]> git.saurik.com Git - redis.git/log
redis.git
14 years agoAdd sds function that can be called with va_list
Pieter Noordhuis [Thu, 2 Sep 2010 17:35:07 +0000 (19:35 +0200)] 
Add sds function that can be called with va_list

14 years agoChanged reply buildup internals
Pieter Noordhuis [Thu, 2 Sep 2010 17:18:55 +0000 (19:18 +0200)] 
Changed reply buildup internals

14 years agoRemoved unneeded function
Pieter Noordhuis [Thu, 2 Sep 2010 12:31:25 +0000 (14:31 +0200)] 
Removed unneeded function

14 years agoUse existing reply functions where possible
Pieter Noordhuis [Thu, 2 Sep 2010 12:30:56 +0000 (14:30 +0200)] 
Use existing reply functions where possible

14 years agoFree the sds in addReplySds when it cannot be added to the reply
Pieter Noordhuis [Thu, 2 Sep 2010 12:19:15 +0000 (14:19 +0200)] 
Free the sds in addReplySds when it cannot be added to the reply

14 years agoIntialize bufpos in the fake client
Pieter Noordhuis [Thu, 2 Sep 2010 12:17:53 +0000 (14:17 +0200)] 
Intialize bufpos in the fake client

14 years agoUse specialized function to add multi bulk reply length
Pieter Noordhuis [Thu, 2 Sep 2010 10:38:34 +0000 (12:38 +0200)] 
Use specialized function to add multi bulk reply length

14 years agoDon't build a reply when replaying the AOF
Pieter Noordhuis [Mon, 30 Aug 2010 14:51:39 +0000 (16:51 +0200)] 
Don't build a reply when replaying the AOF

14 years agoWrapper for adding unknown multi bulk length to reply list
Pieter Noordhuis [Mon, 30 Aug 2010 14:02:06 +0000 (16:02 +0200)] 
Wrapper for adding unknown multi bulk length to reply list

14 years agoRefactor reply buildup for speed on large multi bulk replies
Pieter Noordhuis [Mon, 30 Aug 2010 12:44:34 +0000 (14:44 +0200)] 
Refactor reply buildup for speed on large multi bulk replies

14 years agoShow the current throughput while benchmarking
Pieter Noordhuis [Mon, 30 Aug 2010 09:25:02 +0000 (11:25 +0200)] 
Show the current throughput while benchmarking

14 years agoRefactor reply parsing code in redis-benchmark for efficiency
Pieter Noordhuis [Mon, 30 Aug 2010 09:14:54 +0000 (11:14 +0200)] 
Refactor reply parsing code in redis-benchmark for efficiency

14 years agoFix for a race in BGSAVE that may result in some data not being saved as soon as...
antirez [Mon, 30 Aug 2010 08:32:32 +0000 (10:32 +0200)] 
Fix for a race in BGSAVE that may result in some data not being saved as soon as possible (when the configured saving triggers should fire). Also known as Issue 313, more details there in the google code issue.

14 years agofix for the prev fix
antirez [Fri, 27 Aug 2010 15:06:36 +0000 (17:06 +0200)] 
fix for the prev fix

14 years agoFix for bug 312, yet to verify in a couple of minutes...
antirez [Fri, 27 Aug 2010 15:04:26 +0000 (17:04 +0200)] 
Fix for bug 312, yet to verify in a couple of minutes...

14 years agoFixed another instace of the Issue 173
antirez [Fri, 27 Aug 2010 10:46:10 +0000 (12:46 +0200)] 
Fixed another instace of the Issue 173

14 years agoFixed segfault in freeMemoryIfNeeded due to the fact that keys are now sds strings...
antirez [Fri, 27 Aug 2010 09:01:03 +0000 (11:01 +0200)] 
Fixed segfault in freeMemoryIfNeeded due to the fact that keys are now sds strings and not objects in the main hash table, thanks to Anthony Lauzon for spotting the bug and providing a patch.

