----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: http://reviews.gem5.org/r/3256/#review7744 -----------------------------------------------------------
I'm confused... if the request says it needs a writable copy, it will always get a writable copy (invalidating other copies as necessary), so I don't see how we'd end up with a packet that has the until-recently-known-as-shared flag set. - Steve Reinhardt On Dec. 22, 2015, 8:28 a.m., Andreas Hansson wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > http://reviews.gem5.org/r/3256/ > ----------------------------------------------------------- > > (Updated Dec. 22, 2015, 8:28 a.m.) > > > Review request for Default. > > > Repository: gem5 > > > Description > ------- > > Changeset 11282:bf2ca8cda251 > --------------------------- > mem: Only forward the non-writable flag if truly needed > > This patch ensures a cache does not forward the non-writable flag from > a CPU-side request unless it has to. Thus, we avoid being overly > conservative and returning non-writable responses even when we know > that the response is the only writable copy in the system (the request > was sent out with NeedsWritable set). > > > Diffs > ----- > > src/mem/cache/cache.cc d9a0136ab8cc > > Diff: http://reviews.gem5.org/r/3256/diff/ > > > Testing > ------- > > > Thanks, > > Andreas Hansson > > _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
