Allegro.cc - Online Community

Allegro.cc Forums » Allegro Development » Building profiling version of allegro fails with undefined reference to mcount

This thread is locked; no one can reply to it. rss feed Print
Building profiling version of allegro fails with undefined reference to mcount
Edgar Reynaldo
Member #8,592
May 2007
avatar

I wanted to try and figure out why bitmap locking is so dreadfully slow in OpenGL on my laptop but I can't build the profiling version of allegro due to undefined references to mcount.

I believe both the source modules and the library need to be linked with -pg for this to work correctly. I don't know where to look in the cmake files to fix this though. Any ideas?

Quote:

#SelectExpand
1/* ... */ 2 3Linking CXX shared library lib\allegro_monolith-profile-5.1.dll 4Creating library file: lib\liballegro_monolith-profile.dll.a 5CMakeFiles\allegro_monolith.dir/objects.a(allegro.c.obj):allegro.c:(.text+0x4): undefined reference to `mcount' 6CMakeFiles\allegro_monolith.dir/objects.a(allegro.c.obj):allegro.c:(.text+0x16): undefined reference to `mcount' 7CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x7): undefined reference to `mcount' 8CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x183): undefined reference to `mcount' 9CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x1f7): undefined reference to `mcount' 10CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x4ef): more undefined references to `mcount' follow 11collect2: ld returned 1 exit status 12mingw32-make[2]: *** [lib/allegro_monolith-profile-5.1.dll] Error 1 13mingw32-make[1]: *** [CMakeFiles/allegro_monolith.dir/all] Error 2 14mingw32-make: *** [all] Error 2 15 16c:\mingw\LIBS\A5GIT\allegro\build>

SiegeLord
Member #7,827
October 2006
avatar

There is a Profile build type. Did you try it?

"For in much wisdom is much grief: and he that increases knowledge increases sorrow."-Ecclesiastes 1:18
[SiegeLord's Abode][Codes]:[DAllegro5]:[RustAllegro]

Edgar Reynaldo
Member #8,592
May 2007
avatar

Sorry, yes. I should have mentioned the cmake parameters I used :

#SelectExpand
1c:\ctwoplus\progcode\allegro5>cd %a5git% 2 3c:\mingw\LIBS\A5GIT\allegro\build>cmake -G "MinGW Makefiles" -DSHARED=On -DCMAKE_BUILD_TYPE=Profile -DWANT_DOCS=Off -DWANT_DEMO=Off -DWANT_EXAMPLES=Off -DWANT_MONOLITH=On .. 4-- Using MINGDIR: C:/mingw 5-- Allowing GCC/Clang to use SSE instructions 6-- Could NOT find PkgConfig (missing: PKG_CONFIG_EXECUTABLE) 7-- Could NOT find OPENSL (missing: OPENSL_INCLUDE_DIR OPENSL_LIBRARY) 8WARNING: Dynamic loading will be disabled for FLAC as static library was found: C:/mingw/lib/libFLAC.a 9WARNING: Dynamic loading will be disabled for DUMB as static library was found: C:/mingw/lib/libdumb.a 10-- DLL name for C:/mingw/lib/libvorbisfile.dll.a: libvorbisfile-3.dll 11-- Configuring done 12-- Generating done 13-- Build files have been written to: C:/mingw/LIBS/A5GIT/allegro/build 14 15c:\mingw\LIBS\A5GIT\allegro\build>mingw32-make 16[ 0%] Building C object CMakeFiles/allegro_monolith.dir/src/allegro.c.obj 17 18/*...*/ 19 20[ 99%] Building C object CMakeFiles/allegro_monolith.dir/addons/main/generic_main.c.obj 21Linking CXX shared library lib\allegro_monolith-profile-5.1.dll 22Creating library file: lib\liballegro_monolith-profile.dll.a 23CMakeFiles\allegro_monolith.dir/objects.a(allegro.c.obj):allegro.c:(.text+0x4): undefined reference to `mcount' 24CMakeFiles\allegro_monolith.dir/objects.a(allegro.c.obj):allegro.c:(.text+0x16): undefined reference to `mcount' 25CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x7): undefined reference to `mcount' 26CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x183): undefined reference to `mcount' 27CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x1f7): undefined reference to `mcount' 28CMakeFiles\allegro_monolith.dir/objects.a(bitmap.c.obj):bitmap.c:(.text+0x4ef): more undefined references to `mcount' follow 29collect2: ld returned 1 exit status 30mingw32-make[2]: *** [lib/allegro_monolith-profile-5.1.dll] Error 1 31mingw32-make[1]: *** [CMakeFiles/allegro_monolith.dir/all] Error 2 32mingw32-make: *** [all] Error 2 33 34c:\mingw\LIBS\A5GIT\allegro\build>

Yeah, I also tried it without the monolith and it fails the same way.