14 years agointset loading bug fixed
antirez [Thu, 26 Aug 2010 17:10:40 +0000 (19:10 +0200)] 
intset loading bug fixed

14 years agotest for intset integer encodability test and some small refactoring
antirez [Thu, 26 Aug 2010 16:47:03 +0000 (18:47 +0200)] 
test for intset integer encodability test and some small refactoring

14 years agotranslated a few long logn into int64_t for correctness and to avoid compilation...
antirez [Thu, 26 Aug 2010 16:11:26 +0000 (18:11 +0200)] 
translated a few long logn into int64_t for correctness and to avoid compilation warnings as well

14 years agoredis-cli tests commented out since there is a problem with the CI server and this...
antirez [Thu, 26 Aug 2010 15:52:02 +0000 (17:52 +0200)] 
redis-cli tests commented out since there is a problem with the CI server and this tests, will be activated again once the problem is fixed

14 years agoSORT stress testing against bigger aggregate values
antirez [Thu, 26 Aug 2010 15:29:13 +0000 (17:29 +0200)] 
SORT stress testing against bigger aggregate values

14 years agoVersion is now 2.1.3
antirez [Thu, 26 Aug 2010 14:58:02 +0000 (16:58 +0200)] 
Version is now 2.1.3

14 years agoMerge remote branch 'pietern/redis-cli-fix'
antirez [Thu, 26 Aug 2010 14:19:06 +0000 (16:19 +0200)] 
Merge remote branch 'pietern/redis-cli-fix'

14 years agoforgot the traceleaks var in tests set to 1, reverted to the default, 0
antirez [Thu, 26 Aug 2010 11:18:56 +0000 (13:18 +0200)] 
forgot the traceleaks var in tests set to 1, reverted to the default, 0

14 years agoFix parenthesis error on decrementing *argc
Pieter Noordhuis [Thu, 26 Aug 2010 11:18:19 +0000 (13:18 +0200)] 
Fix parenthesis error on decrementing *argc

14 years agofixed a memory leak in the new Set code
antirez [Thu, 26 Aug 2010 11:18:24 +0000 (13:18 +0200)] 
fixed a memory leak in the new Set code

14 years agocomputeObjectSwappability is now able to compute swappability for intset encoded...
antirez [Thu, 26 Aug 2010 10:28:53 +0000 (12:28 +0200)] 
computeObjectSwappability is now able to compute swappability for intset encoded sets

14 years agoFix type that was not renamed and compiler warning
Pieter Noordhuis [Thu, 26 Aug 2010 10:13:51 +0000 (12:13 +0200)] 
Fix type that was not renamed and compiler warning

14 years agoremoved a duplicated ERRNO checking that is useless at all
antirez [Thu, 26 Aug 2010 10:10:16 +0000 (12:10 +0200)] 
removed a duplicated ERRNO checking that is useless at all

14 years agoMerge remote branch 'pietern/intset-split'
antirez [Thu, 26 Aug 2010 10:04:24 +0000 (12:04 +0200)] 
Merge remote branch 'pietern/intset-split'

14 years agoMake the function intsetUpgrade self-contained
Pieter Noordhuis [Thu, 26 Aug 2010 09:22:58 +0000 (11:22 +0200)] 
Make the function intsetUpgrade self-contained

14 years agoExpand macro's to functions for readability
Pieter Noordhuis [Thu, 26 Aug 2010 09:06:30 +0000 (11:06 +0200)] 
Expand macro's to functions for readability

14 years agoUse fstat to detect if stdin was redirected
Pieter Noordhuis [Wed, 25 Aug 2010 12:48:50 +0000 (14:48 +0200)] 
Use fstat to detect if stdin was redirected

14 years agoMake helper functions simpler
Pieter Noordhuis [Wed, 25 Aug 2010 12:15:41 +0000 (14:15 +0200)] 
Make helper functions simpler

14 years agoComments in redis-cli tests
Pieter Noordhuis [Wed, 25 Aug 2010 12:08:32 +0000 (14:08 +0200)] 
Comments in redis-cli tests

