802616 - don't use software floating point math on ARMv6

Status: RESOLVED FIXED

(Firefox Build System :: General, defect)

Description

[Brad Lassey [:blassey] (use needinfo?)]

Attachment: patch

this is OMG slowing us down, I think we just need to make floating point hardware a min system requirement

Comment 1

[Mike Hommey [:glandium]]

Comment on attachment 672305 [details] [diff] [review]
patch

Please make the change in the mozconfigs.

Comment 2

[Brad Lassey [:blassey] (use needinfo?)]

(In reply to Mike Hommey [:glandium] from comment #1)
> Comment on attachment 672305 [details] [diff] [review]
> patch
> 
> Please make the change in the mozconfigs.

I don't want to. I want this to be the default so that developers will be building the same thing we are shipping

Comment 3

[Mike Hommey [:glandium]]

(In reply to Brad Lassey [:blassey] from comment #2)
> (In reply to Mike Hommey [:glandium] from comment #1)
> > Comment on attachment 672305 [details] [diff] [review]
> > patch
> > 
> > Please make the change in the mozconfigs.
> 
> I don't want to. I want this to be the default so that developers will be
> building the same thing we are shipping

The change in build/autoconf/arch.m4 impacts more than developers. Developers already have to add something to do armv6 build, i don't see why it would be a problem for them to add something more.

Comment 4

[Mike Hommey [:glandium]]

Comment on attachment 672305 [details] [diff] [review]
patch

Review of attachment 672305 [details] [diff] [review]:
-----------------------------------------------------------------

Make it at least limited to Android.

Comment 5

[Justin Wood (:Callek)]

Without looking at how high up in the configure we need/want this, my suspicion is a change in http://mxr.mozilla.org/comm-central/source/mozilla/mobile/android/confvars.sh would do what brad and glandium want, without being intrusive

Comment 6

[Brad Lassey [:blassey] (use needinfo?)]

Attachment: patch

Comment 7

[Brad Lassey [:blassey] (use needinfo?)]

Comment on attachment 672330 [details] [diff] [review]
patch

[Approval Request Comment]
Bug caused by (feature/regressing bug #): 
User impact if declined: for some reason (we will investigate) soft float math is killing us on region operations, which is making display list building very slow, which makes panning very slow. Not taking this will make panning slower than we want it to be.
Testing completed (on m-c, etc.): just landed on m-i, should probably get a couple nightlies before uplift
Risk to taking this patch (and alternatives if risky): 
String or UUID changes made by this patch: 



https://hg.mozilla.org/integration/mozilla-inbound/rev/c4119fd2fb37

Comment 8

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/c4119fd2fb37

Comment 9

[Lukas Blakk [:lsblakk] use ?needinfo]

Triage drive-by, will wait for a couple of nightlies.

Comment 10

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/1bb9cedfb386
https://hg.mozilla.org/releases/mozilla-beta/rev/121213ddf98b

Comment 11

[Tracy Walker [:tracy]]

mass remove verifyme requests greater than 4 months old