Implement the PAPR (pSeries) virtualized interrupt controller (xics)
PAPR defines an interrupt control architecture which is logically divided
into ICS (Interrupt Control Presentation, each unit is responsible for
presenting interrupts to a particular "interrupt server", i.e. CPU) and
ICS (Interrupt Control Source, each unit responsible for one or more
hardware interrupts as numbered globally across the system). All PAPR
virtual IO devices expect to deliver interrupts via this mechanism. In
Linux, this interrupt controller system is handled by the "xics" driver.
On pSeries systems, access to the interrupt controller is virtualized via
hypercalls and RTAS methods. However, the virtualized interface is very
similar to the underlying interrupt controller hardware, and similar PICs
exist un-virtualized in some other systems.
This patch implements both the ICP and ICS sides of the PAPR interrupt
controller. For now, only the hypercall virtualized interface is provided,
however it would be relatively straightforward to graft an emulated
register interface onto the underlying interrupt logic if we want to add
a machine with a hardware ICS/ICP system in the future.
There are some limitations in this implementation: it is assumed for now
that only one instance of the ICS exists, although a full xics system can
have several, each responsible for a different group of hardware irqs.
ICP/ICS can handle both level-sensitve (LSI) and message signalled (MSI)
interrupt inputs. For now, this implementation supports only MSI
interrupts, since that is used by PAPR virtual IO devices.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-04-01 08:15:25 +04:00
|
|
|
/*
|
|
|
|
* QEMU PowerPC pSeries Logical Partition (aka sPAPR) hardware System Emulator
|
|
|
|
*
|
|
|
|
* PAPR Virtualized Interrupt System, aka ICS/ICP aka xics
|
|
|
|
*
|
|
|
|
* Copyright (c) 2010,2011 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.
|
|
|
|
*
|
|
|
|
*/
|
2016-06-29 14:47:03 +03:00
|
|
|
|
|
|
|
#ifndef XICS_H
|
|
|
|
#define XICS_H
|
Implement the PAPR (pSeries) virtualized interrupt controller (xics)
PAPR defines an interrupt control architecture which is logically divided
into ICS (Interrupt Control Presentation, each unit is responsible for
presenting interrupts to a particular "interrupt server", i.e. CPU) and
ICS (Interrupt Control Source, each unit responsible for one or more
hardware interrupts as numbered globally across the system). All PAPR
virtual IO devices expect to deliver interrupts via this mechanism. In
Linux, this interrupt controller system is handled by the "xics" driver.
On pSeries systems, access to the interrupt controller is virtualized via
hypercalls and RTAS methods. However, the virtualized interface is very
similar to the underlying interrupt controller hardware, and similar PICs
exist un-virtualized in some other systems.
This patch implements both the ICP and ICS sides of the PAPR interrupt
controller. For now, only the hypercall virtualized interface is provided,
however it would be relatively straightforward to graft an emulated
register interface onto the underlying interrupt logic if we want to add
a machine with a hardware ICS/ICP system in the future.
There are some limitations in this implementation: it is assumed for now
that only one instance of the ICS exists, although a full xics system can
have several, each responsible for a different group of hardware irqs.
ICP/ICS can handle both level-sensitve (LSI) and message signalled (MSI)
interrupt inputs. For now, this implementation supports only MSI
interrupts, since that is used by PAPR virtual IO devices.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-04-01 08:15:25 +04:00
|
|
|
|
2017-03-07 12:23:40 +03:00
|
|
|
#include "hw/qdev.h"
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
|
Implement the PAPR (pSeries) virtualized interrupt controller (xics)
PAPR defines an interrupt control architecture which is logically divided
into ICS (Interrupt Control Presentation, each unit is responsible for
presenting interrupts to a particular "interrupt server", i.e. CPU) and
ICS (Interrupt Control Source, each unit responsible for one or more
hardware interrupts as numbered globally across the system). All PAPR
virtual IO devices expect to deliver interrupts via this mechanism. In
Linux, this interrupt controller system is handled by the "xics" driver.
