4e8e4580a2
Fix macro usage, use Dv for SA_FLAG_STACKINFO, and English nits.
2004-03-14 10:19:59 +00:00
bcba66f310
Minor English nits.
2004-03-14 10:18:33 +00:00
e71ca4ef99
some manpages of scheduler activation syscalls.
2004-03-14 09:18:45 +00:00
4604becb77
Sprinkle cast to prevent unwanted sign extension. Some of those actually
...
makes sed invoked during build process an out of bound array access.
2004-03-12 22:34:09 +00:00
ca430e5111
Really, there is no need to use the old int0x80 syscall entry just for
...
getcontext(). There are some registers not properly saved with the
new syscall entry, but these are calee-saved, scratch or meaningless.
It's the setcontext() return path what matters.
2004-03-09 20:21:22 +00:00
294b4fa9f2
we also have to reset bits in the i387 status
...
as required
2004-03-09 17:16:13 +00:00
b79602af35
-clear status bits in mxcsr as required
...
-return the old accumulated status bits
2004-03-05 16:27:56 +00:00
a19008b17f
Explicitly use base 10 when parsing numbers, to enforce the "decimal number"
...
semantics that the comments referred to and as per dd(1) (which was the
origin of this functionality).
2004-03-05 05:58:29 +00:00
06588a00dd
Sort SEE ALSO, drop trailing space.
2004-03-04 23:49:31 +00:00
61e64d47d4
Sort SEE ALSO.
2004-03-04 23:47:56 +00:00
a19b383778
libc.so.12.113: isinf(), isnan().
2004-03-04 23:43:56 +00:00
8e54f10bf7
* Turn isinf(3) and isnan(3) into C99-style macros.
...
* Make it possible for ports to override these (i.e., VAX).
* Remove isnanl(), which was internal to libc only.
2004-03-04 23:42:38 +00:00
119b5f631a
Don't call end{serv,proto}ent_r() on return. It results dangling pointer.
2004-03-04 02:30:41 +00:00
eeabd92cc2
Add UC_STACK and stack_t related definitions.
2004-03-04 00:16:47 +00:00
71f2db14c8
Use this to symbolically get into ucontext_t and struct sigcontext.
2004-03-03 22:58:34 +00:00
3bd8b208cf
Add support to generate an assym.h if corresponding ARCHDIR contains a
...
genassym.cf.
2004-03-03 22:57:26 +00:00
afc5614967
Copy this from libpthread so arch-specific .S files can use assym.h
2004-03-03 22:56:04 +00:00
4634f0716d
Don't zero a floating point variable on entry -- this causes the FPU to
...
be allocated needlessly by almost every program (at least on i386/amd64).
2004-03-03 11:19:05 +00:00
ca0ee478c7
use "%s" for err(), do not pass pointer directly. Pedro Martelletto
2004-03-01 23:06:21 +00:00
1dd7501db3
blow this up by sticking a copyright notice to it
2004-03-01 20:56:05 +00:00
c5bc0dada8
optimize this: we can use %ra for relocation directly
...
(saves 1 insn)
2004-03-01 20:43:19 +00:00
9560ce9b0f
we need a silly trampoline to make _resumecontext work
2004-03-01 19:26:29 +00:00
293e83e97e
get the offset to gp right
2004-03-01 18:32:51 +00:00
e9ee155221
Get the stack alignment right: the stack arguments must start
...
at a 16-byte aligned address, the return address is at an
"odd" address.
approved by fvdl
2004-02-28 15:57:21 +00:00
a839ccc650
alignment trouble: uc_link is at offset 8
2004-02-27 21:23:05 +00:00
5a257984e5
Fix indent and other whitespace usage.
2004-02-26 23:01:15 +00:00
37d3fd292f
the char we fill with shouldn't be sign-extended
...
(the way the length argument is dealt with looks suspicious
too, but this would only have an effect for >=2G fields)
this can probably be optimized...
2004-02-26 20:50:06 +00:00
3cac582c48
fix copyback from fp stack - we need 64 bits
2004-02-26 19:19:06 +00:00
5509874609
Semantic fix for allocation size for malloc() and realloc() previously
...
modified. Allocation unit isn't sizeof(char **) but sizeof(char *).
Noted by Geoff Wing.
2004-02-25 06:28:10 +00:00
e79fe4ef48
Fix allocation size bug from Takahiro Kambe.
2004-02-24 15:28:07 +00:00
d3516aeed5
- fix a memory leak on the current line buffer.
...
- fix a memory leak on realloc failure.
2004-02-23 16:06:52 +00:00
ad788cab09
In gcc3 world the "millicode" must be provided by libgcc*.a, so don't
...
include it in libc.
2004-02-22 01:05:49 +00:00
2b03200f1f
servent_r/protoent_r additions
2004-02-19 19:31:06 +00:00
d68ecdf1c9
getservent_r/getprotoent_r
2004-02-19 19:27:53 +00:00
ec210d4a19
Implement the non-reentrant ones using the reentrant ones.
