Max Okumoto [Sat, 14 May 2005 22:53:17 +0000 (22:53 +0000)]
FreeBSD-Date: 2005/05/12 17:40:16
Author: jmallett
Include <err.h>. On my system CVS Make(1) has an implicit declaration of
warnc(3) without it.
Max Okumoto [Sat, 14 May 2005 22:52:18 +0000 (22:52 +0000)]
FreeBSD-Date: 2005/05/12 15:16:44
Author: harti
Fix the fix for bin/72510 applied in job.c:1.70. Actually there have
been two maxJobs variables: one static in job.c and one global used in
main.c and parse.c. Makeing one global out of these was the wrong way
to fix the problem. Instead rename the global one to jobLimit and keep
maxJobs static in job.c.
Suggested by: rwatson
PR: bin/72510
Max Okumoto [Sat, 14 May 2005 22:49:16 +0000 (22:49 +0000)]
Use strpbrk() instead of array to match characters for meta characters.
This patch was mixed in with the one I submited
FreeBSD-Date: 2005/05/12 11:40:57
Change in FreeBSD: harti (idea taken, but done differently)
Joerg Sonnenberger [Sat, 14 May 2005 09:29:22 +0000 (09:29 +0000)]
Remove an incorrect free. The code path should normally not be hit, but
better safe than sorry.
Joerg Sonnenberger [Sat, 14 May 2005 08:21:54 +0000 (08:21 +0000)]
Reorder initialisation by calling if_initname before vr_reset.
Nuke long unused variable.
Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
Max Okumoto [Sat, 14 May 2005 06:26:22 +0000 (06:26 +0000)]
Remove extranious blankline.
Caught by harti.
Max Okumoto [Sat, 14 May 2005 05:52:28 +0000 (05:52 +0000)]
Fix spelling error.
Caught by harti
Max Okumoto [Sat, 14 May 2005 04:51:22 +0000 (04:51 +0000)]
Constify ProcExec()
Noticed by Harti.
Joerg Sonnenberger [Sat, 14 May 2005 01:33:35 +0000 (01:33 +0000)]
Move if_initname call before xl_reset to ensure correct initialisation.
Submitted-by: Sepherosa Ziehau <sepherosa@gmail.com>
Joerg Sonnenberger [Fri, 13 May 2005 22:30:11 +0000 (22:30 +0000)]
Add an explicit abort in printaddr to avoid unused variable warnings.
This is in line with the surrounding code.
Joerg Sonnenberger [Fri, 13 May 2005 22:28:24 +0000 (22:28 +0000)]
Fix pointer arithemetic.
Joerg Sonnenberger [Fri, 13 May 2005 21:31:49 +0000 (21:31 +0000)]
WARNS=6 clean already.
Joerg Sonnenberger [Fri, 13 May 2005 17:40:44 +0000 (17:40 +0000)]
Don't activate -funit-at-a-time with -O2 and higher by default, when
propolice is active. There are some bad interactions which can lead
to false negatives under rare circumstances.
Sascha Wildner [Fri, 13 May 2005 15:51:10 +0000 (15:51 +0000)]
Perl is gone from base. You can use your own style again.
David Rhodus [Thu, 12 May 2005 01:41:07 +0000 (01:41 +0000)]
Fix kernel build issue. Missing function argument.
Joerg Sonnenberger [Wed, 11 May 2005 20:58:35 +0000 (20:58 +0000)]
Use device_printf / if_printf and stop abusing the ifnet fields for
internal purposes.
Joerg Sonnenberger [Wed, 11 May 2005 20:36:06 +0000 (20:36 +0000)]
Remove unused second argument to tulip_intr_handler.
Joerg Sonnenberger [Wed, 11 May 2005 20:05:56 +0000 (20:05 +0000)]
Replace local tulip_mbuf_compress with m_defrag call.
Matthew Dillon [Wed, 11 May 2005 19:58:19 +0000 (19:58 +0000)]
Document additional work in last commit. Bumped library to ld-elf.so.2
Matthew Dillon [Wed, 11 May 2005 19:47:09 +0000 (19:47 +0000)]
Major TLS cleanups.
* The RTLD no longer attempts to set up an initial TLS. The job has been
moved to libc. The RTLD does not use TLS itself. The
allocate_initial_tls() function has been removed.
* The _init_tls() call made by CRT1 is no mandatory.