On pSeries systems, access to the interrupt controller is virtualized via
hypercalls and RTAS methods. However, the virtualized interface is very
similar to the underlying interrupt controller hardware, and similar PICs
exist un-virtualized in some other systems.
This patch implements both the ICP and ICS sides of the PAPR interrupt
controller. For now, only the hypercall virtualized interface is provided,
however it would be relatively straightforward to graft an emulated
register interface onto the underlying interrupt logic if we want to add
a machine with a hardware ICS/ICP system in the future.
There are some limitations in this implementation: it is assumed for now
that only one instance of the ICS exists, although a full xics system can
have several, each responsible for a different group of hardware irqs.
ICP/ICS can handle both level-sensitve (LSI) and message signalled (MSI)
interrupt inputs. For now, this implementation supports only MSI
interrupts, since that is used by PAPR virtual IO devices.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-04-01 08:15:25 +04:00
|
|
|
#define XICS_IPI 0x2
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
#define XICS_BUID 0x1
|
|
|
|
#define XICS_IRQ_BASE (XICS_BUID << 12)
|
|
|
|
|
|
|
|
/*
|
|
|
|
* We currently only support one BUID which is our interrupt base
|
|
|
|
* (the kernel implementation supports more but we don't exploit
|
|
|
|
* that yet)
|
|
|
|
*/
|
2013-09-26 10:18:39 +04:00
|
|
|
typedef struct ICPStateClass ICPStateClass;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
typedef struct ICPState ICPState;
|
2017-04-03 10:46:00 +03:00
|
|
|
typedef struct PnvICPState PnvICPState;
|
2013-09-26 10:18:39 +04:00
|
|
|
typedef struct ICSStateClass ICSStateClass;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
typedef struct ICSState ICSState;
|
|
|
|
typedef struct ICSIRQState ICSIRQState;
|
2017-02-27 17:29:17 +03:00
|
|
|
typedef struct XICSFabric XICSFabric;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
|
|
|
|
#define TYPE_ICP "icp"
|
|
|
|
#define ICP(obj) OBJECT_CHECK(ICPState, (obj), TYPE_ICP)
|
|
|
|
|
2017-04-03 10:46:00 +03:00
|
|
|
#define TYPE_PNV_ICP "pnv-icp"
|
|
|
|
#define PNV_ICP(obj) OBJECT_CHECK(PnvICPState, (obj), TYPE_PNV_ICP)
|
|
|
|
|
2013-09-26 10:18:39 +04:00
|
|
|
#define ICP_CLASS(klass) \
|
|
|
|
OBJECT_CLASS_CHECK(ICPStateClass, (klass), TYPE_ICP)
|
|
|
|
#define ICP_GET_CLASS(obj) \
|
|
|
|
OBJECT_GET_CLASS(ICPStateClass, (obj), TYPE_ICP)
|
|
|
|
|
|
|
|
struct ICPStateClass {
|
|
|
|
DeviceClass parent_class;
|
|
|
|
|
2018-06-20 13:24:13 +03:00
|
|
|
DeviceRealize parent_realize;
|
2013-09-26 10:18:39 +04:00
|
|
|
};
|
|
|
|
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
struct ICPState {
|
|
|
|
/*< private >*/
|
|
|
|
DeviceState parent_obj;
|
|
|
|
/*< public >*/
|
2013-09-26 10:18:44 +04:00
|
|
|
CPUState *cs;
|
2016-10-03 10:24:46 +03:00
|
|
|
ICSState *xirr_owner;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
uint32_t xirr;
|
|
|
|
uint8_t pending_priority;
|
|
|
|
uint8_t mfrr;
|
|
|
|
qemu_irq output;
|
2016-10-18 11:06:33 +03:00
|
|
|
