lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20220314114749.28646-1-mark-pk.tsai@mediatek.com>
Date:   Mon, 14 Mar 2022 19:47:49 +0800
From:   Mark-PK Tsai <mark-pk.tsai@...iatek.com>
To:     <rostedt@...dmis.org>
CC:     <linux-arm-kernel@...ts.infradead.org>,
        <linux-kernel@...r.kernel.org>,
        <linux-mediatek@...ts.infradead.org>, <mark-pk.tsai@...iatek.com>,
        <matthias.bgg@...il.com>, <mingo@...hat.com>,
        <yj.chiang@...iatek.com>
Subject: Re: [PATCH] tracing: make tracer_init_tracefs initcall asynchronous

> On Fri, 11 Mar 2022 19:26:56 +0800
> Mark-PK Tsai <mark-pk.tsai@...iatek.com> wrote:
> 
> > tracer_init_tracefs() is slow especially when there are
> > lots of trace events.
> > Create a kthread to do tracer_init_tracefs() asynchronously
> 
> When making comments like this, please provide the benchmarks you used,
> with the numbers before and after.

I've retest it with kernel 5.17-rc7 on my arm64 board, and I found that
the critical path is trace_eval_sync which spend about 430 ms.
It's almost half of the time the do_initcalls spends.
Below is the test result.

			before		after
tracer_init_tracefs	29872 us	66 us
trace_eval_sync		429695 us	459370 us
do_initcalls		797818 us	799890 us

I locally skip trace_eval_sync and got below result.

			before		after		diff
do_initcalls		359252 us	341725 us	-17527 us

So beside this patch, could we add a kernel parameter or a
option to skip it when it doesn't used right after kernel boot?

> 
> > to speed up the initialization of kernel and move the
> > related functions and variables out of init section.
> 
> Thus we sacrifice memory for boot time. I'd like to also see how much
> memory is freed from init before and after this patch.
> 

Below is the INIT_TEXT and INIT_DATA diff:

		before	after	diff
INIT_TEXT	7F290	7EDAC	-0x4e4	bytes
INIT_DATA	116FE8	116FE8	0	bytes

And the init section is 64K aligned on arm64 so that when I test
on my platform, the actual memory freed by initmem_free() have no
diffrence after apply this patch.

