kernel using gcc-5.3 fails to boot right

Adrian Chadd adrian at freebsd.org
Mon Aug 15 06:39:05 UTC 2016


Hi,

So to follow follow up - it looks like there's some reasonably
interesting garbage behaviour going on. Adding printfs changes the
values of those 64 bit variables being printed, and the loop
behaviour.

eg:

ipfw2 initialized, divert loadable, nat loadable, default to deny,
logging disabled^M
CALLING add_sopt_handler: from
/usr/home/adrian/work/freebsd/head-embedded/src/sys/modules/ipfw/../../netpfil/ipfw/ip_fw_sockopt.c:3388^M
ipfw_add_sopt_handler: called; sh=0xc021dad0, count=11^M
ipfw_add_sopt_handler: called; gencnt=1969842025, ctl3_hsize=0, sz=11,
count=11, (176 bytes)^M
ipfw_add_sopt_handler: retrying; gencnt=8388358300892480083, tgen=1969842025^M
ipfw_add_sopt_handler: called; gencnt=1969842025, ctl3_hsize=0, sz=11,
count=11, (176 bytes)^M
ipfw_add_sopt_handler: retrying; gencnt=8388358300892480083, tgen=1969842025^M
ipfw_add_sopt_handler: called; gencnt=1969842025, ctl3_hsize=0, sz=11,
count=11, (176 bytes)^M
ipfw_add_sopt_handler: retrying; gencnt=8388358300892480083, tgen=1969842025^M
ipfw_add_sopt_handler: called; gencnt=1969842025, ctl3_hsize=0, sz=11,
count=11, (176 bytes)^M

... all I'm doing is printing, and tgen is the value of gencnt when
the loop starts.

So, something smells fishy.


-adrian


More information about the freebsd-mips mailing list