Crash tool issue since printk re-implementation

Crash tool issue since printk re-implementation

[Sebastien Fabre]

Hello,

Since kernel v5.0, PREEMPT_RT patches contain printk re-implementation. This re-implementation removes some log symbols in 
VMCOREINFO and crash tool cannot retrieve kernel logs (because symbols no more exist).
Is this impact on crash expected? If so, how can we retrieve a functional crash tool with kernel v5.0 (and following) ?

Thanks,
Fabre Sebastien

Re: Crash tool issue since printk re-implementation

[Sebastian Andrzej Siewior]

On 2020-02-12 12:10:23 [+0000], Sebastien Fabre wrote:
> Hello,
Hi,

> Since kernel v5.0, PREEMPT_RT patches contain printk re-implementation. This re-implementation removes some log symbols in 
> VMCOREINFO and crash tool cannot retrieve kernel logs (because symbols no more exist).
> Is this impact on crash expected? If so, how can we retrieve a functional crash tool with kernel v5.0 (and following) ?

It is known, yes. I'm aware of crash and the gdb scritps should also
suffer from this. 
I'm not sure anyone took a look at it (yet) and I don't think that the
`crash' tool will accept the patches unless final printk rework is
merged upstream.

> Thanks,
> Fabre Sebastien

Sebastian

Re: Crash tool issue since printk re-implementation

[John Ogness]

On 2020-02-14, Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:
> On 2020-02-12 12:10:23 [+0000], Sebastien Fabre wrote:
>> Since kernel v5.0, PREEMPT_RT patches contain printk
>> re-implementation. This re-implementation removes some log symbols in
>> VMCOREINFO and crash tool cannot retrieve kernel logs (because
>> symbols no more exist).  Is this impact on crash expected? If so, how
>> can we retrieve a functional crash tool with kernel v5.0 (and
>> following) ?
>
> It is known, yes. I'm aware of crash and the gdb scritps should also
> suffer from this. 

Perhaps this should be listed under the "Known issues" of the PREEMPT_RT
release announcements.

John Ogness

Re: Crash tool issue since printk re-implementation

[Gene Heskett]

On Friday 14 February 2020 08:42:57 John Ogness wrote:

> On 2020-02-14, Sebastian Andrzej Siewior <bigeasy@linutronix.de> wrote:
> > On 2020-02-12 12:10:23 [+0000], Sebastien Fabre wrote:
> >> Since kernel v5.0, PREEMPT_RT patches contain printk
> >> re-implementation. This re-implementation removes some log symbols
> >> in VMCOREINFO and crash tool cannot retrieve kernel logs (because
> >> symbols no more exist).  Is this impact on crash expected? If so,
> >> how can we retrieve a functional crash tool with kernel v5.0 (and
> >> following) ?
> >
> > It is known, yes. I'm aware of crash and the gdb scritps should also
> > suffer from this.
>
> Perhaps this should be listed under the "Known issues" of the
> PREEMPT_RT release announcements.
>
> John Ogness

I would like to agree, as I think that explains why an attempt to build 
5+ in this preempt-rt flavor, on an rpi4, for an rpi4 has been a no boot 
failure. 

Plus it appears I'd have to replace the bootloader because raspian puts 
the overlays in a /boot/overlays subdir whereas you are putting them 
in /boot (I think, I haven't found the .config magic twanger that puts 
them in the overlays subdir.)

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.
 - Louis D. Brandeis
Genes Web page <http://geneslinuxbox.net:6309/gene>