Bruce Evans [Sun, 13 Apr 1997 12:36:42 +0000 (12:36 +0000)]
Fixed synopsis. There was a missing arg, 3 args with the wrong type,
and many apparent dependendcies on <sys/types.h> from use of u_char.
<dialog.h> avoids u_char, so avoid it here too.
Add more bootstrapping bandaids: copy over some of the host system
dependent utilities to /bootstrap inside the chroot tree, so the
hosting and targeted environment are not required the same. By now,
mount/umount (due to the Lite2 changes) and cpio (due to lchown) make
it there. This should allow building a 3.0-current release from
inside a 2.2 vintage system, and visa verse.
Isn't lchown also a reason for a shared lib version number bump?
The function union_fsync tries to lock overlaying vnode object when
dolock is not set (that is, targetvp == overlaying vnode object).
Current code use FIXUP macro to do this, and never unlocks overlaying
vnode object in union_fsync. So, the vnode object will be locked
twice and never unlocked.
The path name buffer, cn->cn_pnbuf, is FREEed by VOP_MKDIR when
relookup() in union_relookup() is succeeded. However, if relookup()
returns non-zero value, that is relookup fails, VOP_MKDIR is never
called (c.f. union_mkshadow). Thus, pathname buffer is never FREEed.
John Dyson [Sun, 13 Apr 1997 04:07:24 +0000 (04:07 +0000)]
Decrease the amount of memory allocated for bouncing. This will
allow large systems to boot successfully with bounce buffers compiled
in. We are now limiting bounce space to 512K. The 8MB allocated for
a 512MB system is very bogus -- and that is now fixed.
John Dyson [Sun, 13 Apr 1997 03:35:30 +0000 (03:35 +0000)]
The pmap code was too generous in the allocation of kva space for
the pv entries. This problem has become obvious due to the increase
in the size of the pv entries. We need to create a more intelligent
policy for pv entry management eventually.
Submitted by: David Greenman <dg@freebsd.org>
John Dyson [Sun, 13 Apr 1997 03:33:25 +0000 (03:33 +0000)]
Improve the buffer cache memory policy by moving pages over to the
cache queue more often. The pageout daemon had to be waken up
more often than necessary since pages were not put on the
cache queue, when they should have been.
Submitted by: David Greenman <dg@freebsd.org>
John Dyson [Sun, 13 Apr 1997 01:48:35 +0000 (01:48 +0000)]
Fully implement vfork. Vfork is now much much faster than even our
fork. (On my machine, fork is about 240usecs, vfork is 78usecs.)
Implement rfork(!RFPROC !RFMEM), which allows a thread to divorce its memory
from the other threads of a group.
Implement rfork(!RFPROC RFCFDG), which closes all file descriptors, eliminating
possible existing shares with other threads/processes.
Implement rfork(!RFPROC RFFDG), which divorces the file descriptors for a
thread from the rest of the group.
Fix the case where a thread does an exec. It is almost nonsense for a thread
to modify the other threads address space by an exec, so we
now automatically divorce the address space before modifying it.
Brian Somers [Sun, 13 Apr 1997 01:07:00 +0000 (01:07 +0000)]
Allow forcing of local_ip:remote_ip in pap-secrets
and chap-secrets files. This allows specific users
to have pre-allocated IP numbers while others get
assigned dynamically.
Submitted by: David Nugent <davidn@unique.usn.blaze.net.au>
Brian Somers [Sun, 13 Apr 1997 00:54:45 +0000 (00:54 +0000)]
Make -background option attempt each phone number at most
once. Make -background ignore redial. Output exit value
to syslog with "PPP Terminated" message.
Brian Somers [Sat, 12 Apr 1997 22:58:41 +0000 (22:58 +0000)]
Only wait for the redial timeout when the last phone number in the
list has been dialed. Alternate number dialing has no "pause".
Suggested by: joerg
Document this behaviour. Document that the number of dial attempts
applies to the number of phone calls rather than the number of times
each number is dialed. Add a missing .El. Give a decent description
of how to connect to an ISP.
Minor fixups. Mention that DDB now has a history, and understands
emacs-style editing (which it used to have for a long time already).
Also mention the `gdb' and `help' commands. Other commands need an
overhaul, too (like the various `show' subcommands), but i don't feel
very competent for these.
Everyone's favorite, i think: make DDB understand the arrow keys for the
basic cursor movements. Assumes ANSI/DEC tty, but you can still resort
to plain emacs ^p/^n etc anyway.
Peter Wemm [Sat, 12 Apr 1997 07:09:18 +0000 (07:09 +0000)]
Make tclsh conditional on the src/contrib/tcl directory existing, and
allow it to be disabled via NOTCL in /etc/make.conf (similar to other
stuff in src/Makefile, eg: NOGAMES)
Peter Wemm [Sat, 12 Apr 1997 07:06:08 +0000 (07:06 +0000)]
Put on my flame proof suit and make libtcl build conditional on
the src/contrib/tcl directory existing, and also have an /etc/make.conf
override (NOTCL) to stop building libtcl. This is in similar other
things from src/Makefile, eg: NOGAMES, NOLKM, etc.
This is so that people can put in a refuse entry in their cvsup files and
not fetch the tcl code, and have it not built automatically. I'll do
something similar for perl.
Brian Somers [Sat, 12 Apr 1997 04:23:13 +0000 (04:23 +0000)]
Support input and output filters with remote printing.
Output filters are executed on a per-file basis as it's
necessary to supply the file size to the "other side".
