mariadb/newbrt
Bradley C. Kuszmaul 6b654211c0 free diskblocks. Addresses #1195.
git-svn-id: file:///svn/toku/tokudb.1195@7679 c7de825b-a66e-492c-adef-691d508d4ae1
2013-04-16 23:57:25 -04:00
..
checksum-benchmarks
compression-ratio
hash-benchmarks
locking-benchmarks
omt-with-o1-cursors
tests free diskblocks. Addresses #1195. 2013-04-16 23:57:25 -04:00
worker-thread-benchmarks
block_allocator.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
block_allocator.h
bread.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
bread.h
brt-internal.h free diskblocks. Addresses #1195. 2013-04-16 23:57:25 -04:00
brt-old.c Simplified splitting code (don't split fifo. Instead push everything to the child.) Tests run. Addresses #1195. 2013-04-16 23:57:24 -04:00
brt-search.h compile newbrt with icc. addresses #1185 2013-04-16 23:57:19 -04:00
brt-serialize.c free diskblocks. Addresses #1195. 2013-04-16 23:57:25 -04:00
brt-test-helpers.c newbrt tests run. Still no deletes. But the code is smaller than it was (2734 according to sloccount vs 3137 before, which is 13 percent reduction). Addresses #1195. 2013-04-16 23:57:24 -04:00
brt-verify.c Make some tests run better. Addresses #1195. 2013-04-16 23:57:24 -04:00
brt-XY.c
brt.c free diskblocks. Addresses #1195. 2013-04-16 23:57:25 -04:00
brt.h free diskblocks. Addresses #1195. 2013-04-16 23:57:25 -04:00
brtdump.c free diskblocks. Addresses #1195. 2013-04-16 23:57:25 -04:00
brttypes.h problem including brt-types.h in c++ code. c is not always compatible with c++. addresses #1185 2013-04-16 23:57:22 -04:00
cachetable-rwlock.h Get rid of most CIL warnings. Addresses #1185. 2013-04-16 23:57:20 -04:00
cachetable-writequeue.h
cachetable.c turn on the cachetable writer threads. use fast asserts. addresses #1195 2013-04-16 23:57:25 -04:00
cachetable.h newbrt tests run (had to get the keylen computation right.) Still does not reuse blocknums and diskspace. Addresses #1195. 2013-04-16 23:57:24 -04:00
fifo.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
fifo.h Get rid of icc warnings (with port-win warnings.) Addresss #1185. 2013-04-16 23:57:21 -04:00
fingerprint.c Get rid of dead code in fingerprint. Addresses #1195. 2013-04-16 23:57:24 -04:00
hashfun.h
hclip.c
includes.h Addresses #1185 2013-04-16 23:57:21 -04:00
key.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
key.h
kv-pair.h
leafentry.c Make it work better with icc. Addresses #1185. 2013-04-16 23:57:21 -04:00
leafentry.h Try to build a LESWITCHCALL that works with both icc on windows, and on icc and gcc on linux. Addresses #1185. 2013-04-16 23:57:22 -04:00
list.h
lock.h
log-internal.h Addresses #1185 2013-04-16 23:57:21 -04:00
log.c compile with icc 32 bit linux. addresses #1195 2013-04-16 23:57:25 -04:00
log.h Make it work better with icc. Addresses #1185. 2013-04-16 23:57:21 -04:00
logformat.c Addresses #1185 2013-04-16 23:57:21 -04:00
Makefile minor changes to compile with icc and make the windows port easier. addresses #1195 2013-04-16 23:57:25 -04:00
Makefile.include
Makefile.new
memarena.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
memarena.h
memory-debug.c
memory.c Addresses #1185 2013-04-16 23:57:22 -04:00
memory.h Get rid of some compiler warnings. Addresses #1185. 2013-04-16 23:57:20 -04:00
mempool.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
mempool.h compile newbrt with icc. addresses #1185 2013-04-16 23:57:19 -04:00
omt.c Addresses #1185 2013-04-16 23:57:22 -04:00
omt.h
portability.h Start simplification of tree code, in preparation for merge on delete. Addresses #1195. 2013-04-16 23:57:23 -04:00
rbuf.h Get rid of icc warnings (with port-win warnings.) Addresss #1185. 2013-04-16 23:57:21 -04:00
recover.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
roll.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
rwlock.c
tdb-recover.c Get rid of more cil warnings. Addresses #1185. 2013-04-16 23:57:20 -04:00
tdb_logprint.c Get rid of icc warnings (with port-win warnings.) Addresss #1185. 2013-04-16 23:57:21 -04:00
threadpool.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
threadpool.h CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
toku_assert.c Addresses #1185 2013-04-16 23:57:22 -04:00
toku_assert.h turn on the cachetable writer threads. use fast asserts. addresses #1195 2013-04-16 23:57:25 -04:00
trace_mem.c Make it work better with icc. Addresses #1185. 2013-04-16 23:57:21 -04:00
trace_mem.h Get rid of almost all compiler warnings in newbrt. Addresses #1185. 2013-04-16 23:57:20 -04:00
valgrind.suppressions compile with gcc 4.3. addresses #1185 2013-04-16 23:57:21 -04:00
wbuf.h Get rid of some compiler warnings. Addresses #1185. 2013-04-16 23:57:19 -04:00
x1764.c compile using gcc 4.3 (fedora 9) addresses #1195 2013-04-16 23:57:25 -04:00
x1764.h compile using gcc 4.3 (fedora 9) addresses #1195 2013-04-16 23:57:25 -04:00
ybt.c CIL gets upset if includes are not in the same order, because int64_t gets typedefed to different things depending on what you include. So do all the includes in a standard way. Addresses #1185. 2013-04-16 23:57:20 -04:00
ybt.h