Commit Graph

41 Commits

Author SHA1 Message Date
~hatteb-mitlyd
a1a62f2d1a eviscerate u2_ho_cash 2014-05-12 10:51:53 -07:00
~hatteb-mitlyd
c51892324f convert nash to store value pointers 2014-05-12 10:51:52 -07:00
C. Guy Yarvin
44dfc2118a Style cleanups. 2014-05-07 11:31:54 -07:00
~hatteb-mitlyd
af465d30d5 use u2_none instead of u2_nul in nash 2014-05-07 11:21:16 -07:00
~hatteb-mitlyd
7d9c363d65 Use a bit-mapped patricia trie in nash
this is much faster than the world's worst hashtable, or will be when it
actually works...
2014-05-07 10:17:37 -07:00
~hatteb-mitlyd
e2bf68ab7e Revert "Revert "use nash for cue""
This reverts commit 04caeff12f5e471519082b1c5f3020943df136db.

Medical science made some advances by leaps and bounds. The nash code is
more aggressive with the allocator and leaves more garbage around for
e.g. u2_term_io_init to trip over.
2014-04-30 11:34:43 -07:00
~hatteb-mitlyd
2c93a4efff Revert "use nash for cue"
This reverts commit f691d25bd23716e9411288d3eb9bd4c9b9d33ae1.

Oddly, this causes a failure to boot, on some linux, for some reason
heretofore unknown to medical science.
2014-04-30 11:15:45 -07:00
~hatteb-mitlyd
61c0a6ed2c use nash for cue 2014-04-29 10:03:31 -07:00
~hatteb-mitlyd
b04852fbc4 comments, debugging 2014-04-24 12:42:09 -07:00
~hatteb-mitlyd
aacbe8f4da use a non-noun-based hashtable to jam 2014-04-23 19:55:53 -07:00
~hatteb-mitlyd
955c5a8fb5 make u2_chop deterministic
u2_atom_word happily reads garbage past the end of the buffer. Make it
not do that.
2014-04-18 10:07:40 -07:00
C. Guy Yarvin
1317ac3c1e Clean up requests from dead connections. 2014-04-14 11:01:18 -07:00
Steve Dee
6ae1d505a7 Check return on getrlimit 2014-04-10 18:23:47 -07:00
~hatteb-mitlyd
a0c18f56b2 change to asserting malloc 2014-04-01 17:48:26 -07:00
~hatteb-mitlyd
309f4b31fd try to raise limits 2014-03-29 12:35:40 -07:00
C. Guy Yarvin
0e5645687f Some syntax changes. 2014-03-25 10:46:30 -07:00
C. Guy Yarvin
b0248fb89a Testing web proxying. 2014-03-23 11:44:23 -07:00
C. Guy Yarvin
fe7bd138ee Moar filez pls. 2014-03-20 10:18:45 -07:00
Steve Dee
ba71f7ca11 Trailing whitespace cleanup, C level 2014-03-14 10:50:12 -07:00
C. Guy Yarvin
3b98f316a0 Various fixes and improvements. 2014-03-04 11:17:34 -08:00
C. Guy Yarvin
ce4daf0586 Reverse basket resize. 2014-02-18 00:17:31 -08:00
C. Guy Yarvin
6c22acbfa7 Some memory palliatives. 2014-02-14 17:35:22 -08:00
C. Guy Yarvin
98006a2851 Last checkin on funbreach. 2014-01-06 12:37:42 -08:00
C. Guy Yarvin
e7ae289002 Removed unused runes. 2014-01-04 20:11:15 -08:00
C. Guy Yarvin
0b38cd36c5 Eliminate %wtcn in favor of %wtts. 2013-12-26 21:43:25 -05:00
C. Guy Yarvin
6faa9c453a Merge branch 'master' of github.com:urbit/urbit 2013-12-20 18:49:25 -08:00
C. Guy Yarvin
c57ec891d3 Merge branch 'funky'
Conflicts:
	urb/urbit.pill
	urb/zod/arvo/ames.hoon
2013-12-20 18:45:17 -08:00
C. Guy Yarvin
c3965d0ad3 On the way to funk. 2013-12-20 09:34:13 -08:00
Steven Dee
f2a839426b Ye olde whitespace cleanup part 2
sed -i, glanced at.
2013-12-18 13:17:47 -08:00
Steve Dee
a75e974af9 Fix remaining non-libuv memory leaks
The only leaks remaining are in libuv's fsevents code. They could
be our fault, but it's unlikely.
2013-12-16 18:36:32 -08:00
C. Guy Yarvin
33430d9fbc Fix bus errors with proper truncation. 2013-11-13 10:57:43 -08:00
C. Guy Yarvin
a4baea40fd Fix a bunch of memory leaks and stuff. 2013-11-11 23:09:11 -08:00
C. Guy Yarvin
8bd25d8b32 Use ftruncate() to map checkpoint images as sparse files. 2013-11-11 11:13:47 -08:00
Curtis Yarvin
fd4e5f9e5f Remove debug printf. 2013-11-09 04:05:48 +00:00
C. Guy Yarvin
91fe182f81 Various fixes and improvements... 2013-11-08 15:37:38 -08:00
Steven Dee
c489d2dc7d Remove remainder of sprintf & strcpy calls
Note that there are still some in libuv.
2013-11-01 15:10:41 -04:00
Steven Dee
324c6a235d Remove most instances of strcpy and sprintf
OpenBSD whines about these and recommends using strlcpy / snprintf
instead. Since strlcpy isn't quite universal yet, we use strncpy instead
and be careful about terminating the string. We could implement a
portable strlcpy in terms of strncpy, but that'd add another function to
the namespace.

Yes, usually the length is obviously bounded. Still, pretending
strcpy/sprintf don't exist seems like a great strategy.

N.B. there are still a few occurrences of strcpy and sprintf under f/
and in libuv, but I don't have time to tackle them right now.
2013-10-30 15:25:22 -04:00
Aaron Sokoloski
3755a7620e loom.c -- drop unnecessary fsync in _loom_write that causes lag on some systems.
u2_loom_save syncs after writing anyway.
2013-10-14 21:02:32 +01:00
Steven Dee
1a6f18d09d Consider both FreeBSD and OpenBSD "bsd" 2013-10-07 12:31:41 -04:00
Lev Serebryakov
018cfe5eaf Quick'n'dirty port to FreeBSD (checked on FreeBSD 9.2/amd64) 2013-10-02 18:44:22 +04:00
Christian Carter
3af3130bdc Cleaning up old code 2013-09-28 13:21:18 -07:00