12007-12-11 Philippe Elie <phil.el@wanadoo.fr> 2 3 * libutil++/op_bfd.h: remove sym_offset(), we can get what did 4 sym_offset() in a simpler way at call site 5 * libutil++/op_bfd.cpp: 6 * libpp/callgraph_container.cpp: 7 * libpp/profile_container.cpp: 8 * libpp/profile_container.h: 9 102007-11-25 Philippe Elie <phil.el@wanadoo.fr> 11 12 * daemon/opd_stats.c: dump invalid pc count, added to oprofile stats 13 fs 2.6.24-rc 14 152007-11-15 Dave Nomura <dcnltc@us.ibm.com> 16 17 * fixed a some typos in manually patched event files 18 192007-11-15 Dave Nomura <dcnltc@us.ibm.com> 20 21 * libpp/format_output.cpp: output XML SYMBOL_DATA for callers/callees 22 * libpp/format.h: 23 242007-11-15 Dave Nomura <dcnltc@us.ibm.com> 25 26 * events/ppc64/power6/event_mappings: updated Power6 event files 27 * events/ppc64/power6/events: 28 292007-11-15 Philippe Elie <phil.el@wanadoo.fr> 30 31 handle --root which act as a replacement (a prefix) for the / fs 32 * libpp/locate_images.cpp: root path search taking care if a 33 root path is given we never search through "/" 34 * libpp/locate_images.h: 35 * libpp/populate.cpp: 36 * libpp/populate_for_spu.cpp: use the right path to check last 37 modification time, this is yet again a change in behavior, module 38 was never found a this point, even if we found them later, and we 39 didn't try to check_mtime(), so opreport is again a bit more verbose 40 * libutil++/op_bfd.cpp: be more cautious when checking if we can try to 41 open the bfd file, don't open it in the '/' fs if a --root is given 42 * libpp/profile_spec.cpp: convey the root_path all over the code. 43 * libpp/profile_spec.h: 44 * pp/common_option.cpp: 45 * pp/common_option.h: 46 * pp/opannotate_options.cpp: 47 * pp/oparchive.cpp: 48 * pp/oparchive_options.cpp: 49 * pp/opgprof_options.cpp: 50 * pp/opreport_options.cpp: 51 * doc/opannotate.1.in: document --root 52 * doc/oparchive.1.in: 53 * doc/opgprof.1.in: 54 * doc/opreport.1.in: 55 * doc/oprofile.xml: 56 572007-11-10 Dave Nomura <dcnltc@us.ibm.com> 58 59 * libutil++/bfd_support.cpp: is_correct_function() is itended to deal 60 with missing debug information, if linenr != 0, don't call it since 61 we know we have valid debug info 62 632007-11-08 Haavard Skinnemoen <hskinnemoen@atmel.com> 64 65 * utils/opcontrol: Busybox's implementation of "kill" doesn't 66 understand the "-s SIG" option. Use "-SIG" instead. 67 682007-11-07 Philippe Elie <phil.el@wanadoo.fr> 69 70 Compile fix for gcc 2.95.3 71 * libpp/format_output.cpp: unused local var, then unused params 72 * libpp/format_output.h: 73 * libpp/op_header.cpp: missing include 74 * libutil++/bfd_spu_support.cpp: gcc 2.95.3 do not like struct bfd, 75 2.95 get it wrong probably but anyway all these struct are useless. 76 772007-11-06 Philippe Elie <phil.el@wanadoo.fr> 78 79 * pp/opannotate.cpp: for module we didn't use the right image name, 80 this bug appeared after 0.9.3 release 81 822007-11-05 Richard Purdie <rpurdie@openedhand.com> 83 84 * events/arm/armv6/events: Fix armv6 events to match mainline kernels 85 862007-11-04 Philippe Elie <phil.el@wanadoo.fr> 87 88 * utils/opcontrol: nmi watchdog is now cleanly handled in the kernel 89 but it can be on by default even if CONFIG_WATCHDOG=n, a sysctl 90 exists to put it off. We need to check this in opcontrol since we 91 returned success if a counter is missing and starting oprofile with 92 oprof_start silently failed. 