Commit Graph

87153 Commits

Author SHA1 Message Date
lukem 5b5d0d23b8 oops, remove temporary comment 2001-11-01 08:21:57 +00:00
lukem 7f675a25b7 fix -Wshadow warnings 2001-11-01 08:21:07 +00:00
lukem 3a17632dfb remove arg name in prototype decl 2001-11-01 08:18:01 +00:00
lukem 7623928b3a fix a couple of -Wshadow warnings 2001-11-01 08:06:57 +00:00
lukem 4c4307e3ce fix -Wshadow warnings 2001-11-01 08:03:03 +00:00
lukem fa02ec4ccb fix a -Wshadow warning 2001-11-01 07:48:08 +00:00
lukem 3d6fc56715 fix -Wshadow warning by moving "int version" from global to main() scope 2001-11-01 07:44:05 +00:00
lukem d3656d428d fix a couple of -Wshadow warnings 2001-11-01 07:39:38 +00:00
chs fa6e18a029 in pmap_extract(), detect unmapped users addresses too. 2001-11-01 07:37:36 +00:00
lukem 4475e1747e fix a couple of -Wshadow warnings 2001-11-01 07:31:22 +00:00
lukem 2d79659157 move -Wwrite-strings to WARNS=3 too ... (yech!) 2001-11-01 07:27:35 +00:00
lukem 0e643f61a5 Move -Wcast-qual from WARNS=2 to WARNS=3 (a new level), because it
generates too many false positives that require gross amounts of
workaround, and the other WARNS=2 stuff is quite useful. Per discussion
with simonb & christos.
2001-11-01 07:17:17 +00:00
lukem 058fbb840d opendisk(3) checks the device type for us 2001-11-01 07:04:18 +00:00
lukem b3bf0bffe1 Enforce is iscooked arg;
if zero, a matching path must be of type S_IFCHR, and
	if non-zero, a matching path must be of type S_IFBLK.
2001-11-01 06:53:24 +00:00
chs 550caf0ce3 allow SWAP_GETDUMPDEV for all users.
use {LIST,TAILQ}_FOREACH where appropriate.
2001-11-01 03:49:30 +00:00
imp fdf7743003 Add Intel PRO WLAN 2011. 2001-11-01 03:00:51 +00:00
tv 70a4e4b481 Revert previous until it's fixed a more generic way ... I didn't realize
this was pulled in via reachover from elsewhere.
2001-11-01 02:15:41 +00:00
thorpej d8c2abc4aa Make *really* sure that MACHINE is set to an empty string so that
it doesn't infect the linker script build (i.e. override any MACHINE=
that might already be in MAKEFLAGS).
2001-11-01 01:52:45 +00:00
thorpej a4a68fe775 When doing an only-tools build, don't use the makewrapper. 2001-11-01 01:49:44 +00:00
wiz 496a336cb0 Slight markup change. 2001-11-01 01:15:42 +00:00
wiz 73680f8d68 Drop a spare .Pp. 2001-11-01 01:13:43 +00:00
wiz 27bebba05c Sort SEE ALSO, drop some unnecessary .Pp's, whitespace and misc fixes. 2001-11-01 01:10:59 +00:00
thorpej f6614f977b Document build.sh -t 2001-11-01 00:14:45 +00:00
thorpej f55ffb9482 Add a -t option that builds and installs the toolchain only
(and thus implies -b).  To make this work, we also check do_buildsystem
when doing the DESTDIR sanity check.
2001-11-01 00:14:17 +00:00
tv 44c7a1639c Document new build.sh options, and remove the temporary note on USETOOLS
("no" functions as documented now).
2001-10-31 23:20:34 +00:00
reinoud 1ee1d4a7ad Use the -F writing to a normal file feature of newfs to avoid problems with
rewriting the disklabel. The newfs is moved up in the sequence after the
creation of the file and the disklabel is written after it.
2001-10-31 22:30:47 +00:00
fvdl f8ff451a0a Fix reversed wait/nowait logic. 2001-10-31 22:23:18 +00:00
thorpej 44f1c938e8 Don't actually need elf64 support in installboot(8). 2001-10-31 22:13:03 +00:00
christos 961acd7fc2 PR/10266: t_getstr() leaks memory. This PR will stay in feedback
until the problem gets addressed properly. The following fix
is a stopgap measure to stop the leaking :-(

I fixed the t_getstr() memory leak problem, but that instantly
revealed a problem in t_agetstr() which is an extremely broken
interface. It realloc's memory, potentially moving the area where
it returned pointers into in previous calls. This function needs
to be removed and or changed. I added a horrible work-around for
now, but I will revisit the problem shortly. In the meantime nobody
should be using the t_agetstr() API, and I'll be fixing the rest
of the programs and or the API when I figure out the best solution...
This is t_agetstr() is used by:

	games/hack/hack.termcap.c
	games/larn/io.c
	games/tetris/screen.c
	lib/libterm/termcap.c
	lib/libterm/termcap.h
	libexec/getty/main.c
	usr.bin/top/screen.c
	usr.bin/ul/ul.c
2001-10-31 21:52:17 +00:00
thorpej 66f972ab70 Need byteorder.c for loadfile_elf*.c 2001-10-31 21:39:02 +00:00
thorpej 0971f93345 Make the ELF loadfile routines byte-order independent when used in
a non-_STANDALONE environment (e.g. installboot(8)): internalize and
externalize the exec, program, and section headers as necessary.

