Closed
Bug 564194
Opened 15 years ago
Closed 6 years ago
MDC MMGC Documentation Review / Document GC rules
Categories
(Tamarin Graveyard :: Garbage Collection (mmGC), defect, P3)
Tamarin Graveyard
Garbage Collection (mmGC)
Tracking
(Not tracked)
RESOLVED
WONTFIX
Q2 12 - Cyril
People
(Reporter: dansmith, Unassigned)
References
Details
Attachments
(1 file)
1.89 KB,
text/plain
|
Details |
The Tamarin MMGC documentation should be reviewed for accuracy and updated following the changes in FP 10.1.
https://developer.mozilla.org/en/MMgc
Summary: Public MMGC Documentation Review → MMGC Documentation Review
Summary: MMGC Documentation Review → MDC MMGC Documentation Review
Assignee: nobody → treilly
Flags: flashplayer-qrb+
Priority: -- → P3
Target Milestone: --- → flash10.2
Updated•14 years ago
|
Whiteboard: Tracking
Updated•14 years ago
|
Assignee: treilly → lhansen
Comment 1•14 years ago
|
||
See also bug #617395 about a "GC rules" document.
Comment 2•14 years ago
|
||
We should kill the MDC docs and move everything into html files in the docs/mmgc directory in TR. There can be a link from the MDC page into the hg repository so no functionality will be lost.
Comment 3•14 years ago
|
||
Also:
- object invariants for exact tracing
- delete == bad
- safegc rules / usage
Priority: P3 → P2
Updated•14 years ago
|
Flags: flashplayer-bug-
Comment 4•14 years ago
|
||
From bug #617395:
"We need to document (text file in docs directory) all the MMgc rules. We just
removed the global new operator and have a new rule that all GC allocations
done with new have to inherit from GCObject (enforced with compile error).
Also we just tightened up the rules around what can be passed to a write
barrier (assertion enforced).
Something that listed all the rules and how/if they are enforced would be
great. In an ideal world we could prove that if none of the rules were broken
that a program couldn't crash from a GC problem. Really though I just want to
point people at something when they say "I didn't know that rule! Where are the
rules written down?"
--------
GC.h needs to include a clear and visible comment at the beginning that points
to that file.
It's not clear to me why that documentation should not be part of the MMgc docs
on MDC, though, since that would allow us to have all the docs in one place and
structured in some useful way. (Alternatively kill the MDC docs and move
everything into the docs directory.) See also bug #564194 which is about
cleaning up the MDC docs.
--------
I'm now firmly in favor of killing the MDC docs and moving everything into the
docs directory."
Updated•14 years ago
|
Summary: MDC MMGC Documentation Review → MDC MMGC Documentation Review / Document GC rules
Updated•14 years ago
|
Priority: P2 → P3
Comment 6•14 years ago
|
||
I want to redo this as an HTML document with a table that show for each rule how its enforced (ie compiler, assert, not, telepathy etc).
Comment 7•14 years ago
|
||
Updated•14 years ago
|
Whiteboard: Tracking
Updated•14 years ago
|
Target Milestone: Q3 11 - Serrano → Q4 11 - Anza
Comment 8•14 years ago
|
||
I just landed a fairly extensive "rules" document in doc/mmgc in TR (without review - I figure we can review when it's finished, but there seemed value in landing it today, ahead of the GC tutorial tomorrow).
I also landed a stub "mmgc cookbook", but it really is a stub - hardly anything there.
Comment 9•14 years ago
|
||
Fixme: allocation of arrays is mentioned under the "dont use GC::Alloc or GC::Calloc", but otherwise unaddressed.
Comment 10•14 years ago
|
||
Discussions with A Greenblatt on DL-MMgc: We probably want to have a well-developed document on how to use MMgc (notably GC'd storage) effectively.
Updated•13 years ago
|
Assignee: lhansen → nobody
Comment 12•6 years ago
|
||
Tamarin is a dead project now. Mass WONTFIX.
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → WONTFIX
Comment 13•6 years ago
|
||
Tamarin isn't maintained anymore. WONTFIX remaining bugs.
You need to log in
before you can comment on or make changes to this bug.
Description
•