From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg Kurz Date: Tue, 12 Nov 2019 12:01:09 +0000 Subject: Re: [PATCH] KVM: PPC: Book3S HV: XIVE: Free previous EQ page when setting up a new one Message-Id: <20191112130109.4d7bcf05@bahia.lan> List-Id: References: <157346576671.818016.10401178701091199969.stgit@bahia.lan> <3373a85a-09bb-3345-ef27-68177c360786@kaod.org> In-Reply-To: <3373a85a-09bb-3345-ef27-68177c360786@kaod.org> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: =?UTF-8?B?Q8OpZHJpYw==?= Le Goater Cc: Paul Mackerras , Michael Ellerman , Benjamin Herrenschmidt , David Gibson , Lijun Pan , Satheesh Rajendran , kvm-ppc@vger.kernel.org, linuxppc-dev@lists.ozlabs.org, stable@vger.kernel.org, linux-kernel@vger.kernel.org On Mon, 11 Nov 2019 12:26:25 +0100 C=C3=A9dric Le Goater wrote: > On 11/11/2019 10:49, Greg Kurz wrote: > > The EQ page is allocated by the guest and then passed to the hypervisor > > with the H_INT_SET_QUEUE_CONFIG hcall. A reference is taken on the page > > before handing it over to the HW. This reference is dropped either when > > the guest issues the H_INT_RESET hcall or when the KVM device is releas= ed. > > But, the guest can legitimately call H_INT_SET_QUEUE_CONFIG several tim= es > > to reset the EQ (vCPU hot unplug) or set a new EQ (guest reboot). In bo= th > > cases the EQ page reference is leaked. This is especially visible when > > the guest memory is backed with huge pages: start a VM up to the guest > > userspace, either reboot it or unplug a vCPU, quit QEMU. The leak is > > observed by comparing the value of HugePages_Free in /proc/meminfo befo= re > > and after the VM is run. > >=20 > > Note that the EQ reset path seems to be calling put_page() but this is > > done after xive_native_configure_queue() which clears the qpage field > > in the XIVE queue structure, ie. the put_page() block is a nop and the > > previous page pointer was just overwritten anyway. In the other case of > > configuring a new EQ page, nothing seems to be done to release the old > > one. >=20 > Yes. Nice catch. I think we should try to fix the problem differently.=20 >=20 > The routine xive_native_configure_queue() is only suited for XIVE=20 > drivers doing their own EQ page allocation: Linux PowerNV and the=20 > KVM XICS-over-XIVE device. The KVM XIVE device acts as a proxy for=20 > the guest OS doing the allocation and it has different needs. >=20 Well xive_native_configure_queue() is at least partially suited for all thr= ee drivers since they use it to configure the EQ. But it doesn't address the p= age allocation/de-allocation which is indeed different. > Having a specific xive_native_configure_queue() for the KVM XIVE=20 > device seems overkill. May be, we could introduce a helper routine=20 > in KVM XIVE device calling xive_native_configure_queue() and handling=20 > the page reference how it should be ? That is to drop the previous > page reference in case of a change on q->qpage. >=20 Yes, that seems better. I'll post a v2 with the helper you've mailed me. >=20 > Also, we should try to preserve the previous setting until the whole=20 > configuration is in place. That seems possible up to the call to=20 > xive_native_configure_queue(). If kvmppc_xive_attach_escalation() > fails I think it is too late, as the HW has been configured by=20 > xive_native_configure_queue(), and we should just cleanup everything.=20 >=20 > Thanks, >=20 > C.=20 >=20 >=20 > > Fix both cases by always calling put_page() on the existing EQ page in > > kvmppc_xive_native_set_queue_config(). This is a seemless change for the > > EQ reset case. However this causes xive_native_configure_queue() to be > > called twice for the new EQ page case: one time to reset the EQ and ano= ther > > time to configure the new page. This is needed because we cannot release > > the EQ page before calling xive_native_configure_queue() since it may s= till > > be used by the HW. We cannot modify xive_native_configure_queue() to dr= op > > the reference either because this function is also used by the XICS-on-= XIVE > > device which requires free_pages() instead of put_page(). This isn't a = big > > deal anyway since H_INT_SET_QUEUE_CONFIG isn't a hot path. > >=20 > > Reported-by: Satheesh Rajendran > > Cc: stable@vger.kernel.org # v5.2 > > Fixes: 13ce3297c576 ("KVM: PPC: Book3S HV: XIVE: Add controls for the E= Q configuration") > > Signed-off-by: Greg Kurz > > --- > > arch/powerpc/kvm/book3s_xive_native.c | 21 ++++++++++++--------- > > 1 file changed, 12 insertions(+), 9 deletions(-) > >=20 > > diff --git a/arch/powerpc/kvm/book3s_xive_native.c b/arch/powerpc/kvm/b= ook3s_xive_native.c > > index 34bd123fa024..8ab908d23dc2 100644 > > --- a/arch/powerpc/kvm/book3s_xive_native.c > > +++ b/arch/powerpc/kvm/book3s_xive_native.c > > @@ -570,10 +570,12 @@ static int kvmppc_xive_native_set_queue_config(st= ruct kvmppc_xive *xive, > > __func__, server, priority, kvm_eq.flags, > > kvm_eq.qshift, kvm_eq.qaddr, kvm_eq.qtoggle, kvm_eq.qindex); > > =20 > > - /* reset queue and disable queueing */ > > - if (!kvm_eq.qshift) { > > - q->guest_qaddr =3D 0; > > - q->guest_qshift =3D 0; > > + /* > > + * Reset queue and disable queueing. It will be re-enabled > > + * later on if the guest is configuring a new EQ page. > > + */ > > + if (q->guest_qshift) { > > + page =3D virt_to_page(q->qpage); > > =20 > > rc =3D xive_native_configure_queue(xc->vp_id, q, priority, > > NULL, 0, true); > > @@ -583,12 +585,13 @@ static int kvmppc_xive_native_set_queue_config(st= ruct kvmppc_xive *xive, > > return rc; > > } > > =20 > > - if (q->qpage) { > > - put_page(virt_to_page(q->qpage)); > > - q->qpage =3D NULL; > > - } > > + put_page(page); > > =20 > > - return 0; > > + if (!kvm_eq.qshift) { > > + q->guest_qaddr =3D 0; > > + q->guest_qshift =3D 0; > > + return 0; > > + } > > } > > =20 > > /* > >=20 >=20