qemu-e2k/hw/ppc/spapr_pci.c

2533 lines
80 KiB
C
Raw Normal View History

/*
* QEMU sPAPR PCI host originated from Uninorth PCI host
*
* Copyright (c) 2011 Alexey Kardashevskiy, IBM Corporation.
* Copyright (C) 2011 David Gibson, IBM Corporation.
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
#include "qemu/osdep.h"
2016-03-14 09:01:28 +01:00
#include "qapi/error.h"
#include "hw/irq.h"
#include "hw/sysbus.h"
#include "migration/vmstate.h"
#include "hw/pci/pci.h"
#include "hw/pci/msi.h"
#include "hw/pci/msix.h"
#include "hw/pci/pci_host.h"
#include "hw/ppc/spapr.h"
#include "hw/pci-host/spapr.h"
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
#include "exec/ram_addr.h"
#include <libfdt.h>
#include "trace.h"
#include "qemu/error-report.h"
#include "qemu/module.h"
#include "qapi/qmp/qerror.h"
#include "hw/ppc/fdt.h"
#include "hw/pci/pci_bridge.h"
#include "hw/pci/pci_bus.h"
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
#include "hw/pci/pci_ids.h"
#include "hw/ppc/spapr_drc.h"
#include "hw/qdev-properties.h"
#include "sysemu/device_tree.h"
#include "sysemu/kvm.h"
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
#include "sysemu/hostmem.h"
#include "sysemu/numa.h"
#include "hw/ppc/spapr_numa.h"
#include "qemu/log.h"
/* Copied from the kernel arch/powerpc/platforms/pseries/msi.c */
#define RTAS_QUERY_FN 0
#define RTAS_CHANGE_FN 1
#define RTAS_RESET_FN 2
#define RTAS_CHANGE_MSI_FN 3
#define RTAS_CHANGE_MSIX_FN 4
/* Interrupt types to return on RTAS_CHANGE_* */
#define RTAS_TYPE_MSI 1
#define RTAS_TYPE_MSIX 2
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *spapr_pci_find_phb(SpaprMachineState *spapr, uint64_t buid)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
QLIST_FOREACH(sphb, &spapr->phbs, list) {
if (sphb->buid != buid) {
continue;
}
return sphb;
}
return NULL;
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
PCIDevice *spapr_pci_find_dev(SpaprMachineState *spapr, uint64_t buid,
uint32_t config_addr)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = spapr_pci_find_phb(spapr, buid);
PCIHostState *phb = PCI_HOST_BRIDGE(sphb);
int bus_num = (config_addr >> 16) & 0xFF;
int devfn = (config_addr >> 8) & 0xFF;
if (!phb) {
return NULL;
}
return pci_find_device(phb->bus, bus_num, devfn);
}
static uint32_t rtas_pci_cfgaddr(uint32_t arg)
{
/* This handles the encoding of extended config space addresses */
return ((arg >> 20) & 0xf00) | (arg & 0xff);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void finish_read_pci_config(SpaprMachineState *spapr, uint64_t buid,
uint32_t addr, uint32_t size,
target_ulong rets)
{
PCIDevice *pci_dev;
uint32_t val;
if ((size != 1) && (size != 2) && (size != 4)) {
/* access must be 1, 2 or 4 bytes */
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
pci_dev = spapr_pci_find_dev(spapr, buid, addr);
addr = rtas_pci_cfgaddr(addr);
if (!pci_dev || (addr % size) || (addr >= pci_config_size(pci_dev))) {
/* Access must be to a valid device, within bounds and
* naturally aligned */
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
val = pci_host_config_read_common(pci_dev, addr,
pci_config_size(pci_dev), size);
rtas_st(rets, 0, RTAS_OUT_SUCCESS);
rtas_st(rets, 1, val);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void rtas_ibm_read_pci_config(PowerPCCPU *cpu, SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args,
uint32_t nret, target_ulong rets)
{
uint64_t buid;
uint32_t size, addr;
if ((nargs != 4) || (nret != 2)) {
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
buid = rtas_ldq(args, 1);
size = rtas_ld(args, 3);
addr = rtas_ld(args, 0);
finish_read_pci_config(spapr, buid, addr, size, rets);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void rtas_read_pci_config(PowerPCCPU *cpu, SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args,
uint32_t nret, target_ulong rets)
{
uint32_t size, addr;
if ((nargs != 2) || (nret != 2)) {
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
size = rtas_ld(args, 1);
addr = rtas_ld(args, 0);
finish_read_pci_config(spapr, 0, addr, size, rets);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void finish_write_pci_config(SpaprMachineState *spapr, uint64_t buid,
uint32_t addr, uint32_t size,
uint32_t val, target_ulong rets)
{
PCIDevice *pci_dev;
if ((size != 1) && (size != 2) && (size != 4)) {
/* access must be 1, 2 or 4 bytes */
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
pci_dev = spapr_pci_find_dev(spapr, buid, addr);
addr = rtas_pci_cfgaddr(addr);
if (!pci_dev || (addr % size) || (addr >= pci_config_size(pci_dev))) {
/* Access must be to a valid device, within bounds and
* naturally aligned */
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
pci_host_config_write_common(pci_dev, addr, pci_config_size(pci_dev),
val, size);
rtas_st(rets, 0, RTAS_OUT_SUCCESS);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void rtas_ibm_write_pci_config(PowerPCCPU *cpu, SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args,
uint32_t nret, target_ulong rets)
{
uint64_t buid;
uint32_t val, size, addr;
if ((nargs != 5) || (nret != 1)) {
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
buid = rtas_ldq(args, 1);
val = rtas_ld(args, 4);
size = rtas_ld(args, 3);
addr = rtas_ld(args, 0);
finish_write_pci_config(spapr, buid, addr, size, val, rets);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void rtas_write_pci_config(PowerPCCPU *cpu, SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args,
uint32_t nret, target_ulong rets)
{
uint32_t val, size, addr;
if ((nargs != 3) || (nret != 1)) {
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
val = rtas_ld(args, 2);
size = rtas_ld(args, 1);
addr = rtas_ld(args, 0);
finish_write_pci_config(spapr, 0, addr, size, val, rets);
}
/*
* Set MSI/MSIX message data.
* This is required for msi_notify()/msix_notify() which
* will write at the addresses via spapr_msi_write().
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
*
* If hwaddr == 0, all entries will have .data == first_irq i.e.
* table will be reset.
*/
spapr-pci: rework MSI/MSIX On the sPAPR platform a guest allocates MSI/MSIX vectors via RTAS hypercalls which return global IRQ numbers to a guest so it only operates with those and never touches MSIMessage. Therefore MSIMessage handling is completely hidden in QEMU. Previously every sPAPR PCI host bridge implemented its own MSI window to catch msi_notify()/msix_notify() calls from QEMU devices (virtio-pci or vfio) and route them to the guest via qemu_pulse_irq(). MSIMessage used to be encoded as: .addr - address within the PHB MSI window; .data - the device index on PHB plus vector number. The MSI MR write function translated this MSIMessage to a global IRQ number and called qemu_pulse_irq(). However the total number of IRQs is not really big (at the moment it is 1024 IRQs starting from 4096) and even 16bit data field of MSIMessage seems to be enough to store an IRQ number there. This simplifies MSI handling in sPAPR PHB. Specifically, this does: 1. remove a MSI window from a PHB; 2. add a single memory region for all MSIs to sPAPREnvironment and spapr_pci_msi_init() to initialize it; 3. encode MSIMessage as: * .addr - a fixed address of SPAPR_PCI_MSI_WINDOW==0x40000000000ULL; * .data as an IRQ number. 4. change IRQ allocator to align first IRQ number in a block for MSI. MSI uses lower bits to specify the vector number so the first IRQ has to be aligned. MSIX does not need any special allocator though. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Reviewed-by: Anthony Liguori <aliguori@us.ibm.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2013-07-12 09:38:24 +02:00
static void spapr_msi_setmsg(PCIDevice *pdev, hwaddr addr, bool msix,
unsigned first_irq, unsigned req_num)
{
unsigned i;
spapr-pci: rework MSI/MSIX On the sPAPR platform a guest allocates MSI/MSIX vectors via RTAS hypercalls which return global IRQ numbers to a guest so it only operates with those and never touches MSIMessage. Therefore MSIMessage handling is completely hidden in QEMU. Previously every sPAPR PCI host bridge implemented its own MSI window to catch msi_notify()/msix_notify() calls from QEMU devices (virtio-pci or vfio) and route them to the guest via qemu_pulse_irq(). MSIMessage used to be encoded as: .addr - address within the PHB MSI window; .data - the device index on PHB plus vector number. The MSI MR write function translated this MSIMessage to a global IRQ number and called qemu_pulse_irq(). However the total number of IRQs is not really big (at the moment it is 1024 IRQs starting from 4096) and even 16bit data field of MSIMessage seems to be enough to store an IRQ number there. This simplifies MSI handling in sPAPR PHB. Specifically, this does: 1. remove a MSI window from a PHB; 2. add a single memory region for all MSIs to sPAPREnvironment and spapr_pci_msi_init() to initialize it; 3. encode MSIMessage as: * .addr - a fixed address of SPAPR_PCI_MSI_WINDOW==0x40000000000ULL; * .data as an IRQ number. 4. change IRQ allocator to align first IRQ number in a block for MSI. MSI uses lower bits to specify the vector number so the first IRQ has to be aligned. MSIX does not need any special allocator though. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Reviewed-by: Anthony Liguori <aliguori@us.ibm.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2013-07-12 09:38:24 +02:00
MSIMessage msg = { .address = addr, .data = first_irq };
if (!msix) {
msi_set_message(pdev, msg);
trace_spapr_pci_msi_setup(pdev->name, 0, msg.address);
return;
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
for (i = 0; i < req_num; ++i) {
msix_set_message(pdev, i, msg);
trace_spapr_pci_msi_setup(pdev->name, i, msg.address);
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
if (addr) {
++msg.data;
}
}
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void rtas_ibm_change_msi(PowerPCCPU *cpu, SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineClass *smc = SPAPR_MACHINE_GET_CLASS(spapr);
uint32_t config_addr = rtas_ld(args, 0);
uint64_t buid = rtas_ldq(args, 1);
unsigned int func = rtas_ld(args, 3);
unsigned int req_num = rtas_ld(args, 4); /* 0 == remove all */
unsigned int seq_num = rtas_ld(args, 5);
unsigned int ret_intr_type;
unsigned int irq, max_irqs = 0;
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *phb = NULL;
PCIDevice *pdev = NULL;
SpaprPciMsi *msi;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
int *config_addr_key;
Error *err = NULL;
int i;
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
/* Fins SpaprPhbState */
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
phb = spapr_pci_find_phb(spapr, buid);
if (phb) {
pdev = spapr_pci_find_dev(spapr, buid, config_addr);
}
if (!phb || !pdev) {
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
return;
}
switch (func) {
case RTAS_CHANGE_FN:
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
if (msi_present(pdev)) {
ret_intr_type = RTAS_TYPE_MSI;
} else if (msix_present(pdev)) {
ret_intr_type = RTAS_TYPE_MSIX;
} else {
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
return;
}
break;
case RTAS_CHANGE_MSI_FN:
if (msi_present(pdev)) {
ret_intr_type = RTAS_TYPE_MSI;
} else {
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
return;
}
break;
case RTAS_CHANGE_MSIX_FN:
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
if (msix_present(pdev)) {
ret_intr_type = RTAS_TYPE_MSIX;
} else {
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
return;
}
break;
default:
error_report("rtas_ibm_change_msi(%u) is not implemented", func);
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
return;
}
msi = (SpaprPciMsi *) g_hash_table_lookup(phb->msi, &config_addr);
/* Releasing MSIs */
if (!req_num) {
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
if (!msi) {
trace_spapr_pci_msi("Releasing wrong config", config_addr);
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
if (msi_present(pdev)) {
spapr_msi_setmsg(pdev, 0, false, 0, 0);
}
if (msix_present(pdev)) {
spapr_msi_setmsg(pdev, 0, true, 0, 0);
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
g_hash_table_remove(phb->msi, &config_addr);
trace_spapr_pci_msi("Released MSIs", config_addr);
rtas_st(rets, 0, RTAS_OUT_SUCCESS);
rtas_st(rets, 1, 0);
return;
}
/* Enabling MSI */
/* Check if the device supports as many IRQs as requested */
if (ret_intr_type == RTAS_TYPE_MSI) {
max_irqs = msi_nr_vectors_allocated(pdev);
} else if (ret_intr_type == RTAS_TYPE_MSIX) {
max_irqs = pdev->msix_entries_nr;
}
if (!max_irqs) {
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
error_report("Requested interrupt type %d is not enabled for device %x",
ret_intr_type, config_addr);
rtas_st(rets, 0, -1); /* Hardware error */
return;
}
/* Correct the number if the guest asked for too many */
if (req_num > max_irqs) {
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
trace_spapr_pci_msi_retry(config_addr, req_num, max_irqs);
req_num = max_irqs;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
irq = 0; /* to avoid misleading trace */
goto out;
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
/* Allocate MSIs */
if (smc->legacy_irq_allocation) {
irq = spapr_irq_find(spapr, req_num, ret_intr_type == RTAS_TYPE_MSI,
&err);
} else {
irq = spapr_irq_msi_alloc(spapr, req_num,
ret_intr_type == RTAS_TYPE_MSI, &err);
}
if (err) {
error_reportf_err(err, "Can't allocate MSIs for device %x: ",
config_addr);
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
for (i = 0; i < req_num; i++) {
spapr_irq_claim(spapr, irq + i, false, &err);
if (err) {
if (i) {
spapr_irq_free(spapr, irq, i);
}
if (!smc->legacy_irq_allocation) {
spapr_irq_msi_free(spapr, irq, req_num);
}
error_reportf_err(err, "Can't allocate MSIs for device %x: ",
config_addr);
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
}
/* Release previous MSIs */
if (msi) {
g_hash_table_remove(phb->msi, &config_addr);
}
/* Setup MSI/MSIX vectors in the device (via cfgspace or MSIX BAR) */
spapr_msi_setmsg(pdev, SPAPR_PCI_MSI_WINDOW, ret_intr_type == RTAS_TYPE_MSIX,
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
irq, req_num);
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
/* Add MSI device to cache */
msi = g_new(SpaprPciMsi, 1);
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
msi->first_irq = irq;
msi->num = req_num;
config_addr_key = g_new(int, 1);
*config_addr_key = config_addr;
g_hash_table_insert(phb->msi, config_addr_key, msi);
out:
rtas_st(rets, 0, RTAS_OUT_SUCCESS);
rtas_st(rets, 1, req_num);
rtas_st(rets, 2, ++seq_num);
if (nret > 3) {
rtas_st(rets, 3, ret_intr_type);
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
trace_spapr_pci_rtas_ibm_change_msi(config_addr, func, req_num, irq);
}
static void rtas_ibm_query_interrupt_source_number(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token,
uint32_t nargs,
target_ulong args,
uint32_t nret,
target_ulong rets)
{
uint32_t config_addr = rtas_ld(args, 0);
uint64_t buid = rtas_ldq(args, 1);
unsigned int intr_src_num = -1, ioa_intr_num = rtas_ld(args, 3);
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *phb = NULL;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
PCIDevice *pdev = NULL;
SpaprPciMsi *msi;
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
/* Find SpaprPhbState */
phb = spapr_pci_find_phb(spapr, buid);
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
if (phb) {
pdev = spapr_pci_find_dev(spapr, buid, config_addr);
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
}
if (!phb || !pdev) {
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
return;
}
/* Find device descriptor and start IRQ */
msi = (SpaprPciMsi *) g_hash_table_lookup(phb->msi, &config_addr);
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
if (!msi || !msi->first_irq || !msi->num || (ioa_intr_num >= msi->num)) {
trace_spapr_pci_msi("Failed to return vector", config_addr);
rtas_st(rets, 0, RTAS_OUT_HW_ERROR);
return;
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
intr_src_num = msi->first_irq + ioa_intr_num;
trace_spapr_pci_rtas_ibm_query_interrupt_source_number(ioa_intr_num,
intr_src_num);
rtas_st(rets, 0, RTAS_OUT_SUCCESS);
rtas_st(rets, 1, intr_src_num);
rtas_st(rets, 2, 1);/* 0 == level; 1 == edge */
}
static void rtas_ibm_set_eeh_option(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
uint32_t addr, option;
uint64_t buid;
int ret;
if ((nargs != 4) || (nret != 1)) {
goto param_error_exit;
}
buid = rtas_ldq(args, 1);
addr = rtas_ld(args, 0);
option = rtas_ld(args, 3);
sphb = spapr_pci_find_phb(spapr, buid);
if (!sphb) {
goto param_error_exit;
}
if (!spapr_phb_eeh_available(sphb)) {
goto param_error_exit;
}
ret = spapr_phb_vfio_eeh_set_option(sphb, addr, option);
rtas_st(rets, 0, ret);
return;
param_error_exit:
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
}
static void rtas_ibm_get_config_addr_info2(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
PCIDevice *pdev;
uint32_t addr, option;
uint64_t buid;
if ((nargs != 4) || (nret != 2)) {
goto param_error_exit;
}
buid = rtas_ldq(args, 1);
sphb = spapr_pci_find_phb(spapr, buid);
if (!sphb) {
goto param_error_exit;
}
if (!spapr_phb_eeh_available(sphb)) {
goto param_error_exit;
}
/*
* We always have PE address of form "00BB0001". "BB"
* represents the bus number of PE's primary bus.
*/
option = rtas_ld(args, 3);
switch (option) {
case RTAS_GET_PE_ADDR:
addr = rtas_ld(args, 0);
pdev = spapr_pci_find_dev(spapr, buid, addr);
if (!pdev) {
goto param_error_exit;
}
rtas_st(rets, 1, (pci_bus_num(pci_get_bus(pdev)) << 16) + 1);
break;
case RTAS_GET_PE_MODE:
rtas_st(rets, 1, RTAS_PE_MODE_SHARED);
break;
default:
goto param_error_exit;
}
rtas_st(rets, 0, RTAS_OUT_SUCCESS);
return;
param_error_exit:
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
}
static void rtas_ibm_read_slot_reset_state2(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
uint64_t buid;
int state, ret;
if ((nargs != 3) || (nret != 4 && nret != 5)) {
goto param_error_exit;
}
buid = rtas_ldq(args, 1);
sphb = spapr_pci_find_phb(spapr, buid);
if (!sphb) {
goto param_error_exit;
}
if (!spapr_phb_eeh_available(sphb)) {
goto param_error_exit;
}
ret = spapr_phb_vfio_eeh_get_state(sphb, &state);
rtas_st(rets, 0, ret);
if (ret != RTAS_OUT_SUCCESS) {
return;
}
rtas_st(rets, 1, state);
rtas_st(rets, 2, RTAS_EEH_SUPPORT);
rtas_st(rets, 3, RTAS_EEH_PE_UNAVAIL_INFO);
if (nret >= 5) {
rtas_st(rets, 4, RTAS_EEH_PE_RECOVER_INFO);
}
return;
param_error_exit:
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
}
static void rtas_ibm_set_slot_reset(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
uint32_t option;
uint64_t buid;
int ret;
if ((nargs != 4) || (nret != 1)) {
goto param_error_exit;
}
buid = rtas_ldq(args, 1);
option = rtas_ld(args, 3);
sphb = spapr_pci_find_phb(spapr, buid);
if (!sphb) {
goto param_error_exit;
}
if (!spapr_phb_eeh_available(sphb)) {
goto param_error_exit;
}
ret = spapr_phb_vfio_eeh_reset(sphb, option);
rtas_st(rets, 0, ret);
return;
param_error_exit:
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
}
static void rtas_ibm_configure_pe(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
uint64_t buid;
int ret;
if ((nargs != 3) || (nret != 1)) {
goto param_error_exit;
}
buid = rtas_ldq(args, 1);
sphb = spapr_pci_find_phb(spapr, buid);
if (!sphb) {
goto param_error_exit;
}
if (!spapr_phb_eeh_available(sphb)) {
goto param_error_exit;
}
ret = spapr_phb_vfio_eeh_configure(sphb);
rtas_st(rets, 0, ret);
return;
param_error_exit:
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
}
/* To support it later */
static void rtas_ibm_slot_error_detail(PowerPCCPU *cpu,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr,
uint32_t token, uint32_t nargs,
target_ulong args, uint32_t nret,
target_ulong rets)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
int option;
uint64_t buid;
if ((nargs != 8) || (nret != 1)) {
goto param_error_exit;
}
buid = rtas_ldq(args, 1);
sphb = spapr_pci_find_phb(spapr, buid);
if (!sphb) {
goto param_error_exit;
}
if (!spapr_phb_eeh_available(sphb)) {
goto param_error_exit;
}
option = rtas_ld(args, 7);
switch (option) {
case RTAS_SLOT_TEMP_ERR_LOG:
case RTAS_SLOT_PERM_ERR_LOG:
break;
default:
goto param_error_exit;
}
/* We don't have error log yet */
rtas_st(rets, 0, RTAS_OUT_NO_ERRORS_FOUND);
return;
param_error_exit:
rtas_st(rets, 0, RTAS_OUT_PARAM_ERROR);
}
static void pci_spapr_set_irq(void *opaque, int irq_num, int level)
{
/*
* Here we use the number returned by pci_swizzle_map_irq_fn to find a
* corresponding qemu_irq.
