* Synchronize advances the write sequence and returns when all * readers have observed it. * * If your application can cache a sequence number returned from * smr_advance() and poll or wait at a later time there will * be less chance of busy looping while waiting for readers. */
| 249 | * be less chance of busy looping while waiting for readers. |
| 250 | */ |
| 251 | static inline void |
| 252 | smr_synchronize(smr_t smr) |
| 253 | { |
| 254 | |
| 255 | smr_wait(smr, smr_advance(smr)); |
| 256 | } |
| 257 | |
| 258 | /* Only at startup. */ |
| 259 | void smr_init(void); |
no test coverage detected