7-stable: linux_dist-gentoo-stage3 wonn't bootstrap

[Arno J. Klaassen]

Roman Divacky <rdivacky at freebsd.org> writes:

> On Sat, Dec 27, 2008 at 09:55:05PM +0100, Arno J. Klaassen wrote:
> > 
> > Hello,
> > 
> > I'm having serious problems to get MedInria (a medical imaging
> > toolchain) run under Linux emulation
> > ( http://gforge.inria.fr/frs/?group_id=727&release_id=2488
> > if ever someone would like to try : at my place even the
> > simplest graphical action (change tool, open file, etc) 
> > literally exploses the X-server ...)
> > 
> > so I thought to run it in debugger etc in the
> > chroot gentoo : but its bootstrap.sh will not complete,
> > for the time being because of :
> > 
> > 
> >  - compat.linux.osrelease 2.4.2 is not supported
> > 
> >  - so I installed fc6 and set compat.linux.osrelease to 2.6.16
> >    but then the chroot complains with many errors like :
> > 
> >     - (find): syscall fstatat64 not implemented
> >     - (rm): syscall unlinkat not implemented
> >     - (tar): syscall futimesat not implemented
> >     - (chmod): syscall fchmodat not implemented
> > 
> > So :
> >   - does someone have a patch with these at-functions
> >     (or the svn command to create it)
> 
> that would not be easy.... but 8-current implements those

OK, but the maybe the linux_dist-gentoo-stage3 should be marked
BROKEN on <8
 
> >   - or can I even bluntly copy the linuxulator from head
> >     to releng_7 ?
> 
> no... I doubt that would work.
> 
> 
> try 8-current and stick with that if it works

OK, I upgraded this box to 8-current; now the bootstrap.sh
'hangs' when compiling gconv_simple.c (I can provide
whole command line ) :

  top(1) shows :

   17547 root          1  76    0  4640K  4048K ppwait   0:05  0.00% gmake
   23865 root          1  76    0  4596K  4176K ppwait   0:02  0.00% gmake
   23899 root          1  76    0 18152K 16808K pipewr   0:01  0.00% cc1
   15076 root          1  76    0  1672K   872K wait     0:00  0.00% sandbox
   23898 root          1  76    0  2296K  1428K ppwait   0:00  0.00% i486-pc-linux

  and dmesg(1) says :

  t_delta 15.fcd4d7d810818e80 too short
lock order reversal:
 1st 0xffffff0024436ba8 pseudofs (pseudofs) @ /raid1/bsd/src-current/sys/kern/vfs_vnops.c:531
 2nd 0xffffffff80729360 sysctl lock (sysctl lock) @ /raid1/bsd/src-current/sys/kern/kern_sysctl.c:1088
KDB: stack backtrace:
db_trace_self_wrapper() at db_trace_self_wrapper+0x2a
_witness_debugger() at _witness_debugger+0x65
witness_checkorder() at witness_checkorder+0x859
_sx_xlock() at _sx_xlock+0x55
kernel_sysctl() at kernel_sysctl+0xbc
linprocfs_docpuinfo() at linprocfs_docpuinfo+0x84
pfs_read() at pfs_read+0x53c
vn_read() at vn_read+0x267
dofileread() at dofileread+0xa1
kern_readv() at kern_readv+0x60
read() at read+0x54
ia32_syscall() at ia32_syscall+0x1ab
Xint0x80_syscall() at Xint0x80_syscall+0x60
--- syscall (3, Linux ELF32, read), rip = 0x2812ae7e, rsp = 0xffff9c24, rbp = 0xffff9c38 ---


NB, linprocfs is mounted on /compat/linux/proc as well
as /usr/local/gentoo-stage3/proc

Let me know if you need more info; I'll leave the box in this state
for a while

Arno