linux-kernel - Re: [PATCH 0/7] HWPoison: Refactor get page interface

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]

Date:   Wed, 2 Dec 2020 14:41:39 +0100
From:   Oscar Salvador <osalvador@...e.de>
To:     Qian Cai <qcai@...hat.com>
Cc:     akpm@...ux-foundation.org, n-horiguchi@...jp.nec.com,
        linux-mm@...ck.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 0/7] HWPoison: Refactor get page interface

On Wed, Dec 02, 2020 at 08:34:57AM -0500, Qian Cai wrote:
> On Thu, 2020-11-19 at 11:57 +0100, Oscar Salvador wrote:
> > Hi,
> > 
> > following up on previous fix-ups an refactors, this patchset simplifies
> > the get page interface and removes the MF_COUNT_INCREASED trick we have
> > for soft offline.
> 
> Well, the madvise() EIO is back. I don't understand why we can't test it on a
> NUMA system before posting this over and over again.
> 
> # git clone https://e.coding.net/cailca/linux/mm
> # cd mm; make
> # ./ranbug 1 
> - start: migrate_huge_offline
> - use NUMA nodes 0,3.
> - mmap and free 8388608 bytes hugepages on node 0
> - mmap and free 8388608 bytes hugepages on node 3
> madvise: Input/output error

I tried it out myself enlarging the window race artificially but I was not able
to get -EIO anymore.
But as Vlastimil pointed out in the respective patch, it is better to keep the
page pinned for madvise.

I am planning to re-post leaving out the patches that remove the pinning.

Anyway, thanks for the report.

-- 
Oscar Salvador
SUSE L3