14 years agoAdd a newline to tty output after every reply
Pieter Noordhuis [Wed, 25 Aug 2010 11:39:11 +0000 (13:39 +0200)] 
Add a newline to tty output after every reply

14 years agoRe-introduce the interactive field so we can reconnect in interactive mode
Pieter Noordhuis [Wed, 25 Aug 2010 08:05:50 +0000 (10:05 +0200)] 
Re-introduce the interactive field so we can reconnect in interactive mode

14 years agoMerge master and move argument splitting patch to sds.c
Pieter Noordhuis [Wed, 25 Aug 2010 07:54:02 +0000 (09:54 +0200)] 
Merge master and move argument splitting patch to sds.c

14 years agoredis-cli now supports automatically reconnection in interactive mode
antirez [Tue, 24 Aug 2010 16:39:34 +0000 (18:39 +0200)] 
redis-cli now supports automatically reconnection in interactive mode

14 years agoIssue 179 fixed, now redis-cli is able to parse correctly multi bulk replies with...
antirez [Tue, 24 Aug 2010 16:08:09 +0000 (18:08 +0200)] 
Issue 179 fixed, now redis-cli is able to parse correctly multi bulk replies with elements that are errors

14 years agoThe pid file is now created only after the server is correctly initialied. It is...
antirez [Tue, 24 Aug 2010 15:09:25 +0000 (17:09 +0200)] 
The pid file is now created only after the server is correctly initialied. It is also removed on sigterm and when the stack trace is produced after a sigbus or a sigsegv. This two changes should fix the Issue 175

14 years agoslave now detect lost connection during SYNC, fixing Issue 173
antirez [Tue, 24 Aug 2010 14:25:00 +0000 (16:25 +0200)] 
slave now detect lost connection during SYNC, fixing Issue 173

14 years agoslave with attached slaves now close the conection to all the slaves when the connect...
antirez [Tue, 24 Aug 2010 14:04:13 +0000 (16:04 +0200)] 
slave with attached slaves now close the conection to all the slaves when the connection to the master is lost. Now a slave without a connected link to the master will refuse SYNC from other slaves. Enhanced the replication error reporting. All this will fix Issue 156

14 years agoBLPOPping clients are no longer subject to connection timeouts, fixing issues 155
antirez [Tue, 24 Aug 2010 10:10:59 +0000 (12:10 +0200)] 
BLPOPping clients are no longer subject to connection timeouts, fixing issues 155

14 years agoadded tests for invalid bulk argument
antirez [Tue, 24 Aug 2010 09:49:05 +0000 (11:49 +0200)] 
added tests for invalid bulk argument

14 years agosanity check for the bulk argument in protocol parsing code, fixing issue 146
antirez [Tue, 24 Aug 2010 09:45:05 +0000 (11:45 +0200)] 
sanity check for the bulk argument in protocol parsing code, fixing issue 146

14 years agochanged the comments on top of redis-copy.rb to reflect what the program really does
antirez [Tue, 24 Aug 2010 08:10:01 +0000 (10:10 +0200)] 
changed the comments on top of redis-copy.rb to reflect what the program really does

14 years agoFixed overflow detection in argument to long convertion function in general, and...
antirez [Mon, 23 Aug 2010 15:06:38 +0000 (17:06 +0200)] 
Fixed overflow detection in argument to long convertion function in general, and in expire/ttl pairs specifically, addressing issue 54

14 years agoRename iterator to setTypeIterator for consistency
Pieter Noordhuis [Sat, 21 Aug 2010 09:25:13 +0000 (11:25 +0200)] 
Rename iterator to setTypeIterator for consistency

14 years agoMake SORT use the hybrid set accessors to allow sorting intsets
Pieter Noordhuis [Sat, 21 Aug 2010 09:15:31 +0000 (11:15 +0200)] 
Make SORT use the hybrid set accessors to allow sorting intsets

14 years agoSORT tests with hash table encoded set as input
Pieter Noordhuis [Sat, 21 Aug 2010 09:03:56 +0000 (11:03 +0200)] 
SORT tests with hash table encoded set as input