93 942007-11-03 Richard Purdie <rpurdie@openedhand.com> 95 96 * pp/oparchive.cpp: Check list_files option before calling 97 create_dir() for debug files. 98 992007-11-03 Philippe Elie <phil.el@wanadoo.fr> 100 101 with --xml and --details, bfd open/close was done one time per 102 symbols, not one time per application. Solved by moving open/close 103 in the caller and cache bfd object, this work because --xml imply 104 symbols are sorted by app name then lib name. This is not perfect as 105 we can open/close multiple time bfd object with --separate=library 106 but still a real win. Quick test showed an improvement by over 40 107 times. Improvement grows as number of symbols per binary. 108 * libpp/xml_utils.cpp: move bfd open ... 109 * libpp/xml_utils.h: 110 * libpp/format_output.cpp: ... here 111 * libpp/format_output.h: 112 * pp/opreport.cpp: 113 1142007-11-02 Philippe Elie <phil.el@wanadoo.fr> 115 116 * pp/opreport_options.cpp: --xml is not compatible with --sort since 117 xml output need some predefined sort options, we warned against 118 the use of --sort but didn't reset the sort options to default. 119 1202007-11-01 Philippe Elie <phil.el@wanadoo.fr> 121 122 * pp/oparchive_options.cpp: oparchive need to be run as root if you 123 want to get save all binaries, generally I don't like to restrict 124 user to do silly things but --output-directory / is an easy typo 125 to do and will completely screw up the box. 126 1272007-10-30 Philippe Elie <phil.el@wanadoo.fr> 128 129 * libpp/format_output.cpp: Fix #931882, xml output not changed at 130 the moment. 131 1322007-10-30 Philippe Elie <phil.el@wanadoo.fr> 133 134 Prepare #931882 fix, continue to move down real filename 135 retrieval by extending image_name_storage which can hold now the 136 real image name and the image as derived from the sample filename. 137 Use the new api in a compatible way with the old behavior, no output 138 change intended. 139 * libpp/format_output.cpp: 140 * libpp/format_output.h: 141 * libpp/locate_images.cpp: 142 * libpp/locate_images.h: 143 * libpp/name_storage.cpp: 144 * libpp/name_storage.h: 145 * libpp/symbol.cpp: 146 * libpp/symbol.h: 147 * libpp/xml_utils.cpp: 148 * libpp/xml_utils.h: 149 * pp/opreport.cpp: 150 1512007-10-29 Philippe Elie <phil.el@wanadoo.fr> 152 153 * libpp/image_errors.cpp: Pass to image_errors the archive_path through 154 an extra_images to ensure no duplicate error message occur nor we 155 miss error, this fix two bugs, one with differential profile 156 an error message was missing when two binary are missing, another 157 with opreport where the same error message was duplicate because in 158 on case we passed the real binary name (prefixed with archive_path) 159 while latter we passed the short form of the binary name. 160 * libpp/image_errors.h: 161 * libpp/callgraph_container.cpp: 162 * libpp/populate.cpp: 163 * libpp/populate_for_spu.cpp: 164 * libpp/xml_utils.cpp: 165 * libutil++/op_bfd.h: 166 167 * pp/opannotate.cpp: global var archive_path is no longer used 168 * pp/opannotate_options.cpp: 169 * pp/opannotate_options.h: 170 * pp/oparchive.cpp: 171 * pp/opgprof.cpp: 172 * pp/opgprof_options.h: 173 * pp/opreport.cpp: 174 * pp/opreport_options.cpp: 175 * pp/opreport_options.h: 176 1772007-10-29 Philippe Elie <phil.el@wanadoo.fr> 178 179 Cleanup the way per spec archive_path is handled, as extra_images 180 is built using an archive_path it make no sense to pass archive_path 181 as parameter all over the place, rather make it a member of 182 extra_images and use it, anyway it was required the archive_path 183 passed to find_image_path() was the same as used to build the 184 extra_images. 