[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <YUHd3kGUngaB758q@8bytes.org>
Date: Wed, 15 Sep 2021 13:49:50 +0200
From: Joerg Roedel <joro@...tes.org>
To: Mike Rapoport <rppt@...nel.org>
Cc: x86@...nel.org, Thomas Gleixner <tglx@...utronix.de>,
Ingo Molnar <mingo@...hat.com>, Borislav Petkov <bp@...en8.de>,
hpa@...or.com, jroedel@...e.de,
Andrew Morton <akpm@...ux-foundation.org>,
Brijesh Singh <brijesh.singh@....com>,
linux-kernel@...r.kernel.org, stable@...r.kernel.org
Subject: Re: [PATCH] x86/64/mm: Map all kernel memory into trampoline_pgd
Hi Mike,
On Tue, Sep 14, 2021 at 10:52:39AM +0300, Mike Rapoport wrote:
> On Mon, Sep 13, 2021 at 11:52:36AM +0200, Joerg Roedel wrote:
> > + for (i = pgd_index(__PAGE_OFFSET); i < PTRS_PER_PGD; i++)
> > + trampoline_pgd[i] = init_top_pgt[i].pgd;
>
> Don't we need to update the trampoline_pgd in sync_global_pgds() as well?
No, the trampoline_pgd is setup after preallocate_vmalloc_pages(), so
everything that would need synchronization is already in the reference
page-table.
Regards,
Joerg
Powered by blists - more mailing lists