*/
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *phb = opaque;
SpaprMachineState *spapr = SPAPR_MACHINE(qdev_get_machine());
trace_spapr_pci_lsi_set(phb->dtbusname, irq_num, phb->lsi_table[irq_num].irq);
qemu_set_irq(spapr_qirq(spapr, phb->lsi_table[irq_num].irq), level);
}
static PCIINTxRoute spapr_route_intx_pin_to_irq(void *opaque, int pin)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(opaque);
PCIINTxRoute route;
route.mode = PCI_INTX_ENABLED;
route.irq = sphb->lsi_table[pin].irq;
return route;
}
static uint64_t spapr_msi_read(void *opaque, hwaddr addr, unsigned size)
{
qemu_log_mask(LOG_GUEST_ERROR, "%s: invalid access\n", __func__);
return 0;
}
/*
* MSI/MSIX memory region implementation.
* The handler handles both MSI and MSIX.
* The vector number is encoded in least bits in data.
*/
static void spapr_msi_write(void *opaque, hwaddr addr,
uint64_t data, unsigned size)
{
SpaprMachineState *spapr = opaque;
spapr-pci: rework MSI/MSIX On the sPAPR platform a guest allocates MSI/MSIX vectors via RTAS hypercalls which return global IRQ numbers to a guest so it only operates with those and never touches MSIMessage. Therefore MSIMessage handling is completely hidden in QEMU. Previously every sPAPR PCI host bridge implemented its own MSI window to catch msi_notify()/msix_notify() calls from QEMU devices (virtio-pci or vfio) and route them to the guest via qemu_pulse_irq(). MSIMessage used to be encoded as: .addr - address within the PHB MSI window; .data - the device index on PHB plus vector number. The MSI MR write function translated this MSIMessage to a global IRQ number and called qemu_pulse_irq(). However the total number of IRQs is not really big (at the moment it is 1024 IRQs starting from 4096) and even 16bit data field of MSIMessage seems to be enough to store an IRQ number there. This simplifies MSI handling in sPAPR PHB. Specifically, this does: 1. remove a MSI window from a PHB; 2. add a single memory region for all MSIs to sPAPREnvironment and spapr_pci_msi_init() to initialize it; 3. encode MSIMessage as: * .addr - a fixed address of SPAPR_PCI_MSI_WINDOW==0x40000000000ULL; * .data as an IRQ number. 4. change IRQ allocator to align first IRQ number in a block for MSI. MSI uses lower bits to specify the vector number so the first IRQ has to be aligned. MSIX does not need any special allocator though. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Reviewed-by: Anthony Liguori <aliguori@us.ibm.com> Acked-by: Michael S. Tsirkin <mst@redhat.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2013-07-12 09:38:24 +02:00
uint32_t irq = data;
trace_spapr_pci_msi_write(addr, data, irq);
qemu_irq_pulse(spapr_qirq(spapr, irq));
}
static const MemoryRegionOps spapr_msi_ops = {
/*
* .read result is undefined by PCI spec.
* define .read method to avoid assert failure in memory_region_init_io
*/
.read = spapr_msi_read,
.write = spapr_msi_write,
.endianness = DEVICE_LITTLE_ENDIAN
};
/*
* PHB PCI device
*/
static AddressSpace *spapr_pci_dma_iommu(PCIBus *bus, void *opaque, int devfn)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *phb = opaque;
return &phb->iommu_as;
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static char *spapr_phb_vfio_get_loc_code(SpaprPhbState *sphb, PCIDevice *pdev)
{
g_autofree char *path = NULL;
g_autofree char *host = NULL;
g_autofree char *devspec = NULL;
char *buf = NULL;
/* Get the PCI VFIO host id */
host = object_property_get_str(OBJECT(pdev), "host", NULL);
if (!host) {
return NULL;
}
/* Construct the path of the file that will give us the DT location */
path = g_strdup_printf("/sys/bus/pci/devices/%s/devspec", host);
if (!g_file_get_contents(path, &devspec, NULL, NULL)) {
return NULL;
}
/* Construct and read from host device tree the loc-code */
path = g_strdup_printf("/proc/device-tree%s/ibm,loc-code", devspec);
if (!g_file_get_contents(path, &buf, NULL, NULL)) {
return NULL;
}
return buf;
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static char *spapr_phb_get_loc_code(SpaprPhbState *sphb, PCIDevice *pdev)
{
char *buf;
const char *devtype = "qemu";
uint32_t busnr = pci_bus_num(PCI_BUS(qdev_get_parent_bus(DEVICE(pdev))));
if (object_dynamic_cast(OBJECT(pdev), "vfio-pci")) {
buf = spapr_phb_vfio_get_loc_code(sphb, pdev);
if (buf) {
return buf;
}
devtype = "vfio";
}
/*
* For emulated devices and VFIO-failure case, make up
* the loc-code.
*/
buf = g_strdup_printf("%s_%s:%04x:%02x:%02x.%x",
devtype, pdev->name, sphb->index, busnr,
PCI_SLOT(pdev->devfn), PCI_FUNC(pdev->devfn));
return buf;
}
/* Macros to operate with address in OF binding to PCI */
#define b_x(x, p, l) (((x) & ((1<<(l))-1)) << (p))
#define b_n(x) b_x((x), 31, 1) /* 0 if relocatable */
#define b_p(x) b_x((x), 30, 1) /* 1 if prefetchable */
#define b_t(x) b_x((x), 29, 1) /* 1 if the address is aliased */
#define b_ss(x) b_x((x), 24, 2) /* the space code */
#define b_bbbbbbbb(x) b_x((x), 16, 8) /* bus number */
#define b_ddddd(x) b_x((x), 11, 5) /* device number */
#define b_fff(x) b_x((x), 8, 3) /* function number */
#define b_rrrrrrrr(x) b_x((x), 0, 8) /* register number */
/* for 'reg' OF properties */
#define RESOURCE_CELLS_SIZE 2
#define RESOURCE_CELLS_ADDRESS 3
typedef struct ResourceFields {
uint32_t phys_hi;
uint32_t phys_mid;
uint32_t phys_lo;
uint32_t size_hi;
uint32_t size_lo;
} QEMU_PACKED ResourceFields;
typedef struct ResourceProps {
ResourceFields reg[8];
uint32_t reg_len;
} ResourceProps;
/* fill in the 'reg' OF properties for
* a PCI device. 'reg' describes resource requirements for a
* device's IO/MEM regions.
*
* the property is an array of ('phys-addr', 'size') pairs describing
* the addressable regions of the PCI device, where 'phys-addr' is a
* RESOURCE_CELLS_ADDRESS-tuple of 32-bit integers corresponding to
* (phys.hi, phys.mid, phys.lo), and 'size' is a
* RESOURCE_CELLS_SIZE-tuple corresponding to (size.hi, size.lo).
*
* phys.hi = 0xYYXXXXZZ, where:
* 0xYY = npt000ss
* ||| |
* ||| +-- space code
* ||| |
* ||| + 00 if configuration space
* ||| + 01 if IO region,
* ||| + 10 if 32-bit MEM region
* ||| + 11 if 64-bit MEM region
* |||
* ||+------ for non-relocatable IO: 1 if aliased
* || for relocatable IO: 1 if below 64KB
* || for MEM: 1 if below 1MB
* |+------- 1 if region is prefetchable
* +-------- 1 if region is non-relocatable
* 0xXXXX = bbbbbbbb dddddfff, encoding bus, slot, and function
* bits respectively
* 0xZZ = rrrrrrrr, the register number of the BAR corresponding
* to the region
*
* phys.mid and phys.lo correspond respectively to the hi/lo portions
* of the actual address of the region.
*
* note also that addresses defined in this property are, at least
* for PAPR guests, relative to the PHBs IO/MEM windows, and
* correspond directly to the addresses in the BARs.
*
* in accordance with PCI Bus Binding to Open Firmware,
* IEEE Std 1275-1994, section 4.1.1, as implemented by PAPR+ v2.7,
* Appendix C.
*/
static void populate_resource_props(PCIDevice *d, ResourceProps *rp)
{
int bus_num = pci_bus_num(PCI_BUS(qdev_get_parent_bus(DEVICE(d))));
uint32_t dev_id = (b_bbbbbbbb(bus_num) |
b_ddddd(PCI_SLOT(d->devfn)) |
b_fff(PCI_FUNC(d->devfn)));
ResourceFields *reg;
int i, reg_idx = 0;
/* config space region */
reg = &rp->reg[reg_idx++];
reg->phys_hi = cpu_to_be32(dev_id);
reg->phys_mid = 0;
reg->phys_lo = 0;
reg->size_hi = 0;
reg->size_lo = 0;
for (i = 0; i < PCI_NUM_REGIONS; i++) {
if (!d->io_regions[i].size) {
continue;
}
reg = &rp->reg[reg_idx++];
reg->phys_hi = cpu_to_be32(dev_id | b_rrrrrrrr(pci_bar(d, i)));
if (d->io_regions[i].type & PCI_BASE_ADDRESS_SPACE_IO) {
reg->phys_hi |= cpu_to_be32(b_ss(1));
} else if (d->io_regions[i].type & PCI_BASE_ADDRESS_MEM_TYPE_64) {
reg->phys_hi |= cpu_to_be32(b_ss(3));
} else {
reg->phys_hi |= cpu_to_be32(b_ss(2));
}
reg->phys_mid = 0;
reg->phys_lo = 0;
reg->size_hi = cpu_to_be32(d->io_regions[i].size >> 32);
reg->size_lo = cpu_to_be32(d->io_regions[i].size);
}
rp->reg_len = reg_idx * sizeof(ResourceFields);
}
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
typedef struct PCIClass PCIClass;
typedef struct PCISubClass PCISubClass;
typedef struct PCIIFace PCIIFace;
struct PCIIFace {
int iface;
const char *name;
};
struct PCISubClass {
int subclass;
const char *name;
const PCIIFace *iface;
};
struct PCIClass {
const char *name;
const PCISubClass *subc;
};
static const PCISubClass undef_subclass[] = {
{ PCI_CLASS_NOT_DEFINED_VGA, "display", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass mass_subclass[] = {
{ PCI_CLASS_STORAGE_SCSI, "scsi", NULL },
{ PCI_CLASS_STORAGE_IDE, "ide", NULL },
{ PCI_CLASS_STORAGE_FLOPPY, "fdc", NULL },
{ PCI_CLASS_STORAGE_IPI, "ipi", NULL },
{ PCI_CLASS_STORAGE_RAID, "raid", NULL },
{ PCI_CLASS_STORAGE_ATA, "ata", NULL },
{ PCI_CLASS_STORAGE_SATA, "sata", NULL },
{ PCI_CLASS_STORAGE_SAS, "sas", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass net_subclass[] = {
{ PCI_CLASS_NETWORK_ETHERNET, "ethernet", NULL },
{ PCI_CLASS_NETWORK_TOKEN_RING, "token-ring", NULL },
{ PCI_CLASS_NETWORK_FDDI, "fddi", NULL },
{ PCI_CLASS_NETWORK_ATM, "atm", NULL },
{ PCI_CLASS_NETWORK_ISDN, "isdn", NULL },
{ PCI_CLASS_NETWORK_WORLDFIP, "worldfip", NULL },
{ PCI_CLASS_NETWORK_PICMG214, "picmg", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass displ_subclass[] = {
{ PCI_CLASS_DISPLAY_VGA, "vga", NULL },
{ PCI_CLASS_DISPLAY_XGA, "xga", NULL },
{ PCI_CLASS_DISPLAY_3D, "3d-controller", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass media_subclass[] = {
{ PCI_CLASS_MULTIMEDIA_VIDEO, "video", NULL },
{ PCI_CLASS_MULTIMEDIA_AUDIO, "sound", NULL },
{ PCI_CLASS_MULTIMEDIA_PHONE, "telephony", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass mem_subclass[] = {
{ PCI_CLASS_MEMORY_RAM, "memory", NULL },
{ PCI_CLASS_MEMORY_FLASH, "flash", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass bridg_subclass[] = {
{ PCI_CLASS_BRIDGE_HOST, "host", NULL },
{ PCI_CLASS_BRIDGE_ISA, "isa", NULL },
{ PCI_CLASS_BRIDGE_EISA, "eisa", NULL },
{ PCI_CLASS_BRIDGE_MC, "mca", NULL },
{ PCI_CLASS_BRIDGE_PCI, "pci", NULL },
{ PCI_CLASS_BRIDGE_PCMCIA, "pcmcia", NULL },
{ PCI_CLASS_BRIDGE_NUBUS, "nubus", NULL },
{ PCI_CLASS_BRIDGE_CARDBUS, "cardbus", NULL },
{ PCI_CLASS_BRIDGE_RACEWAY, "raceway", NULL },
{ PCI_CLASS_BRIDGE_PCI_SEMITP, "semi-transparent-pci", NULL },
{ PCI_CLASS_BRIDGE_IB_PCI, "infiniband", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass comm_subclass[] = {
{ PCI_CLASS_COMMUNICATION_SERIAL, "serial", NULL },
{ PCI_CLASS_COMMUNICATION_PARALLEL, "parallel", NULL },
{ PCI_CLASS_COMMUNICATION_MULTISERIAL, "multiport-serial", NULL },
{ PCI_CLASS_COMMUNICATION_MODEM, "modem", NULL },
{ PCI_CLASS_COMMUNICATION_GPIB, "gpib", NULL },
{ PCI_CLASS_COMMUNICATION_SC, "smart-card", NULL },
{ 0xFF, NULL, NULL, },
};
static const PCIIFace pic_iface[] = {
{ PCI_CLASS_SYSTEM_PIC_IOAPIC, "io-apic" },
{ PCI_CLASS_SYSTEM_PIC_IOXAPIC, "io-xapic" },
{ 0xFF, NULL },
};
static const PCISubClass sys_subclass[] = {
{ PCI_CLASS_SYSTEM_PIC, "interrupt-controller", pic_iface },
{ PCI_CLASS_SYSTEM_DMA, "dma-controller", NULL },
{ PCI_CLASS_SYSTEM_TIMER, "timer", NULL },
{ PCI_CLASS_SYSTEM_RTC, "rtc", NULL },
{ PCI_CLASS_SYSTEM_PCI_HOTPLUG, "hot-plug-controller", NULL },
{ PCI_CLASS_SYSTEM_SDHCI, "sd-host-controller", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass inp_subclass[] = {
{ PCI_CLASS_INPUT_KEYBOARD, "keyboard", NULL },
{ PCI_CLASS_INPUT_PEN, "pen", NULL },
{ PCI_CLASS_INPUT_MOUSE, "mouse", NULL },
{ PCI_CLASS_INPUT_SCANNER, "scanner", NULL },
{ PCI_CLASS_INPUT_GAMEPORT, "gameport", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass dock_subclass[] = {
{ PCI_CLASS_DOCKING_GENERIC, "dock", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass cpu_subclass[] = {
{ PCI_CLASS_PROCESSOR_PENTIUM, "pentium", NULL },
{ PCI_CLASS_PROCESSOR_POWERPC, "powerpc", NULL },
{ PCI_CLASS_PROCESSOR_MIPS, "mips", NULL },
{ PCI_CLASS_PROCESSOR_CO, "co-processor", NULL },
{ 0xFF, NULL, NULL },
};
static const PCIIFace usb_iface[] = {
{ PCI_CLASS_SERIAL_USB_UHCI, "usb-uhci" },
{ PCI_CLASS_SERIAL_USB_OHCI, "usb-ohci", },
{ PCI_CLASS_SERIAL_USB_EHCI, "usb-ehci" },
{ PCI_CLASS_SERIAL_USB_XHCI, "usb-xhci" },
{ PCI_CLASS_SERIAL_USB_UNKNOWN, "usb-unknown" },
{ PCI_CLASS_SERIAL_USB_DEVICE, "usb-device" },
{ 0xFF, NULL },
};
static const PCISubClass ser_subclass[] = {
{ PCI_CLASS_SERIAL_FIREWIRE, "firewire", NULL },
{ PCI_CLASS_SERIAL_ACCESS, "access-bus", NULL },
{ PCI_CLASS_SERIAL_SSA, "ssa", NULL },
{ PCI_CLASS_SERIAL_USB, "usb", usb_iface },
{ PCI_CLASS_SERIAL_FIBER, "fibre-channel", NULL },
{ PCI_CLASS_SERIAL_SMBUS, "smb", NULL },
{ PCI_CLASS_SERIAL_IB, "infiniband", NULL },
{ PCI_CLASS_SERIAL_IPMI, "ipmi", NULL },
{ PCI_CLASS_SERIAL_SERCOS, "sercos", NULL },
{ PCI_CLASS_SERIAL_CANBUS, "canbus", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass wrl_subclass[] = {
{ PCI_CLASS_WIRELESS_IRDA, "irda", NULL },
{ PCI_CLASS_WIRELESS_CIR, "consumer-ir", NULL },
{ PCI_CLASS_WIRELESS_RF_CONTROLLER, "rf-controller", NULL },
{ PCI_CLASS_WIRELESS_BLUETOOTH, "bluetooth", NULL },
{ PCI_CLASS_WIRELESS_BROADBAND, "broadband", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass sat_subclass[] = {
{ PCI_CLASS_SATELLITE_TV, "satellite-tv", NULL },
{ PCI_CLASS_SATELLITE_AUDIO, "satellite-audio", NULL },
{ PCI_CLASS_SATELLITE_VOICE, "satellite-voice", NULL },
{ PCI_CLASS_SATELLITE_DATA, "satellite-data", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass crypt_subclass[] = {
{ PCI_CLASS_CRYPT_NETWORK, "network-encryption", NULL },
{ PCI_CLASS_CRYPT_ENTERTAINMENT,
"entertainment-encryption", NULL },
{ 0xFF, NULL, NULL },
};
static const PCISubClass spc_subclass[] = {
{ PCI_CLASS_SP_DPIO, "dpio", NULL },
{ PCI_CLASS_SP_PERF, "counter", NULL },
{ PCI_CLASS_SP_SYNCH, "measurement", NULL },
{ PCI_CLASS_SP_MANAGEMENT, "management-card", NULL },
{ 0xFF, NULL, NULL },
};
static const PCIClass pci_classes[] = {
{ "legacy-device", undef_subclass },
{ "mass-storage", mass_subclass },
{ "network", net_subclass },
{ "display", displ_subclass, },
{ "multimedia-device", media_subclass },
{ "memory-controller", mem_subclass },
{ "unknown-bridge", bridg_subclass },
{ "communication-controller", comm_subclass},
{ "system-peripheral", sys_subclass },
{ "input-controller", inp_subclass },
{ "docking-station", dock_subclass },
{ "cpu", cpu_subclass },
{ "serial-bus", ser_subclass },
{ "wireless-controller", wrl_subclass },
{ "intelligent-io", NULL },
{ "satellite-device", sat_subclass },
{ "encryption", crypt_subclass },
{ "data-processing-controller", spc_subclass },
};
static const char *dt_name_from_class(uint8_t class, uint8_t subclass,
uint8_t iface)
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
{
const PCIClass *pclass;
const PCISubClass *psubclass;
const PCIIFace *piface;
const char *name;
if (class >= ARRAY_SIZE(pci_classes)) {
return "pci";
}
pclass = pci_classes + class;
name = pclass->name;
if (pclass->subc == NULL) {
return name;
}
psubclass = pclass->subc;
while ((psubclass->subclass & 0xff) != 0xff) {
if ((psubclass->subclass & 0xff) == subclass) {
name = psubclass->name;
break;
}
psubclass++;
}
piface = psubclass->iface;
if (piface == NULL) {
return name;
}
while ((piface->iface & 0xff) != 0xff) {
if ((piface->iface & 0xff) == iface) {
name = piface->name;
break;
}
piface++;
}
return name;
}
/*
* DRC helper functions
*/
static uint32_t drc_id_from_devfn(SpaprPhbState *phb,
uint8_t chassis, int32_t devfn)
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
{
return (phb->index << 16) | (chassis << 8) | devfn;
}
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
static SpaprDrc *drc_from_devfn(SpaprPhbState *phb,
uint8_t chassis, int32_t devfn)
{
return spapr_drc_by_id(TYPE_SPAPR_DRC_PCI,
drc_id_from_devfn(phb, chassis, devfn));
}
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
static uint8_t chassis_from_bus(PCIBus *bus)
{
if (pci_bus_is_root(bus)) {
return 0;
} else {
PCIDevice *bridge = pci_bridge_get_device(bus);
return object_property_get_uint(OBJECT(bridge), "chassis_nr",
&error_abort);
}
}
static SpaprDrc *drc_from_dev(SpaprPhbState *phb, PCIDevice *dev)
{
uint8_t chassis = chassis_from_bus(pci_get_bus(dev));
return drc_from_devfn(phb, chassis, dev->devfn);
}
static void add_drcs(SpaprPhbState *phb, PCIBus *bus)
{
Object *owner;
int i;
uint8_t chassis;
if (!phb->dr_enabled) {
return;
}
chassis = chassis_from_bus(bus);
if (pci_bus_is_root(bus)) {
owner = OBJECT(phb);
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
} else {
owner = OBJECT(pci_bridge_get_device(bus));
}
for (i = 0; i < PCI_SLOT_MAX * PCI_FUNC_MAX; i++) {
spapr_dr_connector_new(owner, TYPE_SPAPR_DRC_PCI,
drc_id_from_devfn(phb, chassis, i));
}
}
static void remove_drcs(SpaprPhbState *phb, PCIBus *bus)
{
int i;
uint8_t chassis;
if (!phb->dr_enabled) {
return;
}
chassis = chassis_from_bus(bus);
for (i = PCI_SLOT_MAX * PCI_FUNC_MAX - 1; i >= 0; i--) {
SpaprDrc *drc = drc_from_devfn(phb, chassis, i);
if (drc) {
object_unparent(OBJECT(drc));
}
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
}
}
typedef struct PciWalkFdt {
void *fdt;
int offset;
SpaprPhbState *sphb;
int err;
} PciWalkFdt;
static int spapr_dt_pci_device(SpaprPhbState *sphb, PCIDevice *dev,
void *fdt, int parent_offset);
static void spapr_dt_pci_device_cb(PCIBus *bus, PCIDevice *pdev,
void *opaque)
{
PciWalkFdt *p = opaque;
int err;
if (p->err) {
/* Something's already broken, don't keep going */
return;
}
err = spapr_dt_pci_device(p->sphb, pdev, p->fdt, p->offset);
if (err < 0) {
p->err = err;
}
}
/* Augment PCI device node with bridge specific information */
static int spapr_dt_pci_bus(SpaprPhbState *sphb, PCIBus *bus,
void *fdt, int offset)
{
Object *owner;
PciWalkFdt cbinfo = {
.fdt = fdt,
.offset = offset,
.sphb = sphb,
.err = 0,
};
int ret;
_FDT(fdt_setprop_cell(fdt, offset, "#address-cells",
RESOURCE_CELLS_ADDRESS));
