All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jason Baron <jbaron@redhat.com>
To: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Cc: Hendrik Brueckner <brueckner@linux.vnet.ibm.com>,
	Frederic Weisbecker <fweisbec@gmail.com>,
	linux-kernel@vger.kernel.org, mingo@elte.hu,
	laijs@cn.fujitsu.com, rostedt@goodmis.org, peterz@infradead.org,
	jiayingz@google.com, mbligh@google.com, lizf@cn.fujitsu.com,
	Heiko Carstens <heiko.carstens@de.ibm.com>,
	Martin Schwidefsky <schwidefsky@de.ibm.com>
Subject: Re: [PATCH 08/12] add trace events for each syscall entry/exit
Date: Tue, 25 Aug 2009 13:04:41 -0400	[thread overview]
Message-ID: <20090825170441.GD2656@redhat.com> (raw)
In-Reply-To: <20090825162004.GA25058@Krystal>

On Tue, Aug 25, 2009 at 12:20:04PM -0400, Mathieu Desnoyers wrote:
> 
> Uh ? kernel threads can invoke a system call. There are rare places
> where kernel code actually invoke system calls. I don't see why we
> should not deal with them.
> 
> Moreover, the problem you face is more general: if we set the
> TIF_SYSCALL_FTRACE flag of a standard thread right in the middle of its
> system call, x86_64 will cause the syscall exit to execute by re-reading
> the thread flags and run a syscall trace exit.
> 
> We could simply initialize the "saved system calls id" number to
> something like -1, so that if we happen to return from a syscall that
> did not get its id recorded at syscall entry, we know it because it's
> not initialized.
> 
> We would need to carefully put back the -1 value after clearing the
> thread flag when we stop tracing too (while still holding a mutex).
> 
> Mathieu
> 

why can't we have a syscall exit that is unmatched? we calculate
the exit syscall number for the the pt_regs structure at exit, so we
don't need to match it up with an entry to know which syscall it is.

thanks,

-Jason

  parent reply	other threads:[~2009-08-25 17:05 UTC|newest]

