Add the move-aout-libs upgrade target so that people who have already
gone elf can put their libraries through the mincer. Anyone who hasn't
deleted aout libraries from /usr/lib (but has done a make world putting
the new aout libs in /usr/lib/aout) will be asked for confirmation
to delete them one by one.
Add a move-aout-libs target to do just that. It takes the well known
library directories: /usr/lib /usr/lib/compat /usr/local/lib and
/usr/X11R6/lib and adds any other directories that ldconfig knows
about, then removes any that are already aout directories (that is,
with aout as a component of the path) and passes this directory
list to the move_aout_libs.sh script to do the hack (believe it,
it moves libraries that are in use) work.
Some preliminary updates to the release notes, just as I think of
them. By no means a comprehensive list of all the things which should
eventually go in here before 3.0-RELEASE, nor are the CAM issues
reflected here yet.
Add a new library function getobjformat(). It checks all the
standard places ("/etc/objformat", ${OBJFORMAT}, argv) for an
indication of the user's preferred object file format. This
consolidates some code that was starting to be duplicated in more
and more places.
Use the new function in ldconfig.
Note: I don't think that gcc should use getobjformat(), even though
it could. The compiler should limit itself to functions that are
widespread, to ease porting and cross-compilation.
brian [Wed, 9 Sep 1998 00:03:09 +0000 (00:03 +0000)]
Initialise lcp::his_mru to the ``set mtu'' value if it's less
than DEF_MRU, allowing our interface mtu to be decreased
despite negotiation with the peer.
- If the OACTIVE flag is set on entry to xl_start(), check to see if the
transmitter is wedged. If so, try to unwedge it, process any descriptors
that might need to be free()d, then proceed.
- Disable the 'background' autonegotiation performed during bootstrap.
What happens currently is that the driver starts an autoneg session,
the sets a timeout in the ifnet structure and returns. Later, when the
timer expires, the watchdog routine calls the autoneg handler to check
the results of the session. The problem with this is that the session
may not complete until some point after we have started to mount NFS
filesystems, which can cause the mounts to fail. This is especially
troublesome if booting with an NFS rootfs: we need the interface up
and running before reaching the mountroot() code.
The default behavior now is to do the autoneg synchronously, i.e. wait
5 seconds for the autoneg to complete before exiting the driver attach
routine. People who want the old behavior can compile the driver with
XL_BACKGROUND_AUTONEG #defined. This has no effect on autoneg sessions
initiated by 'ifconfig xl0 media autoselect.'
This slows the probe down a little, but it's either that or botching
NFS mounts at bootup.
- If xl_setmode_mii() is called and there's an autoneg session in progress,
cancel it, _then_ set the modes.
Oops, don't assume that the environment is normal in devfs_mount().
It isn't for the hidden mountpoint. The static vfs's haven't been
attached then, so mp->mnt_vfc can't be valid.
If traps are set, they are now executed even when a signal-blocking
foreground child is running. Formerly, traps were exceuted after the
next child exit.
The enables the user to put a breaking wrapper around a blocking
application:
(trap 'echo trap ; exit 1' 2; ./pestyblocker; echo -n)
The "echo -n" after the child call is needed to prevent sh from
optimizing the trap-executing shell away. I'm working on this.
Go back to using MSDOSFS instead of Robert Nordier's stand-alone DOS
library. I have to include MSDOSFS on the boot floppy anyway (so I can
read DOS floppies in a few other contexts) and since both -stable and
-current MSDOSFS handle long filenames, I'm OK on that now. Should
save about 10K of space.
o Set paths internally for pkg_add for tools we might want to invoke.
o Fix bogus suffix handling.
o Tell user when an FTP url is being xferred rather than being silent.
This sort of violates "the unix way" but it stops people from whacking
^C when they think it's hung, too. Sometimes visual indication of
success is important. Doesn't spit out anything if not on a TTY.
Make profiling work for ELF. gprof now autodetects the format of
the executable file, so it will work for both a.out and ELF format
files. I have split the object format specific code into separate
source files. It's cleaner than it was before, but it's still
pretty crufty.
