linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Sergey Senozhatsky <senozhatsky@chromium.org>
To: Petr Mladek <pmladek@suse.com>
Cc: Sergey Senozhatsky <senozhatsky@chromium.org>,
	John Ogness <john.ogness@linutronix.de>,
	Steven Rostedt <rostedt@goodmis.org>,
	Thomas Gleixner <tglx@linutronix.de>,
	linux-kernel@vger.kernel.org,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Subject: Re: [PATCH printk v1 03/13] printk: use percpu flag instead of cpu_online()
Date: Wed, 16 Feb 2022 12:29:29 +0900	[thread overview]
Message-ID: <YgxvmepaTqRxwn/o@google.com> (raw)
In-Reply-To: <YguCuFYeZ52mkr4r@alley>

On (22/02/15 11:38), Petr Mladek wrote:
> On Mon 2022-02-14 16:35:18, Sergey Senozhatsky wrote:
> > On (22/02/11 17:05), Petr Mladek wrote:
> > > On Mon 2022-02-07 20:49:13, John Ogness wrote:
> > [..]
> > > The problem is the commit ac25575203c11145066ea ("[PATCH] CPU hotplug
> > > printk fix"). It suggests that per-CPU data of slab are freed during
> > > hotplug.
> > > 
> > > There are many other things that are manipulated during cpu hotplug.
> > > And there are the two notifiers "printk:dead" and "printk:online",
> > > see printk_late_init(). Maybe, we should use them to decide whether
> > > the non-trivial consoles are callable during CPU hotplug.
> > 
> > Great findings. Looks like we only set __printk_percpu_data_ready to
> > true and never set it back to false, relying on cpu_online() in such
> > cases. But here's the thing: we have printk_percpu_data_ready() in
> > __printk_recursion_counter() and in wake_up_klogd() and in
> > defer_console_output(), but why we never check __printk_percpu_data_ready
> > in __down_trylock_console_sem()/__up_console_sem() and more importantly
> > in console_trylock_spinning() and those do access this_cpu() in printk safe
> > enter/exit. Am I missing something?
> 
> Great point!
> 
> I am not 100% sure. But it seems that static per-CPU variables might
> actually be used since the boot.

Wow, this is great to learn. Thanks!

> This is from mm/percpu.c
> 
>  * There is special consideration for the first chunk which must handle
>  * the static percpu variables in the kernel image as allocation services
>  * are not online yet.  In short, the first chunk is structured like so:
>  *
>  *                  <Static | [Reserved] | Dynamic>
>  *
>  * The static data is copied from the original section managed by the
>  * linker.  The reserved section, if non-zero, primarily manages static
>  * percpu variables from kernel modules.  Finally, the dynamic section
>  * takes care of normal allocations.
> 
> 
> I thought that it might work only for CPU0. But it seems that it
> probably works for each possible cpu, see:
> 
> bool __is_kernel_percpu_address(unsigned long addr, unsigned long *can_addr)
> {
> #ifdef CONFIG_SMP
> 	const size_t static_size = __per_cpu_end - __per_cpu_start;
> 	void __percpu *base = __addr_to_pcpu_ptr(pcpu_base_addr);
> 	unsigned int cpu;
> 
> 	for_each_possible_cpu(cpu) {
> 		void *start = per_cpu_ptr(base, cpu);
> 		void *va = (void *)addr;
> 
> 		if (va >= start && va < start + static_size) {
> [...]
> }
> 
> and
> 
> /**
>  * is_kernel_percpu_address - test whether address is from static percpu area
>  * @addr: address to test
>  *
>  * Test whether @addr belongs to in-kernel static percpu area.  Module
>  * static percpu areas are not considered.  For those, use
>  * is_module_percpu_address().
>  *
>  * RETURNS:
>  * %true if @addr is from in-kernel static percpu area, %false otherwise.
>  */
> bool is_kernel_percpu_address(unsigned long addr)
> {
> 	return __is_kernel_percpu_address(addr, NULL);
> }
> 
> 
> Most likely, only dynamically allocated per-cpu variables have to wait
> until the per-cpu areas are initialized.
> 
> This might explain why there is no generic
> are_per_cpu_variables_ready() callback.
> 
> We should probably revisit the code and remove the fallback to
> normal static variables.

