diff options
| author | Alan Modra <amodra@gmail.com> | 2018-12-01 00:05:25 +1030 |
|---|---|---|
| committer | Alan Modra <amodra@gcc.gnu.org> | 2018-12-01 00:05:25 +1030 |
| commit | b49f1a7e7c85add83637cf7df46b8fc9b5b299e9 (patch) | |
| tree | d3cba1d26194ebe465ba3abbdc1bc4b7e64133af /gcc/cfgexpand.c | |
| parent | ceb8c8c4cae47fd07cd7dda74bb31b98ce4adadc (diff) | |
| download | gcc-b49f1a7e7c85add83637cf7df46b8fc9b5b299e9.zip gcc-b49f1a7e7c85add83637cf7df46b8fc9b5b299e9.tar.gz gcc-b49f1a7e7c85add83637cf7df46b8fc9b5b299e9.tar.bz2 | |
[RS6000] PowerPC64 soft-float
This patch aims to prevent long sequences loading soft-float
constants. For 32-bit, it makes sense to load values inline to a gpr
with lis, addi, but not so much for 64-bit where a 5 insn sequence
might be needed for each gpr. For TFmode in particular, a 10 insn
sequence is reduced to 2 loads from memory plus 1 or 2 address setup
insns.
* config/rs6000/predicates.md (easy_fp_constant): Avoid long
dependent insn sequences.
* config/rs6000/rs6000.c (num_insns_constant): Support long
double constants.
* config/rs6000/rs6000.md (mov<mode>_softfloat <FMOVE128>): Adjust
length attribute.
From-SVN: r266663
Diffstat (limited to 'gcc/cfgexpand.c')
0 files changed, 0 insertions, 0 deletions