14 years agoMake SORT tests use both ziplists and linked lists as input
Pieter Noordhuis [Sat, 21 Aug 2010 09:02:22 +0000 (11:02 +0200)] 
Make SORT tests use both ziplists and linked lists as input

14 years agoMove SORT tests around
Pieter Noordhuis [Sat, 21 Aug 2010 08:55:53 +0000 (10:55 +0200)] 
Move SORT tests around

14 years agoDon't abort test suite when the server block has a return value
Pieter Noordhuis [Sat, 21 Aug 2010 08:54:31 +0000 (10:54 +0200)] 
Don't abort test suite when the server block has a return value

14 years agoCheck if stroll return value was clamped
Pieter Noordhuis [Fri, 20 Aug 2010 11:42:42 +0000 (13:42 +0200)] 
Check if stroll return value was clamped

14 years agoMerge branch 'master' into intset-split
Pieter Noordhuis [Fri, 20 Aug 2010 10:40:29 +0000 (12:40 +0200)] 
Merge branch 'master' into intset-split

Conflicts:
src/Makefile
src/t_set.c

14 years agoredis cli argument splitting is general and is now moved into the sds.c lib
antirez [Thu, 5 Aug 2010 09:36:39 +0000 (11:36 +0200)] 
redis cli argument splitting is general and is now moved into the sds.c lib

14 years agoDeprecate starting interactive mode using the -i flag
Pieter Noordhuis [Wed, 4 Aug 2010 16:36:03 +0000 (18:36 +0200)] 
Deprecate starting interactive mode using the -i flag

14 years agoSet tty before going into interactive mode to get non-pretty output when
Pieter Noordhuis [Wed, 4 Aug 2010 16:16:39 +0000 (18:16 +0200)] 
Set tty before going into interactive mode to get non-pretty output when
the commands are read from stdin.

14 years agoChange output format for non-tty redis-cli execution
Pieter Noordhuis [Wed, 4 Aug 2010 15:46:56 +0000 (17:46 +0200)] 
Change output format for non-tty redis-cli execution

14 years agoLet the output mode depend on having a tty or not
Pieter Noordhuis [Wed, 4 Aug 2010 15:16:05 +0000 (17:16 +0200)] 
Let the output mode depend on having a tty or not

14 years agoTests for redis-cli in non-interactive mode
Pieter Noordhuis [Wed, 4 Aug 2010 15:02:13 +0000 (17:02 +0200)] 
Tests for redis-cli in non-interactive mode

Minor change in redis-cli output for the (multi-)bulk response but this
will be fixed in the next commit.

14 years agoAdd tests for quotation in an interactive redis-cli session
Pieter Noordhuis [Wed, 4 Aug 2010 13:29:28 +0000 (15:29 +0200)] 
Add tests for quotation in an interactive redis-cli session

Patched redis-cli to abort on unexpected quotation. This caused
redis-cli to get into an infinite, memory-consuming loop.

14 years agoTests for the interactive mode of redis-cli
Pieter Noordhuis [Wed, 4 Aug 2010 12:15:52 +0000 (14:15 +0200)] 
Tests for the interactive mode of redis-cli

Changed redis-cli to output the raw response for a bulk reply when it is
run in interactive mode instead of checking isatty.

14 years agoPERSIST: a fix and some basic test
antirez [Tue, 3 Aug 2010 12:25:22 +0000 (14:25 +0200)] 
PERSIST: a fix and some basic test

14 years agoPERSIST command implemented
antirez [Tue, 3 Aug 2010 12:19:20 +0000 (14:19 +0200)] 
PERSIST command implemented

14 years agoreplication test with expires
antirez [Tue, 3 Aug 2010 11:38:39 +0000 (13:38 +0200)] 
replication test with expires

14 years agoMerge remote branch 'pietern/test-encoding-fix' into writeonexpire
antirez [Tue, 3 Aug 2010 11:09:27 +0000 (13:09 +0200)] 
Merge remote branch 'pietern/test-encoding-fix' into writeonexpire

