From: Alexey Kardashevskiy <aik@ozlabs.ru>
To: Alexander Graf <agraf@suse.de>, qemu-devel@nongnu.org
Cc: Alex Williamson <alex.williamson@redhat.com>, qemu-ppc@nongnu.org
Subject: Re: [Qemu-devel] [RFC PATCH 06/10] spapr_rtas: Add Dynamic DMA windows (DDW) RTAS calls support
Date: Tue, 12 Aug 2014 01:34:49 +1000 [thread overview]
Message-ID: <53E8E299.8060201@ozlabs.ru> (raw)
In-Reply-To: <53E8AE4B.9030500@suse.de>
On 08/11/2014 09:51 PM, Alexander Graf wrote:
>
> On 31.07.14 11:34, Alexey Kardashevskiy wrote:
>> This adds support for Dynamic DMA Windows (DDW) option defined by
>> the SPAPR specification which allows to have additional DMA window(s)
>> which can support page sizes other than 4K.
>>
>> The existing implementation of DDW in the guest tries to create one huge
>> DMA window with 64K or 16MB pages and map the entire guest RAM to. If it
>> succeeds, the guest switches to dma_direct_ops and never calls
>> TCE hypercalls (H_PUT_TCE,...) again. This enables VFIO devices to use
>> the entire RAM and not waste time on map/unmap.
>>
>> This adds 4 RTAS handlers:
>> * ibm,query-pe-dma-window
>> * ibm,create-pe-dma-window
>> * ibm,remove-pe-dma-window
>> * ibm,reset-pe-dma-window
>> These are registered from type_init() callback.
>>
>> These RTAS handlers are implemented in a separate file to avoid polluting
>> spapr_iommu.c with PHB.
>>
>> Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
>> ---
>> hw/ppc/Makefile.objs | 3 +
>> hw/ppc/spapr_rtas_ddw.c | 296
>> ++++++++++++++++++++++++++++++++++++++++++++
>> include/hw/pci-host/spapr.h | 18 +++
>> include/hw/ppc/spapr.h | 6 +-
>> trace-events | 4 +
>> 5 files changed, 326 insertions(+), 1 deletion(-)
>> create mode 100644 hw/ppc/spapr_rtas_ddw.c
>>
>> diff --git a/hw/ppc/Makefile.objs b/hw/ppc/Makefile.objs
>> index edd44d0..9773294 100644
>> --- a/hw/ppc/Makefile.objs
>> +++ b/hw/ppc/Makefile.objs
>> @@ -7,6 +7,9 @@ obj-$(CONFIG_PSERIES) += spapr_pci.o
>> ifeq ($(CONFIG_PCI)$(CONFIG_PSERIES)$(CONFIG_LINUX), yyy)
>> obj-y += spapr_pci_vfio.o
>> endif
>> +ifeq ($(CONFIG_PCI)$(CONFIG_PSERIES), yy)
>> +obj-y += spapr_rtas_ddw.o
>> +endif
>> # PowerPC 4xx boards
>> obj-y += ppc405_boards.o ppc4xx_devs.o ppc405_uc.o ppc440_bamboo.o
>> obj-y += ppc4xx_pci.o
>> diff --git a/hw/ppc/spapr_rtas_ddw.c b/hw/ppc/spapr_rtas_ddw.c
>> new file mode 100644
>> index 0000000..943af2c
>> --- /dev/null
>> +++ b/hw/ppc/spapr_rtas_ddw.c
>> @@ -0,0 +1,296 @@
>> +/*
>> + * QEMU sPAPR Dynamic DMA windows support
>> + *
>> + * Copyright (c) 2014 Alexey Kardashevskiy, IBM Corporation.
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License as published by
>> + * the Free Software Foundation; either version 2 of the License,
>> + * or (at your option) any later version.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
>> + * GNU General Public License for more details.
>> + *
>> + * You should have received a copy of the GNU General Public License
>> + * along with this program; if not, see <http://www.gnu.org/licenses/>.
