jcs.org
qemu with hax to log dma reads & writes
jcs.org/2018/11/12/vfio
docs: update ACPI CPU hotplug spec with new protocol
Add description of new CPU hotplug interface.
To switch from from legacy mode into new mode use fact
that write accesses into CPU present bitmap were never
used before and were ignored by QEMU.
So use it to as a way to switch from legacy mode.
That way pc/q35 machine starts in legacy mode and
QEMU generated ACPI tables will switch to new CPU
hotplug interface during runtime.
In case QEMU is started with legacy BIOS (that doesn't
support QEMU generated ACPI tables), legacy CPU hotplug
will remain active and could be used by BIOS built in
ACPI tables for CPU hotplug.
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
authored by
Igor Mammedov
and committed by
Michael S. Tsirkin
abd49bc2
aa1dd39c
+82
-12
+82
-12
docs/specs/acpi_cpu_hotplug.txt
+82
-12
docs/specs/acpi_cpu_hotplug.txt
···
4
4
QEMU supports CPU hotplug via ACPI. This document
5
5
describes the interface between QEMU and the ACPI BIOS.
6
6
7
-
ACPI GPE block (IO ports 0xafe0-0xafe3, byte access):
8
-
-----------------------------------------
9
-
10
-
Generic ACPI GPE block. Bit 2 (GPE.2) used to notify CPU
11
-
hot-add/remove event to ACPI BIOS, via SCI interrupt.
7
+
ACPI BIOS GPE.2 handler is dedicated for notifying OS about CPU hot-add
8
+
and hot-remove events.
12
9
10
+
============================================
11
+
Legacy ACPI CPU hotplug interface registers:
12
+
--------------------------------------------
13
13
CPU present bitmap for:
14
14
ICH9-LPC (IO port 0x0cd8-0xcf7, 1-byte access)
15
15
PIIX-PM (IO port 0xaf00-0xaf1f, 1-byte access)
16
+
One bit per CPU. Bit position reflects corresponding CPU APIC ID. Read-only.
17
+
The first DWORD in bitmap is used in write mode to switch from legacy
18
+
to new CPU hotplug interface, write 0 into it to do switch.
16
19
---------------------------------------------------------------
17
-
One bit per CPU. Bit position reflects corresponding CPU APIC ID.
18
-
Read-only.
20
+
QEMU sets corresponding CPU bit on hot-add event and issues SCI
21
+
with GPE.2 event set. CPU present map is read by ACPI BIOS GPE.2 handler
22
+
to notify OS about CPU hot-add events. CPU hot-remove isn't supported.
23
+
24
+
=====================================
25
+
ACPI CPU hotplug interface registers:
26
+
-------------------------------------
27
+
Register block base address:
28
+
ICH9-LPC IO port 0x0cd8
29
+
PIIX-PM IO port 0xaf00
30
+
Register block size:
31
+
ACPI_CPU_HOTPLUG_REG_LEN = 12
19
32
20
-
CPU hot-add/remove notification:
21
-
-----------------------------------------------------
22
-
QEMU sets/clears corresponding CPU bit on hot-add/remove event.
23
-
CPU present map read by ACPI BIOS GPE.2 handler to notify OS of CPU
24
-
hot-(un)plug events.
33
+
read access:
34
+
offset:
35
+
[0x0-0x3] reserved
36
+
[0x4] CPU device status fields: (1 byte access)
37
+
bits:
38
+
0: Device is enabled and may be used by guest
39
+
1: Device insert event, used to distinguish device for which
40
+
no device check event to OSPM was issued.
41
+
It's valid only when bit 0 is set.
42
+
2: Device remove event, used to distinguish device for which
43
+
no device eject request to OSPM was issued.
44
+
3-7: reserved and should be ignored by OSPM
45
+
[0x5-0x7] reserved
46
+
[0x8] Command data: (DWORD access)
47
+
in case of error or unsupported command reads is 0xFFFFFFFF
48
+
current 'Command field' value:
49
+
0: returns PXM value corresponding to device
50
+
51
+
write access:
52
+
offset:
53
+
[0x0-0x3] CPU selector: (DWORD access)
54
+
selects active CPU device. All following accesses to other
55
+
registers will read/store data from/to selected CPU.
56
+
[0x4] CPU device control fields: (1 byte access)
57
+
bits:
58
+
0: reserved, OSPM must clear it before writing to register.
59
+
1: if set to 1 clears device insert event, set by OSPM
60
+
after it has emitted device check event for the
61
+
selected CPU device
62
+
2: if set to 1 clears device remove event, set by OSPM
63
+
after it has emitted device eject request for the
64
+
selected CPU device
65
+
3: if set to 1 initiates device eject, set by OSPM when it
66
+
triggers CPU device removal and calls _EJ0 method
67
+
4-7: reserved, OSPM must clear them before writing to register
68
+
[0x5] Command field: (1 byte access)
69
+
value:
70
+
0: selects a CPU device with inserting/removing events and
71
+
following reads from 'Command data' register return
72
+
selected CPU (CPU selector value). If no CPU with events
73
+
found, the current CPU selector doesn't change and
74
+
corresponding insert/remove event flags are not set.
75
+
1: following writes to 'Command data' register set OST event
76
+
register in QEMU
77
+
2: following writes to 'Command data' register set OST status
78
+
register in QEMU
79
+
other values: reserved
80
+
[0x6-0x7] reserved
81
+
[0x8] Command data: (DWORD access)
82
+
current 'Command field' value:
83
+
0: OSPM reads value of CPU selector
84
+
1: stores value into OST event register
85
+
2: stores value into OST status register, triggers
86
+
ACPI_DEVICE_OST QMP event from QEMU to external applications
87
+
with current values of OST event and status registers.
88
+
other values: reserved
89
+
90
+
Selecting CPU device beyond possible range has no effect on platform:
91
+
- write accesses to CPU hot-plug registers not documented above are
92
+
ignored
93
+
- read accesses to CPU hot-plug registers not documented above return
94
+
all bits set to 0.