Commit graph

109573 commits

Author SHA1 Message Date
Gleb Smirnoff a1cc90233c Add missed 'the'.
Submitted by:	ru
2004-11-23 12:54:44 +00:00
Gleb Smirnoff 75272f1cf7 Document new changes:
- how node listens to flow control messages
- how node sends flow control messages

Reviewed by:	ru
MFC after:	1 week
2004-11-23 12:27:09 +00:00
Pawel Jakub Dawidek a17dd95f14 - Add missing Giant drop before acquiring the topology lock.
- Move DROP_GIANT()/PICKUP_GIANT() to g_gate_ioctl().
2004-11-23 11:18:26 +00:00
Simon L. B. Nielsen e034d09527 Bump documentation date for last commit, before I get poked by ru
for forgetting it.
2004-11-23 10:56:45 +00:00
Simon L. B. Nielsen bf54f484df Add Dell PERC 4e/Di and Dell PERC 4e/Si to the list of supported
devices.

Submitted by:	Muthu_T@Dell.com
MFC after:	3 days
2004-11-23 10:47:08 +00:00
Gleb Smirnoff 28b24b7b1a Netgraph flow control: change interface status when node receive
LINK_IS_UP/LINK_IS_DOWN messages.

Approved by:	julian (mentor), implicitly
MFC after:	1 week
2004-11-23 09:48:13 +00:00
Gleb Smirnoff 129353b81c Introduce new failure detection algorithm, called NG_ONE2MANY_FAIL_NOTIFY.
It means, that node listens to flow control messages from downstreams
and removes link from list of active links whenever a LINK_IS_DOWN message
is received. If LINK_IS_UP message is received, then links is put
back into list of active links.

Approved by:	julian (mentor), implicitly
MFC after:	1 week
2004-11-23 09:45:47 +00:00
Gleb Smirnoff ea9a9044b1 o Use ng_timeout() instead of timeout(9).
o Implement some netgraph flow control:
- Whenever status of HDLC heartbeat from pear is timed out,
  send NGM_LINK_IS_DOWN message.
- If HDLC link changes status from down to up, send
  NGM_LINK_IS_UP message.

Approved by:	julian (mentor), implicitly
MFC after:	1 week
2004-11-23 09:30:27 +00:00
Ruslan Ermilov d938e8d640 Hopefully fix the "aicasm" build-tool issue when using ${KERNSRCDIR}
different from ${.CURDIR}.

Reported by:	jhb
2004-11-23 09:09:47 +00:00
Marcel Moolenaar 7cb1eb1b27 Unhook the loader subdirectory. The ia64 EFI loader is now build
under ../ia64/efi.
2004-11-23 06:04:51 +00:00
Marcel Moolenaar 25cd518860 Hookup the efi subdirectory. 2004-11-23 06:03:03 +00:00
Marcel Moolenaar 9bf99aadf9 This file was repocopied from src/sys/boot/efi/loader.
Updated for the new build location.
2004-11-23 06:02:03 +00:00
Marcel Moolenaar 5b1f181a89 Visit the efi subdirectory before we visit the machine-specific
subdirectory so that the library built there can be used by the
machine specific boot code.
2004-11-23 05:52:00 +00:00
Marcel Moolenaar f9f47c6e1c This file was repocopied from src/sys/boot/efi/arch/ia64. 2004-11-23 05:38:30 +00:00
Marcel Moolenaar 9c7d032c8b This file was repocopied from src/sys/boot/efi/loader. 2004-11-23 05:37:47 +00:00
Warner Losh 685e700261 It appears that 'kbd' device has never been used and isn't needed.
Build tests show that this isn't used for GENERIC or LINT, and nobody
seemed to know why they existed.
2004-11-23 00:00:43 +00:00
Olivier Houchard df6f945b94 Add arm/mem.c. 2004-11-22 22:20:50 +00:00
Joerg Wunsch 466bc517d1 [Sorry, forgot to commit my source changes in my previous commit.]
Document all options and general usage.

Implement the -a option to bump the annotation_level.  This improves
the Emacs gud behaviour.  You can now supply the following function

(defun gud-gdb-massage-args (file args) (cons "-a" args))

(e.g. by evaluating it from the *scratch* buffer) and get the normal
jump to the source window when browsing the stack.

We should probably eventually supply our own kgdb submode to gud.el.
2004-11-22 16:08:19 +00:00
Joerg Wunsch fa632bb17f Document all options and general usage.
Implement the -a option to bump the annotation_level.  This improves
the Emacs gud behaviour.  You can now supply the following function

(defun gud-gdb-massage-args (file args) (cons "-a" args))

(e.g. by evaluating it from the *scratch* buffer) and get the normal
jump to the source window when browsing the stack.

We should probably eventually supply our own kgdb submode to gud.el.
2004-11-22 16:07:28 +00:00
Hartmut Brandt d96b4f0130 Get rid of the shell table sentinel. It isn't needed anymore because
JobMatchShell returns NULL when no shell name matches since revision 1.51.
2004-11-22 15:45:57 +00:00
Joerg Wunsch 05127e036e Fix the abuse of Ar macros for designating flag options, use Fl instead. 2004-11-22 15:43:46 +00:00
Robert Watson ca127a3e80 Remove "Unlocked read" annotations associated with previously unlocked
use of socket buffer fields in the TCP input code.  These references
are now protected by use of the receive socket buffer lock.

MFC after:	1 week
2004-11-22 13:16:27 +00:00
Yaroslav Tykhiy ebd83647a4 When looking for a virtual host to handle the connection,
stop the search on the first match for efficiency.

