The Official Win32 20110912 builds are out

[Peter(6)]

A couple of notes about the Nightly thread.
Nightly Build thread RSS feed (courtesy DaCypher)

The Official Win32 20110912 NIGHTLY build is out. Changeset: 569a45bfb71c
The Official Win32 20110912 AURORA build is out. Changeset: b1d382004d3f
.The Official Firefox 7.0 beta 6 builds are not yet out Changeset: 85615f1671ec

Previous NIGHTLY 20110911 (FF9.0a1)
Previous AURORA 20110911 (FF8.0a2)
Previous BETA 20110908 (FF7.0b5 build 1)

Hourly NIGHTLY builds : mozilla-central-linux || mozilla-central-macosx || mozilla-central-win32 || Firefox Tinderbox || NIGHTLY Changelog
Hourly NIGHTLY archive for Windows, Mac and Linux (not the final format yet)
AURORA Tinderbox || AURORA changelog

Last update: 2011-09-12 -- 07:00 PDT = 14:00 UTC

Fixed:

NIGHTLY

1. #378577 [Firefox:Preferences]-"Use Bookmark" button needs three dots after the text [Win]
2. #656251 [Core:XSLT]-reduce verbosity of Accept header for fetching XSLT [All]
3. #680113 [Toolkit:Add-ons Manager]-Select Add-ons dialog lists distribution add-ons as installed by the user [Mac]
4. #682338 [Core:XP Toolkit/Widgets: XUL]-nsXULPopupListener::HandleEvent intends to focus contextmenu target on platforms where it's shown on mousedown, but not those where it's shown on mouseup; been doing exact opposite since 2001 [All]
5. #682845 [Core:Networking]-Bug 665706 changes IDL without changing IIDs, changes nsStandardURL serialization format [All]
6. #685464 [Core:XP Toolkit/Widgets: Menus]-DOMMenuItemActive is not consistent menulist items [Win]
7. #685798 [Core:DOM]-Node.parentElement not implemented [All]

AURORA
None

BETA
None

Partial Landings/WIPs/Incoming:

NIGHTLY
None

AURORA
None

BETA
None

Regressions/Annoying/Common bugs:

None

Nightly 9.0a1 fixes since 20110322 (mozilla 2.2) = ~2916
Aurora 8.0a2 fixes since 20110322 (mozilla 2.2) = ~2536
Beta 7.0 fixes since 20110322 (mozilla 2.2) = 1944
Release FF 6.0.2 ~1333,FF 5.0 ~669

Release tracking Firefox 7-10 see HERE

[C@rb0n]

@C@rb0n: - Yeah. I have tried that too and still the same.

Is this with 32-bit of 64-bit Firefox? Do you notice a difference between the two?

If your hardware is sufficiently fast, maybe the only other reasonable possibility is interaction with some other software on your system.

What security software do you have? Have you tried disabling those?

What happens if you run Firefox under XP Mode?

[sciguyryan]

@C@rb0n: - Yeah. I have tried that too and still the same.

Is this with 32-bit of 64-bit Firefox? Do you notice a difference between the two?

If your hardware is sufficiently fast, maybe the only other reasonable possibility is interaction with some other software on your system.

What security software do you have? Have you tried disabling those?

What happens if you run Firefox under XP Mode?

32-bit currently. I haven't used the 64-bit in quite a while as I found it unstable when last I used it.

I'm running on an Intel i7 930 with hyperthreading @ 2.8GHz per core, 12GB of DDR3 1600MHz, a SATAII SSD hard drive and a Gigabyte GTX 560 TI graphics card. I doubt hardware is at fault given that I haven't changed anything recently and this problem surfaced less than a week ago.

I use ESET Smart Security 4. I haven't changed that in a long time so I doubt it is at fault too. I have tested with it disabled as you suggested and the issue still remains.

I haven't tested Firefox under XP mode - wouldn't that affect the hardware acceleration features?

