On Sat, 17 Jun 2000, Matthew Dillon wrote:
> ...
> It looks like sccnattach() is calling scvidprobe() and scvidprobe()
> is then:
>
> 0xc023484b <scvidprobe+7>: cmpl $0x0,0x10(%ebp)
> 0xc023484f <scvidprobe+11>: setne %al
> 0xc0234852 <scvidprobe+14>: movzbl %al,%eax
> 0xc0234855 <scvidprobe+17>: push %eax
> 0xc0234856 <scvidprobe+18>: call 0xc0229c90 <vid_configure> <----****
> 0xc023485b <scvidprobe+23>: push %ebx
> 0xc023485c <scvidprobe+24>: push $0xc0282682
> 0xc0234861 <scvidprobe+29>: call 0xc0229b4c <vid_find_adapter>
>
> Calling vid_configure and vid_configure is dying.
>
> The list is generated from a linker_set ... one of those special linker
> lists.
>
> Something's broken the list.... maybe the DATA_SET macro is something ilke
> that.
Dependencies for setdef.h were broken by removing ioconf.o from ${OBJS}.
Previously, any change in the configuration resulted in a change to
ioconf.c, so setdefs.h got rebuilt. Now, removing whole drivers doesn't
change anything in ${OBJS}, so the old setdefs.h gets used and the wrong
amount of space is allocated for some linker sets. This bug cost me a
few hours when I tried removing console drivers in order to debug the
console initialization bugs.
This quick fix also backs out rev.1.180:
diff -c2 Makefile.i386~ Makefile.i386
*** Makefile.i386~ Sun Jun 18 21:22:28 2000
--- Makefile.i386 Sun Jun 18 21:31:50 2000
***************
*** 173,178 ****
${NORMAL_C}
! setdef0.c setdef1.c setdefs.h: ${OBJS} param.o
! @gensetdefs ${OBJS} param.o
# this rule stops ./assym.s in .depend from causing problems
--- 172,177 ----
${NORMAL_C}
! setdef0.c setdef1.c setdefs.h: Makefile ${OBJS}
! @gensetdefs ${OBJS}
# this rule stops ./assym.s in .depend from causing problems
Bruce
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-current" in the body of the message
