Re: Detailed Stack Information Patch Next Generation

[Andrew Morton <akpm@linux-foundation.org>]

> Subject: Detailed Stack Information Patch Next Generation

That's not a very useful patch title.  Something like

	[patch] procfs: provide stack information for threads

would suit.

On Wed, 03 Jun 2009 22:34:09 +0200
Stefani Seibold <stefani@seibold.net> wrote:

> Hi everybody,
> 
> kernel 2.6.30 is near so i think it is time for the next try.
> 
> This is the new version of the "detailed stack info" patch which give
> you a better overview of the userland application stack usage,
> especially for embedded linux.
> 
> Currently you are only able to dump the main process/thread stack usage
> which is showed in /proc/pid/status by the "VmStk" Value. But you get no
> information about the consumed stack memory of the the threads.
> 
> There is an enhancement in the /proc/<pid>/{task/*,}/*maps and which
> marks the vm mapping where the thread stack pointer reside with "[thread
> stack xxxxxxxx]". xxxxxxxx is the start address of the stack.
> 
> Also there is a new entry "stack usage" in /proc/<pid>/{task/*,}/status
> which will you give the current stack usage in kb.
> 
> I also fixed stack base address in /proc/<pid>/task/*/stat to the base
> address of the associated thread stack and not the one of the main
> process. This makes more sense.
> 
> Changes since last posting:
> 
>  - Redesigned everything what was suggested by Ingo
>  - the thread watch monitor is gone
>  - the /proc/stackmon entry is also gone
>  - slime done
> 
> The patch is against 2.6.30-rc7 and tested with on intel and ppc
> architectures.
> 
> I think it is now ready for inclusion in the main kernel. Ingo what do
> you think?
> 

Missing Signed-off-by:

> 
>  fs/exec.c             |    2 +
>  fs/proc/array.c       |   69 +++++++++++++++++++++++++++++++++++++++++++++++++-
>  fs/proc/task_mmu.c    |   14 ++++++++++
>  include/linux/sched.h |    1 
>  kernel/fork.c         |    3 ++
>  5 files changed, 88 insertions(+), 1 deletion(-)
> 
> -patch begins here--------------------------------------------------------------
> 
> diff -u -N -r linux-2.6.30.orig/fs/exec.c linux-2.6.30/fs/exec.c
> --- linux-2.6.30.orig/fs/exec.c	2009-06-03 17:36:41.000000000 +0200
> +++ linux-2.6.30/fs/exec.c	2009-06-03 17:30:47.000000000 +0200
> @@ -1328,6 +1328,8 @@
>  	if (retval < 0)
>  		goto out;
>  
> +	current->stack_start = current->mm->start_stack;
> +
>  	/* execve succeeded */
>  	current->fs->in_exec = 0;
>  	current->in_execve = 0;
> diff -u -N -r linux-2.6.30.orig/fs/proc/array.c linux-2.6.30/fs/proc/array.c
> --- linux-2.6.30.orig/fs/proc/array.c	2009-06-03 17:36:41.000000000 +0200
> +++ linux-2.6.30/fs/proc/array.c	2009-06-03 17:32:53.000000000 +0200
> @@ -82,6 +82,7 @@
>  #include <linux/pid_namespace.h>
>  #include <linux/ptrace.h>
>  #include <linux/tracehook.h>
> +#include <linux/pfn.h>
>  
>  #include <asm/pgtable.h>
>  #include <asm/processor.h>
> @@ -321,6 +322,71 @@
>  			p->nivcsw);
>  }
>  
> +#ifdef CONFIG_STACK_GROWSUP
> +static inline unsigned long get_stack_pages(struct vm_area_struct *vma,
> +					struct task_struct *p)
> +{
> +	unsigned long	i;
> +	struct page	*page;
> +	unsigned long	stack_usage;
> +	unsigned long	esp;

It's unclear what the units are here.  Are they in bytes?  Pages?  Code
comments and well-chosen identifiers are the way to fix this.

> +	esp = KSTK_ESP(p);

So `esp' is in bytes.

> +	stack_usage = (PFN_ALIGN(esp)-PFN_ALIGN(p->stack_start));

And `stack_usage' measures bytes

> +	for (i = vma->vm_end; i-PAGE_SIZE > esp; i -= PAGE_SIZE) {

And `i' measures bytes

> +		page = follow_page(vma, i-PAGE_SIZE, 0);
> +
> +		if ((!IS_ERR(page) == 0) || (page))

Why not simply

		if (IS_ERR(page) || page)

?

And shouldn't it be !page?

> +			break;
> +	}
> +	return ((i - esp + stack_usage) >> (PAGE_SHIFT)) + 1;

And we return a number-of-pages.  Fair enough.