[C@rb0n]

I haven't tested Firefox under XP mode - wouldn't that affect the hardware acceleration features?

The idea is to eliminate the possibility that something installed on your Windows 7 system is at fault. Using XP Mode should help with this, since it will have access to your hardware -- mostly independent of the host OS.

You could also open Windows Task Manager and click to sort columns on the Processes tab by CPU usage. Then you could check to see if there's anything using high CPU during the pause, based on what process shows up on top. You could also add other columns, like I/O, to monitor.

Alternately, you could use Microsoft Process Monitor to see if Firefox or something else is performing a lot of I/O during that time.

If you're using RAID, that's another possible source of pauses (usually seen with mechanical drives).

[sciguyryan]

I haven't tested Firefox under XP mode - wouldn't that affect the hardware acceleration features?

The idea is to eliminate the possibility that something installed on your Windows 7 system is at fault. Using XP Mode should help with this, since it will have access to your hardware -- mostly independent of the host OS.

I'll test it but given that this started suddenly several nightlies ago... I doubt the OS is at fault here given that very little has changed on my system for a good few months. I'll report back after I've tested it.

I posted a snapshot from Process Explorer on the bug I filled for this (see Bug 686157).

I don't use any RAID at all.

Edit: That was an amazingly short test! It only took me a few minutes to get exactly the same effect on the XP compatibility mode.

[C@rb0n]

Based on your screenshot in bug 686157, and your response above, I suspect you're having I/O problems more so than CPU problems.

One way to confirm this is to set browser.sessionstore.interval to something like 1800000 (i.e. 30 minutes), restart Firefox, and see if you still get pauses.

Have you looked at SMART status for your disk or run any other type of diagnostics?

[sciguyryan]

Based on your screenshot in bug 686157, and your response above, I suspect you're having I/O problems more so than CPU problems.

One way to confirm this is to set browser.sessionstore.interval to something like 600,000 (i.e. 10 minutes), restart Firefox, and see if you still get pauses.

Have you looked at SMART status for your disk or run any other type of diagnostics?

it is a brand new hard drive, less than a month old. Just ran a SMART on it and all is fine.

I have just ran Process Explorer for a good 20 minutes to see if IO load corresponds to the issues and it doesn't seem to correlate except for small IO spikes (of 10KB at a time). Certainly not enough to cause this as far as I can tell.

I'll run Fx with the value as above and see if that fixes it.

[sciguyryan]

Okay. I just tried something given on an idea you said about with the session interval. I reset that and then deleted the session restore files... and now the problem has gone. I'm not sure what it was in the file that caused it but I have the exact same tabs open now and no issue at all.

[C@rb0n]

If that's the case, I'm surprised you experienced it with a new profile. You didn't use that session file in the new profile did you?

In any case, you might have experienced something similar to bug 624409, caused by a bad session file.

[sciguyryan]

If that's the case, I'm surprised you experienced it with a new profile. You didn't use that session file in the new profile did you?

In any case, you might have experienced something similar to bug 624409, caused by a bad session file.

No. I didn't. But I did import the tab set via sync and the problem emerged a few days later - likely whatever caused it is a cumulative issue and simply resurfaced after further use. I just looked at the old session restore file and it was over 5MB. No wonder that was causing issues.

[WhtDwrf]

Nightly's out:

Mozilla/5.0 (Windows NT 5.1; rv:9.0a1) Gecko/20110912 Firefox/9.0a1 ID:20110912030829
CSet: 569a45bfb71c

[Str0ngwun]

Did anyone file a bug, that on the last 2 nightlies - music players don't work? examples are soundcloud and grooveshark
nvm, it was a local issue (I forgot that I've recently attempted to install a newer version of audio drivers, and forgot to double restart)

[alex_mayorga]

Did anyone file a bug, that on the last 2 nightlies - music players don't work? examples are soundcloud and grooveshark

Grooveshark WFM at this very moment.

[Peter(6)]

This thread is closed, please use the next one.