...
Name files by function name.
2004-02-19 19:27:26 +00:00
7782337bb7
_r functions for services and protocols; also add the inet_cidr_* function
...
which will come next.
2004-02-19 19:22:51 +00:00
8059eed1bc
Re-entrant versions of services and prorocols functions.
2004-02-19 19:21:44 +00:00
aff9207208
document svcudp_create
2004-02-16 02:39:36 +00:00
e3fc4b66c3
Spell the plural of suffix "suffixes", not "suffices".
...
Inspired by PR 24400 by Todd Vierling.
2004-02-13 17:56:17 +00:00
d141a7411d
Forgot to build signbit() library parts in previous; noticed by mrg.
2004-02-12 10:28:28 +00:00
6dbd72c053
MAKEVERBOSE messages
2004-02-08 22:52:41 +00:00
fa3c270527
Minor KNF.
2004-01-31 16:56:12 +00:00
9af86fc6f2
Fix two typos, and remove a superfluous comma.
2004-01-30 16:03:50 +00:00
7fcc4d6c6b
Sort SEE ALSO.
2004-01-30 15:38:47 +00:00
fa6c746b59
cross reference sticky(7)
2004-01-30 15:31:27 +00:00
4348bc0770
whitespace KNF
2004-01-29 08:25:28 +00:00
72489e1ea0
uuidgen(2) syscall. Originally from FreeBSD, ported by John Franklin in
...
PR#23470, with minor updates by me. This is only the syscall support
from that PR, for now.
Changes: port over fix from FreeBSD for multicast address generation.
Changed bcopy to memcpy. For now, #ifdef notyet the portions of
kern_uuid.c that are meant to be used by (currently nonexistent) other
things in the kernel. Added syscall to COMPAT_FREEBSD as well, though
that's currently not useful, as any program new enough to use this call
also uses other syscalls we don't (yet) emulate.
2004-01-29 02:00:02 +00:00
b763e0987d
Use the correct parameters when building the YP map name in _yp_getnetbyaddr().
...
Fixes getnetbyaddr() when reading from YP. From Brian Ginsbach in email.
2004-01-29 01:39:15 +00:00
b776c8e6bb
Retain the euid of the target user whilst reading the
...
.rhosts file. This allows it to have a mode of 600
and still function correctly even when used on an NFS
client with uid 0 mapped to another (anonymous) UID.
Patch from Brian Ginsbach <ginsbach@cray.com>,
and fixes [bin/14880] from Darren Reed <darrenr@>.
2004-01-27 11:46:34 +00:00
c5a53802fe
Use syslog(3) instead of err(3) to log errors.
...
Based on email conversation with Brian Ginsbach.
Add '%option never-interactive' to nslexer.l, to prevent isatty() from
being called unnecessarily.
Per discussion on current-users@ between Christian Biere and Greg Woods.
2004-01-25 16:38:15 +00:00
cef0afe940
Lots of grammar and wording changes from jmc@openbsd, with some
...
additional consistency changes by me.
2004-01-24 16:58:54 +00:00
36f4f7f5c3
Misc fixes from jmc@openbsd.
2004-01-24 15:33:43 +00:00
9ee271adfd
If the syscall fails, the carry bit will be set - not the zero bit.
2004-01-23 13:45:43 +00:00
be85fac5e9
Fix typo: 'Each characters' -> 'Each character'
2004-01-21 12:40:14 +00:00
1373fa4222
s/sgn/sign (compiles now)
2004-01-20 19:41:30 +00:00
8863e1ea05
Make these compile too.
2004-01-20 19:33:19 +00:00
2758365e99
Make these compile.
2004-01-20 19:29:43 +00:00
dff5a86b96
Too sad - inlining syscall stubs on sparc64 is so much easier than doing
...
PIC calls to the libc stubs, but unfortunately we need to do the latter-
otherwise we break weak symbol semantics and make libpthread fail.
Thanks to Christian Limpach for the analyzis.
2004-01-19 18:33:21 +00:00
f04f3a62db
Forgot to mention in last commit message: remove some accidently commited
...
debug code.
2004-01-19 18:20:35 +00:00
fd70c1464d
Unfortunately we can not restore the signal mask with setcontext() - this
...
breaks libpthread, which tries to intercept the weak sigprocmask symbol.
So use an explicit sigprocmask call instead.
2004-01-19 18:19:33 +00:00
e032233866
Store the "signal mask saved" flag at the end of the jmp_buf - we now use
...
more of it than before (and I bet the old place just is a non translated
32bit sparc value).
2004-01-19 18:17:21 +00:00
3d2a90fb95
Make this compile.
2004-01-18 19:33:01 +00:00
c5bbd9dca0
_citrus_NONE_stdenc_wctomb: return E2BIG correctly.
2004-01-18 03:57:30 +00:00
1d81763357
print max rather than min in error message when number
...
parsed by strsuftollx is greater than max
2004-01-17 23:02:51 +00:00
2f3e9d2cf8
Avoid using the depreceated __sigreturn14 for longjmp by adapting
...