185 186 As a side effect it fix a corner case: oparchive archive:tmp1 -o tmp2 187 was not working for module, archiving from an archive should work now. 188 * libpp/arrange_profiles.cpp: 189 * libpp/arrange_profiles.h: 190 * libpp/callgraph_container.cpp: 191 * libpp/callgraph_container.h: 192 * libpp/locate_images.cpp: 193 * libpp/locate_images.h: 194 * libpp/populate.cpp: 195 * libpp/populate.h: 196 * libpp/populate_for_spu.cpp: 197 * libpp/populate_for_spu.h: 198 * libpp/profile_spec.cpp: 199 * libpp/xml_utils.cpp: 200 * libpp/xml_utils.h: 201 * libutil++/op_bfd.cpp: 202 * libutil++/op_bfd.h: 203 * libutil++/op_spu_bfd.cpp: 204 * pp/opannotate.cpp: 205 * pp/oparchive.cpp: 206 * pp/oparchive_options.cpp: 207 * pp/oparchive_options.h: 208 * pp/opgprof.cpp: 209 * pp/opgprof_options.cpp: 210 * pp/opreport.cpp: 211 2122007-10-29 Philippe Elie <phil.el@wanadoo.fr> 213 214 * libutil++/file_manip.cpp: do not create the output file if the input 215 file can not be read. It change oparchive behavior which created 216 empty file with input file like -rws--x--x 1 root root 217 2182007-10-28 Philippe Elie <phil.el@wanadoo.fr> 219 220 * libpp/format_output.cpp: minor overkill 221 2222007-10-27 Philippe Elie <phil.el@wanadoo.fr> 223 224 Fix #1819350, it turned out this implied another fix for differential 225 profile, extra_images must be built per profile specification not 226 globally so most of this patch pass an extra_images down to op_bfd. 227 The way image name resolution has changed. We fixup an image name 228 only when really needed, when opening a bfd file, when checking a 229 binary exists and when we try matching an image name with a profile 230 specification, this robustify differential profile. This patch also 231 allow a profile_spec ala image:*oprofile.ko which was broken too. 232 It's now mandatory to use the same -p option used with oparchive when 233 using an archive with other pp tools, it was already the case but was 234 broken, this patch does not clarify the documentation about this 235 point because I suspect oparchive should be modified to save at the 236 top level directory the contents of -p option, pp tools will use that 237 automatically and --image_path will be obsoleted when used with an 238 archive. 239 240 This patch except clarifying than --image-path with archive: is a bit 241 special does not add any user level visible change. 