public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking
@ 2022-05-27 8:49 dcb314 at hotmail dot com
2022-05-27 9:14 ` [Bug middle-end/105747] " pinskia at gcc dot gnu.org
` (9 more replies)
0 siblings, 10 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-05-27 8:49 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
Bug ID: 105747
Summary: Possible resource hog with -O2 -fno-var-tracking
Product: gcc
Version: unknown
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: dcb314 at hotmail dot com
Target Milestone: ---
Created attachment 53041
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53041&action=edit
C source code
The attached code seems to take a suspiciously long time on
the raspberry pi.
dcb@raspberrypi:~/csmith $ (ulimit -t 600; time /home/dcb/gcc/results/bin/gcc
-c -w -O2 -fno-var-tracking bug814.c )
real 7m4.234s
user 7m2.175s
sys 0m0.549s
dcb@raspberrypi:~/csmith $
Time on the x86_64 is about 20 seconds.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug middle-end/105747] Possible resource hog with -O2 -fno-var-tracking
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
@ 2022-05-27 9:14 ` pinskia at gcc dot gnu.org
2022-05-27 9:22 ` dcb314 at hotmail dot com
` (8 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: pinskia at gcc dot gnu.org @ 2022-05-27 9:14 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Component|c |middle-end
Keywords| |compile-time-hog
--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Can you give the output of gcc -v?
Also maybe add -ftime-report and give that output too
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug middle-end/105747] Possible resource hog with -O2 -fno-var-tracking
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
2022-05-27 9:14 ` [Bug middle-end/105747] " pinskia at gcc dot gnu.org
@ 2022-05-27 9:22 ` dcb314 at hotmail dot com
2022-05-27 9:31 ` [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes pinskia at gcc dot gnu.org
` (7 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-05-27 9:22 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #2 from David Binderman <dcb314 at hotmail dot com> ---
(In reply to Andrew Pinski from comment #1)
> Can you give the output of gcc -v?
Sure.
dcb@raspberrypi:~/csmith $ /home/dcb/gcc/results/bin/gcc -v
Using built-in specs.
COLLECT_GCC=/home/dcb/gcc/results/bin/gcc
COLLECT_LTO_WRAPPER=/home/dcb/gcc/results.20220525/libexec/gcc/arm-linux-gnueabihf/13.0.0/lto-wrapper
Target: arm-linux-gnueabihf
Configured with: ../trunk/configure --prefix=/home/dcb/gcc/results.20220525
--disable-bootstrap --disable-multilib --disable-werror
--with-pkgversion=850a9ce8bcca59c7 --enable-checking=yes
--enable-languages=c,c++ --with-cpu=cortex-a72 --with-fpu=neon-fp-armv8
--with-float=hard --build=arm-linux-gnueabihf --host=arm-linux-gnueabihf
--target=arm-linux-gnueabihf
Thread model: posix
Supported LTO compression algorithms: zlib
