- Notifications
You must be signed in to change notification settings - Fork5
Commitdad1f46
committed
TransferPredicateLocksToNewTarget should initialize a new lock
entry's commitSeqNo to that of the old one being transferred, or takethe minimum commitSeqNo if it is merging two lock entries.Also, CreatePredicateLock should initialize commitSeqNo for toInvalidSerCommitSeqNo instead of to 0. (I don't think using 0 wouldactually affect anything, but we should be consistent.)I also added a couple of assertions I used to track this down: alock's commitSeqNo should never be zero, and it should beInvalidSerCommitSeqNo if and only if the lock is not held byOldCommittedSxact.Dan Ports, to fix leak of predicate locks reported by YAMAMOTO Takashi.1 parent7c797e7 commitdad1f46
1 file changed
+16
-2
lines changedLines changed: 16 additions & 2 deletions
Original file line number | Diff line number | Diff line change | |
---|---|---|---|
| |||
2074 | 2074 |
| |
2075 | 2075 |
| |
2076 | 2076 |
| |
2077 |
| - | |
| 2077 | + | |
2078 | 2078 |
| |
2079 | 2079 |
| |
2080 | 2080 |
| |
| |||
2508 | 2508 |
| |
2509 | 2509 |
| |
2510 | 2510 |
| |
| 2511 | + | |
2511 | 2512 |
| |
2512 | 2513 |
| |
2513 | 2514 |
| |
| |||
2552 | 2553 |
| |
2553 | 2554 |
| |
2554 | 2555 |
| |
2555 |
| - | |
| 2556 | + | |
2556 | 2557 |
| |
| 2558 | + | |
| 2559 | + | |
| 2560 | + | |
| 2561 | + | |
| 2562 | + | |
| 2563 | + | |
| 2564 | + | |
| 2565 | + | |
| 2566 | + | |
2557 | 2567 |
| |
2558 | 2568 |
| |
2559 | 2569 |
| |
| |||
3137 | 3147 |
| |
3138 | 3148 |
| |
3139 | 3149 |
| |
| 3150 | + | |
| 3151 | + | |
3140 | 3152 |
| |
3141 | 3153 |
| |
3142 | 3154 |
| |
| |||
3261 | 3273 |
| |
3262 | 3274 |
| |
3263 | 3275 |
| |
| 3276 | + | |
| 3277 | + | |
3264 | 3278 |
| |
3265 | 3279 |
| |
3266 | 3280 |
| |
|
0 commit comments
Comments
(0)