From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:51473) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1b5nCI-0007V7-6W for qemu-devel@nongnu.org; Thu, 26 May 2016 00:47:04 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1b5nCF-0007MX-4g for qemu-devel@nongnu.org; Thu, 26 May 2016 00:47:01 -0400 Date: Thu, 26 May 2016 13:39:02 +1000 From: David Gibson Message-ID: <20160526033902.GA17226@voom.fritz.box> References: <1462344751-28281-1-git-send-email-aik@ozlabs.ru> <1462344751-28281-9-git-send-email-aik@ozlabs.ru> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha1; protocol="application/pgp-signature"; boundary="6JAe/4X0qssghT2u" Content-Disposition: inline In-Reply-To: <1462344751-28281-9-git-send-email-aik@ozlabs.ru> Subject: Re: [Qemu-devel] [PATCH qemu v16 08/19] spapr_iommu: Introduce "enabled" state for TCE table List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Alexey Kardashevskiy Cc: qemu-devel@nongnu.org, qemu-ppc@nongnu.org, Alexander Graf , Alex Williamson , Paolo Bonzini --6JAe/4X0qssghT2u Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Wed, May 04, 2016 at 04:52:20PM +1000, Alexey Kardashevskiy wrote: > Currently TCE tables are created once at start and their sizes never > change. We are going to change that by introducing a Dynamic DMA windows > support where DMA configuration may change during the guest execution. >=20 > This changes spapr_tce_new_table() to create an empty zero-size IOMMU > memory region (IOMMU MR). Only LIOBN is assigned by the time of creation. > It still will be called once at the owner object (VIO or PHB) creation. >=20 > This introduces an "enabled" state for TCE table objects with two > helper functions - spapr_tce_table_enable()/spapr_tce_table_disable(). > - spapr_tce_table_enable() receives TCE table parameters, allocates > a guest view of the TCE table (in the user space or KVM) and > sets the correct size on the IOMMU MR. > - spapr_tce_table_disable() disposes the table and resets the IOMMU MR > size. >=20 > This changes the PHB reset handler to do the default DMA initialization > instead of spapr_phb_realize(). This does not make differenct now but > later with more than just one DMA window, we will have to remove them all > and create the default one on a system reset. >=20 > No visible change in behaviour is expected except the actual table > will be reallocated every reset. We might optimize this later. >=20 > The other way to implement this would be dynamically create/remove > the TCE table QOM objects but this would make migration impossible > as the migration code expects all QOM objects to exist at the receiver > so we have to have TCE table objects created when migration begins. >=20 > spapr_tce_table_do_enable() is separated from from spapr_tce_table_enable= () > as later it will be called at the sPAPRTCETable post-migration stage when > it already has all the properties set after the migration; the same is > done for spapr_tce_table_disable(). >=20 > Signed-off-by: Alexey Kardashevskiy > Reviewed-by: David Gibson > --- > Changes: > v15: > * made adjustments after removing spapr_phb_dma_window_enable() >=20 > v14: > * added spapr_tce_table_do_disable(), will make difference in following > patch with fully dynamic table migration >=20 > # Conflicts: > # hw/ppc/spapr_pci.c > --- > hw/ppc/spapr_iommu.c | 86 ++++++++++++++++++++++++++++++++++++--------= ------ > hw/ppc/spapr_pci.c | 8 +++-- > hw/ppc/spapr_vio.c | 8 ++--- > include/hw/ppc/spapr.h | 10 +++--- > 4 files changed, 75 insertions(+), 37 deletions(-) >=20 > diff --git a/hw/ppc/spapr_iommu.c b/hw/ppc/spapr_iommu.c > index 8132f64..9bcd3f6 100644 > --- a/hw/ppc/spapr_iommu.c > +++ b/hw/ppc/spapr_iommu.c > @@ -17,6 +17,7 @@ > * License along with this library; if not, see . > */ > #include "qemu/osdep.h" > +#include "qemu/error-report.h" > #include "hw/hw.h" > #include "sysemu/kvm.h" > #include "hw/qdev.h" > @@ -174,15 +175,9 @@ static int spapr_tce_table_realize(DeviceState *dev) > sPAPRTCETable *tcet =3D SPAPR_TCE_TABLE(dev); > =20 > tcet->fd =3D -1; > - tcet->table =3D spapr_tce_alloc_table(tcet->liobn, > - tcet->page_shift, > - tcet->nb_table, > - &tcet->fd, > - tcet->need_vfio); > - > + tcet->need_vfio =3D false; > memory_region_init_iommu(&tcet->iommu, OBJECT(dev), &spapr_iommu_ops, > - "iommu-spapr", > - (uint64_t)tcet->nb_table << tcet->page_shif= t); > + "iommu-spapr", 0); > =20 > QLIST_INSERT_HEAD(&spapr_tce_tables, tcet, list); > =20 > @@ -224,14 +219,10 @@ void spapr_tce_set_need_vfio(sPAPRTCETable *tcet, b= ool need_vfio) > tcet->table =3D newtable; > } > =20 > -sPAPRTCETable *spapr_tce_new_table(DeviceState *owner, uint32_t liobn, > - uint64_t bus_offset, > - uint32_t page_shift, > - uint32_t nb_table, > - bool need_vfio) > +sPAPRTCETable *spapr_tce_new_table(DeviceState *owner, uint32_t liobn) > { > sPAPRTCETable *tcet; > - char tmp[64]; > + char tmp[32]; > =20 > if (spapr_tce_find_by_liobn(liobn)) { > fprintf(stderr, "Attempted to create TCE table with duplicate" > @@ -239,16 +230,8 @@ sPAPRTCETable *spapr_tce_new_table(DeviceState *owne= r, uint32_t liobn, > return NULL; > } > =20 > - if (!nb_table) { > - return NULL; > - } > - > tcet =3D SPAPR_TCE_TABLE(object_new(TYPE_SPAPR_TCE_TABLE)); > tcet->liobn =3D liobn; > - tcet->bus_offset =3D bus_offset; > - tcet->page_shift =3D page_shift; > - tcet->nb_table =3D nb_table; > - tcet->need_vfio =3D need_vfio; > =20 > snprintf(tmp, sizeof(tmp), "tce-table-%x", liobn); > object_property_add_child(OBJECT(owner), tmp, OBJECT(tcet), NULL); > @@ -258,14 +241,69 @@ sPAPRTCETable *spapr_tce_new_table(DeviceState *own= er, uint32_t liobn, > return tcet; > } > =20 > +static void spapr_tce_table_do_enable(sPAPRTCETable *tcet) > +{ > + if (!tcet->nb_table) { > + return; > + } > + > + tcet->table =3D spapr_tce_alloc_table(tcet->liobn, > + tcet->page_shift, > + tcet->nb_table, > + &tcet->fd, > + tcet->need_vfio); > + > + memory_region_set_size(&tcet->iommu, > + (uint64_t)tcet->nb_table << tcet->page_shift); > + > + tcet->enabled =3D true; > +} > + > +void spapr_tce_table_enable(sPAPRTCETable *tcet, > + uint32_t page_shift, uint64_t bus_offset, > + uint32_t nb_table) > +{ > + if (tcet->enabled) { > + error_report("Warning: trying to enable already enabled TCE tabl= e"); > + return; > + } > + > + tcet->bus_offset =3D bus_offset; > + tcet->page_shift =3D page_shift; > + tcet->nb_table =3D nb_table; > + > + spapr_tce_table_do_enable(tcet); > +} > + > +static void spapr_tce_table_do_disable(sPAPRTCETable *tcet) > +{ > + memory_region_set_size(&tcet->iommu, 0); > + > + spapr_tce_free_table(tcet->table, tcet->fd, tcet->nb_table); > + tcet->fd =3D -1; > + tcet->table =3D NULL; > + tcet->enabled =3D false; > + tcet->bus_offset =3D 0; > + tcet->page_shift =3D 0; > + tcet->nb_table =3D 0; > +} > + > +static void spapr_tce_table_disable(sPAPRTCETable *tcet) > +{ > + if (!tcet->enabled) { > + error_report("Warning: trying to disable already disabled TCE ta= ble"); > + return; > + } > + spapr_tce_table_do_disable(tcet); > +} > + > static void spapr_tce_table_unrealize(DeviceState *dev, Error **errp) > { > sPAPRTCETable *tcet =3D SPAPR_TCE_TABLE(dev); > =20 > QLIST_REMOVE(tcet, list); > =20 > - spapr_tce_free_table(tcet->table, tcet->fd, tcet->nb_table); > - tcet->fd =3D -1; > + spapr_tce_table_disable(tcet); This should probably be do_disable(), or you'll get a spurious error if you start and stop a VM but don't enable the table in between, or if the guest disables all the tables before the shutdown. > } > =20 > MemoryRegion *spapr_tce_get_iommu(sPAPRTCETable *tcet) > diff --git a/hw/ppc/spapr_pci.c b/hw/ppc/spapr_pci.c > index 742d127..beeac06 100644 > --- a/hw/ppc/spapr_pci.c > +++ b/hw/ppc/spapr_pci.c > @@ -1464,8 +1464,7 @@ static void spapr_phb_realize(DeviceState *dev, Err= or **errp) > } > =20 > nb_table =3D sphb->dma_win_size >> SPAPR_TCE_PAGE_SHIFT; > - tcet =3D spapr_tce_new_table(DEVICE(sphb), sphb->dma_liobn, > - 0, SPAPR_TCE_PAGE_SHIFT, nb_table, false); > + tcet =3D spapr_tce_new_table(DEVICE(sphb), sphb->dma_liobn); > if (!tcet) { > error_setg(errp, "Unable to create TCE table for %s", > sphb->dtbusname); > @@ -1473,7 +1472,10 @@ static void spapr_phb_realize(DeviceState *dev, Er= ror **errp) > } > =20 > /* Register default 32bit DMA window */ > - memory_region_add_subregion(&sphb->iommu_root, sphb->dma_win_addr, > + spapr_tce_table_enable(tcet, SPAPR_TCE_PAGE_SHIFT, sphb->dma_win_add= r, > + nb_table); > + > + memory_region_add_subregion(&sphb->iommu_root, tcet->bus_offset, > spapr_tce_get_iommu(tcet)); > =20 > sphb->msi =3D g_hash_table_new_full(g_int_hash, g_int_equal, g_free,= g_free); > diff --git a/hw/ppc/spapr_vio.c b/hw/ppc/spapr_vio.c > index 8aa021f..a7d49a0 100644 > --- a/hw/ppc/spapr_vio.c > +++ b/hw/ppc/spapr_vio.c > @@ -482,11 +482,9 @@ static void spapr_vio_busdev_realize(DeviceState *qd= ev, Error **errp) > memory_region_add_subregion_overlap(&dev->mrroot, 0, &dev->mrbyp= ass, 1); > address_space_init(&dev->as, &dev->mrroot, qdev->id); > =20 > - dev->tcet =3D spapr_tce_new_table(qdev, liobn, > - 0, > - SPAPR_TCE_PAGE_SHIFT, > - pc->rtce_window_size >> > - SPAPR_TCE_PAGE_SHIFT, false); > + dev->tcet =3D spapr_tce_new_table(qdev, liobn); > + spapr_tce_table_enable(dev->tcet, SPAPR_TCE_PAGE_SHIFT, 0, > + pc->rtce_window_size >> SPAPR_TCE_PAGE_SH= IFT); > dev->tcet->vdev =3D dev; > memory_region_add_subregion_overlap(&dev->mrroot, 0, > spapr_tce_get_iommu(dev->tce= t), 2); > diff --git a/include/hw/ppc/spapr.h b/include/hw/ppc/spapr.h > index 815d5ee..0140810 100644 > --- a/include/hw/ppc/spapr.h > +++ b/include/hw/ppc/spapr.h > @@ -534,6 +534,7 @@ typedef struct sPAPRTCETable sPAPRTCETable; > =20 > struct sPAPRTCETable { > DeviceState parent; > + bool enabled; > uint32_t liobn; > uint32_t nb_table; > uint64_t bus_offset; > @@ -561,11 +562,10 @@ void spapr_events_fdt_skel(void *fdt, uint32_t epow= _irq); > int spapr_h_cas_compose_response(sPAPRMachineState *sm, > target_ulong addr, target_ulong size, > bool cpu_update, bool memory_update); > -sPAPRTCETable *spapr_tce_new_table(DeviceState *owner, uint32_t liobn, > - uint64_t bus_offset, > - uint32_t page_shift, > - uint32_t nb_table, > - bool need_vfio); > +sPAPRTCETable *spapr_tce_new_table(DeviceState *owner, uint32_t liobn); > +void spapr_tce_table_enable(sPAPRTCETable *tcet, > + uint32_t page_shift, uint64_t bus_offset, > + uint32_t nb_table); > void spapr_tce_set_need_vfio(sPAPRTCETable *tcet, bool need_vfio); > =20 > MemoryRegion *spapr_tce_get_iommu(sPAPRTCETable *tcet); --=20 David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson --6JAe/4X0qssghT2u Content-Type: application/pgp-signature; name="signature.asc" -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iQIcBAEBAgAGBQJXRm/WAAoJEGw4ysog2bOS3d0QALWKSNFvHDG0kXp25iqwXdVR uJVzc/HWfIF1l8fxU+DiIGqLb85aJTYzV0HsMfncPBUJJHFWNuNkp3i2zTfj1jNd V/ixV2O8ftifGd9z9HnfEIRZgbMT4pKQNZ+y0GpwLznGUedwruEyTibMrwUwgEKB /n0ReeqTA2d/Xa3A1ZwzNvNqkceNCoVTXI2g1kpvflzTi+51KLJjvZpbr0m3F8i9 dUuMYB4Ncv0N4oMzoNm0n0cBuSgNT8IEkD9xi6wFlIrFdLie5c4RnH171Pxu/R2D kEZBwr29RtAzV6e1NWh3cKhshbx8OzBQJwV5HViQC69kvAAmXtPOFXTalnnbkeb6 X1uVatSezezl9UaoRfHzQ5TOsWBDwkG5av4ybNuLQ/2HDSG1dsxMZhO+M6/mU+FK MdHcvpWCqyEu9/dild0ybZpkK/eTQBOahZ9fVx3yCy20E5iy5pnapOynJFeH0nhB 4uPUzI8qFTpRo3hizKL92hSxYnGqORBkHyg4AW0vfH74sD6gi9P9LtS5yN1Ml8kP IrVpeNwfGngpoRe4o7HxWndZAOTbugTLYTysSsZVXX/KY+uYgsEbBMpobcT6xIM+ 4NAoD8NNWiYVH3QpcZCPdPdVZButsK6nfmjeVyKDGpigbjxGD9J5kpV+sf/guRJp HSbwxWrdEBYZiz9trIPZ =PT26 -----END PGP SIGNATURE----- --6JAe/4X0qssghT2u--