TODO: - Consider porting the 4.4bsd gprof (adding all the dylib stuff and order stuff). - Add multi-arch support for findcalls() and make work better (use disasm). Changes for the 5.10 release (the cctools-421 release): - Changed new_sample_set() in gprof.c to not read the samples one at a time but to allocate a buffer for the samples and read them all at once. This makes running gprof faster. Radar bug #2876478. Changes for the 5.10 release (the cctools-400 release): - Changed the Makefile back to again use the -dependency-file with gcc. Changes for the 5.10 release (the cctools-396 release): - Changed the Makefile to not use the -dependency-file with gcc as well as mwccppc. Changes for the 5.10 release (the cctools-385 release): - Fixed a bug in creating the gprof(1) -S order file in not picking up the file names for symbols. Changed count_N_SO_stabs() and load_files() in getnfile.c to ignore the N_SO stabs for directory names that end in '/'. These stabs were created by the compiler starting after DP4 (from Cheetah on). Radar bug #2748170. Changes for the 5.9 release (the cctools-373 release): - Maded the nl variable in gprof.c into a __private_extern__ to avoid the warning about "unused multiple definitions" when building with -twolevel_namespace as the default. Changes for the 5.9 release (the cctools-359 release): - Added -Wno-long-double to shutup the compiler for <architecture/m88k/fp_regs.h>. Changes for the 5.7 release (the cctools-327 release): - Changed back where gprof(1) picks up its the System Framework's private headers. It will continue to look in /System/Library/Frameworks/System.framework/PrivateHeaders and may someday will lookin in /MacOSX/System/Frameworks/System.framework/PrivateHeaders in some post-DP4 release. Changes for the 5.6 release (the cctools-323 release): - Changed back where gprof picks up its the System Framework's private headers back to /System/Library/Frameworks/System.framework/PrivateHeaders from /MacOSX/System/Frameworks/System.framework/PrivateHeaders for the Space release. This won't change till post-DP4. Changes the 5.3 release (the cctools-303 release): - Changed the 6 uses of short in the structs in scatter.c to int and 0x7fff to INT_MAX. The 'w' field was overflowing and causing it to crash. Radar bug #2377933. Changes the 5.3 release (the cctools-292 release): - Changed the return type of main() in gprof.c from void to int to remove a warning from the egcs compiler. Also changed the exit() call at the end to return(). - Added -I$(NEXT_ROOT)/System/Library/Frameworks/System.framework/PrivateHeaders to the Makefile as the egcs compiler's cpp no longer searches this by default. This is needed to pick up <streams/streams.h> for use with the rld interface. Changes for the 5.3 release (the cctools-285 release) - Removed the old sccsid and rcsid's from the source that was #if !defined(NeXT) - Picked up the fix for Radar bug #2287470 (see below). Updated this change to include ifdef __OPENSTEP__'s in the code and Makefile changes to use $(RC_OS). Changes for the 5.3 release, MacOS X bring up (the cctools-282 release): - Changed ifdef's __SLICK__ to __OPENSTEP__. - Changed the Makefile to allow for RC_OS=macos for MacOS X builds. - Added a few casts in places to get it to compile with the MetroWerks compiler without -relax_pointers. Changes for the 5.2 release (the cctools-281.2 release) - Moved text files gprof.flat and gprof.callg from /usr/lib to usr/share . The files gprof/Makefile and gprof/gprof.h were changed (Radar bug #2287470). Changes for the 5.2 release (the cctools-267 release): - Fixed a warning in new_sample_set() in gprof.c in a print format. Radar bug #2227608. - Removed the #ifdef notdef code in gprof.c that multiplied the hertz by 100 which had a comment about the kernel being broken. The kernel was fixed in Titan1P to sample correctly. Radar bug #2227608. Changes for the 5.1 release (the cctools-260 release): - Added -c to all the install commands in the Makefile. Changes for the 5.1 release (the cctools-258 release): - Added code for picking up the profrate from the 4.4bsd header. - Added code for checking the version in the 4.4bsd header. - Fixed a bug in new_sample_set() in setting the scale value which is a double from interger arithmetic of (a-b)/c when a-b = c-1 which causes the value to round to zero rather than getting .999999 which is much closer to the approxmate value of 1.0 . The fix was to cast the integer values to double in the calculation ((double)a-(double)b)/(double)c . - All of the above changes are under Radar bug #2204555. Changes for the 5.1 release (the cctools-253 release): - Changed the Makefile so /usr/ucb is not created for Rhapsody builds. Changes for the 5.1 release (the cctools-248 release): - Removed the copy of gmon.h in this directory. For Rhapsody gmon.h will be owned by some other project than cctools. That is the version in <mach-o/gmon.h> will not be installed. The new gmon.h has a different structure for the profile header. This effected gprof.h and gprof.c and #ifdef __SLICK__ code was added. Changes for the 5.1 release (the cctools-247 