918321 - Remove some dead code from the JS engine

Status: RESOLVED FIXED

(Core :: JavaScript Engine, defect)

Description

[(no longer active)]

Attachment: Patch (v1)

Discovered by clang.

Comment 1

[Luke Wagner [:luke]]

Comment on attachment 807190 [details] [diff] [review]
Patch (v1)

Thanks!

Comment 2

[(no longer active)]

https://hg.mozilla.org/integration/mozilla-inbound/rev/04bf94160337

Comment 3

[Steve Fink [:sfink] [:s:]]

Comment on attachment 807190 [details] [diff] [review]
Patch (v1)

Review of attachment 807190 [details] [diff] [review]:
-----------------------------------------------------------------

::: js/src/jit/MIR.h
@@ +8454,5 @@
>  
>  class MAsmJSLoadGlobalVar : public MNullaryInstruction
>  {
>      MAsmJSLoadGlobalVar(MIRType type, unsigned globalDataOffset, bool isConstant)
> +      : globalDataOffset_(globalDataOffset)

Seems like the isConstant parameter is dead now too.

Comment 4

[Steve Fink [:sfink] [:s:]]

Attachment: Eliminate dead isConstant param

Or would this be useful information in the future?

Comment 5

[Luke Wagner [:luke]]

Let's check with Douglas: are there some pending patches that will be using isConstant?

Comment 6

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/04bf94160337

Comment 7

[Douglas Crosher [:dougc]]

The lack of use of isConstant might be a regression from
bug 877338, or it might no longer be necessary after the
patch in 877338.   It might still be possible to relax
the constraints on movement of a global load when it is
a constant.  I'll need to study the code further.