https://gcc.gnu.org/bugzilla/show_bug.cgi?id=90595
Bug ID: 90595
Summary: LRA liveness analysis is slow
Product: gcc
Version: 10.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: rtl-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: rguenth at gcc dot gnu.org
Target Milestone: ---
Split out from PR88440. As of today you'll see a big compile-time increase
for compiling module_configure.fppized.f90 of 521.wrf_r at -O2.
Time profile before/after:
╔══════════════════════════╤════════╤════════╤═════════╗
║ PASS │ Before │ After │ Change ║
╠══════════════════════════╪════════╪════════╪═════════╣
║ backwards jump threading │ 6.29 │ 6.16 │ 97.93% ║
║ integrated RA │ 6.76 │ 6.41 │ 94.82% ║
║ tree SSA incremental │ 9.01 │ 11.16 │ 123.86% ║
║ LRA create live ranges │ 15.68 │ 40.02 │ 255.23% ║
║ PRE │ 23.24 │ 32.32 │ 139.07% ║
║ alias stmt walking │ 27.69 │ 28.75 │ 103.83% ║
║ phase opt and generate │ 124.13 │ 163.95 │ 132.08% ║
║ TOTAL │ 125.39 │ 165.17 │ 131.73% ║
so LRA live ranges is already slow before. perf profiling after the change
shows
Samples: 579 of event 'cycles:ppp', Event count (approx.): 257134187434191
Overhead Command Shared Object Symbol
22.26% f951 f951 [.] process_bb_lives
15.06% f951 f951 [.] ix86_hard_regno_call_part_clobbered
8.55% f951 f951 [.] concat
6.88% f951 f951 [.] find_base_term
3.60% f951 f951 [.] get_ref_base_and_extent
3.27% f951 f951 [.] find_base_term
2.95% f951 f951 [.] make_hard_regno_dead
which IMHO points at
static inline void
check_pseudos_live_through_calls (int regno,
HARD_REG_SET last_call_used_reg_set,
rtx_insn *call_insn)
{
...
for (hr = 0; HARD_REGISTER_NUM_P (hr); hr++)
if (targetm.hard_regno_call_part_clobbered (call_insn, hr,
PSEUDO_REGNO_MODE (regno)))
add_to_hard_reg_set (&lra_reg_info[regno].conflict_hard_regs,
PSEUDO_REGNO_MODE (regno), hr);
where we do a lot of redundant work because we call this function in a loop
possibly very many times with the same arguments besides regno (but with
same PSEUDO_REGNO_MODE).