|
2017-02-27 17:29:17 +03:00
|
|
|
XICSFabric *xics;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
};
|
|
|
|
|
2017-06-08 16:42:41 +03:00
|
|
|
#define ICP_PROP_XICS "xics"
|
2017-06-08 16:42:59 +03:00
|
|
|
#define ICP_PROP_CPU "cpu"
|
2017-06-08 16:42:41 +03:00
|
|
|
|
2017-04-03 10:46:00 +03:00
|
|
|
struct PnvICPState {
|
|
|
|
ICPState parent_obj;
|
|
|
|
|
|
|
|
MemoryRegion mmio;
|
|
|
|
uint32_t links[3];
|
|
|
|
};
|
|
|
|
|
2016-10-03 10:24:47 +03:00
|
|
|
#define TYPE_ICS_BASE "ics-base"
|
|
|
|
#define ICS_BASE(obj) OBJECT_CHECK(ICSState, (obj), TYPE_ICS_BASE)
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
|
2016-10-03 10:24:47 +03:00
|
|
|
/* Retain ics for sPAPR for migration from existing sPAPR guests */
|
|
|
|
#define TYPE_ICS_SIMPLE "ics"
|
|
|
|
#define ICS_SIMPLE(obj) OBJECT_CHECK(ICSState, (obj), TYPE_ICS_SIMPLE)
|
2013-09-26 10:18:44 +04:00
|
|
|
|
2016-10-03 10:24:47 +03:00
|
|
|
#define ICS_BASE_CLASS(klass) \
|
|
|
|
OBJECT_CLASS_CHECK(ICSStateClass, (klass), TYPE_ICS_BASE)
|
|
|
|
#define ICS_BASE_GET_CLASS(obj) \
|
|
|
|
OBJECT_GET_CLASS(ICSStateClass, (obj), TYPE_ICS_BASE)
|
2013-09-26 10:18:39 +04:00
|
|
|
|
|
|
|
struct ICSStateClass {
|
|
|
|
DeviceClass parent_class;
|
|
|
|
|
2018-06-25 12:17:14 +03:00
|
|
|
DeviceRealize parent_realize;
|
2018-06-25 12:17:16 +03:00
|
|
|
DeviceReset parent_reset;
|
2018-06-25 12:17:14 +03:00
|
|
|
|
2016-10-03 10:24:47 +03:00
|
|
|
void (*reject)(ICSState *s, uint32_t irq);
|
|
|
|
void (*resend)(ICSState *s);
|
|
|
|
void (*eoi)(ICSState *s, uint32_t irq);
|
2013-09-26 10:18:39 +04:00
|
|
|
};
|
|
|
|
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
struct ICSState {
|
|
|
|
/*< private >*/
|
|
|
|
DeviceState parent_obj;
|
|
|
|
/*< public >*/
|
|
|
|
uint32_t nr_irqs;
|
|
|
|
uint32_t offset;
|
|
|
|
ICSIRQState *irqs;
|
2017-02-27 17:29:25 +03:00
|
|
|
XICSFabric *xics;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
};
|
Implement the PAPR (pSeries) virtualized interrupt controller (xics)
PAPR defines an interrupt control architecture which is logically divided
into ICS (Interrupt Control Presentation, each unit is responsible for
presenting interrupts to a particular "interrupt server", i.e. CPU) and
ICS (Interrupt Control Source, each unit responsible for one or more
hardware interrupts as numbered globally across the system). All PAPR
virtual IO devices expect to deliver interrupts via this mechanism. In
Linux, this interrupt controller system is handled by the "xics" driver.
On pSeries systems, access to the interrupt controller is virtualized via
hypercalls and RTAS methods. However, the virtualized interface is very
similar to the underlying interrupt controller hardware, and similar PICs
exist un-virtualized in some other systems.
This patch implements both the ICP and ICS sides of the PAPR interrupt
controller. For now, only the hypercall virtualized interface is provided,
however it would be relatively straightforward to graft an emulated
register interface onto the underlying interrupt logic if we want to add
a machine with a hardware ICS/ICP system in the future.
