Re: [PATCH 2/3] mm: Use line-buffered printk() for show_free_areas().

[Petr Mladek]

On Fri 2018-11-02 22:31:56, Tetsuo Handa wrote:
> syzbot is sometimes getting mixed output like below due to concurrent
> printk(). Mitigate such output by using line-buffered printk() API.
> 
>   Node 0 DMA: 1*4kB (U) 0*8kB 0*16kB 1*32kB 
>   syz-executor0: page allocation failure: order:0, mode:0x484020(GFP_ATOMIC|__GFP_COMP), nodemask=(null)
>   (U) 
>   syz-executor0 cpuset=
>   2*64kB 
>   syz0
>   (U) 
>    mems_allowed=0
>   1*128kB 
>   CPU: 0 PID: 7592 Comm: syz-executor0 Not tainted 4.19.0-rc6+ #118
>   (U) 
>   Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
>   1*256kB (U) 
>   Call Trace:
>   0*512kB 
>    <IRQ>
>   1*1024kB 
>    __dump_stack lib/dump_stack.c:77 [inline]
>    dump_stack+0x1c4/0x2b4 lib/dump_stack.c:113
>   (U) 
>   1*2048kB 
>    warn_alloc.cold.119+0xb7/0x1bd mm/page_alloc.c:3426
>   (M) 
> 
> Signed-off-by: Tetsuo Handa <penguin-kernel@xxxxxxxxxxxxxxxxxxx>
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> ---
>  mm/page_alloc.c | 32 +++++++++++++++++---------------
>  1 file changed, 17 insertions(+), 15 deletions(-)
> 
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index a919ba5..4411d5a 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -4694,10 +4694,10 @@ unsigned long nr_free_pagecache_pages(void)
>  	return nr_free_zone_pages(gfp_zone(GFP_HIGHUSER_MOVABLE));
>  }
>  
> -static inline void show_node(struct zone *zone)
> +static inline void show_node(struct zone *zone, struct printk_buffer *buf)
>  {
>  	if (IS_ENABLED(CONFIG_NUMA))
> -		printk("Node %d ", zone_to_nid(zone));
> +		printk_buffered(buf, "Node %d ", zone_to_nid(zone));

The conversion looks fine to me. I just think about renaming
printk_buffered to bprintk().

Best Regards,
Petr