Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
138,607 Commits 606 Branches 0 Tags 3.1 GiB
Commit Graph

138607 Commits

Author SHA1 Message Date
he d36a483f83 Make this build with WARNS=3, and enable it. 
Renames local variable syscall to syscall_ent to placate -Wshadow,
and adds const to struct field initialized to string literals.
 2005-07-17 09:45:50 +00:00
he 92e13de495 Only define HAVE_STDBOOL_H if we are using __GNUC__ >= 3. 
Fixes build problem for vax.
 2005-07-17 09:22:03 +00:00
yamt 2a6dc9d02d - introduce PGO_NOBLOCKALLOC and use it for ubc mapping 
to prevent unnecessary block allocations in the case that
  page size > block size.

- ufs_balloc_range: use VM_PROT_WRITE+PGO_NOBLOCKALLOC rather than
  VM_PROT_READ.
 2005-07-17 09:13:35 +00:00
he 7b9dcb30de Follow up the constification of ioctlname() in ktrace.h. 2005-07-17 08:44:08 +00:00
tls 0ac3367f2a Improvements to xbdback (the domain 0 driver that provides "xbd" virtual 
disks to other domains) from Jed Davis, <jld@panix.com>:

	* Issue multiple requests when necessary rather than
	  assuming that arbitrary requests can be mapped into single
	  contiguous virtual address ranges.

	* Don't assume that all data for a request is consecutive
	  in memory.  With some client OSes, it's not.

	The above two changes fix data corruption issues with Linux
	clients with certain filesystem block sizes.

	* Gracefully handle memory or pool allocation failures after
	  beginning to handle a request from the ring.

	* Merge contiguous requests to avoid the "64K turns into 44K + 20K
	  and doubles the transactions per second at the disk" problem
	  caused by the 11-page limit caused by the structure of Xen
	  ring entries.  This causes a very slight performance decrease
	  for sequential 64K I/O if the disk is not already saturated with
	  requests (about 1%) but halves the transactions per second we
	  hit the disk with -- or better.  It even compensates for bizarre
	  Linux behaviour like breaking long requests up into 5.5K pieces.

	* Probably some stuff I forgot to mention.

Disk throughput (though not latency) is now much, much closer to the
"raw hardware" case than it was before.
 2005-07-17 08:23:43 +00:00
christos 553881914b WARNS=3, knf, ansi prototypes 2005-07-17 07:36:26 +00:00
skrll 16044f3d38 Support immediate binding on x86_64. 
Closes PR 30570. Thanks to TAMURA Kent for testing.
 2005-07-17 05:57:21 +00:00
christos 210844e5e7 Increase the default datasize limit to 2GB. 2005-07-17 05:29:54 +00:00
christos 9213de1869 catch up with verified exec defopt changes. 2005-07-17 05:22:34 +00:00
isaki e70e55fa02 Correct a typo. 2005-07-17 05:06:57 +00:00
macallan 440d58aac9 added tslot.4 and dbri.4 2005-07-17 01:27:15 +00:00
macallan 1039a142ea add manpages for tslot and dbri, update tctrl 2005-07-17 01:00:31 +00:00
rpaulo ca2db1fe2b If the DEC board name is not known, remove the leading whitespace in 
sc_name.
This closes PR#30458.

ok'ed Jason Thorpe.
 2005-07-17 00:44:13 +00:00
hubertf 673730506c Add support for reading cloop2 compressed filesystem image, 
enable by putting VND_COMPRESSION into kernel config file.
Written by Cliff Wright, polished up slightly by me.
 2005-07-17 00:08:27 +00:00
wiz beb8063c6a Remove duplicate IPSEC_NAT_T line introduced in previous. 2005-07-16 23:20:00 +00:00
nathanw 0774992407 Add cancellation protection to pthread_once() as per the standard; if 
the once routine is cancelled, the effect on once_control is as if
pthread_once() was never called.