_FDT(fdt_setprop_cell(fdt, offset, "#size-cells",
RESOURCE_CELLS_SIZE));
assert(bus);
pci_for_each_device_under_bus_reverse(bus, spapr_dt_pci_device_cb, &cbinfo);
if (cbinfo.err) {
return cbinfo.err;
}
if (pci_bus_is_root(bus)) {
owner = OBJECT(sphb);
} else {
owner = OBJECT(pci_bridge_get_device(bus));
}
ret = spapr_dt_drc(fdt, offset, owner,
SPAPR_DR_CONNECTOR_TYPE_PCI);
if (ret) {
return ret;
}
return offset;
}
spapr: Adjust firmware path of PCI devices It is currently not possible to perform a strict boot from USB storage: $ qemu-system-ppc64 -accel kvm -nodefaults -nographic -serial stdio \ -boot strict=on \ -device qemu-xhci \ -device usb-storage,drive=disk,bootindex=0 \ -blockdev driver=file,node-name=disk,filename=fedora-ppc64le.qcow2 SLOF ********************************************************************** QEMU Starting Build Date = Jul 17 2020 11:15:24 FW Version = git-e18ddad8516ff2cf Press "s" to enter Open Firmware. Populating /vdevice methods Populating /vdevice/vty@71000000 Populating /vdevice/nvram@71000001 Populating /pci@800000020000000 00 0000 (D) : 1b36 000d serial bus [ usb-xhci ] No NVRAM common partition, re-initializing... Scanning USB XHCI: Initializing USB Storage SCSI: Looking for devices 101000000000000 DISK : "QEMU QEMU HARDDISK 2.5+" Using default console: /vdevice/vty@71000000 Welcome to Open Firmware Copyright (c) 2004, 2017 IBM Corporation All rights reserved. This program and the accompanying materials are made available under the terms of the BSD License available at http://www.opensource.org/licenses/bsd-license.php Trying to load: from: /pci@800000020000000/usb@0/storage@1/disk@101000000000000 ... E3405: No such device E3407: Load failed Type 'boot' and press return to continue booting the system. Type 'reset-all' and press return to reboot the system. Ready! 0 > The device tree handed over by QEMU to SLOF indeed contains: qemu,boot-list = "/pci@800000020000000/usb@0/storage@1/disk@101000000000000 HALT"; but the device node is named usb-xhci@0, not usb@0. This happens because the firmware names of PCI devices returned by get_boot_devices_list() come from pcibus_get_fw_dev_path(), while the sPAPR PHB code uses a different naming scheme for device nodes. This inconsistency has always been there but it was hidden for a long time because SLOF used to rename USB device nodes, until this commit, merged in QEMU 4.2.0 : commit 85164ad4ed9960cac842fa4cc067c6b6699b0994 Author: Alexey Kardashevskiy <aik@ozlabs.ru> Date: Wed Sep 11 16:24:32 2019 +1000 pseries: Update SLOF firmware image This fixes USB host bus adapter name in the device tree to match QEMU's one. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Fortunately, sPAPR implements the firmware path provider interface. This provides a way to override the default firmware paths. Just factor out the sPAPR PHB naming logic from spapr_dt_pci_device() to a helper, and use it in the sPAPR firmware path provider hook. Fixes: 85164ad4ed99 ("pseries: Update SLOF firmware image") Signed-off-by: Greg Kurz <groug@kaod.org> Message-Id: <20210122170157.246374-1-groug@kaod.org> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2021-01-22 18:01:57 +01:00
char *spapr_pci_fw_dev_name(PCIDevice *dev)
{
const gchar *basename;
int slot = PCI_SLOT(dev->devfn);
int func = PCI_FUNC(dev->devfn);
uint32_t ccode = pci_default_read_config(dev, PCI_CLASS_PROG, 3);
basename = dt_name_from_class((ccode >> 16) & 0xff, (ccode >> 8) & 0xff,
ccode & 0xff);
if (func != 0) {
return g_strdup_printf("%s@%x,%x", basename, slot, func);
} else {
return g_strdup_printf("%s@%x", basename, slot);
}
}
/* create OF node for pci device and required OF DT properties */
static int spapr_dt_pci_device(SpaprPhbState *sphb, PCIDevice *dev,
void *fdt, int parent_offset)
{
int offset;
spapr: Adjust firmware path of PCI devices It is currently not possible to perform a strict boot from USB storage: $ qemu-system-ppc64 -accel kvm -nodefaults -nographic -serial stdio \ -boot strict=on \ -device qemu-xhci \ -device usb-storage,drive=disk,bootindex=0 \ -blockdev driver=file,node-name=disk,filename=fedora-ppc64le.qcow2 SLOF ********************************************************************** QEMU Starting Build Date = Jul 17 2020 11:15:24 FW Version = git-e18ddad8516ff2cf Press "s" to enter Open Firmware. Populating /vdevice methods Populating /vdevice/vty@71000000 Populating /vdevice/nvram@71000001 Populating /pci@800000020000000 00 0000 (D) : 1b36 000d serial bus [ usb-xhci ] No NVRAM common partition, re-initializing... Scanning USB XHCI: Initializing USB Storage SCSI: Looking for devices 101000000000000 DISK : "QEMU QEMU HARDDISK 2.5+" Using default console: /vdevice/vty@71000000 Welcome to Open Firmware Copyright (c) 2004, 2017 IBM Corporation All rights reserved. This program and the accompanying materials are made available under the terms of the BSD License available at http://www.opensource.org/licenses/bsd-license.php Trying to load: from: /pci@800000020000000/usb@0/storage@1/disk@101000000000000 ... E3405: No such device E3407: Load failed Type 'boot' and press return to continue booting the system. Type 'reset-all' and press return to reboot the system. Ready! 0 > The device tree handed over by QEMU to SLOF indeed contains: qemu,boot-list = "/pci@800000020000000/usb@0/storage@1/disk@101000000000000 HALT"; but the device node is named usb-xhci@0, not usb@0. This happens because the firmware names of PCI devices returned by get_boot_devices_list() come from pcibus_get_fw_dev_path(), while the sPAPR PHB code uses a different naming scheme for device nodes. This inconsistency has always been there but it was hidden for a long time because SLOF used to rename USB device nodes, until this commit, merged in QEMU 4.2.0 : commit 85164ad4ed9960cac842fa4cc067c6b6699b0994 Author: Alexey Kardashevskiy <aik@ozlabs.ru> Date: Wed Sep 11 16:24:32 2019 +1000 pseries: Update SLOF firmware image This fixes USB host bus adapter name in the device tree to match QEMU's one. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Fortunately, sPAPR implements the firmware path provider interface. This provides a way to override the default firmware paths. Just factor out the sPAPR PHB naming logic from spapr_dt_pci_device() to a helper, and use it in the sPAPR firmware path provider hook. Fixes: 85164ad4ed99 ("pseries: Update SLOF firmware image") Signed-off-by: Greg Kurz <groug@kaod.org> Message-Id: <20210122170157.246374-1-groug@kaod.org> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2021-01-22 18:01:57 +01:00
g_autofree gchar *nodename = spapr_pci_fw_dev_name(dev);
PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(dev);
ResourceProps rp;
SpaprDrc *drc = drc_from_dev(sphb, dev);
uint32_t vendor_id = pci_default_read_config(dev, PCI_VENDOR_ID, 2);
uint32_t device_id = pci_default_read_config(dev, PCI_DEVICE_ID, 2);
uint32_t revision_id = pci_default_read_config(dev, PCI_REVISION_ID, 1);
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
uint32_t ccode = pci_default_read_config(dev, PCI_CLASS_PROG, 3);
uint32_t irq_pin = pci_default_read_config(dev, PCI_INTERRUPT_PIN, 1);
uint32_t subsystem_id = pci_default_read_config(dev, PCI_SUBSYSTEM_ID, 2);
uint32_t subsystem_vendor_id =
pci_default_read_config(dev, PCI_SUBSYSTEM_VENDOR_ID, 2);
uint32_t cache_line_size =
pci_default_read_config(dev, PCI_CACHE_LINE_SIZE, 1);
uint32_t pci_status = pci_default_read_config(dev, PCI_STATUS, 2);
gchar *loc_code;
_FDT(offset = fdt_add_subnode(fdt, parent_offset, nodename));
/* in accordance with PAPR+ v2.7 13.6.3, Table 181 */
_FDT(fdt_setprop_cell(fdt, offset, "vendor-id", vendor_id));
_FDT(fdt_setprop_cell(fdt, offset, "device-id", device_id));
_FDT(fdt_setprop_cell(fdt, offset, "revision-id", revision_id));
spapr: generate DT node names When DT node names for PCI devices are generated by SLOF, they are generated according to the type of the device (for instance, ethernet for virtio-net-pci device). Node name for hotplugged devices is generated by QEMU. This patch adds the mechanic to QEMU to create the node name according to the device type too. The data structure has been roughly copied from OpenBIOS/OpenHackware, node names from SLOF. Example: Hotplugging some PCI cards with QEMU monitor: device_add virtio-tablet-pci device_add virtio-serial-pci device_add virtio-mouse-pci device_add virtio-scsi-pci device_add virtio-gpu-pci device_add ne2k_pci device_add nec-usb-xhci device_add intel-hda What we can see in linux device tree: for dir in /proc/device-tree/pci@800000020000000/*@*/; do echo $dir cat $dir/name echo done WITHOUT this patch: /proc/device-tree/pci@800000020000000/pci@0/ pci /proc/device-tree/pci@800000020000000/pci@1/ pci /proc/device-tree/pci@800000020000000/pci@2/ pci /proc/device-tree/pci@800000020000000/pci@3/ pci /proc/device-tree/pci@800000020000000/pci@4/ pci /proc/device-tree/pci@800000020000000/pci@5/ pci /proc/device-tree/pci@800000020000000/pci@6/ pci /proc/device-tree/pci@800000020000000/pci@7/ pci WITH this patch: /proc/device-tree/pci@800000020000000/communication-controller@1/ communication-controller /proc/device-tree/pci@800000020000000/display@4/ display /proc/device-tree/pci@800000020000000/ethernet@5/ ethernet /proc/device-tree/pci@800000020000000/input-controller@0/ input-controller /proc/device-tree/pci@800000020000000/mouse@2/ mouse /proc/device-tree/pci@800000020000000/multimedia-device@7/ multimedia-device /proc/device-tree/pci@800000020000000/scsi@3/ scsi /proc/device-tree/pci@800000020000000/usb-xhci@6/ usb-xhci Signed-off-by: Laurent Vivier <lvivier@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2017-02-17 14:31:34 +01:00
_FDT(fdt_setprop_cell(fdt, offset, "class-code", ccode));
if (irq_pin) {
_FDT(fdt_setprop_cell(fdt, offset, "interrupts", irq_pin));
}
if (subsystem_id) {
_FDT(fdt_setprop_cell(fdt, offset, "subsystem-id", subsystem_id));
}
if (subsystem_vendor_id) {
_FDT(fdt_setprop_cell(fdt, offset, "subsystem-vendor-id",
subsystem_vendor_id));
}
_FDT(fdt_setprop_cell(fdt, offset, "cache-line-size", cache_line_size));
/* the following fdt cells are masked off the pci status register */
_FDT(fdt_setprop_cell(fdt, offset, "devsel-speed",
PCI_STATUS_DEVSEL_MASK & pci_status));
if (pci_status & PCI_STATUS_FAST_BACK) {
_FDT(fdt_setprop(fdt, offset, "fast-back-to-back", NULL, 0));
}
if (pci_status & PCI_STATUS_66MHZ) {
_FDT(fdt_setprop(fdt, offset, "66mhz-capable", NULL, 0));
}
if (pci_status & PCI_STATUS_UDF) {
_FDT(fdt_setprop(fdt, offset, "udf-supported", NULL, 0));
}
loc_code = spapr_phb_get_loc_code(sphb, dev);
_FDT(fdt_setprop_string(fdt, offset, "ibm,loc-code", loc_code));
g_free(loc_code);
if (drc) {
_FDT(fdt_setprop_cell(fdt, offset, "ibm,my-drc-index",
spapr_drc_index(drc)));
}
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
if (msi_present(dev)) {
uint32_t max_msi = msi_nr_vectors_allocated(dev);
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
if (max_msi) {
_FDT(fdt_setprop_cell(fdt, offset, "ibm,req#msi", max_msi));
}
}
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
if (msix_present(dev)) {
uint32_t max_msix = dev->msix_entries_nr;
spapr_pci: fix MSI/MSIX selection In various place we don't correctly check if the device supports MSI or MSI-X. This can cause devices to be advertised with MSI support, even if they only support MSI-X (like virtio-pci-* devices for example): ethernet@0 { ibm,req#msi = <0x1>; <--- wrong! . ibm,loc-code = "qemu_virtio-net-pci:0000:00:00.0"; . ibm,req#msi-x = <0x3>; }; Worse, this can also cause the "ibm,change-msi" RTAS call to corrupt the PCI status and cause migration to fail: qemu-system-ppc64: get_pci_config_device: Bad config data: i=0x6 read: 0 device: 10 cmask: 10 wmask: 0 w1cmask:0 ^^ PCI_STATUS_CAP_LIST bit which is assumed to be constant This patch changes spapr_populate_pci_child_dt() to properly check for MSI support using msi_present(): this ensures that PCIDevice::msi_cap was set by msi_init() and that msi_nr_vectors_allocated() will look at the right place in the config space. Checking PCIDevice::msix_entries_nr is enough for MSI-X but let's add a call to msix_present() there as well for consistency. It also changes rtas_ibm_change_msi() to select the appropriate MSI type in Function 1 instead of always selecting plain MSI. This new behaviour is compliant with LoPAPR 1.1, as described in "Table 71. ibm,change-msi Argument Call Buffer": Function 1: If Number Outputs is equal to 3, request to set to a new number of MSIs (including set to 0). If the “ibm,change-msix-capable” property exists and Number Outputs is equal to 4, request is to set to a new number of MSI or MSI-X (platform choice) interrupts (including set to 0). Since MSI is the the platform default (LoPAPR 6.2.3 MSI Option), let's check for MSI support first. And finally, it checks the input parameters are valid, as described in LoPAPR 1.1 "R1–7.3.10.5.1–3": For the MSI option: The platform must return a Status of -3 (Parameter error) from ibm,change-msi, with no change in interrupt assignments if the PCI configuration address does not support MSI and Function 3 was requested (that is, the “ibm,req#msi” property must exist for the PCI configuration address in order to use Function 3), or does not support MSI-X and Function 4 is requested (that is, the “ibm,req#msi-x” property must exist for the PCI configuration address in order to use Function 4), or if neither MSIs nor MSI-Xs are supported and Function 1 is requested. This ensures that the ret_intr_type variable contains a valid MSI type for this device, and that spapr_msi_setmsg() won't corrupt the PCI status. Signed-off-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2018-01-26 23:25:24 +01:00
if (max_msix) {
_FDT(fdt_setprop_cell(fdt, offset, "ibm,req#msi-x", max_msix));
}
}
populate_resource_props(dev, &rp);
_FDT(fdt_setprop(fdt, offset, "reg", (uint8_t *)rp.reg, rp.reg_len));
if (sphb->pcie_ecs && pci_is_express(dev)) {
_FDT(fdt_setprop_cell(fdt, offset, "ibm,pci-config-space-type", 0x1));
}
spapr: Support NVIDIA V100 GPU with NVLink2 NVIDIA V100 GPUs have on-board RAM which is mapped into the host memory space and accessible as normal RAM via an NVLink bus. The VFIO-PCI driver implements special regions for such GPUs and emulates an NVLink bridge. NVLink2-enabled POWER9 CPUs also provide address translation services which includes an ATS shootdown (ATSD) register exported via the NVLink bridge device. This adds a quirk to VFIO to map the GPU memory and create an MR; the new MR is stored in a PCI device as a QOM link. The sPAPR PCI uses this to get the MR and map it to the system address space. Another quirk does the same for ATSD. This adds additional steps to sPAPR PHB setup: 1. Search for specific GPUs and NPUs, collect findings in sPAPRPHBState::nvgpus, manage system address space mappings; 2. Add device-specific properties such as "ibm,npu", "ibm,gpu", "memory-block", "link-speed" to advertise the NVLink2 function to the guest; 3. Add "mmio-atsd" to vPHB to advertise the ATSD capability; 4. Add new memory blocks (with extra "linux,memory-usable" to prevent the guest OS from accessing the new memory until it is onlined) and npuphb# nodes representing an NPU unit for every vPHB as the GPU driver uses it for link discovery. This allocates space for GPU RAM and ATSD like we do for MMIOs by adding 2 new parameters to the phb_placement() hook. Older machine types set these to zero. This puts new memory nodes in a separate NUMA node to as the GPU RAM needs to be configured equally distant from any other node in the system. Unlike the host setup which assigns numa ids from 255 downwards, this adds new NUMA nodes after the user configures nodes or from 1 if none were configured. This adds requirement similar to EEH - one IOMMU group per vPHB. The reason for this is that ATSD registers belong to a physical NPU so they cannot invalidate translations on GPUs attached to another NPU. It is guaranteed by the host platform as it does not mix NVLink bridges or GPUs from different NPU in the same IOMMU group. If more than one IOMMU group is detected on a vPHB, this disables ATSD support for that vPHB and prints a warning. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [aw: for vfio portions] Acked-by: Alex Williamson <alex.williamson@redhat.com> Message-Id: <20190312082103.130561-1-aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-12 09:21:03 +01:00
spapr_phb_nvgpu_populate_pcidev_dt(dev, fdt, offset, sphb);
if (!pc->is_bridge) {
/* Properties only for non-bridges */
uint32_t min_grant = pci_default_read_config(dev, PCI_MIN_GNT, 1);
uint32_t max_latency = pci_default_read_config(dev, PCI_MAX_LAT, 1);
_FDT(fdt_setprop_cell(fdt, offset, "min-grant", min_grant));
_FDT(fdt_setprop_cell(fdt, offset, "max-latency", max_latency));
return offset;
} else {
PCIBus *sec_bus = pci_bridge_get_sec_bus(PCI_BRIDGE(dev));
return spapr_dt_pci_bus(sphb, sec_bus, fdt, offset);
}
}
/* Callback to be called during DRC release. */
void spapr_phb_remove_pci_device_cb(DeviceState *dev)
{
HotplugHandler *hotplug_ctrl = qdev_get_hotplug_handler(dev);
hotplug_handler_unplug(hotplug_ctrl, dev, &error_abort);