gcc version 13.0.0 20220525 (experimental) (850a9ce8bcca59c7)
dcb@raspberrypi:~/csmith $
> Also maybe add -ftime-report and give that output too
dcb@raspberrypi:~/csmith $ (ulimit -t 600; time /home/dcb/gcc/results/bin/gcc
-c -w -O2 -fno-var-tracking -ftime-report bug814.c )
Time variable usr sys wall
GGC
phase setup : 0.05 ( 0%) 0.01 ( 0%) 0.10 ( 0%)
2166k ( 2%)
phase parsing : 0.92 ( 0%) 1.68 ( 50%) 2.64 ( 1%)
3162k ( 3%)
phase opt and generate : 413.81 (100%) 1.65 ( 49%) 416.39 ( 99%)
108M ( 95%)
phase last asm : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
2432 ( 0%)
garbage collection : 0.95 ( 0%) 0.03 ( 1%) 0.99 ( 0%)
0 ( 0%)
dump files : 0.02 ( 0%) 0.00 ( 0%) 0.06 ( 0%)
0 ( 0%)
callgraph construction : 0.17 ( 0%) 0.03 ( 1%) 0.21 ( 0%)
1731k ( 1%)
callgraph optimization : 0.11 ( 0%) 0.01 ( 0%) 0.11 ( 0%)
2360 ( 0%)
callgraph functions expansion : 379.05 ( 91%) 1.00 ( 30%) 380.77 ( 91%)
102M ( 90%)
callgraph ipa passes : 34.08 ( 8%) 0.64 ( 19%) 34.92 ( 8%)
3324k ( 3%)
ipa function summary : 0.03 ( 0%) 0.00 ( 0%) 0.04 ( 0%)
30k ( 0%)
ipa dead code removal : 0.02 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
ipa cp : 0.07 ( 0%) 0.00 ( 0%) 0.07 ( 0%)
98k ( 0%)
ipa inlining heuristics : 0.04 ( 0%) 0.00 ( 0%) 0.03 ( 0%)
105k ( 0%)
ipa function splitting : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
36k ( 0%)
ipa pure const : 0.05 ( 0%) 0.00 ( 0%) 0.02 ( 0%)
1800 ( 0%)
ipa icf : 0.04 ( 0%) 0.00 ( 0%) 0.04 ( 0%)
0 ( 0%)
ipa free inline summary : 0.02 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
0 ( 0%)
ipa modref : 0.01 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
11k ( 0%)
cfg construction : 0.00 ( 0%) 0.00 ( 0%) 0.02 ( 0%)
9952 ( 0%)
cfg cleanup : 0.06 ( 0%) 0.00 ( 0%) 0.05 ( 0%)
13k ( 0%)
CFG verifier : 0.95 ( 0%) 0.00 ( 0%) 1.08 ( 0%)
0 ( 0%)
trivially dead code : 0.18 ( 0%) 0.00 ( 0%) 0.20 ( 0%)
0 ( 0%)
df scan insns : 0.10 ( 0%) 0.00 ( 0%) 0.11 ( 0%)
672 ( 0%)
df reaching defs : 0.05 ( 0%) 0.00 ( 0%) 0.10 ( 0%)
0 ( 0%)
df live regs : 0.83 ( 0%) 0.01 ( 0%) 0.91 ( 0%)
0 ( 0%)
df live&initialized regs : 0.14 ( 0%) 0.01 ( 0%) 0.16 ( 0%)
0 ( 0%)
df use-def / def-use chains : 0.04 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
df live reg subwords : 0.02 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
df reg dead/unused notes : 0.36 ( 0%) 0.00 ( 0%) 0.33 ( 0%)
505k ( 0%)
register information : 0.15 ( 0%) 0.00 ( 0%) 0.19 ( 0%)
0 ( 0%)
alias analysis : 0.30 ( 0%) 0.01 ( 0%) 0.37 ( 0%)
2877k ( 2%)
alias stmt walking : 68.16 ( 16%) 0.46 ( 14%) 69.10 ( 16%)
114k ( 0%)
register scan : 0.07 ( 0%) 0.00 ( 0%) 0.06 ( 0%)
31k ( 0%)
rebuild jump labels : 0.05 ( 0%) 0.00 ( 0%) 0.05 ( 0%)
0 ( 0%)
preprocessing : 0.20 ( 0%) 0.33 ( 10%) 0.57 ( 0%)
240k ( 0%)
lexical analysis : 0.36 ( 0%) 0.62 ( 19%) 0.90 ( 0%)
0 ( 0%)
parser (global) : 0.03 ( 0%) 0.18 ( 5%) 0.24 ( 0%)
784k ( 1%)
parser function body : 0.32 ( 0%) 0.53 ( 16%) 0.92 ( 0%)
2059k ( 2%)