14 years agono longer passing tests due to the new write-on-volatile semantics modified/removed
antirez [Tue, 3 Aug 2010 11:08:32 +0000 (13:08 +0200)] 
no longer passing tests due to the new write-on-volatile semantics modified/removed

14 years agoallow to set a new EXPIRE of an existing volatile key
antirez [Tue, 3 Aug 2010 10:26:30 +0000 (12:26 +0200)] 
allow to set a new EXPIRE of an existing volatile key

14 years agomemory leak removed from expire propagation code
antirez [Mon, 2 Aug 2010 19:37:39 +0000 (21:37 +0200)] 
memory leak removed from expire propagation code

14 years agosupport for write operations against expiring keys, by master-controlled expiring...
antirez [Mon, 2 Aug 2010 16:13:39 +0000 (18:13 +0200)] 
support for write operations against expiring keys, by master-controlled expiring in replication and AOF synthesizing DEL operations

14 years agoFix assertion function on value encoding
Pieter Noordhuis [Sun, 1 Aug 2010 09:20:26 +0000 (11:20 +0200)] 
Fix assertion function on value encoding

14 years agoMerge remote branch 'pietern/list-test'
antirez [Sat, 31 Jul 2010 11:55:06 +0000 (13:55 +0200)] 
Merge remote branch 'pietern/list-test'

14 years agoChange getDoubleFromObject to fail on NaN.
Pieter Noordhuis [Thu, 29 Jul 2010 20:13:31 +0000 (22:13 +0200)] 
Change getDoubleFromObject to fail on NaN.

Return an error when the resulting value is not a number (NaN). Fix
ZUNIONSTORE/ZINTERSTORE to clean up when a weight argument is not a
double value.

14 years agoFix ZUNIONSTORE/ZINTERSTORE to never store a NaN score.
Pieter Noordhuis [Thu, 29 Jul 2010 19:31:58 +0000 (21:31 +0200)] 
Fix ZUNIONSTORE/ZINTERSTORE to never store a NaN score.

When +inf and -inf are added, the result is NaN. We don't want NaN
scores in a sorted set, so agreed on the result of this operation being
zero.

14 years agoUse a large value to consistently trigger a list encoding,
Pieter Noordhuis [Thu, 29 Jul 2010 11:53:52 +0000 (13:53 +0200)] 
Use a large value to consistently trigger a list encoding,
even when the list is swapped out and in again.

14 years agoensure the value is swapped in before testing its encoding
Pieter Noordhuis [Thu, 29 Jul 2010 11:31:24 +0000 (13:31 +0200)] 
ensure the value is swapped in before testing its encoding

14 years agofix of the fix for the replication bug
antirez [Wed, 28 Jul 2010 16:56:52 +0000 (18:56 +0200)] 
fix of the fix for the replication bug

14 years agoFixed a replication bug in ZINTERSTORE.
antirez [Wed, 28 Jul 2010 16:42:02 +0000 (18:42 +0200)] 
Fixed a replication bug in ZINTERSTORE.

In order to trigger the bug what's needed is to call ZINTERSTORE
resulting into an empty set created, bug against a key that already
existed. The command was not propagated, so the replica ended with the
key that the master removed. Sequence of command to reproduce:

redis-cli hset 446 34 905
redis-cli hset 446 393 911
redis-cli zadd 966 0.085412045980529885 652
redis-cli zadd 645 0.25081839284432045 280
redis-cli zinterstore 446 2 966 645

14 years agobetter random dataset creation function in test. master-slave replication test now...
antirez [Wed, 28 Jul 2010 12:08:46 +0000 (14:08 +0200)] 
better random dataset creation function in test. master-slave replication test now is able to save the two datasets in CSV when an inconsistency is detected.

14 years agofixed a ziplist bug about encoding of integer values overflowing 64 bit
antirez [Tue, 27 Jul 2010 13:26:08 +0000 (15:26 +0200)] 
fixed a ziplist bug about encoding of integer values overflowing 64 bit

14 years agoremoved test code having bad effects...
antirez [Tue, 27 Jul 2010 12:46:39 +0000 (14:46 +0200)] 
removed test code having bad effects...