Thread overview: 88+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-10 20:52 [PATCH 00/12] add syscall tracepoints V3 Jason Baron
2009-08-10 20:52 ` [PATCH 01/12] map syscall name to number Jason Baron
2009-08-10 20:52 ` [PATCH 02/12] call arch_init_ftrace_syscalls at boot Jason Baron
2009-08-10 20:52 ` [PATCH 03/12] add DECLARE_TRACE_WITH_CALLBACK() macro Jason Baron
2009-08-10 20:52 ` [PATCH 04/12] add syscall tracepoints Jason Baron
2009-08-10 20:52 ` [PATCH 05/12] update FTRACE_SYSCALL_MAX Jason Baron
2009-08-11 11:00   ` Frederic Weisbecker
2009-08-11 19:39     ` Matt Fleming
2009-08-24 13:41     ` Paul Mundt
2009-08-24 14:06       ` Jason Baron
2009-08-24 14:15         ` Paul Mundt
2009-08-24 14:34           ` Frederic Weisbecker
2009-08-24 14:37             ` Paul Mundt
2009-08-24 14:42           ` Jason Baron
2009-08-24 14:50             ` Paul Mundt
2009-08-24 18:34               ` Ingo Molnar
2009-08-10 20:52 ` [PATCH 06/12] trace_event - raw_init bailout Jason Baron
2009-08-10 20:52 ` [PATCH 07/12] add ftrace_event_call void * 'data' field Jason Baron
2009-08-11 10:09   ` Frederic Weisbecker
2009-08-17 22:19     ` Steven Rostedt
2009-08-17 23:09       ` Frederic Weisbecker
2009-08-18  0:06         ` Steven Rostedt
2009-08-10 20:52 ` [PATCH 08/12] add trace events for each syscall entry/exit Jason Baron
2009-08-11 10:50   ` Frederic Weisbecker
2009-08-11 11:45     ` Ingo Molnar
2009-08-11 12:01       ` Frederic Weisbecker
2009-08-25 12:50   ` Hendrik Brueckner
2009-08-25 14:15     ` Frederic Weisbecker
2009-08-25 16:02       ` Hendrik Brueckner
2009-08-25 16:20         ` Mathieu Desnoyers
2009-08-25 16:59           ` Frederic Weisbecker
2009-08-25 17:31             ` Frederic Weisbecker
2009-08-25 18:31               ` Mathieu Desnoyers
2009-08-25 19:42                 ` Frederic Weisbecker
2009-08-25 19:51                   ` Mathieu Desnoyers
2009-08-26  0:19                     ` Frederic Weisbecker
2009-08-26  0:42                       ` Mathieu Desnoyers
2009-08-26  7:28                         ` Ingo Molnar
2009-08-26 17:11                           ` Mathieu Desnoyers
2009-08-26  6:48                   ` Peter Zijlstra
2009-08-25 22:04                 ` Martin Schwidefsky
2009-08-26  7:38                   ` Heiko Carstens
2009-08-26 12:32                     ` Frederic Weisbecker
2009-08-26  6:21                 ` Peter Zijlstra
2009-08-26 17:08                   ` Mathieu Desnoyers
2009-08-26 18:41                     ` Christoph Hellwig
2009-08-26 18:42                       ` Christoph Hellwig
2009-08-26 19:01                         ` Mathieu Desnoyers
2009-08-26  7:10                 ` Peter Zijlstra
2009-08-26 17:10                   ` Mathieu Desnoyers
2009-08-26 17:24                   ` H. Peter Anvin
2009-08-25 17:04           ` Jason Baron [this message]
2009-08-25 18:15             ` Mathieu Desnoyers
2009-08-26 12:35         ` Frederic Weisbecker
2009-08-26 12:59           ` Heiko Carstens
2009-08-26 13:30             ` Frederic Weisbecker
2009-08-26 13:48               ` Steven Rostedt
2009-08-26 13:53                 ` Frederic Weisbecker
2009-08-26 14:44                   ` Steven Rostedt
2009-08-26 13:56                 ` Peter Zijlstra
2009-08-26 14:41                   ` Steven Rostedt
2009-08-26 14:10               ` Heiko Carstens
2009-08-26 14:27                 ` Frederic Weisbecker
2009-08-26 14:43                   ` Steven Rostedt
2009-08-26 16:14                     ` Frederic Weisbecker
2009-08-26 14:43                 ` Steven Rostedt
2009-08-26 14:41           ` Hendrik Brueckner
2009-08-28 12:28         ` [tip:tracing/core] tracing: Don't trace kernel thread syscalls tip-bot for Hendrik Brueckner
2009-08-25 21:40     ` [PATCH 08/12] add trace events for each syscall entry/exit Frederic Weisbecker
2009-08-25 22:09       ` Frederic Weisbecker
2009-08-26  7:47         ` Heiko Carstens
2009-08-28 12:27     ` [tip:tracing/core] tracing: Check invalid syscall nr while tracing syscalls tip-bot for Hendrik Brueckner
2009-08-10 20:52 ` [PATCH 09/12] add support traceopint ids Jason Baron
2009-08-11 11:28   ` Frederic Weisbecker
2009-08-10 20:53 ` [PATCH 10/12] add perf counter support Jason Baron
2009-08-11 12:12   ` Frederic Weisbecker
2009-08-11 12:17     ` Ingo Molnar
2009-08-11 12:25       ` Frederic Weisbecker
2009-08-10 20:53 ` [PATCH 11/12] add more namespace area to 'perf list' output Jason Baron
2009-08-10 20:53 ` [PATCH 12/12] convert x86_64 mmap and uname to use DEFINE_SYSCALL Jason Baron
2009-08-25 12:31 ` [PATCH 00/12] add syscall tracepoints V3 - s390 arch update Hendrik Brueckner
2009-08-25 13:52   ` Frederic Weisbecker
2009-08-25 14:39     ` Heiko Carstens
2009-08-25 19:52       ` Frederic Weisbecker
2009-08-25 15:38     ` Hendrik Brueckner
2009-08-26 16:53   ` Frederic Weisbecker
2009-08-27  7:27     ` [PATCH]: tracing: s390 arch updates for tracing syscalls Hendrik Brueckner
2009-08-28 12:27   ` [tip:tracing/core] tracing: Add syscall tracepoints - s390 arch update tip-bot for Hendrik Brueckner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20090825170441.GD2656@redhat.com \
    --to=jbaron@redhat.com \
    --cc=brueckner@linux.vnet.ibm.com \
    --cc=fweisbec@gmail.com \
    --cc=heiko.carstens@de.ibm.com \
    --cc=jiayingz@google.com \
    --cc=laijs@cn.fujitsu.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=lizf@cn.fujitsu.com \
    --cc=mathieu.desnoyers@polymtl.ca \
    --cc=mbligh@google.com \
    --cc=mingo@elte.hu \
    --cc=peterz@infradead.org \
    --cc=rostedt@goodmis.org \
    --cc=schwidefsky@de.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.