parser inl. func. body : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
46k ( 0%)
early inlining heuristics : 0.02 ( 0%) 0.01 ( 0%) 0.01 ( 0%)
141k ( 0%)
inline parameters : 0.12 ( 0%) 0.01 ( 0%) 0.04 ( 0%)
77k ( 0%)
integration : 0.07 ( 0%) 0.02 ( 1%) 0.07 ( 0%)
629k ( 1%)
tree gimplify : 0.20 ( 0%) 0.00 ( 0%) 0.21 ( 0%)
2735k ( 2%)
tree CFG construction : 0.01 ( 0%) 0.01 ( 0%) 0.01 ( 0%)
182k ( 0%)
tree CFG cleanup : 0.14 ( 0%) 0.00 ( 0%) 0.10 ( 0%)
75k ( 0%)
tree VRP : 0.13 ( 0%) 0.00 ( 0%) 0.13 ( 0%)
94k ( 0%)
tree Early VRP : 0.17 ( 0%) 0.00 ( 0%) 0.20 ( 0%)
154k ( 0%)
tree copy propagation : 0.03 ( 0%) 0.00 ( 0%) 0.06 ( 0%)
152 ( 0%)
tree PTA : 0.20 ( 0%) 0.01 ( 0%) 0.19 ( 0%)
56k ( 0%)
tree SSA rewrite : 0.04 ( 0%) 0.06 ( 2%) 0.09 ( 0%)
557k ( 0%)
tree SSA incremental : 0.14 ( 0%) 0.01 ( 0%) 0.21 ( 0%)
343k ( 0%)
tree operand scan : 0.15 ( 0%) 0.11 ( 3%) 0.25 ( 0%)
666k ( 1%)
dominator optimization : 0.21 ( 0%) 0.00 ( 0%) 0.20 ( 0%)
778k ( 1%)
backwards jump threading : 0.11 ( 0%) 0.02 ( 1%) 0.12 ( 0%)
165k ( 0%)
tree SRA : 0.09 ( 0%) 0.00 ( 0%) 0.06 ( 0%)
30k ( 0%)
tree CCP : 0.17 ( 0%) 0.00 ( 0%) 0.20 ( 0%)
60k ( 0%)
tree reassociation : 0.02 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
320 ( 0%)
tree PRE : 0.80 ( 0%) 0.01 ( 0%) 0.84 ( 0%)
742k ( 1%)
tree FRE : 2.00 ( 0%) 0.02 ( 1%) 1.77 ( 0%)
1294k ( 1%)
tree code sinking : 0.11 ( 0%) 0.00 ( 0%) 0.08 ( 0%)
11k ( 0%)
tree linearize phis : 0.07 ( 0%) 0.01 ( 0%) 0.07 ( 0%)
5848 ( 0%)
tree backward propagate : 0.00 ( 0%) 0.01 ( 0%) 0.00 ( 0%)
0 ( 0%)
tree forward propagate : 0.11 ( 0%) 0.00 ( 0%) 0.10 ( 0%)
33k ( 0%)
tree phiprop : 0.02 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
0 ( 0%)
tree conservative DCE : 0.07 ( 0%) 0.02 ( 1%) 0.07 ( 0%)
128 ( 0%)
tree aggressive DCE : 0.04 ( 0%) 0.02 ( 1%) 0.04 ( 0%)
65k ( 0%)
tree DSE : 42.71 ( 10%) 0.00 ( 0%) 42.76 ( 10%)
11k ( 0%)
tree loop optimization : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
tree loop invariant motion : 0.02 ( 0%) 0.00 ( 0%) 0.03 ( 0%)
17k ( 0%)
tree canonical iv : 0.02 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
13k ( 0%)
scev constant prop : 0.00 ( 0%) 0.00 ( 0%) 0.03 ( 0%)
9456 ( 0%)
complete unrolling : 0.08 ( 0%) 0.00 ( 0%) 0.05 ( 0%)
71k ( 0%)
tree vectorization : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
25k ( 0%)
tree slp vectorization : 0.24 ( 0%) 0.00 ( 0%) 0.22 ( 0%)
1519k ( 1%)
tree loop distribution : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
6288 ( 0%)
tree iv optimization : 0.08 ( 0%) 0.00 ( 0%) 0.09 ( 0%)
223k ( 0%)
predictive commoning : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
14k ( 0%)
tree copy headers : 0.03 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
36k ( 0%)
tree SSA verifier : 2.27 ( 1%) 0.07 ( 2%) 2.18 ( 1%)
0 ( 0%)
tree STMT verifier : 4.90 ( 1%) 0.12 ( 4%) 5.14 ( 1%)
0 ( 0%)
tree switch lowering : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
