linux-ext4 - Re: [RFC 02/10] mm/hmm: Remove dead check for HugeTLB and FS DAX

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date: Fri, 12 Apr 2024 11:28:47 +1000
From: Alistair Popple <apopple@...dia.com>
To: Peter Xu <peterx@...hat.com>
Cc: Jason Gunthorpe <jgg@...dia.com>, linux-mm@...ck.org,
 david@...morbit.com, dan.j.williams@...el.com, jhubbard@...dia.com,
 rcampbell@...dia.com, willy@...radead.org, linux-fsdevel@...r.kernel.org,
 jack@...e.cz, djwong@...nel.org, hch@....de, david@...hat.com,
 ruansy.fnst@...itsu.com, nvdimm@...ts.linux.dev,
 linux-xfs@...r.kernel.org, linux-ext4@...r.kernel.org, jglisse@...hat.com
Subject: Re: [RFC 02/10] mm/hmm: Remove dead check for HugeTLB and FS DAX


Peter Xu <peterx@...hat.com> writes:

> On Thu, Apr 11, 2024 at 09:25:30AM -0300, Jason Gunthorpe wrote:
>> On Thu, Apr 11, 2024 at 10:57:23AM +1000, Alistair Popple wrote:
>> > pud_huge() returns true only for a HugeTLB page. pud_devmap() is only
>> > used by FS DAX pages. These two things are mutually exclusive so this
>> > code is dead code and can be removed.
>> 
>> I'm not sure this is true.. pud_huge() is mostly a misspelling of pud_leaf()..
>> 
>> > -	if (pud_huge(pud) && pud_devmap(pud)) {
>> 
>> I suspect this should be written as:
>> 
>>    if (pud_leaf(pud) && pud_devmap(pud)) {

Oh that makes a lot more sense. I'd taken the comment for pud_huge() at
face value (ie. that it's a hugetlbfs page) without digging further.

>> In line with Peter's work here:
>> 
>> https://lore.kernel.org/linux-mm/20240321220802.679544-1-peterx@redhat.com/
>
> Just to provide more information for Alistair, this patch already switched
> that over to a _leaf():

Got it, thanks (and apologies for missing my Cc on that).

> https://lore.kernel.org/r/20240318200404.448346-12-peterx@redhat.com
>
> That's in mm-unstable now, so should see that in a rebase.
>
> And btw it's great to see that pxx_devmap() can go away.

Yep, AFAICT pxx_devmap only exists to do this special FS DAX
refcounting. Once that is fixed it can go away.

> Thanks,