Submitted by:	Nick Leuta
2004-11-22 11:10:04 +00:00
Yaroslav Tykhiy 4cbc4ad644 Calling pam_chauthtok() isn't really needed since
an FTP user has no chance to change password anyway.

Submitted by:	Nick Leuta
2004-11-22 11:02:42 +00:00
Yaroslav Tykhiy e897216f45 Don't log the chroot dir on every command since it's constant for a session.
Log it once at the beginning of the session instead.  OTOH, log wd each
time for the sake of better auditing and consistent log format.

Proposed by:	Nick Leuta <skynick -at- mail.sc.ru>
2004-11-22 10:48:29 +00:00
Yaroslav Tykhiy 7cdd3cb70b Always log remote IP.
PR:		bin/59773
2004-11-22 10:27:16 +00:00
Yaroslav Tykhiy 04683b2c35 Treat host name buffers consistently. 2004-11-22 10:16:43 +00:00
Pyun YongHyeon 38fa13a6a2 Make hme(4) mpsafe
- Let hme_start()/hme_init() acquire lock and then call
   hme_start_locked()/hme_init_locked() respectivly.
 - Teardown interrupt handler before hme_detach().
 - Remove IFF_NEEDSGIANT flag and mark interrupt handler INTR_MPSAFE.
 - Set callout handler to CALLOUT_MPSAFE.
 - Add locks in hme MII interface.

Reviewed by:	jake
Tested by:	Julian C. Dunn  <jdunn at opentrend dot net>
MFC after:	2 weeks
2004-11-22 06:46:30 +00:00
David Schultz 40ca336b7c Remove some information that only applies to the old a.out core dump format. 2004-11-22 03:15:21 +00:00
Greg Lehey b4425fbe38 Print dates the right way round. 2004-11-21 23:05:06 +00:00
Marcel Moolenaar 2ba0042660 Remove struct ia64_itir and use a plain old uint64_t instead. 2004-11-21 21:40:08 +00:00
Olivier Houchard b1ff74ebb3 Cleanup. 2004-11-21 19:41:27 +00:00
Olivier Houchard 7fc1bf0371 Set the frame pointer to 0 in fork_trampoline(). 2004-11-21 19:33:47 +00:00
Olivier Houchard 9ebe15a1d7 Implement breakpoints and single stepping on arm.
Obtained from:	NetBSD
2004-11-21 18:11:39 +00:00
Olivier Houchard 75f88c7ca4 Do not attempt to skip a breakpoint that is a result of a software single step,
or bad things happen.
2004-11-21 18:11:02 +00:00
Warner Losh 4b87c653f9 Properly wither the geom container on detach. This will allow one to
then later reload fdc and not have duplicate fd devices in dev.

# Maybe this should be moved to a convenience function.

Reviewed by: phk
2004-11-21 16:25:21 +00:00
Yoshihiro Takahashi ea54c3694c uart_i8251_ops is gone. 2004-11-21 15:18:35 +00:00
Robert Watson 98734750b4 s/send/sent/ in comment describing TCPS_SYN_RECEIVED. 2004-11-21 14:38:04 +00:00
David Schultz c17ff94938 Neither of the arguments to closef() can be NULL anymore, so don't
check for that.
2004-11-21 11:06:24 +00:00
Scott Long 4c10e55d26 Remove an extra #include 2004-11-21 06:28:35 +00:00
Scott Long 938eaf8e6b MFC amd64:
Consolidate all of the bounce tests into the BUS_DMA_COULD_BOUNCE flag.
  Allocate the bounce zone at either tag creation or map creation to help
  avoid null-pointer derefs later on.  Track total pages per zone so that
  each zone can get a minimum allocation at tag creation time instead of
  being defeated by mis-behaving tags that suck up the max amount.
2004-11-21 04:43:28 +00:00
Scott Long 25590d09b9 Consolidate all of the bounce tests into the BUS_DMA_COULD_BOUNCE flag.
Allocate the bounce zone at either tag creation or map creation to help
avoid null-pointer derefs later on.  Track total pages per zone so that
each zone can get a minimum allocation at tag creation time instead of
being defeated by mis-behaving tags that suck up the max amount.
2004-11-21 04:15:26 +00:00
Scott Long 1248408df3 Don't force busdma to pre-allocate bounce pages for static allocations. 2004-11-21 04:03:29 +00:00
Scott Long 8a40c10eb9 Don't force busdma to pre-allocate bounce pages for the parent tag. 2004-11-21 04:02:36 +00:00
Tim J. Robbins c2089f05ef Remove regression tests for the obsolete rune interface. 2004-11-21 03:21:22 +00:00
Tim J. Robbins bd9643b135 Fix computation of the 'n' argument to mbrtowc (through XMBRTOWC) to avoid
reading past 'stop' in various places when converting multibyte characters.
Reading too far caused truncation to not be detected when it should have
been, eventually causing regexec() to loop infinitely in with certain
combinations of patterns and strings in multibyte locales.

PR:		74020
MFC after:	4 weeks
2004-11-21 03:14:32 +00:00
Max Laier ed0cf68945 Commit ALTQ-patch for ed(4).
Requested and tested by:	pav
MFC after:			1 week
2004-11-21 02:42:09 +00:00
Marcel Moolenaar 57a8f01dc1 o Support for the i8251 wasn't there. Remove the files.
o  Remove the headers with IC register definitions. The headers are
   now taken from sys/dev/ic
2004-11-21 01:51:37 +00:00
Marcel Moolenaar 0204ed7a3f Include the header with the register definitions from sys/dev/ic. They
are shared now.
2004-11-21 01:43:27 +00:00
Marcel Moolenaar 43f0d5705e This file was repocopied from sys/dev/uart/uart_dev_z8530.h. 2004-11-21 01:34:15 +00:00