tree strlen optimization : 0.63 ( 0%) 0.00 ( 0%) 0.63 ( 0%)
19k ( 0%)
tree modref : 0.03 ( 0%) 0.00 ( 0%) 0.02 ( 0%)
18k ( 0%)
callgraph verifier : 0.03 ( 0%) 0.00 ( 0%) 0.05 ( 0%)
0 ( 0%)
dominance computation : 0.14 ( 0%) 0.00 ( 0%) 0.24 ( 0%)
0 ( 0%)
out of ssa : 0.01 ( 0%) 0.00 ( 0%) 0.03 ( 0%)
776 ( 0%)
expand vars : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
103k ( 0%)
expand : 0.37 ( 0%) 0.00 ( 0%) 0.37 ( 0%)
4487k ( 4%)
post expand cleanups : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
21k ( 0%)
varconst : 0.00 ( 0%) 0.02 ( 1%) 0.00 ( 0%)
9000 ( 0%)
lower subreg : 0.09 ( 0%) 0.00 ( 0%) 0.08 ( 0%)
46k ( 0%)
jump : 0.01 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
0 ( 0%)
forward prop : 0.35 ( 0%) 0.00 ( 0%) 0.34 ( 0%)
237k ( 0%)
CSE : 8.07 ( 2%) 0.00 ( 0%) 8.09 ( 2%)
2159k ( 2%)
dead code elimination : 0.11 ( 0%) 0.00 ( 0%) 0.11 ( 0%)
0 ( 0%)
dead store elim1 : 1.24 ( 0%) 0.00 ( 0%) 1.23 ( 0%)
797k ( 1%)
dead store elim2 : 10.56 ( 3%) 0.07 ( 2%) 10.68 ( 3%)
36M ( 32%)
loop init : 0.10 ( 0%) 0.02 ( 1%) 0.16 ( 0%)
443k ( 0%)
loop invariant motion : 0.02 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
7800 ( 0%)
loop doloop : 0.07 ( 0%) 0.00 ( 0%) 0.07 ( 0%)
5488 ( 0%)
CPROP : 0.52 ( 0%) 0.00 ( 0%) 0.46 ( 0%)
330k ( 0%)
PRE : 12.48 ( 3%) 0.13 ( 4%) 12.63 ( 3%)
37M ( 33%)
auto inc dec : 0.04 ( 0%) 0.00 ( 0%) 0.04 ( 0%)
3336 ( 0%)
CSE 2 : 6.13 ( 1%) 0.01 ( 0%) 6.13 ( 1%)
798k ( 1%)
branch prediction : 0.03 ( 0%) 0.02 ( 1%) 0.02 ( 0%)
17k ( 0%)
combiner : 0.56 ( 0%) 0.00 ( 0%) 0.59 ( 0%)
879k ( 1%)
scheduling : 210.15 ( 51%) 0.08 ( 2%) 210.56 ( 50%)
1351k ( 1%)
integrated RA : 4.34 ( 1%) 0.00 ( 0%) 4.27 ( 1%)
4165k ( 4%)
LRA non-specific : 0.77 ( 0%) 0.01 ( 0%) 0.77 ( 0%)
757k ( 1%)
LRA virtuals elimination : 0.13 ( 0%) 0.00 ( 0%) 0.13 ( 0%)
85k ( 0%)
LRA reload inheritance : 0.19 ( 0%) 0.00 ( 0%) 0.21 ( 0%)
419k ( 0%)
LRA create live ranges : 0.33 ( 0%) 0.00 ( 0%) 0.32 ( 0%)
141k ( 0%)
LRA hard reg assignment : 7.85 ( 2%) 0.02 ( 1%) 7.93 ( 2%)
0 ( 0%)
LRA coalesce pseudo regs : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
LRA rematerialization : 0.08 ( 0%) 0.00 ( 0%) 0.08 ( 0%)
72 ( 0%)
reload CSE regs : 14.01 ( 3%) 0.00 ( 0%) 14.02 ( 3%)
413k ( 0%)
thread pro- & epilogue : 0.11 ( 0%) 0.00 ( 0%) 0.12 ( 0%)
22k ( 0%)
if-conversion 2 : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
800 ( 0%)
peephole 2 : 0.03 ( 0%) 0.00 ( 0%) 0.04 ( 0%)
912 ( 0%)
scheduling fusion : 0.90 ( 0%) 0.01 ( 0%) 0.91 ( 0%)
103k ( 0%)
hard reg cprop : 0.16 ( 0%) 0.00 ( 0%) 0.15 ( 0%)
408 ( 0%)
scheduling 2 : 1.13 ( 0%) 0.01 ( 0%) 1.16 ( 0%)
154k ( 0%)
machine dep reorg : 0.06 ( 0%) 0.00 ( 0%) 0.05 ( 0%)
9976 ( 0%)
reorder blocks : 0.05 ( 0%) 0.00 ( 0%) 0.02 ( 0%)
21k ( 0%)
shorten branches : 0.04 ( 0%) 0.00 ( 0%) 0.04 ( 0%)
2304 ( 0%)
final : 0.10 ( 0%) 0.03 ( 1%) 0.13 ( 0%)
3064 ( 0%)
tree if-combine : 0.00 ( 0%) 0.00 ( 0%) 0.02 ( 0%)
192 ( 0%)