There are some limitations in this implementation: it is assumed for now
that only one instance of the ICS exists, although a full xics system can
have several, each responsible for a different group of hardware irqs.
ICP/ICS can handle both level-sensitve (LSI) and message signalled (MSI)
interrupt inputs. For now, this implementation supports only MSI
interrupts, since that is used by PAPR virtual IO devices.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-04-01 08:15:25 +04:00
|
|
|
|
2017-06-08 16:42:41 +03:00
|
|
|
#define ICS_PROP_XICS "xics"
|
|
|
|
|
2016-06-28 22:05:13 +03:00
|
|
|
static inline bool ics_valid_irq(ICSState *ics, uint32_t nr)
|
|
|
|
{
|
2019-01-02 08:57:41 +03:00
|
|
|
return (nr >= ics->offset) && (nr < (ics->offset + ics->nr_irqs));
|
2016-06-28 22:05:13 +03:00
|
|
|
}
|
|
|
|
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
struct ICSIRQState {
|
|
|
|
uint32_t server;
|
|
|
|
uint8_t priority;
|
|
|
|
uint8_t saved_priority;
|
|
|
|
#define XICS_STATUS_ASSERTED 0x1
|
|
|
|
#define XICS_STATUS_SENT 0x2
|
|
|
|
#define XICS_STATUS_REJECTED 0x4
|
|
|
|
#define XICS_STATUS_MASKED_PENDING 0x8
|
2017-04-27 09:32:03 +03:00
|
|
|
#define XICS_STATUS_PRESENTED 0x10
|
|
|
|
#define XICS_STATUS_QUEUED 0x20
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
uint8_t status;
|
2014-05-30 13:34:12 +04:00
|
|
|
/* (flags & XICS_FLAGS_IRQ_MASK) == 0 means the interrupt is not allocated */
|
|
|
|
#define XICS_FLAGS_IRQ_LSI 0x1
|
|
|
|
#define XICS_FLAGS_IRQ_MSI 0x2
|
|
|
|
#define XICS_FLAGS_IRQ_MASK 0x3
|
|
|
|
uint8_t flags;
|
xics: rename types to be sane and follow coding style
Basically, in HW the layout of the interrupt network is:
- One ICP per processor thread (the "presenter"). This contains the
registers to fetch a pending interrupt (ack), EOI, and control the
processor priority.
- One ICS per logical source of interrupts (ie, one per PCI host
bridge, and a few others here or there). This contains the per-interrupt
source configuration (target processor(s), priority, mask) and the
per-interrupt internal state.
Under PAPR, there is a single "virtual" ICS ... somewhat (it's a bit
oddball what pHyp does here, arguably there are two but we can ignore
that distinction). There is no register level access. A pair of firmware
(RTAS) calls is used to configure each virtual interrupt.
So our model here is somewhat the same. We have one ICS in the emulated
XICS which arguably *is* the emulated XICS, there's no point making it a
separate "device", that would just be gross, and each VCPU has an
associated ICP.
Yet we call the "XICS" struct icp_state and then the ICPs
'struct icp_server_state'. It's particularly confusing when all of the
functions have xics_prefixes yet take *icp arguments.
Rename:
struct icp_state -> XICSState
struct icp_server_state -> ICPState
struct ics_state -> ICSState
struct ics_irq_state -> ICSIRQState
Signed-off-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Message-id: 1374175984-8930-12-git-send-email-aliguori@us.ibm.com
[aik: added ics_resend() on post_load]
Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
2013-07-18 23:33:04 +04:00
|
|
|
};
|
Implement the PAPR (pSeries) virtualized interrupt controller (xics)
PAPR defines an interrupt control architecture which is logically divided
into ICS (Interrupt Control Presentation, each unit is responsible for
presenting interrupts to a particular "interrupt server", i.e. CPU) and
ICS (Interrupt Control Source, each unit responsible for one or more
hardware interrupts as numbered globally across the system). All PAPR
virtual IO devices expect to deliver interrupts via this mechanism. In
Linux, this interrupt controller system is handled by the "xics" driver.
On pSeries systems, access to the interrupt controller is virtualized via
hypercalls and RTAS methods. However, the virtualized interface is very
similar to the underlying interrupt controller hardware, and similar PICs
exist un-virtualized in some other systems.
This patch implements both the ICP and ICS sides of the PAPR interrupt
controller. For now, only the hypercall virtualized interface is provided,
however it would be relatively straightforward to graft an emulated
register interface onto the underlying interrupt logic if we want to add
a machine with a hardware ICS/ICP system in the future.
There are some limitations in this implementation: it is assumed for now
that only one instance of the ICS exists, although a full xics system can
have several, each responsible for a different group of hardware irqs.
ICP/ICS can handle both level-sensitve (LSI) and message signalled (MSI)
interrupt inputs. For now, this implementation supports only MSI
interrupts, since that is used by PAPR virtual IO devices.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-04-01 08:15:25 +04:00
|
|
|
|
2017-03-03 19:28:50 +03:00
|
|
|
struct XICSFabric {
|
2017-02-27 17:29:14 +03:00
|
|
|
Object parent;
|
2017-03-03 19:28:50 +03:00
|
|
|
};
|
2017-02-27 17:29:14 +03:00
|
|
|
|
|
|
|
#define TYPE_XICS_FABRIC "xics-fabric"
|
|
|
|
#define XICS_FABRIC(obj) \
|
|
|
|
OBJECT_CHECK(XICSFabric, (obj), TYPE_XICS_FABRIC)
|
|
|
|
#define XICS_FABRIC_CLASS(klass) \
|
|
|
|
OBJECT_CLASS_CHECK(XICSFabricClass, (klass), TYPE_XICS_FABRIC)
|
|
|
|
#define XICS_FABRIC_GET_CLASS(obj) \
|
|
|
|
OBJECT_GET_CLASS(XICSFabricClass, (obj), TYPE_XICS_FABRIC)
|
|
|
|
|
|
|
|
typedef struct XICSFabricClass {
|
|
|
|
InterfaceClass parent;
|
|
|
|
ICSState *(*ics_get)(XICSFabric *xi, int irq);
|
|
|
|
void (*ics_resend)(XICSFabric *xi);
|
2017-02-27 17:29:21 +03:00
|
|
|
ICPState *(*icp_get)(XICSFabric *xi, int server);
|
2017-02-27 17:29:14 +03:00
|
|
|
} XICSFabricClass;
|
|
|
|
|
2017-02-27 17:29:25 +03:00
|
|
|
ICPState *xics_icp_get(XICSFabric *xi, int server);
|
Implement the PAPR (pSeries) virtualized interrupt controller (xics)
PAPR defines an interrupt control architecture which is logically divided
into ICS (Interrupt Control Presentation, each unit is responsible for
presenting interrupts to a particular "interrupt server", i.e. CPU) and
ICS (Interrupt Control Source, each unit responsible for one or more
hardware interrupts as numbered globally across the system). All PAPR
virtual IO devices expect to deliver interrupts via this mechanism. In
Linux, this interrupt controller system is handled by the "xics" driver.
On pSeries systems, access to the interrupt controller is virtualized via
hypercalls and RTAS methods. However, the virtualized interface is very
similar to the underlying interrupt controller hardware, and similar PICs
exist un-virtualized in some other systems.
This patch implements both the ICP and ICS sides of the PAPR interrupt
controller. For now, only the hypercall virtualized interface is provided,
however it would be relatively straightforward to graft an emulated
register interface onto the underlying interrupt logic if we want to add
a machine with a hardware ICS/ICP system in the future.
