Commit Graph

92052 Commits

Author SHA1 Message Date
ross
d750bb9154 TARGETS is defined by bsd.own.mk, so use LOCALTARGETS instead 2002-02-06 23:29:42 +00:00
jdolecek
104f7c34d6 joyvar.h is now at <dev/ic/joyvar.h> 2002-02-06 21:32:16 +00:00
thorpej
d17cc4f043 * Move some generic ARM OpenFirmware stuff into arch/arm/ofw (still
more can move, but not yet).
* Garbage-collect some cruft from arch/arm32 in preparation for
  renaming to arch/shark.
2002-02-06 21:30:25 +00:00
wiz
6c8b3885d6 Some nits. 2002-02-06 21:18:51 +00:00
jdolecek
cf523d6e7c Move the joy(4) manpage to MI location, too - it's not i386-only.
Pointed out by Thomas Klausner, thanks :)
2002-02-06 21:06:32 +00:00
thorpej
886eff9ae7 Rename to reflect the MACHINE_ARCH that will be used on ARM ELF
platforms.
2002-02-06 20:39:00 +00:00
thorpej
535e775dd5 MACHINE_ARCH for ELF ARM platforms is "arm", not "arm32". 2002-02-06 20:36:18 +00:00
kleink
3b7a9506f7 Printf the 601's HID0. 2002-02-06 20:00:48 +00:00
thorpej
690f10fea7 Use arm/conf/Makefile.arm, and add support for building ELF kernels
and converting them to a.out.
2002-02-06 19:59:41 +00:00
kleink
69e30815cf Add a printf bitmask for HID0. 2002-02-06 19:59:30 +00:00
thorpej
c9a1155bcf Adjust for new location of elf2aout.sh, and echo the command before
executing it (like we do for other commands).
2002-02-06 19:59:01 +00:00
wiz
2e56fbc7e0 Another <tab> -> <sp><tab> noted by Patrick Welche. 2002-02-06 19:58:02 +00:00
thorpej
5113cbfb17 A slightly cleaned up version of chris/nick's elf2aout.sh for ARM
platforms.
2002-02-06 19:54:47 +00:00
wiz
7d3c2b8400 Drop some trailing whitespace. 2002-02-06 19:29:06 +00:00
wiz
c0acf9a5ea Punctuation nits, sort sections, sort SEE ALSO. 2002-02-06 19:24:52 +00:00
wiz
eae645b02b Fix some typos, uppercase tcp. 2002-02-06 19:17:01 +00:00
wiz
c36eec1c26 Fix .Dd (no abbreviations, please). 2002-02-06 19:09:22 +00:00
wiz
eb0430f0ee Fix a typo, and correct an abbreviation (Mhz -> MHz). 2002-02-06 19:07:27 +00:00
thorpej
eb79ee01a8 ip_mloopback(): process the delayed checksum on the copy, not
the original mbuf.
2002-02-06 18:00:01 +00:00
thorpej
da13cb2fb5 Back out all the vm_page_md changes. They are causing some
mysterious problems (a similar change to the i386 pmap causes
mysterious problems there, as well), and the issue needs to
be investigated more.
2002-02-06 17:41:42 +00:00
thorpej
4611193917 Efficiency tweaks, some made possible by vm_page_md. 2002-02-06 17:32:35 +00:00
pk
086b942780 Add tu' and tl' variable expansion modifiers, which transform the
value to uppercase and lowercase, respectively. From Kevin Neal from FreeBSD.
2002-02-06 16:26:12 +00:00
pk
86c078cb11 JobPrintCommand: avoid stale pointers on a node's command list.
The debugging value of the code that did this is limited, so drop it
altogether.
2002-02-06 16:20:25 +00:00
lukem
339d59ee8a Changes to the tiny images so that an mfs /dev isn't required (there might
not be enough RAM to support that), at the expense of more inodes and a
slightly larger md file system (which, when compressed down, still fits in
a kernel and that kernel still fits on a 1.2MB floppy):
- Define MAKEDEV_DEVS to 'std md0 fd0 fd1 wd0 wd1 wd2 wd3 cd0 cd1 pty0 ttyv0'.
- Call makefs with -f 10, to add a few spare inodes.
- Crank imagesize from 1344k to 1440k to handle extra space needed by the
  inodes containing the devices.
2002-02-06 16:15:56 +00:00
lukem
f551aa0ea4 If MAKEDEV_DEVS is defined, run etc.i386/MAKEDEV through MAKEDEV.wrapper and
makedev2spec.awk to generate another mtree specfile for use when building
the image, and don't copy dev/MAKEDEV to the target image.  This is primarily
for the tiny images which might not have enough memory for an mfs /dev.
Otherwise (MAKEDEV_DEVS not defined), continue to copy dev/MAKEDEV.
2002-02-06 16:12:18 +00:00
lukem
876b3e8195 no need to list etc/passwd anymore 2002-02-06 16:09:14 +00:00
lukem
1a64af8d1b MAKEDEV.wrapper
generate list of commands from provided MAKEDEV script with
	given arguments, to feed into makedev2spec.awk

