Comments on: Unittests on PPC and Non-SSE2 Machines

By: John Ford

John Ford — Mon, 23 Nov 2009 03:38:46 +0000

that is awesome news!

By: Ted Mielczarek

Ted Mielczarek — Tue, 10 Nov 2009 14:14:12 +0000

dbaron flipped the pref in reftest to disable the slow script dialogs just recently in bug 523934, because we were hitting them on debug unittests.

By: Justin Dolske

Justin Dolske — Sun, 08 Nov 2009 04:40:59 +0000

GeriatricMasters? Heh, I approve!

By: Jesse Ruderman

Jesse Ruderman — Sat, 07 Nov 2009 18:50:18 +0000

Are you using runreftest.py? It should set that pref for you automatically.

By: William J. Edney

William J. Edney — Sat, 07 Nov 2009 16:35:46 +0000

John -

Here is the preference that should do what you want regarding having script timeouts:

http://kb.mozillazine.org/Dom.max_script_run_time

Don’t know to set this in an automated fashion, except to have a process update the ‘preferences’ file.

Cheers,

- Bill

By: -

Sat, 07 Nov 2009 16:11:43 +0000

This is important, specially on the AMD side. Original Athlons had neither SSE nor SSE2, and were sold until 2002 at the very least and got to 1.4GHz speeds (and were more powerful than some mobile x86 processors being sold today!). Then you have the Athlon XP, which had SSE but not SSE2, sold until 2005 I think with speeds in excess of 2GHz.

So yeah, expecting MMX nowadays is reasonable. SSE is very questionable, SSE2 is probably crazy.