Commit Graph

2436 Commits

Author SHA1 Message Date
03a5c0ecb5 Correct cleanup when aborting AOF loading.
Because of the new ability to start with a truncated AOF, we need
to correctly release all the memory on EOF error. Otherwise there is a
small leak, that is not really a problem, but causes a false positive in
the tests that detect memory leaks.
2014-09-08 10:58:00 +02:00
683cf5b872 AOF ability to load truncated files. 2014-09-08 10:57:57 +02:00
537f3053fa AOF loading: split handling of format errors from unexpected EOF. 2014-09-08 10:57:10 +02:00
667e2991dc Return empty string if GETRANGE of empty string
Previously, GETRANGE of a key containing nothing ("")
would allocate a large (size_t)-1 return value causing
crashes on 32bit builds when it tried to allocate the
4 GB return string.
2014-09-04 11:47:21 +02:00
4284b07220 Increase size of range request in getrange
32 bit builds don't have a big enough long to capture
the same range as a 64 bit build.  If we use "long long"
we get proper size limits everywhere.

Also updates size of unsigned comparison to fit new size of `end`.

Fixes #1981
2014-09-04 11:46:57 +02:00
db6e874a25 Redis 2.8.14. 2014-09-01 17:00:38 +02:00
dd36f18c76 Sentinel: Abort Hello quicker if not connected
We can save a little work by aborting when we enter the function
if we're disconnected.
2014-09-01 16:35:04 +02:00
59ad68ca4f Store the length of the static argv when first allocated. 2014-09-01 10:40:04 +02:00
8fb8a474fd Fix dictRehash assert casting type.
Also related to #1929.
2014-08-27 10:31:52 +02:00
85b3db4065 Cast to right type in dictNext().
This closes issue #1929, the other part was fixed in the context of issue
2014-08-27 10:31:48 +02:00
3a706ba792 Use long for rehash and iterator index in dict.h.
This allows to support datasets with more than 2 billion of keys
(possible in very large memory instances, this bug was actually
reported).

Closes issue #1814.
2014-08-27 10:31:44 +02:00
c3eac3ddff No longer useful assert removed from sdsIncrLen(). 2014-08-27 10:31:32 +02:00
fdcadd16c6 Fix redis-benchmark abort output 2014-08-27 10:31:29 +02:00
ac48aada30 Deny CLIENT command in scripts
We don't want scripts doing CLIENT SETNAME
or CLIENT KILL or CLIENT LIST or CLIENT PAUSE.

Originally reported by Chris Wj then proper
action inspired by Itamar Haber.

Reference: https://groups.google.com/forum/#!topic/redis-db/09B2EYwyVgk
2014-08-27 10:31:20 +02:00
abe29c83e6 redis-benchmark: Remove unused buffer
This just deletes old code that didn't get removed when
logic changed.  We were setting offsets that never
got read anywhere.

Since clients are now just cloned, we don't need to track
per-client buffer offsets anywhere because they are all
the same from the original client.
2014-08-27 10:30:49 +02:00
8fcce90b25 redis-benchmark: add auth option
Closes #1097
2014-08-27 10:30:49 +02:00
4ac8472cac Rename two 'buf' vars to 'ip' for better clarity
Clearly ip[32] is wrong, but it's less clear that buf[32] was wrong
without further reading.
2014-08-27 10:30:42 +02:00
f7b5e2d11e Sentinel: fix bufsize to support IPv6 address
Closes #1914
2014-08-27 10:30:37 +02:00
32168a5e8a Sentinel: Reject config from STDIN
Sentinel needs to die with a more accurate error message
when attempted to open a config from STDIN.

See: https://groups.google.com/forum/#!topic/redis-db/sYx7VNMWaNM
2014-08-27 10:30:26 +02:00
4bf7c5d0f8 Remove unused global variable
It has an important name, but nothing uses it...
2014-08-27 10:30:22 +02:00
33b37b1400 Fix invalid expire error for SET family commands. 2014-08-27 10:30:19 +02:00
ae2cd2e16f Remove unused function
Closes #878
2014-08-27 10:30:11 +02:00
8c0743d80e Extend range of bytesToHuman to TB and PB
Also adds a fallthrough case for when given
large values (like overflow numbers of 2^64 by mistake).

Closes #858
2014-08-27 10:30:08 +02:00
ef12b4c3eb Add error check for writing RDB checksum
Closes #857
2014-08-27 10:30:04 +02:00
1580a378b8 redis-cli: fix prompt after shutdown command
Fix redis-cli prompt to state "not connected" after a SHUTDOWN command
is sent.
2014-08-27 10:30:01 +02:00
467848460a Clarify argument to dict macro
d is more clear because the type of argument is dict not dictht

Closes #513
2014-08-27 10:29:57 +02:00
5e707c66dd Fix undefined behavior in ll2string().
The bug was found by @CAFxX, thanks!
See issue #1940.
2014-08-27 10:29:53 +02:00
65d47452f8 Remove warnings and improve integer sign correctness. 2014-08-27 10:29:26 +02:00
09757420a4 Add -W to compilation flags. 2014-08-27 10:25:57 +02:00
134345e332 Use unsigned integers in SDS header.
This raises the max string to 4GB without any downside.
2014-08-27 10:25:54 +02:00
f89a7bbe11 va_copy must be matched by va_end
Hat tip to @rfuchs. See: https://github.com/redis/hiredis/pull/178.

