A recent patch increased GCC's memory consumption in some cases!

gcctest@suse.de gcctest@suse.de
Tue Nov 11 01:09:00 GMT 2008


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 empty function compilation at -O0 level:
    Overall memory needed: 8183k -> 8208k
    Peak memory use before GGC: 1291k
    Peak memory use after GGC: 1217k
    Maximum of released memory in single GGC run: 134k
    Garbage: 218k
    Leak: 1221k
    Overhead: 136k
    GGC runs: 4
    Pre-IPA-Garbage: 207k
    Pre-IPA-Leak: 1224k
    Pre-IPA-Overhead: 135k
    Post-IPA-Garbage: 207k
    Post-IPA-Leak: 1224k
    Post-IPA-Overhead: 135k

comparing empty function compilation at -O0 -g level:
    Overall memory needed: 8427k -> 8452k
    Peak memory use before GGC: 1319k
    Peak memory use after GGC: 1245k
    Maximum of released memory in single GGC run: 133k
    Garbage: 220k
    Leak: 1254k
    Overhead: 141k
    GGC runs: 4
    Pre-IPA-Garbage: 207k
    Pre-IPA-Leak: 1224k
    Pre-IPA-Overhead: 135k
    Post-IPA-Garbage: 207k
    Post-IPA-Leak: 1224k
    Post-IPA-Overhead: 135k

comparing empty function compilation at -O1 level:
    Overall memory needed: 8239k -> 8236k
    Peak memory use before GGC: 1291k
    Peak memory use after GGC: 1217k
    Maximum of released memory in single GGC run: 134k
    Garbage: 221k
    Leak: 1221k
    Overhead: 137k
    GGC runs: 4
    Pre-IPA-Garbage: 207k
    Pre-IPA-Leak: 1224k
    Pre-IPA-Overhead: 135k
    Post-IPA-Garbage: 207k
    Post-IPA-Leak: 1224k
    Post-IPA-Overhead: 135k

comparing empty function compilation at -O2 level:
    Overall memory needed: 8451k -> 8452k
    Peak memory use before GGC: 1291k
    Peak memory use after GGC: 1218k
    Maximum of released memory in single GGC run: 135k
    Garbage: 226k
    Leak: 1221k
    Overhead: 138k
    GGC runs: 4
    Pre-IPA-Garbage: 207k
    Pre-IPA-Leak: 1224k
    Pre-IPA-Overhead: 135k
    Post-IPA-Garbage: 207k
    Post-IPA-Leak: 1224k
    Post-IPA-Overhead: 135k

comparing empty function compilation at -O3 level:
    Overall memory needed: 8455k -> 8456k
    Peak memory use before GGC: 1291k
    Peak memory use after GGC: 1218k
    Maximum of released memory in single GGC run: 135k
    Garbage: 226k
    Leak: 1221k
    Overhead: 138k
    GGC runs: 4
    Pre-IPA-Garbage: 207k
    Pre-IPA-Leak: 1224k
    Pre-IPA-Overhead: 135k
    Post-IPA-Garbage: 207k
    Post-IPA-Leak: 1224k
    Post-IPA-Overhead: 135k

comparing combine.c compilation at -O0 level:
    Overall memory needed: 32039k -> 31988k
    Peak memory use before GGC: 18016k -> 18018k
    Peak memory use after GGC: 17797k -> 17801k
    Maximum of released memory in single GGC run: 1840k -> 1839k
    Garbage: 39485k -> 39406k
    Leak: 5800k -> 5800k
    Overhead: 5232k -> 5220k
    GGC runs: 337
    Pre-IPA-Garbage: 12405k -> 12408k
    Pre-IPA-Leak: 19346k -> 19349k
    Pre-IPA-Overhead: 2558k -> 2559k
    Post-IPA-Garbage: 12405k -> 12408k
    Post-IPA-Leak: 19346k -> 19349k
    Post-IPA-Overhead: 2558k -> 2559k

comparing combine.c compilation at -O0 -g level:
    Overall memory needed: 34063k -> 34008k
    Peak memory use before GGC: 20016k -> 19938k
    Peak memory use after GGC: 19653k -> 19657k
    Maximum of released memory in single GGC run: 1850k -> 1849k
    Garbage: 39790k -> 39703k
    Leak: 9089k -> 9082k
    Overhead: 6050k -> 6038k
    GGC runs: 321
    Pre-IPA-Garbage: 12504k -> 12507k
    Pre-IPA-Leak: 21619k -> 21623k
    Pre-IPA-Overhead: 3049k -> 3049k
    Post-IPA-Garbage: 12504k -> 12507k
    Post-IPA-Leak: 21619k -> 21623k
    Post-IPA-Overhead: 3049k -> 3049k

comparing combine.c compilation at -O1 level:
  Amount of produced GGC garbage increased from 46626k to 46801k, overall 0.38%
    Overall memory needed: 30751k -> 30752k
    Peak memory use before GGC: 15675k -> 15682k
    Peak memory use after GGC: 15504k -> 15507k
    Maximum of released memory in single GGC run: 1380k -> 1340k
    Garbage: 46626k -> 46801k
    Leak: 5775k -> 5780k
    Overhead: 5975k -> 6014k
    GGC runs: 401 -> 402
    Pre-IPA-Garbage: 13144k -> 13147k
    Pre-IPA-Leak: 16841k -> 16845k
    Pre-IPA-Overhead: 2471k -> 2472k
    Post-IPA-Garbage: 13144k -> 13147k
    Post-IPA-Leak: 16841k -> 16845k
    Post-IPA-Overhead: 2471k -> 2472k

comparing combine.c compilation at -O2 level:
    Overall memory needed: 31403k -> 31340k
    Peak memory use before GGC: 15820k -> 15823k
    Peak memory use after GGC: 15657k -> 15661k
    Maximum of released memory in single GGC run: 1356k -> 1355k
    Garbage: 60688k -> 60528k
    Leak: 5950k -> 5809k
    Overhead: 8013k -> 8023k
    GGC runs: 471 -> 469
    Pre-IPA-Garbage: 13307k -> 13310k
    Pre-IPA-Leak: 16923k -> 16927k
    Pre-IPA-Overhead: 2491k -> 2491k
    Post-IPA-Garbage: 13307k -> 13310k
    Post-IPA-Leak: 16923k -> 16927k
    Post-IPA-Overhead: 2491k -> 2491k

comparing combine.c compilation at -O3 level:
  Peak amount of GGC memory allocated before garbage collecting increased from 15819k to 15911k, overall 0.58%
  Peak amount of GGC memory still allocated after garbage collecting increased from 15658k to 15752k, overall 0.60%
  Amount of produced GGC garbage decreased from 76176k to 72955k, overall -4.41%
  Amount of memory still referenced at the end of compilation increased from 5822k to 7136k, overall 22.55%
    Overall memory needed: 31935k -> 31968k
    Peak memory use before GGC: 15819k -> 15911k
    Peak memory use after GGC: 15658k -> 15752k
    Maximum of released memory in single GGC run: 1809k -> 1629k
    Garbage: 76176k -> 72955k
    Leak: 5822k -> 7136k
    Overhead: 9792k -> 9459k
    GGC runs: 506 -> 496
    Pre-IPA-Garbage: 13307k -> 13310k
    Pre-IPA-Leak: 16923k -> 16927k
    Pre-IPA-Overhead: 2491k -> 2491k
    Post-IPA-Garbage: 13307k -> 13310k
    Post-IPA-Leak: 16923k -> 16927k
    Post-IPA-Overhead: 2491k -> 2491k

comparing insn-attrtab.c compilation at -O0 level:
    Overall memory needed: 156827k -> 155236k
    Peak memory use before GGC: 65231k -> 65230k
    Peak memory use after GGC: 53275k
    Maximum of released memory in single GGC run: 27425k -> 27424k
    Garbage: 130451k -> 130437k
    Leak: 8497k
    Overhead: 15725k -> 15723k
    GGC runs: 263
    Pre-IPA-Garbage: 38214k -> 38215k
    Pre-IPA-Leak: 55487k -> 55487k
    Pre-IPA-Overhead: 8223k -> 8223k
    Post-IPA-Garbage: 38214k -> 38215k
    Post-IPA-Leak: 55487k -> 55487k
    Post-IPA-Overhead: 8223k -> 8223k

comparing insn-attrtab.c compilation at -O0 -g level:
    Overall memory needed: 158099k -> 156508k
    Peak memory use before GGC: 66505k -> 66504k
    Peak memory use after GGC: 54547k -> 54546k
    Maximum of released memory in single GGC run: 27426k -> 27425k
    Garbage: 130929k -> 130915k
    Leak: 10147k
    Overhead: 16180k -> 16179k
    GGC runs: 255
    Pre-IPA-Garbage: 38271k -> 38272k
    Pre-IPA-Leak: 57029k -> 57029k
    Pre-IPA-Overhead: 8558k -> 8558k
    Post-IPA-Garbage: 38271k -> 38272k
    Post-IPA-Leak: 57029k -> 57029k
    Post-IPA-Overhead: 8558k -> 8558k

comparing insn-attrtab.c compilation at -O1 level:
    Overall memory needed: 134911k -> 133328k
    Peak memory use before GGC: 50202k -> 50200k
    Peak memory use after GGC: 43295k -> 43294k
    Maximum of released memory in single GGC run: 22951k
    Garbage: 180962k -> 180978k
    Leak: 7875k -> 7873k
    Overhead: 24525k -> 24528k
    GGC runs: 302 -> 301
    Pre-IPA-Garbage: 43191k -> 43193k
    Pre-IPA-Leak: 43086k -> 43086k
    Pre-IPA-Overhead: 7642k -> 7642k
    Post-IPA-Garbage: 43191k -> 43193k
    Post-IPA-Leak: 43086k -> 43086k
    Post-IPA-Overhead: 7642k -> 7642k

comparing insn-attrtab.c compilation at -O2 level:
  Peak amount of GGC memory still allocated after garbage collecting increased from 44284k to 45013k, overall 1.65%
  Amount of produced GGC garbage decreased from 212172k to 204660k, overall -3.67%
  Amount of memory still referenced at the end of compilation increased from 7882k to 15536k, overall 97.11%
    Overall memory needed: 149807k -> 148824k
    Peak memory use before GGC: 50205k -> 50204k
    Peak memory use after GGC: 44284k -> 45013k
    Maximum of released memory in single GGC run: 21469k -> 17966k
    Garbage: 212172k -> 204660k
    Leak: 7882k -> 15536k
    Overhead: 29998k -> 30014k
    GGC runs: 333 -> 326
    Pre-IPA-Garbage: 43263k -> 43265k
    Pre-IPA-Leak: 43092k -> 43092k
    Pre-IPA-Overhead: 7651k -> 7651k
    Post-IPA-Garbage: 43263k -> 43265k
    Post-IPA-Leak: 43092k -> 43092k
    Post-IPA-Overhead: 7651k -> 7651k

comparing insn-attrtab.c compilation at -O3 level:
  Amount of memory still referenced at the end of compilation increased from 7889k to 7899k, overall 0.12%
    Overall memory needed: 164031k -> 164772k
    Peak memory use before GGC: 61823k -> 61826k
    Peak memory use after GGC: 58723k -> 58725k
    Maximum of released memory in single GGC run: 23616k -> 23617k
    Garbage: 242406k -> 242519k
    Leak: 7889k -> 7899k
    Overhead: 33439k -> 33464k
    GGC runs: 339
    Pre-IPA-Garbage: 43263k -> 43265k
    Pre-IPA-Leak: 43092k -> 43092k
    Pre-IPA-Overhead: 7651k -> 7651k
    Post-IPA-Garbage: 43263k -> 43265k
    Post-IPA-Leak: 43092k -> 43092k
    Post-IPA-Overhead: 7651k -> 7651k

comparing Gerald's testcase PR8361 compilation at -O0 level:
    Overall memory needed: 151284k -> 151215k
    Peak memory use before GGC: 82965k -> 82970k
    Peak memory use after GGC: 82143k -> 82148k
    Maximum of released memory in single GGC run: 14701k -> 14702k
    Garbage: 205364k -> 205239k
    Leak: 52055k -> 52062k
    Overhead: 26934k -> 26925k
    GGC runs: 415
    Pre-IPA-Garbage: 111126k -> 111129k
    Pre-IPA-Leak: 88384k -> 88390k
    Pre-IPA-Overhead: 14817k -> 14820k
    Post-IPA-Garbage: 111126k -> 111129k
    Post-IPA-Leak: 88384k -> 88390k
    Post-IPA-Overhead: 14817k -> 14820k

comparing Gerald's testcase PR8361 compilation at -O0 -g level:
    Overall memory needed: 169216k -> 169155k
    Peak memory use before GGC: 96591k -> 96595k
    Peak memory use after GGC: 95634k -> 95639k
    Maximum of released memory in single GGC run: 15129k -> 15130k
    Garbage: 210983k -> 210902k
    Leak: 78617k -> 78624k
    Overhead: 33598k -> 33589k
    GGC runs: 387
    Pre-IPA-Garbage: 111748k -> 111751k
    Pre-IPA-Leak: 104899k -> 104905k
    Pre-IPA-Overhead: 18320k -> 18323k
    Post-IPA-Garbage: 111748k -> 111751k
    Post-IPA-Leak: 104899k -> 104905k
    Post-IPA-Overhead: 18320k -> 18323k

comparing Gerald's testcase PR8361 compilation at -O1 level:
    Overall memory needed: 111196k -> 111223k
    Peak memory use before GGC: 84209k -> 84225k
    Peak memory use after GGC: 83371k -> 83386k
    Maximum of released memory in single GGC run: 14981k -> 14982k
    Garbage: 282253k -> 282306k
    Leak: 49359k -> 49364k
    Overhead: 31648k -> 31646k
    GGC runs: 502 -> 503
    Pre-IPA-Garbage: 159782k -> 159802k
    Pre-IPA-Leak: 88068k -> 88086k
    Pre-IPA-Overhead: 19881k -> 19886k
    Post-IPA-Garbage: 159782k -> 159802k
    Post-IPA-Leak: 88068k -> 88086k
    Post-IPA-Overhead: 19881k -> 19886k

comparing Gerald's testcase PR8361 compilation at -O2 level:
  Amount of produced GGC garbage increased from 336506k to 337301k, overall 0.24%
    Overall memory needed: 112692k -> 112647k
    Peak memory use before GGC: 85982k -> 85989k
    Peak memory use after GGC: 85115k -> 85136k
    Maximum of released memory in single GGC run: 14965k
    Garbage: 336506k -> 337301k
    Leak: 49437k -> 49420k
    Overhead: 38151k -> 38215k
    GGC runs: 568 -> 570
    Pre-IPA-Garbage: 163775k -> 163792k
    Pre-IPA-Leak: 88430k -> 88452k
    Pre-IPA-Overhead: 20357k -> 20364k
    Post-IPA-Garbage: 163775k -> 163792k
    Post-IPA-Leak: 88430k -> 88452k
    Post-IPA-Overhead: 20357k -> 20364k

comparing Gerald's testcase PR8361 compilation at -O3 level:
  Amount of produced GGC garbage increased from 367606k to 368780k, overall 0.32%
    Overall memory needed: 113768k -> 113715k
    Peak memory use before GGC: 86670k -> 86611k
    Peak memory use after GGC: 85729k -> 85750k
    Maximum of released memory in single GGC run: 14965k
    Garbage: 367606k -> 368780k
    Leak: 49424k -> 49422k
    Overhead: 41427k -> 41456k
    GGC runs: 597 -> 601
    Pre-IPA-Garbage: 163855k -> 163872k
    Pre-IPA-Leak: 89088k -> 89110k
    Pre-IPA-Overhead: 20409k -> 20416k
    Post-IPA-Garbage: 163855k -> 163872k
    Post-IPA-Leak: 89088k -> 89110k
    Post-IPA-Overhead: 20409k -> 20416k

comparing PR rtl-optimization/28071 testcase compilation at -O0 level:
    Overall memory needed: 365547k -> 361915k
    Peak memory use before GGC: 78518k
    Peak memory use after GGC: 49452k -> 49453k
    Maximum of released memory in single GGC run: 38186k
    Garbage: 144652k -> 144651k
    Leak: 7110k -> 7110k
    Overhead: 24889k -> 24889k
    GGC runs: 87
    Pre-IPA-Garbage: 12561k
    Pre-IPA-Leak: 20190k
    Pre-IPA-Overhead: 2241k
    Post-IPA-Garbage: 12561k
    Post-IPA-Leak: 20190k
    Post-IPA-Overhead: 2241k

comparing PR rtl-optimization/28071 testcase compilation at -O0 -g level:
    Overall memory needed: 366343k -> 362703k
    Peak memory use before GGC: 79215k
    Peak memory use after GGC: 50149k
    Maximum of released memory in single GGC run: 38171k -> 38170k
    Garbage: 144753k -> 144752k
    Leak: 9152k -> 9152k
    Overhead: 25473k -> 25473k
    GGC runs: 93
    Pre-IPA-Garbage: 12569k
    Pre-IPA-Leak: 20439k
    Pre-IPA-Overhead: 2295k
    Post-IPA-Garbage: 12569k
    Post-IPA-Leak: 20439k
    Post-IPA-Overhead: 2295k

comparing PR rtl-optimization/28071 testcase compilation at -O1 level:
  Amount of memory still referenced at the end of compilation decreased from 7807k to 7551k, overall -3.39%
    Overall memory needed: 245251k -> 242332k
    Peak memory use before GGC: 73624k
    Peak memory use after GGC: 66143k
    Maximum of released memory in single GGC run: 34721k -> 34735k
    Garbage: 222457k -> 221628k
    Leak: 7807k -> 7551k
    Overhead: 30583k -> 30480k
    GGC runs: 97
    Pre-IPA-Garbage: 48348k
    Pre-IPA-Leak: 63005k
    Pre-IPA-Overhead: 8797k
    Post-IPA-Garbage: 48348k
    Post-IPA-Leak: 63005k
    Post-IPA-Overhead: 8797k

comparing PR rtl-optimization/28071 testcase compilation at -O2 level:
  Amount of produced GGC garbage increased from 250446k to 251372k, overall 0.37%
  Amount of memory still referenced at the end of compilation decreased from 7901k to 7553k, overall -4.61%
    Overall memory needed: 372135k -> 369256k
    Peak memory use before GGC: 73624k
    Peak memory use after GGC: 66143k
    Maximum of released memory in single GGC run: 36044k -> 36061k
    Garbage: 250446k -> 251372k
    Leak: 7901k -> 7553k
    Overhead: 36662k -> 36845k
    GGC runs: 106
    Pre-IPA-Garbage: 107058k
    Pre-IPA-Leak: 75901k -> 75901k
    Pre-IPA-Overhead: 14919k -> 14919k
    Post-IPA-Garbage: 107058k
    Post-IPA-Leak: 75901k -> 75901k
    Post-IPA-Overhead: 14919k -> 14919k

comparing PR rtl-optimization/28071 testcase compilation at -O3 -fno-tree-pre -fno-tree-fre level:
  Amount of produced GGC garbage increased from 362325k to 365075k, overall 0.76%
  Amount of memory still referenced at the end of compilation decreased from 9383k to 9099k, overall -3.12%
    Overall memory needed: 1026651k -> 1026608k
    Peak memory use before GGC: 141898k
    Peak memory use after GGC: 129175k
    Maximum of released memory in single GGC run: 62686k -> 62763k
    Garbage: 362325k -> 365075k
    Leak: 9383k -> 9099k
    Overhead: 44297k -> 44856k
    GGC runs: 103
    Pre-IPA-Garbage: 107058k
    Pre-IPA-Leak: 75901k -> 75901k
    Pre-IPA-Overhead: 14919k -> 14919k
    Post-IPA-Garbage: 107058k
    Post-IPA-Leak: 75901k -> 75901k
    Post-IPA-Overhead: 14919k -> 14919k

Head of the ChangeLog is:

--- /usr/src/SpecTests/sandbox-britten-memory/x86_64/mem-result/ChangeLog	2008-10-09 16:51:44.000000000 +0000
+++ /usr/src/SpecTests/sandbox-britten-memory/gcc/gcc/ChangeLog	2008-11-10 20:51:37.000000000 +0000
@@ -1,5 +1,1632 @@
+2008-11-10  Kevin Buettner  <kevinb@redhat.com>
+
+	* config/m32c/prologue.md (prologue_enter_16): Set FB to SP - 2.
+	(prologue_enter_32): Set FB to SP - 4.
+
+2008-11-10  DJ Delorie  <dj@redhat.com>
+
+	* config/m32c/m32c.c (m32c_const_ok_for_constraint_p): Add ImB
+	constraint for single-bit-clear in lower byte of HI constant, vs
+	Imb which just ignores the upper byte.
+	* config/m32c/predicates.md (m32c_1mask8_operand): Use it.
+	* config/m32c/bitops.md (andhi3_16, andhi3_24): Use it.
+	(iorhi3_16): Check for single bit set, not single bit clear.
+
+2008-11-10  Janis Johnson  <janis187@us.ibm.com>
+
+	* doc/sourcebuild.texi (Torture Tests): Fix formatting for bullets.
+
+	* doc/sourcebuild.texi (Torture Tests): Fix formatting in example.
+
+2008-11-10  Eric Botcazou  <ebotcazou@adacore.com>
+
+	* calls.c (store_unaligned_arguments_into_pseudos): Deal only with
+	values living in memory and use more precise alignment information.
+
+2008-11-10  Jakub Jelinek  <jakub@redhat.com>
+
+	PR middle-end/35314
+	* tree-mudflap.c (mf_build_check_statement_for): Split then_block
+	after __mf_check call if the call must end a bb.
+
+2008-11-10  Ralph Loader  <suckfish@ihug.co.nz>
+
+	PR middle-end/37807
+	PR middle-end/37809
+	* combine.c (force_to_mode): Do not process vector types.
+
+	* rtlanal.c (nonzero_bits1): Do not process vector types.
+	(num_sign_bit_copies1): Likewise.
+
+2008-11-09  Thomas Schwinge  <tschwinge@gnu.org>
+
+	* config/i386/gnu.h: Add copyright and licensing header.
+	* config/gnu.h: Likewise.
+
+2008-11-07  Andrew Pinski  <andrew_pinski@playstation.sony.com>
+
+	* fwprop.c (fwprop_done): Call df_remove_problem.
+
+2008-11-07  Sebastian Pop  <sebastian.pop@amd.com>
+
+	PR middle-end/37379
+	* tree-scalar-evolution.c (instantiate_scev_1): Return
+	chrec_dont_know for VL_EXP_CLASS_P.
+	
+2008-11-06  Uros Bizjak  <ubizjak@gmail.com>
+
+	* reg-stack.c (reg_to_stack): Add missing braces in memset
+	size calculation.
+
+2008-11-06  Andrew Stubbs  <ams@codesourcery.com>
+
+	* config/arm/bpabi.h (PROFILE_HOOK): New undef.
+	(NO_PROFILE_COUNTERS): New define.
+
+2008-11-06  David Edelsohn  <edelsohn@gnu.org>
+
+	PR target/26397
+	* config/rs6000/aix.h (LIBSTDCXX_STATIC): Define.
+	
+2008-11-06  Kazu Hirata  <kazu@codesourcery.com>
+
+	PR target/35574
+	* config/sparc/constraints.md (D): New.
+	* config/sparc/predicates.md (const_double_or_vector_operand): New.
+	* config/sparc/sparc.c (sparc_extra_constraint_check): Handle the
+	'D' constraint.
+	* config/sparc/sparc.md (*movdf_insn_sp32_v9, *movdf_insn_sp64):
+	Use the 'D' constraint in addition to 'F' in some alternatives.
+	(DF splitter): Generalize for V64mode.
+	* doc/md.texi (SPARC): Document the 'D' constraint.
+
+2008-11-06  Uros Bizjak  <ubizjak@gmail.com>
+
+	* reg-stack.c (reg_to_stack): Generate +QNaN using real_nan.
+
+2008-11-06  Ben Elliston  <bje@au.ibm.com>
+
+	* gcc.c: Remove ancient comment about a bug in Sun cc.
+
+2008-11-05  Janis Johnson  <janis187@us.ibm.com>
+
+	* doc/sourcebuild.texi (Torture Tests): Add ADDITIONAL_TORTURE_OPTIONS.
+
+2008-11-06  Ben Elliston  <bje@au.ibm.com>	
+
+	* config/rs6000/rs6000.c (rs6000_override_options): Comment fix.
+
+2008-11-06  Ben Elliston  <bje@au.ibm.com>
+
+	* config/rs6000/rs6000.opt (mwarn-cell-microcode): Improve option
+	description.
+
+2008-11-05  Hans-Peter Nilsson  <hp@axis.com>
+
+	PR target/38016
+	* config/cris/cris.c (cris_order_for_addsi3): Test for !REG_P, not
+	just MEM_P.
+
+2008-11-05  Martin Jambor  <mjambor@suse.cz>
+
+	PR middle-end/37861
+	* tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Don't turn
+	pointer arithmetics into array_ref if the array is accessed
+	through an indirect_ref.
+
+2008-11-05  Richard Guenther  <rguenther@suse.de>
+
+	PR middle-end/37742
+	* tree-ssa.c (useless_type_conversion_p_1): Check different restrict
+	qualified pointer conversion before stripping qualifiers.
+	* gimplify.c (create_tmp_from_val): Use correctly qualified type.
+	* tree-flow.h (may_propagate_address_into_dereference): Declare.
+	* tree-ssa-ccp.c (may_propagate_address_into_dereference): New
+	function.
+	(ccp_fold): Use it.
+	* tree-ssa-forwprop.c (rhs_to_tree): Remove useless conversions,
+	properly canonicalize binary ops.
+	(forward_propagate_addr_expr_1): Use
+	may_propagate_address_into_dereference.
+
+2008-11-05  Uros Bizjak  <ubizjak@gmail.com>
+
+	PR middle-end/37286
+	* reg-stack.c (subst_stack_regs_pat) [MINUS, DIV, MULT, PLUS]:
+	Initialize uninitialized input registers with a NaN.
+
+2008-11-05  Tobias Grosser  <grosser@fim.uni-passau.de>
+
+	PR middle-end/37833
+
+	* graphite.c (scan_tree_for_params): Add POINTER_PLUS_EXPR.
+
+2008-11-05  Tobias Grosser  <grosser@fim.uni-passau.de>
+
+	PR middle-end/37943
+
+	* graphite.c (scopdet_basic_block_info): Fix loops with multiple
+	exits and conditions.
+	* testsuite/gcc.dg/graphite/pr37943.c: New.
+
+2008-11-04  Jakub Jelinek  <jakub@redhat.com>
+	    Andrew Pinski  <andrew_pinski@playstation.sony.com>
+
+	PR c/35430
+	* c-common.c (warn_for_sign_compare): For complex result_type
+	use component's type.
+
+2008-11-04  Jakub Jelinek  <jakuB@redhat.com>
+
+	* print-tree.c (print_node): Use code instead of TREE_CODE (node).
+
+	* print-tree.c (print_node): Don't print DECL_INITIAL
+	for PARM_DECL.
+
+	PR c/37106
+	* c-common.c (parse_optimize_options): Save and restore
+	flag_strict_aliasing around decode_options call.
+
+2008-11-04  Uros Bizjak  <ubizjak@gmail.com>
+
+	* config/i386/driver-i386.c (enum vendor_signatures): New enum.
+	(host_detect_local_cpu): Use it instead of casted strings to
+	compare vendor signatures.
+
+2008-11-03  Mikael Pettersson  <mikpe@it.uu.se>
+
+	PR target/37989
+	* config/i386/mingw32.h (REAL_LIBGCC_SPEC): Only add libgcc_s.a
+	or libgcc_eh.a to spec if ENABLE_SHARED_LIBGCC.
+
+2008-11-03  Catherine Moore  <clm@codesourcery.com>
+
+	* config/mips.c (mips_conditional_register_usage):  Handle the
+	DSP control register.
+	* doc/extend.texi: Document the DSP control register.
+
+2008-11-03  Steve Ellcey <sje@cup.hp.com>
+	    Jakub Jelinek  <jakub@redhat.com>
+
+	* omp-low.c (expand_omp_for_generic): If iter_type has different
+	precision than type and type is a pointer type, cast n1 and n2
+	first to an integer type with the same precision as pointers
+	and only afterwards to iter_type.
+
+2008-11-03  Richard Sandiford  <rdsandiford@googlemail.com>
+
+	* config/arm/arm.md (UNSPEC_PIC_BASE): Update documentation.
+	(UNSPEC_GOTSYM_OFF): New unspec.
+	(pic_add_dot_plus_four): Simplify the UNSPEC_PIC_BASE operands.
+	(pic_add_dot_plus_eight): Likewise.
+	(tls_load_dot_plus_eight): Likewise.  Update peephole2.
+	* config/arm/thumb2.md (pic_load_dot_plus_four): Simplify the
+	UNSPEC_PIC_BASE operands.
+	* config/arm/arm.c (arm_load_pic_register): Use UNSPEC_GOTSYM_OFF.
+	(arm_output_addr_const_extra): Handle it.
+
+2008-11-03  Eric Botcazou  <ebotcazou@adacore.com>
+
