irq Sharing

[Cardy13]

Hi, I randomly get BSOD's about IRQ_NOT_LESS_OR_EQUAL

I checked my IRQ's and I noticed a couple things...

High Definition Audio Controller is on the same IRQ as my Nvidia GeForce 8800 GTS.

I also noticed another High Definition Audio Controller on the same IRQ as Nvidia Nforce Serial ATA Controller

Are these bad things to be sharing?

My specs:

EVGA Nforce 680i SLi LT Mobo
4x 1gb Adata DDR2-800mhz
EVGA Geforce 8800 GTS SSC edition
Intel Core 2 Quad q6600
Soundblaster X-fi PCI Expressx1
500gb Seagate barracude 7200rpm HDD (SATA)
52x LG Dvd rw/cdrw lightscribe drive (SATA)


Thanks for any help you can offer.

[screwballl]

What OS?

The MS website has all sorts of solutions from driver issues to BIOS settings to Fast User switching...

[Tech Geek Deluxe]

Are you using XP Pro or ?

Likely it isn't an IRQ issue.

More likely a Ram or file corruption issue.
If it persists try swapping Ram Modules and/or testing them.


However; since you asked; it isn't a good thing to have those on the same IRQ.

I was one of a handful of peeps in the US to figure out how to change and set IRQ's for Win 2000 Pro [without a re-install or bios settings] ... I don't know however if XP Pro works the same way.

I'm not sure if the Bios in the later gen motherboards force IRQ priority and make the OS recognize the settings.


I can tell you that having very active hardware such as a video card and a sound card and/or a network share the same IRQ ... well it can cause problems or lockups ... or sound hummm ... or various irritating phenomena.

However the multicore processors and faster ram and hardware greatly reduce those issues now to where it usually isn't an issue.

[Cardy13]

I am having these problems on Vista Home Premium 64 bit. I have a dual boot of XP pro, but I haven't used it in a while and I can't directly duplicate the BSOD as it is usually random.

the High Definition Audio Device I believe is onboard, but there is two of them so I don't know if maybe one of those is my soundcard? Or would my soundcard come up as Sounblaster X-fi or something similiar.

Thanks

[Hitman]

Sounds like a driver issue.

[Boy'nBlack]

Nah, tech has it. It's not so exacting in the literal IRQ that we had control of a long time ago. It's a memory error. An IRQ is nothing more than an interrupt request. It could be a driver, but it can be anything hardware butting in pretty much. Just like talking on a CB without a "breaker". You step on someone, that's an IRQ error. No refetch outside of the bus.

The numbers later aim you at what it is. It's usually first faulted in the memory, but that can cloak what it really is as well. It's just the indication if it's not a hex of a 128-bit number. Most just happen to be 128-bit numbers though. If it's not, THEN you can identify the slot to narrow it down to a specific piece of hardware and/or driver.

[Cardy13]

Hmm, OK...

If I post some dumps would anyone be able to analyze them for me? Try to isolate the problem?

I remember a while back the BSOD would have an Nvidia driver at the bottom... can't remember what one but I know it was an Nvidia driver. Sometimes it's there, sometimes it isn't.. but I've always been using the latest drivers. I reseted the CMOS and reseated all my ram, and I couldn't duplicate it doing the exact same thing that I was doing before, so it seems a bit better now, but I'm always expecting it to come up randomly again... no peace of mind here. I THINK... not 100% sure, but I think sometimes i would get the PAGE_FAULT_IN_NON_PAGED_AREA BSOD. I don't know if that has to do with the page file or not. I have one 500gb HDD, split up into 3 partitions: Vista, XP, and Files. I have the page file on the Vista partition @ 6144MB (1.5x the amount of my RAM).

Anyways I also ran Memtest86, there were errors on test 5, on all four sticks of my ram.. Is that probable? All four sticks being bad? I ran memtest a while back (4 months ago I would guess) and I came up clean, no errors on 12 hours of testing.

After I noticed errors on test 5 I ran just test 5 over and over, and came back to Module 0 having infinite errors.. It just kept going up and up and up. The time tested was frozen, and module 0s error count was going through the roof. This concerned me a bit.. Like I said I've reseted the CMOS and I set the BIOS to default as well as reseating and switching around my RAM (4 sticks in 4 slots, so I couldn't really change slots, just swapped the RAM around)

I'm gonna run memtest again right now and see if I still have errors. The whole thing seems a bit loopy, I haven't dealt with a problem this ambigious ever. It's not like it happens ALOT, but I never know when its going to happen, and it always keeps me on edge, lol.

Anyways let me know if I should post my dumps and maybe one of you guys could help me out :).

Thanks,
Chris.

Edit: I also would like to mention that you guys are very helpful here :) I really appreciate the help.

[screwballl]

memory testing should be done certain ways. You took care of step 1. Step 2 is to take out all sticks except for 1 in the first slot and run the memtest again. Still get errors? Try the same single stick in slot 2... then slot 3, then slot 4. Try each stick by itself in slot 1, then slot 2, then 3 and 4. Also try 2 sticks in slot 1 and 2, 1 and 3, 1 and 4, 2 and 3, 2 and 4, 3 and 4...
Narrow it down so you can find out if it is a single bad RAM slot or a single bad stick or a bad pair of sticks... if all the memory gives no errors on any tests in slots 2, 3 and 4, then slot 1 may be bad and need to RMA the motherboard. If all sticks come up bad in all slots then something zapped all 4 sticks of memory, but there is a possibility that the board itself did it. Using another stick of memory that is known to be good can be tested starting backwards, from slot 4, to 3, 2, then 1. This lets you know if the board itself is bad.
If you have another system, try the 4 sticks of memory in there running the same tests, that usually tells you if the memory itself is bad.

I know the memory tests take time but that is the best way to find out if it is the memory itself or the board itself.

[Boy'nBlack]

When threads like this die, I get worried. But it's the holiday, so...bump?

I'm not a fan of memory tests, as even good units can fail a test and lead one off into the dark forest. If we had a Sig to look at, see a 680/780i, it'd be quick to say what the problem is to me. If not overclocked and on an Intel and we had just one fail, then we'd have an actual step to progress from.