>> + */
>> +
>> +#include "hw/ppc/spapr.h"
>> +#include "hw/pci-host/spapr.h"
>> +#include "trace.h"
>> +
>> +static inline uint32_t spapr_iommu_fixmask(uint32_t cur_mask,
>> + struct ppc_one_seg_page_size
>> *sps,
>> + uint32_t query_mask,
>> + int shift,
>> + uint32_t add_mask)
>> +{
>> + if ((sps->page_shift == shift) && (query_mask & add_mask)) {
>> + cur_mask |= add_mask;
>> + }
>> + return cur_mask;
>> +}
>> +
>> +static void rtas_ibm_query_pe_dma_window(PowerPCCPU *cpu,
>> + sPAPREnvironment *spapr,
>> + uint32_t token, uint32_t nargs,
>> + target_ulong args,
>> + uint32_t nret, target_ulong rets)
>> +{
>> + CPUPPCState *env = &cpu->env;
>> + sPAPRPHBState *sphb;
>> + sPAPRPHBClass *spc;
>> + uint64_t buid;
>> + uint32_t addr, pgmask = 0;
>> + uint32_t windows_available = 0, page_size_mask = 0;
>> + long ret, i;
>> +
>> + if ((nargs != 3) || (nret != 5)) {
>> + goto param_error_exit;
>> + }
>> +
>> + buid = ((uint64_t)rtas_ld(args, 1) << 32) | rtas_ld(args, 2);
>> + addr = rtas_ld(args, 0);
>> + sphb = spapr_pci_find_phb(spapr, buid);
>> + if (!sphb) {
>> + goto param_error_exit;
>> + }
>> +
>> + spc = SPAPR_PCI_HOST_BRIDGE_GET_CLASS(sphb);
>> + if (!spc->ddw_query) {
>> + goto hw_error_exit;
>> + }
>> +
>> + ret = spc->ddw_query(sphb, &windows_available, &page_size_mask);
>> + trace_spapr_iommu_ddw_query(buid, addr, windows_available,
>> + page_size_mask, pgmask, ret);
>> + if (ret) {
>> + goto hw_error_exit;
>> + }
>> +
>> + /* DBG! */
>> + if (!(page_size_mask & DDW_PGSIZE_16M)) {
>> + goto hw_error_exit;
>> + }
>> +
>> + /* Work out biggest possible page size */
>> + for (i = 0; i < PPC_PAGE_SIZES_MAX_SZ; i++) {
>> + int j;
>> + struct ppc_one_seg_page_size *sps = &env->sps.sps[i];
>> + const struct { int shift; uint32_t mask; } masks[] = {
>> + { 12, DDW_PGSIZE_4K },
>> + { 16, DDW_PGSIZE_64K },
>> + { 24, DDW_PGSIZE_16M },
>> + { 25, DDW_PGSIZE_32M },
>> + { 26, DDW_PGSIZE_64M },
>> + { 27, DDW_PGSIZE_128M },
>> + { 28, DDW_PGSIZE_256M },
>> + { 34, DDW_PGSIZE_16G },
>> + };
>> + for (j = 0; j < ARRAY_SIZE(masks); ++j) {
>> + pgmask = spapr_iommu_fixmask(pgmask, sps, page_size_mask,
>> + masks[j].shift, masks[j].mask);
>> + }
>> + }
>> +
>> + rtas_st(rets, 0, RTAS_OUT_SUCCESS);
>> + rtas_st(rets, 1, windows_available);
>> + /* Return maximum number as all RAM was 4K pages */
>> + rtas_st(rets, 2, ram_size >> SPAPR_TCE_PAGE_SHIFT);
>> + rtas_st(rets, 3, pgmask);
>> + rtas_st(rets, 4, pgmask); /* DMA migration mask */
>> + return;
>> +
>> +hw_error_exit:
>> + rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
>> + return;
>> +
>> +param_error_exit:
>> + rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
>> +}
>> +
>> +static void rtas_ibm_create_pe_dma_window(PowerPCCPU *cpu,
>> + sPAPREnvironment *spapr,
>> + uint32_t token, uint32_t nargs,
>> + target_ulong args,
>> + uint32_t nret, target_ulong rets)
>> +{
>> + sPAPRPHBState *sphb;
>> + sPAPRPHBClass *spc;
>> + sPAPRTCETable *tcet = NULL;
>> + uint32_t addr, page_shift, window_shift, liobn;
>> + uint64_t buid;
>> + long ret;
>> +
>> + if ((nargs != 5) || (nret != 4)) {
>> + goto param_error_exit;
>> + }
>> +
>> + buid = ((uint64_t)rtas_ld(args, 1) << 32) | rtas_ld(args, 2);
>> + addr = rtas_ld(args, 0);
>> + sphb = spapr_pci_find_phb(spapr, buid);
>> + if (!sphb) {
>> + goto param_error_exit;
>> + }
>> +
>> + spc = SPAPR_PCI_HOST_BRIDGE_GET_CLASS(sphb);
>> + if (!spc->ddw_create) {
>> + goto hw_error_exit;
>> + }
>> +
>> + page_shift = rtas_ld(args, 3);
>> + window_shift = rtas_ld(args, 4);
>> + liobn = sphb->dma_liobn + 0x10000;
>
> What offset is this?
