Text 2090, 77 rader
Skriven 2011-07-12 02:31:08 av Dariusz Piatkowski
Ärende: Re: [New] FF and SM with Native Printing Support
========================================================
p2f31e31560005d75370fdda921ad383fa595805209e2d6a57d1dba32d6236b7e.newsdawg.com
From: "Dariusz Piatkowski" <dariusz@_NO-SPAM_mnsi.net>

On Fri, 8 Jul 2011 03:33:22 UTC, "Rich Walsh" <spamyourself@127.0.0.1> wrote:

> On Fri, 8 Jul 2011 01:57:37 UTC, "Dariusz Piatkowski"
<dariusz@_NO-SPAM_mnsi.net> wrote:
> 
> > Well guys, I tried every single option I could think of, such as the
following:
> > 
> > 1) no RWS
> > 2) no hrdwr speedup
> > 3) "PSD=OS2APIC.PSD /P=4 /APIC" (normal mode) replaced with
"PSD=OS2APIC.PSD 
> > /P=4 /PIC"
> > 4) Firefox started with 'safemode' (all add-ons and personalization
shut-off)
> > 5) this version (4.0.2b) appears to be better behaved, but once 3-4 windows
are 
> > started up (about 2-3 tabs each) the 4 CPU cores start getting maxed
out....
> > 
> > I really am empty handed at this point in time, although I can tell that
the 
> > Firefox configuration did not change (other then 4.0-pre to 4.0.1 & 4.0.2)
and 
> > that the issue only appeared once multi-core system was activated.
> 
> It seems fairly clear that the problem lies in your system's setup, so I'd
> strip it to the bare bones.
> 
> First, back out the PMMERGE from FP6 - IIRC, it had problems.
> 
> Next, disable all of your add-ons:  empty your startup folder, rem-out RUN=
> statements in config.sys, and deregister questionable WPS classes.
> 
> FWIW... a few years ago, all my compiles started failing.  After I disabled
> everything then started adding stuff back, I discovered the cause was FT2LIB.
> It certainly wasn't what I expected because compilers, linkers, etc, are all
> text-mode apps while FT2LIB only loads into PM processes.  Still, it _was_
> the cause of the problem.  The moral here:  leave no stone unturned.

I agree with you...certainly somewhere in there I do have a problem...the big 
question is how the heck do I get at it in a somewhat of an planned manner, as 
opposed to just trying things on the fly so to speak.

Yes, I did hear enough reports that PMMERGE.DLL was a problem area in 
FP6...strangely enough, doing a binary compare to the FP5 version only shows 
differences where th BLDLEVEL info comes in:

1) FP5, 12-29-04   7:33p   1270499          49  PMMERGE.DLL

IBM:14.103

2) FP6,  7-05-06  10:28p   1270499          49  PMMERGE.DLL
IBM:14.105

Sure enough...size wise, they are the same.

I am currently running the matching FP6 SMP krnl & DOSCALL1.DLL files.

In my attempts to narrow this down I had started executing something similar to

your idea regarding the CONFIG.SYS statments, classes, etc....so far I'm really

going through Firefox related stuff though.

Regarding the particular Firefox/Mozialla hi-res timer value (SET 
NSPR_OS2_NO_HIRES_TIMER=1), can you point out what specifically this causes 
Firefox to do (or avoid doing)??? I'm curious if I can use that explanation to 
further narrow down better what might be the root cause of the problem here?

Thanks!

--- Internet Rex 2.31
 * Origin: NewsGuy - Unlimited Usenet $19.95 (1:261/20.999)