On 8/23/23 18:42, Vineet Gupta wrote:
Seriously, I detest it too, but the irony is I've now made my 2nd change
in there and keep adding to ugliness :-(
Happens to all of us sometimes.
So I think your change makes sense. But I think it can be refined to
simplify the larger chunk of
On 8/23/23 13:04, Jeff Law wrote:
Thanks for your patience on this. I needed a bit of time to gather my
thoughts and review some code.
No worries at all.
index 8b7256108157..1802eef908fc 100644
--- a/gcc/config/riscv/riscv.cc
+++ b/gcc/config/riscv/riscv.cc
@@ -2464,14 +2464,9 @@ riscv_rtx
On 8/9/23 18:30, Vineet Gupta wrote:
Hoist want_to_gcse_p () calls rtx_cost () to compute max distance for
hoist candidates . For a const with cost 1 backend currently returns 0,
causing Hoist to bail and elide GCSE.
Note that constants requiring more than 1 insns to setup were working
alread
Hoist want_to_gcse_p () calls rtx_cost () to compute max distance for
hoist candidates . For a const with cost 1 backend currently returns 0,
causing Hoist to bail and elide GCSE.
Note that constants requiring more than 1 insns to setup were working
already since backend is returning 1 as well. Ar