* The RTLD no longer calls the library _init() functions early. Instead it
sets up a callback that allows CRT1 to call the library _init() functions
after it has called _init_tls().
* LIBC now has a _libc_init_tls() function which is weakly bound as
_init_tls(). This function initializes the TLS for the program (generally
by calling the _rtld_allocate_tls() and then doing any required
post-allocation initialization). In the absence of a threading library
this function is basically it.
Threading libraries may replace _init_tls() with their own version. This
does not yet occur as of this commit. The threading library would then
almost certainly call _libc_init_tls() and then make further modifications
to the TCB as necessary (e.g. populate the tcb_pthread field).
* RTLD's allocate_tls() no longer does anything fancy. The 'old_tcb' argument
and concept has been removed.
* We have a new tcb_errno_p field in the TCB, but as of this commit it
is not yet being used.
Reviewed-by: Joerg Sonnenberger <joerg@britannica.bec.de>,
David Xu <davidxu@crater.dragonflybsd.org>
Joerg Sonnenberger [Wed, 11 May 2005 19:40:00 +0000 (19:40 +0000)]
Simplify code by using IF_DRAIN.
Joerg Sonnenberger [Wed, 11 May 2005 15:50:14 +0000 (15:50 +0000)]
typo
Joerg Sonnenberger [Wed, 11 May 2005 15:45:45 +0000 (15:45 +0000)]
Merge from vendor branch GCC:
Remove two unnecessary entries.
Joerg Sonnenberger [Wed, 11 May 2005 15:45:45 +0000 (15:45 +0000)]
Remove two unnecessary entries.
David Xu [Wed, 11 May 2005 12:59:58 +0000 (12:59 +0000)]
I am lost. if thread is blocked on mutex, _thread_kern_sched_state()
will rewrite fname and lineno. :(
David Xu [Wed, 11 May 2005 12:50:09 +0000 (12:50 +0000)]
Override _flockfile_debug.
David Xu [Wed, 11 May 2005 12:45:57 +0000 (12:45 +0000)]
o Fix obsolete comment.
o Fix stub function names.
Joerg Sonnenberger [Wed, 11 May 2005 12:10:43 +0000 (12:10 +0000)]
Remove obsolete comment.
Pointed-out-by: Sepherosa Ziehau <sepherosa@gmail.com>
Joerg Sonnenberger [Wed, 11 May 2005 12:07:21 +0000 (12:07 +0000)]
Remove sc_unit and use if_printf / device_printf instead.
Max Okumoto [Tue, 10 May 2005 23:26:17 +0000 (23:26 +0000)]
Remove some extra white space
Jeffrey Hsu [Tue, 10 May 2005 15:48:10 +0000 (15:48 +0000)]
Implement TCP Appropriate Byte Counting.
Reviewed by Noritoshi Demizu, demizu@dd.iij4u.or.jp.
Misunderstanding of spec clarified by Mark Allman.
Joerg Sonnenberger [Tue, 10 May 2005 15:29:53 +0000 (15:29 +0000)]
Replace __offsetof with similiar pointer expression and use "m"
contraint. This generates the correct assembly as before, but removes
the hard coded offsets.
Work-around-by: YONETANI Tomokazu <qhwt+dfly@les.ath.cx>
Matthew Dillon [Mon, 9 May 2005 17:36:59 +0000 (17:36 +0000)]
Do not allow installworld to proceed if the currently running kernel
contains the TLS fork bug.
Matthew Dillon [Mon, 9 May 2005 17:24:14 +0000 (17:24 +0000)]
Give the DEVELOPMENT branch a subversion so we can do build/installworld
checks on the kernel rev currently running.
Joerg Sonnenberger [Mon, 9 May 2005 16:05:54 +0000 (16:05 +0000)]
Typo in '(' description.
Joerg Sonnenberger [Mon, 9 May 2005 13:37:27 +0000 (13:37 +0000)]
0x7f is only a control character.
David Xu [Mon, 9 May 2005 13:28:40 +0000 (13:28 +0000)]
Eliminate PS_FILE_WAIT state, instead use FILE locking code in libc.
David Xu [Mon, 9 May 2005 12:43:40 +0000 (12:43 +0000)]
Add locking for FILE.
Obtained from: FreeBSD
YONETANI Tomokazu [Mon, 9 May 2005 09:53:07 +0000 (09:53 +0000)]
Fix duplicate bitmask for _A and _G.