qdev: Let the hotplug_handler_unplug() caller delete the device When unplugging a device, at one point the device will be destroyed via object_unparent(). This will, one the one hand, unrealize the removed device hierarchy, and on the other hand, destroy/free the device hierarchy. When chaining hotplug handlers, we want to overwrite a bus hotplug handler by the machine hotplug handler, to be able to perform some part of the plug/unplug and to forward the calls to the bus hotplug handler. For now, the bus hotplug handler would trigger an object_unparent(), not allowing us to perform some unplug action on a device after we forwarded the call to the bus hotplug handler. The device would be gone at that point. machine_unplug_handler(dev) /* eventually do unplug stuff */ bus_unplug_handler(dev) /* dev is gone, we can't do more unplug stuff */ So move the object_unparent() to the original caller of the unplug. For now, keep the unrealize() at the original places of the object_unparent(). For implicitly chained hotplug handlers (e.g. pc code calling acpi hotplug handlers), the object_unparent() has to be done by the outermost caller. So when calling hotplug_handler_unplug() from inside an unplug handler, nothing is to be done. hotplug_handler_unplug(dev) -> calls machine_unplug_handler() machine_unplug_handler(dev) { /* eventually do unplug stuff */ bus_unplug_handler(dev) -> calls unrealize(dev) /* we can do more unplug stuff but device already unrealized */ } object_unparent(dev) In the long run, every unplug action should be factored out of the unrealize() function into the unplug handler (especially for PCI). Then we can get rid of the additonal unrealize() calls and object_unparent() will properly unrealize the device hierarchy after the device has been unplugged. hotplug_handler_unplug(dev) -> calls machine_unplug_handler() machine_unplug_handler(dev) { /* eventually do unplug stuff */ bus_unplug_handler(dev) -> only unplugs, does not unrealize /* we can do more unplug stuff */ } object_unparent(dev) -> will unrealize The original approach was suggested by Igor Mammedov for the PCI part, but I extended it to all hotplug handlers. I consider this one step into the right direction. To summarize: - object_unparent() on synchronous unplugs is done by common code -- "Caller of hotplug_handler_unplug" - object_unparent() on asynchronous unplugs ("unplug requests") has to be done manually -- "Caller of hotplug_handler_unplug" Reviewed-by: Igor Mammedov <imammedo@redhat.com> Acked-by: Cornelia Huck <cohuck@redhat.com> Signed-off-by: David Hildenbrand <david@redhat.com> Message-Id: <20190228122849.4296-2-david@redhat.com> Reviewed-by: Greg Kurz <groug@kaod.org> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
2019-02-28 13:28:47 +01:00
object_unparent(OBJECT(dev));
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
int spapr_pci_dt_populate(SpaprDrc *drc, SpaprMachineState *spapr,
void *fdt, int *fdt_start_offset, Error **errp)
{
HotplugHandler *plug_handler = qdev_get_hotplug_handler(drc->dev);
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(plug_handler);
PCIDevice *pdev = PCI_DEVICE(drc->dev);
*fdt_start_offset = spapr_dt_pci_device(sphb, pdev, fdt, 0);
return 0;
}
static void spapr_pci_bridge_plug(SpaprPhbState *phb,
PCIBridge *bridge)
{
PCIBus *bus = pci_bridge_get_sec_bus(bridge);
add_drcs(phb, bus);
}
spapr_pci: Robustify support of PCI bridges Some recent error handling cleanups unveiled issues with our support of PCI bridges: 1) QEMU aborts when using non-standard PCI bridge types, unveiled by commit 7ef1553dac "spapr_pci: Drop some dead error handling" $ qemu-system-ppc64 -M pseries -device pcie-pci-bridge Unexpected error in object_property_find() at qom/object.c:1240: qemu-system-ppc64: -device pcie-pci-bridge: Property '.chassis_nr' not found Aborted (core dumped) This happens because we assume all PCI bridge types to have a "chassis_nr" property. This property only exists with the standard PCI bridge type "pci-bridge" actually. We could possibly revert 7ef1553dac but it seems much simpler to check the presence of "chassis_nr" earlier. 2) QEMU abort if same "chassis_nr" value is used several times, unveiled by commit d2623129a7de "qom: Drop parameter @errp of object_property_add() & friends" $ qemu-system-ppc64 -M pseries -device pci-bridge,chassis_nr=1 \ -device pci-bridge,chassis_nr=1 Unexpected error in object_property_try_add() at qom/object.c:1167: qemu-system-ppc64: -device pci-bridge,chassis_nr=1: attempt to add duplicate property '40000100' to object (type 'container') Aborted (core dumped) This happens because we assume that "chassis_nr" values are unique, but nobody enforces that and we end up generating duplicate DRC ids. The PCI code doesn't really care for duplicate "chassis_nr" properties since it is only used to initialize the "Chassis Number Register" of the bridge, with no functional impact on QEMU. So, even if passing the same value several times might look weird, it never broke anything before, so I guess we don't necessarily want to enforce strict checking in the PCI code now. Workaround both issues in the PAPR code: check that the bridge has a unique and non null "chassis_nr" when plugging it into its parent bus. Fixes: 05929a6c5dfe ("spapr: Don't use bus number for building DRC ids") Fixes: 7ef1553dac ("spapr_pci: Drop some dead error handling") Fixes: d2623129a7de ("qom: Drop parameter @errp of object_property_add() & friends") Reported-by: Thomas Huth <thuth@redhat.com> Signed-off-by: Greg Kurz <groug@kaod.org> Message-Id: <159431476748.407044.16711294833569014964.stgit@bahia.lan> [dwg: Move check slightly to a better place] Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2020-07-09 19:12:47 +02:00
/* Returns non-zero if the value of "chassis_nr" is already in use */
static int check_chassis_nr(Object *obj, void *opaque)
{
int new_chassis_nr =
object_property_get_uint(opaque, "chassis_nr", &error_abort);
int chassis_nr =
object_property_get_uint(obj, "chassis_nr", NULL);
if (!object_dynamic_cast(obj, TYPE_PCI_BRIDGE)) {
return 0;
}
/* Skip unsupported bridge types */
if (!chassis_nr) {
return 0;
}
/* Skip self */
if (obj == opaque) {
return 0;
}
return chassis_nr == new_chassis_nr;
}
static bool bridge_has_valid_chassis_nr(Object *bridge, Error **errp)
{
int chassis_nr =
object_property_get_uint(bridge, "chassis_nr", NULL);
/*
* slotid_cap_init() already ensures that "chassis_nr" isn't null for
* standard PCI bridges, so this really tells if "chassis_nr" is present
* or not.
*/
if (!chassis_nr) {
error_setg(errp, "PCI Bridge lacks a \"chassis_nr\" property");
error_append_hint(errp, "Try -device pci-bridge instead.\n");
return false;
}
/* We want unique values for "chassis_nr" */
if (object_child_foreach_recursive(object_get_root(), check_chassis_nr,
bridge)) {
error_setg(errp, "Bridge chassis %d already in use", chassis_nr);
return false;
}
return true;
}
static void spapr_pci_pre_plug(HotplugHandler *plug_handler,
DeviceState *plugged_dev, Error **errp)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *phb = SPAPR_PCI_HOST_BRIDGE(DEVICE(plug_handler));
PCIDevice *pdev = PCI_DEVICE(plugged_dev);
PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(plugged_dev);
SpaprDrc *drc = drc_from_dev(phb, pdev);
PCIBus *bus = PCI_BUS(qdev_get_parent_bus(DEVICE(pdev)));
uint32_t slotnr = PCI_SLOT(pdev->devfn);
if (!phb->dr_enabled) {
/* if this is a hotplug operation initiated by the user
* we need to let them know it's not enabled
*/
if (plugged_dev->hotplugged) {
error_setg(errp, QERR_BUS_NO_HOTPLUG,
object_get_typename(OBJECT(phb)));
return;
}
}
if (pc->is_bridge) {
spapr_pci: Robustify support of PCI bridges Some recent error handling cleanups unveiled issues with our support of PCI bridges: 1) QEMU aborts when using non-standard PCI bridge types, unveiled by commit 7ef1553dac "spapr_pci: Drop some dead error handling" $ qemu-system-ppc64 -M pseries -device pcie-pci-bridge Unexpected error in object_property_find() at qom/object.c:1240: qemu-system-ppc64: -device pcie-pci-bridge: Property '.chassis_nr' not found Aborted (core dumped) This happens because we assume all PCI bridge types to have a "chassis_nr" property. This property only exists with the standard PCI bridge type "pci-bridge" actually. We could possibly revert 7ef1553dac but it seems much simpler to check the presence of "chassis_nr" earlier. 2) QEMU abort if same "chassis_nr" value is used several times, unveiled by commit d2623129a7de "qom: Drop parameter @errp of object_property_add() & friends" $ qemu-system-ppc64 -M pseries -device pci-bridge,chassis_nr=1 \ -device pci-bridge,chassis_nr=1 Unexpected error in object_property_try_add() at qom/object.c:1167: qemu-system-ppc64: -device pci-bridge,chassis_nr=1: attempt to add duplicate property '40000100' to object (type 'container') Aborted (core dumped) This happens because we assume that "chassis_nr" values are unique, but nobody enforces that and we end up generating duplicate DRC ids. The PCI code doesn't really care for duplicate "chassis_nr" properties since it is only used to initialize the "Chassis Number Register" of the bridge, with no functional impact on QEMU. So, even if passing the same value several times might look weird, it never broke anything before, so I guess we don't necessarily want to enforce strict checking in the PCI code now. Workaround both issues in the PAPR code: check that the bridge has a unique and non null "chassis_nr" when plugging it into its parent bus. Fixes: 05929a6c5dfe ("spapr: Don't use bus number for building DRC ids") Fixes: 7ef1553dac ("spapr_pci: Drop some dead error handling") Fixes: d2623129a7de ("qom: Drop parameter @errp of object_property_add() & friends") Reported-by: Thomas Huth <thuth@redhat.com> Signed-off-by: Greg Kurz <groug@kaod.org> Message-Id: <159431476748.407044.16711294833569014964.stgit@bahia.lan> [dwg: Move check slightly to a better place] Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2020-07-09 19:12:47 +02:00
if (!bridge_has_valid_chassis_nr(OBJECT(plugged_dev), errp)) {
return;
}
}
/* Following the QEMU convention used for PCIe multifunction
* hotplug, we do not allow functions to be hotplugged to a
* slot that already has function 0 present
*/
if (plugged_dev->hotplugged && bus->devices[PCI_DEVFN(slotnr, 0)] &&
PCI_FUNC(pdev->devfn) != 0) {
error_setg(errp, "PCI: slot %d function 0 already occupied by %s,"
" additional functions can no longer be exposed to guest.",
slotnr, bus->devices[PCI_DEVFN(slotnr, 0)]->name);
}
if (drc && drc->dev) {
error_setg(errp, "PCI: slot %d already occupied by %s", slotnr,
pci_get_function_0(PCI_DEVICE(drc->dev))->name);
return;
}
}
static void spapr_pci_plug(HotplugHandler *plug_handler,
DeviceState *plugged_dev, Error **errp)
{
SpaprPhbState *phb = SPAPR_PCI_HOST_BRIDGE(DEVICE(plug_handler));
PCIDevice *pdev = PCI_DEVICE(plugged_dev);
PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(plugged_dev);
SpaprDrc *drc = drc_from_dev(phb, pdev);
uint32_t slotnr = PCI_SLOT(pdev->devfn);
/*
* If DR is disabled we don't need to do anything in the case of
* hotplug or coldplug callbacks.
*/
if (!phb->dr_enabled) {
return;
}
g_assert(drc);
if (pc->is_bridge) {
spapr_pci_bridge_plug(phb, PCI_BRIDGE(plugged_dev));
}
/* spapr_pci_pre_plug() already checked the DRC is attachable */
spapr_drc_attach(drc, DEVICE(pdev));
/* If this is function 0, signal hotplug for all the device functions.
* Otherwise defer sending the hotplug event.
*/
spapr: Treat devices added before inbound migration as coldplugged When migrating a guest which has already had devices hotplugged, libvirt typically starts the destination qemu with -incoming defer, adds those hotplugged devices with qmp, then initiates the incoming migration. This causes problems for the management of spapr DRC state. Because the device is treated as hotplugged, it goes into a DRC state for a device immediately after it's plugged, but before the guest has acknowledged its presence. However, chances are the guest on the source machine *has* acknowledged the device's presence and configured it. If the source has fully configured the device, then DRC state won't be sent in the migration stream: for maximum migration compatibility with earlier versions we don't migrate DRCs in coldplug-equivalent state. That means that the DRC effectively changes state over the migrate, causing problems later on. In addition, logging hotplug events for these devices isn't what we want because a) those events should already have been issued on the source host and b) the event queue should get wiped out by the incoming state anyway. In short, what we really want is to treat devices added before an incoming migration as if they were coldplugged. To do this, we first add a spapr_drc_hotplugged() helper which determines if the device is hotplugged in the sense relevant for DRC state management. We only send hotplug events when this is true. Second, when we add a device which isn't hotplugged in this sense, we force a reset of the DRC state - this ensures the DRC is in a coldplug-equivalent state (there isn't usually a system reset between these device adds and the incoming migration). This is based on an earlier patch by Laurent Vivier, cleaned up and extended. Signed-off-by: Laurent Vivier <lvivier@redhat.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Greg Kurz <groug@kaod.org> Tested-by: Daniel Barboza <danielhb@linux.vnet.ibm.com>
2017-06-09 13:08:10 +02:00
if (!spapr_drc_hotplugged(plugged_dev)) {
spapr_drc_reset(drc);
} else if (PCI_FUNC(pdev->devfn) == 0) {
int i;
uint8_t chassis = chassis_from_bus(pci_get_bus(pdev));
for (i = 0; i < 8; i++) {
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprDrc *func_drc;
SpaprDrcClass *func_drck;
SpaprDREntitySense state;
func_drc = drc_from_devfn(phb, chassis, PCI_DEVFN(slotnr, i));
func_drck = SPAPR_DR_CONNECTOR_GET_CLASS(func_drc);
state = func_drck->dr_entity_sense(func_drc);
if (state == SPAPR_DR_ENTITY_SENSE_PRESENT) {
spapr_hotplug_req_add_by_index(func_drc);
}
}
}
}
static void spapr_pci_bridge_unplug(SpaprPhbState *phb,
PCIBridge *bridge)
{
PCIBus *bus = pci_bridge_get_sec_bus(bridge);
remove_drcs(phb, bus);
}
static void spapr_pci_unplug(HotplugHandler *plug_handler,
DeviceState *plugged_dev, Error **errp)
{
PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(plugged_dev);
SpaprPhbState *phb = SPAPR_PCI_HOST_BRIDGE(DEVICE(plug_handler));
/* some version guests do not wait for completion of a device
* cleanup (generally done asynchronously by the kernel) before
* signaling to QEMU that the device is safe, but instead sleep
* for some 'safe' period of time. unfortunately on a busy host
* this sleep isn't guaranteed to be long enough, resulting in
* bad things like IRQ lines being left asserted during final
* device removal. to deal with this we call reset just prior
* to finalizing the device, which will put the device back into
* an 'idle' state, as the device cleanup code expects.
*/
pci_device_reset(PCI_DEVICE(plugged_dev));
if (pc->is_bridge) {
spapr_pci_bridge_unplug(phb, PCI_BRIDGE(plugged_dev));
return;
}
qdev_unrealize(plugged_dev);
}
static void spapr_pci_unplug_request(HotplugHandler *plug_handler,
DeviceState *plugged_dev, Error **errp)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *phb = SPAPR_PCI_HOST_BRIDGE(DEVICE(plug_handler));
PCIDevice *pdev = PCI_DEVICE(plugged_dev);
SpaprDrc *drc = drc_from_dev(phb, pdev);
if (!phb->dr_enabled) {
error_setg(errp, QERR_BUS_NO_HOTPLUG,
object_get_typename(OBJECT(phb)));
return;
}
g_assert(drc);
g_assert(drc->dev == plugged_dev);
if (!spapr_drc_unplug_requested(drc)) {
PCIDeviceClass *pc = PCI_DEVICE_GET_CLASS(plugged_dev);
uint32_t slotnr = PCI_SLOT(pdev->devfn);
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprDrc *func_drc;
SpaprDrcClass *func_drck;
SpaprDREntitySense state;
int i;
uint8_t chassis = chassis_from_bus(pci_get_bus(pdev));
if (pc->is_bridge) {
error_setg(errp, "PCI: Hot unplug of PCI bridges not supported");
return;
}
if (object_property_get_uint(OBJECT(pdev), "nvlink2-tgt", NULL)) {
error_setg(errp, "PCI: Cannot unplug NVLink2 devices");
return;
}
/* ensure any other present functions are pending unplug */
if (PCI_FUNC(pdev->devfn) == 0) {
for (i = 1; i < 8; i++) {
func_drc = drc_from_devfn(phb, chassis, PCI_DEVFN(slotnr, i));
func_drck = SPAPR_DR_CONNECTOR_GET_CLASS(func_drc);
state = func_drck->dr_entity_sense(func_drc);
if (state == SPAPR_DR_ENTITY_SENSE_PRESENT
&& !spapr_drc_unplug_requested(func_drc)) {
spapr_pci: remove all child functions in function zero unplug There is nothing wrong with how sPAPR handles multifunction PCI hot unplugs. The problem is that x86 does it simpler. Instead of removing each non-zero function and then removing function zero, x86 can remove any function of the slot to trigger the hot unplug. Libvirt will be directly impacted by this difference, in the (hopefully soon) PCI Multifunction hot plug/unplug support. For hot plugs, both x86 and sPAPR will operate the same way: a XML with all desired functions to be added, then consecutive hotplugs of all non-zero functions first, zero last. For hot unplugs, at least in the current state, a XML with the devices to be removed must also be provided because of how sPAPR operates - x86 does not need it - since any function unplug will unplug the whole PCIe slot. This difference puts extra strain in the management layer, which needs to either handle both archs differently in the unplug scenario or choose treat x86 like sPAPR, forcing x86 users to cope with sPAPR internals. This patch changes spapr_pci_unplug_request to handle the unplug of function zero differently. When removing function zero, instead of error-ing out if there are any remaining function DRCs which needs detaching, detach those. This has no effect in any existing scripts that are detaching the non-zero functions before function zero, and can be used by management as a shortcut to remove the whole PCI multifunction device without specifying each child function. Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com> Message-Id: <20190822195918.3307-1-danielhb413@gmail.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-08-22 21:59:18 +02:00
/*
* Attempting to remove function 0 of a multifunction
* device will will cascade into removing all child
* functions, even if their unplug weren't requested
* beforehand.