+	* config/i386/i386.c (classify_argument) <ARRAY_TYPE>: Promote partial
+	integer class to full integer class if the offset is not word-aligned.
+
+2008-11-03  Andrew Pinski  <andrew_pinski@playstation.sony.com>
+
+	PR rtl-opt/37782
+	* loop-doloop.c (doloop_modify): Add from_mode argument that says what
+	mode count is in.
+	(doloop_optimize): Update call to doloop_modify.
+
+2008-11-03  Eric Botcazou  <ebotcazou@adacore.com>
+
+	* tree-sra.c (bitfield_overlaps_p): Fix oversight.
+
+2008-11-03  Rainer Orth  <ro@TechFak.Uni-Bielefeld.DE>
+
+	PR other/37463
+	* configure.ac (gcc_cv_ld_ro_rw_mix): Move before
+	gcc_cv_as_cfi_directive.
+	(gcc_cv_as_cfi_directive) [*-*-solaris*]: Check if linker supports
+	merging read-only and read-write sections or assembler emits
+	read-write .eh_frame sections.
+	* configure: Regenerate.
+	
+2008-11-03  Nathan Froyd  <froydnj@codesourcery.com>
+
+	Revert:
+	2008-10-31  Nathan Froyd  <froydnj@codesourcery.com>
+
+	* config/rs6000/rs6000.c (no_global_regs_above): Fix precedence
+	problem.
+	(rs6000_emit_prologue): Invert logic.
+	* config/rs6000/rs6000.md (*save_gpregs_<mode>): Use explicit
+	(reg:P 11) instead of match_operand.
+	(*save_fpregs_<mode>): Likewise.
+	(*restore_gpregs_<mode>): Likewise.
+	(*return_and_restore_gpregs_<mode>): Likewise.
+	(*return_and_restore_fpregs_<mode>): Likewise.
+	* config/rs6000/spe.md (*save_gpregs_spe): Use explicit
+	(reg:P 11) insted of match_operand.
+	(*restore_gpregs_spe): Likewise.
+	(*return_and_restore_gpregs_spe): Likewise.
+
+2008-11-03  Harsha Jagasia  <harsha.jagasia@amd.com>
+
+	PR tree-optimization/37684
+	* gcc.dg/graphite/pr37684.c: New.
+	* graphite.c (exclude_component_ref): New.
+	(is_simple_operand): Call exclude_component_ref.
+
+2008-11-03  Sebastian Pop  <sebastian.pop@amd.com>
+
+	PR tree-optimization/36908
+	* testsuite/gcc.dg/tree-ssa/pr36908.c: New.
+	* tree-loop-distribution.c (number_of_rw_in_rdg): New.
+	(number_of_rw_in_partition): New.
+	(partition_contains_all_rw): New.
+	(ldist_gen): Do not distribute when one of the partitions
+	contains all the memory operations.
+
+2008-11-03  Sebastian Pop  <sebastian.pop@amd.com>
+
+	* cfghooks.c (split_block): Set BB_IRREDUCIBLE_LOOP and
+	EDGE_IRREDUCIBLE_LOOP.
+
+2008-11-03  Bernd Schmidt  <bernd.schmidt@analog.com>
+
+	* config/bfin/bfin.c (bfin_optimize_loop): Properly handle case
+	where we have one entry point in the loop which isn't the head.
+
+2008-11-03  Richard Guenther  <rguenther@suse.de>
+
+	PR middle-end/37573
+	* tree-data-ref.c (split_constant_offset_1): Fix tuplification.
+
+2008-11-03  Jakub Jelinek  <jakub@redhat.com>
+
+	PR middle-end/37858
+	* passes.c (execute_one_pass): Don't look at cfun->curr_properties
+	for ipa and simple ipa passes.
+
+2008-11-02  Richard Guenther  <rguenther@suse.de>
+
+	PR tree-optimization/37542
+	* tree-ssa-pre.c (fully_constant_expression): Handle more cases.
+	* tree-ssa-sccvn.c (vn_get_expr_for): Fix typo.
+	(vn_nary_op_lookup_stmt): Adjust for unary reference trees.
+	(vn_nary_op_insert_stmt): Likewise.
+	(visit_use): Likewise.
+
+2008-11-02  Anatoly Sokolov  <aesok@post.ru>
+
+	* config/avr/avr.md (UNSPEC_SWAP): Remove constants.
+	(*swap): Remove.
+	(rotlqi3, *rotlqi3_4, rotlhi3, *rotlhi3_8,
+	rotlsi3, *rotlsi3_8, *rotlsi3_16, *rotlsi3_24): New patterns.
+	(ashlqi3_const4, ashlqi3_const5, ashlqi3_const6, ashlqi3_l_const4,
+	ashlqi3_l_const5, ashlqi3_l_const6, lshrqi3_const4, lshrqi3_const5,
+	lshrqi3_const6, lshrqi3_l_const4, lshrqi3_l_const4, lshrqi3_l_const6
+	peephole2's): Replace unspec with rotate.
+	* config/avr/avr.c (avr_rtx_costs): Add ROTATE.
+
+2008-11-02  Richard Guenther  <rguenther@suse.de>
+
+	PR tree-optimization/37991
+	* tree-ssa-sccvn.h (copy_vuses_from_stmt): Remove.
+	* tree-ssa-sccvn.c (copy_vuses_from_stmt): Make static.
+	(set_ssa_val_to): Print if the value changed.
+	(simplify_binary_expression): Strip useless conversions.
+
+2008-11-01  Hans-Peter Nilsson  <hp@axis.com>
+
+	PR target/37939
+	* config/cris/cris.c (cris_rtx_costs) <MULT>: Return 0 for an ADDI
+	operand.
+
+2008-11-01  Richard Guenther  <rguenther@suse.de>
+
+	PR middle-end/37976
+	* builtins.c (fold_builtin_strspn): Return a size_t.
+	(fold_builtin_strcspn): Likewise.
+
+2008-10-31  Nathan Froyd  <froydnj@codesourcery.com>
+
+	* config/rs6000/rs6000.c (rs6000_file_start): Output gnu
+	attribute for struct return convention.
+
+2008-10-31  Nathan Froyd  <froydnj@codesourcery.com>
+
+	* config/rs6000/crtsavres.asm: Really, really delete.
+
+2008-10-31  Nathan Froyd  <froydnj@codesourcery.com>
+
+	* config/rs6000/rs6000.c (no_global_regs_above): Fix precedence
+	problem.
+	(rs6000_emit_prologue): Invert logic.
+	* config/rs6000/rs6000.md (*save_gpregs_<mode>): Use explicit
+	(reg:P 11) instead of match_operand.
+	(*save_fpregs_<mode>): Likewise.
+	(*restore_gpregs_<mode>): Likewise.
+	(*return_and_restore_gpregs_<mode>): Likewise.
+	(*return_and_restore_fpregs_<mode>): Likewise.
+	* config/rs6000/spe.md (*save_gpregs_spe): Use explicit
+	(reg:P 11) insted of match_operand.
+	(*restore_gpregs_spe): Likewise.
+	(*return_and_restore_gpregs_spe): Likewise.
+
+2008-10-28  Luis Machado  <luisgpm@br.ibm.com>
+
+	* gcc.h (DEFAULT_WORD_SWITCH_TAKES_ARG): Add "dumpbase" to the
+	list of word switches that take args.
+
+2008-10-30  Jakub Jelinek  <jakub@redhat.com>
+
+	* config/s390/s390.c (s390_mark_symbol_ref_as_used): New function.
+	(s390_output_pool_entry): Call it through for_each_rtx.
+
+	PR middle-end/37730
+	* expr.c (store_constructor): For vectors, if target is a MEM, use
+	target's MEM_ALIAS_SET instead of elttype alias set.
+
+2008-10-29  Kaz Kojima  <kkojima@gcc.gnu.org>
+
+	PR target/37909
+	* config/sh/sh.c (untangle_mova): Return -1 when NEW_MOVA has
+	no address.
+
+2008-10-29  David Edelsohn  <edelsohn@gnu.org>
+
+	PR target/37878
+	* config/rs6000/predicates.md (word_offset_memref_operand):
+	Restructure code and look inside auto-inc/dec addresses.
+	
+2008-10-29  Steve Ellcey  <sje@cup.hp.com>
+
+	PR target/32277
+	* libgcov.c ( __gcov_indirect_call_profiler): Check
+	TARGET_VTABLE_USES_DESCRIPTORS.
+
+2008-10-29  Stefan Schulze Frielinghaus  <xxschulz@de.ibm.com>
+
+	* config/spu/spu.h (FRAME_GROWS_DOWNWARD): Define.
+	(INITIAL_FRAME_POINTER_OFFSET): Remove.
+	* config/spu/spu.c (spu_initial_elimination_offset): Calculate new
+	offset if eliminating soft frame pointer.
+	* config/spu/spu.md (stack_protect_set, stack_protect_test)
+	(stack_protect_test_si): Add initial machine description
+	for Stack Smashing Protector
+
+2008-10-29  Manuel Lopez-Ibanez  <manu@gcc.gnu.org>
+
+	PR 11492
+	* c-common.c (min_precision): Move to...
+	* tree.c (tree_int_cst_min_precision): ... to here. Renamed.
+	* tree.h (tree_int_cst_min_precision): Declare.
+	* c-common.h (min_precision): Delete declaration.
+	* fold-const.c (tree_binary_nonnegative_warnv_p): Handle
+	multiplication of non-negative integer constants.
+	* c-decl.c (check_bitfield_type_and_width): Rename min_precision to
+	tree_int_cst_min_precision.
+	(finish_enum): Likewise.
+
+2008-10-29  Joseph Myers  <joseph@codesourcery.com>
+
+	PR middle-end/36578
+	* convert.c (convert_to_real): Do not optimize conversions of
+	binary arithmetic operations between binary and decimal
+	floating-point types.  Consider mode of target type in determining
+	decimal type for arithmetic.  Unless
+	flag_unsafe_math_optimizations, do not optimize binary conversions
+	where this may change rounding behavior.
+	* real.c (real_can_shorten_arithmetic): New.
+	* real.h (real_can_shorten_arithmetic): Declare.
+
+2008-10-29  Bernd Schmidt  <bernd.schmidt@analog.com>
+	
+	* config/bfin/bfin-protos.h (WA_05000257, WA_05000283, WA_05000315,
+	ENABLE_WA_05000257, ENABLE_WA_05000283, ENABLE_WA_05000315): New.
+	* config/bfin/bfin.c (bfin_cpus): Add these workaround bits as
+	appropriate.
+	(must_save_p): For some workarounds, interrupts need to clobber a
+	P register.
+	(expand_prologue_reg_save, expand_epilogue_reg_restore): Save LC0
+	and LC1 for WA_05000257.
+	(expand_interrupt_handler_prologue): Add dummy read of CHIPID for
+	WA_05000283 and WA_05000315.
+	* config/bfin/bfin.md (UNSPEC_VOLATILE_DUMMY): New constant.
+	(movbi): Add alternative to set CC to 1; improve code for setting
+	CC to 0.
+	(dummy_load): New pattern.
+
+2008-10-29  Jakub Jelinek  <jakub@redhat.com>
+
+	PR middle-end/37870
+	* expmed.c (extract_bit_field_1): If int_mode_for_mode returns
+	BLKmode for non-memory, convert using a wider MODE_INT mode
+	or through memory.
+
+	PR middle-end/37913
+	* tree-cfgcleanup.c (split_bbs_on_noreturn_calls): Only split bbs
+	that haven't been removed yet.
+
+2008-10-29  Bernd Schmidt  <bernd.schmidt@analog.com>
+
+	* config/bfin/bfin.c (struct machine_function): New member
+	has_loopreg_clobber.
+	(bfin_expand_movmem): Set it when generating memcpy insns.
+	(n_regs_saved_by_prologue, expand_prologue_reg_save,
+	expand_epilogue_reg_restore): If we have hardware loops,
+	memcpy insns (indicated by has_loopreg_clobber) or function
+	calls, we need to save the loop registers.
+
+2008-10-28  H.J. Lu  <hongjiu.lu@intel.com>
+
+	* config/i386/i386.c (core2_cost): Fix typos in comments.
+
+2008-10-28  Jakub Jelinek  <jakub@redhat.com>
+
+	PR c/37924
+	* combine.c (make_compound_operation): Don't call make_extraction with
+	non-positive length.
+	(simplify_shift_const_1): Canonicalize count even if complement_p.
+
+2008-10-28  Joseph Myers  <joseph@codesourcery.com>
+
+	* convert.c (strip_float_extensions): Do not remove or introduce
+	conversions between binary and decimal floating-point types.
+
+2008-10-28  Jakub Jelinek  <jakub@redhat.com>
+
+	PR middle-end/37931
+	* fold-const.c (distribute_bit_expr): Convert common, left and
+	right arguments to type.
+
+2008-10-28  Nick Clifton  <nickc@redhat.com>
+
+	* config/mn10300/mn10300.h (CALL_REALLY_USED_REGISTERS): Define.
+	* config/mn10300/mn10300.c (fp_regs_to_save): Test the
+	call_really_used_regs array rather than the call_used_regs array.
+	(mn10300_get_live_callee_saved_regs, expand_prologue,
+	expand_epilogue, output_tst): Likewise.
+
+2008-10-27  Jakub Jelinek  <jakub@redhat.com>
+
+	PR target/37378
+	* df-scan.c (df_bb_refs_collect): Don't handle EH_USES here.
+	(df_get_entry_block_def_set): Neither here.
+	(df_get_regular_block_artificial_uses): Add EH_USES registers.
+
+	PR tree-optimization/37879
+	* predict.c (tree_estimate_probability): Check if last_stmt is
+	non-NULL before dereferencing it.
+
+2008-10-27  Vladimir Makarov  <vmakarov@redhat.com>
+
+	* ira-int.h (ira_allocno): Add member updated_cover_class_cost.
+	(ALLOCNO_UPDATED_COVER_CLASS_COST): New.
+	(ira_fast_allocation): Remove the prototype.
+	
+	* ira-color.c (update_copy_costs, allocno_cost_compare_func,
+	assign_hard_reg, calculate_allocno_spill_cost): Use updated costs.
+	(color_pass): Modify the updated costs.
+	(ira_color): Rename to color.  Make it static.
+	(ira_fast_allocation): Rename to fast_allocation.  Make it static.
+	(ira_color): New function.
+	
+	* ira-conflicts.c (process_regs_for_copy): Propagate hard reg cost
+	change.
+
+	* ira-lives.c (last_call_num, allocno_saved_at_call): New
+	variables.
+	(set_allocno_live, clear_allocno_live, mark_ref_live,
+	mark_ref_dead): Invalidate corresponding element of
+	allocno_saved_at_call.
+	(process_bb_node_lives): Increment last_call_num.  Setup
+	allocno_saved_at_call.  Don't increase ALLOCNO_CALL_FREQ if the
+	allocno was already saved.
+	(ira_create_allocno_live_ranges): Initiate last_call_num and
+	allocno_saved_at_call.
+
+	* ira-build.c (ira_create_allocno): Initiate
+	ALLOCNO_UPDATED_COVER_CLASS_COST.
+	(create_cap_allocno, propagate_allocno_info,
+	remove_unnecessary_allocnos): Remove setting updated costs.
+	(ira_flattening): Set up ALLOCNO_UPDATED_COVER_CLASS_COST.
+
+	* ira.c (ira):  Don't call ira_fast_allocation.
+
+	* ira-costs.c (setup_allocno_cover_class_and_costs): Don't set up
+	updated costs.
+	
+2008-10-27  Vladimir Makarov  <vmakarov@redhat.com>
+
+	PR middle-end/37813
+	* ira-conflicts.c (process_regs_for_copy): Remove class subset check.
+
+	* ira-int.h (ira_hard_regno_cover_class): New.
+
+	* ira-lives.c (mark_reg_live, mark_reg_dead,
+	process_bb_node_lives): Use ira_hard_regno_cover_class.
+
+	* ira.c (reg_class ira_hard_regno_cover_class): New global variable.
+	(setup_hard_regno_cover_class): New function.
+	(ira_init): Call setup_hard_regno_cover_class.
+
+	* ira-costs.c (cost_class_nums): Add comment.
+	(find_allocno_class_costs): Initiate cost_class_nums.
+	(setup_allocno_cover_class_and_costs): Check cost_class_nums.
+
+2008-10-27  Vladimir Makarov  <vmakarov@redhat.com>
+
+	PR middle-end/37884
+	* ira-build.c (copy_live_ranges_to_removed_store_destinations):
+	Rename to copy_info_to_removed_store_destinations.  Propagate
+	conflict hard regs and register stack attribute.
+
+2008-10-26  John David Anglin  <dave.anglin@nrc-cnrc.gc.ca>
+
+	PR middle-end/37316
+	* pa.c (function_arg_padding):  Pad complex and vector types
+	upward in 64-bit runtime.
+	(function_arg): Use BLKmode for PARALLEL in 64-bit runtime.
+
+2008-10-26  Matthias Klose  <doko@ubuntu.com>
+
+	* doc/install.texi: Document requirements on antlr.
+
+2008-10-25  Richard Sandiford  <rdsandiford@googlemail.com>
+
+	* config/mips/mips.h (REG_ALLOC_ORDER): Put call-clobbered registers
+	first.
+
+2008-10-24  Michael Meissner  <meissner@linux.vnet.ibm.com>
+
+	PR target/37841
+	* function.c (locate_and_pad_parm): If FUNCTION_ARG_ADJUST_OFFSET
+	is defined, use it to modify the constant offset.
+
+	* doc/tm.texi (FUNCTION_ARG_OFFSET): Document new macro.
+
+	* config/spu/spu.h (FUNCTION_ARG_OFFSET): New macro to move char
+	and short arguments to the correct location as mandated by the ABI.
+
+2008-10-24  Kaz Kojima  <kkojima@gcc.gnu.org>
+
+	PR rtl-optimization/37769
+	* regmove.c (optimize_reg_copy_2): Update REG_INC note if needed.
+
+2008-10-24  Kaz Kojima  <kkojima@gcc.gnu.org>
+
+	* config/sh/t-sh: Use $(MULTILIB_CFLAGS) when compiling to
+	unwind-dw2-Os-4-200.o.
+
+2008-10-24  Joseph Myers  <joseph@codesourcery.com>
+
+	* c-typeck.c (enum impl_conv): Remove ic_argpass_nonproto.
+	(convert_for_assignment): Remove ic_argpass_nonproto cases.
+
+2008-10-24  Jakub Jelinek  <jakub@redhat.com>
+
+	PR tree-optimization/36038
+	* tree-ssa-loop-ivopts.c (add_old_iv_candidates): For pointer bases