Joerg Sonnenberger [Sun, 8 May 2005 18:17:05 +0000 (18:17 +0000)]
Use ifq_handoff instead of the hand-rolled version with inlined
tunstart.
Matthew Dillon [Sun, 8 May 2005 18:14:56 +0000 (18:14 +0000)]
Initial commit of the generation utility and configuration files for the
new system call mapping subsystem.
Joerg Sonnenberger [Sun, 8 May 2005 18:11:02 +0000 (18:11 +0000)]
Use ifq_handoff instead of handrolling it.
Joerg Sonnenberger [Sun, 8 May 2005 17:52:06 +0000 (17:52 +0000)]
atm_output did almost the same as ifq_handoff, it just skipped
the handling of M_MCAST. Therefore let it call the real thing and
get that case for free.
Joerg Sonnenberger [Sun, 8 May 2005 17:42:35 +0000 (17:42 +0000)]
Replace ifq_handoff like code with a call to ifq_handoff.
Joerg Sonnenberger [Sun, 8 May 2005 16:10:37 +0000 (16:10 +0000)]
Catch up with ctype.h.
Joerg Sonnenberger [Sun, 8 May 2005 15:55:15 +0000 (15:55 +0000)]
Nuke ctypeio.c and associated ctype to rune table conversion,
we never used it anyway.
Joerg Sonnenberger [Sun, 8 May 2005 15:52:05 +0000 (15:52 +0000)]
Bite the bullet and add real masks isprint and isgraph. Use this as
excuse to make the flags identical to the rune version by adding _A
for the alpha class and by making the hexdigit class self contained.
Joerg Sonnenberger [Sun, 8 May 2005 13:11:06 +0000 (13:11 +0000)]
Add handling of R_386_TLS_TPOFF32.
Submitted-by: Nathaniel W. Filardo <nwf@andrew.cmu.edu>
Joerg Sonnenberger [Sun, 8 May 2005 11:45:32 +0000 (11:45 +0000)]
ALTQ
Joerg Sonnenberger [Sun, 8 May 2005 11:01:26 +0000 (11:01 +0000)]
Left-over from the old mbuf chain tagging.
Fix-by: YONETANI Tomokazu <qhwt+dfly@les.ath.cx>
Simon Schubert [Sun, 8 May 2005 00:44:59 +0000 (00:44 +0000)]
Clean generated directories
Simon Schubert [Sun, 8 May 2005 00:35:48 +0000 (00:35 +0000)]
Fix compliation with gcc40: Move structure definition within headers.
Taken-from: binutils-2.16
Matthew Dillon [Sun, 8 May 2005 00:12:22 +0000 (00:12 +0000)]
incore() is used to detect logical block number collisions, and other
callers will check B_INVAL on return. Do not return a false negative
if the buffer we find happens to be B_INVAL as this could result in
duplicate buffers in the buffer cache. As of the red-black tree code,
which detects duplicate entries, the case will immediately panic the machine.
MFC: 2 weeks
Simon Schubert [Sun, 8 May 2005 00:07:39 +0000 (00:07 +0000)]
Fix compilation with gcc40: Move function prototype from function to
file scope.
Simon Schubert [Sun, 8 May 2005 00:06:46 +0000 (00:06 +0000)]
Clean generated files/directories.
Simon Schubert [Sat, 7 May 2005 23:42:39 +0000 (23:42 +0000)]
Fix compilation with gcc40: Don't declare static variable extern.
Simon Schubert [Sat, 7 May 2005 23:39:51 +0000 (23:39 +0000)]
Fix compilation with gcc40: Don't cast lvalue for postincrement.
Simon Schubert [Sat, 7 May 2005 23:35:13 +0000 (23:35 +0000)]
Fix compliation with gcc40: Don't export ROOTKEY with wrong type at all.
Simon Schubert [Sat, 7 May 2005 23:32:57 +0000 (23:32 +0000)]
Fix compliation with gcc40: Don't cast the lvalue, instead cast the rvalue.
Simon Schubert [Sat, 7 May 2005 23:26:12 +0000 (23:26 +0000)]
Fix compilation with gcc40: *_hash() take (unsigned char *) as
first parameter.
Simon Schubert [Sat, 7 May 2005 23:24:44 +0000 (23:24 +0000)]
Fix compilation with gcc40: Make `padchar' char *, like all other
char pointers in this function.
