summaryrefslogtreecommitdiff
Commit message (Expand)AuthorAgeFilesLines
* Move the write-generation out of the block header and into the pageKeith Bostic2012-11-2817-218/+230
* Re-seed the random number once (and only once), for each new test/formatKeith Bostic2012-11-271-14/+15
* Remove commented out code from test_checkpoint02Alex Gorrod2012-11-281-7/+0
* Incorporate reconciliation tracking items in page and cache memoryKeith Bostic2012-11-272-3/+20
* Add statistic to track bytes changed: for cursor.insert count key andKeith Bostic2012-11-275-52/+71
* And now, lint, which is why I started whacking on it.Keith Bostic2012-11-271-1/+1
* Typo in the last commit, size can't be modified in place.Keith Bostic2012-11-271-2/+3
* updateKeith Bostic2012-11-271-0/+1
* size_t/int, char/u_char lint.Keith Bostic2012-11-271-6/+7
* Arithmetic shift of a negative number is undefined by ISO/IEC 9899.Keith Bostic2012-11-271-7/+20
* Extent list block allocation was only half as big as it needs to be,Keith Bostic2012-11-271-1/+1
* Have test/format seed the random number generator once, using time(3) instead...Michael Cahill2012-11-271-5/+4
* Assert that the write buffer is aligned for direct I/O and large enoughKeith Bostic2012-11-261-2/+6
* Remove unused local variable.Alex Gorrod2012-11-271-1/+0
* Fix lint.Alex Gorrod2012-11-271-1/+3
* Michael says (and I agree): Compression shouldn't have to know about theKeith Bostic2012-11-263-96/+130
* Turn off prefix compression 80% of the time, otherwise raw compressionKeith Bostic2012-11-261-2/+2
* Silently ignore WT_SESSION::compress_raw for file types and configurationsKeith Bostic2012-11-264-36/+17
* lint, don't include <sys/time.h> explicitly, tested on FreeBSD,Keith Bostic2012-11-261-1/+0
* lint (don't include <sys/time.h> explicitly, it's a repeated includeKeith Bostic2012-11-261-1/+0
* Merge pull request #382 from wiredtiger/rawMichael Cahill2012-11-2552-1065/+2043
|\
| * Merge branch 'develop' into rawKeith Bostic2012-11-257-38/+34
| |\ | |/ |/|
* | We're using the WT_REC_SINGLE flag for two purposes ("called for eviction"Keith Bostic2012-11-257-38/+32
| * Make the compaction trigger a configuration string.Keith Bostic2012-11-257-19/+36
| * Clarify the huffman docs a bit, no real change.Keith Bostic2012-11-251-19/+20
| * Remove unnecessary test: addr can't be NULL at that point.Keith Bostic2012-11-241-2/+1
| * Tighten up the extent list offset checks: no extent block should appearKeith Bostic2012-11-244-16/+23
| * signed/unsigned lint.Keith Bostic2012-11-231-3/+3
| * signed/unsigned lint.Keith Bostic2012-11-231-1/+1
| * Maintain bytes read/written as well as I/O counts, reference #381.Keith Bostic2012-11-237-114/+148
| * Add a check for buffer alignment in the write path.Keith Bostic2012-11-232-3/+7
| * typo in a commentKeith Bostic2012-11-231-1/+1
| * Check the page header's flags field as part of verification.Keith Bostic2012-11-231-0/+11
| * Make WT_SESSION::create's checksum configuration 3-state: on, off, orKeith Bostic2012-11-2311-33/+79
| * If we turn off checksums because we're depending on decompression failures toKeith Bostic2012-11-231-1/+11
| * We need to checksum the data if it's not compressed, even if checksumsKeith Bostic2012-11-2314-50/+82
| * Move the checksum code into support/cksum, it's at least potentially generalKeith Bostic2012-11-233-2/+2
| * Merge branch 'develop' into rawKeith Bostic2012-11-237-29/+65
| |\ | |/ |/|
* | Update LSM worker to have cleaner error handling.Alex Gorrod2012-11-231-22/+19
* | Cut release 1.3.81.3.8Alex Gorrod2012-11-226-7/+44
| * Pack the extent list offset/size pairs for portability.Keith Bostic2012-11-221-15/+23
| * lint: src_len no longer used in decompress, we never return an error.Keith Bostic2012-11-221-1/+1
| * Merge branch 'develop' into rawKeith Bostic2012-11-226-10/+65
| |\ | |/ |/|
* | Update the documentation around statistics keys.Michael Cahill2012-11-224-2/+16
* | Don't try to merge with a chunk that is much larger than a small chunk.Michael Cahill2012-11-211-3/+12
* | After an LSM merge, fault in some pages before the new tree goes live to avoi...Michael Cahill2012-11-212-3/+33
* | whitespaceMichael Cahill2012-11-211-1/+1
| * Move compression out of the block-manager layer to the btree layer.Keith Bostic2012-11-2223-541/+545
| * __wt_err() returns a void, no reason to cast it.Keith Bostic2012-11-221-6/+4
| * Add support for a checksum configuration.Keith Bostic2012-11-223-0/+7