Christian Limpachs arch/m68k implementation using setcontext().
2004-01-17 22:01:15 +00:00
ead753d26d
Fix typo fpclassifd.c -> fpclassifyd.c, reported by Bertram Barth.
2004-01-16 10:46:11 +00:00
3ce85dd47a
inifities are probably infinities.
2004-01-15 20:31:40 +00:00
51008fe87a
I guess ture means true.
2004-01-15 20:30:46 +00:00
bb37e9c818
libc.so.12.110: fpclassify(), isfinite(), isnormal(), signbit().
2004-01-15 19:44:46 +00:00
d82e732345
Add C99 fpclassify(), isfinite(), isnormal(), and signbit() macros.
2004-01-15 19:43:44 +00:00
3c36c70c82
fix a typo
2004-01-11 16:05:27 +00:00
7d6187a689
PR/24023: Min Sik Kim: Don't deadlock while closing mappers. The module
...
uninit function can call close again, which will try to obtain a held
lock. Unlock the lock before calling the actual close function, since
we already disassociated cm from the rest of the data structures.
2004-01-08 19:23:19 +00:00
70cfe67d9f
There is no sysctl(9) (yet?), comment it out;
...
fix a typo.
2004-01-08 09:21:35 +00:00
6807f73559
Fix some spacing nits, an Xref section, and a long line.
...
Bump date for previous.
2004-01-08 09:12:57 +00:00
76f7bb35fd
Add descriptions of flags.
2004-01-07 21:25:14 +00:00
46f02625cb
- Describe all system threads.
...
- Mention all MI threads that are not device specific; i.e. mention scsibus,
usb, raidframe, nfs, but not pms.
2004-01-07 16:26:23 +00:00
40069685c3
Make this compile, and, as a side-effect, LP64 clean.
2004-01-06 11:29:30 +00:00
12eeb195bb
kill strcpy
2004-01-06 08:08:44 +00:00
8a293ad181
Document F_MAXFD, F_CLOSEM
2004-01-05 00:34:34 +00:00
f6ee845eac
Fix a cast for little endian archs
2004-01-03 23:50:47 +00:00
a2c6783de5
When writing/reading longs use explicit 32bit temporary values (this is
...
what "long" means in xdr context).
Fixes PR lib/23960.
2004-01-03 23:01:41 +00:00
a004e52a39
Uppercase CPU; mark up NULL; remove superfluous word; simplify macro usage.
2004-01-03 13:09:44 +00:00
9cfe841137
At long last, add descriptions of new "dynamic" operations.
2004-01-03 06:06:36 +00:00
ba54915063
Fix two inverted comparisions and a few comments.
...
This makes sparc64 pass the sigsetjmp regression tests.
2004-01-02 22:30:36 +00:00
35f883b08a
KNF
2004-01-02 21:49:35 +00:00
90e89977d4
- libc.so.12.109 and libpthread.so.0.4
...
- add ssize_t stackinfo_offset argument to sa_register syscall
- remove sa_unblockyield syscall
- make __sigprocmask14 syscall weak
- in pthread__sa_start: catch up with 4-argument sa_register syscall
2004-01-02 18:56:39 +00:00
8ab300d124
Fix typo; from Aidan Kehoe in PR lib/23951.
2004-01-02 17:06:21 +00:00
2fec59e8d0
Fix typo; from Aidan Kehoe in PR lib/23950.
2004-01-02 17:04:52 +00:00
ce5186b58b
Restore stack frames on error return.
2004-01-02 14:31:31 +00:00
97ab4e62e3
Restore stack frames on error exit. This makes us pass dynamic and static
...
linked versions of the clonetest regression test.
While here, some cosmetics: branch prediction, avoid depreceated instructions.
2004-01-02 14:06:51 +00:00
855852a04b
comment on values from 0x80000000 to 0xffffffff
2004-01-02 12:27:41 +00:00
5d489a8a2a
safer realloc
2004-01-02 12:25:46 +00:00
64d2443254
safer realloc, some KNF. noted by deraadt@openbsd
2004-01-02 12:19:25 +00:00
586bc581af
Use CC64FSZ instead of magic constant for stack frame size.
...
Frob the clone's stack pointer to be BIASed - it is 64 bit code too.
Use "carry set" as test for a failed syscall.
2003-12-31 15:56:26 +00:00
47923e6bf5
In case of underflow, return -0.0 or +0.0, depending on the sign of
...
the argument.
This makes sparc64 pass regress/lib/libc/ldexp.
2003-12-30 23:19:07 +00:00
7fd2c31d41
Add support for scanf "hh" format modifiers - we use it in int_fmtio.h
...
for SCNd8 and friends - so we should support it properly (and not only
by pure luck on alignment blind little endian archs - it used the
"h" code path, storing a 16bit int via a byte pointer).
This makes sparc64 pass regress/lib/libc/int_fmtio.