Edit
I looked through CMakeLists.txt and I found this :

Quote:

CMakeLists.txt#SelectExpand
393# Profiling. 394 395list(APPEND CMAKE_BUILD_CONFIGURATIONS Profile) 396if(COMPILER_GCC_OR_CLANG) 397 set(CMAKE_C_FLAGS_PROFILE "-pg" 398 CACHE STRING "profiling flags") 399 set(CMAKE_CXX_FLAGS_PROFILE "${CMAKE_C_FLAGS_PROFILE}" 400 CACHE STRING "profiling flags") 401 set(CMAKE_EXE_LINKER_FLAGS_PROFILE "-pg" 402 CACHE STRING "profiling flags") 403 mark_as_advanced( 404 CMAKE_C_FLAGS_PROFILE 405 CMAKE_CXX_FLAGS_PROFILE 406 CMAKE_EXE_LINKER_FLAGS_PROFILE 407 ) 408endif(COMPILER_GCC_OR_CLANG)

It sets a flag for the source modules and for the linker so I'm not sure what is wrong. Perhaps the library needs a separate linker parameter in cmake?

Thomas Fjellstrom
Member #476
June 2000
avatar

There is a separate linker flag for shared libs iirc.

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

Edgar Reynaldo
Member #8,592
May 2007
avatar

From
http://www.cmake.org/pipermail/cmake/2011-December/048040.html

Quote:

On 12/12/2011 09:00 PM, Robert Dailey wrote:

Quote:

I need a version of CMAKE_EXE_LINKER_FLAGS for shared libraries. I need
to specify /NODEFAULTLIB for a shared library (DLL) that I'm building,
but it doesn't seem that CMAKE_EXE_LINKER_FLAGS is used by shared
library projects generated by CMake for visual studio.

Any suggestions?

Use CMAKE_SHARED_LINKER_FLAGS. Strangely, this family of variables is
missing from the docs...

Michael

So we need to use CMAKE_SHARED_LINKER_FLAGS somewhere? Like in the profiling section above? cmake confuses me. :P

Thomas Fjellstrom
Member #476
June 2000
avatar

Pretty much yeah.

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

pkrcel
Member #14,001
February 2012

so we need to use CMAKE_SHARED_LINKER_FLAGS somewhere? Like in the profiling section above? cmake confuses me. :P

You should use CMAKE_SHARED_LINKER_FLAGS instead of CMAKE_EXE_LINKER_FLAGS.

Try to use the CMAKE_VERBOSE_MAKEFILE option so we see the actual link command passed to the command line (warning: LOTS of text).

It is unlikely that Google shares your distaste for capitalism. - Derezo
If one had the eternity of time, one would do things later. - Johan Halmén

Thomas Fjellstrom
Member #476
June 2000
avatar

You could just pass: VERBOSE=1 to make. if you want to see the commands make runs.

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

pkrcel
Member #14,001
February 2012

I think it does just the same....but I'll try it out ;D

It is unlikely that Google shares your distaste for capitalism. - Derezo
If one had the eternity of time, one would do things later. - Johan Halmén

Edgar Reynaldo
Member #8,592
May 2007
avatar

pkrcel said:

You should use CMAKE_SHARED_LINKER_FLAGS instead of CMAKE_EXE_LINKER_FLAGS.

I would assume we still need CMAKE_EXE_LINKER_FLAGS though, as the examples and demos would all need to be linked with -pg.

I got it to build by changing the above section from CMakeLists.txt to the below :

CMakeLists.txt#SelectExpand
393# Profiling. 394 395list(APPEND CMAKE_BUILD_CONFIGURATIONS Profile) 396if(COMPILER_GCC_OR_CLANG) 397 set(CMAKE_C_FLAGS_PROFILE "-pg" 398 CACHE STRING "profiling flags") 399 set(CMAKE_CXX_FLAGS_PROFILE "${CMAKE_C_FLAGS_PROFILE}" 400 CACHE STRING "profiling flags") 401 set(CMAKE_EXE_LINKER_FLAGS_PROFILE "-pg" 402 CACHE STRING "profiling flags") 403 set(CMAKE_SHARED_LINKER_FLAGS_PROFILE "-pg" 404 CACHE STRING "profiling flags") 405 mark_as_advanced( 406 CMAKE_C_FLAGS_PROFILE 407 CMAKE_CXX_FLAGS_PROFILE 408 CMAKE_EXE_LINKER_FLAGS_PROFILE 409 CMAKE_SHARED_LINKER_FLAGS_PROFILE 410 ) 411endif(COMPILER_GCC_OR_CLANG)

I kept CMAKE_EXE_LINKER_FLAGS_PROFILE and added CMAKE_SHARED_LINKER_FLAGS_PROFILE as -pg and added it to the mark_as_advanced section below (I have no idea what that does).

