Commit Graph

86080 Commits

Author SHA1 Message Date
Poul-Henning Kamp
f0e185d705 Implement a bio-taskqueue to reduce number of context switches in
disk I/O processing.

        The intent is that the disk driver in its hardware interrupt
        routine will simply schedule the bio on the task queue with
        a routine to finish off whatever needs done.

        The g_up thread will then schedule this routine, the likely
        outcome of which is a biodone() which queues the bio on
        g_up's regular queue where it will be picked up and processed.

        Compared to the using the regular taskqueue, this saves one
        contextswitch.

Change our scheduling of the g_up and g_down queues to be water-tight,
at the cost of breaking the userland regression test-shims.

Input and ideas from:   scottl
2003-02-11 22:30:26 +00:00
Dag-Erling Smørgrav
55b84e8a54 Say hello to Tinderbox 2.0, the choice of a new generation! 2003-02-11 22:17:49 +00:00
Mike Barcroft
8e9b28311e Implement C99's signbit() macro. 2003-02-11 21:56:21 +00:00
Sam Leffler
ee7785a7f8 PFIL_HOOKS optimization: check if at least one hook is present before
munging the IP header to pass to the hooks
2003-02-11 21:48:20 +00:00
Wes Peters
24860c5a80 Add Warner Losh quote.
Submitted by:	Alexandr Kovalenko <never@nevermind.kiev.ua>
2003-02-11 21:38:02 +00:00
Poul-Henning Kamp
2c71b5639f Announce our ability to do DFLTPHYS sized transfers. 2003-02-11 21:30:28 +00:00
Poul-Henning Kamp
7e579a7c0a Announce our ability to do MAXPHYS transfers. 2003-02-11 21:29:53 +00:00
Poul-Henning Kamp
0364fe2cb3 Advertise MAXPHYS upwards, we will split as necessary before we get to the
bottom of things.
2003-02-11 21:24:25 +00:00
Poul-Henning Kamp
99af707c39 Check disk->d_maxsize/dev->si_iosize_max at open time rather than in strategy.
Printf a warning and use DFLTPHYS if the drive has not set a size.
2003-02-11 21:23:34 +00:00
Joerg Wunsch
1e35598c3a Deregister the dev_clone event handler when unloading the module. Bad
things might happen otherwise.

Noticed by:	Michael Reifenberger <root@nihil.reifenberger.com>
2003-02-11 21:12:43 +00:00
Jens Schweikhardt
00ad14c702 Improve a comment and undo a bogus s/a/an/ in a comment. An asm guru
should add more comments explaining which registers hold which variables.

Spotted by:	bde
2003-02-11 20:15:11 +00:00
Tom Rhodes
dc82395af0 Add NOMAN here also. 2003-02-11 19:45:44 +00:00
Sean Chittenden
a3b3a1cf6c Update random(6) to have the ability to randomize a file/stdin based
off of lines or words.  See the man page for details.

Reviewed by:	markm
MFC after:	3 days
2003-02-11 19:32:18 +00:00
Tom Rhodes
24cc1d58d5 Add -DNOMAN to the list. 2003-02-11 19:21:13 +00:00
Hideyuki KURASHINA
7a48e6deb8 Catch up the following with the English version:
1.11 ->  1.12	early-adopter/article.sgml
	1.493 -> 1.495	relnotes/common/new.sgml

Approved by:	hrs (mentor)
2003-02-11 18:56:46 +00:00
Poul-Henning Kamp
15c672a31d Make a mutex to stop the race coming into geom_disk's done routine.
Cut up requests into smaller bits if they are longer than the drivers
disk->d_maxsize or dev->si_iosize_max.

Properly handle the race condition when using g_clone_bio() is used
without having the single-threadedness of g_down/g_up secure locking.
2003-02-11 18:32:31 +00:00
Warner Losh
184210ff8f Update man page to match the new world order after Sam updated us to
NetBSD's 802.11 infrastructure.
2003-02-11 17:40:56 +00:00
Warner Losh
7f511526c2 Commit a nearly real man page describing the current state of devctl.
More work likely needs to happen.  This describes things better than
the old "this man page intentionally left blank" style man page that
I'd committed previously.

Nitpickers: comb nits and commit!
2003-02-11 17:36:38 +00:00
Brian Feldman
d00d7d5c39 Sound devices were being leaked by dsp_open() not actually releasing the
channels it allocates if chn_reset() on them resulted in failure.
ARROW'D!
2003-02-11 16:58:54 +00:00
Yaroslav Tykhiy
e937a222a6 MFCs noted: ftpd(8) -h, -P, "/./"; ftpchroot(5). 2003-02-11 16:54:16 +00:00
Poul-Henning Kamp
8028f16b1b Don't divide by zero if there is no stripewidth specified. 2003-02-11 15:23:41 +00:00
Poul-Henning Kamp
200c8a23cd Typo in last commit. 2003-02-11 15:20:34 +00:00
Dag-Erling Smørgrav
a559222eb5 Force cvsup mirrors to refetch this file. 2003-02-11 15:19:31 +00:00
Poul-Henning Kamp
8a63edc3a1 Better names for struct disk elements: d_maxsize, d_stripeoffset
and d_stripesisze;

Introduce si_stripesize and si_stripeoffset in struct cdev so we
can make the visible to clustering code.