2003-12-30 22:10:20 +00:00
9f96807631
When storing data size as 32bit int in an overflow page, copy the
...
length to a temporary u_int32_t variable and memmove from there to the
final destination - instead of copying the first four 0 bytes from a
64bit variable directly on big endian 64 bit platforms.
This makes sparc64 pass regress/lib/libc/db.
2003-12-30 21:20:16 +00:00
9a4bbff85d
_citrus_iconv_std_iconv_convert:
...
- allow 0 byte input. PR/23889 (Valeriy E. Ushakov)
- remove redundant E2BIG check, which should be done by
each encoding modules anyway.
2003-12-29 18:05:01 +00:00
95a5e260a9
_citrus_MSKanji_wcrtomb_priv: return E2BIG correctly.
2003-12-29 17:59:03 +00:00
556685e5a3
_citrus_NONE_stdenc_cstomb: return E2BIG correctly.
2003-12-29 17:49:22 +00:00
08f5fb02e9
Remove sentence about this being an alpha
...
version -- after twelve years it is at least beta ;)
The BUGS section is extensive enough even without it.
2003-12-29 17:36:12 +00:00
d46bc94200
Niels Provos kindly agreed to drop clauses 3 and 4 from the
...
license -- thanks.
Based on OpenBSD commit and hints by itojun.
2003-12-26 19:04:55 +00:00
f9f4341a9f
Don't ever add a ".0" to a single digit number.
...
XXX: Should (for example) 1024 be 1.0K or 1K when the HN_DECIMAL flag
is passed? Should there be a separate option that says "use the
HN_DECIMAL behaviour unless we are exactly equal to the suffix"?
2003-12-26 11:30:36 +00:00
903622333a
Use an unsigned comparison against minbrk.
...
From Steve Woodford, inspired by a similar change to other archs from
OpenBSD.
2003-12-26 11:23:44 +00:00
269c3e3d2b
Use an unsigned comparison against minbrk.
...
From OpenBSD, changed by Miod Vallat.
2003-12-26 11:21:48 +00:00
a2c17b5c1e
One merge-o missed in previous.
2003-12-20 00:21:00 +00:00
84a2a5ca6e
Merge tzcode2003e.
2003-12-20 00:12:05 +00:00
b717391e1a
Note: the previous revision was an update for tzcode2003_e_.
2003-12-19 23:52:44 +00:00
4a2741cba1
Update for tzcode2003d.
2003-12-19 22:25:42 +00:00
ef1b175332
Bump date for previous (hi christos!).
2003-12-17 23:04:33 +00:00
4c4fa74153
Say what happens with pgrp == 0.
2003-12-17 21:05:30 +00:00
a8c35a134f
remove superfluous assignment
2003-12-14 23:56:28 +00:00
7db1d33cba
Modify the licences of code written by Theo De Raadt from a 4-clause
...
to a 2-clause licence (retaining UCB clauses (1) and (2)), per PR
22409 from Joel Baker, approved by Theo de Raadt, and ratified by
myself - the only discrepancy being the handling of the original
clause 3 in src/usr.sbin/yppoll/yppoll.c.
2003-12-10 12:06:25 +00:00
c81f32fe6c
comment from niels provos;
...
- seed2 is necessary, but use it as "seed2 + x" not "seed2 ^ x".
- skipping number is not needed, so disable it for 16bit generator (makes
the repetition period to 30000)
2003-12-10 05:22:18 +00:00
9510307938
simplify a sentence
2003-12-09 21:40:54 +00:00
af68367011
Reference kfilter_register(9) rather than the non-existent kfilter_register(2).
2003-12-09 19:49:53 +00:00
02f06f3790
Bump date for previous (hi andrew!).
2003-12-09 16:00:40 +00:00
31f1dc4a27
Update man page to correct PROC_* names (addresses PR lib/23645) and
...
also describe the new PROC_PID_STOPEXIT variable.
Man page still not updated to include new features. I'm still working
on that.
2003-12-09 01:39:03 +00:00
aa6e9c56c3
Make this compile on non-IEEE GCC3 targets (aka VAX).
2003-12-08 06:18:16 +00:00
1492b0bcb0
"int flags" is correct in the prototype, but not in the English
...
language. s/int/in/
2003-12-08 00:55:57 +00:00
3c12447a00
Make building IPv6 support controllable by MKINET6
2003-12-07 21:57:22 +00:00
29b106093d
Swap WEAK_ALIAS args to how they should be.
2003-12-07 19:37:23 +00:00
c9c20433ad
Revert last change, as this is a 3rd-party file, noted by kleink@.
2003-12-05 06:56:19 +00:00
0b30fd3d69
Spaces.
2003-12-05 00:57:36 +00:00
39d96c1f34
netbsd.org -> NetBSD.org
...
NetBSD.ORG -> NetBSD.org
Now src/lib is done.
