public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Christoph Hellwig <hch@infradead.org>
To: Takao Indoh <indou.takao@soft.fujitsu.com>
Cc: linux-kernel@vger.kernel.org, Ingo Molnar <mingo@elte.hu>,
	Christoph Hellwig <hch@infradead.org>, Andi Kleen <ak@muc.de>
Subject: Re: [3/4] [PATCH]Diskdump - yet another crash dump function
Date: Thu, 17 Jun 2004 13:00:20 +0100	[thread overview]
Message-ID: <20040617120020.GA30691@infradead.org> (raw)
In-Reply-To: <C7C4545F11DFBEindou.takao@soft.fujitsu.com>

On Thu, Jun 17, 2004 at 08:34:37PM +0900, Takao Indoh wrote:
> Instead of redefinition, I add hooks into the timer/tasklet routines.
> 
> ex.
> 
>  int mod_timer(struct timer_list *timer, unsigned long expires)
>  {
>  	BUG_ON(!timer->function);
> 
> +	if (crashdump_mode()) {
> +		return diskdump_mod_timer(timer, expires);
> +	}
>  
>  	 check_timer(timer);
> 
> Please see the following patches for details.
> How do you think? 

Looks good in principle, some more comments on the actual patch below:

> diff -Nur linux-2.6.6.org/include/linux/diskdumplib.h linux-2.6.6/include/linux/diskdumplib.h
> --- linux-2.6.6.org/include/linux/diskdumplib.h	1970-01-01 09:00:00.000000000 +0900
> +++ linux-2.6.6/include/linux/diskdumplib.h	2004-06-17 18:28:01.000000000 +0900

Should probably just go to linux/dump.h or dump_priv.h

> diff -Nur linux-2.6.6.org/include/linux/interrupt.h linux-2.6.6/include/linux/interrupt.h
> --- linux-2.6.6.org/include/linux/interrupt.h	2004-06-04 21:22:09.000000000 +0900
> +++ linux-2.6.6/include/linux/interrupt.h	2004-06-17 18:28:01.000000000 +0900
> @@ -7,6 +7,7 @@
>  #include <linux/linkage.h>
>  #include <linux/bitops.h>
>  #include <linux/preempt.h>
> +#include <linux/diskdumplib.h>
>  #include <asm/atomic.h>
>  #include <asm/hardirq.h>
>  #include <asm/ptrace.h>
> @@ -172,6 +173,11 @@
>  
>  static inline void tasklet_schedule(struct tasklet_struct *t)
>  {
> +	if (crashdump_mode()) {
> +		diskdump_tasklet_schedule(t);
> +		return;
> +	}
> +
>  	if (!test_and_set_bit(TASKLET_STATE_SCHED, &t->state))
>  		__tasklet_schedule(t);
>  }

Please sprinclde unlikely's all over here.  Also the above could be
shortened to

static inline void tasklet_schedule(struct tasklet_struct *t)
{
	if (unlikely(crashdump_mode()))
		diskdump_tasklet_schedule(t);
	else if (!test_and_set_bit(TASKLET_STATE_SCHED, &t->state))
		__tasklet_schedule(t);
}

> +int diskdump_schedule_work(struct work_struct *work)
> +{
> +	list_add_tail(&work->entry, &diskdump_workq);
> +	return 1;
> +}

Should probably just inlined, I guess the function call is bigger
than all of list_add_tail.

> +void diskdump_add_timer(struct timer_list *timer)
> +{
> +	timer->base = (void *)1;
> +	list_add(&timer->entry, &diskdump_timers);
> +}

dito.

> +	/* run timers */
> +	list_for_each_safe(t, n, &diskdump_timers) {
> +		timer = list_entry(t, struct timer_list, entry);

list_for_each_entry_safe here please.

> @@ -255,6 +255,10 @@
>  {
>  	BUG_ON(!timer->function);
>  
> +	if (crashdump_mode()) {
> +		return diskdump_mod_timer(timer, expires);
> +	}

please remove the superflous braces here.


  reply	other threads:[~2004-06-17 12:00 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-05-27  9:33 [PATCH]Diskdump - yet another crash dump function Takao Indoh
2004-05-27 12:36 ` [1/4] " Takao Indoh
2004-05-27 12:37 ` [2/4] " Takao Indoh
2004-05-27 13:48   ` Christoph Hellwig
2004-05-28  2:13     ` Takao Indoh
2004-05-27 12:39 ` [3/4] " Takao Indoh
2004-05-27 13:51   ` Christoph Hellwig
2004-05-27 21:04     ` Ingo Molnar
2004-06-17 11:34       ` Takao Indoh
2004-06-17 12:00         ` Christoph Hellwig [this message]
2004-06-17 12:45           ` Takao Indoh
2004-06-17 12:13         ` Ingo Molnar
2004-06-17 12:18           ` Christoph Hellwig
2004-06-17 12:32             ` Ingo Molnar
2004-06-17 14:56               ` Jeff Moyer
2004-06-17 15:45                 ` Nobuhiro Tachino
2004-06-17 13:04           ` Takao Indoh
2004-06-17 13:10             ` Ingo Molnar
2004-06-17 13:11               ` Ingo Molnar
2004-06-17 13:15                 ` Ingo Molnar
2004-06-17 14:00                 ` Takao Indoh
2004-06-17 14:45                   ` Nobuhiro Tachino
2004-06-17 14:53                     ` Takao Indoh
2004-06-18 12:02           ` Takao Indoh
2004-06-21 20:40             ` Nobuhiro Tachino
2004-06-22 10:19               ` Ingo Molnar
2004-06-23 12:11                 ` Takao Indoh
2004-06-23 13:00                   ` Takao Indoh
2004-06-21  5:46           ` Keith Owens
2004-06-21  6:25             ` Takao Indoh
2004-06-22  4:21             ` Rob Landley
2004-06-22  7:56             ` Ingo Molnar
2004-05-28  9:38     ` Takao Indoh
2004-05-27 12:40 ` [4/4] " Takao Indoh
2004-05-27 13:34 ` [Document][PATCH]Diskdump " Takao Indoh
2004-06-03 13:10 ` [PATCH]Diskdump " Pavel Machek
2004-06-04  0:44   ` Takao Indoh
2004-06-04  9:33     ` Pavel Machek
     [not found] <20pwP-55v-5@gated-at.bofh.it>
     [not found] ` <20suK-7C5-11@gated-at.bofh.it>
     [not found]   ` <20tAB-5c-31@gated-at.bofh.it>
     [not found]     ` <20AiB-69m-17@gated-at.bofh.it>
2004-05-27 21:31       ` [3/4] " Andi Kleen
2004-05-28  7:38         ` Ingo Molnar

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=20040617120020.GA30691@infradead.org \
    --to=hch@infradead.org \
    --cc=ak@muc.de \
    --cc=indou.takao@soft.fujitsu.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    /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