242 * libpp/arrange_profiles.cpp: 243 * libpp/arrange_profiles.h: 244 * libpp/callgraph_container.cpp: 245 * libpp/callgraph_container.h: 246 * libpp/locate_images.cpp: 247 * libpp/locate_images.h: 248 * libpp/populate.cpp: 249 * libpp/populate_for_spu.cpp: 250 * libpp/profile_container.cpp: 251 * libpp/profile_container.h: 252 * libpp/profile_spec.cpp: 253 * libpp/profile_spec.h: 254 * libpp/xml_utils.cpp: 255 * libpp/xml_utils.h: 256 * libutil++/Makefile.am: 257 * libutil++/op_bfd.cpp: 258 * libutil++/op_bfd.h: 259 * libutil++/op_spu_bfd.cpp: 260 * pp/common_option.cpp: 261 * pp/common_option.h: 262 * pp/opannotate.cpp: 263 * pp/opannotate_options.cpp: 264 * pp/oparchive.cpp: 265 * pp/oparchive_options.cpp: 266 * pp/opgprof.cpp: 267 * pp/opgprof_options.cpp: 268 * pp/opreport.cpp: 269 * pp/opreport_options.cpp: 270 2712007-10-27 Philippe Elie <phil.el@wanadoo.fr> 272 273 * libpp/profile_spec.cpp: look like an obvious typo, we must fixup 274 the string we are parsing not a previously seen or empty string 275 2762007-10-25 Philippe Elie <phil.el@wanadoo.fr> 277 278 * libutil++/sparse_array.h: size() return the max index while we 279 want the max index + 1, the only user is diff_container, 280 differential container is broken in 0.9.3. This fix bug #1820202 281 2822007-10-25 Philippe Elie <phil.el@wanadoo.fr> 283 284 * utils/opcontrol: fix the previous opcontrol --reset patch, it was 285 broken if the daemon is running because --reset imply --dump so 286 we can use this --reset shortcut only if the daemon is not running. 287 2882007-10-25 Philippe Elie <phil.el@wanadoo.fr> 289 290 * utils/opcontrol: if objdump is missing the error message is 291 obscure, this fix bug #1564920 by checking if objdump exists 292 and is an executable before using it 293 2942007-10-25 Philippe Elie <phil.el@wanadoo.fr> 295 296 * utils/opcontrol: --reset doesn't need to load the module, it was 297 counter-intuitive than the sequence opcontrol --deinit; 298 opcontrol --reset must be followed by another opcontrol --deinit 299 3002007-10-25 Philippe Elie <phil.el@wanadoo.fr> 301 302 * utils/opcontrol: error out earlier if the module is not loaded 303 even for --list-events or --dump to avoid obscure error message 304 3052007-10-24 Philippe Elie <phil.el@wanadoo.fr> 306 307 * libop/op_config.h: bump sample filename format, should have been 308 done before 0.9.3 309 * libpp/op_header.cpp: do not use odb_open() to read the sample file 310 header, if it fails it can return EINVAL if sanity checking fails 311 so we can't know if the failure came from a sample file format 312 change or some other failure. This explain some "sample_filename: 313 Invalid argument" we got in bug report when user didn't cleanup 314 the sample file directory. 315 * libpp/profile.cpp: use read_header() first to check the sample file 316 format to get better error message. 317 3182007-10-19 Philippe Elie <phil.el@wanadoo.fr> 319 320 * libop/op_events.c: force the use of hexa notation in event file for 321 the field intended to be in hexacidemal so make check will catch 322 attempt to use decimal notation for them. 323 324 * events/mips/20K/events: all event number were in decimal but parsed 325 as hexadecimal, this fix bug #1717298 326 * events/mips/24K/events: 327 * events/mips/25K/events: 328 * events/mips/5K/events: 329 * events/mips/r12000/events: 330 * events/mips/sb1/events: 331 * events/mips/vr5432/events: 332 * events/mips/vr5500/events: 333 * events/mips/34K/events: comment a bunch of overlapping event. 