Simon Schubert [Sat, 7 May 2005 23:20:43 +0000 (23:20 +0000)]
Fix compilation with gcc40: Move function prototypes from funtion to
file scope.
This stone-old C sucks the skin off a dead goat. Or even off a living one.
Simon Schubert [Sat, 7 May 2005 22:50:05 +0000 (22:50 +0000)]
Fix compilation with gcc40: Swap #include order
Simon Schubert [Sat, 7 May 2005 22:34:51 +0000 (22:34 +0000)]
Fix compilation with gcc40: Move structure definitions up so that array
type in included header "extern.h" isn't incomplete.
Simon Schubert [Sat, 7 May 2005 22:29:39 +0000 (22:29 +0000)]
Fix compilation with gcc40: un-static-ize variable to match extern
declaration.
Simon Schubert [Sat, 7 May 2005 22:05:41 +0000 (22:05 +0000)]
Fix compilation with gcc40: Move declaration of variable so that type isn't
incomplete.
Simon Schubert [Sat, 7 May 2005 21:22:32 +0000 (21:22 +0000)]
Fix compilation with gcc40: Only use -fno-stack-protector with gcc versions
which have propolice support.
Simon Schubert [Sat, 7 May 2005 20:00:20 +0000 (20:00 +0000)]
Make getty compile with gcc40: reorder/include headers to resolve
incomplete type.
Simon Schubert [Sat, 7 May 2005 19:55:31 +0000 (19:55 +0000)]
Clean generated files
Simon Schubert [Sat, 7 May 2005 19:48:16 +0000 (19:48 +0000)]
Fix compilation for gcc40: Don't declare variable with conflicting types.
Simon Schubert [Sat, 7 May 2005 18:33:25 +0000 (18:33 +0000)]
Fix compilation with gcc40: Remove -fwritable-strings from CFLAGS.
Simon Schubert [Sat, 7 May 2005 18:05:14 +0000 (18:05 +0000)]
Make this compile with gcc40: rename variables so that they don't collide
with those from ncurses. Don't ask me why ncurses exports PC, BC and UP.
Simon Schubert [Sat, 7 May 2005 18:03:39 +0000 (18:03 +0000)]
Make this compile with gcc40: Move function prototypes from function scope
to file scope.
Simon Schubert [Sat, 7 May 2005 18:02:35 +0000 (18:02 +0000)]
Make hack compile with gcc40: Remove previously deprecated, now
unsupported option -fwritable-strings.
This might break hack in some way, though a quick check didn't show
anything. Takers for cleaning this code?
Sascha Wildner [Sat, 7 May 2005 17:38:34 +0000 (17:38 +0000)]
Remove PCVT kernel part and mop up.
Simon Schubert [Sat, 7 May 2005 16:22:42 +0000 (16:22 +0000)]
Fix gcc40 compilation of lib/libsvg by merging from FreeBSD:
bde 2004/03/11 05:38:54 PST
FreeBSD src repository
Modified files:
sys/amd64/include cpufunc.h
sys/i386/include cpufunc.h
Log:
Don't implement anything in the ffs family in <machine/cpufunc.h>
in the non-_KERNEL case. This "fixes" applications that include
this "kernel-only" header and also include <strings.h> (or get
<strings.h> via the default _BSD_VISIBLE pollution in <string.h>.
In C++ there was a fatal error: the declaration specifies C linkage
but the implementation gives C++ linkage. In C there was only a
static/extern mismatch if the headers were included in a certain order
order, and a partially redundant declaration for all include orders;
gcc emits incomplete or wrong diagnostics for these, but only for
compiling with -Wsystem-headers and certain other warning options, so
the problem was usually not seen for C.
Ports breakage reported by: kris
Revision Changes Path
1.142 +4 -0 src/sys/amd64/include/cpufunc.h
1.139 +4 -0 src/sys/i386/include/cpufunc.h
Simon Schubert [Sat, 7 May 2005 15:33:05 +0000 (15:33 +0000)]
Fix gcc40 compilation by removing casts of lvalues.
This gives us back one warning (with gcc34, gcc40 emits much more
anyways). The correct fix would be sys/cdefs.h and __DECONST, but as this
is contrib source anyways I don't feel like fixing warning messages here.
