linux-kernel - Re: [PATCH 2/3] GHES: Move memory initialisation to ghes

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: <87fuctp1eq.fsf@e105922-lin.cambridge.arm.com>
Date:   Tue, 15 Aug 2017 11:31:57 +0100
From:   Punit Agrawal <punit.agrawal@....com>
To:     Borislav Petkov <bp@...e.de>
Cc:     linux-acpi@...r.kernel.org, lorenzo.pieralisi@....com,
        sudeep.holla@....com, linux-kernel@...r.kernel.org,
        "Rafael J . Wysocki" <rjw@...ysocki.net>,
        James Morse <james.morse@....com>
Subject: Re: [PATCH 2/3] GHES: Move memory initialisation to ghes_probe()

[ thanks for dropping the bad email! ]

Borislav Petkov <bp@...e.de> writes:

> On Tue, Aug 15, 2017 at 11:10:05AM +0100, Punit Agrawal wrote:
>> In the absence of any GHES entries these are wasted.
>
> I know. This whole APEI init code would need a proper cleanup, like
> acpi_pci_root_init() calls acpi_hest_init() and it shouldn't have to
> communicate through a variable with GHES whether to init or not but it
> should initialize GHES itself. And ghes_init() being a device_initcall()
> is just yuck.

I'd seen the link between acpi_pci_root_init() and acpi_hest_init() but
didn't want to open another can of worms... :)

$SUBJECT was my attempt (small) at improving the situation but I guess I
didn't go far enough.

>
> Something for the todo list I guess...
>
>> The system does not provide the Hardware Error Source Table (HEST) which
>> is checked in the hest driver (drivers/acpi/apei/hest.c).
>> 
>> I think I'll go with your original suggestion to change hest_disable
>> from a boolean to something with more states. Re-checking for the HEST
>> table again feels like duplication.
>> 
>> Makes sense?
>
> Right, and then depending on the setting of hest_disable, either issue
> the "HEST is not enabled" message or not.

Ack!

>
> Thanks.
>
> -- 
> Regards/Gruss,
>     Boris.
>
> SUSE Linux GmbH, GF: Felix Imendörffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nürnberg)