straight-line strength reduction : 0.02 ( 0%) 0.00 ( 0%) 0.03 ( 0%)
648 ( 0%)
store merging : 0.37 ( 0%) 0.01 ( 0%) 0.39 ( 0%)
320k ( 0%)
initialize rtl : 0.01 ( 0%) 0.01 ( 0%) 0.01 ( 0%)
4984 ( 0%)
tree loop if-conversion : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
29k ( 0%)
access analysis : 0.03 ( 0%) 0.01 ( 0%) 0.03 ( 0%)
96 ( 0%)
early local passes : 0.00 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
0 ( 0%)
rest of compilation : 0.22 ( 0%) 0.02 ( 1%) 0.33 ( 0%)
527k ( 0%)
remove unused locals : 0.07 ( 0%) 0.01 ( 0%) 0.08 ( 0%)
0 ( 0%)
address taken : 0.05 ( 0%) 0.07 ( 2%) 0.17 ( 0%)
0 ( 0%)
verify loop closed : 0.01 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
0 ( 0%)
verify RTL sharing : 2.36 ( 1%) 0.01 ( 0%) 2.34 ( 1%)
0 ( 0%)
rebuild frequencies : 0.01 ( 0%) 0.00 ( 0%) 0.00 ( 0%)
0 ( 0%)
repair loop structures : 0.01 ( 0%) 0.00 ( 0%) 0.01 ( 0%)
216 ( 0%)
TOTAL : 414.79 3.34 419.14
113M
Extra diagnostic checks enabled; compiler may run slowly.
Configure with --enable-checking=release to disable checks.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
2022-05-27 9:14 ` [Bug middle-end/105747] " pinskia at gcc dot gnu.org
2022-05-27 9:22 ` dcb314 at hotmail dot com
@ 2022-05-27 9:31 ` pinskia at gcc dot gnu.org
2022-05-27 11:14 ` rguenth at gcc dot gnu.org
` (6 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: pinskia at gcc dot gnu.org @ 2022-05-27 9:31 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
Andrew Pinski <pinskia at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Summary|Possible resource hog with |Scheduler can take a long
|-O2 -fno-var-tracking |time on arm-linux sometimes
Target| |arm-linux-gnueabihf
Component|middle-end |rtl-optimization
--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Not shocking really the scheduler is taking a long time.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (2 preceding siblings ...)
2022-05-27 9:31 ` [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes pinskia at gcc dot gnu.org
@ 2022-05-27 11:14 ` rguenth at gcc dot gnu.org
2022-06-07 8:28 ` dcb314 at hotmail dot com
` (5 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: rguenth at gcc dot gnu.org @ 2022-05-27 11:14 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Version|unknown |13.0
--- Comment #4 from Richard Biener <rguenth at gcc dot gnu.org> ---
Confirmed with a cross from x86_64:
alias stmt walking : 2.47 ( 27%)
tree DSE : 1.52 ( 17%)
scheduling : 1.36 ( 15%)
TOTAL : 9.15
so it's not only scheduling. perf (release checking) shows:
Samples: 37K of event 'cycles', Event count (approx.): 41984911906
Overhead Samples Command Shared Object Symbol
26.20% 9765 cc1 cc1 [.] get_ref_base_and_extent
4.37% 1635 cc1 cc1 [.] find_base_term
3.75% 1392 cc1 cc1 [.] rank_for_schedule
3.13% 1168 cc1 cc1 [.] wi::lshift_large
2.52% 938 cc1 cc1 [.] canonize
2.36% 876 cc1 cc1 [.] sd_lists_size
2.27% 844 cc1 cc1 [.]
