On Sun, 18 Jul 2004, Harald Schmalzbauer wrote: > Am Sonntag, 18. Juli 2004 01:06 schrieb Robert Watson: > > On Sun, 18 Jul 2004, Harald Schmalzbauer wrote: > > > Yes, it's "cam_periph.c,v 1.56 2003/11/08 10:56:57 scottl" I didn't any > > > local modification! I just removed my obj directory but that didn't > > > help. CVsup done half an hour ago. Btw, can anybody exlain me what > > > timezone the timestamp is from? Shouldn't it be UTC (or at least include > > > the TZ code)? > > > > Can you check to see if your local copy contains an include of mutex.h? > > The above time stamp looks right, FYI. > > Ok, since nobody else seems to have this problem I started poking int > the dark and found the "error": When I remove #options INVARIANTS from > my kernel.config it doesn't stop at the cam any more. Now I have a pf > show stopper but I'm sure this will be fixed soon. Still odd on the CAM front. The patch I'm building with to get pf working is attached. I also had to remove a reference to "caddr_t sg" from a function prototype in the Linux compat socket code. Seems to be a bad kernel compile day. Index: pf.c =================================================================== RCS file: /data/ncvs/src/sys/contrib/pf/net/pf.c,v retrieving revision 1.13 diff -u -r1.13 pf.c --- pf.c 17 Jul 2004 17:15:15 -0000 1.13 +++ pf.c 17 Jul 2004 23:42:44 -0000 _at__at_ -5161,7 +5161,7 _at__at_ if ((m0 = m_copym2(*m, 0, M_COPYALL, M_NOWAIT)) == NULL) #endif return; - if ((mtag = m_tag_copy(mtag)) == NULL) + if ((mtag = m_tag_copy(mtag, M_DONTWAIT)) == NULL) goto bad; m_tag_prepend(m0, mtag); } else { _at__at_ -5444,7 +5444,7 _at__at_ if ((m0 = m_copym2(*m, 0, M_COPYALL, M_NOWAIT)) == NULL) #endif return; - if ((mtag = m_tag_copy(mtag)) == NULL) + if ((mtag = m_tag_copy(mtag, M_DONTWAIT)) == NULL) goto bad; m_tag_prepend(m0, mtag); } else {Received on Sat Jul 17 2004 - 22:37:35 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:02 UTC