Fixes #1187
2014-08-27 10:25:44 +02:00
dbcf381099 Remove duplicate prototypes in redis.h
Also moves acceptHandler() to be near the other accept...() functions.

Closes #1105
2014-08-27 10:25:40 +02:00
6bc74b4393 Use resolv library in Solaris
For some Solaris flavours, the inet_aton in in resolv library.
Not linking this library will introduce link error.

Improves compatability with older Solaris and still
works on new Solaris.

Closes #1092
2014-08-27 10:25:37 +02:00
866f6bcc80 Cleanup double semicolons
Closes #1161
2014-08-27 10:25:31 +02:00
29ed617ac4 redis-check-dump: Prevent segfault if can't malloc
Found by The Mayhem Team (Alexandre Rebert, Thanassis Avgerinos,
Sang Kil Cha, David Brumley, Manuel Egele) Cylab, Carnegie Mellon
University. See http://bugs.debian.org/716259 for more.

Signed-off-by: Chris Lamb <lamby@debian.org>

Fixes #1191
2014-08-27 10:25:28 +02:00
a43e45e391 Remove unused LINE_BUFLEN definition
Closes #1129
2014-08-27 10:25:25 +02:00
25a4699a9e Reject MOVE to non-integer DBs
Previously, "MOVE key somestring" would move the key to
DB 0 which is just unexpected and wrong.
String as DB == error.

Test added too.

Modified by @antirez in order to use the getLongLongFromObject() API
instead of strtol().

Fixes #1428
2014-08-27 10:25:22 +02:00
09a3976478 pubsub: Return integers for NUMSUB, not strings
Also adds test for numsub — due to tcl being tcl,
it doesn't capture the "numberness" of the fix,
but now we at least have one test case for numsub.

Closes #1561
2014-08-27 10:25:15 +02:00
60c748628b Avoid unnecessary decoding in ziplist.c
Closes #1519
2014-08-27 10:25:12 +02:00
96614f5866 Improve accuracy of HAVE_ATOMIC dependency check
[I had to split out the clang check due to
 clang *really* not liking the __GLIBC_PREREQ macro; -matt]

Closes #1456
2014-08-27 10:25:07 +02:00
d1df50ad59 Remove redundant event loop fd processing
Reason:
https://github.com/antirez/redis/pull/1450#issuecomment-30432925

Closes #1450
2014-08-27 10:25:04 +02:00
282c6106ce memtest: Add missing free()
Closes #1614
2014-08-27 10:25:00 +02:00
25b5abe8cf src/Makefile: Reword "to run make test" message
Closes #1610
2014-08-27 10:24:56 +02:00
a7c46df612 Use 'void' for zero-argument functions
According to the C standard,
it is desirable to give the type 'void'
to functions have no argument.

Closes #1631
2014-08-27 10:24:52 +02:00
8dbd0134e5 Fix assert technical correctness
dictAdd returns DICT_OK, not REDIS_OK. They both
have the same underlying values, so it works even though
the code is technically wrong.

Fixes #1512
2014-08-27 10:24:46 +02:00
4be50f7fbb redis-check-dump: use names instead of magic nums
Use constants to avoid magic numbers in `types`, which is an array
that stores the names of `REDIS` types.

Closes #1681
2014-08-27 10:24:36 +02:00
8fc2247c9b Fix issues raised by clang analyzer
Modified by @antirez since the original fix to genInfoString() looked
weak. Probably the clang analyzer complained about `section` being
possibly NULL, and strcasecmp() called with a NULL pointer. In the
practice this can never happen, still for the sake of correctness
the right fix is not to modify only the first call, but to set `section`
to the value of "default" if it happens to be NULL.

Closes #1660
2014-08-27 10:24:33 +02:00
ebd25710dd Force quit when receiving a second SIGINT.
Also quit ASAP when we are still loading a DB, since care is not needed
in this special condition, especially for a SIGINT.
2014-08-27 10:24:29 +02:00
524b2626f7 Add graceful exit when Ctrl-C is received 2014-08-27 10:24:26 +02:00
78bde472c0 Fix intset midpoint selection
The classic (min+max)/2 is provably unsafe.  Fixed
as recommended in research:
http://googleresearch.blogspot.com/2006/06/extra-extra-read-all-about-it-nearly.html

Fix inspired by @wjin, but I used a different approach.
(later, I found @kuebler fixed the same issue too).

Fixes #1741, #1602
2014-08-27 10:24:22 +02:00