*/
spapr: rename spapr_drc_detach() to spapr_drc_unplug_request() spapr_drc_detach() is not the best name for what the function does. The function does not detach the DRC, it makes an uncommited attempt to do it. It'll mark the DRC as pending unplug, via the 'unplug_request' flag, and only if the DRC state is drck->empty_state it will detach the DRC, via spapr_drc_release(). This is a contrast with its pair spapr_drc_attach(), where the function is indeed creating the DRC QOM object. If you know what spapr_drc_attach() does, you can be misled into thinking that spapr_drc_detach() is removing the DRC from QEMU internal state, which isn't true. The current role of this function is better described as a request for detach, since there's no guarantee that we're going to detach the DRC in the end. Rename the function to spapr_drc_unplug_request to reflect what is is doing. The initial idea was to change the name to spapr_drc_detach_request(), and later on change the unplug_request flag to detach_request. However, unplug_request is a migratable boolean for a long time now and renaming it is not worth the trouble. spapr_drc_unplug_request() setting drc->unplug_request is more natural than spapr_drc_detach_request setting drc->unplug_request. Reviewed-by: Greg Kurz <groug@kaod.org> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com> Message-Id: <20210222194531.62717-3-danielhb413@gmail.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2021-02-22 20:45:28 +01:00
spapr_drc_unplug_request(func_drc);
}
}
}
spapr: rename spapr_drc_detach() to spapr_drc_unplug_request() spapr_drc_detach() is not the best name for what the function does. The function does not detach the DRC, it makes an uncommited attempt to do it. It'll mark the DRC as pending unplug, via the 'unplug_request' flag, and only if the DRC state is drck->empty_state it will detach the DRC, via spapr_drc_release(). This is a contrast with its pair spapr_drc_attach(), where the function is indeed creating the DRC QOM object. If you know what spapr_drc_attach() does, you can be misled into thinking that spapr_drc_detach() is removing the DRC from QEMU internal state, which isn't true. The current role of this function is better described as a request for detach, since there's no guarantee that we're going to detach the DRC in the end. Rename the function to spapr_drc_unplug_request to reflect what is is doing. The initial idea was to change the name to spapr_drc_detach_request(), and later on change the unplug_request flag to detach_request. However, unplug_request is a migratable boolean for a long time now and renaming it is not worth the trouble. spapr_drc_unplug_request() setting drc->unplug_request is more natural than spapr_drc_detach_request setting drc->unplug_request. Reviewed-by: Greg Kurz <groug@kaod.org> Reviewed-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com> Message-Id: <20210222194531.62717-3-danielhb413@gmail.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2021-02-22 20:45:28 +01:00
spapr_drc_unplug_request(drc);
/* if this isn't func 0, defer unplug event. otherwise signal removal
* for all present functions
*/
if (PCI_FUNC(pdev->devfn) == 0) {
for (i = 7; i >= 0; i--) {
func_drc = drc_from_devfn(phb, chassis, PCI_DEVFN(slotnr, i));
func_drck = SPAPR_DR_CONNECTOR_GET_CLASS(func_drc);
state = func_drck->dr_entity_sense(func_drc);
if (state == SPAPR_DR_ENTITY_SENSE_PRESENT) {
spapr_hotplug_req_remove_by_index(func_drc);
}
}
}
} else {
error_setg(errp,
"PCI device unplug already in progress for device %s",
drc->dev->id);
}
}
static void spapr_phb_finalizefn(Object *obj)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(obj);
g_free(sphb->dtbusname);
sphb->dtbusname = NULL;
}
qdev: Unrealize must not fail Devices may have component devices and buses. Device realization may fail. Realization is recursive: a device's realize() method realizes its components, and device_set_realized() realizes its buses (which should in turn realize the devices on that bus, except bus_set_realized() doesn't implement that, yet). When realization of a component or bus fails, we need to roll back: unrealize everything we realized so far. If any of these unrealizes failed, the device would be left in an inconsistent state. Must not happen. device_set_realized() lets it happen: it ignores errors in the roll back code starting at label child_realize_fail. Since realization is recursive, unrealization must be recursive, too. But how could a partly failed unrealize be rolled back? We'd have to re-realize, which can fail. This design is fundamentally broken. device_set_realized() does not roll back at all. Instead, it keeps unrealizing, ignoring further errors. It can screw up even for a device with no buses: if the lone dc->unrealize() fails, it still unregisters vmstate, and calls listeners' unrealize() callback. bus_set_realized() does not roll back either. Instead, it stops unrealizing. Fortunately, no unrealize method can fail, as we'll see below. To fix the design error, drop parameter @errp from all the unrealize methods. Any unrealize method that uses @errp now needs an update. This leads us to unrealize() methods that can fail. Merely passing it to another unrealize method cannot cause failure, though. Here are the ones that do other things with @errp: * virtio_serial_device_unrealize() Fails when qbus_set_hotplug_handler() fails, but still does all the other work. On failure, the device would stay realized with its resources completely gone. Oops. Can't happen, because qbus_set_hotplug_handler() can't actually fail here. Pass &error_abort to qbus_set_hotplug_handler() instead. * hw/ppc/spapr_drc.c's unrealize() Fails when object_property_del() fails, but all the other work is already done. On failure, the device would stay realized with its vmstate registration gone. Oops. Can't happen, because object_property_del() can't actually fail here. Pass &error_abort to object_property_del() instead. * spapr_phb_unrealize() Fails and bails out when remove_drcs() fails, but other work is already done. On failure, the device would stay realized with some of its resources gone. Oops. remove_drcs() fails only when chassis_from_bus()'s object_property_get_uint() fails, and it can't here. Pass &error_abort to remove_drcs() instead. Therefore, no unrealize method can fail before this patch. device_set_realized()'s recursive unrealization via bus uses object_property_set_bool(). Can't drop @errp there, so pass &error_abort. We similarly unrealize with object_property_set_bool() elsewhere, always ignoring errors. Pass &error_abort instead. Several unrealize methods no longer handle errors from other unrealize methods: virtio_9p_device_unrealize(), virtio_input_device_unrealize(), scsi_qdev_unrealize(), ... Much of the deleted error handling looks wrong anyway. One unrealize methods no longer ignore such errors: usb_ehci_pci_exit(). Several realize methods no longer ignore errors when rolling back: v9fs_device_realize_common(), pci_qdev_unrealize(), spapr_phb_realize(), usb_qdev_realize(), vfio_ccw_realize(), virtio_device_realize(). Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <20200505152926.18877-17-armbru@redhat.com>
2020-05-05 17:29:24 +02:00
static void spapr_phb_unrealize(DeviceState *dev)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr = SPAPR_MACHINE(qdev_get_machine());
SysBusDevice *s = SYS_BUS_DEVICE(dev);
PCIHostState *phb = PCI_HOST_BRIDGE(s);
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(phb);
SpaprTceTable *tcet;
int i;
const unsigned windows_supported = spapr_phb_windows_supported(sphb);
spapr: Support NVIDIA V100 GPU with NVLink2 NVIDIA V100 GPUs have on-board RAM which is mapped into the host memory space and accessible as normal RAM via an NVLink bus. The VFIO-PCI driver implements special regions for such GPUs and emulates an NVLink bridge. NVLink2-enabled POWER9 CPUs also provide address translation services which includes an ATS shootdown (ATSD) register exported via the NVLink bridge device. This adds a quirk to VFIO to map the GPU memory and create an MR; the new MR is stored in a PCI device as a QOM link. The sPAPR PCI uses this to get the MR and map it to the system address space. Another quirk does the same for ATSD. This adds additional steps to sPAPR PHB setup: 1. Search for specific GPUs and NPUs, collect findings in sPAPRPHBState::nvgpus, manage system address space mappings; 2. Add device-specific properties such as "ibm,npu", "ibm,gpu", "memory-block", "link-speed" to advertise the NVLink2 function to the guest; 3. Add "mmio-atsd" to vPHB to advertise the ATSD capability; 4. Add new memory blocks (with extra "linux,memory-usable" to prevent the guest OS from accessing the new memory until it is onlined) and npuphb# nodes representing an NPU unit for every vPHB as the GPU driver uses it for link discovery. This allocates space for GPU RAM and ATSD like we do for MMIOs by adding 2 new parameters to the phb_placement() hook. Older machine types set these to zero. This puts new memory nodes in a separate NUMA node to as the GPU RAM needs to be configured equally distant from any other node in the system. Unlike the host setup which assigns numa ids from 255 downwards, this adds new NUMA nodes after the user configures nodes or from 1 if none were configured. This adds requirement similar to EEH - one IOMMU group per vPHB. The reason for this is that ATSD registers belong to a physical NPU so they cannot invalidate translations on GPUs attached to another NPU. It is guaranteed by the host platform as it does not mix NVLink bridges or GPUs from different NPU in the same IOMMU group. If more than one IOMMU group is detected on a vPHB, this disables ATSD support for that vPHB and prints a warning. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [aw: for vfio portions] Acked-by: Alex Williamson <alex.williamson@redhat.com> Message-Id: <20190312082103.130561-1-aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-12 09:21:03 +01:00
spapr_phb_nvgpu_free(sphb);
if (sphb->msi) {
g_hash_table_unref(sphb->msi);
sphb->msi = NULL;
}
/*
* Remove IO/MMIO subregions and aliases, rest should get cleaned
* via PHB's unrealize->object_finalize
*/
for (i = windows_supported - 1; i >= 0; i--) {
tcet = spapr_tce_find_by_liobn(sphb->dma_liobn[i]);
if (tcet) {
memory_region_del_subregion(&sphb->iommu_root,
spapr_tce_get_iommu(tcet));
}
}
remove_drcs(sphb, phb->bus);
for (i = PCI_NUM_PINS - 1; i >= 0; i--) {
if (sphb->lsi_table[i].irq) {
spapr_irq_free(spapr, sphb->lsi_table[i].irq, 1);
sphb->lsi_table[i].irq = 0;
}
}
QLIST_REMOVE(sphb, list);
memory_region_del_subregion(&sphb->iommu_root, &sphb->msiwindow);
2019-06-21 11:27:33 +02:00
/*
* An attached PCI device may have memory listeners, eg. VFIO PCI. We have
* unmapped all sections. Remove the listeners now, before destroying the
* address space.
*/
address_space_remove_listeners(&sphb->iommu_as);
address_space_destroy(&sphb->iommu_as);
qbus_set_hotplug_handler(BUS(phb->bus), NULL);
pci_unregister_root_bus(phb->bus);
memory_region_del_subregion(get_system_memory(), &sphb->iowindow);
if (sphb->mem64_win_pciaddr != (hwaddr)-1) {
memory_region_del_subregion(get_system_memory(), &sphb->mem64window);
}
memory_region_del_subregion(get_system_memory(), &sphb->mem32window);
}
static void spapr_phb_destroy_msi(gpointer opaque)
{
SpaprMachineState *spapr = SPAPR_MACHINE(qdev_get_machine());
SpaprMachineClass *smc = SPAPR_MACHINE_GET_CLASS(spapr);
SpaprPciMsi *msi = opaque;
if (!smc->legacy_irq_allocation) {
spapr_irq_msi_free(spapr, msi->first_irq, msi->num);
}
spapr_irq_free(spapr, msi->first_irq, msi->num);
g_free(msi);
}
static void spapr_phb_realize(DeviceState *dev, Error **errp)
{
ERRP_GUARD();
/* We don't use SPAPR_MACHINE() in order to exit gracefully if the user
* tries to add a sPAPR PHB to a non-pseries machine.
*/
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineState *spapr =
(SpaprMachineState *) object_dynamic_cast(qdev_get_machine(),
TYPE_SPAPR_MACHINE);
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprMachineClass *smc = spapr ? SPAPR_MACHINE_GET_CLASS(spapr) : NULL;
SysBusDevice *s = SYS_BUS_DEVICE(dev);
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(s);
PCIHostState *phb = PCI_HOST_BRIDGE(s);
MachineState *ms = MACHINE(spapr);
char *namebuf;
int i;
PCIBus *bus;
uint64_t msi_window_size = 4096;
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprTceTable *tcet;
const unsigned windows_supported = spapr_phb_windows_supported(sphb);
if (!spapr) {
error_setg(errp, TYPE_SPAPR_PCI_HOST_BRIDGE " needs a pseries machine");
return;
}
assert(sphb->index != (uint32_t)-1); /* checked in spapr_phb_pre_plug() */
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
if (sphb->mem64_win_size != 0) {
if (sphb->mem_win_size > SPAPR_PCI_MEM32_WIN_SIZE) {
error_setg(errp, "32-bit memory window of size 0x%"HWADDR_PRIx
" (max 2 GiB)", sphb->mem_win_size);
return;
}
/* 64-bit window defaults to identity mapping */
sphb->mem64_win_pciaddr = sphb->mem64_win_addr;
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
} else if (sphb->mem_win_size > SPAPR_PCI_MEM32_WIN_SIZE) {
/*
* For compatibility with old configuration, if no 64-bit MMIO
* window is specified, but the ordinary (32-bit) memory
* window is specified as > 2GiB, we treat it as a 2GiB 32-bit
* window, with a 64-bit MMIO window following on immediately
* afterwards
*/
sphb->mem64_win_size = sphb->mem_win_size - SPAPR_PCI_MEM32_WIN_SIZE;
sphb->mem64_win_addr = sphb->mem_win_addr + SPAPR_PCI_MEM32_WIN_SIZE;
sphb->mem64_win_pciaddr =
SPAPR_PCI_MEM_WIN_BUS_OFFSET + SPAPR_PCI_MEM32_WIN_SIZE;
sphb->mem_win_size = SPAPR_PCI_MEM32_WIN_SIZE;
}
if (spapr_pci_find_phb(spapr, sphb->buid)) {
SpaprPhbState *s;
error_setg(errp, "PCI host bridges must have unique indexes");
error_append_hint(errp, "The following indexes are already in use:");
QLIST_FOREACH(s, &spapr->phbs, list) {
error_append_hint(errp, " %d", s->index);
}
error_append_hint(errp, "\nTry another value for the index property\n");
return;
}
spapr_pci: advertise explicit numa IDs even when there's 1 node With the addition of "numa_node" properties for PHBs we began advertising NUMA affinity in cases where nb_numa_nodes > 1. Since the default on the guest side is to make no assumptions about PHB NUMA affinity (defaulting to -1), there is still a valid use-case for explicitly defining a PHB's NUMA affinity even when there's just one node. In particular, some workloads make faulty assumptions about /sys/bus/pci/<devid>/numa_node being >= 0, warranting the use of this property as a workaround even if there's just 1 PHB or NUMA node. Enable this use-case by always advertising the PHB's NUMA affinity if "numa_node" has been explicitly set. We could achieve this by relaxing the check to simply be nb_numa_nodes > 0, but even safer would be to check numa_info[nodeid].present explicitly, and to fail at start time for cases where it does not exist. This has an additional affect of no longer advertising PHB NUMA affinity unconditionally if nb_numa_nodes > 1 and "numa_node" property is unset/-1, but since the default value on the guest side for each PHB is also -1, the behavior should be the same for that situation. We could still retain the old behavior if desired, but the decision seems arbitrary, so we take the simpler route. Cc: Alexey Kardashevskiy <aik@ozlabs.ru> Cc: Shivaprasad G. Bhat <shivapbh@in.ibm.com> Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-10-18 22:50:23 +02:00
if (sphb->numa_node != -1 &&
(sphb->numa_node >= MAX_NODES ||
!ms->numa_state->nodes[sphb->numa_node].present)) {
spapr_pci: advertise explicit numa IDs even when there's 1 node With the addition of "numa_node" properties for PHBs we began advertising NUMA affinity in cases where nb_numa_nodes > 1. Since the default on the guest side is to make no assumptions about PHB NUMA affinity (defaulting to -1), there is still a valid use-case for explicitly defining a PHB's NUMA affinity even when there's just one node. In particular, some workloads make faulty assumptions about /sys/bus/pci/<devid>/numa_node being >= 0, warranting the use of this property as a workaround even if there's just 1 PHB or NUMA node. Enable this use-case by always advertising the PHB's NUMA affinity if "numa_node" has been explicitly set. We could achieve this by relaxing the check to simply be nb_numa_nodes > 0, but even safer would be to check numa_info[nodeid].present explicitly, and to fail at start time for cases where it does not exist. This has an additional affect of no longer advertising PHB NUMA affinity unconditionally if nb_numa_nodes > 1 and "numa_node" property is unset/-1, but since the default value on the guest side for each PHB is also -1, the behavior should be the same for that situation. We could still retain the old behavior if desired, but the decision seems arbitrary, so we take the simpler route. Cc: Alexey Kardashevskiy <aik@ozlabs.ru> Cc: Shivaprasad G. Bhat <shivapbh@in.ibm.com> Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-10-18 22:50:23 +02:00
error_setg(errp, "Invalid NUMA node ID for PCI host bridge");
return;
}
sphb->dtbusname = g_strdup_printf("pci@%" PRIx64, sphb->buid);
/* Initialize memory regions */
namebuf = g_strdup_printf("%s.mmio", sphb->dtbusname);
memory_region_init(&sphb->memspace, OBJECT(sphb), namebuf, UINT64_MAX);
g_free(namebuf);
namebuf = g_strdup_printf("%s.mmio32-alias", sphb->dtbusname);
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
memory_region_init_alias(&sphb->mem32window, OBJECT(sphb),
namebuf, &sphb->memspace,
SPAPR_PCI_MEM_WIN_BUS_OFFSET, sphb->mem_win_size);
g_free(namebuf);
memory_region_add_subregion(get_system_memory(), sphb->mem_win_addr,
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
&sphb->mem32window);
if (sphb->mem64_win_size != 0) {
namebuf = g_strdup_printf("%s.mmio64-alias", sphb->dtbusname);
memory_region_init_alias(&sphb->mem64window, OBJECT(sphb),
namebuf, &sphb->memspace,
sphb->mem64_win_pciaddr, sphb->mem64_win_size);
g_free(namebuf);
memory_region_add_subregion(get_system_memory(),
sphb->mem64_win_addr,
&sphb->mem64window);
}
/* Initialize IO regions */
namebuf = g_strdup_printf("%s.io", sphb->dtbusname);
memory_region_init(&sphb->iospace, OBJECT(sphb),
namebuf, SPAPR_PCI_IO_WIN_SIZE);
g_free(namebuf);
namebuf = g_strdup_printf("%s.io-alias", sphb->dtbusname);
memory_region_init_alias(&sphb->iowindow, OBJECT(sphb), namebuf,
&sphb->iospace, 0, SPAPR_PCI_IO_WIN_SIZE);
g_free(namebuf);
memory_region_add_subregion(get_system_memory(), sphb->io_win_addr,
&sphb->iowindow);
spapr_pci: Fix broken naming of PCI bus Recent commit 5cf0d326a0fe fixed a regression which was preventing the guest to access the extended config space of a PCIe device. This was done by introducing a new PCI bus subtype for PAPR. The original fix was causing PCI busses to be named "spapr-pci-host-bridge-root-bus.N" instead of "pci.N", which was making upper layers unhappy of course. This got worked around by hardcoding the PCI bus name to "pci.0", but this only works for the default PHB. And we're now hitting: # qemu-system-ppc64 \ -device spapr-pci-host-bridge,index=1 \ -device e1000e,bus=pci.0 \ -device e1000e,bus=pci.1 qemu-system-ppc64: -device e1000e,bus=pci.1: Bus 'pci.1' not found David already posted some patches [1] to control PCI extended config space accesses with a new flag in the base PCI bus class instead of subtyping. These patches are a bit more intrusive though, and are targetted for 4.1. When no name is passed to pci_register_bus(), the core device code generates a lowercase name based on the QOM typename. The typename for the base PCI bus class is "PCI", hence the "pci.0", "pci.1" bus names. Rename the type of the PAPR PCI bus to "pci", so that the QOM code can generate proper names. This is a hack but it is enough to fix the regression. And all this will be reworked properly in 4.1. [1] https://patchwork.ozlabs.org/project/qemu-devel/list/?series=100486 Fixes: 5cf0d326a0fe Signed-off-by: Greg Kurz <groug@kaod.org> Message-Id: <155500034416.646888.1307366522340665522.stgit@bahia.lab.toulouse-stg.fr.ibm.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-04-11 18:32:24 +02:00
bus = pci_register_root_bus(dev, NULL,
pci_spapr_set_irq, pci_swizzle_map_irq_fn, sphb,
&sphb->memspace, &sphb->iospace,
PCI_DEVFN(0, 0), PCI_NUM_PINS,
TYPE_PCI_BUS);
/*
* Despite resembling a vanilla PCI bus in most ways, the PAPR
* para-virtualized PCI bus *does* permit PCI-E extended config
* space access
*/
if (sphb->pcie_ecs) {
bus->flags |= PCI_BUS_EXTENDED_CONFIG_SPACE;
}
phb->bus = bus;
qbus_set_hotplug_handler(BUS(phb->bus), OBJECT(sphb));
/*
* Initialize PHB address space.
