linux-kernel - Re: Boot warning at rcu_check_gp_start

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <6A6B0325-64C4-4470-91B4-37104CF8DA1A@lca.pw>
Date:   Tue, 21 Jan 2020 14:09:05 -0500
From:   Qian Cai <cai@....pw>
To:     "Paul E. McKenney" <paulmck@...nel.org>
Cc:     rcu@...r.kernel.org, LKML <linux-kernel@...r.kernel.org>
Subject: Re: Boot warning at rcu_check_gp_start_stall()



> On Jan 21, 2020, at 11:15 AM, Paul E. McKenney <paulmck@...nel.org> wrote:
> 
> On Tue, Jan 21, 2020 at 09:37:13AM -0500, Qian Cai wrote:
>> 
>> 
>>> On Jan 21, 2020, at 9:19 AM, Paul E. McKenney <paulmck@...nel.org> wrote:
>>> 
>>> One approach would be to boot with rcupdate.rcu_cpu_stall_timeout=300,
>>> which would allow more time.
>> 
>> It works for me if once that warning triggered,  give a bit information about adjusting the parameter when debugging options are on to suppress the warning due to expected long boot.
> 
> Indeed.  300 seconds as shown above is currently the maximum, but
> please let me know if it needs to be increased.  This module parameter
> is writable after boot via sysfs, so maybe that could be part of the
> workaround.
> 
>>> Longer term, I could suppress this warning during boot when
>>> CONFIG_EFI_PGT_DUMP=y, but that sounds quite specific.  Alternatively,
>>> I could provide a Kconfig option that suppressed this during boot
>>> that was selected by whatever long-running boot-time Kconfig option
>>> needed it.  Yet another approach would be for long-running operations
>>> like efi_dump_pagetable() to suppress stalls on entry and re-enable them
>>> upon exit.
>>> 
>>> Thoughts?
>> 
>> None of the options sounds particularly better for me because there could come up with other options may trigger this, memtest comes in mind, for example. Then, it is a bit of pain to maintain of unknown.
> 
> I was afraid of that.  ;-)
> 
> Could you please send me the full dmesg up to that point?  No promises,
> but it might well be that I can make some broad-spectrum adjustment
> within RCU.  Only one way to find out…

https://cailca.github.io/files/dmesg.txt