diff options
| author | Kyrylo Tkachov <kyrylo.tkachov@arm.com> | 2016-08-26 10:23:20 +0000 |
|---|---|---|
| committer | Kyrylo Tkachov <ktkachov@gcc.gnu.org> | 2016-08-26 10:23:20 +0000 |
| commit | 83c7402aa5d210a822e025f353c8f9bce027a4a7 (patch) | |
| tree | 8b417e4dff33baa354bd9e65b9e2d10d690b522a | |
| parent | 5791f55d02239dc7172884f9d5143632b8ca602d (diff) | |
| download | gcc-83c7402aa5d210a822e025f353c8f9bce027a4a7.zip gcc-83c7402aa5d210a822e025f353c8f9bce027a4a7.tar.gz gcc-83c7402aa5d210a822e025f353c8f9bce027a4a7.tar.bz2 | |
[ARM] PR target/70473: Reduce size of Cortex-A8 automaton
PR target/70473
* config/arm/cortex-a8-neon.md (cortex_a8_vfp_muld): Reduce
reservation duration to 15 cycles.
(cortex_a8_vfp_macs): Likewise.
(cortex_a8_vfp_macd): Likewise.
(cortex_a8_vfp_divs): Likewise.
(cortex_a8_vfp_divd): Likewise.
From-SVN: r239772
| -rw-r--r-- | gcc/ChangeLog | 10 | ||||
| -rw-r--r-- | gcc/config/arm/cortex-a8-neon.md | 14 |
2 files changed, 19 insertions, 5 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 9c82deb..dc3cb73 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -1,5 +1,15 @@ 2016-08-26 Kyrylo Tkachov <kyrylo.tkachov@arm.com> + PR target/70473 + * config/arm/cortex-a8-neon.md (cortex_a8_vfp_muld): Reduce + reservation duration to 15 cycles. + (cortex_a8_vfp_macs): Likewise. + (cortex_a8_vfp_macd): Likewise. + (cortex_a8_vfp_divs): Likewise. + (cortex_a8_vfp_divd): Likewise. + +2016-08-26 Kyrylo Tkachov <kyrylo.tkachov@arm.com> + * config/arm/arm.c (arm_sets_movw_movt_fusible_p): New function. (aarch_macro_fusion_pair_p): Use above to avoid early return. diff --git a/gcc/config/arm/cortex-a8-neon.md b/gcc/config/arm/cortex-a8-neon.md index 45f861f..b16c299 100644 --- a/gcc/config/arm/cortex-a8-neon.md +++ b/gcc/config/arm/cortex-a8-neon.md @@ -357,30 +357,34 @@ (eq_attr "type" "fmuls")) "cortex_a8_vfp,cortex_a8_vfplite*11") +;; Don't model a reservation for more than 15 cycles as this explodes the +;; state space of the automaton for little gain. It is unlikely that the +;; scheduler will find enough instructions to hide the full latency of the +;; instructions. (define_insn_reservation "cortex_a8_vfp_muld" 17 (and (eq_attr "tune" "cortexa8") (eq_attr "type" "fmuld")) - "cortex_a8_vfp,cortex_a8_vfplite*16") + "cortex_a8_vfp,cortex_a8_vfplite*15") (define_insn_reservation "cortex_a8_vfp_macs" 21 (and (eq_attr "tune" "cortexa8") (eq_attr "type" "fmacs,ffmas")) - "cortex_a8_vfp,cortex_a8_vfplite*20") + "cortex_a8_vfp,cortex_a8_vfplite*15") (define_insn_reservation "cortex_a8_vfp_macd" 26 (and (eq_attr "tune" "cortexa8") (eq_attr "type" "fmacd,ffmad")) - "cortex_a8_vfp,cortex_a8_vfplite*25") + "cortex_a8_vfp,cortex_a8_vfplite*15") (define_insn_reservation "cortex_a8_vfp_divs" 37 (and (eq_attr "tune" "cortexa8") (eq_attr "type" "fdivs, fsqrts")) - "cortex_a8_vfp,cortex_a8_vfplite*36") + "cortex_a8_vfp,cortex_a8_vfplite*15") (define_insn_reservation "cortex_a8_vfp_divd" 65 (and (eq_attr "tune" "cortexa8") (eq_attr "type" "fdivd, fsqrtd")) - "cortex_a8_vfp,cortex_a8_vfplite*64") + "cortex_a8_vfp,cortex_a8_vfplite*15") ;; Comparisons can actually take 7 cycles sometimes instead of four, ;; but given all the other instructions lumped into type=ffarith that |