PANIC: blockable slep lock (sx) msi @ ....msi.c:374

Attilio Rao attilio at
Fri May 4 20:15:18 UTC 2007

Scott Long wrote:
> The real problem is that infrastructure code like MSI should not hold 
> locks across calls to consumer code.  That's what needs to be fixed, not
> hacks to WITNESS.

Actually, I think nobody proposed to hack WITNESS, mine was just a 
question about a related topic.
I'm still trying to figure out why msi currently uses an sx (I suspect 
there are some sleeping point).


More information about the freebsd-current mailing list