This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug middle-end/10776] [3.3/3.4 regression] Large aggregate initializers with a single relocated entry causes excessive compile time regression
- From: "steven at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 25 Jul 2003 16:58:12 -0000
- Subject: [Bug middle-end/10776] [3.3/3.4 regression] Large aggregate initializers with a single relocated entry causes excessive compile time regression
- References: <20030514062600.10776.gonz@ratloop.com>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=10776
steven at gcc dot gnu dot org changed:
What |Removed |Added
----------------------------------------------------------------------------
Last reconfirmed|0000-00-00 00:00:00 |2003-07-25 16:58:11
date| |
------- Additional Comments From steven at gcc dot gnu dot org 2003-07-25 16:58 -------
Compiler:
GNU C++ version 3.4 20030725 (experimental) (i686-pc-linux-gnu)
compiled by GNU C version 3.4 20030725 (experimental).
GGC heuristics: --param ggc-min-expand=47 --param ggc-min-heapsize=31916
Flags:
-O -quiet
File: pr10776.cc
Flat profile:
Each sample counts as 0.01 seconds.
% cumulative self self total
time seconds seconds calls ms/call ms/call name
27.10 34.24 34.24 142911 0.24 0.24 cselib_lookup
20.58 60.24 26.00 12691 2.05 2.44 find_equiv_reg
19.54 84.93 24.69 91778260 0.00 0.00 read_dependence
13.26 101.68 16.75 91778261 0.00 0.00 main
4.09 106.85 5.17 178078426 0.00 0.00 rtx_equal_p
3.12 110.79 3.94 85515155 0.00 0.00 canon_true_dependence
2.16 113.53 2.73 6432761 0.00 0.00 for_each_rtx
1.80 115.80 2.27 85515155 0.00 0.00 anti_dependence
1.74 118.00 2.20 248931 0.01 0.01 rtx_equal_for_cselib_p
1.20 119.50 1.51 81076436 0.00 0.00 find_reg_note
1.07 120.86 1.35 91778260 0.00 0.00 alias_sets_conflict_p
(all others <1%)
So this is another example of find_reg_equiv being one of the serious bottle
necks, just like Bug 2692. Maybe the two are related...