linux-kernel - Re: [PATCH 7/8] efi: pstore: Follow convention for the efi-pstore backend name

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]

Message-ID: <28a9290e-b8af-93bb-26d3-3b12864ada4e@igalia.com>
Date: Tue, 4 Jun 2024 16:05:18 -0300
From: "Guilherme G. Piccoli" <gpiccoli@...lia.com>
To: Stephen Boyd <swboyd@...omium.org>
Cc: kernel-dev@...lia.com, kernel@...ccoli.net,
 linux-fsdevel@...r.kernel.org, keescook@...omium.org, anton@...msg.org,
 ccross@...roid.com, tony.luck@...el.com, linux-efi@...r.kernel.org,
 Ard Biesheuvel <ardb@...nel.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 7/8] efi: pstore: Follow convention for the efi-pstore
 backend name

On 03/06/2024 20:02, Stephen Boyd wrote:
> [...]
> This patch broke ChromeOS' crash reporter when running on EFI[1], which
> luckily isn't the typical mode of operation for Chromebooks. The problem
> was that we had hardcoded something like dmesg-efi-<number> into the
> regex logic that finds EFI pstore records. I didn't write the original
> code but I think the idea was to speed things up by parsing the
> filenames themselves to collect the files related to a crash record
> instead of opening and parsing the header from the files to figure out
> which file corresponds to which record.
> 
> I suspect the fix is pretty simple (make the driver name match either
> one via a regex) but I just wanted to drop a note here that this made
> some lives harder, not easier.

Oh, many apologies for that Stephen - of course if I was aware of the
hardcoding in the tool, I'd not mess it up or would fix the tooling
first, before changing the kernel code.

At least, as a bright side here you found the tool's limitation and
there's the obvious improvement/fix for that =)

Cheers,

Guilherme