Usual play with transcode and xawtv.
Then usual oops but *before that* something which hasn't happened before.
Slab corruption: start=c57c2000, len=4096
000: 6e 72 6d 71 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b
I also have a "cat /proc/slabinfo" taken *after* the oops.
Available on request because it's big (25k)
see atached dmesg file for more details.
Config, lspci ... atached
Bye
Calin
--
"A mouse is a device used to point at
the xterm you want to type in".
Kim Alm on a.s.r.
Calin wrote:
>Then usual oops but *before that* something which hasn't happened before.
>Slab corruption: start=c57c2000, len=4096
>000: 6e 72 6d 71 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b
>
>
Is that reproducable? The selftest code in slab noticed a change in a
buffer that is not in use: All values should be 0x6b, but the first 4
bytes are different.
Could you try to boot with CONFIG_DEBUG_PAGEALLOC enabled? Perhaps that
identifies who corrupts the page. Unfortunately it can't detect DMA writes.
>------------ kernel BUG at include/linux/mm.h:275!
>
put_page_testzero noticed that the page count was already 0. It seems
that someone uses memory that was already freed.
--
Manfred