sd_find_dep_between_no_cache
1.89% 701 cc1 cc1 [.]
operand_compare::operand_equal_p
1.80% 671 cc1 cc1 [.] wi::mul_internal
1.71% 637 cc1 cc1 [.]
simplify_context::simplify_plus_minus
1.63% 613 cc1 cc1 [.] memrefs_conflict_p
note with a non-optimized compiler and checking enabled (but -fno-checking)
things are much slower.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (3 preceding siblings ...)
2022-05-27 11:14 ` rguenth at gcc dot gnu.org
@ 2022-06-07 8:28 ` dcb314 at hotmail dot com
2022-06-07 9:30 ` dcb314 at hotmail dot com
` (4 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-06-07 8:28 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #5 from David Binderman <dcb314 at hotmail dot com> ---
Created attachment 53097
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53097&action=edit
C source code
This source code file takes something over ten minutes,
with compiler flags -fno-var-tracking -g -Wall -O3 -march=native
on a 32 bit ARM machine, like raspberry PI 3.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (4 preceding siblings ...)
2022-06-07 8:28 ` dcb314 at hotmail dot com
@ 2022-06-07 9:30 ` dcb314 at hotmail dot com
2022-06-11 9:45 ` dcb314 at hotmail dot com
` (3 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-06-07 9:30 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #6 from David Binderman <dcb314 at hotmail dot com> ---
Here are the results from -ftime-report, with the 0.0% lines removed:
Time variable usr sys wall
GGC
phase opt and generate : 199.90 (100%) 0.92 ( 87%) 206.57 (100%)
581M ( 99%)
callgraph functions expansion : 193.45 ( 97%) 0.81 ( 76%) 199.89 ( 97%)
568M ( 96%)
callgraph ipa passes : 6.32 ( 3%) 0.10 ( 9%) 6.53 ( 3%)
5915k ( 1%)
CSE : 1.18 ( 1%) 0.01 ( 1%) 1.22 ( 1%)
3346k ( 1%)
dead store elim2 : 2.62 ( 1%) 0.07 ( 7%) 2.75 ( 1%)
85M ( 14%)
PRE : 14.61 ( 7%) 0.35 ( 33%) 15.48 ( 7%)
385M ( 65%)
scheduling : 142.72 ( 71%) 0.12 ( 11%) 146.64 ( 71%)
5015k ( 1%)
integrated RA : 1.08 ( 1%) 0.02 ( 2%) 1.15 ( 1%)
26M ( 4%)
TOTAL : 200.06 1.06 206.91
589M
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (5 preceding siblings ...)
2022-06-07 9:30 ` dcb314 at hotmail dot com
@ 2022-06-11 9:45 ` dcb314 at hotmail dot com
2022-06-11 9:48 ` dcb314 at hotmail dot com
` (2 subsequent siblings)
9 siblings, 0 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-06-11 9:45 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #7 from David Binderman <dcb314 at hotmail dot com> ---
Created attachment 53119
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53119&action=edit
C source code
Another one. Over 15 minutes this time.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (6 preceding siblings ...)
2022-06-11 9:45 ` dcb314 at hotmail dot com
@ 2022-06-11 9:48 ` dcb314 at hotmail dot com
2022-06-12 9:40 ` dcb314 at hotmail dot com
2024-02-19 13:29 ` rguenth at gcc dot gnu.org
9 siblings, 0 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-06-11 9:48 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #8 from David Binderman <dcb314 at hotmail dot com> ---
(In reply to David Binderman from comment #7)
> Created attachment 53119 [details]
> C source code
>
> Another one. Over 15 minutes this time.
Flag -O2 required.
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (7 preceding siblings ...)
