Improve documentation

This commit is contained in:
Julien Cretin
2020-10-13 12:46:25 +02:00
parent eede767fe3
commit f512645837
2 changed files with 24 additions and 7 deletions

View File

@@ -456,8 +456,15 @@ bitfield! {
LEN_COMPACT: Length,
}
// Overview of the first word of the different kind of entries:
// Overview of the first word of the different kind of entries.
//
// Each column represents a bit of the word. The first 2 lines give the position in hexadecimal of
// the bit in the word (the exponent of 2 when the word is written in binary). Each entry starts
// with the sequence of bits of its identifier. The dots following the identifier are the number of
// bits necessary to hold the information of the entry (including the checksum). The remaining free
// bits after the dots are not used by the entry.
//
// 0 1
// 0123456789abcdef0123456789abcdef
// padding 0
// header 10..............................
@@ -466,7 +473,8 @@ bitfield! {
// marker 11010..........
// remove 11011.................
//
// NOTE: We could pad the internal entries to the right.
// NOTE: We could pad the internal entries to the right by extending their identifier. This permits
// to free some space for shorter identifier for future kind of entries.
// The fields of a padding entry.
bitfield! {

View File

@@ -95,7 +95,7 @@
//! The _total lifetime_ of the store is below `L = ((E + 1) * N - 1) * (P - 2)` and
//! above `L - M` words, where `E` is the maximum number of erase cycles. The
//! lifetime is used when capacity is used, including transiently, as well as when
//! compaction occurs. Compaction frequency and litetime consumption are positively
//! compaction occurs. Compaction frequency and lifetime consumption are positively
//! correlated to the store load factor (the ratio of used capacity to total capacity).
//!
//! It is possible to approximate the cost of transient words in terms of capacity:
@@ -125,8 +125,8 @@
//! It is preferred to use `Clear` with a threshold of 0 to keep the lifetime
//! tracking.
//!
//! The store properties may still hold outside some of those assumptions but with
//! weaker probabilities as the usage diverges from them.
//! The store properties may still hold outside some of those assumptions, but with
//! an increasing chance of failure.
//!
//! # Implementation
//!
@@ -301,9 +301,18 @@
//! erased word. Marking an entry as padding or deleted is a single bit operation,
//! so the property trivially holds. For those cases, the proof relies on the fact
//! that there is exactly one bit equal to 0 in the 3 first bits. Either the 3 first
//! bits are still `111` in which case we expect the remaining bits to bit equal
//! bits are still `111` in which case we expect the remaining bits to be equal
//! to 1. Otherwise we can use the checksum of the given type of entry because those
//! 2 types of entries are not reachable from each other.
//! 2 types of entries are not reachable from each other. Here is a visualization of
//! the partitioning based on the first 3 bits:
//!
//! | First 3 bits | Description | How to check |
//! | ------------:| ------------------ | ---------------------------- |
//! | `111` | Erased word | All bits set to `1` |
//! | `101` | User entry | Contains a checksum |
//! | `110` | Internal entry | Contains a checksum |
//! | `100` | Deleted user entry | No check, atomically written |
//! | `0??` | Padding entry | No check, atomically written |
//!
//! To show that valid entries of a given type are not reachable from each other, we
//! show 3 lemmas: