12008-12-23 Maynard Johnson <maynardj@us.ibm.com> 2 3 * utils/opcontrol: reverse the logic in 4 is_non_cell_ppc64_variant to avoid the need 5 for maintenance as new ppc64 variants are addded 6 72008-12-12 Maynard Johnson <maynardj@us.ibm.com> 8 9 * libop/Makefile.am: Add Andi Kleen's new op_hw_specific.h 10 file to to Makefile.am 11 122008-11-24 Robert Richter <robert.richter@amd.com> 13 14 * utils/opcontrol: Correct spelling error 15 162008-11-24 Robert Richter <robert.richter@amd.com> 17 18 * daemon/opd_events.c: 19 * daemon/opd_events.h: 20 * daemon/opd_mangling.c: 21 * daemon/opd_mangling.h: 22 * daemon/opd_sfile.c: 23 * daemon/opd_trans.c: 24 * events/x86-64/family10/events: 25 * utils/opcontrol: fix whitespaces 26 272008-11-12 Andi Kleen <andi@firstfloor.org> 28 29 * libop/op_events.c 30 * libop/op_events.h 31 * libpp/op_header.cpp: Fix regression in arch perfmon code 32 (see bug #2161762) 33 342008-11-12 Andi Kleen <andi@firstfloor.org> 35 36 * events/i386/arch_perfmon/events: Fix event name to avoid 37 parsing error 38 392008-11-05 Maynard Johnson <maynardj@us.ibm.com> 40 41 * libutil++/file_manip.cpp: Add mode arg to open 42 432008-10-30 William Cohen <wcohen@redhat.com> 44 45 * m4/binutils.m4: Defeat compiler optimization in configure 46 check for bfd_get_synthetic_symtab 47 482008-10-27 Maynard Johnson <maynardj@us.ibm.com> 49 50 * libutil++/op_spu_bfd.cpp: Initialize anon_obj to false for op_bfd 51 objects for Cell SPE embedded binaries 52 532008-10-23 Dave Nomura <dcnltc@us.ibm.com> 54 55 * libutil++/xml_output.cpp: 56 * libutil/xml_out.c: 57 * libutil/xml_out.h: 58 * libutil/xml_events.c: 59 * libutil/xml_events.h: 60 * libutil/Makefile.am: 61 * libop/op_xml_out.c: 62 * libop/op_xml_out.h: 63 * libop/op_xml_events.c: 64 * libop/op_xml_events.h: 65 * libop/Makefile.am: moved xml util routines to libop and fixed some 66 other problems in the previous patch 67 682008-10-15 Dave Nomura <dcnltc@us.ibm.com> 69 70 * utils/ophelp.c: 71 * libutil++/xml_output.cpp: 72 * libutil++/xml_output.h: 73 * libutil/xml_out.c: 74 * libutil/xml_out.h: 75 * libutil/xml_events.c: 76 * libutil/xml_events.h: 77 * libutil/Makefile.am: added -X option to ophelp to generate XML 78 792008-10-06 Maynard Johnson <maynardj@us.ibm.com> 80 81 * configure.in: 82 * m4/binutils.m4: 83 * m4/cellspubfdsupport.m4: Handle BFD's new dependency on libz 84 852008-08-28 Maynard Johnson <maynardj@us.ibm.com> 86 87 * configure.in: Add configure option for non-standard binutils lib 88 892008-08-28 Andi Kleen <andi@firstfloor.org> 90 91 * events/i386/arch_perfmon/events: 92 * events/i386/arch_perfmon/unit_masks: 93 * events/Makefile.am: 94 * libop/op_cpu_type.c: 95 * libop/op_cpu_type.h: 96 * libop/op_events.c: 97 * libop/op_events.h: 98 * libop/op_hw_specific.h: 99 * libpp/op_header.cpp: 100 * utils/ophelp.c: support for Intel arch perfmon 101 1022008-08-28 Maynard Johnson <maynardj@us.ibm.com> 103 104 * configure.in: Change AM_INIT_AUTOMAKE to 0.9.5cvs; enables -werror 105 gcc option to stop build on warnings 106 1072008-08-27 Jatin Nansi <jnansi@redhat.com> 108 109 * utils/opcontrol: vecho parameters after "--verbose" option processed 110 1112008-08-08 Daniel Hansel <daniel.hansel@linux.vnet.ibm.com> 112 113 * daemon/init.c: Fixed call of execvp() to execute opjitconv if 114 it is installed in a custom directory 115 1162008-08-07 Maynard Johnson <maynardj@us.ibm.com> 117 118 * libutil++/bfd_support.cpp: 119 * libutil++/op_bfd.cpp: 120 * libutil++/op_bfd.h: Fix a couple problems relating 121 to overlay symbols for Cell SPE applications 122 1232008-07-31 Maynard Johnson <maynardj@us.ibm.com> 124 125 * configure.in: Fix to allow libtool to recognize 126 alternate binutils dir passed via --with-binutils 127 1282008-07-24 Jean Pihet <jpihet@mvista.com> 129 130 * events/arm/armv7/events: 131 * events/arm/armv7/unit_masks: 132 * libop/op_cpu_type.c: 133 * libop/op_cpu_type.h: 134 * libop/op_events.c: 135 * utils/ophelp.c: Added ARMv7 support to be consistent with the 136 kernel, added ARMv7 specific events. 137 Tested on OMAP3430 and OMAP3530 chipsets. 138 1392008-07-17 Maynard Johnson <maynardj@us.ibm.com> 140 141 * configure.in: bump version in AM_INIT_AUTOMAKE to 0.9.4 142 1432008-07-17 Will Cohen <wcohen@redhat.com> 144 145 * libop/op_alloc_counter.c: Assume perfmon managing PMU hw 146 when no counters 147 1482008-07-07 Maynard Johnson <maynardj@us.ibm.com> 149 150 * HACKING: Ask contributors to include Signed-off-by 151 line with their patches 152 1532008-07-03 Richard Purdie <rpurdie@openedhand.com> 154 155 * libutil++/bfd_support.cpp: 156 * libutil++/bfd_support.h: 157 * libutil++/op_bfd.cpp: 158 * libutil++/op_bfd.h: 159 * libutil++/op_spu_bfd.cpp: 160 * pp/oparchive.cpp: Switch separate_debug_file_exists() 161 and hence find_separate_debug_file() to use 162 extra_images::find_image_path() so debug files in --root 163 directories are correctly found 164 1652008-07-03 Maynard Johnson <maynardj@us.ibm.com> 166 167 * ChangeLog: Correct date of previous entry 168 * README_PACKAGERS: Add more advice to packagers 169 regarding the new JIT support libraries 170 1712008-07-03 Maynard Johnson <maynardj@us.ibm.com> 172 173 * libopagent/Makefile.am: 174 * libopagent/opagent.c: 175 * remove libopagent/bfddefines.c: Move code from bfddefines 176 to libopagent to fix cross-compile error 177 1782008-06-30 Shunichi Fuji <palglowr@gmail.com> 179 180 * libpp/callgraph_container.cpp: Fix compile warning 181 using gcc 4.3.1 for uninitialized variable 182 1832008-06-23 Maynard Johnson <maynardj@us.ibm.com> 184 185 * daemon/opd_stats.c: Do not force printing of sample_invalid_eip 186 since this is not available from pre-2.6.24 kernels 187 1882008-06-23 Maynard Johnson <maynardj@us.ibm.com> 189 190 * opjitconv/opjitconv.c: Switch to special user account 191 after determining whether or not we have JIT dump files 192 to process in order to avoid unnecessary error messages 193 in oprofiled.log 194 1952008-06-11 Maynard Johnson <maynardj@us.ibm.com> 196 197 * configure.in: 198 * Makefile.am: 199 * agents/Makefile.am: 200 * doc/oprofile.xml: Change make install error to warning 201 if special user account does not exist 202 2032008-06-11 Maynard Johnson <maynardj@us.ibm.com> 204 205 * Makefile.am: Add README_PACKAGERS to EXTRA_DIST 206 2072008-06-03 Will Cohen <wcohen@redhat.com> 208 209 * libop/op_alloc_counter.c: Use counters actually available. 210 2112008-05-21 George McCollister <georgem@novatech-llc.com> 212 213 * daemon/opd_cookie.c: Fix ARM big-endian syscall 214 2152008-05-20 Jason Yeh <jason.yeh@amd.com> 216 217 * events/x86_64/hammer/events: 218 * events/x86_64/hammer/unit_masks: 219 * events/x86_64/family10/events: 220 * events/x86_64/family10/unit_masks: update for recent BKDG 221 revisions 222 2232008-05-19 Maynard Johnson <maynardj@us.ibm.com> 224 225 * libpp/profile_spec.cpp: Fix bug in differential profile 226 when using archive spec and an image spec 227 2282008-05-19 Maynard Johnson <maynardj@us.ibm.com> 229 230 * libpp/format_output.cpp: 231 * libpp/xml_utils.cpp: 232 * libutil++/op_bfd.cpp: 233 * libutil++/op_bfd.h: Do not to try to get symbol contents 234 for artificial symbols and other cases where no symbol