From: Steven Rostedt <rostedt@goodmis.org>
To: "Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com>
Cc: Michael Ellerman <mpe@ellerman.id.au>,
Michal Suchanek <msuchanek@suse.de>, Torsten Duwe <duwe@suse.de>,
<linuxppc-dev@lists.ozlabs.org>, <linux-kernel@vger.kernel.org>
Subject: Re: [RFC PATCH 1/6] trace/stack: Move code to save the stack trace into a separate function
Date: Tue, 1 Jun 2021 11:28:24 -0400 [thread overview]
Message-ID: <20210601112824.29c5f168@oasis.local.home> (raw)
In-Reply-To: <6a8b68f8bd64f8c16d97ef943534c639781e7f77.1621577151.git.naveen.n.rao@linux.vnet.ibm.com>
On Fri, 21 May 2021 12:18:36 +0530
"Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com> wrote:
> In preparation to add support for stack tracer to powerpc, move code to
> save stack trace and to calculate the frame sizes into a separate weak
> function. Also provide access to some of the data structures used by the
> stack trace code so that architectures can update those.
>
> Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com>
> ---
> include/linux/ftrace.h | 8 ++++
> kernel/trace/trace_stack.c | 98 ++++++++++++++++++++------------------
> 2 files changed, 60 insertions(+), 46 deletions(-)
>
> diff --git a/include/linux/ftrace.h b/include/linux/ftrace.h
> index a69f363b61bf73..8263427379f05c 100644
> --- a/include/linux/ftrace.h
> +++ b/include/linux/ftrace.h
> @@ -368,10 +368,18 @@ static inline void arch_ftrace_set_direct_caller(struct pt_regs *regs,
>
> #ifdef CONFIG_STACK_TRACER
>
> +#define STACK_TRACE_ENTRIES 500
> +
> +extern unsigned long stack_dump_trace[STACK_TRACE_ENTRIES];
> +extern unsigned stack_trace_index[STACK_TRACE_ENTRIES];
> +extern unsigned int stack_trace_nr_entries;
> +extern unsigned long stack_trace_max_size;
> extern int stack_tracer_enabled;
>
> int stack_trace_sysctl(struct ctl_table *table, int write, void *buffer,
> size_t *lenp, loff_t *ppos);
> +void stack_get_trace(unsigned long traced_ip, unsigned long *stack_ref,
> + unsigned long stack_size, int *tracer_frame);
>
> /* DO NOT MODIFY THIS VARIABLE DIRECTLY! */
> DECLARE_PER_CPU(int, disable_stack_tracer);
> diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c
> index 63c28504205162..5b63dbd37c8c25 100644
> --- a/kernel/trace/trace_stack.c
> +++ b/kernel/trace/trace_stack.c
> @@ -19,13 +19,11 @@
>
> #include "trace.h"
>
> -#define STACK_TRACE_ENTRIES 500
> +unsigned long stack_dump_trace[STACK_TRACE_ENTRIES];
> +unsigned stack_trace_index[STACK_TRACE_ENTRIES];
>
> -static unsigned long stack_dump_trace[STACK_TRACE_ENTRIES];
> -static unsigned stack_trace_index[STACK_TRACE_ENTRIES];
> -
> -static unsigned int stack_trace_nr_entries;
> -static unsigned long stack_trace_max_size;
> +unsigned int stack_trace_nr_entries;
> +unsigned long stack_trace_max_size;
> static arch_spinlock_t stack_trace_max_lock =
> (arch_spinlock_t)__ARCH_SPIN_LOCK_UNLOCKED;
>
> @@ -152,49 +150,19 @@ static void print_max_stack(void)
> * Although the entry function is not displayed, the first function (sys_foo)
> * will still include the stack size of it.
> */
> -static void check_stack(unsigned long ip, unsigned long *stack)
I just got back from PTO and have a ton of other obligations to attend
to before I can dig deeper into this. I'm not opposed to this change,
but the stack_tracer has not been getting the love that it deserves and
I think you hit one of the issues that needs to be addressed. I'm not
sure this is a PPC only issue, and would like to see if I can get more
time (or someone else can) to reevaluate the way stack tracer works,
and see if it can be made a bit more robust.
-- Steve
next prev parent reply other threads:[~2021-06-01 15:28 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-05-21 6:48 [RFC PATCH 0/6] powerpc: Stack tracer fixes Naveen N. Rao
2021-05-21 6:48 ` [RFC PATCH 1/6] trace/stack: Move code to save the stack trace into a separate function Naveen N. Rao
2021-06-01 15:28 ` Steven Rostedt [this message]
2021-06-02 10:35 ` Naveen N. Rao
2021-06-02 14:09 ` Steven Rostedt
2021-05-21 6:48 ` [RFC PATCH 2/6] powerpc/trace: Add support for stack tracer Naveen N. Rao
2021-06-01 13:51 ` Naveen N. Rao
2021-05-21 6:48 ` [RFC PATCH 3/6] powerpc: Indicate traced function name in show_stack() Naveen N. Rao
2021-05-21 6:48 ` [RFC PATCH 4/6] powerpc/perf: Include traced function in the callchain Naveen N. Rao
2021-05-21 6:48 ` [RFC PATCH 5/6] powerpc/stacktrace: Include ftraced function in arch_stack_walk_reliable() Naveen N. Rao
2021-05-21 6:48 ` [RFC PATCH 6/6] powerpc/stacktrace: Include ftraced function in arch_stack_walk() Naveen N. Rao
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=20210601112824.29c5f168@oasis.local.home \
--to=rostedt@goodmis.org \
--cc=duwe@suse.de \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=mpe@ellerman.id.au \
--cc=msuchanek@suse.de \
--cc=naveen.n.rao@linux.vnet.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).