makedev2spec.awk
	take output of MAKEDEV.wrapper and generate an mtree(8) specfile.
2002-02-06 16:08:16 +00:00
lukem
caa29fae38 #undef DIRBLKSIZ before #define-ing it 2002-02-06 15:44:49 +00:00
lukem
64d97711a2 remove btodb() hack now that makefs doesn't need it anymore.
the #define was wrong anyway; it was x<<9 instead of x>>9
2002-02-06 15:39:01 +00:00
lukem
5f26fac683 use (size / DEV_BSIZE) instead of btodb(size). fixes cross build issue. 2002-02-06 15:36:30 +00:00
lukem
02647fd20d don't need <ufs/ufs/dir.h> here 2002-02-06 14:58:15 +00:00
pooka
8c1a2832b7 Remove inclusion of unused machine/cpufunc.h, it doesn't belong here anyways. 2002-02-06 14:50:42 +00:00
ross
0271045b3e type in ^ directly, instead of \(ua which isn't in all the fonts 2002-02-06 12:51:13 +00:00
ross
ee869b65aa s/-width Fn xx/-width xx/
The -width argument isn't parsed and can't have attached markup.
2002-02-06 12:49:05 +00:00
tsutsui
552bb9895a TAB/space cosmetics 2002-02-06 12:43:05 +00:00
kleink
8785df9eae Oops, typo in previous, too. 2002-02-06 12:26:02 +00:00
kleink
ec9a0b843e * Fix typos and misusage !PMAP_MAP_POOLPAGE mb_map allocation.
* Use pmap_kenter_pa() for mapiodev().
2002-02-06 12:24:48 +00:00
christos
285fe0ec7a Make sure that we set *objp back to the decoded argument; pointed by
by Ben Harris.
2002-02-06 02:23:45 +00:00
lukem
6d5ff7c496 - clean up some debugging and verbose output
- use fs_cstotal.cs_nifree rather than fs->fs_ncg * fs->fs_ipg to
  determine the number of free inodes available in the file system
- don't count '.' (the root node for a level) in inode sizing, since
  the parent has already been counted
- some more todo items
2002-02-06 02:17:14 +00:00
mrg
171ba175d1 fix typo: s/hard/hand/ 2002-02-06 01:38:53 +00:00
chris
ffd3311a48 LINKADDRESS != LOADADDRESS, so fix that typo.
Also move the elf2aout conversion to be last in the build sequence, this means that strip is run on the elf kernel.
Also have dbsym run on the kernel, this should allow us to have symbols in the kernel.
This now means kernels boot and run, and can also have ddb symbols.
2002-02-06 00:55:56 +00:00
fredb
554e7c4e9d Forgot one -- `options COMPAT_15'. 2002-02-06 00:48:53 +00:00
fredb
69538a48bf Update the GENERIC and GENERICSBC kernels to conform to current fashion:
- - Disable `options DIAGNOSTIC' by default, as the checks are reported to
    be not really cheap.
- - Add `options COMPAT_AOUT_M68K', commented out, in preparation for ELF.
- - Remove `options EXEC_ELF32' from the kernel config. You can't just
    build an ELF kernel -- you need to update your toolchain, and then,
    once you go ELF, it's not an option, so it needs to go in std.mac68k,
    as on the other m68k ports.
- - Comment out `options COMPAT_SUNOS'. This hasn't been reported to work
    for a long time, so it's not very useful in GENERIC.
- - Enable `options IPFILTER_LOG', to go along with the `options GATEWAY'
    and `pseudo-device ipfilter' motif.
2002-02-06 00:45:53 +00:00
lukem
8b407af734 - allow ${MTREECONF} to contain multiple files, and build ${WORKSPEC} up by
catting these together with "/unset all" between each file.
- when populating the tree (${WORKBUILT} rule), depend upon ${WORKSPEC}
  rather than ${MTREECONF}
- support ${MAKEFS_FLAGS} as optional flags for ${MAKEFS}
- use CLEANFILES to remove files
2002-02-05 23:29:43 +00:00
lukem
5bb092ee45 replace /unset mode with /unset all 2002-02-05 23:16:51 +00:00
lukem
e07af841a2 use CLEANFILES to remove cruft 2002-02-05 23:16:16 +00:00
wormey
cef7fb637b Fix from Takeshi Shibagaki (ie9t-sbgk@asahi-net.or.jp):
(effect of)  fixed a bug when it makes install image.
(reason for) We cannot apply to newfs no disklabel device, so must
             write temporary disklabel. refer to i386.
2002-02-05 23:10:02 +00:00
christos
169617572d Fix u_int <-> size_t confusion. 2002-02-05 23:00:59 +00:00
christos
22a842763e Deal with other sizeof() return issues. We usually cast it to u_int, because
this is what the rpc function argument passing expects, or use socklen_t
as appropriate...
2002-02-05 22:41:47 +00:00
christos
91caf38a2a - Cast pointer casts to (void *) first to shut-up lint
- Deal with variable size enums, by generating assignment to an enum_t variable.
- Use f_print everywhere instead of fprintf - when in Rome...
- I don't understand the value of the unexplained hack. Use IXDR_{U_,}INT32
  instead of IXDR_{U_,}LONG.

With those changes our generated librpcsvc files pass lint.
2002-02-05 22:05:42 +00:00