Bruce Evans [Fri, 11 Apr 1997 17:55:19 +0000 (17:55 +0000)]
Cleaned up synopsis:
- don't quote function names, since quoting them is unnecessary and
unusual and confuses my synopsis checker.
- include <sys/types.h> instead of <sys/param.h>. It is normal to
(have to) include <sys/param.h> instead of <sys/types.h>, but it
is more useful for man pages to document minimal prerequisites.
Bruce Evans [Fri, 11 Apr 1997 17:08:56 +0000 (17:08 +0000)]
Don't print nonexistent library names for `ld -f [-Lfoo] -lbar'. A
dependency on `bar' is very unlikely to be correct.
This is a quick fix for broken dependencies in gdb and many other
places. The dependencies on internal libraries are now missing
instead of wrong when `make depend' is run before the libraries
are created.
Deleted ddb_inb and ddb_outb functions that provide I/O access
routines as function for DDB. The inb and outb are provided as
functions in machdep.c when DDB is defined.
Bruce Evans [Thu, 10 Apr 1997 16:14:09 +0000 (16:14 +0000)]
Fixed bogus existence test related to searching for a nearby obj
directory. manpath.1 is always in the current (= object) directory,
so don't search for it.
Bruce Evans [Thu, 10 Apr 1997 16:13:17 +0000 (16:13 +0000)]
Fixed bogus existence test related to searching for a nearby obj
directory. man.1 is always in the current (= object) directory,
so don't search for it.
Bruce Evans [Thu, 10 Apr 1997 16:04:43 +0000 (16:04 +0000)]
Fixed bogus existence test related to searching for a nearby obj
directory. config.h is always in the current (= object) directory,
so don't search for it.
config.h is not a source for the library, so don't put it in SRCS and
don't make the library depend on it.
Bill Paul [Thu, 10 Apr 1997 14:12:51 +0000 (14:12 +0000)]
Add a sanity test to ypproc_xfr_2_svc(): check that the map and domain
in the transfer request actually exist. Technically ypxfr can do this too,
but why waste the cycles getting ypxfr off the ground for a transfer we
already know is going to fail.
Also apply stricter access control rules; ypproc_xfr_2_svc() is in a
different class than the normal map access procedures procedures.
Bruce Evans [Thu, 10 Apr 1997 13:54:19 +0000 (13:54 +0000)]
Deprecated <sys/dir.h> some more - changed the comment saying that
<dirent.h> should be used instead to a warning. If this causes too
many warnings in ports then it should be changed back after checking
some ports for related configuration errors.
Moved the definition of DIRSIZ() from <sys/dir.h> to <sys/dirent.h>
so that it can be used in the kernel without including <sys/dir.h>.
Renamed it in some cases to avoid new namespace pollution.
Bruce Evans [Thu, 10 Apr 1997 13:17:09 +0000 (13:17 +0000)]
Use smalllblktosize() instead of multiplying small block numbers
by fs->fs_bsize. The macro is usually faster and makes it clearer
that the multiplication can't overflow.
John Polstra [Wed, 9 Apr 1997 20:31:15 +0000 (20:31 +0000)]
Use "-pg" instead of "-p" for generating profiled object files.
The two options differ only when linking, but "-p" is incorrect
because it is associated with the old-style "prof" command, which we
do not support.
John Polstra [Wed, 9 Apr 1997 19:14:31 +0000 (19:14 +0000)]
Declare the constructor/destructor linker sets as extern rather than
common. Add one do-nothing element to each set. This ensures that
the linker realizes that they are linker sets rather than simple
commons, and makes it possible to link c++rt0.o into every shared
library regardless of whether it is a C++ library or not. Without
this change, the constructors and destructors in the main program
could be executed multiple times.
This change is going to make it possible to get rid of the
CPLUSPLUSLIB makefile variable once and for all. It is a piece of
the solution to PR gnu/3505 (gcc -shared). Finally, it fixes a
heretofore unreported bug: If CPLUSPLUSLIB was set in a makefile
for a C++ shared library that had no static constructors or
destructors in it, then the main program's constructors and
destructors would be executed multiple times.
John Polstra [Wed, 9 Apr 1997 18:59:36 +0000 (18:59 +0000)]
Process "lib/csu/i386" at the beginning of the "libraries" target,
instead of in the middle of it. Soon, "c++rt0.o" will be linked
into all shared libraries. An up-to-date version of it must be in
place before any shared libraries are built.
This is one piece of the solution for PR gnu/3035 (gcc -shared).
Ignore SIGPIPE or any user can quit ppp just telnetting to it
and quitting telnet immediately (while phone number dialed)
Log client connection/disconnection with PHASE_BIT now.
Add more error recovery on client disconnection
Bruce Evans [Wed, 9 Apr 1997 16:53:45 +0000 (16:53 +0000)]
Removed support for OLD_PIPE. <sys/stat.h> is now missing the hack that
supported nameless pipes being indistinguishable from fifos. We're not
going back.
Bruce Evans [Wed, 9 Apr 1997 16:10:27 +0000 (16:10 +0000)]
Use `ld -f' (actually `cc -Wl,-f') at `make depend' time to generate
almost perfect dependencies on crt0's and libraries. DPADD and
bsd.libnames.mk should go away soon. Use a new _EXTRADEPEND target
to implement this and to avoid editing of .depend when .depend isn;t
being rebuilt. The afterdepend target doesn't seem to be good for
anything and is now unused.
Fixed LDDESTDIR for the DESTDIR case when ${SHLIBDIR} != /usr/lib.
Added commented-out -nostdlib to LDDESTDIR for the DESTDIR case.
The wrong libraries may be used without this; however it breaks
linkage to crt0 and libc.