linux-kernel - RE: [RFC][PATCH 2/2] write callback: Check if existing entry is erasable

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] [day] [month] [year] [list]

Date:	Fri, 6 Jul 2012 00:24:42 +0000
From:	Seiji Aguchi <seiji.aguchi@....com>
To:	"Luck, Tony" <tony.luck@...el.com>,
	Don Zickus <dzickus@...hat.com>,
	"Matthew Garrett (mjg@...hat.com)" <mjg@...hat.com>
CC:	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	"mikew@...gle.com" <mikew@...gle.com>,
	"dle-develop@...ts.sourceforge.net" 
	<dle-develop@...ts.sourceforge.net>,
	Satoru Moriya <satoru.moriya@....com>
Subject: RE: [RFC][PATCH 2/2] write callback: Check if existing entry is
 erasable

> 10K is almost certainly more than we need for 99.9% of problems ... I set the default there to test out that pstore would correctly
> break a dump into more than one back-end ERST record (about 7K) and never changed it back. So don't treat 10K with any magic
> reverence. It's easy to argue that a smaller number is good enough.

OK.

> There are certainly less over-write worries if you can handle a few (4, 5, 6) simultaneously logged errors of sufficient size to be useful
> (must capture all of the panic strings, backtrace and register dump plus "enough" lines before the panic to see any obvious issues).
>

To handle multiple logs, I will probably introduce a new kernel parameter, like efi_pstore_max_log_num.
Users can calculate  overall consumption of NVRAM for kmsg logging with it.
 
> If you only get to store two errors, then perhaps one non-over writable panic type entry, and one other "most recent" type entry?
> 
> With just one, like current EFI, then there are certainly hard choices that might not be the best for certain pathological situations.

I will consider a policy for multiple logging from now.
(In case where oops happens multiple times and kernel hangs, oldest oops may be informative.... I can't decide the policy at this time.)

Seiji


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/