File: PLAN.md

package info (click to toggle)
chromium 139.0.7258.127-1
  • links: PTS, VCS
  • area: main
  • in suites:
  • size: 6,122,068 kB
  • sloc: cpp: 35,100,771; ansic: 7,163,530; javascript: 4,103,002; python: 1,436,920; asm: 946,517; xml: 746,709; pascal: 187,653; perl: 88,691; sh: 88,436; objc: 79,953; sql: 51,488; cs: 44,583; fortran: 24,137; makefile: 22,147; tcl: 15,277; php: 13,980; yacc: 8,984; ruby: 7,485; awk: 3,720; lisp: 3,096; lex: 1,327; ada: 727; jsp: 228; sed: 36
file content (26 lines) | stat: -rw-r--r-- 1,209 bytes parent folder | download | duplicates (11) 
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
 
# //components/webcrypto plan

This document outlines the plan for the code living in this directory. The
current code is more or less in maintenance mode.

## Code Changes

* Investigate whether CryptoThreadPool is necessary, whether it is necessary
  for all operations, and whether it can be use some existing more generic
  thread pool rather than requiring a dedicated worker thread
  (https://crbug.com/623700)
* Fix the semantic mismatches in JWK importing with usage masks
  (https://crbug.com/1136147)
* Make AES key scheduling more efficient by not recomputing keys all the
  time (https://crbug.com/1049916)
* Move the entire component into blink and get rid of a bunch of abstraction
  layers (https://crbug.com/614385)
* Remove the unnecessary "threadsafety caches" in WebCrypto keys
  (https://crbug.com/1180244)

## Behavior / Spec Changes (will require spec work)
* Either finish the X25519 / ed25519 implementation (which will probably require
  standardizing it) or remove it (https://crbug.com/1032821)
* Make most of the operations synchronous instead of asynchronous - they are
  fundamentally fast and do not need to be farmed out to worker threads, except
  for *maybe* key generation.