Everything seems to build fine, all the demo, examples and library worked fine. When run, the exe's produce gmon.out like they're supposed to but gprof is giving me some messed up results, and doesn't even include the allegro functions in the profile, which I can't explain. I'm trying to upgrade my compiler as we speak, prob won't finish til tomorrow though.

Yeah, the output from gprof is not making any sense at all. For example, I ran ex_threads and then ran gprof on it :

#SelectExpand
1Flat profile: 2 3Each sample counts as 0.01 seconds. 4 no time accumulated 5 6 % cumulative self self total 7 time seconds seconds calls Ts/call Ts/call name 8 0.00 0.00 0.00 26797 0.00 0.00 log_printf 9 10 % the percentage of the total running time of the 11time program used by this function. 12 13cumulative a running sum of the number of seconds accounted 14 seconds for by this function and those listed above it. 15 16 self the number of seconds accounted for by this 17seconds function alone. This is the major sort for this 18 listing. 19 20calls the number of times this function was invoked, if 21 this function is profiled, else blank. 22 23 self the average number of milliseconds spent in this 24ms/call function per call, if this function is profiled, 25 else blank. 26 27 total the average number of milliseconds spent in this 28ms/call function and its descendents per call, if this 29 function is profiled, else blank. 30 31name the name of the function. This is the minor sort 32 for this listing. The index shows the location of 33 the function in the gprof listing. If the index is 34 in parenthesis it shows where it would appear in 35 the gprof listing if it were to be printed. 36 37 Call graph (explanation follows) 38 39 40granularity: each sample hit covers 4 byte(s) no time propagated 41 42index % time self children called name 43 0.00 0.00 279/26797 animate_square [140] 44 0.00 0.00 2109/26797 gen_square [154] 45 0.00 0.00 24409/26797 thread_func [177] 46[2] 0.0 0.00 0.00 26797 log_printf [2] 47----------------------------------------------- 48 49 This table describes the call tree of the program, and was sorted by 50 the total amount of time spent in each function and its children. 51 52 Each entry in this table consists of several lines. The line with the 53 index number at the left hand margin lists the current function. 54 The lines above it list the functions that called this function, 55 and the lines below it list the functions this one called. 56 This line lists: 57 index A unique number given to each element of the table. 58 Index numbers are sorted numerically. 59 The index number is printed next to every function name so 60 it is easier to look up where the function in the table. 61 62 % time This is the percentage of the `total' time that was spent 63 in this function and its children. Note that due to 64 different viewpoints, functions excluded by options, etc, 65 these numbers will NOT add up to 100%. 66 67 self This is the total amount of time spent in this function. 68 69 children This is the total amount of time propagated into this 70 function by its children. 71 72 called This is the number of times the function was called. 73 If the function called itself recursively, the number 74 only includes non-recursive calls, and is followed by 75 a `+' and the number of recursive calls. 76 77 name The name of the current function. The index number is 78 printed after it. If the function is a member of a 79 cycle, the cycle number is printed between the 80 function's name and the index number. 81 82 83 For the function's parents, the fields have the following meanings: 84 85 self This is the amount of time that was propagated directly 86 from the function into this parent. 87 88 children This is the amount of time that was propagated from 89 the function's children into this parent. 90 91 called This is the number of times this parent called the 92 function `/' the total number of times the function 93 was called. Recursive calls to the function are not 94 included in the number after the `/'. 95 96 name This is the name of the parent. The parent's index 97 number is printed after it. If the parent is a 98 member of a cycle, the cycle number is printed between 99 the name and the index number. 100 101 If the parents of the function cannot be determined, the word 102 `<spontaneous>' is printed in the `name' field, and all the other 103 fields are blank. 104 105 For the function's children, the fields have the following meanings: 106 107 self This is the amount of time that was propagated directly 108 from the child into the function. 109 110 children This is the amount of time that was propagated from the 111 child's children to the function. 112 113 called This is the number of times the function called 114 this child `/' the total number of times the child 115 was called. Recursive calls by the child are not 116 listed in the number after the `/'. 117 118 name This is the name of the child. The child's index 119 number is printed after it. If the child is a 120 member of a cycle, the cycle number is printed 121 between the name and the index number. 122 123 If there are any cycles (circles) in the call graph, there is an 124 entry for the cycle-as-a-whole. This entry shows who called the 125 cycle (as parents) and the members of the cycle (as children.) 126 The `+' recursive calls entry shows the number of function calls that 127 were internal to the cycle, and the calls entry for each member shows, 128 for that member, how many times it was called from other members of 129 the cycle. 130 131Index by function name 132 133 [2] log_printf

Only log_printf is listed in the flat profile, and none of the allegro functions even show up!

