1441205 - XPTMethodDescriptor::result is unused

Status: RESOLVED FIXED

(Core :: XPCOM, enhancement)

Description

[Andrew McCreight [:mccr8]]

I noticed this and was somewhat alarmed, but I think what is going on here is that a method either:
(1) is not notxpcom, in which case the result type is always nsresult so nothing reads it
(2) is xpcom, in which case it can only be called from C++, so the XPT information is irrelevant.

(2) suggests that we might be able to skip storing more data for notxpcom methods, unless there's something weird like script can enumerate over methods that it can't call.

Comment 1

[Andrew McCreight [:mccr8]]

Unsurprisingly, this does not have any measurable impact on memory usage, to the extent measured by about:memory.

I filed bug 1441222 on loading less information on methods not exposed to script.

Comment 2

[Andrew McCreight [:mccr8]]

Attachment: Bug 1441205 - XPTMethodDescriptor::result is unused.

This sounds weird, but either:
1) A method is notxpcom, so it can't be called from script and the XPT
information is unused.
2) A method is not notxpcom, in which case the result type is nsresult.

Review commit: https://reviewboard.mozilla.org/r/223228/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/223228/

Comment 3

[Nicholas Nethercote [inactive]]

Comment on attachment 8954076 [details]
Bug 1441205 - XPTMethodDescriptor::result is unused.

https://reviewboard.mozilla.org/r/223228/#review229402

Comment 4

[Nicholas Nethercote [inactive]]

> Unsurprisingly, this does not have any measurable impact on memory usage, to
> the extent measured by about:memory.

Not even a single byte? That seems surprising...

Comment 5

[Andrew McCreight [:mccr8]]

(In reply to Nicholas Nethercote [:njn] from comment #4)
> Not even a single byte? That seems surprising...

This was just with two digits of precision. I forgot that if you check "verbose" it gives you the actual bytes... (Also I wonder if the arena allocation means that you only get some kind of arena page level granularity.)

Comment 6

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/f9658f61fbe2
XPTMethodDescriptor::result is unused. r=njn

Comment 7

[Nicholas Nethercote [inactive]]

> (Also I wonder if the arena
> allocation means that you only get some kind of arena page level
> granularity.)

Ah, that might explain it.

Comment 8

[Arthur Iakab [arthur_iakab]]

https://hg.mozilla.org/mozilla-central/rev/f9658f61fbe2