Re: [exim] Another strange log about tainting

On 2019-12-15 11:31, Jeremy Harris wrote:

> > devuan linux "ascii", kernel 4.9.0-9-amd64, libc 2.24-11+deb9u4, exim
> > self-built from 4.93 source.
> >
> > Looking how is_tainted is implemented, I see that its answer on
> > constant strings would in any case depend on the order of heap
> > versus initialized memory virtual addresses, which doesn't sound
> > like the best thing to rely on.
>
> Yes, if Devuan plays odd tricks with memory layout the you'll need to
> define TAINT_CHECK_SLOW (for the whole build, so in a Makefile) to
> avoid the optimisation that other Linuxen seem to be able to use.

Color me an unbeliever that Devuan changes Debian on this level.

These (from my Local/Makefile) may be of interest, I wonder what the
result would be on other Linuxen using these flags.

EXTRALIBS = -lcorkipset -lcork
EXTRALIBS_EXIM = -lspf2 -lwrap -ldl -export-dynamic
DEFS = -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE -D_FORTIFY_SOURCE=2
WARNINGS = -Wdate-time -Wformat
CFLAGS = -O2 -fstack-protector-strong $(WARNINGS) $(DEFS)
LDFLAGS = -Wl,-z,relro

BTW, is there a test in the suite covering is_tainted()? Looks like
having one may be a good idea. If there isn't one I volunteer to come
up with a patch.

--
Please don't Cc: me privately on mailing lists and Usenet,
if you also post the followup to the list or newsgroup.
To reply privately _only_ on Usenet and on broken lists
which rewrite From, fetch the TXT record for no-use.mooo.com.