* By default there will be at least one subregion for default
* 32bit DMA window.
* Later the guest might want to create another DMA window
* which will become another memory subregion.
*/
namebuf = g_strdup_printf("%s.iommu-root", sphb->dtbusname);
memory_region_init(&sphb->iommu_root, OBJECT(sphb),
namebuf, UINT64_MAX);
g_free(namebuf);
address_space_init(&sphb->iommu_as, &sphb->iommu_root,
sphb->dtbusname);
/*
* As MSI/MSIX interrupts trigger by writing at MSI/MSIX vectors,
* we need to allocate some memory to catch those writes coming
* from msi_notify()/msix_notify().
* As MSIMessage:addr is going to be the same and MSIMessage:data
* is going to be a VIRQ number, 4 bytes of the MSI MR will only
* be used.
*
* For KVM we want to ensure that this memory is a full page so that
* our memory slot is of page size granularity.
*/
if (kvm_enabled()) {
msi_window_size = qemu_real_host_page_size();
}
memory_region_init_io(&sphb->msiwindow, OBJECT(sphb), &spapr_msi_ops, spapr,
"msi", msi_window_size);
memory_region_add_subregion(&sphb->iommu_root, SPAPR_PCI_MSI_WINDOW,
&sphb->msiwindow);
pci_setup_iommu(bus, spapr_pci_dma_iommu, sphb);
pci_bus_set_route_irq_fn(bus, spapr_route_intx_pin_to_irq);
QLIST_INSERT_HEAD(&spapr->phbs, sphb, list);
/* Initialize the LSI table */
for (i = 0; i < PCI_NUM_PINS; i++) {
int irq = SPAPR_IRQ_PCI_LSI + sphb->index * PCI_NUM_PINS + i;
if (smc->legacy_irq_allocation) {
irq = spapr_irq_findone(spapr, errp);
if (irq < 0) {
error_prepend(errp, "can't allocate LSIs: ");
/*
* Older machines will never support PHB hotplug, ie, this is an
* init only path and QEMU will terminate. No need to rollback.
*/
return;
}
}
if (spapr_irq_claim(spapr, irq, true, errp) < 0) {
error_prepend(errp, "can't allocate LSIs: ");
goto unrealize;
}
sphb->lsi_table[i].irq = irq;
}
/* allocate connectors for child PCI devices */
add_drcs(sphb, phb->bus);
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
/* DMA setup */
for (i = 0; i < windows_supported; ++i) {
tcet = spapr_tce_new_table(DEVICE(sphb), sphb->dma_liobn[i]);
if (!tcet) {
error_setg(errp, "Creating window#%d failed for %s",
i, sphb->dtbusname);
goto unrealize;
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
}
memory_region_add_subregion(&sphb->iommu_root, 0,
spapr_tce_get_iommu(tcet));
}
sphb->msi = g_hash_table_new_full(g_int_hash, g_int_equal, g_free,
spapr_phb_destroy_msi);
return;
unrealize:
qdev: Unrealize must not fail Devices may have component devices and buses. Device realization may fail. Realization is recursive: a device's realize() method realizes its components, and device_set_realized() realizes its buses (which should in turn realize the devices on that bus, except bus_set_realized() doesn't implement that, yet). When realization of a component or bus fails, we need to roll back: unrealize everything we realized so far. If any of these unrealizes failed, the device would be left in an inconsistent state. Must not happen. device_set_realized() lets it happen: it ignores errors in the roll back code starting at label child_realize_fail. Since realization is recursive, unrealization must be recursive, too. But how could a partly failed unrealize be rolled back? We'd have to re-realize, which can fail. This design is fundamentally broken. device_set_realized() does not roll back at all. Instead, it keeps unrealizing, ignoring further errors. It can screw up even for a device with no buses: if the lone dc->unrealize() fails, it still unregisters vmstate, and calls listeners' unrealize() callback. bus_set_realized() does not roll back either. Instead, it stops unrealizing. Fortunately, no unrealize method can fail, as we'll see below. To fix the design error, drop parameter @errp from all the unrealize methods. Any unrealize method that uses @errp now needs an update. This leads us to unrealize() methods that can fail. Merely passing it to another unrealize method cannot cause failure, though. Here are the ones that do other things with @errp: * virtio_serial_device_unrealize() Fails when qbus_set_hotplug_handler() fails, but still does all the other work. On failure, the device would stay realized with its resources completely gone. Oops. Can't happen, because qbus_set_hotplug_handler() can't actually fail here. Pass &error_abort to qbus_set_hotplug_handler() instead. * hw/ppc/spapr_drc.c's unrealize() Fails when object_property_del() fails, but all the other work is already done. On failure, the device would stay realized with its vmstate registration gone. Oops. Can't happen, because object_property_del() can't actually fail here. Pass &error_abort to object_property_del() instead. * spapr_phb_unrealize() Fails and bails out when remove_drcs() fails, but other work is already done. On failure, the device would stay realized with some of its resources gone. Oops. remove_drcs() fails only when chassis_from_bus()'s object_property_get_uint() fails, and it can't here. Pass &error_abort to remove_drcs() instead. Therefore, no unrealize method can fail before this patch. device_set_realized()'s recursive unrealization via bus uses object_property_set_bool(). Can't drop @errp there, so pass &error_abort. We similarly unrealize with object_property_set_bool() elsewhere, always ignoring errors. Pass &error_abort instead. Several unrealize methods no longer handle errors from other unrealize methods: virtio_9p_device_unrealize(), virtio_input_device_unrealize(), scsi_qdev_unrealize(), ... Much of the deleted error handling looks wrong anyway. One unrealize methods no longer ignore such errors: usb_ehci_pci_exit(). Several realize methods no longer ignore errors when rolling back: v9fs_device_realize_common(), pci_qdev_unrealize(), spapr_phb_realize(), usb_qdev_realize(), vfio_ccw_realize(), virtio_device_realize(). Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Paolo Bonzini <pbonzini@redhat.com> Message-Id: <20200505152926.18877-17-armbru@redhat.com>
2020-05-05 17:29:24 +02:00
spapr_phb_unrealize(dev);
}
static int spapr_phb_children_reset(Object *child, void *opaque)
{
DeviceState *dev = (DeviceState *) object_dynamic_cast(child, TYPE_DEVICE);
if (dev) {
device_legacy_reset(dev);
}
return 0;
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
void spapr_phb_dma_reset(SpaprPhbState *sphb)
{
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
int i;
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprTceTable *tcet;
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
for (i = 0; i < SPAPR_PCI_DMA_MAX_WINDOWS; ++i) {
tcet = spapr_tce_find_by_liobn(sphb->dma_liobn[i]);
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
if (tcet && tcet->nb_table) {
spapr_tce_table_disable(tcet);
}
}
/* Register default 32bit DMA window */
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
tcet = spapr_tce_find_by_liobn(sphb->dma_liobn[0]);
spapr_tce_table_enable(tcet, SPAPR_TCE_PAGE_SHIFT, sphb->dma_win_addr,
sphb->dma_win_size >> SPAPR_TCE_PAGE_SHIFT);
tcet->def_win = true;
}
static void spapr_phb_reset(DeviceState *qdev)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(qdev);
Error *err = NULL;
spapr_phb_dma_reset(sphb);
spapr: Support NVIDIA V100 GPU with NVLink2 NVIDIA V100 GPUs have on-board RAM which is mapped into the host memory space and accessible as normal RAM via an NVLink bus. The VFIO-PCI driver implements special regions for such GPUs and emulates an NVLink bridge. NVLink2-enabled POWER9 CPUs also provide address translation services which includes an ATS shootdown (ATSD) register exported via the NVLink bridge device. This adds a quirk to VFIO to map the GPU memory and create an MR; the new MR is stored in a PCI device as a QOM link. The sPAPR PCI uses this to get the MR and map it to the system address space. Another quirk does the same for ATSD. This adds additional steps to sPAPR PHB setup: 1. Search for specific GPUs and NPUs, collect findings in sPAPRPHBState::nvgpus, manage system address space mappings; 2. Add device-specific properties such as "ibm,npu", "ibm,gpu", "memory-block", "link-speed" to advertise the NVLink2 function to the guest; 3. Add "mmio-atsd" to vPHB to advertise the ATSD capability; 4. Add new memory blocks (with extra "linux,memory-usable" to prevent the guest OS from accessing the new memory until it is onlined) and npuphb# nodes representing an NPU unit for every vPHB as the GPU driver uses it for link discovery. This allocates space for GPU RAM and ATSD like we do for MMIOs by adding 2 new parameters to the phb_placement() hook. Older machine types set these to zero. This puts new memory nodes in a separate NUMA node to as the GPU RAM needs to be configured equally distant from any other node in the system. Unlike the host setup which assigns numa ids from 255 downwards, this adds new NUMA nodes after the user configures nodes or from 1 if none were configured. This adds requirement similar to EEH - one IOMMU group per vPHB. The reason for this is that ATSD registers belong to a physical NPU so they cannot invalidate translations on GPUs attached to another NPU. It is guaranteed by the host platform as it does not mix NVLink bridges or GPUs from different NPU in the same IOMMU group. If more than one IOMMU group is detected on a vPHB, this disables ATSD support for that vPHB and prints a warning. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [aw: for vfio portions] Acked-by: Alex Williamson <alex.williamson@redhat.com> Message-Id: <20190312082103.130561-1-aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-12 09:21:03 +01:00
spapr_phb_nvgpu_free(sphb);
spapr_phb_nvgpu_setup(sphb, &err);
if (err) {
error_report_err(err);
spapr: Support NVIDIA V100 GPU with NVLink2 NVIDIA V100 GPUs have on-board RAM which is mapped into the host memory space and accessible as normal RAM via an NVLink bus. The VFIO-PCI driver implements special regions for such GPUs and emulates an NVLink bridge. NVLink2-enabled POWER9 CPUs also provide address translation services which includes an ATS shootdown (ATSD) register exported via the NVLink bridge device. This adds a quirk to VFIO to map the GPU memory and create an MR; the new MR is stored in a PCI device as a QOM link. The sPAPR PCI uses this to get the MR and map it to the system address space. Another quirk does the same for ATSD. This adds additional steps to sPAPR PHB setup: 1. Search for specific GPUs and NPUs, collect findings in sPAPRPHBState::nvgpus, manage system address space mappings; 2. Add device-specific properties such as "ibm,npu", "ibm,gpu", "memory-block", "link-speed" to advertise the NVLink2 function to the guest; 3. Add "mmio-atsd" to vPHB to advertise the ATSD capability; 4. Add new memory blocks (with extra "linux,memory-usable" to prevent the guest OS from accessing the new memory until it is onlined) and npuphb# nodes representing an NPU unit for every vPHB as the GPU driver uses it for link discovery. This allocates space for GPU RAM and ATSD like we do for MMIOs by adding 2 new parameters to the phb_placement() hook. Older machine types set these to zero. This puts new memory nodes in a separate NUMA node to as the GPU RAM needs to be configured equally distant from any other node in the system. Unlike the host setup which assigns numa ids from 255 downwards, this adds new NUMA nodes after the user configures nodes or from 1 if none were configured. This adds requirement similar to EEH - one IOMMU group per vPHB. The reason for this is that ATSD registers belong to a physical NPU so they cannot invalidate translations on GPUs attached to another NPU. It is guaranteed by the host platform as it does not mix NVLink bridges or GPUs from different NPU in the same IOMMU group. If more than one IOMMU group is detected on a vPHB, this disables ATSD support for that vPHB and prints a warning. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [aw: for vfio portions] Acked-by: Alex Williamson <alex.williamson@redhat.com> Message-Id: <20190312082103.130561-1-aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-12 09:21:03 +01:00
}
/* Reset the IOMMU state */
object_child_foreach(OBJECT(qdev), spapr_phb_children_reset, NULL);
if (spapr_phb_eeh_available(SPAPR_PCI_HOST_BRIDGE(qdev))) {
spapr_phb_vfio_reset(qdev);
}
g_hash_table_remove_all(sphb->msi);
}
static Property spapr_phb_properties[] = {
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_UINT32("index", SpaprPhbState, index, -1),
DEFINE_PROP_UINT64("mem_win_size", SpaprPhbState, mem_win_size,
spapr: Improved placement of PCI host bridges in guest memory map Currently, the MMIO space for accessing PCI on pseries guests begins at 1 TiB in guest address space. Each PCI host bridge (PHB) has a 64 GiB chunk of address space in which it places its outbound PIO and 32-bit and 64-bit MMIO windows. This scheme as several problems: - It limits guest RAM to 1 TiB (though we have a limited fix for this now) - It limits the total MMIO window to 64 GiB. This is not always enough for some of the large nVidia GPGPU cards - Putting all the windows into a single 64 GiB area means that naturally aligning things within there will waste more address space. In addition there was a miscalculation in some of the defaults, which meant that the MMIO windows for each PHB actually slightly overran the 64 GiB region for that PHB. We got away without nasty consequences because the overrun fit within an unused area at the beginning of the next PHB's region, but it's not pretty. This patch implements a new scheme which addresses those problems, and is also closer to what bare metal hardware and pHyp guests generally use. Because some guest versions (including most current distro kernels) can't access PCI MMIO above 64 TiB, we put all the PCI windows between 32 TiB and 64 TiB. This is broken into 1 TiB chunks. The first 1 TiB contains the PIO (64 kiB) and 32-bit MMIO (2 GiB) windows for all of the PHBs. Each subsequent TiB chunk contains a naturally aligned 64-bit MMIO window for one PHB each. This reduces the number of allowed PHBs (without full manual configuration of all the windows) from 256 to 31, but this should still be plenty in practice. We also change some of the default window sizes for manually configured PHBs to saner values. Finally we adjust some tests and libqos so that it correctly uses the new default locations. Ideally it would parse the device tree given to the guest, but that's a more complex problem for another time. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-16 03:04:15 +02:00
SPAPR_PCI_MEM32_WIN_SIZE),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_UINT64("mem64_win_size", SpaprPhbState, mem64_win_size,
spapr: Improved placement of PCI host bridges in guest memory map Currently, the MMIO space for accessing PCI on pseries guests begins at 1 TiB in guest address space. Each PCI host bridge (PHB) has a 64 GiB chunk of address space in which it places its outbound PIO and 32-bit and 64-bit MMIO windows. This scheme as several problems: - It limits guest RAM to 1 TiB (though we have a limited fix for this now) - It limits the total MMIO window to 64 GiB. This is not always enough for some of the large nVidia GPGPU cards - Putting all the windows into a single 64 GiB area means that naturally aligning things within there will waste more address space. In addition there was a miscalculation in some of the defaults, which meant that the MMIO windows for each PHB actually slightly overran the 64 GiB region for that PHB. We got away without nasty consequences because the overrun fit within an unused area at the beginning of the next PHB's region, but it's not pretty. This patch implements a new scheme which addresses those problems, and is also closer to what bare metal hardware and pHyp guests generally use. Because some guest versions (including most current distro kernels) can't access PCI MMIO above 64 TiB, we put all the PCI windows between 32 TiB and 64 TiB. This is broken into 1 TiB chunks. The first 1 TiB contains the PIO (64 kiB) and 32-bit MMIO (2 GiB) windows for all of the PHBs. Each subsequent TiB chunk contains a naturally aligned 64-bit MMIO window for one PHB each. This reduces the number of allowed PHBs (without full manual configuration of all the windows) from 256 to 31, but this should still be plenty in practice. We also change some of the default window sizes for manually configured PHBs to saner values. Finally we adjust some tests and libqos so that it correctly uses the new default locations. Ideally it would parse the device tree given to the guest, but that's a more complex problem for another time. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-16 03:04:15 +02:00
SPAPR_PCI_MEM64_WIN_SIZE),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_UINT64("io_win_size", SpaprPhbState, io_win_size,
SPAPR_PCI_IO_WIN_SIZE),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_BOOL("dynamic-reconfiguration", SpaprPhbState, dr_enabled,
true),
/* Default DMA window is 0..1GB */
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_UINT64("dma_win_addr", SpaprPhbState, dma_win_addr, 0),
DEFINE_PROP_UINT64("dma_win_size", SpaprPhbState, dma_win_size, 0x40000000),
DEFINE_PROP_UINT64("dma64_win_addr", SpaprPhbState, dma64_win_addr,
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
0x800000000000000ULL),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_BOOL("ddw", SpaprPhbState, ddw_enabled, true),
DEFINE_PROP_UINT64("pgsz", SpaprPhbState, page_size_mask,
(1ULL << 12) | (1ULL << 16)
| (1ULL << 21) | (1ULL << 24)),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_UINT32("numa_node", SpaprPhbState, numa_node, -1),
DEFINE_PROP_BOOL("pre-2.8-migration", SpaprPhbState,
spapr: Fix 2.7<->2.8 migration of PCI host bridge daa2369 "spapr_pci: Add a 64-bit MMIO window" subtly broke migration from qemu-2.7 to the current version. It split the device's MMIO window into two pieces for 32-bit and 64-bit MMIO. The patch included backwards compatibility code to convert the old property into the new format. However, the property value was also transferred in the migration stream and compared with a (probably unwise) VMSTATE_EQUAL. So, the "raw" value from 2.7 is compared to the new style converted value from (pre-)2.8 giving a mismatch and migration failure. Along with the actual field that caused the breakage, there are several other ill-advised VMSTATE_EQUAL()s. To fix forwards migration, we read the values in the stream into scratch variables and ignore them, instead of comparing for equality. To fix backwards migration, we populate those scratch variables in pre_save() with adjusted values to match the old behaviour. To permit the eventual possibility of removing this cruft from the stream, we only include these compatibility fields if a new 'pre-2.8-migration' property is set. We clear it on the pseries-2.8 machine type, which obviously can't be migrated backwards, but set it on earlier machine type versions. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
2016-11-23 00:26:38 +01:00
pre_2_8_migration, false),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
DEFINE_PROP_BOOL("pcie-extended-configuration-space", SpaprPhbState,
pcie_ecs, true),
spapr: Support NVIDIA V100 GPU with NVLink2 NVIDIA V100 GPUs have on-board RAM which is mapped into the host memory space and accessible as normal RAM via an NVLink bus. The VFIO-PCI driver implements special regions for such GPUs and emulates an NVLink bridge. NVLink2-enabled POWER9 CPUs also provide address translation services which includes an ATS shootdown (ATSD) register exported via the NVLink bridge device. This adds a quirk to VFIO to map the GPU memory and create an MR; the new MR is stored in a PCI device as a QOM link. The sPAPR PCI uses this to get the MR and map it to the system address space. Another quirk does the same for ATSD. This adds additional steps to sPAPR PHB setup: 1. Search for specific GPUs and NPUs, collect findings in sPAPRPHBState::nvgpus, manage system address space mappings; 2. Add device-specific properties such as "ibm,npu", "ibm,gpu", "memory-block", "link-speed" to advertise the NVLink2 function to the guest; 3. Add "mmio-atsd" to vPHB to advertise the ATSD capability; 4. Add new memory blocks (with extra "linux,memory-usable" to prevent the guest OS from accessing the new memory until it is onlined) and npuphb# nodes representing an NPU unit for every vPHB as the GPU driver uses it for link discovery. This allocates space for GPU RAM and ATSD like we do for MMIOs by adding 2 new parameters to the phb_placement() hook. Older machine types set these to zero. This puts new memory nodes in a separate NUMA node to as the GPU RAM needs to be configured equally distant from any other node in the system. Unlike the host setup which assigns numa ids from 255 downwards, this adds new NUMA nodes after the user configures nodes or from 1 if none were configured. This adds requirement similar to EEH - one IOMMU group per vPHB. The reason for this is that ATSD registers belong to a physical NPU so they cannot invalidate translations on GPUs attached to another NPU. It is guaranteed by the host platform as it does not mix NVLink bridges or GPUs from different NPU in the same IOMMU group. If more than one IOMMU group is detected on a vPHB, this disables ATSD support for that vPHB and prints a warning. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [aw: for vfio portions] Acked-by: Alex Williamson <alex.williamson@redhat.com> Message-Id: <20190312082103.130561-1-aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-12 09:21:03 +01:00
