Commit Graph

159556 Commits

Author SHA1 Message Date
tsutsui 7df023a2e5 Unroll cache flush/purge ops per 8 cachelines (128bytes).
Improve MI Sonic xfer speed ~2%.
2007-06-08 16:07:49 +00:00
tsutsui f04ca69e5f Use pointers for variables assigned to %a0 by __asm("%a0") so that
gcc may generate silightly better inlined code on optimization.
2007-06-08 15:57:16 +00:00
tsutsui dadbb85281 DCFA(), DCPA(), ICPL(), ICPP(), DCPL(), DCPP, DCFL(), and DCFP()
cache ops are identical between M68040 and M68060 and they are not
available on M68020 and M68030, so no need to check cputype even on
multi CPU kernels and they can always be inlined.

This improves MI Sonic xfer speed ~5%.
http://mail-index.netbsd.org/port-mac68k/2007/06/07/0003.html
2007-06-08 15:44:34 +00:00
tsutsui 792b0b8700 On my LC630 64KB DMA xfer causes timeout errors
so limit sc_maxxfer size to 63KB.
2007-06-08 15:03:51 +00:00
apb 57d159b747 Add a test for the `true && ! true | false` bug
which dsl fixed in revision 1.65 of src/bin/sh/parser.c.
2007-06-08 11:03:47 +00:00
dsl 8da3353686 Fix sh -c 'true && ! true | false; echo $?'
Add some more TRACE((...)) calls to aid such debugging.
Fixes PR bin/36435
Clearly no one tried this test when the changes of rev 1.31 and 1.44 were done!
2007-06-07 20:57:59 +00:00
manu 72fe4c3a84 From Paul Winder <Paul.Winder@tadpole.com>:
Fix ignored INTERNAL_DNS4_LIST
2007-06-07 20:04:26 +00:00
he 47b1d7c5c4 Commit generated file after correcting to use the "standardized"
variable for overriding SSP functionality location test.
2007-06-07 15:56:48 +00:00
tsutsui 36f4b2e43f Enable _pmap_set_page_cacheable(), _pmap_set_page_cacheinhibit() and
_pmap_page_is_cacheable() on all pmap_motorola.c ports and add
BUS_DMA_COHERENT support (from mvme68k) to common m68k bus_dma.c.

Tested with (not-yet-integrated) MI Sonic Ethernet driver on mac68k
http://mail-index.netbsd.org/port-mac68k/2007/06/07/0000.html
and ok'ed by thorpej.
2007-06-07 15:54:08 +00:00
he 9f3262d205 After ${_FILENAME} got assigned its value when it's "-", we can
no longer use it as a copy of ${_F} in the subsequent test for
additional targets under the MDSET_SUFFIXES test.

This brings back builds of srec & bin files for evbarm and img
files for evbppc.
2007-06-07 15:12:34 +00:00
he 678b5429b3 Instead of inventing a new environment variable to override the
SSP functionality location test, use gcc_cv_libc_provides_ssp instead,
as hinted by freza@ and uwe@.
2007-06-07 14:36:27 +00:00
tron 14f772ea69 Don't use alloca(3) inside the X11 libraries if we compile with
stack protection. XFree86 can now be buildt fine with "USE_SSP=yes".
2007-06-07 11:49:17 +00:00
pooka 673ffe67db file 4.21 2007-06-07 11:41:20 +00:00
pooka b27fe77588 regen 2007-06-07 11:38:08 +00:00
pooka 9afe6367b3 crunch cornflakes 2007-06-07 11:32:44 +00:00
pooka 40ad27db84 file 4.21. bugfixes 2007-06-07 11:24:27 +00:00
hannken 6087f7cc14 Dounmount(): rearrange mountlist_slock. vfs_allocate_syncvnode() may sleep
getting a new vnode so it must not be called with this simple_lock taken.

Fixes PR #36395
2007-06-07 10:03:12 +00:00
he 20237a6fdb Commit autoconf-generated file after modifying configure.ac to allow
override of the TARGET_LIBC_PROVIDES_SSP test.
2007-06-07 08:59:53 +00:00
he bd8f620dfb Modify the gcc configure script to allow override of the "target libc
has ssp functions built-in" test via the TARGET_LIBC_PROVIDES_SSP environment
variable, to allow us to configure a cross-compiler appropriately without
having to try to find out by looking in the target's source directory.

Tweak our build to tell gcc that the ssp bits are now in libc.

The native compiler appears to already think that the ssp bits
live in libc, so no change appears to be needed there.