14 years agoConsistency test improved
antirez [Tue, 27 Jul 2010 12:42:11 +0000 (14:42 +0200)] 
Consistency test improved

14 years agomalloc definition with deprecated attribute was duplicated, one removed
antirez [Tue, 27 Jul 2010 12:30:02 +0000 (14:30 +0200)] 
malloc definition with deprecated attribute was duplicated, one removed

14 years agoSTRLEN command implemented
antirez [Tue, 27 Jul 2010 08:09:26 +0000 (10:09 +0200)] 
STRLEN command implemented

14 years agohash table example commented out in dict.c
antirez [Tue, 27 Jul 2010 08:00:38 +0000 (10:00 +0200)] 
hash table example commented out in dict.c

14 years agouse the function deprecated attribute if compiling with GCC to get warnings for mallo...
antirez [Tue, 27 Jul 2010 07:36:42 +0000 (09:36 +0200)] 
use the function deprecated attribute if compiling with GCC to get warnings for malloc/free usages. We always want to use our zmalloc/zfree versions for memory usage tracking

14 years agoAdd zcalloc and use it where appropriate
Benjamin Kramer [Sat, 24 Jul 2010 21:20:00 +0000 (23:20 +0200)] 
Add zcalloc and use it where appropriate

calloc is more effecient than malloc+memset when the system uses mmap to
allocate memory. mmap always returns zeroed memory so the memset can be
avoided.  The threshold to use mmap is 16k in osx libc and 128k in bsd
libc and glibc. The kernel can lazily allocate the pages, this reduces
memory usage when we have a page table or hash table that is mostly
empty.

This change is most visible when you start a new redis instance with vm
enabled.  You'll see no increased memory usage no matter how big your
page table is.

14 years agoRemove _dictAlloc and friends
Benjamin Kramer [Sat, 24 Jul 2010 21:10:42 +0000 (23:10 +0200)] 
Remove _dictAlloc and friends

zmalloc calls abort() so _dictPanic will never be called.

14 years agoReduce code duplication
Benjamin Kramer [Sat, 24 Jul 2010 20:37:01 +0000 (22:37 +0200)] 
Reduce code duplication

14 years agoexit with non-zero status when there are failed tests
Pieter Noordhuis [Fri, 23 Jul 2010 11:08:35 +0000 (13:08 +0200)] 
exit with non-zero status when there are failed tests

14 years agodon't open/close log file if log level is not matched
antirez [Thu, 22 Jul 2010 21:31:40 +0000 (23:31 +0200)] 
don't open/close log file if log level is not matched

14 years agofix rare condition where 'key' would already be destroyed while is was needed later on
Pieter Noordhuis [Thu, 22 Jul 2010 14:06:27 +0000 (16:06 +0200)] 
fix rare condition where 'key' would already be destroyed while is was needed later on

14 years agodefensive programming: set o->ptr to NULL before freeing objects
antirez [Thu, 22 Jul 2010 13:48:57 +0000 (15:48 +0200)] 
defensive programming: set o->ptr to NULL before freeing objects

14 years agoother shared objects where created in the I/O thread in createStringObjectFromLongLon...
antirez [Thu, 22 Jul 2010 12:48:45 +0000 (14:48 +0200)] 
other shared objects where created in the I/O thread in createStringObjectFromLongLong. Fixed as well.

14 years agominor typo fixed in a comment
antirez [Thu, 22 Jul 2010 11:12:24 +0000 (13:12 +0200)] 
minor typo fixed in a comment

14 years agodon't use object sharing inside I/O threads, as a fix for a well known instability...
antirez [Thu, 22 Jul 2010 11:08:02 +0000 (13:08 +0200)] 
don't use object sharing inside I/O threads, as a fix for a well known instability of VM introduced with the new object sharing code

14 years agovm_blocked_clients count fixed in INFO, thanks to Pietern Noordhuis
antirez [Wed, 21 Jul 2010 11:16:26 +0000 (13:16 +0200)] 
vm_blocked_clients count fixed in INFO, thanks to Pietern Noordhuis