Fast boot
地域：
CA
CA Campbell
摘要
Initializing a computing system using dormant pages includes marking a set of guest physical addresses as dormant. It further includes, for each node in a plurality of physical nodes, designating a set of real physical addresses for zeroing. An operating system is executing collectively across the physical nodes.
說明書
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
This application claims priority to U.S. Provisional Patent Application No. 62/586,052 entitled FAST BOOT filed Nov. 14, 2017 which is incorporated herein by reference for all purposes.
The time to perform memory testing and initialization is proportional to the amount of memory presented to the operating system. On conventional servers with large memories, the booting up of such large memory systems may take a substantial amount of time. Further, if the operating system is performing the initialization, then normal applications are typically not run until this initialization is complete. Improvements to memory initialization of large memory systems are needed.
Various embodiments of the invention are disclosed in the following detailed description and the accompanying drawings.
權(quán)利要求
What is claimed is:1. A computer system, comprising:a plurality of physical nodes, wherein an operating system is executing collectively across the plurality of physical nodes; wherein a set of guest physical addresses is marked as dormant; and wherein, for each node in the plurality, a set of real physical addresses is designated for zeroing. 2. The computer system of claim 1 , wherein the set of guest physical addresses marked as dormant are initially located on a first node in the plurality of physical nodes. 3. The computer system of claim 1 , wherein each node in the plurality comprises a list of real physical addresses to be zeroed. 4. The computer system of claim 3 , wherein a portion of physical memory corresponding to a real physical address in the list of real physical addresses to be zeroed is allocated and zeroed. 5. The computer system of claim 4 , wherein subsequent to the allocation and zeroing of the is portion of physical memory corresponding to the real physical address, the real physical address is removed from the list of real physical addresses to be zeroed, and wherein the real physical address is included in a list of real physical addresses that have been zeroed. 6. The computer system of claim 4 , wherein the portion of physical memory is allocated and zeroed on-demand. 7. A computer system, comprising:a plurality of physical nodes, wherein an operating system is executing collectively across the plurality of physical nodes; wherein an indication of a stall on a guest physical address marked as dormant is received at a first node in the plurality; wherein an available portion of physical memory is obtained; and wherein the guest physical address is mapped to the obtained available portion of physical memory. 8. The computer system of claim 7 , wherein obtaining the available portion of physical memory comprises checking a list of real physical addresses local to the first node that have been zeroed. 9. The computer system of claim 8 , wherein obtaining the available portion of physical memory comprises obtaining a portion of physical memory corresponding to a real physical address in the list of real physical addresses that have been zeroed. 10. The computer system of claim 9 , wherein the real physical address is removed from the list of real physical addresses that have been zeroed. 11. The computer system of claim 7 , wherein obtaining the available portion of physical memory comprises checking a list of real physical addresses local to the first node that are to be to zeroed. 12. The computer system of claim 11 , wherein obtaining the available portion of physical memory comprises allocating a portion of physical memory corresponding to a real physical address in the list of real physical addresses local to the first node that are to be zeroed, wherein the allocated portion of physical memory is zeroed. 13. The computer system of claim 7 , wherein a virtualized processor associated with the stall is migrated to another node in the plurality that has local memory that is available to be mapped to the guest physical address. 14. A computer system, comprising:a plurality of physical nodes, wherein an operating system is executing collectively across the plurality of physical nodes; wherein an instruction to zero a portion of memory associated with a guest physical address is detected; wherein the guest physical address is marked as dormant; and wherein the instruction to zero the portion of memory is skipped. 15. The computer system of claim 14 , wherein a portion of physical memory is unmapped from the guest physical address marked as dormant. 16. The computer system of claim 15 , wherein the portion of physical memory is included in a list of portions of physical memory to be zeroed. 17. The computer system of claim 14 , wherein the instruction is detected at a first node in the plurality of physical nodes, and wherein the guest physical address is defined on a second node. 18. The computer system of claim 17 , wherein a message is sent from the first node to the second node indicating that the guest physical address is to be marked as dormant. 19. The computer system of claim 17 , wherein ownership of the guest physical address is transferred from the second node to the first node, and wherein subsequent to the transfer, the guest physical address is marked as dormant. 20. The computer system of claim 19 , wherein the guest physical address is unmapped from the second node.
意見反饋
使用該功能遇到問題