+	add sizetype IV with initial value zero instead of pointer type.
+
+2008-10-24  Manuel L??pez-Ib????ez  <manu@gcc.gnu.org>
+
+	PR c/7543
+	* value-prof.c (gimple_stringop_fixed_value): Use parentheses
+	around bit operation.
+	* profile.c (is_edge_inconsistent): Likewise.
+	* fold-const.c (truth_value_p): Move from here...
+	* tree.h (truth_value_p): ... to here.
+	* c-tree.h (c_expr): Update description of original_code.
+	* c-typeck.c (parser_build_unary_op): Set original_code.
+	(parser_build_binary_op): Update call to warn_about_parentheses.
+	* c-common.c (warn_about_parentheses): Take two additional
+	arguments of the operands. Use a switch. Quote operators
+	appropriately. Define macro APPEARS_TO_BE_BOOLEAN_EXPR_P.
+	Add warning about !x | y and !x & y.
+	* c-common.h (warn_about_parentheses): Update declaration.
+
+2008-10-24  Jakub Jelinek  <jakub@redhat.com>
+
+	* Makefile.in (lang_checks_parallelized, check_gcc_parallelize,
+	check_p_tool, check_p_vars, check_p_subno, check_p_comma,
+	check_p_subwork, check_p_numbers, check_p_subdir, check_p_subdirs):
+	New variables.
+	(check-subtargets, check-%-subtargets, check-parallel-%): New targets.
+	(check-%): For test targets listed in lang_checks_parallelized
+	if -j is used and RUNTESTFLAGS doesn't specify tests to execute,
+	run the testing in multiple make goals, possibly parallel, and
+	afterwards run dg-extract-results.sh to merge the sum and log files.
+
+2008-10-24  Richard Sandiford  <rdsandiford@googlemail.com>
+
+	* config/mips/mips.c (mips_canonicalize_move_class): New function.
+	(mips_move_to_gpr_cost): Likewise.
+	(mips_move_from_gpr_cost): Likewise.
+	(mips_register_move_cost): Make more fine-grained.
+
+2008-10-23  Tobias Grosser  <grosser@fim.uni-passau.de>
+
+	* graphite.c (graphite_apply_transformations): Check for
+	-fgraphite-identity.
+	* toplev.c (process_options): Add graphite_identity.
+	* tree-ssa-loop.c (gate_graphite_transforms): Add graphite_identity.
+
+2008-10-23  Bernd Schmidt  <bernd.schmidt@analog.com>
+
+	* config/bfin/bfin.c (bdesc_2arg): Add mulhisill, mulhisilh,
+	mulhisihl and mulhisihh builtins.
+
+	From Jie Zhang  <jie.zhang@analog.com>
+	* config/bfin/bfin.md (composev2hi): Put operands into vector
+	with correct order.
+	
+2008-10-22  Chao-ying Fu  <fu@mips.com>
+
+	* config/mips/mips.opt (msmartmips): Accept -mno-smartmips.
+
+2008-10-22  Bernd Schmidt  <bernd.schmidt@analog.com>
+
+	From Mike Frysinger  <michael.frysinger@analog.com>
+	* config/bfin/bfin-protos.h (bfin_cpu_type): Add BFIN_CPU_BF512,
+	BFIN_CPU_BF514, BFIN_CPU_BF516, and BFIN_CPU_BF518.
+	* config/bfin/bfin.c (bfin_cpus[]): Add 0.0 for bf512, bf514, bf516,
+	and bf518.  Add 0.2 for bf522, bf523, bf524, bf526, and bf527.
+	Add 0.6 for bf533, bf532, and bf531.  Add 0.5 for bf538 and bf539.
+	Add 0.2 for bf542, bf544, bf547, bf548, and bf549.
+	* config/bfin/bfin.h (TARGET_CPU_CPP_BUILTINS): Define __ADSPBF512__
+	for BFIN_CPU_BF512, __ADSPBF514__ for BFIN_CPU_BF514, __ADSPBF516__
+	for BFIN_CPU_BF516, and __ADSPBF518__ for BFIN_CPU_BF518.  Define
+	__ADSPBF51x__ for all of them.
+	* config/bfin/elf.h (LIB_SPEC): Select proper linker scripts for
+	-mcpu bf512, bf514, bf516, and bf518.
+	* config/bfin/t-bfin-elf (MULTILIB_MATCHES): Select bf532-none for
+	bf512-none, bf514-none, bf516-none, and bf518-none.
+	* config/bfin/t-bfin-linux (MULTILIB_MATCHES): Likewise.
+	* config/bfin/t-bfin-uclinux (MULTILIB_MATCHES): Likewise.
+	* doc/invoke.texi (Blackfin Options): Document that
+	-mcpu now accepts bf512, bf514, bf516, and bf518.
+
+2008-10-22  Jakub Jelinek  <jakub@redhat.com>
+
+	PR middle-end/37882
+	* fold-const.c (build_range_type): For 1 .. signed_max
+	range call build_nonstandard_inter_type if signed_type_for
+	returned a type with bigger precision.
+
+2008-10-22  Richard Guenther  <rguenther@suse.de>
+
+	* tree.def (COMPLEX_TYPE): Constrain element type.
+	* tree.c (build_complex_type): Assert that.
+	* tree-ssa-structalias.c (could_have_pointers): Complex types
+	cannot have pointers.
+
+2008-10-22  Manuel L??pez-Ib????ez  <manu@gcc.gnu.org>
+
+	PR c/30949
+	* c-typeck.c (convert_for_assignment): Do not give declaration's
+	location for builtins. Spell out which type was expected and which
+	was given.
+
+2008-10-22  Nick Clifton  <nickc@redhat.com>
+
+	* config/frv/frv.h (HARD_REGNO_RENAME_OK): Define.  Do not allow
+	the fdpic register to be a target when running in fdpic mode.
+
+2008-10-22 Rafael Espindola  <espindola@google.com>
+
+	* ipa-struct-reorg.c (create_new_alloc_sites): Properly insert the
+	newly created statements.
+
+2008-10-22  Kaz Kojima  <kkojima@gcc.gnu.org>
+
+	PR target/37633
+	* config/sh/sh.c (sh_hard_regno_mode_ok): New function.
+	* config/sh/sh.h (HARD_REGNO_MODE_OK): Use it.
+	* config/sh/sh-protos.h (sh_hard_regno_mode_ok): Declare.
+
+2008-10-22  Jakub Jelinek  <jakub@redhat.com>
+
+	PR target/37880
+	* doc/invoke.texi: Fix spelling of -mlarge-data-threshold option.
+	Adjust -mcmodel=medium description for 2005-07-31 changes.
+
+2008-10-22  Jan Hubicka  <jh@suse.cz>
+	    Jakub Jelinek  <jakub@redhat.com>
+
+	PR middle-end/35853
+	* doc/invoke.texi: Remove references to obsoleted -d dumps.
+
+2008-10-21  Richard Henderson  <rth@redhat.com>
+
+	PR 37815
+	* emit-rtl.c (get_spill_slot_decl): Export.
+	* emit-rtl.h (get_spill_slot_decl): Declare.
+	* var-tracking.c (vt_add_function_parameters): Relax assertion
+	on the contents of MEM_EXPR in a PARM_DECL to include a spill slot.
+
+2008-10-21  Bob Wilson  <bob.wilson@acm.org>
+
+	* var-tracking.c (insn_stack_adjust_offset_pre_post): If insn has a
+	REG_FRAME_RELATED_EXPR note, examine the pattern from the note instead
+	of from insn.
+
+2008-10-21  Adam Nemet  <anemet@caviumnetworks.com>
+
+	PR middle-end/37669
+	* tree-ssa-ccp.c (ccp_fold_builtin): Only allow one argument to be
+	processed by get_maxval_strlen.
+
+2008-10-21  David Edelsohn  <edelsohn@gnu.org>
+
+	PR target/35485
+	* tree.c (get_file_function_name): Copy first_global_object_name.
+	Centralize call to clean_symbol_name.
+
+2008-10-21  Sandra Loosemore  <sandra@codesourcery.com>
+
+	* config.gcc (powerpc-*): Make t-ppcgas imply usegas.h.
+	* config/svr4.h (SVR4_ASM_SPEC): New.
+	(ASM_SPEC): Inherit from SVR4_ASM_SPEC.
+	* config/rs6000/sysv4.h (ASM_SPEC): Inherit from SVR4_ASM_SPEC.
+
+	* doc/invoke.texi (Option Summary): Add -T to linker options.
+	(Link Options): Document -T.
+	
+2008-10-21  Jakub Jelinek  <jakub@redhat.com>
+
+	* tree-inline.c (tree_inlinable_function_p): Remove tuples
+	debugging hack.
+
+	* gimplify.c (gimplify_expr): Drop TREE_OVERFLOW from INTEGER_CSTs.
+
+	PR debug/37020
+	* c-decl.c (merge_decls): Don't call outlining_inline_function hook.
+
+2008-10-20  Daniel Berlin  <dberlin@dberlin.org>
+
+	* tree-ssa-pre.c (insert_into_preds_of_block): Don't rewrite constant
+	part of expression.
+
+2008-10-20  Tobias Schl??ter  <tobi@gcc.gnu.org>
+
+	* doc/install.texi: Fix typos in previous patch.
+
+2008-10-19  Tobias Schl??ter  <tobi@gcc.gnu.org>
+
+	* doc/install.texi: Document in-tree building of gcc and mpfr.
+
+2008-10-19  Richard Guenther  <rguenther@suse.de>
+
+	* tree-ssa-alias.c (may_alias_p): Remove bogus shortcut.