Add stripesize and stripeoffset to providers.

DTRT with stripesize and stripeoffset in various places in GEOM.
2003-02-11 14:57:34 +00:00
Poul-Henning Kamp
bb17c5f69f Propagate DISKFLAG_CANDELETE from struct disk to G_PF_CANDELETE on the
provider.
2003-02-11 14:12:06 +00:00
Yaroslav Tykhiy
f29288c05e Kill unnecessary vertical whitespace. 2003-02-11 14:10:48 +00:00
Poul-Henning Kamp
11eb149478 Wrap a long line. 2003-02-11 14:09:48 +00:00
Yoshihiro Takahashi
5628bf6970 Sort. 2003-02-11 13:19:01 +00:00
Poul-Henning Kamp
392d56b4bd Don't short-circuit zero-length requests of they are BIO_[SG]ETATTR. 2003-02-11 13:13:10 +00:00
Poul-Henning Kamp
a0ca480ce6 Retire D_CANFREE flag. 2003-02-11 12:51:51 +00:00
Poul-Henning Kamp
88b1af7730 Use the SI_CANDELETE flag on the dev_t rather than the D_CANFREE flag
on the cdevsw to determine ability to handle the BIO_DELETE request.
2003-02-11 12:49:58 +00:00
Poul-Henning Kamp
cadb987095 Unconditionally make our provider with G_PF_CANDELETE. 2003-02-11 12:37:04 +00:00
Poul-Henning Kamp
faa7b8b96e Propagate G_PF_CANDELETE to our own providers from the provider we attach to. 2003-02-11 12:36:33 +00:00
Poul-Henning Kamp
b3b3d1b7fe Mark our provider with G_PF_CANDELETE in the cases where this is actually
the case.
2003-02-11 12:35:44 +00:00
Poul-Henning Kamp
eb3543e4cc Mark our struct disk with DISKFLAG_CANDELETE instead of the cdevsw with
D_CANFREE.
2003-02-11 12:35:01 +00:00
Dag-Erling Smørgrav
9794bba8a8 document the current default value for VersionAddendum. 2003-02-11 12:11:15 +00:00
Yaroslav Tykhiy
b8939f6fa6 Use LOG_AUTHPRIV to hide the username attempted during an invalid login
from everyone but sysadmins.

PR:		bin/29487
MFC after:	3 days
2003-02-11 11:58:33 +00:00
Poul-Henning Kamp
83333d4569 Introduce flag field and G_PF_CANDELETE field on providers. 2003-02-11 11:55:40 +00:00
Poul-Henning Kamp
6cadf88d1c Introduce SI_CANDELETE flag on dev_t. 2003-02-11 11:55:02 +00:00
Poul-Henning Kamp
876fbedb4c Switch to use the TSC code i386/i386/tsc.c 2003-02-11 11:43:58 +00:00
Poul-Henning Kamp
70d8e2e9aa Switch to using the TSC code in i386/i386/tsc.c. 2003-02-11 11:43:25 +00:00
Poul-Henning Kamp
dc803e37fb Remove another printf which does not say anything we didn't already know. 2003-02-11 11:02:27 +00:00
Poul-Henning Kamp
81e9eba33a Turn the "updating" flag (back) into two sequence number fields at
either ends of the structure so we have a way to determine if a
snapshot is consistent.
2003-02-11 11:01:26 +00:00
Poul-Henning Kamp
9358a8009a Remove a debugging printf. 2003-02-11 10:59:43 +00:00
Seigo Tanimura
cd6c2a8874 s/IPSSEC/IPSEC/ 2003-02-11 10:51:56 +00:00
Alfred Perlstein
5d1999eeae Fixup printf format. 2003-02-11 08:19:52 +00:00
Yoshihiro Takahashi
4a29c3a740 MFi386: Use DIOCGPC98 ioctl. 2003-02-11 07:32:45 +00:00
Yoshihiro Takahashi
afb7a66e08 Use DIOCGPC98 ioctl to write boot block. 2003-02-11 07:29:14 +00:00
Alfred Perlstein
969e2ea158 Actually link in the attr_{set,get}stack. 2003-02-11 07:28:28 +00:00
Alfred Perlstein
42e1b74af2 Don't lock FILEDESC under PROC.
The locking here needs to be revisited, but this ought to get rid of the
LOR messages that people are complaining about for now.  I imagine either
I or someone else interested with smp will eventually clear this up.
2003-02-11 07:20:52 +00:00