2003-12-04 23:39:18 +00:00
6fc962bc06
oops
2003-12-04 19:45:19 +00:00
93375edaa0
Bump for sysctl() update
2003-12-04 19:42:32 +00:00
fcc5d1a731
Adapt userland sysctl goop to new world order, permitting dynamic
...
discovery.
2003-12-04 19:40:55 +00:00
97f5444c8a
MAKEVERBOSE tweaks
2003-12-04 01:02:44 +00:00
b02f563aeb
improve random number generation by using largest 31-bit prime
...
as denominator; regenerate the randtbl[] with new generator
patch provided in lib/22853 by KIHARA Hideto; code actually present
in libkern/random.c since 1994/06/03
matches similar change done in glibc in 1995, and FreeBSD & OpenBSD in 1996
2003-11-26 20:44:40 +00:00
ea56bcee84
Hide the register number constants behind an _R_ prefix, and also
...
rename FPBASE to _FPBASE, so that we avoid polluting the user's
name space when e.g. <sys/ptrace.h> is included. Previously, the
PC symbol in mips/regnum.h would conflict with the declaration of
the external variable by the same name in termcap.h, as discovered
by the ``okheaders'' regression test.
2003-11-26 08:36:49 +00:00
18239b5b77
Fix compile problem if -UYP -DHESIOD.
...
Noted on current-users by Srinivasa Kanduru.
2003-11-26 00:48:59 +00:00
492a994a19
Bump date for previous.
2003-11-25 23:28:31 +00:00
fd02e804dd
take "Skip a random number of ids" into consideration, correct the rotation
...
period number. simonb
2003-11-25 23:11:54 +00:00
22a07298e9
- Document NO_DEFER
...
- Turn sa_flags into a list to make it easier to find/read
- Sort the sa_flags lists.
2003-11-25 21:09:14 +00:00
b6a2cef9dc
"seed2" was ruining the non-repeating property of this function; remove it.
...
discussed on tech-net for ip_id.c (thanks for all the analysis).
2003-11-25 18:13:06 +00:00
1d408df31d
Catch up with namei: An empty string is not a valid pathname.
...
Noted by SAITOH Akinori in PR misc/23565.
2003-11-25 15:33:58 +00:00
9d0d2c7461
Implement siginfo for sh3.
2003-11-23 23:13:11 +00:00
53a406794d
s/!=/==/
...
The last fix wasn't quite right, but happened to work for me due
to stale dependencies...
Reported by Markus W Kilbinger.
2003-11-23 09:10:20 +00:00
6b3d472b3a
New sentence, new line. Remove trailing spaces.
2003-11-23 07:22:19 +00:00
e54d574bd0
Bump Dd.
...
better explain the interactions between ip.mtudisc and tcp.mssdflt,
with regards to the TCP segment size.
ip.mtudisc is not read-only, as was implied by "returns"
XXX: this ("returns") should be corrected for other parameters, too.
Xr tcp(4)
2003-11-23 07:11:04 +00:00
19552574ee
Don't compile fp*.c if doing a soft-float build.
2003-11-21 11:39:46 +00:00
c09770cac0
Use Rv for return values section; bump date.
2003-11-19 16:50:09 +00:00
cce38fddc2
document RETURN VALUES
2003-11-19 15:20:32 +00:00
9b7e93e2c6
s/modified/modifies/ in the description of makecontext().
2003-11-18 23:03:26 +00:00
faedb5db06
Fix typo in previous.
2003-11-18 08:49:18 +00:00
88ed237274
Bump libc to 12.107 and libpthread to 0.3 for fsync_range(2).
2003-11-18 01:00:19 +00:00
137c808bb8
Userland portion of fsync_range(2), written by Bill Studenmund, and
...
contributed by Wasabi Systems, Inc.
2003-11-18 00:56:56 +00:00
f73b119e04
s/modifed/modified/
...
And while here...
s/NetBSD.ORG/NetBSD.org/
2003-11-14 16:43:57 +00:00
bc4e986724
-traditional-cpp and spaces after commas is not a happy combination.
2003-11-13 06:29:39 +00:00
dc7f22dd3e
Need weak aliases here.
2003-11-11 19:18:14 +00:00
7239f40663
Add #ifdef __weak_alias around use of __weak_alias.
2003-11-11 19:10:02 +00:00
faf6ed6d51
Fix tyop.
2003-11-11 18:29:58 +00:00
ee1b406595
Spell address with two d's. Inspired by similar changes in OpenBSD,
...
originating from Jonathon Gray and forwarded by jmc@openbsd.
2003-11-10 08:51:51 +00:00
e0406466be
This is the last part of siginfo support for pc532. Simon will make this
...
work.
2003-11-06 02:46:44 +00:00
368d3f3d0e
siginfo trampoline addition. This is not used until simon verifies that it
...
works.
2003-11-06 02:28:00 +00:00
c1a19161a9
From Ragge's wishlist: rename infinity -> huge_val.
2003-11-05 13:46:51 +00:00
b01cb286b8
Add userland portion of siginfo for mips. Many thanks to simon for letting
...
me use his machines down under.
