[PATCH] D63628: AMD K10 (Barcelona) Initial Scheduler model
Roman Lebedev via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Fri Jun 21 01:29:55 PDT 2019
lebedev.ri marked 2 inline comments as done.
lebedev.ri added inline comments.
================
Comment at: include/llvm/TableGen/Record.h:1271
+ << " with non-record type!\n";
+ assert(false && "FieldInit");
+ }
----------------
courbet wrote:
> `llvm_unreachable` ?
I'll split this int a separate patch
================
Comment at: lib/Target/X86/X86ScheduleBarcelona.td:17
+// * https://www.realworldtech.com/barcelona/
+//
+//===----------------------------------------------------------------------===//
----------------
courbet wrote:
> `* Measurements from llvm-exegesis` ?
Yeah, i should add that :)
================
Comment at: lib/Target/X86/X86ScheduleBarcelona.td:760
+////////////////////////////////////////////////////////////////////////////////
+
+} // SchedModel
----------------
courbet wrote:
> Does K10 have any zeroing/dependency breaking idioms ? Maybe add a FIXME to implement them.
Well, it's a good question.
Agner says it does, but unless i'm doing something horribly wrong even the trivial `xor %eax, %eax` doesn't work.
So i've left it for later.
Repository:
rL LLVM
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D63628/new/
https://reviews.llvm.org/D63628
More information about the llvm-commits
mailing list