2022-06-11 9:48 ` dcb314 at hotmail dot com
@ 2022-06-12 9:40 ` dcb314 at hotmail dot com
2024-02-19 13:29 ` rguenth at gcc dot gnu.org
9 siblings, 0 replies; 11+ messages in thread
From: dcb314 at hotmail dot com @ 2022-06-12 9:40 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #9 from David Binderman <dcb314 at hotmail dot com> ---
(In reply to David Binderman from comment #8)
> (In reply to David Binderman from comment #7)
> > Created attachment 53119 [details]
> > C source code
> >
> > Another one. Over 15 minutes this time.
>
> Flag -O2 required.
Runs in 21 minutes on an RP3.
dcb@raspberrypi:~/csmith $ (ulimit -t 1800;time /home/dcb/gcc/results/bin/gcc
-I/home/dcb/csmith/results/include/csmith-2.3.0/
-c -g -w -O2 -freport-bug -ftime-report destDir/testFile.23139.c )
Time variable usr sys wall
GGC
phase parsing : 0.77 ( 0%) 1.55 ( 33%) 2.31 ( 0%)
3597k ( 2%)
phase opt and generate :1261.54 (100%) 3.05 ( 66%)1267.62 (100%)
188M ( 97%)
callgraph functions expansion :1212.97 ( 96%) 2.24 ( 48%)1218.13 ( 96%)
179M ( 93%)
callgraph ipa passes : 47.88 ( 4%) 0.79 ( 17%) 48.78 ( 4%)
4512k ( 2%)
alias stmt walking : 125.32 ( 10%) 0.70 ( 15%) 127.13 ( 10%)
126k ( 0%)
preprocessing : 0.20 ( 0%) 0.44 ( 9%) 0.58 ( 0%)
941k ( 0%)
lexical analysis : 0.25 ( 0%) 0.47 ( 10%) 0.87 ( 0%)
0 ( 0%)
parser function body : 0.23 ( 0%) 0.49 ( 11%) 0.68 ( 0%)
1585k ( 1%)
tree operand scan : 0.31 ( 0%) 0.33 ( 7%) 0.71 ( 0%)
830k ( 0%)
tree SSA verifier : 3.12 ( 0%) 0.10 ( 2%) 3.12 ( 0%)
0 ( 0%)
dead store elim2 : 16.05 ( 1%) 0.13 ( 3%) 16.30 ( 1%)
62M ( 32%)
PRE : 19.65 ( 2%) 0.25 ( 5%) 19.90 ( 2%)
46M ( 24%)
scheduling : 865.21 ( 69%) 0.20 ( 4%) 867.61 ( 68%)
4186k ( 2%)
address taken : 0.16 ( 0%) 0.14 ( 3%) 0.22 ( 0%)
0 ( 0%)
Extra diagnostic checks enabled; compiler may run slowly.
Configure with --enable-checking=release to disable checks.
real 21m10.578s
user 21m2.776s
sys 0m4.705s
^ permalink raw reply [flat|nested] 11+ messages in thread
* [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
` (8 preceding siblings ...)
2022-06-12 9:40 ` dcb314 at hotmail dot com
@ 2024-02-19 13:29 ` rguenth at gcc dot gnu.org
9 siblings, 0 replies; 11+ messages in thread
From: rguenth at gcc dot gnu.org @ 2024-02-19 13:29 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105747
--- Comment #10 from Richard Biener <rguenth at gcc dot gnu.org> ---
There were some scheduling fixes for GCC 14, possibly time to re-evaluate.
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2024-02-19 13:29 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-05-27 8:49 [Bug c/105747] New: Possible resource hog with -O2 -fno-var-tracking dcb314 at hotmail dot com
2022-05-27 9:14 ` [Bug middle-end/105747] " pinskia at gcc dot gnu.org
2022-05-27 9:22 ` dcb314 at hotmail dot com
2022-05-27 9:31 ` [Bug rtl-optimization/105747] Scheduler can take a long time on arm-linux sometimes pinskia at gcc dot gnu.org
2022-05-27 11:14 ` rguenth at gcc dot gnu.org
2022-06-07 8:28 ` dcb314 at hotmail dot com
2022-06-07 9:30 ` dcb314 at hotmail dot com
2022-06-11 9:45 ` dcb314 at hotmail dot com
2022-06-11 9:48 ` dcb314 at hotmail dot com
2022-06-12 9:40 ` dcb314 at hotmail dot com
2024-02-19 13:29 ` rguenth at gcc dot gnu.org
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).