(More info) Sudden/Frequent 7.2 lockups out of nowhere

Tuc at T-B-O-H.NET ml at t-b-o-h.net
Fri Sep 28 16:30:56 PDT 2007


	I've got some more info to add...

	I originally thought I was nuts when I was writing this, so I 
didn't say anything... But since then its happened enough times that it
needs to be mentioned. About 80% of the time it locks up, in an area
about the size of a standard square US postage stamp, appears and
disappears a set of lines almost like a bar code/UPC with the cursor in
the upper left of the area. It fades in and out inconsistently.

	So for the ones that haven't totally brushed me off yet... I
did another ktrace, and oddly enough it appeared that as soon as that
part from GMAIL occurred, it went into its loop again. So I renamed
my .xinitrc in expectations it'd just run some xterms, xclock, and twm.
This should have kept the most minimal of things running. Still locks
up. Ok...

	So I decided that even though my X wasn't compiled with
debug, I'd run it under gdb and see if it gave anything of any
value. Oddly, it did give something more.... When it locked up,
I went back to GDB and got :

Program received signal SIGINT, Interrupt.
0x284ad0d1 in NVDmaWait ()
   from /usr/local/lib/xorg/modules/drivers//nv_drv.so
(gdb) bt
#0  0x284ad0d1 in NVDmaWait ()
   from /usr/local/lib/xorg/modules/drivers//nv_drv.so
#1  0x284ae158 in NVSubsequentColorExpandScanline ()
   from /usr/local/lib/xorg/modules/drivers//nv_drv.so
#2  0x295d8518 in XAATEGlyphRendererScanlineLSBFirst ()
   from /usr/local/lib/xorg/modules//libxaa.so
#3  0x295bb80f in XAAGlyphBltTEColorExpansion ()
   from /usr/local/lib/xorg/modules//libxaa.so
#4  0x295bb429 in XAAImageText8TEColorExpansion ()
   from /usr/local/lib/xorg/modules//libxaa.so
#5  0x0813df7d in DamageExtensionInit ()
#6  0x0808483c in doImageText ()
#7  0x08084966 in ImageText ()
#8  0x0807fafa in ProcImageText8 ()
#9  0x081238f1 in XaceHook ()
#10 0x0807be31 in Dispatch ()
#11 0x0806c025 in main ()

	So this is telling me something is happening with my nvidia
(Geoforce 4 GO or something silly like that) card? DMA issues?
Just for giggles, I decided to check that the nv_drv.so hadn't been
updated. Heres what I find :

total 4804
-rwxr-xr-x  1 root  wheel     790 Aug 25 15:57 nv_drv.la
-rwxr-xr-x  1 root  wheel  229324 Aug 25 15:57 nv_drv.so
-rwxr-xr-x  1 root  wheel     813 Jul 25 15:56 voodoo_drv.la
-rwxr-xr-x  1 root  wheel   30493 Jul 25 15:56 voodoo_drv.so
-rwxr-xr-x  1 root  wheel     813 Jul 25 15:55 vmware_drv.la
-rwxr-xr-x  1 root  wheel   56701 Jul 25 15:55 vmware_drv.so
-rwxr-xr-x  1 root  wheel     795 Jul 25 15:54 via_drv.la
-rwxr-xr-x  1 root  wheel  266265 Jul 25 15:54 via_drv.so
-rwxr-xr-x  1 root  wheel     795 Jul 25 15:53 vga_drv.la

	Huh?! The drivers are updated as of EXACTLY one month
and 1 minute later?!?!?!?!!? I don't remember doing this for
any reason. 

	Should I re-install xf86-video-nv-2.1.3 to make sure they
weren't corrupted somehow? Is there something else I should do?

		Thanks, Tuc

More information about the xorg mailing list