334 * events/ppc64/cell-be/unit_masks: well, 0 == 0x0 but it's easier 335 to catch real error by forcing hexa notation for default unit mask. 336 337 * libop/tests/alloc_counter_tests.c: fix mips/34K test, the used event 338 has been remoded 339 3402007-10-17 Philippe Elie <phil.el@wanadoo.fr> 341 342 * doc/oprofile.xml: fix dead url 343 * utils/ophelp.c: 344 3452007-10-17 Jason Yeh <jason.yeh@amd.com> 346 347 * events/x86-64/family10/events: This patch updates the events and 348 unit_masks files to match the BIOS and Kernel Developer's Guide 349 For AMD Family 10h Processors Rev 3.00 released on September 07, 350 2007. The notable changes are L3 cache events and updates to 351 corresponding unit masks. 352 * events/x86-64/family10/unit_masks: 353 3542007-10-17 Philippe Elie <phil.el@wanadoo.fr> 355 356 * utils/opcontrol: newer shell accept if -z ... but older need 357 if test -z. This broke opcontrol --dump with bash 2.x and probably 358 other shell. With bash the error when running opcontrol as non root 359 was: 360 /usr/local/bin/opcontrol: line 1670: -z: command not found 361 3622007-10-12 Philippe Elie <phil.el@wanadoo.fr> 363 364 * libop/op_events.c: check for '.' inside event name, they can't work 365 * events/i386/core_2/events: change '.' by '_' in events name 366 * events/ppc64/cell-be/events: 367 3682007-10-11 Maynard Johnson <maynardj@us.ibm.com> 369 370 * events/ppc64/970MP/event_mappings: Turn off profiling in 371 hypervisor on 970MP to prevent lost interrupts 372 3732007-10-10 Maynard Johnson <maynardj@us.ibm.com> 374 375 * events/ppc64/970MP/events: 376 * events/ppc64/970MP/event_mappings: Fix MMCR values 377 and counter-to-event mappings on a few 970MP groups 378 3792007-10-09 Melchior Franz <mfranz@aon.at> 380 381 * libregex/op_regex.cpp: Fix broken string concatenation 382 3832007-10-05 Will Cohen <wcohen@redhat.com> 384 385 * utils/opcontrol: Modify opcontrol to allow short form -l and -d for 386 normal users. 387 3882007-09-25 Brian Twichell <tbrian@us.ibm.com> 389 390 * utils/opcontrol: Modify opcontrol to use a more inclusive 391 kernel range 392 3932007-08-23 Maynard Johnson <maynardj@us.ibm.com> 394 395 * daemon/opd_anon.c: Make anonymous sample recording 396 work with 32-bit OProfile and 64-bit anonymous code 397 3982007-08-14 John Levon <levon@movementarian.org> 399 400 * doc/oprofile.xml: 401 * doc/opcontrol.1.in: clarify behaviour of opcontrol -e 402 4032007-08-02 Jens Wilke <jens.wilke@de.ibm.com> 404 405 * ChangeLog: Correct the date from previous patch (maynardj) 406 407 * libutil++/op_bfd.cpp: Fix sym_offset calculation. 408 Bug triggers on 64 bit systems with code mapped 409 with an address >4G. 410 4112007-08-02 Jens Wilke <jens.wilke@de.ibm.com> 412 413 * libpp/profile_container.cpp: Remove unnecessary offset 414 calculation. 415 4162007-07-18 Maynard Johnson <maynardj@us.ibm.com> 417 418 * doc/Makefile.am: Include buffers diagram in dist 419 420 * doc/oprofile.xml: Correct filename reference to opreport.xsd 421 4222007-07-16 Maynard Johnson <maynardj@us.ibm.com> 423 424 * configure.in: Bump to 0.9.4cvs. 425 4262007-07-16 Maynard Johnson <maynardj@us.ibm.com> 427 428 * configure.in: 429 * Makefile.am: Bump release to 0.9.3 and add a line to 430 EXTRA_DIST to distribute ChangeLog-2006. 431 4322007-07-09 Maynard Johnson <maynardj@us.ibm.com> 433 434 * doc/opreport.xsd: 435 * libpp/xml_utils.cpp: Correct schema version bump 436 4372007-07-02 Maynard Johnson <maynardj@us.ibm.com> 438 439 * libpp/populate_for_spu.cpp: Fix logic in is_spu_profile() 440 to handle '--separate=thread' option. 