search for: hybrid_ls_rr_sort

On Tue, 2011-10-25 at 21:00 -0700, Andrew Trick wrote: Now, to generate the best PPC schedules, there is one thing you may > want to override. The scheduler's priority function has a > HasReadyFilter attribute (enum). It can be overriden by specializing > hybrid_ls_rr_sort. Setting this to "true" enables proper ILP > scheduling, and maximizes the instructions that can issue in one > group, regardless of register pressure. We still care about register > pressure enough in ARM to avoid enabling this. I'm really not sure how > much it will help...

...wrote: > On Tue, 2011-10-25 at 21:00 -0700, Andrew Trick wrote: > Now, to generate the best PPC schedules, there is one thing you may >> want to override. The scheduler's priority function has a >> HasReadyFilter attribute (enum). It can be overriden by specializing >> hybrid_ls_rr_sort. Setting this to "true" enables proper ILP >> scheduling, and maximizes the instructions that can issue in one >> group, regardless of register pressure. We still care about register >> pressure enough in ARM to avoid enabling this. I'm really not sure how >> m...

..., 2011-10-25 at 21:00 -0700, Andrew Trick wrote: > > Now, to generate the best PPC schedules, there is one thing you may > >> want to override. The scheduler's priority function has a > >> HasReadyFilter attribute (enum). It can be overriden by specializing > >> hybrid_ls_rr_sort. Setting this to "true" enables proper ILP > >> scheduling, and maximizes the instructions that can issue in one > >> group, regardless of register pressure. We still care about register > >> pressure enough in ARM to avoid enabling this. I'm really not sure...

...at 21:00 -0700, Andrew Trick wrote: >>> Now, to generate the best PPC schedules, there is one thing you may >>>> want to override. The scheduler's priority function has a >>>> HasReadyFilter attribute (enum). It can be overriden by specializing >>>> hybrid_ls_rr_sort. Setting this to "true" enables proper ILP >>>> scheduling, and maximizes the instructions that can issue in one >>>> group, regardless of register pressure. We still care about register >>>> pressure enough in ARM to avoid enabling this. I'm really...

...rew Trick wrote: > >>> Now, to generate the best PPC schedules, there is one thing you may > >>>> want to override. The scheduler's priority function has a > >>>> HasReadyFilter attribute (enum). It can be overriden by specializing > >>>> hybrid_ls_rr_sort. Setting this to "true" enables proper ILP > >>>> scheduling, and maximizes the instructions that can issue in one > >>>> group, regardless of register pressure. We still care about register > >>>> pressure enough in ARM to avoid enabling this....

On Dec 19, 2011, at 10:53 PM, Hal Finkel wrote: > Here's my "thought experiment" (from PR11589): I have a bunch of > load-fadd-store chains to schedule. A store takes two cycles to clear > its last pipeline stage. The fadd takes longer to compute its result > (say 5 cycles), but can sustain a rate of 1 independent add per cycle. > As the scheduling is bottom-up, it

...-RA-sched > > and debug your itinerary. > > Andy, I've already looked at the debug output quite a bit; please help > me understand what I'm missing... > > First, looking at the code does seem to confirm my suspicion. This is > certainly is low-pressure mode, and so hybrid_ls_rr_sort::operator() > will return the result of BUCompareLatency. That function first checks > for stalls and returns 1 or -1. Only after that does it look at the > relative latencies. Looking at this more carefully, I think that I see the problem. The heights are set to account for the latencies...

...d for bottom-up, or you could start by simply disabling it instead of the entire hazard recognizer. Now, to generate the best PPC schedules, there is one thing you may want to override. The scheduler's priority function has a HasReadyFilter attribute (enum). It can be overriden by specializing hybrid_ls_rr_sort. Setting this to "true" enables proper ILP scheduling, and maximizes the instructions that can issue in one group, regardless of register pressure. We still care about register pressure enough in ARM to avoid enabling this. I'm really not sure how much it will help on modern PPC imple...

Is there documentation somewhere for the bottom-up scheduling? I'm trying to figure out what changes are necessary in order to support it in the PPC backend. Thanks in advance, Hal On Thu, 2011-10-20 at 10:21 -0700, Evan Cheng wrote: > > On Oct 19, 2011, at 7:29 PM, Hal Finkel <hfinkel at anl.gov> wrote: > > > Evan, > > > > Thanks for the heads up! Is

...debug your itinerary. >> >> Andy, I've already looked at the debug output quite a bit; please help >> me understand what I'm missing... >> >> First, looking at the code does seem to confirm my suspicion. This is >> certainly is low-pressure mode, and so hybrid_ls_rr_sort::operator() >> will return the result of BUCompareLatency. That function first checks >> for stalls and returns 1 or -1. Only after that does it look at the >> relative latencies. > > Looking at this more carefully, I think that I see the problem. The > heights are set t...

...may need to look at -debug-only=pre-RA-sched > and debug your itinerary. Andy, I've already looked at the debug output quite a bit; please help me understand what I'm missing... First, looking at the code does seem to confirm my suspicion. This is certainly is low-pressure mode, and so hybrid_ls_rr_sort::operator() will return the result of BUCompareLatency. That function first checks for stalls and returns 1 or -1. Only after that does it look at the relative latencies. In addition, the stall computation is done using BUHasStall, and that function only checks the current cycle. Without looking f...

...called. Could you please explain the calling sequence? Thanks again, Hal > > Now, to generate the best PPC schedules, there is one thing you may > want to override. The scheduler's priority function has a > HasReadyFilter attribute (enum). It can be overriden by specializing > hybrid_ls_rr_sort. Setting this to "true" enables proper ILP > scheduling, and maximizes the instructions that can issue in one > group, regardless of register pressure. We still care about register > pressure enough in ARM to avoid enabling this. I'm really not sure how > much it will help...

...ain, Hal > > Thanks again, > Hal > > > > > Now, to generate the best PPC schedules, there is one thing you may > > want to override. The scheduler's priority function has a > > HasReadyFilter attribute (enum). It can be overriden by specializing > > hybrid_ls_rr_sort. Setting this to "true" enables proper ILP > > scheduling, and maximizes the instructions that can issue in one > > group, regardless of register pressure. We still care about register > > pressure enough in ARM to avoid enabling this. I'm really not sure how > &g...