Bug report and fix from PR lib/30734.
 2005-07-16 23:14:53 +00:00
nathanw 5baf7dff84 Oops, don't include an extra bonus deadlock. 2005-07-16 23:12:02 +00:00
nathanw ae6f5e4bba Add once3 test. 2005-07-16 23:02:24 +00:00
nathanw 5f424e4d7c Add a test for pthread_once() calling a function which is cancelled, 
as described in PR lib/30734.
 2005-07-16 23:02:11 +00:00
christos 15d0a3d10e Say what sbp is. 2005-07-16 22:54:34 +00:00
christos b5890013d6 Defopt VERIFIED_EXEC* 2005-07-16 22:50:15 +00:00
christos 1510fe1543 defopt verified_exec. 2005-07-16 22:47:18 +00:00
christos da48256d5a Don't depend on #ifdef VERIFIED_EXEC 2005-07-16 22:44:59 +00:00
christos 7fc92b0291 Explain why table is not const. 2005-07-16 22:32:36 +00:00
christos 58fc70c4d6 WARNS=3 2005-07-16 22:00:01 +00:00
christos ba9d68d01a Add __UNCONST. 2005-07-16 21:59:12 +00:00
christos fd2a8295bb - make input errors print the lint source line number they are called from. 
- simplify a conditional.
 2005-07-16 19:54:00 +00:00
christos 41d3ce981c Fix more lint problems. 2005-07-16 19:50:32 +00:00
christos ed7ae36357 Fix linted comments and use __UNCONST. 2005-07-16 19:48:09 +00:00
christos aeec445c81 Use CPPFLAGS for -D flags otherwise lint does not get it. 2005-07-16 19:36:25 +00:00
macallan 4adcd89999 added dbri audio driver 2005-07-16 19:02:45 +00:00
macallan 3f858509a1 add dbri audio driver 2005-07-16 19:00:39 +00:00
macallan 71b0921a17 driver for the audio part of SUNW,DBRI ISDN/audio controllers 2005-07-16 18:58:49 +00:00
christos c13c780daf Ignore indirect references when linting. 
XXX: All this is done backwards, and is a mess. Really the _foo.c files should
include the _foo function definition, not the foo.c files, like everywhere
else.
 2005-07-16 18:01:38 +00:00
christos 47bfe475f7 Add missing _. 2005-07-16 18:00:40 +00:00
christos ec94556f97 Revert previous; this needs to be fixed differently. 2005-07-16 17:53:36 +00:00
christos 94087acf22 Don't leak kernel external variables to userland. This file is included 
from <sys/sysctl.h>.
 2005-07-16 17:50:26 +00:00
christos d4ddc2a3d4 Don't define __indr_reference() for lint. 2005-07-16 17:41:05 +00:00
christos fce61e5249 revert previous, my include files were out of day. 2005-07-16 17:32:09 +00:00
christos 76e7e3ea51 Cast large constant to int. 2005-07-16 17:20:12 +00:00
christos b5626a3a22 Fix prototype inconsistency, discovered by lint2 fixes. 2005-07-16 17:15:46 +00:00
christos 55b453fa3c don't play #define games with lint. Breaks pass2. 2005-07-16 17:12:36 +00:00
christos 10ca8c1c36 We don't need to define fork anymore (pthread_atfork.c) or setlogin 
(setlogin.c). Lint pass 2 works now.
 2005-07-16 17:12:12 +00:00
uwe 065e2422b3 Match HP 620LX. Tested by Christer Andersson. 
XXX: platid_mask_MACH_HP_LX also matches 360LX.  It's not confirmed
whether touch panel in 360LX is connected in the same way.  We may
need to regroup platid masks.
 2005-07-16 16:36:23 +00:00
christos 0aa8a97934 WARNS=3 and remove default values. 2005-07-16 16:21:05 +00:00
christos 140eaa6eeb - KNF 
- better handling of errors and messages
- pass lint
 2005-07-16 16:20:35 +00:00
christos 078c5c1a41 From David Sainty: If a process dissappears while we are signalling it, don't 
count it as a match/error.
 2005-07-16 15:53:56 +00:00
rtr f44fd3e324 not -> no typo 2005-07-16 14:45:18 +00:00
rpaulo 0024ded950 Remove the -L parameter from the first line of the SYNOPSIS 
and bump the date.

ok Hubert Feyrer.
 2005-07-16 12:35:59 +00:00
hannken 5b2600ba63 Add option `ioaddr=<ioaddr>' to amd64. Already noted in man page. 
Approved by: Frank van der Linden <fvdl@netbsd.org>
 2005-07-16 10:43:00 +00:00