2003-11-02 08:37:48 +00:00
8fc6819ddd
Corrent map order to match sparc/ieeefp.h.
2003-11-01 14:26:29 +00:00
9356ead951
Pasto.
2003-10-30 23:34:16 +00:00
617742615c
It turns out the hppa ABI we're using does not support an extended-
...
precision type, so drop the library support for now but leave in place
(#ifdef _LP64) the header definitions.
2003-10-30 22:46:17 +00:00
a1e0a00965
"The fgets() and functions gets() do not..." ->
...
"The fgets() and gets() functions do not..."
2003-10-30 12:59:23 +00:00
2ddae9dc8a
Add manual pages for fpclassify and isfinite.
2003-10-29 22:59:23 +00:00
26198324b4
Merge tzcode2003d.
2003-10-29 20:43:27 +00:00
dc2bc4a89d
Import tzcode2003d.
2003-10-29 20:19:13 +00:00
0a97d0875b
C99 7.12.3.2: add library portions of isfinite, but don't build these just
...
yet.
2003-10-29 20:03:51 +00:00
3991bf60bc
Fix the exponent value we're comparing with.
2003-10-29 19:34:11 +00:00
70d2e11feb
caps "Jonathan STone" in comment less dramatically
2003-10-29 12:28:33 +00:00
6bffe7fb50
Upon rereading the implementation suggestions in the standard, rename the
...
double-precision version to __fpclassifyd(). No need to diverge, here.
2003-10-28 22:50:42 +00:00
75327729a2
C99 7.12.3.1: add IEEE library portions of fpclassify. Don't build these
...
yet as the VAX implementation is still subject to discussion.
2003-10-28 22:05:37 +00:00
72f016b386
Retire FPA support from this file at last; suggested by Richard Earnshaw
...
and not objected to by port-arm.
2003-10-28 21:44:04 +00:00
782579c57a
SUSE3 says htonl and friends come in via arpa/inet.h. This will actually not
...
hurt on NetBSD and makes hosting on non-native for tools/ work correctly
2003-10-28 04:00:51 +00:00
a54bd8931a
Fix tpyo. (ieee854 -> ieee754)
2003-10-28 01:24:06 +00:00
fdb80dd054
Ignore the explicit integer bit here, too; otherwise an infinity with the
...
explicit integer bit set could be classified as NaN.
2003-10-28 00:13:22 +00:00
f81f06350d
One ';' is enough.
2003-10-27 09:27:03 +00:00
dcce429e93
fix a typo that caused the build to fail. mmm commit before compile
2003-10-27 03:09:43 +00:00
5f4474073a
Explain the stack layout correctly now that that I understand it :-)
2003-10-27 02:33:06 +00:00
477697b1f2
Overhaul how `build.sh tools' are used:
...
* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h
* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.
These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).
2003-10-27 02:17:18 +00:00
964d6747b5
Err, rename some members added in previous to make them reflect their
...
semantics better.
2003-10-27 01:11:46 +00:00
6664eb5f2d
Add __sigtramp2
2003-10-27 00:46:05 +00:00
d9676dda89
Signal trampoline stuff for siginfo.
...
XXX: I don't understand why I am off by 48 bytes.
2003-10-27 00:45:36 +00:00
171d653219
Overhaul how `build.sh tools' are used:
...
* Rename "config.h" to "nbtool_config.h" and
HAVE_CONFIG_H to HAVE_NBTOOL_CONFIG_H.
This makes in more obvious in the source when we're using
tools/compat/config.h versus "standard autoconf" config.h
* Consistently move the inclusion of nbtool_config.h to before
<sys/cdefs.h> so that the former can provide __RCSID() (et al),
and there's no need to protect those macros any more.
These changes should make it easier to "tool-ify" a program by adding:
#if HAVE_NBTOOL_CONFIG_H
#include "nbtool_config.h"
#endif
to the top of the source files (for the general case).
2003-10-27 00:12:41 +00:00
5976f8aa56
Retire local union { floating-type; struct ieee_prec; } in favor of those
...
available from <machine/ieee.h>
2003-10-27 00:05:46 +00:00
55337d2d6f
On FPA, switch back `long double' to double-precision. According to
...
Richard Earnshaw, extended-precision was never supported by the compiler.
2003-10-26 16:00:17 +00:00
be5cd162d0
Use the generic `long double' == `double' versions on m68000.
2003-10-26 10:03:11 +00:00
a98b4bac65
Add __infinityf and __infinityl, float respectively long double analogs
...
of __infinity.
2003-10-26 00:48:14 +00:00
e9788d63be
Whitespace cosmetics.
2003-10-25 22:43:26 +00:00
142c4159e8
libc.so.12.106: __infinityf, __infinityl.
2003-10-25 22:33:14 +00:00
9f3dbee22c
Add __infinityf and __infinityl, float respectively long double analogs
...
of __infinity.
2003-10-25 22:31:20 +00:00
8e2428c2c6
Remove an unnecessary line continuation backslash from previous.