441 4422007-06-27 Maynard Johnson <maynardj@us.ibm.com> 443 444 * ChangeLog: ChangeLog cleanup 445 4462007-06-15 Maynard Johnson <maynardj@us.ibm.com> 447 448 * doc/opreport.xsd: 449 * libpp/xml_utils.cpp: Bump schema version to 2.0 to coincide with 450 callgraph elements added recently to the schema 451 4522007-06-13 Maynard Johnson <maynardj@us.ibm.com> 453 454 * doc/oprofile.xml: 455 * events/Makefile.am: 456 * events/ppc64/power5++/event_mappings: 457 * events/ppc64/power5++/events: 458 * events/ppc64/power5++/unit_masks: 459 * libop/op_cpu_type.c: 460 * libop/op_cpu_type.h: 461 * libop/op_events.c: 462 * utils/opcontrol: 463 * utils/ophelp.c: Add support for POWER5++ (i.e., POWER5+ with PMU 464 set to POWER6 mode) 465 4662007-06-12 dean gaudet <dean@arctic.org> 467 468 * events/i386/core_2/events: 469 * events/i386/core_2/unit_masks: updates to match 470 253669-022US ISA vol 3B from Intel. (Including fix 471 for #1580064 from Arun Sharma.) 472 4732007-06-08 dean gaudet <dean@arctic.org> 474 475 * events/x86-64/hammer/events: 476 * events/x86-64/hammer/unit_masks: updates from December 2006 477 AMD update 478 4792007-05-31 Maynard Johnson <maynardj@us.ibm.com> 480 481 * libpp/symbol.h: 482 * libutil++/bfd_support.h: 483 * m4/cellspubfdsupport.m4: Fixups for compile failures on older Linux 484 distributions 485 4862007-05-29 Dave Nomura <dcnltc@us.ibm.com> 487 488 * libregex/demangle_symbol.cpp: Strip off leading '.' generated for elf6r-powerpc targets in mangled symbol names. 489 4902007-05-24 Richard Purdie <rpurdie@openedhand.com> 491 492 * events/arm/xscale1/events: 493 * events/arm/xscale2/events: Add extra Xscale PMU event definitions 494 4952007-05-24 Richard Purdie <rpurdie@openedhand.com> 496 497 * events/arm/armv6/events: 498 * events/arm/armv6/unit_masks: 499 * libop/op_cpu_type.c: 500 * libop/op_cpu_type.h: 501 * libop/op_events.c: 502 * utils/ophelp.c: Name ARM11 support to be consistent with the kernel, 503 remove some duplicate code and add some extra events 504 5052007-05-24 Richard Purdie <rpurdie@openedhand.com> 506 507 * oprofile/pp/oparchive.cpp: If the debug files have the same name 508 as the original binary, the binary gets overwritten. Add in a .debug 509 directory to avoid this. 510 5112007-05-23 Joerg Wagner <wagner@ccrl-nece.de> 512 513 * events/Makefile.am: 514 * events/arm/mpcore/events: 515 * events/arm/mpcore/unit_masks: 516 * libop/op_cpu_type.c: 517 * libop/op_cpu_type.h: 518 * libop/op_events.c: 519 * utils/ophelp.c: ARM MPCore support 520 5212007-05-23 Riku Voipio <riku.voipio@iki.fi> 522 523 * events/Makefile.am: 524 * events/arm/armv6/events: 525 * events/arm/armv6/unit_masks: 526 * libop/op_cpu_type.c: 527 * libop/op_cpu_type.h: 528 * libop/op_events.c: 529 * utils/ophelp.c: ARM11 support 530 5312007-05-21 Richard Purdie <rpurdie@openedhand.com> 532 533 * doc/opreport.xsd: 534 * libpp/format_output.cpp: 535 * libpp/format_output.h: 536 * libpp/xml_utils.cpp: 537 * libutil++/xml_output.cpp: 538 * libutil++/xml_output.h: 539 * pp/opreport.cpp: 540 * pp/opreport_options.cpp: Add callgraph XML output 541 5422007-05-20 Dean Gaudet <dgaudet@users.sourceforge.net> 543 544 * events/i386/core_2/events: ITLB_MISS_RETIRED,MEM_LOAD_RETIRED 545 only work