[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <5dd6e694-8cf9-4b1c-ae83-088b6bd22a17@de.ibm.com>
Date: Thu, 6 Nov 2025 17:22:33 +0100
From: Christian Borntraeger <borntraeger@...ibm.com>
To: Claudio Imbrenda <imbrenda@...ux.ibm.com>, kvm@...r.kernel.org
Cc: linux-kernel@...r.kernel.org, linux-s390@...r.kernel.org,
frankja@...ux.ibm.com, nsg@...ux.ibm.com, nrb@...ux.ibm.com,
seiden@...ux.ibm.com, schlameuss@...ux.ibm.com, hca@...ux.ibm.com,
svens@...ux.ibm.com, agordeev@...ux.ibm.com, gor@...ux.ibm.com,
david@...hat.com, gerald.schaefer@...ux.ibm.com
Subject: Re: [PATCH v3 21/23] KVM: s390: Enable 1M pages for gmap
Am 06.11.25 um 17:11 schrieb Claudio Imbrenda:
> While userspace is allowed to have pages of any size, the new gmap
> would always use 4k pages to back the guest.
>
> Enable 1M pages for gmap.
>
> This allows 1M pages to be used to back a guest when userspace is using
> 1M pages for the corresponding addresses (e.g. THP or hugetlbfs).
>
> Remove the limitation that disallowed having nested guests and
> hugepages at the same time.
Nice. This might allow us to enable hpage=1 as new default as soon as
things stabilize.
We would also be able to use 2GB huge pages for the qemu mapping?
With this patch we then have 1MB pages for the guest mapping.
As a future improvement, maybe also allow guest mapping 2GB pages.
Powered by blists - more mailing lists