show_map_vma(struct seq_file *m, struct vm_area_struct *vma, int is_pid)
goto done; goto done;
} }
tid = pid_of_stack(priv, vma, is_pid); if (is_stack(priv, vma, is_pid))
if (tid != 0) {
/*
* Thread stack in /proc/PID/task/TID/maps or
* the main process stack.
*/
if (!is_pid || (vma->vm_start <= mm->start_stack &&
vma->vm_end >= mm->start_stack)) {
name = "[stack]"; name = "[stack]";
} else {
/* Thread stack in /proc/PID/maps */
seq_pad(m, ' ');
seq_printf(m, "[stack:%d]", tid);
}
}
} }
done: done:
@ -1618,19 +1609,8 @@ static int show_numa_map(struct seq_file *m, void *v, int is_pid)
seq_file_path(m, file, "\n\t= "); seq_file_path(m, file, "\n\t= ");
} else if (vma->vm_start <= mm->brk && vma->vm_end >= mm->start_brk) { } else if (vma->vm_start <= mm->brk && vma->vm_end >= mm->start_brk) {
seq_puts(m, " heap"); seq_puts(m, " heap");
} else { } else if (is_stack(proc_priv, vma, is_pid)) {
pid_t tid = pid_of_stack(proc_priv, vma, is_pid);
if (tid != 0) {
/*
* Thread stack in /proc/PID/task/TID/maps or
* the main process stack.
*/
if (!is_pid || (vma->vm_start <= mm->start_stack &&
vma->vm_end >= mm->start_stack))
seq_puts(m, " stack"); seq_puts(m, " stack");
else
seq_printf(m, " stack:%d", tid);
}
} }
if (is_vm_hugetlb_page(vma)) if (is_vm_hugetlb_page(vma))

View file

@ -123,23 +123,26 @@ unsigned long task_statm(struct mm_struct *mm,
return size; return size;
} }
static pid_t pid_of_stack(struct proc_maps_private *priv, static int is_stack(struct proc_maps_private *priv,
struct vm_area_struct *vma, bool is_pid) struct vm_area_struct *vma, int is_pid)
{ {
struct mm_struct *mm = vma->vm_mm;
int stack = 0;
if (is_pid) {
stack = vma->vm_start <= mm->start_stack &&
vma->vm_end >= mm->start_stack;
} else {
struct inode *inode = priv->inode; struct inode *inode = priv->inode;
struct task_struct *task; struct task_struct *task;
pid_t ret = 0;
rcu_read_lock(); rcu_read_lock();
task = pid_task(proc_pid(inode), PIDTYPE_PID); task = pid_task(proc_pid(inode), PIDTYPE_PID);
if (task) {
task = task_of_stack(task, vma, is_pid);
if (task) if (task)
ret = task_pid_nr_ns(task, inode->i_sb->s_fs_info); stack = vma_is_stack_for_task(vma, task);
}
rcu_read_unlock(); rcu_read_unlock();
}
return ret; return stack;
} }
/* /*
@ -181,21 +184,9 @@ static int nommu_vma_show(struct seq_file *m, struct vm_area_struct *vma,
if (file) { if (file) {
seq_pad(m, ' '); seq_pad(m, ' ');
seq_file_path(m, file, ""); seq_file_path(m, file, "");
} else if (mm) { } else if (mm && is_stack(priv, vma, is_pid)) {
pid_t tid = pid_of_stack(priv, vma, is_pid);
if (tid != 0) {
seq_pad(m, ' '); seq_pad(m, ' ');
/*
* Thread stack in /proc/PID/task/TID/maps or
* the main process stack.
*/
if (!is_pid || (vma->vm_start <= mm->start_stack &&
vma->vm_end >= mm->start_stack))
seq_printf(m, "[stack]"); seq_printf(m, "[stack]");
else
seq_printf(m, "[stack:%d]", tid);
}
} }
seq_putc(m, '\n'); seq_putc(m, '\n');

View file

@ -1341,8 +1341,7 @@ static inline int stack_guard_page_end(struct vm_area_struct *vma,
!vma_growsup(vma->vm_next, addr); !vma_growsup(vma->vm_next, addr);
} }
extern struct task_struct *task_of_stack(struct task_struct *task, int vma_is_stack_for_task(struct vm_area_struct *vma, struct task_struct *t);
struct vm_area_struct *vma, bool in_group);
extern unsigned long move_page_tables(struct vm_area_struct *vma, extern unsigned long move_page_tables(struct vm_area_struct *vma,
unsigned long old_addr, struct vm_area_struct *new_vma, unsigned long old_addr, struct vm_area_struct *new_vma,

View file

@ -230,36 +230,11 @@ void __vma_link_list(struct mm_struct *mm, struct vm_area_struct *vma,
} }
/* Check if the vma is being used as a stack by this task */ /* Check if the vma is being used as a stack by this task */
static int vm_is_stack_for_task(struct task_struct *t, int vma_is_stack_for_task(struct vm_area_struct *vma, struct task_struct *t)
struct vm_area_struct *vma)
{ {
return (vma->vm_start <= KSTK_ESP(t) && vma->vm_end >= KSTK_ESP(t)); return (vma->vm_start <= KSTK_ESP(t) && vma->vm_end >= KSTK_ESP(t));
} }
/*
* Check if the vma is being used as a stack.
* If is_group is non-zero, check in the entire thread group or else
* just check in the current task. Returns the task_struct of the task
* that the vma is stack for. Must be called under rcu_read_lock().
*/
struct task_struct *task_of_stack(struct task_struct *task,
struct vm_area_struct *vma, bool in_group)
{
if (vm_is_stack_for_task(task, vma))
return task;
if (in_group) {
struct task_struct *t;
for_each_thread(task, t) {
if (vm_is_stack_for_task(t, vma))
return t;
}
}
return NULL;
}
#if defined(CONFIG_MMU) && !defined(HAVE_ARCH_PICK_MMAP_LAYOUT) #if defined(CONFIG_MMU) && !defined(HAVE_ARCH_PICK_MMAP_LAYOUT)
void arch_pick_mmap_layout(struct mm_struct *mm) void arch_pick_mmap_layout(struct mm_struct *mm)
{ {