linux-kernel - Re: [PATCH] mm: Fix pud_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: <CA+icZUVvMBTtF+p5obq-eiZzccRo8g8vmiczDGUxZnj-fu8U9Q@mail.gmail.com>
Date:   Thu, 4 Jun 2020 18:58:29 +0200
From:   Sedat Dilek <sedat.dilek@...il.com>
To:     Mike Rapoport <rppt@...nel.org>
Cc:     Joerg Roedel <joro@...tes.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        peterz@...radead.org, jroedel@...e.de,
        Andy Lutomirski <luto@...nel.org>,
        Abdul Haleem <abdhalee@...ux.vnet.ibm.com>,
        Satheesh Rajendran <sathnaga@...ux.vnet.ibm.com>,
        Stephen Rothwell <sfr@...b.auug.org.au>,
        manvanth@...ux.vnet.ibm.com, linux-next@...r.kernel.org,
        Steven Rostedt <rostedt@...dmis.org>,
        linuxppc-dev@...ts.ozlabs.org, hch@....de,
        linux-arch@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-mm@...ck.org
Subject: Re: [PATCH] mm: Fix pud_alloc_track()

On Thu, Jun 4, 2020 at 6:49 PM Mike Rapoport <rppt@...nel.org> wrote:
>
> On Thu, Jun 04, 2020 at 09:44:46AM +0200, Joerg Roedel wrote:
> > From: Joerg Roedel <jroedel@...e.de>
> >
> > The pud_alloc_track() needs to do different checks based on whether
> > __ARCH_HAS_5LEVEL_HACK is defined, like it already does in
> > pud_alloc(). Otherwise it causes boot failures on PowerPC.
> >
> > Provide the correct implementations for both possible settings of
> > __ARCH_HAS_5LEVEL_HACK to fix the boot problems.
>
> There is a patch in mmotm [1] that completely removes
> __ARCH_HAS_5LEVEL_HACK which is a part of the series [2] that updates
> p4d folding accross architectures. This should fix boot on PowerPC and
> the addition of pXd_alloc_track() for __ARCH_HAS_5LEVEL_HACK wouldn't be
> necessary.
>
>
> [1] https://github.com/hnaz/linux-mm/commit/cfae68792af3731ac902ea6ba5ed8df5a0f6bd2f
> [2] https://lore.kernel.org/kvmarm/20200414153455.21744-1-rppt@kernel.org/
>

That link shows an overview of v4 and is easily downloadable as a
single mbox file.
See " Series = mm: remove __ARCH_HAS_5LEVEL_HACK"

- Sedat -

[1] https://lore.kernel.org/patchwork/project/lkml/list/?series=438627

> > Reported-by: Abdul Haleem <abdhalee@...ux.vnet.ibm.com>
> > Tested-by: Abdul Haleem <abdhalee@...ux.vnet.ibm.com>
> > Tested-by: Satheesh Rajendran <sathnaga@...ux.vnet.ibm.com>
> > Fixes: d8626138009b ("mm: add functions to track page directory modifications")
> > Signed-off-by: Joerg Roedel <jroedel@...e.de>
> > ---
> >  include/asm-generic/5level-fixup.h |  5 +++++
> >  include/linux/mm.h                 | 26 +++++++++++++-------------
> >  2 files changed, 18 insertions(+), 13 deletions(-)
> >
> > diff --git a/include/asm-generic/5level-fixup.h b/include/asm-generic/5level-fixup.h
> > index 58046ddc08d0..afbab31fbd7e 100644
> > --- a/include/asm-generic/5level-fixup.h
> > +++ b/include/asm-generic/5level-fixup.h
> > @@ -17,6 +17,11 @@
> >       ((unlikely(pgd_none(*(p4d))) && __pud_alloc(mm, p4d, address)) ? \
> >               NULL : pud_offset(p4d, address))
> >
> > +#define pud_alloc_track(mm, p4d, address, mask)                                      \
> > +     ((unlikely(pgd_none(*(p4d))) &&                                         \
> > +       (__pud_alloc(mm, p4d, address) || ({*(mask)|=PGTBL_P4D_MODIFIED;0;})))?       \
> > +       NULL : pud_offset(p4d, address))
> > +
> >  #define p4d_alloc(mm, pgd, address)          (pgd)
> >  #define p4d_alloc_track(mm, pgd, address, mask)      (pgd)
> >  #define p4d_offset(pgd, start)                       (pgd)
> > diff --git a/include/linux/mm.h b/include/linux/mm.h
> > index 66e0977f970a..ad3b31c5bcc3 100644
> > --- a/include/linux/mm.h
> > +++ b/include/linux/mm.h
> > @@ -2088,35 +2088,35 @@ static inline pud_t *pud_alloc(struct mm_struct *mm, p4d_t *p4d,
> >               NULL : pud_offset(p4d, address);
> >  }
> >
> > -static inline p4d_t *p4d_alloc_track(struct mm_struct *mm, pgd_t *pgd,
> > +static inline pud_t *pud_alloc_track(struct mm_struct *mm, p4d_t *p4d,
> >                                    unsigned long address,
> >                                    pgtbl_mod_mask *mod_mask)
> > -
> >  {
> > -     if (unlikely(pgd_none(*pgd))) {
> > -             if (__p4d_alloc(mm, pgd, address))
> > +     if (unlikely(p4d_none(*p4d))) {
> > +             if (__pud_alloc(mm, p4d, address))
> >                       return NULL;
> > -             *mod_mask |= PGTBL_PGD_MODIFIED;
> > +             *mod_mask |= PGTBL_P4D_MODIFIED;
> >       }
> >
> > -     return p4d_offset(pgd, address);
> > +     return pud_offset(p4d, address);
> >  }
> >
> > -#endif /* !__ARCH_HAS_5LEVEL_HACK */
> > -
> > -static inline pud_t *pud_alloc_track(struct mm_struct *mm, p4d_t *p4d,
> > +static inline p4d_t *p4d_alloc_track(struct mm_struct *mm, pgd_t *pgd,
> >                                    unsigned long address,
> >                                    pgtbl_mod_mask *mod_mask)
> > +
> >  {
> > -     if (unlikely(p4d_none(*p4d))) {
> > -             if (__pud_alloc(mm, p4d, address))
> > +     if (unlikely(pgd_none(*pgd))) {
> > +             if (__p4d_alloc(mm, pgd, address))
> >                       return NULL;
> > -             *mod_mask |= PGTBL_P4D_MODIFIED;
> > +             *mod_mask |= PGTBL_PGD_MODIFIED;
> >       }
> >
> > -     return pud_offset(p4d, address);
> > +     return p4d_offset(pgd, address);
> >  }
> >
> > +#endif /* !__ARCH_HAS_5LEVEL_HACK */
> > +
> >  static inline pmd_t *pmd_alloc(struct mm_struct *mm, pud_t *pud, unsigned long address)
> >  {
> >       return (unlikely(pud_none(*pud)) && __pmd_alloc(mm, pud, address))?
> > --
> > 2.26.2
> >
>
> --
> Sincerely yours,
> Mike.