Simon Schubert [Sat, 7 May 2005 15:01:36 +0000 (15:01 +0000)]
Move function prototype from function scope to file scope to
satisfy gcc40.
Simon Schubert [Sat, 7 May 2005 13:53:47 +0000 (13:53 +0000)]
Attempt to make code more readable
Simon Schubert [Sat, 7 May 2005 13:32:24 +0000 (13:32 +0000)]
K&R -> modern C
Simon Schubert [Sat, 7 May 2005 13:29:35 +0000 (13:29 +0000)]
Move function prototype from function scope to file scope to
satisfy gcc40.
Sascha Wildner [Sat, 7 May 2005 13:06:44 +0000 (13:06 +0000)]
- Use standard types.
- Use TAB for indentation.
- Minor style(9) fixes.
David Xu [Sat, 7 May 2005 10:08:08 +0000 (10:08 +0000)]
Sort names.
David Xu [Sat, 7 May 2005 09:59:58 +0000 (09:59 +0000)]
Remove unused function.
David Xu [Sat, 7 May 2005 09:29:46 +0000 (09:29 +0000)]
1. Fix symbols needed by libthread_db.
2. Add mechanism to report thread events to debugger.
David Xu [Sat, 7 May 2005 09:25:44 +0000 (09:25 +0000)]
Add libthread_db, a thread debugging support library.
David Xu [Sat, 7 May 2005 08:28:06 +0000 (08:28 +0000)]
Add prototypes of proc service API, every debugger must provide the
set of API before it can load thread debugging assistant library:
libthread_db.
David Xu [Sat, 7 May 2005 08:23:51 +0000 (08:23 +0000)]
Define type lwpid_t, it will be used for thread debugger code.
David Xu [Sat, 7 May 2005 07:50:51 +0000 (07:50 +0000)]
Fix comment of error member, now errno is TLS based.
David Xu [Sat, 7 May 2005 07:39:14 +0000 (07:39 +0000)]
Backout revision 1.5, the pthread->error was used as internal
communication variable between threads, two tls_set_tcb calls
are not signal safe.
Sascha Wildner [Sat, 7 May 2005 02:11:25 +0000 (02:11 +0000)]
Correct typo: vender -> vendor
Liam J. Foy [Fri, 6 May 2005 17:16:18 +0000 (17:16 +0000)]
- Produce more informative output to syslog. Generally improve/clean error handling
Ok: dillon
YONETANI Tomokazu [Fri, 6 May 2005 14:25:17 +0000 (14:25 +0000)]
Backout the last commit. _S is also true for a few control codes for which
isprint() should evaluate to false.
Simon Schubert [Fri, 6 May 2005 13:16:08 +0000 (13:16 +0000)]
Fix compilation with gcc-4.0: static int debug can't be extern.
Simon Schubert [Fri, 6 May 2005 11:52:02 +0000 (11:52 +0000)]
Bring in fix from FreeBSD/cperciva:
Log:
If we are going to
1. Copy a NULL-terminated string into a fixed-length buffer, and
2. copyout that buffer to userland,
we really ought to
0. Zero the entire buffer
first.
Security: FreeBSD-SA-05:08.kmem
Thanks to Colin Percival for notifying us!
Simon Schubert [Fri, 6 May 2005 11:27:51 +0000 (11:27 +0000)]
Bring in fix from FreeBSD/cperciva:
Log:
Correct improper permissions on /dev/iir. The earlier permissions
of 0644 allowed for people to do Evil Things via ioctl(2).
Security: FreeBSD-SA-05:06.iir
Thanks to Colin Percival for notifying us!
Sascha Wildner [Thu, 5 May 2005 22:57:45 +0000 (22:57 +0000)]
Remove DEC Alpha support.
Max Okumoto [Thu, 5 May 2005 09:08:42 +0000 (09:08 +0000)]
Cleanup globals in str.c
o Create ArgArray object.
o Rewrite brk_string() and MAKEFLAGS_break() to use ArgArray instead
of global variables.
o Rewrite code that uses brk_string() and MAKEFLAGS_break() to use
ArgArray.
o Remove str_init() code.
Max Okumoto [Thu, 5 May 2005 09:08:09 +0000 (09:08 +0000)]
Cleanup MAKEFLAGS_break()
Max Okumoto [Thu, 5 May 2005 09:07:51 +0000 (09:07 +0000)]
Cleanup brk_string()