Commit Graph

153107 Commits

Author SHA1 Message Date
elad
a6e87ced6c Sync with reality. 2006-10-28 15:34:18 +00:00
elad
40d6264cec Remove bogus sanity check.
The passed size doesn't mean anything really and can only help detect
corrupted configuration files, which should be done in userland anyway.

Note it's possible to trigger a kernel panic by passing a junk
pointer in the 'fingerprint' member of the parameters, but then again
that's true for anything that copies in data from a userland-supplied
pointer. And we have plenty of those.

At the moment, Veriexec only allows the super-user to open the pseudo
device, so it's ~okay. Maybe we should address that in copy(9) or
something?
2006-10-28 15:13:11 +00:00
christos
78921a0243 Initialize decimal_point. 2006-10-28 15:02:02 +00:00
christos
f71e448c3b initialize floating_point as needed. 2006-10-28 14:38:55 +00:00
agc
08c311c029 More cleanup after sha2.h and rmd160.h moved house 2006-10-28 13:05:42 +00:00
elad
066e9189c4 Adapt to changes suggested by yamt@ to get rid of __UNCONST() stuff.
While here, don't leak pathbuf on success.
2006-10-28 11:43:45 +00:00
peter
ba31dc0ca8 - Show statistics immediately, not after a 5 second delay.
- For cbq, cdnr, hfsc and priq: if previous stats aren't available yet
  then don't show the packet/byte rate.
- Place the count check before the delay and after the stats printing.

This fixes PR #32847

ok tron@
2006-10-28 11:43:02 +00:00
peter
9cc4e60c56 Don't allow to initialize an interface with MTU smaller than one.
This prevents a divide by zero and fixes PR #21474.

ok tron@
2006-10-28 11:35:17 +00:00
mrg
8f3077c34d avoid allocating from ksiginfo_pool when we know we aren't going to
use it.  idea partly from yamt.  assert SCHED_ASSERT_UNLOCKED() in
all the places we call ksiginfo_queue() without a ksiginfo.

fixes recent panics detected by LOCKDEBUG.
2006-10-28 08:09:31 +00:00
christos
7016508581 add sys/sha2.h and sys/rmd160.h 2006-10-28 04:25:54 +00:00
tsutsui
1696b0bb2f Put common data for each RX DMA descriptor into a new rxsoft structure
for micro optimization.
2006-10-28 03:42:55 +00:00
freza
ec743521b3 Use curcpu() instead of referencing &cpu_info_primary (which is only
defined by a handful of ports).

OK by christos@
2006-10-28 02:21:36 +00:00
reinoud
8b52dc54b3 Implement another class of `unit not ready' sense handling that is not
fatal.. A `long write in progress' is a retry again later command that is
issued when a device returns immediately after a write request but needs
some time before it can handle read requests.
2006-10-28 01:24:29 +00:00
mrg
4804ddb2d7 what was <crypto/sha2/sha2.h> and <crypto/ripemd160/rmd160.h> is now
<sys/sha2.h> and <sys/rmd160.h>.
2006-10-27 22:50:28 +00:00
uwe
06ddf90514 CONFIGURE_ARGS+=--disable-nls so that tools/gmake doesn't accidentally
pick up tools/gettext that is built when MKMAINTAINERTOOLS=yes

As gettext is built after gmake - this only happens if you either do
an update build and tools/gmake gets rebuilt, or if you use the same
TOOLDIR for several ports.
2006-10-27 22:36:23 +00:00
mrg
3e7961407f make sure we can find <sys/{rmd160,sha2}.h>. make sure the crypto/foo.h
versions of these include something that exists...
2006-10-27 22:32:45 +00:00
uwe
421ab7a955 G/c references to gawk that is no longer in our tree. 2006-10-27 22:31:21 +00:00
mrg
af62440127 add ${.CURDIR}/../../common/lib/libc/hash/{rmd160,sha2} to the .PATH
so that these sources are found after this code moved.

(hi christos!)
2006-10-27 22:25:26 +00:00
elad
145a8382e1 Don't leak handle. 2006-10-27 22:17:09 +00:00
elad
28e2a804c9 For now, also mark "file" entries as "indirect". 2006-10-27 22:14:17 +00:00
uwe
5dd5d540a6 Simplify the __attribute__ ifdef mess now that __used does the right
thing for older gccs.
2006-10-27 22:14:13 +00:00
uwe
b05d0f4d20 Define __used as __unused for gcc's that don't grok __attribute__((__used__))
Those older versions of gcc don't eliminate unused functions/vars
anyway, so it's enough just to suppress the warning.
2006-10-27 22:10:56 +00:00
christos
e1ae6bce13 - don't allocate huge arrays on the stack
- no bogus ; after }; in block statements!
2006-10-27 21:58:59 +00:00
dsl
72c6703a09 Rename 'struct Lst' to 'struct List' and 'struct LstNode' to 'struct 'ListNode'
in lst.d remove a small barrowload of casts from the lst.lib bloatset.
2006-10-27 21:37:25 +00:00
christos
5c60f22be5 need to negate before masking; from abp. 2006-10-27 21:36:50 +00:00
christos
d029b25938 restruct the include files to look like the other hash functions. 2006-10-27 21:25:21 +00:00
christos
a3fadcee99 this moved to common/lib/libc/hash 2006-10-27 21:24:35 +00:00
christos
77c9e41904 this is shared with the kernel now. 2006-10-27 21:23:15 +00:00
christos
31a62606ea Merge kernel and userland rmd160 and sha2 implementation.
XXX: We still install rmd160.h and sha2.h in /usr/include/crypto, unlike
the other hash functions which get installed in /usr/include for compatibility.
2006-10-27 21:20:48 +00:00
dsl
16c8499ed2 Since 'ClientData' is 'void *', nuke almost all the (ClientData) casts. 2006-10-27 21:00:18 +00:00
garbled
3dc1c33272 Add the crypto swcrypto and veriexec psuedo devices. Also add FILEASSOC
and VERIEXEC, following elad's lead.
2006-10-27 20:37:17 +00:00
elad
52bab3d8ab Don't allocate MAXPATHLEN on the stack.
Prompted by and initial diff okay yamt@
2006-10-27 20:16:10 +00:00
christos
c259a653fb add the _ss versions of syslog(3) functions. 2006-10-27 20:05:27 +00:00
christos
1e468fbf99 - flags should be unsigned short instead of short.
- add __SAFE flag, to indicate that we want async-signal-safe functionality
  only.
- add snprintf_ss, and vsnprintf_ss; these are undocumented on purpose.
2006-10-27 20:03:21 +00:00
christos
7dae219b5b add syslog_ss, vsyslog_ss. 2006-10-27 20:01:33 +00:00
christos
d15d9f8a71 add async-signal-safe versions: syslog_ss and vsyslog_ss
multithread-safe syslog_r and vsyslog_r are not async-signal-safe.
2006-10-27 20:00:55 +00:00
christos
abca035cf3 Add signal-safe versions of snprintf and vsnprintf 2006-10-27 19:59:58 +00:00
pooka
eb6da7fc35 fix checkalias true branch: don't unlock or lock twice 2006-10-27 19:54:34 +00:00
garbled
e5bff3e418 A few changes so pnpbus devices can do isadma. Note that isadma still
does not seem to work properly on prep.
2006-10-27 19:52:51 +00:00
drochner
65b9988ba4 gently add some "const", and avoid some casts
compiles with WARNS=4 now (on i386 and alpha)
2006-10-27 19:39:11 +00:00
garbled
8685914d8f Change an instance of curaddr to PHYS_TO_BUS_MEM(t, curaddr) in
_bus_dmamap_load_buffer, otherwise when attempting to DMA a buffer larger
than PAGE_SIZE on an arch where PHYS != BUS it will allways fail.
2006-10-27 19:36:15 +00:00
pooka
0035113cf8 Use spec_fsync for specops vop_fsync: it knows about vflushbuf(), which
is more than what puffs currently knows.  makes e.g. ffs unmount for a
puffs-based device node work.
2006-10-27 19:01:48 +00:00
dogcow
6cb622186f Sync up with rtk 'common macro' changes. 2006-10-27 18:47:07 +00:00
drochner
50ea67a5eb As discussed back in June, remove hmac_sha1 and hmac_md5 from libc, to
reduce bloat and namespace trouble. (both were never public; hmac_sha1
is moved to libcrypt now because it is used locally there)
2006-10-27 18:29:21 +00:00
drochner
d16ceb033e As discussed back in June, move the implementation of hmac_sha1 here
because this is not a public function and crypt-sha1.c is the only
place where it is used (It must have been that way formerly, because
there was a stale prototype in crypt.h all the time.)
Being here, do some RCSID and prototype cleanup, and add as much "const"
as needed to compile these files with WARNS=3.
2006-10-27 18:22:56 +00:00
perry
360359b671 fix a bit of trailing whitespace, and do the space tab dance for a
couple of options so that they work well when commented out
2006-10-27 17:37:44 +00:00
christos
26bb1685bd don't allocate large buffers on the stack. 2006-10-27 16:49:01 +00:00
christos
78a48d28a3 Use strncpy to copy out interface names so that the trailing part of the
buffer is zeroed, and check for overflow.
2006-10-27 15:33:11 +00:00
tnozaki
0c4c29f46b add missing locale syspkg stuffs. 2006-10-27 15:09:46 +00:00
yamt
b89279d039 revert malloc -> kmem_alloc part of uvm_swap.c rev.1.110 because
the current implementation of kmem_free can sleep.
2006-10-27 15:06:04 +00:00