linux-kernel - Re: [PATCH] mm/vmalloc: Check absolute error return from vmap_[p4d|pud|pmd|pte]

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: <20190614134220.GL32656@bombadil.infradead.org>
Date:   Fri, 14 Jun 2019 06:42:20 -0700
From:   Matthew Wilcox <willy@...radead.org>
To:     Anshuman Khandual <anshuman.khandual@....com>
Cc:     Roman Penyaev <rpenyaev@...e.de>, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org,
        Rick Edgecombe <rick.p.edgecombe@...el.com>,
        Andrey Ryabinin <aryabinin@...tuozzo.com>,
        Mike Rapoport <rppt@...ux.ibm.com>,
        Roman Gushchin <guro@...com>, Michal Hocko <mhocko@...e.com>,
        "Uladzislau Rezki (Sony)" <urezki@...il.com>,
        Andrew Morton <akpm@...ux-foundation.org>
Subject: Re: [PATCH] mm/vmalloc: Check absolute error return from
 vmap_[p4d|pud|pmd|pte]_range()

On Fri, Jun 14, 2019 at 10:57:42AM +0530, Anshuman Khandual wrote:
> 
> 
> On 06/13/2019 09:01 PM, Matthew Wilcox wrote:
> > On Thu, Jun 13, 2019 at 08:51:17PM +0530, Anshuman Khandual wrote:
> >> acceptable ? What we have currently is wrong where vmap_pmd_range() could
> >> just wrap EBUSY as ENOMEM and send up the call chain.
> > 
> > It's not wrong.  We do it in lots of places.  Unless there's a caller
> > which really needs to know the difference, it's often better than
> > returning the "real error".
> 
> I can understand the fact that because there are no active users of this
> return code, the current situation has been alright. But then I fail to
> understand how can EBUSY be made ENOMEM and let the caller to think that
> vmap_page_rage() failed because of lack of memory when it is clearly not
> the case. It is really surprising how it can be acceptable inside kernel
> (init_mm) page table functions which need to be thorough enough.

It's a corollary of Steinbach's Guideline for Systems Programming.
There is no possible way to handle this error because this error is
never supposed to happen.  So we may as well return a different error
that will still lead to the caller doing the right thing.