Change the ELF registration/unregistration scheme to be less error prone.
Adding a new brand requires a single addition to linux_brandlist instead of
modifying linux_load(), linux_unload(), and linux_elf_init().
Approved by: jkh
Reviewed by: msmith
MD4Update(), but our version in libmd expects a byte count.
This code is not currently compiled or linked into pppd, so I'm
reasonably sure I didn't break anything. :-) I added the necessary
statements to the Makefile, but left them commented out because we
are in feature freeze. When the code is enabled, we must be careful
to build it only if the DES library is available.
Formerly, the heuristic involving the interpreter path took
precedence.
Also, print a better error message if the brand is missing or not
recognized. If there is no brand at all, give the user a hint that
"brandelf" needs to be run.
Remove /sys/boot from legacy-build.
Add btxld to build-tools.
In src/sys/Makefile:
Add /sys/boot for i386 ELF.
I'm still not sure why the new boot code was being built along with the
legacy stuff, which meant a completely wrong default environment for it.
This may well still be the wrong way to go about this, but it can't work
all that much worse than it has been.
o make install ; make install now works
o make all ; make all is quiet the second time
o Dependancies are properliy debugged; this means that make -jN has a
far hihjer likelyhood of working.
o a proper 'link farm' has been constructed for the build. This
dramatically simplifies the dependancy tangle.
o for perldoc's use, all the .pod files are installed.
o the man3 docs are properly compressed.
o the man pages and libary code are installed by the makefiles, not
by a perl script.
o at the end, h2ph is run.
- the directory was wrong if ${SHLIBDIR} != ${LIBDIR}. It's still wrong
if the installation of the obsolete library was done before /aout was
appended to LIBDIR.
- the version would have become wrong when the default in ../Makefile.inc
is changed from 2.0.
- the comment mostly described moving of libraries to /usr/lib/compat, but
we don't do that.
difference, but might later on when we implement some sort of multi-head
console mode. Select a console after probing them all.
Don't strdup a potentially NULL return from getenv().
If we don't select an active console, choose the first regardless.
Call the console init function, at startup time and on a manual change.
The env_setenv() function needs EV_VOLATILE because it's pointing to
data that isn't malloc'ed and will cause a fault if it's freed later.
Cosmetic change to the init-time character eater (like, make it increment
the index counter - if there's a problem, it would sit there in an infinite
loop instead of only running 10 times).
Also, make sure we set %dx each time around otherwise the commands
suddenly start trying to work on things like com92 instead of com1.
Make sure comc_init() is only run once.
Cosmetic change to init-time character eater.
Prompted by: bde
Also. Don't build & install legacy lkm's when NOLKM has been set.
Otherwise it gets built in the src tree rather than obj, because the
previous 'make obj' at the start of legacy-build does respect NOLKM.
fxp_stop is called as the first thing in fxp_init, and if the tx desc
list has junk in it, the system may panic. This bug showed up as a side
effect of the changes in rev 1.56, but has been in the code since the
beginning.
shared versions should not need to add -lm unless the program uses libm
itself. Strictly speaking, libg++ depends on libstdc++, but libstdc++
has dynamic dependencies on some exception tables and binutils doesn't
seem to like it when a secondary library has undefined references. It
doesn't seem to care when -lstdc++ is added on the command line to ld
though. Anyway, the c++ driver adds -lstdc++ explicitly, so that should
be OK. c++ also adds -lm explicitly too, even though it wouldn't need
to now. [except for statically linked binaries as .a files don't have an
equivalent of automatic internal shared object dependencies.]
Linux and FreeBSD signal numbers. Also, check signal numbers passed
in from application programs for validity. Without these checks,
it is trivial to panic the system from a Linux program.
in the wrong places for a while.
Also, the the libtermlib.so -> libtermcap.so manually for elf, otherwise
the hard link follows the symlink and the result looks rather wierd. The
*.a files are still hard linked under elf as before.
have been linked against it. Try and clean up the leftovers. Also, put
the a.out libs in /usr/lib/compat/aout since that's where the default
a.out ldconfig compat path points to.
vfork() can't be used. We could use alloca() in execl() so that
it can be called between vfork() and execve(), but a "portable"
popen() shouldn't depend on this. Calling execle() instead of
execl() should be fairly safe, since execle() is supposed to be
callable from signal handlers and signal handlers can't call
malloc(). However, execle() is broken.
popen(), but worse. The child calls execvp(), which calls malloc()
a bit more than execl(), and it calls non-library functions that call
malloc() and who-knows-what else (stdio is called in at least some
error cases).
temporary file names were uninitialized if TMPDIR was set and 1 too
small otherwise.
Fixed style bugs in previous commit.
Fixed missing checks for malloc failure in previous commit.
Report malloc failure consistently, at least in temp.c.
