Commit graph

23 commits

Author SHA1 Message Date
Bradley C. Kuszmaul
17e5214663 Refs #1529.
git-svn-id: file:///svn/toku/tokudb@11223 c7de825b-a66e-492c-adef-691d508d4ae1
2013-04-16 23:57:48 -04:00
Bradley C. Kuszmaul
1ed3f26d65 Add #ident Id info to all the newbrt/*.c and newbrt/*.h files. Refs #1529.
git-svn-id: file:///svn/toku/tokudb@11221 c7de825b-a66e-492c-adef-691d508d4ae1
2013-04-16 23:57:48 -04:00
Yoni Fogel
3b42b55489 Closes #1378 Merged all changes into main.
git-svn-id: file:///svn/toku/tokudb@9507 c7de825b-a66e-492c-adef-691d508d4ae1
2013-04-16 23:57:41 -04:00
Bradley C. Kuszmaul
0ed47ea357 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.
git-svn-id: file:///svn/tokudb.1131b+1080a+1185@6439 c7de825b-a66e-492c-adef-691d508d4ae1
2013-04-16 23:57:20 -04:00
Yoni Fogel
b4b9748da8 Addresses #912
Use ybt->size field if ybt->ulen field is unset (for DB_DBT_REALLOC).

git-svn-id: file:///svn/tokudb@4505 c7de825b-a66e-492c-adef-691d508d4ae1
2008-06-12 23:13:04 +00:00
Yoni Fogel
acf688ef2e Closes #912
toku_db_set_* (the dbt setting functions)
no longer require atomicity for ENOMEM cases.
Also, in the case of realloc, we only alloc new memory if
the new size is larger than the old size, or the new size is less than
half of the old size.

Code has some simplifications due to lesser atomicity requirements.

Modified src/tests/test_db_dbt_mem_behavior.c to take into account
that DB_DBT_REALLOC now uses the ulen field.

git-svn-id: file:///svn/tokudb@4504 c7de825b-a66e-492c-adef-691d508d4ae1
2008-06-12 23:07:04 +00:00
Yoni Fogel
fe3cd46d32 Closes #914
git-svn-id: file:///svn/tokudb@4502 c7de825b-a66e-492c-adef-691d508d4ae1
2008-06-12 20:46:09 +00:00
Yoni Fogel
b846ceaa9a Addresses #860
Removed extra mallocs and memcopies for toku_c_get
when the dbts use anything other than DB_DBT_USERMEM

git-svn-id: file:///svn/tokudb@4121 c7de825b-a66e-492c-adef-691d508d4ae1
2008-05-24 17:22:14 +00:00
Yoni Fogel
6ef3961637 Closes #792
Verification of non-null is done before trying a toku-free

git-svn-id: file:///svn/tokudb@3851 c7de825b-a66e-492c-adef-691d508d4ae1
2008-05-08 08:30:50 +00:00
Yoni Fogel
9abb92dc4b Addresses #394
Merge branch tokudb.394 back into trunk

git-svn-id: file:///svn/tokudb@3841 c7de825b-a66e-492c-adef-691d508d4ae1
2008-05-07 20:03:13 +00:00
Bradley C. Kuszmaul
83d416fa09 Merge the 558 branch back to the main with
{{{
svn merge -r3061:3225 https://svn.tokutek.com/tokudb/tokudb.558
}}}

Fixes #630.
Addresses #558.


git-svn-id: file:///svn/tokudb@3226 c7de825b-a66e-492c-adef-691d508d4ae1
2008-04-02 23:40:36 +00:00
Yoni Fogel
3c55e94a95 Addresses #583
We were reading error codes of malloc incorrectly.
errno is not valid unless you know you had an error.

git-svn-id: file:///svn/tokudb@3138 c7de825b-a66e-492c-adef-691d508d4ae1
2008-03-27 21:22:57 +00:00
Rich Prohaska
c6875ae990 set the data and length to 0 for zero length things. dbt's with malloc flags are unchanged. closes #367
git-svn-id: file:///svn/tokudb@2209 c7de825b-a66e-492c-adef-691d508d4ae1
2008-02-08 22:53:08 +00:00
Yoni Fogel
012f706b67 DB_DBT_(MALLOC|REALOC|USERMEM) now behave the same as BDB.
We never touch ulen.  If DB_DBT_USERMEM and ulen is too small, we set size, (do not write anything)
and return DB_BUFFER_SMALL.

This includes test_db_dbt_mem_behavior.c
Closes #146

git-svn-id: file:///svn/tokudb@1034 c7de825b-a66e-492c-adef-691d508d4ae1
2007-12-10 17:04:02 +00:00
Bradley C. Kuszmaul
77ab644ef6 Rename a bunch of symbols to toku_symbols. Addresses #8.
git-svn-id: file:///svn/tokudb@824 c7de825b-a66e-492c-adef-691d508d4ae1
2007-11-29 15:17:46 +00:00
Rich Prohaska
1620ec5815 remove app_private; merge in dup pma;
git-svn-id: file:///svn/tokudb@567 c7de825b-a66e-492c-adef-691d508d4ae1
2007-11-15 14:44:05 +00:00
Bradley C. Kuszmaul
48f0ad74b5 Up
git-svn-id: file:///svn/tokudb@519 c7de825b-a66e-492c-adef-691d508d4ae1
2007-11-14 17:58:38 +00:00
Bradley C. Kuszmaul
0dca556c4f More work on making compare_fun
git-svn-id: file:///svn/tokudb@35 c7de825b-a66e-492c-adef-691d508d4ae1
2007-07-24 11:13:42 +00:00
Bradley C. Kuszmaul
3a649a8ac7 After lots of hacking, the reressions are OK again
git-svn-id: file:///svn/tokudb@34 c7de825b-a66e-492c-adef-691d508d4ae1
2007-07-24 03:28:48 +00:00
Bradley C. Kuszmaul
62cfb06cc9 Things compile again after fiddling with the compare function. There's a malloc bug, however
git-svn-id: file:///svn/tokudb@32 c7de825b-a66e-492c-adef-691d508d4ae1
2007-07-24 01:32:03 +00:00
Bradley C. Kuszmaul
a52b758eb9 pma with compare fun works, and also uses DBTs for most interface work
git-svn-id: file:///svn/tokudb@31 c7de825b-a66e-492c-adef-691d508d4ae1
2007-07-23 20:10:16 +00:00
Bradley C. Kuszmaul
762601f0d2 change my_malloc to toku_malloc. It all works
git-svn-id: file:///svn/tokudb@26 c7de825b-a66e-492c-adef-691d508d4ae1
2007-07-20 18:00:14 +00:00
Bradley C. Kuszmaul
2f8ec84fb8 Rename
git-svn-id: file:///svn/tokudb@4 c7de825b-a66e-492c-adef-691d508d4ae1
2007-07-13 19:37:47 +00:00