Don't cheat on your make world for this update. A lot of things
have to be rebuilt for it to work, including the compiler and all
of the profiled libraries.
brian [Mon, 7 Sep 1998 19:54:08 +0000 (19:54 +0000)]
Remove the extern decl of ``h_errno''. It's already
externed in netdb.h - where all of its #define values
live. If anything requires h_errno here (nothing in
/usr/src seems to) it's almost definitely broken.
Load linux emulator lkm only when not already loaded. This avoids
leftover files in /tmp. Script slightly modified from PR version
to use fewer processes.
PR: i386/7725
Submitted by: Stefan Eggers seggers@semyam.dinoco.de
Removed statically configured mount type numbers (MOUNT_*) and all
references to them.
The change a couple of days ago to ignore these numbers in statically
configured vfsconf structs was slightly premature because the cd9660,
cfs, devfs, ext2fs, nfs vfs's still used MOUNT_* instead of the number
in their vfsconf struct.
Fix problem reported on bugtraq: check permission of device mounted
for non-root users. Fortunately, the default for vfs.usermount is 0.
Tested by: "Jan B. Koum " <jkb@best.com
sef [Mon, 7 Sep 1998 05:49:43 +0000 (05:49 +0000)]
Use a default execution type if none of the listed ones match. This
uses the first type listed in the array as the default type. This isn't
perfect, but I thought it would be better than nothing.
Made unloading of the nfs LKM sort of work. This is mainly to test
detachment of vfs sysctls. Unloading of vfs LKMs doesn't actually
work for any vfs, since it leaves garbage pointers to memory
allocation control structures.
Better handling of word splitting. Don't record the same region
multiple times when performing nested variable expansion, and
preserve some quoting information in order to avoid removing
apparently empty expansion result.
Patch the transmit error handler to avoid following NULL pointers and
generating a trap 12 panic. The code blindly assumed that in the event
of a transmit error, the packet that caused the error would still be
at the head of the driver's transmit queue (sc->xl_cdata.xl_tx_head).
However in the case of error 82 (which indicates that a transmit error
occurred after part of the transmit FIFO memory has been reclaimed)
this is not true: the TX queue has already been flushed, and the
pointer to the head of the queue is NULL, so trying to dereference
the pointer to find the transmit descriptor address causes a crash.
The code now checks for a NULL pointer before trying to reload the
chip's download pointer register. There may still be error messages
printed warning of the transmit error, but no panic should occur.
Note that this eror code is only generated with "cyclone" chipsets
(3c900B, 3c905B, and presumeably the 3c980 server adapter). It should
only appear during periods of heavy traffic, probably only on
non-switched networks.
Problem reported by: Darcy Buskermolen <darcy@ok-connect.com>
RFC 1644 has the status "Experimental Protocol", which means:
4.1.4. Experimental Protocol
A system should not implement an experimental protocol unless it
is participating in the experiment and has coordinated its use of
the protocol with the developer of the protocol.
Change the ELF64_R_SYM() to shift 32 bits instead of 8 bits to match
binutils.
Ensure that three of the structures are the size that binutils writes
them. I just love code that doesn't share header files to avoid
problems like these.
Don't recognize a file as an a.out shared library unless it has at
least 2 version numbers. This fixes the bug where the dynamic
linker would try to load an ELF shared library if it found one.
Note, this change also fixes the same thing in "ld", because the
code is shared.
For "ld" there is still a problem with ".a" libraries, which cannot
be distinguished by name. I haven't decided what, if anything, to
do about that.
Ignore the statically configured vfs type numbers and assign vfs
type numbers in vfs attach order (modulo incomplete reuse of old
numbers after vfs LKMs are unloaded). This requires reinitializing
the sysctl tree (or at least the vfs subtree) for vfs's that support
sysctls (currently only nfs). sysctl_order() already handled
reinitialization reasonably except it checked for annulled self
references in the wrong place.