Reviewed and OK'd by Christos.
2001-10-31 21:24:09 +00:00
jdolecek d58125dd1f kill one more duplicate 2001-10-31 21:15:43 +00:00
tv a988fe0a8a Revert previous. src/include depends on the ability of <bsd.inc.mk> to
provide things to INCS with / in them.  Since nothing in-tree yet uses
INCS{DIR,NAME}_filename, this is not a loss at the moment.
2001-10-31 21:15:41 +00:00
jdolecek f15ae836f3 kill duplicates 2001-10-31 21:13:23 +00:00
jdolecek d66dfc84d0 Xref disklabel and mention that mbrlabel can help setup the disk label
Note that NTFS partition is no longer accessible in NetBSD once
it's marked 'dynamic' in MS Windows XP.

Pointed out by Christos.
2001-10-31 21:04:45 +00:00
tv 01523890ba No need to do the tricky XRPCGEN dance. 2001-10-31 20:58:28 +00:00
tv 89bc80292e No need to do the tricky XRPCGEN dance; simply add the rules after
.include <bsd.lib.mk>.  Defining the explicit compile rules later is fine.
2001-10-31 20:56:09 +00:00
tv 8e97329393 Don't try to do uid or gid lookups from a supplied spec file if -W is given;
the numerical IDs won't be used anyway.
2001-10-31 20:41:17 +00:00
jdolecek 159a2b871c use right number of zeroes for Cyrix i386_cpuid_cpus[] entries
reformat the entries with all zeroes to be more easier to verify
2001-10-31 20:35:21 +00:00
thorpej a43b4351a8 Oops, committed wrong copy of file in previous revision. 2001-10-31 20:22:22 +00:00
thorpej 90aee7593a Add standalone 64-bit to/from {big,little}-endian routines. 2001-10-31 20:19:52 +00:00
atatat 6e885d365a Turn the tun device/network interface into a cloning device. 2001-10-31 20:08:17 +00:00
tv dadd6e7370 Major cleanup and overhaul:
* Allow MACHINE to be set automatically on NetBSD native builds.
  Require -m only if the host is not NetBSD.

* Fail if DESTDIR is set to root (/) and the build is not NetBSD native,
  or the build is attempting a "make release" (option -R).

* Warn the user if DESTDIR is set to root (/) that the kernel must be up
  to date, or else the build might hose the system.  Also, do not remove
  DESTDIR on -r if it is set to root.

* Fail if TOOLDIR is set to / or empty after checking with nbmake.

* Allow DESTDIR and TOOLDIR to be set in mk.conf and/or by the new default
  from <bsd.own.mk>.  Note that if -T is not used to set TOOLDIR, and
  TOOLDIR is not set in the environment, then nbmake will always be
  bootstrapped (rather than looking at nbmake's timestamp).  This is
  because /bin/sh doesn't know how to get TOOLDIR from mk.conf without
  first having nbmake (...which lives in TOOLDIR).

* Do a pass of "make obj" through src/tools before installing nbmake (so
  long as -o/MKOBJDIRS=no is not specified).  This ensures that objdirs
  exist if they are desired, and paves the way for using build.sh to
  build all the src/tools without building everything else.

* Add "-w" option, which allows the builder to specify where the nbmake
  wrapper shell script should be created.  If not specified,
  $TOOLDIR/bin/nbmake-$MACHINE is still created as a default.

* Print the values of DESTDIR and TOOLDIR after bootstrapping nbmake,
  for informational purposes.

It should now be possible to type just "./build.sh" on a NetBSD-current
host and get a new build in / for the appropriate architecture.
However, building in-place like this has *not* been extensively tested
yet, so be careful.
2001-10-31 19:59:43 +00:00
gmcgarry 1d6a2b23b9 vn_marktext.9 becomes vn_markexec.9 2001-10-31 19:57:18 +00:00
gmcgarry 8e84a660f0 Document new VEXECMAP vnode flag and rename vn_marktext() to vn_markexec(). 2001-10-31 19:53:07 +00:00
jmc 100b449518 Add back logic which forces DESTDIR=/ for the distrib-dirs rule if it's not set
or it's blank. The new toolchain rules set it to blank which can run into
the sanity checks in etc/Makefile
2001-10-31 19:41:54 +00:00
tv a75f4fb35e Don't actually delete the default tooldir on "cleandir"; else we'll nuke
the nbmake that is probably currently running.

Just leave the "cleantools" manual target as a shortcut for the user.
2001-10-31 19:38:02 +00:00
jdolecek 1295376b01 Couple changes, mainly to address stuff raised recently on port-i386
('Does i386 support NTFS5 now?'):

add a note both NTFS4 and NTFS5 are supported
reformat the Limitations a bit
assert the write support is not really useful, and should not be used
add a BUGS section, note missing stuff for the write support to be useful,
and the disk label thing spotted by Christos Zoulas
2001-10-31 19:16:36 +00:00
skrll 04b6a0a047 You can build libraries with more than just cc(1) now you know. 2001-10-31 18:44:11 +00:00
jdolecek 76dc75c365 svr4 compat LKMs work on i386 2001-10-31 18:30:18 +00:00