(EE) NVIDIA(0): Error recovery failed. etc with 7300GS

[phelin]

I have an HTPC which was working perfectly until last Saturday, when all the sudden while watching TV the picture froze. I tried to login with ssh to the PC but it didn't respond so I just rebooted the whole PC. But now, the picture froze when X was starting, before the desktop had even loaded properly. I checked with top and saw that xorg was taking 100% of cpu. In the /var/log/Xorg.0.log I saw the following entries:

(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(EE) NVIDIA(0): Error recovery failed.
(EE) NVIDIA(0):  *** Aborting ***
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(II) NVIDIA(0): Initialized GPU GART.
(WW) NVIDIA(0): The NVIDIA X driver has encountered too many errors.  Falling
(WW) NVIDIA(0):     back to write-back cached memory.
(II) NVIDIA(0): Initialized GPU GART.
(EE) NVIDIA(0): Error recovery failed.
(EE) NVIDIA(0):  *** Aborting ***
...

Now I am totally unable to start X, since that same thing happens every time. I had not made any changes to the system in several weeks before this happened, so I am wondering whether this could mean that some piece of the hardware has broken down?

Update: Tried with the nv driver and managed to start. But in 3 minutes the picture froze again and xorg was taking 100% of cpu. Any hint of what's going on would be highly appreciated.

[memeyou]

Originally Posted by phelin

I had not made any changes to the system in several weeks before this happened, so I am wondering whether this could mean that some piece of the hardware has broken down?

This is most likely the case. I've had this error from a junk PSU but it could be something else for you. I'd try a reinstall or livecd or other alternative before replacing hardware, though.

[phelin]

Originally Posted by memeyou

This is most likely the case. I've had this error from a junk PSU but it could be something else for you. I'd try a reinstall or livecd or other alternative before replacing hardware, though.

Tried with a livecd and ended up with the same result. Now I need to know which piece of hardware could be the culprit. I am guessing that it is either the motherboard (Asrock 775Dual-VSTA, 19 months old), the PSU (Antec Neo HE 430, 22 months) or the GPU (MSI NX7300GS-MD256EH, 15 months). In my opinion the PSU is the least probable, but perhaps someone knows better? Any way to test without changing the hardware (don't have any spare parts at the moment)?

[memeyou]

Originally Posted by phelin

Tried with a livecd and ended up with the same result. Now I need to know which piece of hardware could be the culprit. I am guessing that it is either the motherboard (Asrock 775Dual-VSTA, 19 months old), the PSU (Antec Neo HE 430, 22 months) or the GPU (MSI NX7300GS-MD256EH, 15 months). In my opinion the PSU is the least probable, but perhaps someone knows better? Any way to test without changing the hardware (don't have any spare parts at the moment)?

check the psu reviews on newegg - quiet a few of them die over time. it's also a very old product - more like 36 months.

[phelin]

Originally Posted by memeyou

check the psu reviews on newegg - quiet a few of them die over time. it's also a very old product - more like 36 months.

Well, I went and bought a new G35 motherboard with integrated graphics. Haven't yet tried with the old nvidia gpu, but with integrated gpu everything works without a glitch. I will try to arrange time to test the nvidia card to see if it works with this motherboard.

[phelin]

A small update on this matter. I tested the 7300GS with a new motherboard (Asus P5E-V HDMI) and it worked perfectly. So, the culprit seams to have been the old motherboard.