lkml.org 
[lkml]   [2020]   [Apr]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
Date
SubjectRe: [PATCHv2] printk: queue wake_up_klogd irq_work only if per-CPU areas are ready
On (20/04/09 12:25), Simon Kirby wrote:
> This causes "dmesg -w" or "cat /dev/kmsg" to not print new messages after
> dumping the current ring. I hit this on v5.5.9, v5.5.15, v5.6.3, and
> Linus HEAD. This prints no "hi":
>
> (sleep 1;echo hi > /dev/kmsg)& dmesg -w
>
> ...curiously, "strace dmesg -w" shows the data received once ^C is hit.
>
> Jann pointed me to this patch. Applying it or reverting 1b710b1b10eff9d4
> does fix it for me. However, Linus HEAD is still broken and, AFAICS,
> remains unreverted and unfixed in stable/linux-5.4.y through linux-5.6.y.
> It was introduced in 5.6 but backported to those, not just the LTS above.

I'm trying to land this patch. Give me a moment, I'll come back to
you shortly.

-ss

\
 
 \ /
  Last update: 2020-04-10 05:08    [W:0.083 / U:0.000 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site