DEFINE_PROP_UINT64("gpa", SpaprPhbState, nv2_gpa_win_addr, 0),
DEFINE_PROP_UINT64("atsd", SpaprPhbState, nv2_atsd_win_addr, 0),
DEFINE_PROP_BOOL("pre-5.1-associativity", SpaprPhbState,
pre_5_1_assoc, false),
DEFINE_PROP_END_OF_LIST(),
};
static const VMStateDescription vmstate_spapr_pci_lsi = {
.name = "spapr_pci/lsi",
.version_id = 1,
.minimum_version_id = 1,
.fields = (VMStateField[]) {
VMSTATE_UINT32_EQUAL(irq, SpaprPciLsi, NULL),
VMSTATE_END_OF_LIST()
},
};
static const VMStateDescription vmstate_spapr_pci_msi = {
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
.name = "spapr_pci/msi",
.version_id = 1,
.minimum_version_id = 1,
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
.fields = (VMStateField []) {
VMSTATE_UINT32(key, SpaprPciMsiMig),
VMSTATE_UINT32(value.first_irq, SpaprPciMsiMig),
VMSTATE_UINT32(value.num, SpaprPciMsiMig),
VMSTATE_END_OF_LIST()
},
};
static int spapr_pci_pre_save(void *opaque)
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = opaque;
GHashTableIter iter;
gpointer key, value;
int i;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
spapr: Fix 2.7<->2.8 migration of PCI host bridge daa2369 "spapr_pci: Add a 64-bit MMIO window" subtly broke migration from qemu-2.7 to the current version. It split the device's MMIO window into two pieces for 32-bit and 64-bit MMIO. The patch included backwards compatibility code to convert the old property into the new format. However, the property value was also transferred in the migration stream and compared with a (probably unwise) VMSTATE_EQUAL. So, the "raw" value from 2.7 is compared to the new style converted value from (pre-)2.8 giving a mismatch and migration failure. Along with the actual field that caused the breakage, there are several other ill-advised VMSTATE_EQUAL()s. To fix forwards migration, we read the values in the stream into scratch variables and ignore them, instead of comparing for equality. To fix backwards migration, we populate those scratch variables in pre_save() with adjusted values to match the old behaviour. To permit the eventual possibility of removing this cruft from the stream, we only include these compatibility fields if a new 'pre-2.8-migration' property is set. We clear it on the pseries-2.8 machine type, which obviously can't be migrated backwards, but set it on earlier machine type versions. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
2016-11-23 00:26:38 +01:00
if (sphb->pre_2_8_migration) {
sphb->mig_liobn = sphb->dma_liobn[0];
sphb->mig_mem_win_addr = sphb->mem_win_addr;
sphb->mig_mem_win_size = sphb->mem_win_size;
sphb->mig_io_win_addr = sphb->io_win_addr;
sphb->mig_io_win_size = sphb->io_win_size;
if ((sphb->mem64_win_size != 0)
&& (sphb->mem64_win_addr
== (sphb->mem_win_addr + sphb->mem_win_size))) {
sphb->mig_mem_win_size += sphb->mem64_win_size;
}
}
g_free(sphb->msi_devs);
sphb->msi_devs = NULL;
sphb->msi_devs_num = g_hash_table_size(sphb->msi);
if (!sphb->msi_devs_num) {
return 0;
}
sphb->msi_devs = g_new(SpaprPciMsiMig, sphb->msi_devs_num);
g_hash_table_iter_init(&iter, sphb->msi);
for (i = 0; g_hash_table_iter_next(&iter, &key, &value); ++i) {
sphb->msi_devs[i].key = *(uint32_t *) key;
sphb->msi_devs[i].value = *(SpaprPciMsi *) value;
}
return 0;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
}
static int spapr_pci_post_save(void *opaque)
{
SpaprPhbState *sphb = opaque;
g_free(sphb->msi_devs);
sphb->msi_devs = NULL;
sphb->msi_devs_num = 0;
return 0;
}
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
static int spapr_pci_post_load(void *opaque, int version_id)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = opaque;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
gpointer key, value;
int i;
for (i = 0; i < sphb->msi_devs_num; ++i) {
key = g_memdup(&sphb->msi_devs[i].key,
sizeof(sphb->msi_devs[i].key));
value = g_memdup(&sphb->msi_devs[i].value,
sizeof(sphb->msi_devs[i].value));
g_hash_table_insert(sphb->msi, key, value);
}
g_free(sphb->msi_devs);
sphb->msi_devs = NULL;
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
sphb->msi_devs_num = 0;
return 0;
}
spapr: Fix 2.7<->2.8 migration of PCI host bridge daa2369 "spapr_pci: Add a 64-bit MMIO window" subtly broke migration from qemu-2.7 to the current version. It split the device's MMIO window into two pieces for 32-bit and 64-bit MMIO. The patch included backwards compatibility code to convert the old property into the new format. However, the property value was also transferred in the migration stream and compared with a (probably unwise) VMSTATE_EQUAL. So, the "raw" value from 2.7 is compared to the new style converted value from (pre-)2.8 giving a mismatch and migration failure. Along with the actual field that caused the breakage, there are several other ill-advised VMSTATE_EQUAL()s. To fix forwards migration, we read the values in the stream into scratch variables and ignore them, instead of comparing for equality. To fix backwards migration, we populate those scratch variables in pre_save() with adjusted values to match the old behaviour. To permit the eventual possibility of removing this cruft from the stream, we only include these compatibility fields if a new 'pre-2.8-migration' property is set. We clear it on the pseries-2.8 machine type, which obviously can't be migrated backwards, but set it on earlier machine type versions. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
2016-11-23 00:26:38 +01:00
static bool pre_2_8_migration(void *opaque, int version_id)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = opaque;
spapr: Fix 2.7<->2.8 migration of PCI host bridge daa2369 "spapr_pci: Add a 64-bit MMIO window" subtly broke migration from qemu-2.7 to the current version. It split the device's MMIO window into two pieces for 32-bit and 64-bit MMIO. The patch included backwards compatibility code to convert the old property into the new format. However, the property value was also transferred in the migration stream and compared with a (probably unwise) VMSTATE_EQUAL. So, the "raw" value from 2.7 is compared to the new style converted value from (pre-)2.8 giving a mismatch and migration failure. Along with the actual field that caused the breakage, there are several other ill-advised VMSTATE_EQUAL()s. To fix forwards migration, we read the values in the stream into scratch variables and ignore them, instead of comparing for equality. To fix backwards migration, we populate those scratch variables in pre_save() with adjusted values to match the old behaviour. To permit the eventual possibility of removing this cruft from the stream, we only include these compatibility fields if a new 'pre-2.8-migration' property is set. We clear it on the pseries-2.8 machine type, which obviously can't be migrated backwards, but set it on earlier machine type versions. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Reviewed-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Greg Kurz <groug@kaod.org> Reviewed-by: Alexey Kardashevskiy <aik@ozlabs.ru>
2016-11-23 00:26:38 +01:00
return sphb->pre_2_8_migration;
}
static const VMStateDescription vmstate_spapr_pci = {
.name = "spapr_pci",
.version_id = 2,
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
.minimum_version_id = 2,
.pre_save = spapr_pci_pre_save,
.post_save = spapr_pci_post_save,
spapr_pci: Use XICS interrupt allocator and do not cache interrupts in PHB Currently SPAPR PHB keeps track of all allocated MSI (here and below MSI stands for both MSI and MSIX) interrupt because XICS used to be unable to reuse interrupts. This is a problem for dynamic MSI reconfiguration which happens when guest reloads a driver or performs PCI hotplug. Another problem is that the existing implementation can enable MSI on 32 devices maximum (SPAPR_MSIX_MAX_DEVS=32) and there is no good reason for that. This makes use of new XICS ability to reuse interrupts. This reorganizes MSI information storage in sPAPRPHBState. Instead of static array of 32 descriptors (one per a PCI function), this patch adds a GHashTable when @config_addr is a key and (first_irq, num) pair is a value. GHashTable can dynamically grow and shrink so the initial limit of 32 devices is gone. This changes migration stream as @msi_table was a static array while new @msi_devs is a dynamic hash table. This adds temporary array which is used for migration, it is populated in "spapr_pci"::pre_save() callback and expanded into the hash table in post_load() callback. Since the destination side does not know the number of MSI-enabled devices in advance and cannot pre-allocate the temporary array to receive migration state, this makes use of new VMSTATE_STRUCT_VARRAY_ALLOC macro which allocates the array automatically. This resets the MSI configuration space when interrupts are released by the ibm,change-msi RTAS call. This fixed traces to be more informative. This changes vmstate_spapr_pci_msi name from "...lsi" to "...msi" which was incorrect by accident. As the internal representation changed, thus bumps migration version number. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [agraf: drop g_malloc_n usage] Signed-off-by: Alexander Graf <agraf@suse.de>
2014-05-30 11:34:20 +02:00
.post_load = spapr_pci_post_load,
.fields = (VMStateField[]) {
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
VMSTATE_UINT64_EQUAL(buid, SpaprPhbState, NULL),
VMSTATE_UINT32_TEST(mig_liobn, SpaprPhbState, pre_2_8_migration),
VMSTATE_UINT64_TEST(mig_mem_win_addr, SpaprPhbState, pre_2_8_migration),
VMSTATE_UINT64_TEST(mig_mem_win_size, SpaprPhbState, pre_2_8_migration),
VMSTATE_UINT64_TEST(mig_io_win_addr, SpaprPhbState, pre_2_8_migration),
VMSTATE_UINT64_TEST(mig_io_win_size, SpaprPhbState, pre_2_8_migration),
VMSTATE_STRUCT_ARRAY(lsi_table, SpaprPhbState, PCI_NUM_PINS, 0,
vmstate_spapr_pci_lsi, SpaprPciLsi),
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
VMSTATE_INT32(msi_devs_num, SpaprPhbState),
VMSTATE_STRUCT_VARRAY_ALLOC(msi_devs, SpaprPhbState, msi_devs_num, 0,
vmstate_spapr_pci_msi, SpaprPciMsiMig),
VMSTATE_END_OF_LIST()
},
};
pci: Replace pci_find_domain() with more general pci_root_bus_path() pci_find_domain() is used in a number of places where we want an id for a whole PCI domain (i.e. the subtree under a PCI root bus). The trouble is that many platforms may support multiple independent host bridges with no hardware supplied notion of domain number. This patch, therefore, replaces calls to pci_find_domain() with calls to a new pci_root_bus_path() returning a string. The new call is implemented in terms of a new callback in the host bridge class, so it can be defined in some way that's well defined for the platform. When no callback is available we fall back on the qbus name. Most current uses of pci_find_domain() are for error or informational messages, so the change in identifiers should be harmless. The exception is pci_get_dev_path(), whose results form part of migration streams. To maintain compatibility with old migration streams, the PIIX PCI host is altered to always supply "0000" for this path, which matches the old domain number (since the code didn't actually support domains other than 0). For the pseries (spapr) PCI bridge we use a different platform-unique identifier (pseries machines can routinely have dozens of PCI host bridges). Theoretically that breaks migration streams, but given that we don't yet have migration support for pseries, it doesn't matter. Any other machines that have working migration support including PCI devices will need to be updated to maintain migration stream compatibility. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2013-06-06 10:48:49 +02:00
static const char *spapr_phb_root_bus_path(PCIHostState *host_bridge,
PCIBus *rootbus)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb = SPAPR_PCI_HOST_BRIDGE(host_bridge);
pci: Replace pci_find_domain() with more general pci_root_bus_path() pci_find_domain() is used in a number of places where we want an id for a whole PCI domain (i.e. the subtree under a PCI root bus). The trouble is that many platforms may support multiple independent host bridges with no hardware supplied notion of domain number. This patch, therefore, replaces calls to pci_find_domain() with calls to a new pci_root_bus_path() returning a string. The new call is implemented in terms of a new callback in the host bridge class, so it can be defined in some way that's well defined for the platform. When no callback is available we fall back on the qbus name. Most current uses of pci_find_domain() are for error or informational messages, so the change in identifiers should be harmless. The exception is pci_get_dev_path(), whose results form part of migration streams. To maintain compatibility with old migration streams, the PIIX PCI host is altered to always supply "0000" for this path, which matches the old domain number (since the code didn't actually support domains other than 0). For the pseries (spapr) PCI bridge we use a different platform-unique identifier (pseries machines can routinely have dozens of PCI host bridges). Theoretically that breaks migration streams, but given that we don't yet have migration support for pseries, it doesn't matter. Any other machines that have working migration support including PCI devices will need to be updated to maintain migration stream compatibility. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2013-06-06 10:48:49 +02:00
return sphb->dtbusname;
}
static void spapr_phb_class_init(ObjectClass *klass, void *data)
{
pci: Replace pci_find_domain() with more general pci_root_bus_path() pci_find_domain() is used in a number of places where we want an id for a whole PCI domain (i.e. the subtree under a PCI root bus). The trouble is that many platforms may support multiple independent host bridges with no hardware supplied notion of domain number. This patch, therefore, replaces calls to pci_find_domain() with calls to a new pci_root_bus_path() returning a string. The new call is implemented in terms of a new callback in the host bridge class, so it can be defined in some way that's well defined for the platform. When no callback is available we fall back on the qbus name. Most current uses of pci_find_domain() are for error or informational messages, so the change in identifiers should be harmless. The exception is pci_get_dev_path(), whose results form part of migration streams. To maintain compatibility with old migration streams, the PIIX PCI host is altered to always supply "0000" for this path, which matches the old domain number (since the code didn't actually support domains other than 0). For the pseries (spapr) PCI bridge we use a different platform-unique identifier (pseries machines can routinely have dozens of PCI host bridges). Theoretically that breaks migration streams, but given that we don't yet have migration support for pseries, it doesn't matter. Any other machines that have working migration support including PCI devices will need to be updated to maintain migration stream compatibility. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2013-06-06 10:48:49 +02:00
PCIHostBridgeClass *hc = PCI_HOST_BRIDGE_CLASS(klass);
DeviceClass *dc = DEVICE_CLASS(klass);
HotplugHandlerClass *hp = HOTPLUG_HANDLER_CLASS(klass);
pci: Replace pci_find_domain() with more general pci_root_bus_path() pci_find_domain() is used in a number of places where we want an id for a whole PCI domain (i.e. the subtree under a PCI root bus). The trouble is that many platforms may support multiple independent host bridges with no hardware supplied notion of domain number. This patch, therefore, replaces calls to pci_find_domain() with calls to a new pci_root_bus_path() returning a string. The new call is implemented in terms of a new callback in the host bridge class, so it can be defined in some way that's well defined for the platform. When no callback is available we fall back on the qbus name. Most current uses of pci_find_domain() are for error or informational messages, so the change in identifiers should be harmless. The exception is pci_get_dev_path(), whose results form part of migration streams. To maintain compatibility with old migration streams, the PIIX PCI host is altered to always supply "0000" for this path, which matches the old domain number (since the code didn't actually support domains other than 0). For the pseries (spapr) PCI bridge we use a different platform-unique identifier (pseries machines can routinely have dozens of PCI host bridges). Theoretically that breaks migration streams, but given that we don't yet have migration support for pseries, it doesn't matter. Any other machines that have working migration support including PCI devices will need to be updated to maintain migration stream compatibility. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
2013-06-06 10:48:49 +02:00
hc->root_bus_path = spapr_phb_root_bus_path;
dc->realize = spapr_phb_realize;
dc->unrealize = spapr_phb_unrealize;
device_class_set_props(dc, spapr_phb_properties);
dc->reset = spapr_phb_reset;
dc->vmsd = &vmstate_spapr_pci;
sysbus: Set user_creatable=false by default on TYPE_SYS_BUS_DEVICE commit 33cd52b5d7b9adfd009e95f07e6c64dd88ae2a31 unset cannot_instantiate_with_device_add_yet in TYPE_SYSBUS, making all sysbus devices appear on "-device help" and lack the "no-user" flag in "info qdm". To fix this, we can set user_creatable=false by default on TYPE_SYS_BUS_DEVICE, but this requires setting user_creatable=true explicitly on the sysbus devices that actually work with -device. Fortunately today we have just a few has_dynamic_sysbus=1 machines: virt, pc-q35-*, ppce500, and spapr. virt, ppce500, and spapr have extra checks to ensure just a few device types can be instantiated: * virt supports only TYPE_VFIO_CALXEDA_XGMAC, TYPE_VFIO_AMD_XGBE. * ppce500 supports only TYPE_ETSEC_COMMON. * spapr supports only TYPE_SPAPR_PCI_HOST_BRIDGE. This patch sets user_creatable=true explicitly on those 4 device classes. Now, the more complex cases: pc-q35-*: q35 has no sysbus device whitelist yet (which is a separate bug). We are in the process of fixing it and building a sysbus whitelist on q35, but in the meantime we can fix the "-device help" and "info qdm" bugs mentioned above. Also, despite not being strictly necessary for fixing the q35 bug, reducing the list of user_creatable=true devices will help us be more confident when building the q35 whitelist. xen: We also have a hack at xen_set_dynamic_sysbus(), that sets has_dynamic_sysbus=true at runtime when using the Xen accelerator. This hack is only used to allow xen-backend devices to be dynamically plugged/unplugged. This means today we can use -device with the following 22 device types, that are the ones compiled into the qemu-system-x86_64 and qemu-system-i386 binaries: * allwinner-ahci * amd-iommu * cfi.pflash01 * esp * fw_cfg_io * fw_cfg_mem * generic-sdhci * hpet * intel-iommu * ioapic * isabus-bridge * kvmclock * kvm-ioapic * kvmvapic * SUNW,fdtwo * sysbus-ahci * sysbus-fdc * sysbus-ohci * unimplemented-device * virtio-mmio * xen-backend * xen-sysdev This patch adds user_creatable=true explicitly to those devices, temporarily, just to keep 100% compatibility with existing behavior of q35. Subsequent patches will remove user_creatable=true from the devices that are really not meant to user-creatable on any machine, and remove the FIXME comment from the ones that are really supposed to be user-creatable. This is being done in separate patches because we still don't have an obvious list of devices that will be whitelisted by q35, and I would like to get each device reviewed individually. Cc: Alexander Graf <agraf@suse.de> Cc: Alex Williamson <alex.williamson@redhat.com> Cc: Alistair Francis <alistair.francis@xilinx.com> Cc: Beniamino Galvani <b.galvani@gmail.com> Cc: Christian Borntraeger <borntraeger@de.ibm.com> Cc: Cornelia Huck <cornelia.huck@de.ibm.com> Cc: David Gibson <david@gibson.dropbear.id.au> Cc: "Edgar E. Iglesias" <edgar.iglesias@gmail.com> Cc: Eduardo Habkost <ehabkost@redhat.com> Cc: Frank Blaschka <frank.blaschka@de.ibm.com> Cc: Gabriel L. Somlo <somlo@cmu.edu> Cc: Gerd Hoffmann <kraxel@redhat.com> Cc: Igor Mammedov <imammedo@redhat.com> Cc: Jason Wang <jasowang@redhat.com> Cc: John Snow <jsnow@redhat.com> Cc: Juergen Gross <jgross@suse.com> Cc: Kevin Wolf <kwolf@redhat.com> Cc: Laszlo Ersek <lersek@redhat.com> Cc: Marcel Apfelbaum <marcel@redhat.com> Cc: Markus Armbruster <armbru@redhat.com> Cc: Max Reitz <mreitz@redhat.com> Cc: "Michael S. Tsirkin" <mst@redhat.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Pierre Morel <pmorel@linux.vnet.ibm.com> Cc: Prasad J Pandit <pjp@fedoraproject.org> Cc: qemu-arm@nongnu.org Cc: qemu-block@nongnu.org Cc: qemu-ppc@nongnu.org Cc: Richard Henderson <rth@twiddle.net> Cc: Rob Herring <robh@kernel.org> Cc: Shannon Zhao <zhaoshenglong@huawei.com> Cc: sstabellini@kernel.org Cc: Thomas Huth <thuth@redhat.com> Cc: Yi Min Zhao <zyimin@linux.vnet.ibm.com> Acked-by: John Snow <jsnow@redhat.com> Acked-by: Juergen Gross <jgross@suse.com> Acked-by: Marcel Apfelbaum <marcel@redhat.com> Signed-off-by: Eduardo Habkost <ehabkost@redhat.com> Message-Id: <20170503203604.31462-3-ehabkost@redhat.com> Reviewed-by: Markus Armbruster <armbru@redhat.com> [ehabkost: Small changes at sysbus_device_class_init() comments] Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
