linux-kernel - Re: [PATCH v4 09/10] user_events: Optimize writing events by only copying data once

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: <20211108230034.GB1452@kbox>
Date:   Mon, 8 Nov 2021 15:00:34 -0800
From:   Beau Belgrave <beaub@...ux.microsoft.com>
To:     Steven Rostedt <rostedt@...dmis.org>
Cc:     mhiramat@...nel.org, linux-trace-devel@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v4 09/10] user_events: Optimize writing events by only
 copying data once

On Mon, Nov 08, 2021 at 05:45:42PM -0500, Steven Rostedt wrote:
> On Thu,  4 Nov 2021 10:04:32 -0700
> Beau Belgrave <beaub@...ux.microsoft.com> wrote:
> 
> > Pass iterator through to probes to allow copying data directly to the
> > probe buffers instead of taking multiple copies. Enables eBPF user and
> > raw iterator types out to programs for no-copy scenarios.
> > 
> > Signed-off-by: Beau Belgrave <beaub@...ux.microsoft.com>
> > ---
> >  kernel/trace/trace_events_user.c | 97 +++++++++++++++++++++++---------
> >  1 file changed, 69 insertions(+), 28 deletions(-)
> > 
> > diff --git a/kernel/trace/trace_events_user.c b/kernel/trace/trace_events_user.c
> > index b5fe0550b489..d50118b9630a 100644
> > --- a/kernel/trace/trace_events_user.c
> > +++ b/kernel/trace/trace_events_user.c
> > @@ -39,6 +39,10 @@
> >  #define MAX_EVENT_DESC 512
> >  #define EVENT_NAME(user_event) ((user_event)->tracepoint.name)
> >  
> > +#define MAX_BPF_COPY_SIZE PAGE_SIZE
> > +#define MAX_STACK_BPF_DATA 512
> > +#define copy_nofault copy_from_iter_nocache
> > +
> >  static char *register_page_data;
> >  
> >  static DEFINE_MUTEX(reg_mutex);
> > @@ -63,8 +67,7 @@ struct user_event_refs {
> >  	struct user_event *events[];
> >  };
> >  
> > -typedef void (*user_event_func_t) (struct user_event *user,
> > -				   void *data, u32 datalen,
> > +typedef void (*user_event_func_t) (struct user_event *user, struct iov_iter *i,
> >  				   void *tpdata);
> >  
> >  static int user_event_parse(char *name, char *args, char *flags,
> > @@ -491,7 +494,7 @@ static struct user_event *find_user_event(char *name, u32 *outkey)
> >  /*
> >   * Writes the user supplied payload out to a trace file.
> >   */
> > -static void user_event_ftrace(struct user_event *user, void *data, u32 datalen,
> > +static void user_event_ftrace(struct user_event *user, struct iov_iter *i,
> >  			      void *tpdata)
> >  {
> >  	struct trace_event_file *file;
> > @@ -506,41 +509,82 @@ static void user_event_ftrace(struct user_event *user, void *data, u32 datalen,
> >  		return;
> >  
> >  	entry = trace_event_buffer_reserve(&event_buffer, file,
> > -					   sizeof(*entry) + datalen);
> > +					   sizeof(*entry) + i->count);
> >  
> >  	if (unlikely(!entry))
> >  		return;
> >  
> > -	memcpy(entry + 1, data, datalen);
> > +	if (unlikely(!copy_nofault(entry + 1, i->count, i)))
> 
> Need:
> 		__trace_event_discard_commit(event_buffer.buffer, event_buffer.event);
> 
> Because the trace_event_buffer_reserve() will not only allocate space on
> the ring buffer, but may also disable preemption.
> 
> -- Steve
> 

Ah, thank you!

-Beau

> 
> > +		return;
> >  
> >  	trace_event_buffer_commit(&event_buffer);
> >  }
> >  
> >  #ifdef CONFIG_PERF_EVENTS