> +}
> +#else
> +static inline unsigned long get_stack_pages(struct vm_area_struct *vma,
> +					struct task_struct *p)
> +{
> +	unsigned long	i;
> +	struct page	*page;
> +	unsigned long	stack_usage;
> +	unsigned long	esp;
> +
> +	esp = KSTK_ESP(p);
> +
> +	stack_usage = (PFN_ALIGN(p->stack_start) - PFN_ALIGN(esp));
> +
> +	for (i = vma->vm_start; i+PAGE_SIZE <= esp; i += PAGE_SIZE) {
> +
> +		page = follow_page(vma, i, 0);
> +
> +		if ((!IS_ERR(page) == 0) || (page))
> +			break;
> +	}
> +	return ((esp - i + stack_usage) >> (PAGE_SHIFT)) + 1;
> +}

Dittoes.

> +#endif
> +
> +static inline void task_show_stack_usage(struct seq_file *m,
> +						struct task_struct *p)
> +{
> +	struct vm_area_struct	*vma;
> +	struct mm_struct	*mm;
> +
> +	mm = get_task_mm(p);
> +
> +	if (mm) {
> +		vma = find_vma(mm, p->stack_start);
> +
> +		if (vma)
> +			seq_printf(m, "Stack usage:\t%lu kB\n",
> +				get_stack_pages(vma, p) << (PAGE_SHIFT - 10));

So get_stack_pages() did a bytes-to-pages conversion then its sole
caller does a pages-to-bytes conversion.  Can this be simplified?

> +
> +		mmput(mm);
> +	}
> +}
>
>  int proc_pid_status(struct seq_file *m, struct pid_namespace *ns,
>  			struct pid *pid, struct task_struct *task)
>  {
> @@ -340,6 +406,7 @@
>  	task_show_regs(m, task);
>  #endif
>  	task_context_switch_counts(m, task);
> +	task_show_stack_usage(m, task);
>  	return 0;
>  }
>  
> @@ -481,7 +548,7 @@
>  		rsslim,
>  		mm ? mm->start_code : 0,
>  		mm ? mm->end_code : 0,
> -		(permitted && mm) ? mm->start_stack : 0,
> +		(permitted) ? task->stack_start : 0,
>  		esp,
>  		eip,
>  		/* The signal information here is obsolete.
> diff -u -N -r linux-2.6.30.orig/fs/proc/task_mmu.c linux-2.6.30/fs/proc/task_mmu.c
> --- linux-2.6.30.orig/fs/proc/task_mmu.c	2009-06-03 17:36:41.000000000 +0200
> +++ linux-2.6.30/fs/proc/task_mmu.c	2009-06-03 17:30:47.000000000 +0200
> @@ -242,6 +242,20 @@
>  				} else if (vma->vm_start <= mm->start_stack &&
>  					   vma->vm_end >= mm->start_stack) {
>  					name = "[stack]";
> +				} else {
> +					unsigned long stack_start;
> +
> +					stack_start =
> +						((struct proc_maps_private *)
> +						m->private)->task->stack_start;

Like this:

					unsiged long stack_start;
					struct proc_maps_private *pmp;

					pmp = m->private;
					stack_start = pmp->task->stack_start;

> +					if (vma->vm_start <= stack_start &&
> +					    vma->vm_end >= stack_start) {
> +						pad_len_spaces(m, len);
> +						seq_printf(m,
> +						 "[thread stack: %08lx]",
> +						 stack_start);
> +					}
>  				}
>  			} else {
>  				name = "[vdso]";
> diff -u -N -r linux-2.6.30.orig/include/linux/sched.h linux-2.6.30/include/linux/sched.h
> --- linux-2.6.30.orig/include/linux/sched.h	2009-06-03 17:36:41.000000000 +0200
> +++ linux-2.6.30/include/linux/sched.h	2009-06-03 17:30:47.000000000 +0200
> @@ -1429,6 +1429,7 @@
>  	/* state flags for use by tracers */
>  	unsigned long trace;
>  #endif
> +	unsigned long stack_start;
>  };
>  
>  /* Future-safe accessor for struct task_struct's cpus_allowed. */
> diff -u -N -r linux-2.6.30.orig/kernel/fork.c linux-2.6.30/kernel/fork.c
> --- linux-2.6.30.orig/kernel/fork.c	2009-06-03 17:36:42.000000000 +0200
> +++ linux-2.6.30/kernel/fork.c	2009-06-03 17:30:47.000000000 +0200
> @@ -1092,6 +1092,9 @@
>  	if (unlikely(current->ptrace))
>  		ptrace_fork(p, clone_flags);
>  
> +	p->stack_start = (stack_start == KSTK_ESP(current)) ?
> +		current->stack_start : stack_start;
> +

hm.  What's this doing?