pagecache 0.17.0

lock-free pagecache and log for high-performance databases
Documentation 
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
118
119
120
121
122
123
124
125
126
127
128
129

use std::ptr;
use std::sync::Mutex;

use super::*;

/// A simple Lru cache.
pub struct Lru {
    shards: Vec<Mutex<Shard>>,
}

unsafe impl Sync for Lru {}

impl Lru {
    /// Instantiates a new `Lru` cache.
    pub fn new(cache_capacity: u64, cache_bits: usize) -> Lru {
        assert!(
            cache_bits <= 20,
            "way too many shards. use a smaller number of cache_bits"
        );
        let size: usize = 1 << cache_bits;
        let shard_capacity = cache_capacity / size as u64;

        Lru {
            shards: rep_no_copy![
                Mutex::new(Shard::new(shard_capacity));
                size
            ],
        }
    }

    /// Called when a page is accessed. Returns a Vec of pages to
    /// try to page-out. For each one of these, the caller is expected
    /// to call `page_out_succeeded` if the page-out succeeded.
    pub fn accessed(&self, pid: PageId, sz: u64) -> Vec<PageId> {
        let shard_idx = pid % self.shards.len() as u64;
        let rel_idx = pid / self.shards.len() as u64;
        let shard_mu = &self.shards[usize::try_from(shard_idx).unwrap()];
        let mut shard = shard_mu.lock().expect(
            "Lru was poisoned by a \
             thread that panicked \
             inside a critical section",
        );
        let mut rel_ids = shard.accessed(rel_idx, sz);

        for rel_id in &mut rel_ids {
            let real_id = (*rel_id * self.shards.len() as u64) + shard_idx;
            *rel_id = real_id;
        }

        rel_ids
    }
}

#[derive(Clone)]
struct Entry {
    ptr: *mut dll::Node,
    sz: u64,
}

impl Default for Entry {
    fn default() -> Entry {
        Entry {
            ptr: ptr::null_mut(),
            sz: 0,
        }
    }
}

struct Shard {
    list: Dll,
    entries: Vec<Entry>,
    capacity: u64,
    sz: u64,
}

impl Shard {
    fn new(capacity: u64) -> Shard {
        assert!(capacity > 0, "shard capacity must be non-zero");

        Shard {
            list: Dll::default(),
            entries: vec![],
            capacity,
            sz: 0,
        }
    }

    fn accessed(&mut self, rel_idx: PageId, sz: u64) -> Vec<PageId> {
        if PageId::try_from(self.entries.len()).unwrap() <= rel_idx {
            self.entries.resize(
                usize::try_from(rel_idx).unwrap() + 1,
                Entry::default(),
            );
        }

        {
            let entry = &mut self.entries[usize::try_from(rel_idx).unwrap()];

            self.sz -= entry.sz;
            entry.sz = sz;
            self.sz += sz;

            if entry.ptr.is_null() {
                entry.ptr = self.list.push_head(rel_idx);
            } else {
                entry.ptr = self.list.promote(entry.ptr);
            }
        }

        let mut to_evict = vec![];
        while self.sz > self.capacity {
            if self.list.len() == 1 {
                // don't evict what we just added
                break;
            }

            let min_pid = self.list.pop_tail().unwrap();
            self.entries[usize::try_from(min_pid).unwrap()].ptr =
                ptr::null_mut();

            to_evict.push(min_pid);

            self.sz -= self.entries[usize::try_from(min_pid).unwrap()].sz;
            self.entries[usize::try_from(min_pid).unwrap()].sz = 0;
        }

        to_evict
    }
}