This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
GCC memory consumption increased by recent patch!
- From: gcctest at suse dot de
- To: jh at suse dot cz, gcc-regression at gcc dot gnu dot org
- Date: Sat, 05 Mar 2005 03:02:29 +0000
- Subject: GCC memory consumption increased by recent patch!
Hi,
Comparing memory consumption on compilation of combine.i and generate-3.4.ii I got:
comparing combine.c compilation at -O0 level:
Overall memory needed: 24661k -> 24665k
Peak memory use before GGC: 9359k
Peak memory use after GGC: 8673k
Maximum of released memory in single GGC run: 2865k
Garbage: 41675k
Leak: 6395k
Overhead: 5772k
GGC runs: 327
comparing combine.c compilation at -O1 level:
Overall memory needed: 24485k -> 24521k
Peak memory use before GGC: 8673k
Peak memory use after GGC: 8093k
Maximum of released memory in single GGC run: 2026k
Garbage: 58507k -> 58556k
Leak: 6756k -> 6756k
Overhead: 9891k -> 9886k
GGC runs: 500
comparing combine.c compilation at -O2 level:
Overall memory needed: 29137k -> 29149k
Peak memory use before GGC: 12671k
Peak memory use after GGC: 12544k
Maximum of released memory in single GGC run: 2533k
Garbage: 78197k -> 78269k
Leak: 6595k -> 6594k
Overhead: 14161k -> 14170k
GGC runs: 520
comparing combine.c compilation at -O3 level:
Amount of memory still referenced at the end of compilation increased from 7174k to 7189k, overall 0.21%
Overall memory needed: 30833k -> 30817k
Peak memory use before GGC: 12714k -> 12713k
Peak memory use after GGC: 12544k
Maximum of released memory in single GGC run: 3391k -> 3392k
Garbage: 105804k -> 105878k
Leak: 7174k -> 7189k
Overhead: 19019k -> 19022k
GGC runs: 583 -> 582
comparing insn-attrtab.c compilation at -O0 level:
Overall memory needed: 114212k
Peak memory use before GGC: 74821k
Peak memory use after GGC: 45494k
Maximum of released memory in single GGC run: 39413k
Garbage: 152830k
Leak: 10984k
Overhead: 19970k
GGC runs: 273
comparing insn-attrtab.c compilation at -O1 level:
Overall memory needed: 124600k
Peak memory use before GGC: 78641k
Peak memory use after GGC: 70036k
Maximum of released memory in single GGC run: 40782k
Garbage: 369270k
Leak: 11361k
Overhead: 71894k
GGC runs: 398
comparing insn-attrtab.c compilation at -O2 level:
Overall memory needed: 148936k
Peak memory use before GGC: 97678k
Peak memory use after GGC: 83497k
Maximum of released memory in single GGC run: 39163k
Garbage: 486952k
Leak: 11252k
Overhead: 87104k
GGC runs: 342
comparing insn-attrtab.c compilation at -O3 level:
Overall memory needed: 137824k
Peak memory use before GGC: 97679k
Peak memory use after GGC: 83499k
Maximum of released memory in single GGC run: 39162k
Garbage: 487762k
Leak: 11287k
Overhead: 87234k
GGC runs: 348
comparing Gerald's testcase PR8361 compilation at -O0 level:
Overall memory needed: 111952k
Peak memory use before GGC: 87325k
Peak memory use after GGC: 85900k
Maximum of released memory in single GGC run: 19511k
Garbage: 246370k
Leak: 55513k
Overhead: 43275k
GGC runs: 366
comparing Gerald's testcase PR8361 compilation at -O1 level:
Overall memory needed: 104037k -> 104033k
Peak memory use before GGC: 86183k
Peak memory use after GGC: 85119k
Maximum of released memory in single GGC run: 18950k
Garbage: 434991k -> 435088k
Leak: 56864k -> 56863k
Overhead: 65164k -> 65172k
GGC runs: 512 -> 513
comparing Gerald's testcase PR8361 compilation at -O2 level:
Overall memory needed: 104249k
Peak memory use before GGC: 86184k
Peak memory use after GGC: 85119k
Maximum of released memory in single GGC run: 18951k
Garbage: 477138k -> 476277k
Leak: 57427k -> 57427k
Overhead: 74970k -> 74892k
GGC runs: 569
comparing Gerald's testcase PR8361 compilation at -O3 level:
Overall memory needed: 105345k -> 105353k
Peak memory use before GGC: 87323k
Peak memory use after GGC: 86199k
Maximum of released memory in single GGC run: 19400k
Garbage: 482585k -> 481450k
Leak: 57637k -> 57631k
Overhead: 75706k -> 75594k
GGC runs: 557
Head of changelog is:
--- /usr/src/SpecTests/sandbox-britten-memory/x86_64/mem-result/ChangeLog 2005-03-04 21:26:30.000000000 +0000
+++ /usr/src/SpecTests/sandbox-britten-memory/gcc/gcc/ChangeLog 2005-03-05 02:03:32.000000000 +0000
@@ -1,3 +1,54 @@
+2005-03-04 John David Anglin <dave.danglin@nrc-cnrc.gc.ca>
+ Joseph S. Myers <joseph@codesourcery.com>
+
+ PR c++/19797
+ * elf.h, pa64-hpux.h, som.h (ASM_OUTPUT_EXTERNAL_REAL): Define.