2017-05-03 22:35:45 +02:00
/* Supported by TYPE_SPAPR_MACHINE */
dc->user_creatable = true;
set_bit(DEVICE_CATEGORY_BRIDGE, dc->categories);
hp->pre_plug = spapr_pci_pre_plug;
hp->plug = spapr_pci_plug;
hp->unplug = spapr_pci_unplug;
hp->unplug_request = spapr_pci_unplug_request;
}
static const TypeInfo spapr_phb_info = {
.name = TYPE_SPAPR_PCI_HOST_BRIDGE,
.parent = TYPE_PCI_HOST_BRIDGE,
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
.instance_size = sizeof(SpaprPhbState),
.instance_finalize = spapr_phb_finalizefn,
.class_init = spapr_phb_class_init,
.interfaces = (InterfaceInfo[]) {
{ TYPE_HOTPLUG_HANDLER },
{ }
}
};
static void spapr_phb_pci_enumerate_bridge(PCIBus *bus, PCIDevice *pdev,
void *opaque)
{
unsigned int *bus_no = opaque;
PCIBus *sec_bus = NULL;
if ((pci_default_read_config(pdev, PCI_HEADER_TYPE, 1) !=
PCI_HEADER_TYPE_BRIDGE)) {
return;
}
(*bus_no)++;
pci_default_write_config(pdev, PCI_PRIMARY_BUS, pci_dev_bus_num(pdev), 1);
pci_default_write_config(pdev, PCI_SECONDARY_BUS, *bus_no, 1);
pci_default_write_config(pdev, PCI_SUBORDINATE_BUS, *bus_no, 1);
sec_bus = pci_bridge_get_sec_bus(PCI_BRIDGE(pdev));
if (!sec_bus) {
return;
}
pci_for_each_device_under_bus(sec_bus, spapr_phb_pci_enumerate_bridge,
bus_no);
pci_default_write_config(pdev, PCI_SUBORDINATE_BUS, *bus_no, 1);
}
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
static void spapr_phb_pci_enumerate(SpaprPhbState *phb)
{
PCIBus *bus = PCI_HOST_BRIDGE(phb)->bus;
unsigned int bus_no = 0;
pci_for_each_device_under_bus(bus, spapr_phb_pci_enumerate_bridge,
&bus_no);
}
int spapr_dt_phb(SpaprMachineState *spapr, SpaprPhbState *phb,
uint32_t intc_phandle, void *fdt, int *node_offset)
{
int bus_off, i, j, ret;
uint32_t bus_range[] = { cpu_to_be32(0), cpu_to_be32(0xff) };
struct {
uint32_t hi;
uint64_t child;
uint64_t parent;
uint64_t size;
} QEMU_PACKED ranges[] = {
{
cpu_to_be32(b_ss(1)), cpu_to_be64(0),
cpu_to_be64(phb->io_win_addr),
cpu_to_be64(memory_region_size(&phb->iospace)),
},
{
cpu_to_be32(b_ss(2)), cpu_to_be64(SPAPR_PCI_MEM_WIN_BUS_OFFSET),
cpu_to_be64(phb->mem_win_addr),
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
cpu_to_be64(phb->mem_win_size),
},
{
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
cpu_to_be32(b_ss(3)), cpu_to_be64(phb->mem64_win_pciaddr),
cpu_to_be64(phb->mem64_win_addr),
cpu_to_be64(phb->mem64_win_size),
},
};
spapr_pci: Add a 64-bit MMIO window On real hardware, and under pHyp, the PCI host bridges on Power machines typically advertise two outbound MMIO windows from the guest's physical memory space to PCI memory space: - A 32-bit window which maps onto 2GiB..4GiB in the PCI address space - A 64-bit window which maps onto a large region somewhere high in PCI address space (traditionally this used an identity mapping from guest physical address to PCI address, but that's not always the case) The qemu implementation in spapr-pci-host-bridge, however, only supports a single outbound MMIO window, however. At least some Linux versions expect the two windows however, so we arranged this window to map onto the PCI memory space from 2 GiB..~64 GiB, then advertised it as two contiguous windows, the "32-bit" window from 2G..4G and the "64-bit" window from 4G..~64G. This approach means, however, that the 64G window is not naturally aligned. In turn this limits the size of the largest BAR we can map (which does have to be naturally aligned) to roughly half of the total window. With some large nVidia GPGPU cards which have huge memory BARs, this is starting to be a problem. This patch adds true support for separate 32-bit and 64-bit outbound MMIO windows to the spapr-pci-host-bridge implementation, each of which can be independently configured. The 32-bit window always maps to 2G.. in PCI space, but the PCI address of the 64-bit window can be configured (it defaults to the same as the guest physical address). So as not to break possible existing configurations, as long as a 64-bit window is not specified, a large single window can be specified. This will appear the same way to the guest as the old approach, although it's now implemented by two contiguous memory regions rather than a single one. For now, this only adds the possibility of 64-bit windows. The default configuration still uses the legacy mode. Signed-off-by: David Gibson <david@gibson.dropbear.id.au> Reviewed-by: Laurent Vivier <lvivier@redhat.com>
2016-10-11 05:23:33 +02:00
const unsigned sizeof_ranges =
(phb->mem64_win_size ? 3 : 2) * sizeof(ranges[0]);
uint64_t bus_reg[] = { cpu_to_be64(phb->buid), 0 };
uint32_t interrupt_map_mask[] = {
cpu_to_be32(b_ddddd(-1)|b_fff(0)), 0x0, 0x0, cpu_to_be32(-1)};
uint32_t interrupt_map[PCI_SLOT_MAX * PCI_NUM_PINS][7];
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
uint32_t ddw_applicable[] = {
cpu_to_be32(RTAS_IBM_QUERY_PE_DMA_WINDOW),
cpu_to_be32(RTAS_IBM_CREATE_PE_DMA_WINDOW),
cpu_to_be32(RTAS_IBM_REMOVE_PE_DMA_WINDOW)
};
uint32_t ddw_extensions[] = {
cpu_to_be32(2),
cpu_to_be32(RTAS_IBM_RESET_PE_DMA_WINDOW),
cpu_to_be32(1), /* 1: ibm,query-pe-dma-window 6 outputs, PAPR 2.8 */
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
};
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprTceTable *tcet;
SpaprDrc *drc;
Error *err = NULL;
/* Start populating the FDT */
_FDT(bus_off = fdt_add_subnode(fdt, 0, phb->dtbusname));
if (node_offset) {
*node_offset = bus_off;
}
/* Write PHB properties */
_FDT(fdt_setprop_string(fdt, bus_off, "device_type", "pci"));
_FDT(fdt_setprop_string(fdt, bus_off, "compatible", "IBM,Logical_PHB"));
_FDT(fdt_setprop_cell(fdt, bus_off, "#interrupt-cells", 0x1));
_FDT(fdt_setprop(fdt, bus_off, "used-by-rtas", NULL, 0));
_FDT(fdt_setprop(fdt, bus_off, "bus-range", &bus_range, sizeof(bus_range)));
_FDT(fdt_setprop(fdt, bus_off, "ranges", &ranges, sizeof_ranges));
_FDT(fdt_setprop(fdt, bus_off, "reg", &bus_reg, sizeof(bus_reg)));
_FDT(fdt_setprop_cell(fdt, bus_off, "ibm,pci-config-space-type", 0x1));
_FDT(fdt_setprop_cell(fdt, bus_off, "ibm,pe-total-#msi",
spapr_irq_nr_msis(spapr)));
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
/* Dynamic DMA window */
if (phb->ddw_enabled) {
_FDT(fdt_setprop(fdt, bus_off, "ibm,ddw-applicable", &ddw_applicable,
sizeof(ddw_applicable)));
_FDT(fdt_setprop(fdt, bus_off, "ibm,ddw-extensions",
&ddw_extensions, sizeof(ddw_extensions)));
}
/* Advertise NUMA via ibm,associativity */
spapr_pci: advertise explicit numa IDs even when there's 1 node With the addition of "numa_node" properties for PHBs we began advertising NUMA affinity in cases where nb_numa_nodes > 1. Since the default on the guest side is to make no assumptions about PHB NUMA affinity (defaulting to -1), there is still a valid use-case for explicitly defining a PHB's NUMA affinity even when there's just one node. In particular, some workloads make faulty assumptions about /sys/bus/pci/<devid>/numa_node being >= 0, warranting the use of this property as a workaround even if there's just 1 PHB or NUMA node. Enable this use-case by always advertising the PHB's NUMA affinity if "numa_node" has been explicitly set. We could achieve this by relaxing the check to simply be nb_numa_nodes > 0, but even safer would be to check numa_info[nodeid].present explicitly, and to fail at start time for cases where it does not exist. This has an additional affect of no longer advertising PHB NUMA affinity unconditionally if nb_numa_nodes > 1 and "numa_node" property is unset/-1, but since the default value on the guest side for each PHB is also -1, the behavior should be the same for that situation. We could still retain the old behavior if desired, but the decision seems arbitrary, so we take the simpler route. Cc: Alexey Kardashevskiy <aik@ozlabs.ru> Cc: Shivaprasad G. Bhat <shivapbh@in.ibm.com> Signed-off-by: Michael Roth <mdroth@linux.vnet.ibm.com> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-10-18 22:50:23 +02:00
if (phb->numa_node != -1) {
spapr_numa_write_associativity_dt(spapr, fdt, bus_off, phb->numa_node);
}
/* Build the interrupt-map, this must matches what is done
* in pci_swizzle_map_irq_fn
*/
_FDT(fdt_setprop(fdt, bus_off, "interrupt-map-mask",
&interrupt_map_mask, sizeof(interrupt_map_mask)));
for (i = 0; i < PCI_SLOT_MAX; i++) {
for (j = 0; j < PCI_NUM_PINS; j++) {
uint32_t *irqmap = interrupt_map[i*PCI_NUM_PINS + j];
int lsi_num = pci_swizzle(i, j);
irqmap[0] = cpu_to_be32(b_ddddd(i)|b_fff(0));
irqmap[1] = 0;
irqmap[2] = 0;
irqmap[3] = cpu_to_be32(j+1);
irqmap[4] = cpu_to_be32(intc_phandle);
spapr_dt_irq(&irqmap[5], phb->lsi_table[lsi_num].irq, true);
}
}
/* Write interrupt map */
_FDT(fdt_setprop(fdt, bus_off, "interrupt-map", &interrupt_map,
sizeof(interrupt_map)));
spapr_pci/spapr_pci_vfio: Support Dynamic DMA Windows (DDW) This adds support for Dynamic DMA Windows (DDW) option defined by the SPAPR specification which allows to have additional DMA window(s) The "ddw" property is enabled by default on a PHB but for compatibility the pseries-2.6 machine and older disable it. This also creates a single DMA window for the older machines to maintain backward migration. This implements DDW for PHB with emulated and VFIO devices. The host kernel support is required. The advertised IOMMU page sizes are 4K and 64K; 16M pages are supported but not advertised by default, in order to enable them, the user has to specify "pgsz" property for PHB and enable huge pages for RAM. The existing linux guests try creating one additional huge DMA window with 64K or 16MB pages and map the entire guest RAM to. If succeeded, 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 later. This adds a "dma64_win_addr" property which is a bus address for the 64bit window and by default set to 0x800.0000.0000.0000 as this is what the modern POWER8 hardware uses and this allows having emulated and VFIO devices on the same bus. 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 PCI. This changes sPAPRPHBState::dma_liobn to an array to allow 2 LIOBNs and updates all references to dma_liobn. However this does not add 64bit LIOBN to the migration stream as in fact even 32bit LIOBN is rather pointless there (as it is a PHB property and the management software can/should pass LIOBNs via CLI) but we keep it for the backward migration support. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2016-07-04 05:33:07 +02:00
tcet = spapr_tce_find_by_liobn(phb->dma_liobn[0]);
if (!tcet) {
return -1;
}
spapr_dma_dt(fdt, bus_off, "ibm,dma-window",
tcet->liobn, tcet->bus_offset,
tcet->nb_table << tcet->page_shift);
drc = spapr_drc_by_id(TYPE_SPAPR_DRC_PHB, phb->index);
if (drc) {
uint32_t drc_index = cpu_to_be32(spapr_drc_index(drc));
_FDT(fdt_setprop(fdt, bus_off, "ibm,my-drc-index", &drc_index,
sizeof(drc_index)));
}
/* Walk the bridges and program the bus numbers*/
spapr_phb_pci_enumerate(phb);
_FDT(fdt_setprop_cell(fdt, bus_off, "qemu,phb-enumerated", 0x1));
/* Walk the bridge and subordinate buses */
ret = spapr_dt_pci_bus(phb, PCI_HOST_BRIDGE(phb)->bus, fdt, bus_off);
if (ret < 0) {
return ret;
}
spapr_phb_nvgpu_populate_dt(phb, fdt, bus_off, &err);
if (err) {
error_report_err(err);
spapr: Support NVIDIA V100 GPU with NVLink2 NVIDIA V100 GPUs have on-board RAM which is mapped into the host memory space and accessible as normal RAM via an NVLink bus. The VFIO-PCI driver implements special regions for such GPUs and emulates an NVLink bridge. NVLink2-enabled POWER9 CPUs also provide address translation services which includes an ATS shootdown (ATSD) register exported via the NVLink bridge device. This adds a quirk to VFIO to map the GPU memory and create an MR; the new MR is stored in a PCI device as a QOM link. The sPAPR PCI uses this to get the MR and map it to the system address space. Another quirk does the same for ATSD. This adds additional steps to sPAPR PHB setup: 1. Search for specific GPUs and NPUs, collect findings in sPAPRPHBState::nvgpus, manage system address space mappings; 2. Add device-specific properties such as "ibm,npu", "ibm,gpu", "memory-block", "link-speed" to advertise the NVLink2 function to the guest; 3. Add "mmio-atsd" to vPHB to advertise the ATSD capability; 4. Add new memory blocks (with extra "linux,memory-usable" to prevent the guest OS from accessing the new memory until it is onlined) and npuphb# nodes representing an NPU unit for every vPHB as the GPU driver uses it for link discovery. This allocates space for GPU RAM and ATSD like we do for MMIOs by adding 2 new parameters to the phb_placement() hook. Older machine types set these to zero. This puts new memory nodes in a separate NUMA node to as the GPU RAM needs to be configured equally distant from any other node in the system. Unlike the host setup which assigns numa ids from 255 downwards, this adds new NUMA nodes after the user configures nodes or from 1 if none were configured. This adds requirement similar to EEH - one IOMMU group per vPHB. The reason for this is that ATSD registers belong to a physical NPU so they cannot invalidate translations on GPUs attached to another NPU. It is guaranteed by the host platform as it does not mix NVLink bridges or GPUs from different NPU in the same IOMMU group. If more than one IOMMU group is detected on a vPHB, this disables ATSD support for that vPHB and prints a warning. Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru> [aw: for vfio portions] Acked-by: Alex Williamson <alex.williamson@redhat.com> Message-Id: <20190312082103.130561-1-aik@ozlabs.ru> Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-12 09:21:03 +01:00
}
spapr_phb_nvgpu_ram_populate_dt(phb, fdt);
return 0;
}
void spapr_pci_rtas_init(void)
{
spapr_rtas_register(RTAS_READ_PCI_CONFIG, "read-pci-config",
rtas_read_pci_config);
spapr_rtas_register(RTAS_WRITE_PCI_CONFIG, "write-pci-config",
rtas_write_pci_config);
spapr_rtas_register(RTAS_IBM_READ_PCI_CONFIG, "ibm,read-pci-config",
rtas_ibm_read_pci_config);
spapr_rtas_register(RTAS_IBM_WRITE_PCI_CONFIG, "ibm,write-pci-config",
rtas_ibm_write_pci_config);
if (msi_nonbroken) {
spapr_rtas_register(RTAS_IBM_QUERY_INTERRUPT_SOURCE_NUMBER,
"ibm,query-interrupt-source-number",
rtas_ibm_query_interrupt_source_number);
spapr_rtas_register(RTAS_IBM_CHANGE_MSI, "ibm,change-msi",
rtas_ibm_change_msi);
}
spapr_rtas_register(RTAS_IBM_SET_EEH_OPTION,
"ibm,set-eeh-option",
rtas_ibm_set_eeh_option);
spapr_rtas_register(RTAS_IBM_GET_CONFIG_ADDR_INFO2,
"ibm,get-config-addr-info2",
rtas_ibm_get_config_addr_info2);
spapr_rtas_register(RTAS_IBM_READ_SLOT_RESET_STATE2,
"ibm,read-slot-reset-state2",
rtas_ibm_read_slot_reset_state2);
spapr_rtas_register(RTAS_IBM_SET_SLOT_RESET,
"ibm,set-slot-reset",
rtas_ibm_set_slot_reset);
spapr_rtas_register(RTAS_IBM_CONFIGURE_PE,
"ibm,configure-pe",
rtas_ibm_configure_pe);
spapr_rtas_register(RTAS_IBM_SLOT_ERROR_DETAIL,
"ibm,slot-error-detail",
rtas_ibm_slot_error_detail);
}
static void spapr_pci_register_types(void)
{
type_register_static(&spapr_phb_info);
}
type_init(spapr_pci_register_types)
static int spapr_switch_one_vga(DeviceState *dev, void *opaque)
{
bool be = *(bool *)opaque;
if (object_dynamic_cast(OBJECT(dev), "VGA")
|| object_dynamic_cast(OBJECT(dev), "secondary-vga")
|| object_dynamic_cast(OBJECT(dev), "bochs-display")
|| object_dynamic_cast(OBJECT(dev), "virtio-vga")) {
qom: Put name parameter before value / visitor parameter The object_property_set_FOO() setters take property name and value in an unusual order: void object_property_set_FOO(Object *obj, FOO_TYPE value, const char *name, Error **errp) Having to pass value before name feels grating. Swap them. Same for object_property_set(), object_property_get(), and object_property_parse(). Convert callers with this Coccinelle script: @@ identifier fun = { object_property_get, object_property_parse, object_property_set_str, object_property_set_link, object_property_set_bool, object_property_set_int, object_property_set_uint, object_property_set, object_property_set_qobject }; expression obj, v, name, errp; @@ - fun(obj, v, name, errp) + fun(obj, name, v, errp) Chokes on hw/arm/musicpal.c's lcd_refresh() with the unhelpful error message "no position information". Convert that one manually. Fails to convert hw/arm/armsse.c, because Coccinelle gets confused by ARMSSE being used both as typedef and function-like macro there. Convert manually. Fails to convert hw/rx/rx-gdbsim.c, because Coccinelle gets confused by RXCPU being used both as typedef and function-like macro there. Convert manually. The other files using RXCPU that way don't need conversion. Signed-off-by: Markus Armbruster <armbru@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com> Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com> Message-Id: <20200707160613.848843-27-armbru@redhat.com> [Straightforwad conflict with commit 2336172d9b "audio: set default value for pcspk.iobase property" resolved]
2020-07-07 18:05:54 +02:00
object_property_set_bool(OBJECT(dev), "big-endian-framebuffer", be,
&error_abort);
}
return 0;
}
void spapr_pci_switch_vga(SpaprMachineState *spapr, bool big_endian)
{
spapr: Use CamelCase properly The qemu coding standard is to use CamelCase for type and structure names, and the pseries code follows that... sort of. There are quite a lot of places where we bend the rules in order to preserve the capitalization of internal acronyms like "PHB", "TCE", "DIMM" and most commonly "sPAPR". That was a bad idea - it frequently leads to names ending up with hard to read clusters of capital letters, and means they don't catch the eye as type identifiers, which is kind of the point of the CamelCase convention in the first place. In short, keeping type identifiers look like CamelCase is more important than preserving standard capitalization of internal "words". So, this patch renames a heap of spapr internal type names to a more standard CamelCase. In addition to case changes, we also make some other identifier renames: VIOsPAPR* -> SpaprVio* The reverse word ordering was only ever used to mitigate the capital cluster, so revert to the natural ordering. VIOsPAPRVTYDevice -> SpaprVioVty VIOsPAPRVLANDevice -> SpaprVioVlan Brevity, since the "Device" didn't add useful information sPAPRDRConnector -> SpaprDrc sPAPRDRConnectorClass -> SpaprDrcClass Brevity, and makes it clearer this is the same thing as a "DRC" mentioned in many other places in the code This is 100% a mechanical search-and-replace patch. It will, however, conflict with essentially any and all outstanding patches touching the spapr code. Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
2019-03-06 05:35:37 +01:00
SpaprPhbState *sphb;
/*
* For backward compatibility with existing guests, we switch
* the endianness of the VGA controller when changing the guest
* interrupt mode
*/
QLIST_FOREACH(sphb, &spapr->phbs, list) {
BusState *bus = &PCI_HOST_BRIDGE(sphb)->bus->qbus;
qbus_walk_children(bus, spapr_switch_one_vga, NULL, NULL, NULL,
&big_endian);
}
}