tag:blogger.com,1999:blog-1257971411439114772.post2206119448001527739..comments2023-04-10T00:50:19.770+10:00Comments on GHC on SPARC: Wait and performAnonymoushttp://www.blogger.com/profile/08287674468193351664noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-1257971411439114772.post-91915325138395804642009-01-23T02:31:00.000+11:002009-01-23T02:31:00.000+11:00Not having seen the intended usage of MO_writeBarr...Not having seen the intended usage of MO_writeBarrier, I'm going to nonetheless speculate that this will be a no-op under TSO (certainly true if this operation is used *before* a write to ensure consistency); if it is not, a membar #StoreLoad will be sufficient (to ensure that the write occurs before subsequent loads).J-W Maessenhttps://www.blogger.com/profile/01577860012705910561noreply@blogger.com