+
+2008-10-19  Manuel L??pez-Ib????ez  <manu@gcc.gnu.org>
+
+	PR c/30260
+	* c-decl.c (finish_enum): Convert non-integer enumerators to enum type.
+	(build_enumerator): Convert enumerators that fit in integer to
+	integer type.
+
+2008-10-18  Uros Bizjak  <ubizjak@gmail.com>
+
+	* config/i386/i386.md (unnamed peephole2): Do not force memory
+	operands of arith or logical instructions into registers for
+	non-TARGET_READ_MODIFY targets.
+
+	(ffs_cmove): Change operand 0 predicate to register_operand.
+	Change operand 1 predicate to nonimmediate_operand.
+	(ffsdi2): Ditto.
+	(*ffs_no_cmove): Change operand 0 predicate to register_operand.
+
+2008-10-18  Richard Guenther  <rguenther@suse.de>
+
+	* tree-ssa-forwprop.c (forward_propagate_addr_expr_1): Guard
+	VIEW_CONVERT_EXPR case against invalid gimple.
+
+2008-10-17  Andreas Krebbel  <krebbel1@de.ibm.com>
+
+	* c-parser.c (c_parser_binary_expression): Silence the
+	uninitialized variable warning emitted for binary_loc.
+
+2008-10-16  Daniel Berlin  <dberlin@dberlin.org>
+
+	* tree-ssa-pre.c (phi_translate_set): Add constants to phi
+	translation cache.
+
+2008-10-16  Joseph Myers  <joseph@codesourcery.com>
+
+	PR c/33192
+	* c-typeck.c (build_unary_op): Use omit_one_operand for
+	IMAGPART_EXPR of real argument.
+
+2008-10-16  Jakub Jelinek  <jakub@redhat.com>
+
+	PR tree-optimization/37664
+	* fold-const.c (fold_binary): When optimizing comparison with
+	highest or lowest type's value, don't consider TREE_OVERFLOW.
+
+2008-10-16  David Edelsohn  <edelsohn@gnu.org>
+
+	PR target/35483
+	* Makefile.in (coverage.o): Depend on $(TM_P_H).
+	* coverage.c: Include tm_p.h.
+	* config/rs6000/x-aix (jc1): Override LDFLAGS.
+	* config/rs6000/xcoff.h (ASM_GENERATE_INTERNAL_LABEL): Strip
+	dollar signs from PREFIX.
+	* config/rs6000/rs6000.c (output_toc): Use RS6000_OUTPUT_BASENAME
+	instead of manual strip_name_encoding.
+
+2008-10-16  Richard Guenther  <rguenther@suse.de>
+
+	PR middle-end/37418
+	* tree-ssa.c (useless_type_conversion_p_1): Do not treat
+	volatile qualified functions or methods as relevant.
+
+2008-10-16  Jakub Jelinek  <jakub@redhat.com>
+
+	PR tree-optimization/37525
+	* tree.c (int_fits_type_p): Rewrite using double_int.  Zero extend
+	sizetype unsigned constants, both in the case of C and bounds.
+
+2008-10-15  Jan Hubicka  <jh@suse.cz>
+
+	* ira-emit.c (change_regs): Return false when replacing reg by itself.
+
+2008-10-14  Vladimir Makarov  <vmakarov@redhat.com>
+
+	PR target/37633
+	* ira-costs.c (ira_tune_allocno_costs_and_cover_classes): Check
+	HARD_REGNO_CALL_PART_CLOBBERED.
+
+2008-10-15  Vladimir Makarov  <vmakarov@redhat.com>
+
+	PR middle-end/37535
+	* ira-lives.c (mark_early_clobbers): Remove.
+	(make_pseudo_conflict, check_and_make_def_use_conflicts,
+	check_and_make_def_conflicts,
+	make_early_clobber_and_input_conflicts,
+	mark_hard_reg_early_clobbers): New functions.
+	(process_bb_node_lives): Call
+	make_early_clobber_and_input_conflicts and
+	mark_hard_reg_early_clobbers.  Make hard register inputs live again.
+
+	* doc/rtl.texi (clobber): Change descriotion of RA behaviour for
+	early clobbers of pseudo-registers.
+	
+2008-10-15  Vladimir Makarov  <vmakarov@redhat.com>
+
+	PR middle-end/37674
+	* ira-build.c (ira_flattening): Recalculate
+	ALLOCNO_TOTAL_NO_STACK_REG_P and ALLOCNO_TOTAL_CONFLICT_HARD_REGS
+	from the scratch instead of the propagation.
+	
+2008-10-15  Richard Sandiford  <rdsandiford@googlemail.com>
+
+	* doc/rtl.texi (const_double): Remove the "addr" operand.
+	Describe CONST_DOUBLE_* macros under const_double rather
+	than const_vector.
+	(const_fixed): Fix the operand description.
+	(const): Add an @findex directive.
+	(CONST0_RTX, CONST1_RTX, CONST2_RTX): Move description
+	after the constant rtl table.
+	(fix): Combine floating-point and fixed-point descriptions.
+	Fix hyphenation.
+	* sched-deps.c (sched_analyze_2): Remove reference to
+	CONST_DOUBLE_CHAIN.
+
+2008-10-15  Richard Sandiford  <rdsandiford@googlemail.com>
+
+	* config/mn10300/mn10300.h (OUTPUT_ADDR_CONST_EXTRA): Handle
+	UNSPEC_GOTSYM_OFFs.
+	* config/mn10300/mn10300.c (legitimate_pic_operand_p): Return true
+	for UNSPEC_GOTSYM_OFFs.
+	* config/mn10300/mn10300.md (UNSPEC_GOTSYM_OFF): New unspec.
+	(add_GOT_to_pic_reg): Use it.
+	* config/mn10300/constraints.md (S): Allow UNSPEC_GOTSYM_OFF.
+
+2008-10-15  Jan Sjodin  <jan.sjodin@amd.com>
+	    Harsha Jagasia  <harsha.jagasia@amd.com>
+
+	PR tree-optimization/37485
+	* graphite.c (gmp_cst_to_tree): Moved.
+	(iv_stack_entry_is_constant): New.
+	(iv_stack_entry_is_iv): New.
+	(loop_iv_stack_push): Renamed to loop_iv_stack_push_iv.
+	(loop_iv_stack_insert_constant): New.
+	(loop_iv_stack_pop): Use new datatpype.
+	(loop_iv_stack_get_iv): Same.
+	(loop_iv_stack_get_iv_from_name): Same.
+	(loop_iv_stack_debug): Renamed to debug_loop_iv_stack.
+	(loop_iv_stack_patch_for_consts): New.
+	(loop_iv_stack_remove_constants): New.
+	(graphite_create_new_loop): Use loop_iv_stack_push_iv.
+	(translate_clast): Call loop_iv_stack_patch_for_consts and
+	loop_iv_stack_remove_constants.
+	(gloog): Use new datatype.  Redirect construction edge to end
+	block to avoid accidental deletion.
+	* graphite.h (enum iv_stack_entry_kind): New.  Tag for data in
+	iv stack entry.
+	(union iv_stack_entry_data): New.  Data in iv stack entry.
+	(struct iv_stack_entry): New.  Datatype for iv stack entries.
+
+2008-10-15  Sebastian Pop  <sebastian.pop@amd.com>
+
+	* tree-data-ref.c (stmt_simple_memref_p): Fix memleak.
+	* graphite.c (free_scop): Same.
+	(build_scops): Same.
+	(free_scattering): New.
+	(build_cloog_prog): Call free_scattering.
+	(patch_phis_for_virtual_defs): Fix memleak.
+	(graphite_trans_bb_strip_mine): Same.
+	(limit_scops): Same.
+
+2008-10-15  Sebastian Pop  <sebastian.pop@amd.com>
+	    Harsha Jagasia  <harsha.jagasia@amd.com>
+
+	PR tree-optimization/37828
+	* graphite.c (graphite_trans_loop_block): Do not loop block
+	single nested loops.
+
+2008-10-15  Sebastian Pop  <sebastian.pop@amd.com>
+
+	PR tree-optimization/37686
+	* tree-loop-linear.c (linear_transform_loops): Build a
+	loop nest vector.  Pass it to lambda_compute_access_matrices.
+	* tree-data-ref.h (struct access_matrix): Store the loop nest
+	relative to which it encodes the information.
+	(AM_LOOP_NEST_NUM): Renamed AM_LOOP_NEST.
+	(am_vector_index_for_loop): Reimplemented: iterate over the
+	loop nest for finding the loop index in the access matrix.
+	(lambda_compute_access_matrices): Update declaration.
+	* lambda-code.c (build_access_matrix): Pass the loop nest and
+	record it.
+	(lambda_compute_access_matrices): Same.
+
+2008-10-15  Andreas Krebbel  <krebbel1@de.ibm.com>
+
+	* config/s390/s390.h: (TARGET_DFP): This requires TARGET_HARD_FLOAT.
+	* config/s390/s390.md: Replace "TARGET_HARD_FLOAT &&
+	TARGET_(HARD_)DFP" with only TARGET_(HARD_)DFP since this
+	already implies hard float.  Also fix several insn condition
+	with TARGET_DFP which shall require TARGET_HARD_DFP instead.
+
+2008-10-15  David Edelsohn  <edelsohn@gnu.org>
+
+	PR target/35483


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.



More information about the Gcc-regression mailing list