contents 235 exist. Also change to use scoped array for symbol contents 236 instead of automatic allocation from the stack to avoid allocation 237 attempt exceeding max stack size. 238 2392008-05-09 Maynard Johnson <maynardj@us.ibm.com> 240 241 * configure.in: 242 * Makefile.am: Improve error messages for special user account checks 243 2442008-05-08 Maynard Johnson <maynardj@us.ibm.com> 245 246 * libpp/callgraph_container.cpp: 247 * libpp/profile_container.cpp: 248 * libutil++/op_bfd.h: 249 * libutil++/op_bfd.cpp: Fix to allow 32-bit opreport to 250 properly process samples from 64-bit JIT profile agent 251 2522008-04-29 René Rebe <rene@exactcode.de> 253 254 * libabi/opimport.cpp: 255 * libpp/op_header.cpp: 256 * libpp/profile.cpp: 257 * libutil++/bfd_support.cpp: 258 * libutil++/cverb.cpp: 259 * libutil++/op_bfd.cpp: 260 * libutil++/op_spu_bfd.cpp: 261 * pp/common_option.cpp: 262 * pp/opannotate_options.cpp: 263 * pp/oparchive.cpp: 264 * pp/opgprof_options.cpp: added explicit includes for C functions 265 as required for compilation with GCC-4.3 266 2672008-04-28 Daniel Hansel <daniel.hansel@linux.vnet.ibm.com> 268 269 * oprofile/agents/jvmpi/jvmpi_oprofile.cpp: 270 * oprofile/agents/jvmpi/Makefile.am: 271 * oprofile/agents/jvmti/libjvmti_oprofile.c: 272 * oprofile/agents/jvmti/Makefile.am: 273 * oprofile/agents/Makefile.am: 274 * oprofile/autogen.sh: 275 * oprofile/configure.in: 276 * oprofile/daemon/init.c: 277 * oprofile/daemon/Makefile.am: 278 * oprofile/daemon/opd_anon.c: 279 * oprofile/daemon/opd_pipe.c: 280 * oprofile/daemon/opd_pipe.h: 281 * oprofile/daemon/oprofiled.c: 282 * oprofile/daemon/oprofiled.h: 283 * oprofile/doc/Makefile.am: 284 * oprofile/doc/op-jit-devel.xml: 285 * oprofile/doc/oprofile.xml: 286 * oprofile/libop/op_config.c: 287 * oprofile/libop/op_config.h: 288 * oprofile/libopagent/bfddefines.c: 289 * oprofile/libopagent/jitdump.h: 290 * oprofile/libopagent/Makefile.am: 291 * oprofile/libopagent/opagent.c: 292 * oprofile/libopagent/opagent.h: 293 * oprofile/libopagent/opagent_symbols.ver: 294 * oprofile/libpp/arrange_profiles.cpp: 295 * oprofile/libpp/callgraph_container.cpp: 296 * oprofile/libpp/filename_spec.cpp: 297 * oprofile/libpp/filename_spec.h: 298 * oprofile/libpp/op_header.cpp: 299 * oprofile/libpp/op_header.h: 300 * oprofile/libpp/parse_filename.cpp: 301 * oprofile/libpp/parse_filename.h: 302 * oprofile/libpp/profile_container.cpp: 303 * oprofile/libpp/profile.cpp: 304 * oprofile/libpp/profile.h: 305 * oprofile/libpp/profile_spec.cpp: 306 * oprofile/libregex/demangle_java_symbol.cpp: 307 * oprofile/libregex/demangle_java_symbol.h: 308 * oprofile/libregex/demangle_symbol.cpp: 309 * oprofile/libregex/Makefile.am: 310 * oprofile/libregex/tests/java_test.cpp: 311 * oprofile/libregex/tests/Makefile.am: 312 * oprofile/libutil/Makefile.am: 313 * oprofile/libutil/op_file.c: 314 * oprofile/libutil/op_file.h: 315 * oprofile/libutil/op_growable_buffer.c: 316 * oprofile/libutil/op_growable_buffer.h: 317 * oprofile/libutil/op_list.h: 318 * oprofile/libutil++/bfd_support.cpp: 319 * oprofile/libutil++/bfd_support.h: 320 * oprofile/libutil++/file_manip.cpp: 321 * oprofile/libutil++/op_bfd.cpp: 322 * oprofile/libutil++/op_bfd.h: 323 * oprofile/Makefile.am: 324 * oprofile/opjitconv/conversion.c: 325 * oprofile/opjitconv/create_bfd.c: 326 * oprofile/opjitconv/debug_line.c: 327 * oprofile/opjitconv/jitsymbol.c: 328 * oprofile/opjitconv/Makefile.am: 329 * oprofile/opjitconv/opjitconv.c: 330 * oprofile/opjitconv/opjitconv.h: 331 * oprofile/opjitconv/parse_dump.c: 332 * oprofile/pp/opgprof.cpp: 333 * oprofile/README_PACKAGERS: 334 * oprofile/TODO: 335 * oprofile/utils/opcontrol: JIT support 336 3372008-04-18 Maynard Johnson <maynardj@us.ibm.com> 338 339 * m4/binutils.m4: A correct fix for the --with-binutils problem 340 3412008-04-17 Maynard Johnson <maynardj@us.ibm.com> 342 343 * m4/binutils.m4: Revert previous patch as it was buggy 344 3452008-04-16 Maynard Johnson <maynardj@us.ibm.com> 346 347 * m4/binutils.m4: Fix the bfd_get_synthetic_symtab check 348 in this m4 macro to work correctly with '--with-binutils' 349 configure option 350 3512008-04-10 Chris J Arges <arges@linux.vnet.ibm.com> 352 353 * utils/opcontrol: 354 For ppc64 architectures, need to check if all user/kernel flags 355 per counter match. Then use this value to set user/kernel domain 356 profiling. 357 3582008-03-13 Dave Nomura <dcnltc@us.ibm.com> 359 360 * libutil++/op_bfd.cpp: 361 * libutil++/op_bfd.h: 362 * pp/opreport.cpp: 363 gcc 4.3 gives warnings on const function return types 364 3652008-03-03 John Levon <levon@movementarian.org> 366 367 * daemon/oprofiled.c: fix error message typo 368 3692008-02-22 Haavard Skinnemoen <hskinnemoen@atmel.com> 370 371 * events/Makefile.am: 372 * events/avr32/events: 373 * events/avr32/unit_masks: 374 * libop/op_cpu_type.c: 375 * libop/op_cpu_type.h: 376 * libop/op_events.c: 377 * utils/ophelp.c: Add AVR32 support 378 3792008-02-22 Richard Purdie <rpurdie@openedhand.com> 380 381 * libutil++/file_manip.cpp: 382 * pp/oparchive.cpp: 383 Check files actually exist before copying or listing the file as part of 384 oparchive --list-files. 385 386 The copy_file function breaks if passed a directory as a file to copy 387 since it will create a file at the destination causing subtle bugs. 388 3892008-02-22 Andy Fleming <afleming@freescale.com> 390 391 * events/Makefile.am: 392 * events/ppc/e300/events: 393 * events/ppc/e300/unit_masks: 394 * libop/op_cpu_type.c: 395 * libop/op_cpu_type.h: 396 * libop/op_events.c: 397 * utils/ophelp.c: Add e300 support 398 3992008-02-22 Bob Nelson <rrnelson@us.ibm.com> 400 401 * utils/opcontrol: Fix loop in dump code when using 402 --session-dir on a network drive. (clock issues) 403 4042008-02-15 Will Cohen <wcohen@redhat.com> 405 406 * gui/oprof_start_util.cpp 407 * libutil++/file_manip.cpp 408 * libutil++/bfd_support.cpp 409 * libutil++/bfd_spu_support.cpp 410 * libutil++/op_spu_bfd.cpp 411 * libutil++/cverb.cpp 412 * libutil++/child_reader.cpp 413 * pp/opgprof_options.cpp 414 * pp/opannotate_options.cpp 415 * pp/oparchive.cpp 416 * pp/common_option.cpp 417 * libpp/sample_container.cpp 418 * libpp/op_header.cpp 419 * libpp/profile.cpp 420 * libabi/opimport.cpp 421 * libregex/demangle_symbol.cpp: Add includes for gcc 4.3 compatibility. 422 4232008-02-07 Bob Nelson <rrnelson@us.ibm.com> 424 425 * libutil++/op_bfd.h: add bool symb_artificial to op_bfd_symbol 426 * libutil++/op_bfd.cpp: ctor changes, use symb_artificial 427 * libutil++/op_spu_bfd.cpp: profile SPU 'stack' code 428 4292008-02-04 Bob Nelson <rrnelson@us.ibm.com> 430 431 * daemon/opd_spu.c: Fix number of records to check for 432 in an SPU context switch 433 4342008-01-21 Daniel Hansel <daniel.hansel@linux.vnet.ibm.com> 435 436 * various: style cleanups 437 4382008-01-15 Isaku Yamahata <yamahata@valinux.co.jp> 439 440 * daemon/opd_trans.c: fix previous SPU patch 441 4422008-01-11 Isaku Yamahata <yamahata@valinux.co.jp> 443 444 * daemon/opd_perfmon.c: Xen support for IA64 445 4462008-01-11 Markus Armbruster <armbru@redhat.com> 447 448 * daemon/opd_interface.h: 449 * daemon/opd_trans.c: work around SPU/Xen code numbering problem 450 451See ChangeLog-2007 for earlier changelogs. 452