2003-10-25 21:54:30 +00:00
c6126e1095
Sync with i386:
...
Update to account for separate handling of the explicit integer bit;
simplifies a little.
2003-10-25 21:48:34 +00:00
044354301a
Update to account for separate handling of the explicit integer bit;
...
simplifies a little.
2003-10-25 21:42:07 +00:00
72315d3199
Update for FPA 80-bit extended-precision format.
2003-10-25 18:37:43 +00:00
25de9f5e2a
Swap VFP and FPA sections to match the style of most other code;
...
no functional change.
2003-10-25 17:57:15 +00:00
342cbd284f
replace __longjmp14 asm-implementation with __longjmp14 C-implementation
...
uses setcontext instead of __sigreturn14
fixes longjmp/siglongjmp when !COMPAT_16
2003-10-25 17:13:57 +00:00
bb5e0820bc
Redo the VFP/endianness conditional logic of this file, making it easier
...
to read. No functional change except for the side-effect of fixing the
ARMEB FPA case, which wasn't supported anyway.
2003-10-25 16:17:44 +00:00
6afbde6005
Correct another typo.
2003-10-25 14:51:16 +00:00
15e03bd18f
Pasto in previous.
2003-10-25 14:31:21 +00:00
661f76699f
POSIX: when attempting to fdatasync(2) a file which is not open for
...
writing, fail with EBADF.
2003-10-25 01:18:01 +00:00
17bcc01b65
Oops, re-add line continuation.
2003-10-24 19:50:13 +00:00
4ca4f9f41a
Add an error check to ensure these are used on platforms without a
...
distinct long doble type only.
2003-10-24 10:30:36 +00:00
d469976b59
In PIC code "braf" to cerror, not "bsrf" to it!
2003-10-24 04:41:28 +00:00
712c8de209
* Since there is no single portable IEEE 754 format for a long double,
...
keep a common implementation of isinfl() and isnanl() to be used by
platforms where `long double' == `double'; move others into
machine-dependent code.
* In due course, consider __VFP_FP__ on arm.
2003-10-24 00:58:01 +00:00
2ffebceb85
Duh, fix the align-to-32-bits code.
2003-10-23 20:46:40 +00:00
0c336c7f37
Move include for reentrant.h inside CONFIG_H test so this works correctly
...
with tools/compat. Fixes PR#21344
2003-10-23 20:06:16 +00:00
6cded84c4b
Fix an RCS Id botch that occured when renaming.
2003-10-23 00:04:57 +00:00
22396ac6f7
G/c __nanf2, which is neither being used nor the representation of a NaN.
2003-10-22 23:50:47 +00:00
77e57be326
For a double-extended NaN, the exponent must be EXT_EXP_INFNAN, not
...
DBL_EXP_INFNAN.
2003-10-22 21:37:34 +00:00
9122b04d59
the last argument to recvfrom() is a pointer, so use "NULL" for
...
consistency
2003-10-22 15:40:19 +00:00
71baa97c0b
-don't update the RPC server address with the reply's sender address;
...
this breaks server failover under some circumstances, and Solaris doesn't
seem to do it either
-minor cleanup
approved by fvdl
2003-10-22 15:38:31 +00:00
b279ed6377
Don't compare NULL against a character.
2003-10-21 00:07:17 +00:00
bc018099a2
Don't use NULL to compare against a char.
2003-10-21 00:03:47 +00:00
39d7a2e359
Don't use NULL to compare against a character.
2003-10-21 00:00:34 +00:00
9f7abac970
Use the int 0x80 path for context system calls.
2003-10-18 22:47:37 +00:00
7931285ae1
Need a couple more headers for the last change.
2003-10-18 20:11:04 +00:00
6646629469
Sync with the i386 version.
2003-10-18 18:36:35 +00:00
f57e3e8f0a
sync with the x86 one.
2003-10-18 18:33:51 +00:00
c4990e5532
Userland portion of SIGTRAMP_VALID() commit.
2003-10-18 18:16:23 +00:00
2b9ab2aab8
Get ucontext pointer back out of callee-saved register %r15.
2003-10-17 16:23:20 +00:00
58cc8234ca
Slight improvements.
2003-10-15 20:04:48 +00:00
910124a085
Make the plural of `X' `X's instead of `Xs'.
...
Following a suggestion from jmc@openbsd.
2003-10-15 19:44:51 +00:00
538e6e3e1e
Move siginfo(5) and ucontext(5) to siginfo(2) and ucontext(2) resp.,
...
and fix references.
Per a discussion on source-changes around Sep 14 2003.
2003-10-14 14:59:03 +00:00
e752a6f68a
PR/23158: Johan Danielsson: Portable usage of the mk*temp() functions requires
...
six `X's.
2003-10-14 14:29:46 +00:00
faf00d99a2
don't use uninitialized variable ('chlenbak').
...
the problem reported by Markus Illenseer on tech-userlevel@
and analyzed by enami tsugutomo.