with counter 0 546 5472007-05-20 Richard Purdie <rpurdie@openedhand.com> 548 549 * libpp/callgraph_container.cpp: 550 * libpp/callgraph_container.h: 551 * libpp/format_output.cpp: 552 * libpp/format_output.h: 553 * libpp/symbol.h: 554 * libpp/symbol_sort.cpp: 555 * libpp/symbol_sort.h: 556 * pp/opreport.cpp: Convert cg_collection to symbol_collection and use 557 dynamic casting allowing more code reuse for callgraphs 558 5592007-05-19 Richard Purdie <rpurdie@openedhand.com> 560 561 * doc/oparchive.1.in: 562 * doc/oprofile.xml: 563 * pp/oparchive.cpp: 564 * pp/oparchive_options.cpp: 565 * pp/oparchive_options.h: Add --list-files option to list all files 566 that would be handled by oparchive. 567 5682007-05-19 Olof Johansson <olof@lixom.net> 569 570 * events/Makefile.in: 571 * events/ppc64/pa6t/event_mappings: 572 * events/ppc64/pa6t/events: 573 * events/ppc64/pa6t/unit_masks: 574 * libop/op_cpu_type.c: 575 * libop/op_cpu_type.h: 576 * libop/op_events.c: 577 * utils/opcontrol: 578 * utils/ophelp.c: PA Semi 6T support 579 5802007-05-09 Richard Purdie <rpurdie@openedhand.com> 581 582 * oprofile/daemon/opd_cookie.c: Fix syscall for ARM EABI 583 5842007-05-10 Maynard Johnson <maynardj@us.ibm.com> 585 586 * libpp/Makefile.am: 587 * libpp/populate.cpp: 588 * libpp/populate_for_spu.cpp: 589 * libpp/populate_for_spu.h: 590 * libpp/profile.h: 591 * libpp/profile.cpp: 592 * libutil++/Makefile.am: 593 * libutil++/bfd_spu_support.cpp: 594 * libutil++/bfd_support.h: 595 * libutil++/op_bfd.h: 596 * libutil++/op_spu_bfd.cpp: 597 * libpp/xml_utils.cpp: 598 * libpp/profile_container.cpp: 599 * libpp/symbol.h: Patch 3 of 3 for adding support for profiling 600 Cell Broadband Engine SPU 601 6022007-05-10 Maynard Johnson <maynardj@us.ibm.com> 603 604 * daemon/liblegacy/opd_sample_files.c: 605 * daemon/opd_events.c: 606 * daemon/opd_events.h: 607 * daemon/opd_interface.h: 608 * daemon/Makefile.am: 609 * daemon/opd_mangling.c: 610 * daemon/opd_sfile.c: 611 * daemon/opd_sfile.h: 612 * daemon/opd_spu.c: 613 * daemon/opd_trans.c: 614 * daemon/opd_trans.h: 615 * libop/op_sample_file.h: Patch 2 of 3 for adding support for profiling 616 Cell Broadband Engine SPU 617 6182007-05-10 Maynard Johnson <maynardj@us.ibm.com> 619 620 * configure.in: 621 * doc/oprofile.xml: 622 * events/ppc64/cell-be/events: 623 * m4/Makefile.am: 624 * m4/cellspubfdsupport.m4: 625 * utils/opcontrol: Patch 1 of 3 for adding support for profiling 626 Cell Broadband Engine SPU 627 6282007-04-25 Manoj Ekbote <manoj.ekbote@broadcom.com> 629 630 * events/mips/sb1/events: fix SB1 events 631 6322007-04-19 Dave Nomura <dcnltc@us.ibm.com> 633 634 * events/ppc64/power6/events: counter 3 should have been used 635 in the specification of the default CYCLES event, not counter 1 636 6372007-04-13 Will Cohen <wcohen@redhat.com> 638 639 * libop/op_cpu_type.c: 640 * libop/op_cpu_type.h: Move new entries to end of list 641 to improve backward compatibility with enum values. 642 6432007-04-11 John Levon <levon@movementarian.org> 644 645 * doc/oprofile.xml: IRC channel is on OFTC 646 6472007-04-10 Philippe Elie <phil.el@wanadoo.fr> 648 649 * utils/ophelp.c: if userspace doesn't recognize the cpu type 650 warn the user it can use timer mode or upgrade oprofile. 