There are some limitations in this implementation: it is assumed for now
that only one instance of the ICS exists, although a full xics system can
have several, each responsible for a different group of hardware irqs.
ICP/ICS can handle both level-sensitve (LSI) and message signalled (MSI)
interrupt inputs. For now, this implementation supports only MSI
interrupts, since that is used by PAPR virtual IO devices.
Signed-off-by: Paul Mackerras <paulus@samba.org>
Signed-off-by: David Gibson <dwg@au1.ibm.com>
Signed-off-by: Alexander Graf <agraf@suse.de>
2011-04-01 08:15:25 +04:00
|
|
|
|
2016-06-28 22:05:13 +03:00
|
|
|
/* Internal XICS interfaces */
|
2016-10-18 11:06:34 +03:00
|
|
|
void icp_set_cppr(ICPState *icp, uint8_t cppr);
|
|
|
|
void icp_set_mfrr(ICPState *icp, uint8_t mfrr);
|
2016-06-28 22:05:13 +03:00
|
|
|
uint32_t icp_accept(ICPState *ss);
|
2016-06-28 22:05:14 +03:00
|
|
|
uint32_t icp_ipoll(ICPState *ss, uint32_t *mfrr);
|
2016-10-18 11:06:34 +03:00
|
|
|
void icp_eoi(ICPState *icp, uint32_t xirr);
|
2016-06-28 22:05:13 +03:00
|
|
|
|
2016-10-03 10:24:47 +03:00
|
|
|
void ics_simple_write_xive(ICSState *ics, int nr, int server,
|
|
|
|
uint8_t priority, uint8_t saved_priority);
|
2019-01-02 08:57:38 +03:00
|
|
|
void ics_simple_set_irq(void *opaque, int srcno, int val);
|
2016-06-28 22:05:13 +03:00
|
|
|
|
|
|
|
void ics_set_irq_type(ICSState *ics, int srcno, bool lsi);
|
2017-02-27 17:29:32 +03:00
|
|
|
void icp_pic_print_info(ICPState *icp, Monitor *mon);
|
|
|
|
void ics_pic_print_info(ICSState *ics, Monitor *mon);
|
2016-06-28 22:05:13 +03:00
|
|
|
|
2017-02-27 17:29:15 +03:00
|
|
|
void ics_resend(ICSState *ics);
|
2017-02-27 17:29:21 +03:00
|
|
|
void icp_resend(ICPState *ss);
|
2016-06-28 22:05:13 +03:00
|
|
|
|
2017-12-01 19:06:00 +03:00
|
|
|
Object *icp_create(Object *cpu, const char *type, XICSFabric *xi,
|
|
|
|
Error **errp);
|
|
|
|
|
2019-02-15 14:39:48 +03:00
|
|
|
/* KVM */
|
|
|
|
void icp_get_kvm_state(ICPState *icp);
|
|
|
|
int icp_set_kvm_state(ICPState *icp);
|
|
|
|
void icp_synchronize_state(ICPState *icp);
|
2019-02-15 14:40:00 +03:00
|
|
|
void icp_kvm_realize(DeviceState *dev, Error **errp);
|
2019-02-15 14:39:48 +03:00
|
|
|
|
2019-02-15 14:40:18 +03:00
|
|
|
void ics_get_kvm_state(ICSState *ics);
|
2019-02-19 20:18:03 +03:00
|
|
|
int ics_set_kvm_state_one(ICSState *ics, int srcno);
|
2019-02-15 14:40:18 +03:00
|
|
|
int ics_set_kvm_state(ICSState *ics);
|
|
|
|
void ics_synchronize_state(ICSState *ics);
|
2019-02-15 14:40:30 +03:00
|
|
|
void ics_kvm_set_irq(ICSState *ics, int srcno, int val);
|
2019-02-15 14:40:18 +03:00
|
|
|
|
2016-06-29 14:47:03 +03:00
|
|
|
#endif /* XICS_H */
|