2003-10-14 12:50:03 +00:00
789335a49c
Xscale-optimised mem* routines. Contributed by Wasabi Systems.
...
(Note: memcmp/memset improvements also benefit non-Xscale).
memcmp() - Compare 32-bits at a time if possible. Special-case 6-byte
comparisons, for the benefit of the network stack.
memset() - More loop unrolling, plus use of 'strd' instruction,
bzero() results in > 100% speedup on Xscale.
memcpy() - Big-endian support, unrolled loops, 'strd/pld', plus special-
cases for very common length/alignment combinations.
Benchmarks show ~50% improvment on Xscale.
memmove() - Big-endian support. Use fast memcpy(), above, if the regions
bcopy() don't overlap. Otherwise unchanged
XXX: The Xscale optimisations are not enabled by default, unless /etc/mk.conf
XXX: has the right compiler options. The intention is to pull them in via
XXX: something like libxscale.so, selected at runtime by ld.so.conf.
XXX: (Big-endian support is not affected by this).
2003-10-14 07:51:45 +00:00
8295081a25
& -> \*[Am]; mark up FD_SETSIZE; bump date for previous.
2003-10-14 07:44:43 +00:00
5435c421a0
note that programs that use rpc(3) cannot bump FD_SETSIZE.
2003-10-14 07:08:45 +00:00
9f1aac5bb3
Move Jason Downs's code from a 4-clause to a 3-clause licence by
...
removing the advertising clause. Diffs provided in PR 22410 by Joel
Baker, confirmed to the board by Jason Downs.
With additional thanks to Jason Thorpe.
2003-10-13 15:36:33 +00:00
b2b49933f4
Move Matt Green's code from a 4-clause to a 3-clause licence by removing
...
the advertising clause. Diffs provided in PR 22396 by Joel Baker, the changes
were confirmed to the board by Matt Green.
2003-10-13 14:22:20 +00:00
ed6ed8e698
Move Keith Muller's code from a 4-clause to a 3-clause licence by removing
...
the advertising clause. Diffs provided in PR 22397 by Joel Baker, confirmed
to the board by Keith Muller.
2003-10-13 07:41:22 +00:00
f70031fee9
SIGINFO __sigaction & trampoline code.
2003-10-12 14:42:03 +00:00
7152fa1070
use matt's powerpc implementation. It is simpler than mine, and it does
...
matter as far as performance goes.
2003-10-12 04:23:17 +00:00
f431b58f2b
Back out rev. 1.56; `pointer points the byte just after the valid region'
...
is normal case. If the parser wants something more, there should be
another bounds check for it.
2003-10-11 03:35:42 +00:00
62887f0740
Remove another unneeded cast to size_t.
...
Pointed out by Anil Madhavapeddy in private email.
2003-10-10 05:38:45 +00:00
c9be9cafc5
dont use __ARMEB__
...
It changed so that it might distinguish by BYTE_ORDER
2003-10-09 09:44:25 +00:00
b6a7f7b339
support big endian
2003-10-09 09:37:56 +00:00
444e6dce83
No need to cast "len" to a size_t now that "len" is a size_t.
2003-10-09 00:50:34 +00:00
fd04e737c6
Update to use symbolic register names.
2003-10-09 00:17:59 +00:00
1bb83f4a4a
avoid signed->unsigned assignment. openbsd PR 1709
2003-10-08 20:33:17 +00:00
b3ec0a1d45
Siginfo libc changes for alpha. Approved by thorpej@netbsd.org.
2003-10-07 17:08:07 +00:00
8655c7d7eb
Add a MAP_WIRED flag to mmap(2), which causes the new mapping to be
...
wired as if by mlock(2).
2003-10-07 00:17:09 +00:00
4da2d7565d
Userland amd64 siginfo support.
2003-10-06 22:57:23 +00:00
ed9d433616
Update to new assmbler needs. Add getcontext.S stub.
2003-10-06 05:30:21 +00:00
f3198687d6
Add SA_SIGINFO support for ARM.
2003-10-05 19:48:39 +00:00
7b066791c8
Remove references to University of California from my copyright notices.
2003-10-05 17:48:49 +00:00
5238256a02
s/sigcontext/siginfo/
2003-10-05 10:46:41 +00:00
b469b163b9
libc bits for siginfo on sh5.
2003-10-05 10:01:49 +00:00
e3ac3a4d0c
Style fixes to reduce diff size against the FreeBSD version, notably:
...
- Use NULL where appropriate.
- No need to pull in stdint.h.
- Get rid of "register" keywords in the hope that the compiler will
do a better job for us. :-)
- Every message ends with a period.
- MMAP() macro is supposed to return MAP_FAILED rather than -1 if it
fails.
- De-__P.
2003-10-05 04:49:46 +00:00
09c5c31fc8
Convert to use register prefixes.
2003-10-03 22:37:35 +00:00
f462c5301e
paren around macro arg
2003-10-03 22:04:44 +00:00
snj