651 6522007-04-04 Maynard johnson <maynardj@us.ibm.com> 653 654 * libutil++/file_manip.cpp: catch result from chown() to avoid 655 compilation warning (treated as error) 656 6572007-03-23 Jason Yeh <jason.yeh@amd.com> 658 659 * events/Makefile.am: 660 * events/x86-64/family10/events: 661 * events/x86-64/family10/unit_masks: 662 * libop/op_cpu_type.c: 663 * libop/op_cpu_type.h: 664 * libop/op_events.c: 665 * utils/ophelp.c: Add AMD Family 10 support. 666 6672007-02-28 Rob Bradford <rob@o-hand.com> 668 669 * pp/oparchive.cpp: return from main 670 6712007-02-21 Rob Bradford <rob@o-hand.com> 672 673 Reverted previous patches. 674 * pp/oparchive.cpp: Save debuginfo files in .debug sub directory. 675 676 * pp/oparchive.cpp: 677 * pp/oparchive_options.h: 678 * pp/oparchive_options.cpp: List files option. 679 6802007-02-21 Rob Bradford <rob@o-hand.com> 681 682 * pp/oparchive.cpp: Save debuginfo files in .debug sub directory. 683 684 * pp/oparchive.cpp: 685 * pp/oparchive_options.h: 686 * pp/oparchive_options.cpp: List files option. 687 6882007-02-16 Philippe Elie <phil.el@wanadoo.fr> 689 690 * daemon/opd_mangling.c: missing initialisation when mangling an 691 anon and callgraph filename. Fix given by Amitabha Roy. 692 6932007-02-06 Dave Nomura <dcnltc@us.ibm.com> 694 695 * libpp/symbol.h: 696 * libutil++/Makefile.am: 697 * libutil++/sparse_array.h: represent count_array_t as a sparse 698 array type based on std::map rather than growable_vector which is 699 based on std::vector to reduce the excessive heap usage on large 700 profiles. 701 7022007-02-02 Amitabha Roy <amitabha.roy@gmail.com> 703 704 * daemon/opd_anon.c: 705 * daemon/opd_anon.h: 706 * daemon/opd_mangling.c: 707 * libop/op_mangle.c: 708 * libop/op_mangle.h: 709 * libpp/parse_filename.cpp: save and report name of anonymous 710 mapping if there is one (as self-contradictory as that 711 sounds). 712 7132007-02-02 Dave Nomura <dcnltc@us.ibm.com> 714 715 * events/Makefile.am: 716 * events/ppc64/970MP/event_mappings: 717 * events/ppc64/970MP/events: 718 * events/ppc64/970MP/unit_masks: 719 * libop/op_cpu_type.c: 720 * libop/op_cpu_type.h: 721 * libop/op_events.c: 722 * utils/ophelp.c: PPC970MP has different hardware counters than the rest 723 of the PPC family and must be treated as a different architecture 724 by oprofile. 725 7262007-01-31 Dave Nomura <dcnltc@us.ibm.com> 727 728 * libpp/xml_utils.cpp: opreport -X was still getting some incorrect 729 symbols attributed to the <binary> when processing a --separate=lib 730 profile. 731 7322007-01-29 Philippe Elie <phil.el@wanadoo.fr> 733 734 * daemon/oprofiled.c: no need for 0755 for the log file mode, use 0644 735 * utils/opcontrol: don't create the log in opcontrol 736 7372007-01-26 Dave Nomura <dcnltc@us.ibm.com> 738 739 * libpp/format_output.cpp: 740 * libpp/xml_utils.cpp: opreport -X was mixing the symbols associated with 741 an application with those of a library when processing a --separate=lib 742 profile. 743 7442007-01-02 Dave Nomura <dcnltc@us.ibm.com> 745 746 * events/Makefile.am: 747 * events/ppc64/power6/event_mappings: 748 * events/ppc64/power6/events: 749 * events/ppc64/power6/unit_masks: 750 * libop/op_cpu_type.c: 751 * libop/op_cpu_type.h: 752 * libop/op_events.c: 753 * utils/opcontrol: 754 * utils/ophelp.c: Add support for Power6 755 756 757See ChangeLog-2006 for earlier changelogs. 758