linux-kernel - Re: [PATCH] mm/percpu.c: use smarter memory allocation for struct pcpu_alloc

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: <nycvar.YSQ.7.76.1710032011070.5407@knanqh.ubzr>
Date:   Tue, 3 Oct 2017 20:13:37 -0400 (EDT)
From:   Nicolas Pitre <nicolas.pitre@...aro.org>
To:     Dennis Zhou <dennisszhou@...il.com>
cc:     Tejun Heo <tj@...nel.org>, Christoph Lameter <cl@...ux.com>,
        linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] mm/percpu.c: use smarter memory allocation for struct
 pcpu_alloc_info

On Tue, 3 Oct 2017, Dennis Zhou wrote:

> Hi Tejun,
> 
> On Tue, Oct 03, 2017 at 03:36:42PM -0700, Tejun Heo wrote:
> > > Subject: [PATCH] percpu: don't forget to free the temporary struct pcpu_alloc_info
> > 
> > So, IIRC, the error path is either boot fail or some serious bug in
> > arch code.  It really doesn't matter whether we free a page or not.
> >
> 
> In setup_per_cpu_area, a call to either pcpu_embed_first_chunk,
> pcpu_page_first_chunk, or pcpu_setup_first_chunk is made. The first two
> eventually call pcpu_setup_first_chunk with a pairing call to
> pcpu_free_alloc_info right after. This occurs in all implementations. It
> happens we don't have a pairing call to pcpu_free_alloc_info in the UP
> setup_per_cpu_area.

That was my conclusion too (albeit not stated as clearly) and what my 
second patch fixed.


Nicolas