The autoconf-generated configure script will be committed separately shortly.
2007-06-07 08:58:58 +00:00
yamt 7225d589de remove a duplicated definition of FFS_ITIMES. 2007-06-07 05:34:48 +00:00
rjs cc949d333c Make format directive match type of of_buflen. 2007-06-06 23:43:56 +00:00
njoly abfea2f7ae Remove duplicated AMD K8 MSR definitions, already in x86 specialreg.h
file.

ok by xtraeme.
2007-06-06 20:40:11 +00:00
briggs ac798ebce1 Dynamically allocate the buffer used for reading data from the input
handle and default to 64k instead of the 1k (BUFSIZ) static buffer.
This makes a large difference in performance of some applications.
Make the buffer size tunable from the command line.
2007-06-06 17:49:14 +00:00
rjs ef2a6ab865 Wrap ktrace calls with #ifdef KTRACE. 2007-06-06 17:08:27 +00:00
vanhu 6ae0ffb7d9 From Rong-En Fan: fix compilation with gcc 4.2 2007-06-06 15:37:15 +00:00
vanhu cc41629a4c fixed compilation with gcc 4.2 2007-06-06 15:37:14 +00:00
martti d405da7f9d Use "mktemp -d -t xxx" to create the temporary directories. This will use
TMPDIR environment variable if set, otherwise use /tmp.  (misc/35544)
2007-06-06 13:30:47 +00:00
soda 7dd0dd9e6b modify a comment in previous change 2007-06-06 13:21:28 +00:00
vanhu 6817ea28d9 speeds up interfaces update when they changed 2007-06-06 09:47:30 +00:00
vanhu 1ed22670fa From Jianli Liu: speed up interfaces update when they change. 2007-06-06 09:47:29 +00:00
yamt a4a20623eb nfs_getattrcache: simplify. no functional changes. 2007-06-06 09:23:55 +00:00
vanhu 7c53bfe0b6 ignore obsolete lifebyte when validating reloaded configuration 2007-06-06 09:18:16 +00:00
martin 7003cf1903 regen 2007-06-06 08:21:08 +00:00
martin 400bdf5863 Add pseudo-device agr 2007-06-06 08:20:26 +00:00
martin fda9709385 Add pseudo-device agr to more kernels 2007-06-06 08:19:10 +00:00
martin 0cfdef558a regen 2007-06-06 08:16:10 +00:00
martin 094e363596 Grr, do the addition of pseudo-device agr the right way 2007-06-06 08:15:54 +00:00
martin cf9773f5c0 It is "data_start" not "data" - pointed out by Tom Spindler 2007-06-06 07:56:39 +00:00
rillig f1ed085bcf Explained the word "new" in the first paragraph. 2007-06-06 06:18:39 +00:00
rillig 54195a73e6 Increased WARNS to 4, since it passes both gcc333 and gcc411 on i386. 2007-06-06 06:10:50 +00:00
pooka f8326bf98e Move puffs to a two clause license where it already isn't so. And
as agc pointed out, even files with the third clause were already
effectively two clause because of a slight bug in the language...
2007-06-06 01:54:59 +00:00
pooka e2819ad1be In very verbose debug mode, print also return values for operations
(well, at least for those that go through checkop()).
2007-06-06 01:33:10 +00:00
dyoung 4d1beb764e Make ADM5120 boards use the "common" disk subroutines, such as
readdisklabel(9), which are better than the evbmips disk subroutines
at finding a BSD disklabel at a different location than the evbmips
default of sector 0, offset 64.  Helps me boot a RouterBOARD 153
from a CompactFlash containing both a Master Boot Record at sector
0, and a BSD disklabel at sector 1, offset 0.
2007-06-05 21:53:10 +00:00
dyoung 6f3106eba2 Let the operator use the DISKLABELSECTOR and DISKLABELOFFSET
environment variables to override the sector number, and the
bytes-offset in the sector, where disklabel(8) writes a disklabel.

If disklabel(8) was compiled as a host tool, then the defaults
still come from the port's LABELOFFSET and LABELSECTOR #defines.
Otherwise, the defaults still come from the sysctls, kern.labelsector
and kern.labeloffset.

This change helps me create a bootable CompactFlash image with both
a Master Boot Record (MBR) and a BSD disklabel for the RouterBOARD
153.  The RB 153 is an evbmips board with a CompactFlash slot.  It
searches the MBR on a CompactFlash card for a partition containing
an ELF kernel.  When disklabel(8) is built for evbmips, it will
ordinarily overwrite an MBR with the BSD disklabel.
2007-06-05 21:48:46 +00:00
christos 41101b58c8 Add RCSID's to help files and centralize the function to cat the help files. 2007-06-05 17:50:22 +00:00
wiz 8c64f063a1 Apply SODA Nuriyuki's patch for multibyte support.
Based on soda-revised2.patch (in private mail) with a change from
soda on tech-userlevel.

Ok christos@.
2007-06-05 17:48:19 +00:00
yamt da51d139a4 improve post-ubc file overwrite performance in common cases.
ie. when it's safe, actually overwrite blocks rather than doing
read-modify-write.

also fixes PR/33152 and PR/36303.
2007-06-05 12:31:30 +00:00
pooka ae34899ec3 make getopt() actually accept -t also 2007-06-05 12:20:35 +00:00
yamt 962e1166fd add a missing dependency. 2007-06-05 12:14:33 +00:00
mishka d7d886db87 Our VLAN implementation (as it correctly expected) is not allowed
to modify the whole VLAN tags, but it is permitted to change 12-bit
VLAN identificators only. Reflect this fact on the appropriate man
pages.

Antti Kantee and Mihai Chelaru from #netbsd-code were helpful in
better understanding of VLAN stuff. Thank you!
2007-06-05 11:15:15 +00:00