linux-kernel - Re: [PATCH v5 03/12] mm: introduce AS_NO_DIRECT

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: <CA+EHjTxPfzDk=XmwS0uAtjwsYB829s1uZSMC6x3R6KGQ-SqjtQ@mail.gmail.com>
Date: Tue, 2 Sep 2025 08:59:13 +0100
From: Fuad Tabba <tabba@...gle.com>
To: "Roy, Patrick" <roypat@...zon.co.uk>
Cc: "ackerleytng@...gle.com" <ackerleytng@...gle.com>, "david@...hat.com" <david@...hat.com>, 
	"Manwaring, Derek" <derekmn@...zon.com>, "Thomson, Jack" <jackabt@...zon.co.uk>, 
	"Kalyazin, Nikita" <kalyazin@...zon.co.uk>, "kvm@...r.kernel.org" <kvm@...r.kernel.org>, 
	"kvmarm@...ts.linux.dev" <kvmarm@...ts.linux.dev>, 
	"linux-arm-kernel@...ts.infradead.org" <linux-arm-kernel@...ts.infradead.org>, 
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>, "linux-mm@...ck.org" <linux-mm@...ck.org>, 
	"pbonzini@...hat.com" <pbonzini@...hat.com>, "rppt@...nel.org" <rppt@...nel.org>, 
	"seanjc@...gle.com" <seanjc@...gle.com>, "vbabka@...e.cz" <vbabka@...e.cz>, "will@...nel.org" <will@...nel.org>, 
	"Cali, Marco" <xmarcalx@...zon.co.uk>
Subject: Re: [PATCH v5 03/12] mm: introduce AS_NO_DIRECT_MAP

Hi Patrick,

On Mon, 1 Sept 2025 at 15:56, Roy, Patrick <roypat@...zon.co.uk> wrote:
>
> On Mon, 2025-09-01 at 14:54 +0100, "Roy, Patrick" wrote:
> >
> > Hi Fuad!
> >
> > On Thu, 2025-08-28 at 11:21 +0100, Fuad Tabba wrote:
> >> Hi Patrick,
> >>
> >> On Thu, 28 Aug 2025 at 10:39, Roy, Patrick <roypat@...zon.co.uk> wrote:
> >>> diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
> >>> index 12a12dae727d..b52b28ae4636 100644
> >>> --- a/include/linux/pagemap.h
> >>> +++ b/include/linux/pagemap.h
> >>> @@ -211,6 +211,7 @@ enum mapping_flags {
> >>>                                    folio contents */
> >>>         AS_INACCESSIBLE = 8,    /* Do not attempt direct R/W access to the mapping */
> >>>         AS_WRITEBACK_MAY_DEADLOCK_ON_RECLAIM = 9,
> >>> +       AS_NO_DIRECT_MAP = 10,  /* Folios in the mapping are not in the direct map */
> >>>         /* Bits 16-25 are used for FOLIO_ORDER */
> >>>         AS_FOLIO_ORDER_BITS = 5,
> >>>         AS_FOLIO_ORDER_MIN = 16,
> >>> @@ -346,6 +347,21 @@ static inline bool mapping_writeback_may_deadlock_on_reclaim(struct address_spac
> >>>         return test_bit(AS_WRITEBACK_MAY_DEADLOCK_ON_RECLAIM, &mapping->flags);
> >>>  }
> >>>
> >>> +static inline void mapping_set_no_direct_map(struct address_space *mapping)
> >>> +{
> >>> +       set_bit(AS_NO_DIRECT_MAP, &mapping->flags);
> >>> +}
> >>> +
> >>> +static inline bool mapping_no_direct_map(struct address_space *mapping)
> >>> +{
> >>> +       return test_bit(AS_NO_DIRECT_MAP, &mapping->flags);
> >>> +}
> >>> +
> >>> +static inline bool vma_is_no_direct_map(const struct vm_area_struct *vma)
> >>> +{
> >>> +       return vma->vm_file && mapping_no_direct_map(vma->vm_file->f_mapping);
> >>> +}
> >>> +
> >> Any reason vma is const whereas mapping in the function that it calls
> >> (defined above it) isn't?
> >
> > Ah, I cannot say that that was a conscious decision, but rather an artifact of
> > the code that I looked at for reference when writing these two simply did it
> > this way.  Are you saying both should be const, or neither (in my mind, both
> > could be const, but the mapping_*() family of functions further up in this file
> > dont take const arguments, so I'm a bit unsure now)?
>
> Hah, just saw
> https://lore.kernel.org/linux-mm/20250901123028.3383461-3-max.kellermann@ionos.com/.
> Guess that means "both should be const" then :D

I don't have any strong preference regarding which way, as long as
it's consistent. The thing that should be avoided is having one
function with a parameter marked as const, pass that parameter (or
something derived from it), to a non-const function. Instead of
helping, this could cause a lot of headaches when trying to debug
things in the future, and figuring out what something that's supposed
to be "const" is being "corrupted".

Cheers,
/fuad


>
> >> Cheers,
> >> /fuad
> >
> > Best,
> > Patrick
>