Some new LIOBN. Can be +1. May be worth defining as a macro.
>
>> +
>> + ret = spc->ddw_create(sphb, page_shift, window_shift, liobn, &tcet);
>> + trace_spapr_iommu_ddw_create(buid, addr, 1 << page_shift,
>> + 1 << window_shift,
>
> 1ULL? Otherwise 16G pages (and windows) won't work.
Right. Thanks. I'll fix. 16_G_ _pages_ are not supported anyway though.
--
Alexey
next prev parent reply other threads:[~2014-08-11 15:35 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-07-31 9:34 [Qemu-devel] [RFC PATCH 00/10] spapr: vfio: Enable Dynamic DMA windows (DDW) Alexey Kardashevskiy
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 01/10] qom: Make object_child_foreach safe for objects removal Alexey Kardashevskiy
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 02/10] spapr_iommu: Disable in-kernel IOMMU tables for >4GB windows Alexey Kardashevskiy
2014-08-12 1:17 ` David Gibson
2014-08-12 7:32 ` Alexey Kardashevskiy
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 03/10] spapr_pci: Make find_phb()/find_dev() public Alexey Kardashevskiy
2014-08-11 11:39 ` Alexander Graf
2014-08-11 14:56 ` Alexey Kardashevskiy
2014-08-11 17:16 ` Alexander Graf
2014-08-12 1:19 ` David Gibson
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 04/10] spapr_iommu: Make spapr_tce_find_by_liobn() public Alexey Kardashevskiy
2014-08-12 1:19 ` David Gibson
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 05/10] linux headers update for DDW Alexey Kardashevskiy
2014-08-12 1:20 ` David Gibson
2014-08-12 7:16 ` Alexey Kardashevskiy
2014-08-13 3:23 ` David Gibson
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 06/10] spapr_rtas: Add Dynamic DMA windows (DDW) RTAS calls support Alexey Kardashevskiy
2014-08-11 11:51 ` Alexander Graf
2014-08-11 15:34 ` Alexey Kardashevskiy [this message]
2014-08-12 1:45 ` David Gibson
2014-08-12 7:25 ` Alexey Kardashevskiy
2014-08-13 3:27 ` David Gibson
2014-08-14 8:29 ` Alexey Kardashevskiy
2014-08-15 0:04 ` David Gibson
2014-08-15 3:09 ` Alexey Kardashevskiy
2014-08-15 4:20 ` David Gibson
2014-08-15 5:27 ` Alexey Kardashevskiy
2014-08-15 5:30 ` David Gibson
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 07/10] spapr: Add "ddw" machine option Alexey Kardashevskiy
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 08/10] spapr_pci: Enable DDW Alexey Kardashevskiy
2014-08-11 11:59 ` Alexander Graf
2014-08-11 15:26 ` Alexey Kardashevskiy
2014-08-11 17:29 ` Alexander Graf
2014-08-12 0:13 ` Alexey Kardashevskiy
2014-08-12 3:59 ` Alexey Kardashevskiy
2014-08-12 9:36 ` Alexander Graf
2014-08-12 2:10 ` David Gibson
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 09/10] spapr_pci_vfio: " Alexey Kardashevskiy
2014-08-11 12:02 ` Alexander Graf
2014-08-11 15:01 ` Alexey Kardashevskiy
2014-08-11 17:30 ` Alexander Graf
2014-08-12 0:03 ` Alexey Kardashevskiy
2014-08-12 9:37 ` Alexander Graf
2014-08-12 15:10 ` Alexey Kardashevskiy
2014-08-12 15:28 ` Alexander Graf
2014-08-13 0:18 ` Alexey Kardashevskiy
2014-08-14 13:38 ` Alexander Graf
2014-08-15 0:09 ` David Gibson
2014-08-15 3:22 ` Alexey Kardashevskiy
2014-08-15 3:16 ` Alexey Kardashevskiy
2014-08-15 7:37 ` Alexander Graf
2014-08-12 2:14 ` David Gibson
2014-07-31 9:34 ` [Qemu-devel] [RFC PATCH 10/10] vfio: Enable DDW ioctls to VFIO IOMMU driver Alexey Kardashevskiy
2014-08-05 1:30 ` [Qemu-devel] [RFC PATCH 00/10] spapr: vfio: Enable Dynamic DMA windows (DDW) Alexey Kardashevskiy
2014-08-10 23:50 ` Alexey Kardashevskiy
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=53E8E299.8060201@ozlabs.ru \
--to=aik@ozlabs.ru \
--cc=agraf@suse.de \
--cc=alex.williamson@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.