#define SEGMENT_ALIGN SZ_64K

> > 
> > Signed-off-by: Mark-PK Tsai <mark-pk.tsai@...iatek.com>
> > ---
> >  fs/tracefs/inode.c          |  8 ++++----
> >  kernel/trace/ftrace.c       | 12 ++++++------
> >  kernel/trace/trace.c        | 21 ++++++++++++++++-----
> >  kernel/trace/trace_events.c |  2 +-
> >  4 files changed, 27 insertions(+), 16 deletions(-)
> > 
> > diff --git a/fs/tracefs/inode.c b/fs/tracefs/inode.c
> > index de7252715b12..9a713d6bcb7e 100644
> > --- a/fs/tracefs/inode.c
> > +++ b/fs/tracefs/inode.c
> > @@ -561,10 +561,10 @@ struct dentry *tracefs_create_dir(const char *name, struct dentry *parent)
> >   *
> >   * Returns the dentry of the instances directory.
> >   */
> > -__init struct dentry *tracefs_create_instance_dir(const char *name,
> > -					  struct dentry *parent,
> > -					  int (*mkdir)(const char *name),
> > -					  int (*rmdir)(const char *name))
> > +struct dentry *tracefs_create_instance_dir(const char *name,
> > +					   struct dentry *parent,
> > +					   int (*mkdir)(const char *name),
> > +					   int (*rmdir)(const char *name))
> >  {
> >  	struct dentry *dentry;
> >  
> > diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
> > index a4b462b6f944..197630cbd5dd 100644
> > --- a/kernel/trace/ftrace.c
> > +++ b/kernel/trace/ftrace.c
> > @@ -940,7 +940,7 @@ static const struct file_operations ftrace_profile_fops = {
> >  };
> >  
> >  /* used to initialize the real stat files */
> > -static struct tracer_stat function_stats __initdata = {
> > +static struct tracer_stat function_stats = {
> >  	.name		= "functions",
> >  	.stat_start	= function_stat_start,
> >  	.stat_next	= function_stat_next,
> > @@ -949,7 +949,7 @@ static struct tracer_stat function_stats __initdata = {
> >  	.stat_show	= function_stat_show
> >  };
> >  
> > -static __init void ftrace_profile_tracefs(struct dentry *d_tracer)
> > +static void ftrace_profile_tracefs(struct dentry *d_tracer)
> >  {
> >  	struct ftrace_profile_stat *stat;
> >  	struct dentry *entry;
> > @@ -991,7 +991,7 @@ static __init void ftrace_profile_tracefs(struct dentry *d_tracer)
> >  }
> >  
> >  #else /* CONFIG_FUNCTION_PROFILER */
> > -static __init void ftrace_profile_tracefs(struct dentry *d_tracer)
> > +static void ftrace_profile_tracefs(struct dentry *d_tracer)
> >  {
> >  }
> >  #endif /* CONFIG_FUNCTION_PROFILER */
> > @@ -6359,7 +6359,7 @@ void ftrace_destroy_filter_files(struct ftrace_ops *ops)
> >  	mutex_unlock(&ftrace_lock);
> >  }
> >  
> > -static __init int ftrace_init_dyn_tracefs(struct dentry *d_tracer)
> > +static int ftrace_init_dyn_tracefs(struct dentry *d_tracer)
> >  {
> >  
> >  	trace_create_file("available_filter_functions", TRACE_MODE_READ,
> > @@ -7754,8 +7754,8 @@ void ftrace_init_tracefs(struct trace_array *tr, struct dentry *d_tracer)
> >  			  d_tracer, tr, &ftrace_no_pid_fops);
> >  }
> >  
> > -void __init ftrace_init_tracefs_toplevel(struct trace_array *tr,
> > -					 struct dentry *d_tracer)
> > +void ftrace_init_tracefs_toplevel(struct trace_array *tr,
> > +				  struct dentry *d_tracer)
> >  {
> >  	/* Only the top level directory has the dyn_tracefs and profile */
> >  	WARN_ON(!(tr->flags & TRACE_ARRAY_FL_GLOBAL));
> > diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> > index eb44418574f9..f55da82060e2 100644
> > --- a/kernel/trace/trace.c
> > +++ b/kernel/trace/trace.c
> > @@ -9562,10 +9562,10 @@ int tracing_init_dentry(void)
> >  extern struct trace_eval_map *__start_ftrace_eval_maps[];
> >  extern struct trace_eval_map *__stop_ftrace_eval_maps[];
> >  
> > -static struct workqueue_struct *eval_map_wq __initdata;
> > -static struct work_struct eval_map_work __initdata;
> > +static struct workqueue_struct *eval_map_wq;
> > +static struct work_struct eval_map_work;
> >  
> > -static void __init eval_map_work_func(struct work_struct *work)
> > +static void eval_map_work_func(struct work_struct *work)
> >  {
> >  	int len;
> >  
> > @@ -9573,7 +9573,7 @@ static void __init eval_map_work_func(struct work_struct *work)
> >  	trace_insert_eval_map(NULL, __start_ftrace_eval_maps, len);
> >  }
> >  
> > -static int __init trace_eval_init(void)
> > +static int trace_eval_init(void)
> >  {
> >  	INIT_WORK(&eval_map_work, eval_map_work_func);
> >  
> > @@ -9671,7 +9671,7 @@ static struct notifier_block trace_module_nb = {
> >  };
> >  #endif /* CONFIG_MODULES */
> >  
> > -static __init int tracer_init_tracefs(void)
> > +static int tracefs_init(void *data)
> >  {
> >  	int ret;
> >  
> > @@ -9721,6 +9721,17 @@ static __init int tracer_init_tracefs(void)
> >  	return 0;
> >  }
> >  
> > +static __init int tracer_init_tracefs(void)
> > +{
> > +	struct task_struct *thread;
> > +
> > +	thread = kthread_run(tracefs_init, NULL, "tracefs_init");
> > +	if (IS_ERR(thread))
> > +		return PTR_ERR(thread);
> > +
> > +	return 0;
> > +}
> > +
> >  fs_initcall(tracer_init_tracefs);
> >  
> >  static int trace_panic_handler(struct notifier_block *this,
> > diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
> > index 3147614c1812..fe055bef1e8f 100644
> > --- a/kernel/trace/trace_events.c
> > +++ b/kernel/trace/trace_events.c
> > @@ -3687,7 +3687,7 @@ static __init int event_trace_init_fields(void)
> >  	return 0;
> >  }
> >  
> > -__init int event_trace_init(void)
> > +int event_trace_init(void)
> >  {
> >  	struct trace_array *tr;
> >  	struct dentry *entry;
> 
> Hmm, this calls early_event_tracer() which is also in __init. Looks like
> there's going to be a ripple effect due to this change.
> 
> If we want to go this route, then first a change must be made to remove the
> needed functions from init, and then see if we can consolidate it. As there
> are some init functions that are duplicated for init purposes.

Got it!

> 
> -- Steve

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