A recent patch decreased GCC's memory consumption.

[gcctest at suse dot de]

Hi,

I am a friendly script caring about memory consumption in GCC.  Please
contact jh@suse.cz if something is going wrong.

Comparing memory consumption on compilation of combine.i, insn-attrtab.i,
and generate-3.4.ii I got:


comparing combine.c compilation at -O0 level:
  Ovarall memory allocated via mmap and sbrk decreased from 24789k to 23769k, overall -4.29%
    Overall memory needed: 24789k -> 23769k
    Peak memory use before GGC: 9590k
    Peak memory use after GGC: 8937k
    Maximum of released memory in single GGC run: 2754k
    Garbage: 41361k
    Leak: 6690k
    Overhead: 5813k
    GGC runs: 320

comparing combine.c compilation at -O1 level:
    Overall memory needed: 26812k
    Peak memory use before GGC: 17362k
    Peak memory use after GGC: 17173k
    Maximum of released memory in single GGC run: 2402k
    Garbage: 66361k
    Leak: 7195k
    Overhead: 7737k
    GGC runs: 415

comparing combine.c compilation at -O2 level:
    Overall memory needed: 26812k
    Peak memory use before GGC: 17365k
    Peak memory use after GGC: 17173k
    Maximum of released memory in single GGC run: 2504k
    Garbage: 84487k
    Leak: 7436k
    Overhead: 10280k
    GGC runs: 481

comparing combine.c compilation at -O3 level:
    Overall memory needed: 26812k
    Peak memory use before GGC: 17616k
    Peak memory use after GGC: 17173k
    Maximum of released memory in single GGC run: 3353k
    Garbage: 115530k
    Leak: 7506k
    Overhead: 13498k
    GGC runs: 535

comparing insn-attrtab.c compilation at -O0 level:
    Overall memory needed: 81648k
    Peak memory use before GGC: 70263k
    Peak memory use after GGC: 45002k
    Maximum of released memory in single GGC run: 37047k
    Garbage: 148647k
    Leak: 10091k
    Overhead: 19857k
    GGC runs: 250

comparing insn-attrtab.c compilation at -O1 level:
    Overall memory needed: 112156k -> 112552k
    Peak memory use before GGC: 94389k
    Peak memory use after GGC: 83710k
    Maximum of released memory in single GGC run: 32776k
    Garbage: 294413k
    Leak: 10080k
    Overhead: 36804k
    GGC runs: 250

comparing insn-attrtab.c compilation at -O2 level:
    Overall memory needed: 123336k -> 123328k
    Peak memory use before GGC: 113436k
    Peak memory use after GGC: 83674k
    Maximum of released memory in single GGC run: 32461k
    Garbage: 380610k
    Leak: 10194k
    Overhead: 48618k
    GGC runs: 279

comparing insn-attrtab.c compilation at -O3 level:
    Overall memory needed: 123352k -> 123336k
    Peak memory use before GGC: 113461k
    Peak memory use after GGC: 83698k
    Maximum of released memory in single GGC run: 32776k
    Garbage: 381139k
    Leak: 10210k
    Overhead: 48672k
    GGC runs: 282

comparing Gerald's testcase PR8361 compilation at -O0 level:
    Overall memory needed: 118616k
    Peak memory use before GGC: 95321k
    Peak memory use after GGC: 94374k
    Maximum of released memory in single GGC run: 20159k
    Garbage: 227542k
    Leak: 49361k
    Overhead: 36908k
    GGC runs: 344

comparing Gerald's testcase PR8361 compilation at -O1 level:
    Overall memory needed: 105532k
    Peak memory use before GGC: 94274k
    Peak memory use after GGC: 93338k
    Maximum of released memory in single GGC run: 19336k
    Garbage: 599427k
    Leak: 55582k
    Overhead: 67698k
    GGC runs: 524

comparing Gerald's testcase PR8361 compilation at -O2 level:
    Overall memory needed: 105544k
    Peak memory use before GGC: 94274k
    Peak memory use after GGC: 93339k
    Maximum of released memory in single GGC run: 19336k
    Garbage: 681537k
    Leak: 56195k
    Overhead: 78768k
    GGC runs: 587

comparing Gerald's testcase PR8361 compilation at -O3 level:
    Overall memory needed: 108356k
    Peak memory use before GGC: 95464k
    Peak memory use after GGC: 94518k
    Maximum of released memory in single GGC run: 19608k
    Garbage: 727612k
    Leak: 56743k
    Overhead: 82371k
    GGC runs: 596

Head of the ChangeLog is:

--- /usr/src/SpecTests/sandbox-britten-memory/x86_64/mem-result/ChangeLog	2005-07-27 21:42:32.000000000 +0000
+++ /usr/src/SpecTests/sandbox-britten-memory/gcc/gcc/ChangeLog	2005-07-28 07:08:37.000000000 +0000
@@ -1,3 +1,108 @@
+2005-07-28 Jan Beulich <jbeulich@novell.com>
+
+	* config/ia64/ia64.c (ia64_load_pair_ok): New.
+	(ia64_print_operand): Describe and handle 'X'.
+	(ia64_register_move_cost): Also handle FP_REGS.
+	(ia64_preferred_reload_class): Likewise.
+	(ia64_secondary_reload_class): Likewise.
+	(ia64_dependencies_evaluation_hook): New local variable c. Initialize
+	it. Also check for ITANIUM_CLASS_FLDP.
+	* config/ia64/ia64.h (FP_REGNO_P): New.
+	(HARD_REGNO_MODE_OK): Remove explusion of TImode.
+	(reg_class): Add FP_REGS.
+	(REG_CLASS_NAMES): Adjust for it.
+	(REG_CLASS_CONTENTS): Likewise.
+	(REGNO_REG_CLASS): Use FP_REGS where appropriate.
+	(REG_CLASS_FROM_LETTER): Handle 'x'.
+	(CLASS_MAX_NREGS): Handle FP_REGS.
+	(MEMORY_MOVE_COST): Likewise.
+	* config/ia64/ia64.md (itanium_class): Add fldp.
+	(type): Handle fldp.
+	(movti_internal): More allowable operand combinations. Use ldfp8 when
+	splitting unnecessary. Remove predicable attribute. Adjust
+	itanium_class attribute.
+	(smuldi3_highpart): Remove outdated comment.
+	(mulditi3, umulditi3, rotlti3): New expanders.
+	(addti3, subti3, mulditi3_internal, umulditi3_internal, negti2, rotlti3_internal): New insns.
+	(absti2): Disabled new insn for future reference.
+	Respective new splitters.
+	* config/ia64/itanium1.md (1_fldp, 1b_fldp): New insn reservations.
+	* config/ia64/itanium2.md (2_fldp, 2b_fldp): Likewise.
+	* config/ia64/ia64-protos.h (ia64_load_pair_ok): New.
+
+2005-07-25  James A. Morrison  <phython@gcc.gnu.org>
+
+	PR rtl-optimization/23047
+	* simplify-rtx.c (simplify_const_relational_operation): Respect
+	flag_wrapv for comparisons with ABS.
+
+2005-07-27  James A. Morrison  <phython@gcc.gnu.org>
+
+	PR tree-optimization/22493
+	* tree-vrp.c (extract_range_from_unary_expr): Deal with -fwrapv and
+	VR_ANTI_RANGEs properly for NEGATE_EXPRs and ABS_EXPRs.
+
+2005-07-27  Aldy Hernandez  <aldyh@redhat.com>
+
+	* config/frv/frv.opt (moptimize-membar): New.
+
+	* doc/invoke.texi: Document -moptimize-membar and its inverse.
+
+	* config/frv/frv.h: Remove machine_function definition.
+
+        * config/frv/frv.c (struct frv_io): New.
+        (struct machine_function): Moved from frv.h.  Add has_membar_p.
+        (frv_same_doubleword_p, frv_io_fixed_order_p, frv_io_union)
+        (frv_extract_membar, frv_io_check_address, frv_io_handle_set)
+        (frv_io_handle_use_1, frv_io_handle_use, frv_optimize_membar_local)
+        (frv_optimize_membar_global, frv_optimize_membar): New functions.
+        (frv_reorg): Call frv_optimize_membar when appropriate.
+        (bdesc_loads, bdesc_stores): Use the membar code as the icode field.
+        (frv_expand_builtin): Adjust calls accordingly.
+        (frv_io_address_cookie): New function.
+        (frv_expand_load_builtin, frv_expand_store_builtin): Emit a normal
+        load or store rather than a special insn.  Add ccnstant address and
+        io-type operands to the membar.
+	(frv_ifcvt_modify_tests): Unsign regno.
+	(frv_ifcvt_modify_tests): Same.
+
+	* config/frv/frv.md: Remove UNSPEC_BUILTIN_{LOAD,STORE}.  Change
+	UNSPEC_OPTIONAL_MEMBAR constant.
+	(builtin_read_<mode>): Delete.
+	(builtin_write_<mode>): Delete.
+	("optional_membar_<mode>"): Add operand.
+
+	* testsuite/gcc.target/frv/all-builtin-read8.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-read16.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-read32.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-read64.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-write8.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-write16.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-write32.c: Delete.
+	* testsuite/gcc.target/frv/all-builtin-write64.c: Delete.
+	* testsuite/gcc.target/frv/all-read-write-1.c: New.
+
+2005-07-28  Kaz Kojima  <kkojima@gcc.gnu.org>
+
+	* df.c (df_uses_record): Handle SCRATCH.
+
+2005-07-28  Steven Bosscher  <stevenb@suse.de>
+
+	PR debug/20161
+	* passes.c (rest_of_decl_compilation): If decl is a type and
+	we have encountered errors, don't emit debug information.
+
+2005-07-27  Kenneth Zadeck <zadeck@naturalbridge.com>
+
+	* params.def: Fixed comment.
+	
+2005-07-27  Bjoern Haase  <bjoern.m.haase@web.de>
+
+	PR target/19885
+	* config/avr/avr.c (TARGET_ASM_ALIGNED_SI_OP): Add.
+	(TARGET_ASM_UNALIGNED_HI_OP): Add.
+	(TARGET_ASM_UNALIGNED_SI_OP): Add.
+
 2005-07-27  Steven Bosscher  <stevenb@suse.de>
 
 	PR c++/22003


The results can be reproduced by building a 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.

Your testing script.