release): - Added MKDIRS if code to the Makefile to build native on Rhapsody. - Changed the Makefile to use gnutar instead of tar because it works on Rhapsody where tar does not. - Changed the Makefile to install gprof in /usr/bin for RC_OS teflon and in /usr/ucb for RC_OS nextstep (Radar 1682569). - Changed the code ifdef __TEFLON__ to ifndef __SLICK__ (where __TEFLON__ will nolonger be defined for Rhapsody builds) so the default builds will be native Rhapsody builds. The changes were to getnfile.c, gprof.c, scatter.c and gprof.h Changes for the 5.0 release (the cctools-235 release): - Picked up the copyrights from the OpenBSD sources and added them to the files here. No code was changed. Changes for the 4.0 release (the cctools-183 release): - Ifndef'ed __TEFLON__ out all rld uses. Changes for the 4.0 release (the cctools-170 release): - Changed the name of the time based order file from callt.order to time.order. Changes for the 4.0 release (the cctools-168 release): - Added producing a time based order file (callt.order) for use with shared pcsample buffers. - Allowed MH_DYLIB and MH_DYLINKER files to be given for the a.out file. Changes for the 4.0 release (the cctools-167 release): - Added a the following print statement to find_file() in scatter.c: fprintf(stderr, "In producing order files can't find module name " "for functions (make sure all modules are compiled -g and " "the program is not already ordered)\n"); in order to help people understand the common problems (bug #53173). Changes for the 4.0 release (the cctools-160 release): - Fixed two places in getnfile.c that had the wrong if test for the debug flag. Changes for the 4.0 release (the cctools-154 release): - Removed the use of nmedit now that libstuff has __private_extern__'s. Changes for the 4.0 release (the cctools-147 release): - Added support for profiling dyld loaded code. - Removed old code that loaded FVMLIBs symbol tables. Changes for the 3.1 release (the cctools-24 release): - Fixed bug #33796 which turns out every where but in scatter.c the debug stuff is conditional on the -d command line flag as well as the DEBUG macro. The fix was to add #undef DEBUG to scatter.c. Changes for the 3.1 release (the cctools-15 release): - Added multiple profile areas and rld loaded state support (bug #19664). - Major cleanup of gprof code to allow compiling with -Wall cleanly and remove a few vax'ism. Changes for the 3.1 release (the cctools-14 release): - Fixed bug #15775 in which the ncnt field is zero in the gmon.out file and causes gprof to core dump. Changes for the 3.1 release (the cctools-6 release): - Added fat file support (tossed mach-o.c and replaced it). Changes for the 3.0 release (the -52 compiler release): - Added ignoring LC_PREPAGE to mach-o.c (2 places). Changes for the 3.0 release (the -49 compiler release): - Changed the Makefile to meet the RC api. - Removed the mlist stuff. Changes for the 3.0 release (the -47 compiler release): - Changed an include in mach_o.c to match the header file change: m68k/mach/thread_status.h -> mach/m68k/thread_status.h Changes for the 3.0 release (the -44 compiler release): - Switch over to the new header file organization. This also involed changing thread_status structure names and constants. But then ifdef'ed off code in mach_o.c that delt with threads since it got warnings and wasn't used. Also Fixed a few warnings from the compiler. Changes for some 1991 release (the -34 compiler release): - Added installsrc, installIBMsrc and installGNUsrc targets to the Makefile. Changes for the Warp ?? release (the -24 compiler release): - Fixed a bug when the text symbol for the highest value was looked up it would print the message "binary lookup failed" then it returned NULL and later gprof would core dump. Their were two problems the first was in the function getsymtab() in gprof.c in the Mach-O case the last entry in the nl array did not have it's value set to -1. The other problem was that in the function nllookup() in lookup.c was searching to index "nname - 1" and not to nname which would have picked up the last nl entry with the -1 in it. - Changed DFN_DEPTH from 100 to 1000 as per bug report #4851. Changes for the 2.0 impulse X.X release (the -23 compiler release): - Changed the installed mode to not be writeable (755 to 555 and 644 to 444). - Now is linked with libsys. Changes for the 2.0 impulse X.X release (the -19 compiler release): - Fixed the modified mlist function that includes printing categories which did not allocate enough space. Changes for the 2.0 impulse X.X release (the -19 compiler release): - Put in the modified mlist function that includes printing categories. Changes for the 0.91 release (the -12 compiler release): - Added MH_PRELOAD to the types of files gprof will take. - Updated the printing of method names for for the new objective-C runtime. - Ifdef'ed off code in mach_o.c that delt with shared libraries. It was calling itself, read_mach_o(), totally wrong. This code could have never worked and since all the symbols are in the executable and since the kernel can't handle shared libs it is meaningless.