Agreed.

  reply	other threads:[~2022-02-16  3:29 UTC|newest]

Thread overview: 87+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-02-07 19:43 [PATCH printk v1 00/13] implement threaded console printing John Ogness
2022-02-07 19:43 ` [PATCH printk v1 01/13] printk: rename cpulock functions John Ogness
2022-02-11 12:44   ` Petr Mladek
2022-02-11 14:42     ` John Ogness
2022-02-11 20:57       ` Steven Rostedt
2022-02-11 21:04         ` Peter Zijlstra
2022-02-15  9:32           ` Petr Mladek
2022-02-15  9:13       ` Petr Mladek
2022-02-14  6:49     ` Sergey Senozhatsky
2022-02-14  9:45       ` John Ogness
2022-02-15  9:29       ` Petr Mladek
2022-02-16  3:27         ` Sergey Senozhatsky
2022-02-17 14:34         ` John Ogness
2022-02-07 19:43 ` [PATCH printk v1 02/13] printk: cpu sync always disable interrupts John Ogness
2022-02-11 12:58   ` Petr Mladek
2022-02-14  6:36     ` Sergey Senozhatsky
2022-02-07 19:43 ` [PATCH printk v1 03/13] printk: use percpu flag instead of cpu_online() John Ogness
2022-02-11 16:05   ` Petr Mladek
2022-02-14  7:08     ` Sergey Senozhatsky
2022-02-14  7:35     ` Sergey Senozhatsky
2022-02-15 10:38       ` Petr Mladek
2022-02-16  3:29         ` Sergey Senozhatsky [this message]
2022-03-02 14:21         ` John Ogness
2022-03-04 15:56           ` Petr Mladek
2022-03-05 17:05             ` Jason A. Donenfeld
2022-03-07 16:14               ` Petr Mladek
2022-02-16 13:58   ` two locations: was: " Petr Mladek
2022-03-02 14:49     ` John Ogness
2022-03-04 16:14       ` Petr Mladek
2022-03-07 10:06         ` John Ogness
2022-03-08 16:08       ` Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 04/13] printk: get caller_id/timestamp after migration disable John Ogness
2022-02-15  5:53   ` Sergey Senozhatsky
2022-02-15 11:56   ` Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 05/13] printk: call boot_delay_msec() in printk_delay() John Ogness
2022-02-15  5:58   ` Sergey Senozhatsky
2022-02-15 14:59     ` Petr Mladek
2022-02-16  3:21       ` Sergey Senozhatsky
2022-02-15 15:03   ` Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 06/13] printk: refactor and rework printing logic John Ogness
2022-02-16 15:43   ` Petr Mladek
2022-03-02 16:10     ` John Ogness
2022-02-07 19:43 ` [PATCH printk v1 07/13] printk: move buffer definitions into console_emit_next_record() caller John Ogness
2022-02-16 16:10   ` Petr Mladek
2022-03-02 16:25     ` John Ogness
2022-02-07 19:43 ` [PATCH printk v1 08/13] printk: add pr_flush() John Ogness
2022-02-17 10:11   ` Petr Mladek
2022-03-02 17:23     ` John Ogness
2022-03-04 13:24       ` Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 09/13] printk: add functions to allow direct printing John Ogness
2022-02-17 12:52   ` Petr Mladek
2022-02-18  9:00     ` David Laight
2022-02-18 12:52       ` Petr Mladek
2022-03-03 14:37     ` John Ogness
2022-02-07 19:43 ` [PATCH printk v1 10/13] printk: add kthread console printers John Ogness
2022-02-18  9:00   ` early start: was: " Petr Mladek
2022-02-18  9:04   ` start&stop: " Petr Mladek
2022-02-18  9:08   ` main loop: " Petr Mladek
2022-02-18  9:12   ` wake_up_all: " Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 11/13] printk: reimplement console_lock for proper kthread support John Ogness
2022-02-18 16:20   ` Petr Mladek
2022-02-18 21:41     ` John Ogness
2022-02-18 22:03       ` John Ogness
2022-02-22 11:42       ` Petr Mladek
2022-02-23 17:20         ` John Ogness
2022-02-24  8:27           ` Petr Mladek
2022-02-23 10:19   ` Petr Mladek
2022-03-09 13:56     ` John Ogness
2022-03-10 14:34       ` Petr Mladek
2022-03-10 16:08         ` John Ogness
2022-03-11 10:26           ` Petr Mladek
2022-03-11 13:28             ` John Ogness
2022-03-11 16:17               ` Petr Mladek
2022-03-11 22:21                 ` John Ogness
2022-03-14 14:08                   ` Petr Mladek
2022-03-14 14:43                     ` John Ogness
2022-03-14 15:53                       ` Petr Mladek
2022-03-11 18:41               ` Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 12/13] printk: remove @console_locked John Ogness
2022-02-23 12:17   ` Petr Mladek
2022-02-07 19:43 ` [PATCH printk v1 13/13] console: introduce CON_MIGHT_SLEEP for vt John Ogness
2022-02-23 13:37   ` Petr Mladek
2022-02-23 18:31     ` Greg Kroah-Hartman
     [not found] ` <20220208083620.2736-1-hdanton@sina.com>
2022-02-08 11:08   ` [PATCH printk v1 10/13] printk: add kthread console printers John Ogness
2022-02-08 14:53     ` Petr Mladek
2022-02-14  6:12       ` Sergey Senozhatsky
2022-02-14 10:02         ` Petr Mladek

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=YgxvmepaTqRxwn/o@google.com \
    --to=senozhatsky@chromium.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=john.ogness@linutronix.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pmladek@suse.com \
    --cc=rostedt@goodmis.org \
    --cc=tglx@linutronix.de \
    /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).