+ (ASM_OUTPUT_EXTERNAL): Call pa_hpux_asm_output_external.
+ * pa-protos.h (pa_hpux_asm_output_external): Add prototype.
+ * pa.c (get_plabel): Change argument to symbol_ref. Call
+ maybe_get_identifier instead of get_identifier.
+ (pa_hpux_asm_output_external, pa_hpux_file_end): New functions.
+ (struct deferred_plabel): Replace name field with symbol field.
+ (TARGET_ASM_FILE_END): Define as pa_hpux_file_end if
+ ASM_OUTPUT_EXTERNAL_REAL is defined.
+ (output_global_address): Use output_addr_const for output of all
+ symbol_refs.
+ (output_deferred_plabels): Use symbol_ref instead of name for address
+ output.
+ * pa.h (SYMBOL_FLAG_REFERENCED, SYMBOL_REF_REFERENCED_P,
+ ASM_OUTPUT_SYMBOL_REF): New macros.
+ * som.h (ASM_OUTPUT_EXTERNAL_LIBCALL): Use assemble_name_raw. Update
+ comment.
+ (ASM_WEAKEN_LABEL): Use targetm.asm_out.globalize_label to globalize
+ label.
+
+2005-03-05 Joseph S. Myers <joseph@codesourcery.com>
+
+ * config.gcc (hppa*64*-*-hpux11*): Use pa/pa-hpux1111.h for
+ *-*-hpux11.[1-9]*, not just for *-*-hpux11.11.
+
+2005-03-04 Jeff Law <law@redhat.com>
+
+ * basic-block.h (rediscover_loops_after_threading): Declare.
+ * tree-ssa-dom.c: Include cfgloop.h.
+ (tree_ssa_dominator_optimize): Discover loops and some basic
+ properties. Remove forwarder blocks recreated by loop header
+ canonicalization. Also mark backedges in the CFG.
+ * tree-ssa-threadupdate.c: Include cfgloop.h
+ (rediscover_loops_after_threading): Define.
+ (struct local_info): New field, JUMP_THREADED.
+ (prune_undesirable_thread_requests): New function.
+ (redirect_edges): Clear EDGE_ABNORMAL. If edges were threaded
+ then record that fact for the callers of redirct_edges.
+ (thread_block): If BB has incoming backedges, then call
+ prune_undesirable_thraed_requests. Note when we are
+ going to have to rediscover loop information. Return a
+ boolean indicating if any jumps were threaded.
+ (thread_through_all_blocks): Bubble up boolean indicating
+ if any jumps were threaded.
+ * Makefile.in (tree-ssa-dom.o): Depend on cfgloop.h
+ (tree-ssa-threadupdate.o): Similarly.
+
2005-03-04 Kazu Hirata <kazu@cs.umass.edu>
* fold-const.c (fold_ternary): Unroll the "for" loop to
I am friendly script caring about memory consumption in GCC. Please contact
jh@suse.cz if something is going wrong.
The results can be reproduced by building compiler with
--enable-gather-detailed-mem-stats targetting x86-64 and compiling preprocessed
combine.c or testcase from PR8632 with:
-fmem-report --param=ggc-min-heapsize=1024 --param=ggc-min-expand=1 -Ox -Q
The memory consumption summary appears in the dump after detailed listing of
the places they are allocated in. Peak memory consumption is actually computed
by looking for maximal value in {GC XXXX -> YYYY} report.
Yours testing script.