SiegeLord
Member #7,827
October 2006
avatar

You are using WinXP, right? There are some other profiler options but they tend to be Vista+.

"For in much wisdom is much grief: and he that increases knowledge increases sorrow."-Ecclesiastes 1:18
[SiegeLord's Abode][Codes]:[DAllegro5]:[RustAllegro]

pkrcel
Member #14,001
February 2012

Do all the files contain debugging symbols? I am unsure the profile flags include '-g'

EDIT: well okay that should not matter at all. :-X

It is unlikely that Google shares your distaste for capitalism. - Derezo
If one had the eternity of time, one would do things later. - Johan Halmén

Edgar Reynaldo
Member #8,592
May 2007
avatar

My laptop is Vista, my desktop is XP. I do my dev work on my laptop. What other profilers are you talking about SiegeLord?

Allegro builds successfully, and both the source modules and the library are now linked with -pg, but gprof doesn't show any allegro functions in the flat profile. I don't get it, this used to work with allegro 4? What gives?

Edit
Submitted a patch to AD.

Anyone have any suggestions on what to replace gprof with? Or why gprof no longer works? I've read things that suggest gprof doesn't work with shared libraries but like I said this used to work with Allegro 4.

Arthur Kalliokoski
Second in Command
February 2005
avatar

Were the libraries also compiled with -pg?

“Throughout history, poverty is the normal condition of man. Advances which permit this norm to be exceeded — here and there, now and then — are the work of an extremely small minority, frequently despised, often condemned, and almost always opposed by all right-thinking people. Whenever this tiny minority is kept from creating, or (as sometimes happens) is driven out of a society, the people then slip back into abject poverty. This is known as "bad luck.”

― Robert A. Heinlein

Edgar Reynaldo
Member #8,592
May 2007
avatar

Yes, that's what the CMAKE_SHARED_LINKER_FLAGS_PROFILE patch was for. And the output from gprof is still useless. It says there were 6 calls to main, which is ridiculous and obviously wrong.

So, anyone know any good profiling tools for Windows?

Thomas Fjellstrom
Member #476
June 2000
avatar

LINKER_FLAGS aren't for compiling. lib source has to be compiled with -pg as well

--
Thomas Fjellstrom - [website] - [email] - [Allegro Wiki] - [Allegro TODO]
"If you can't think of a better solution, don't try to make a better solution." -- weapon_S
"The less evidence we have for what we believe is certain, the more violently we defend beliefs against those who don't agree" -- https://twitter.com/neiltyson/status/592870205409353730

Edgar Reynaldo
Member #8,592
May 2007
avatar

Yup. It's there already. The flag I mentioned was for building the library, and fixed the issue with unresolved references to mcount.

CMakeLists.txt said:

set(CMAKE_C_FLAGS_PROFILE "-pg" CACHE STRING "profiling flags")

SiegeLord
Member #7,827
October 2006
avatar

I tried compiling Allegro with profiling (i.e. -pg) but none of the examples ran after that treatment, so I'm not sure what is up.

As for other profilers, I just tried Very Sleepy profiler and it picked up Allegro functions (note that I had to compile with -gdwarf-2 to get it to recognize the debug symbols. Perhaps the same will help your profile build?).

"For in much wisdom is much grief: and he that increases knowledge increases sorrow."-Ecclesiastes 1:18
[SiegeLord's Abode][Codes]:[DAllegro5]:[RustAllegro]

Edgar Reynaldo
Member #8,592
May 2007
avatar

I tried VerySleepy last night, and when I tried to attach to a process or launch a new one it just quit, and the program didn't run either.

Support for GCC/mingw. You can now profile executables with embedded DWARF2 data and it should work. No special options are required for this, just compile with “-g” to make sure you have symbols present. You may also wish to use “-fno-omit-frame-pointer” to ensure correct callstacks, although Sleepy will generally work either way. You don’t need to use “-pg” or any rubbish. It can even walk correct stacks between a Microsoft DLL into a GCC one, which was harder than you’d think.

I tried to profile a simple test program with these directions, using -g and -fno-omit-frame-pointer and every time I try to launch the program from verysleepy the program crashes and VerySleepy quits. My test program doesn't crash when run by itself though. The last release of VerySleepy was in November 2011 so I don't have high hopes for a fix unless I figure out what is wrong by myself, but that is another whole project in itself which I don't really want to get into.

I would try CodeXL but that is Win7+.

I couldn't get VerySleepy to work with -gdwarf-2 either. :/

pkrcel
Member #14,001
February 2012

I've succeded in building gperftools in MSYS2...I'll try to play around with the CPU profiler on your example Bitmaplocking.cpp rebuilding it in 32bits exe with all the deps statically linked in.

I don't know if I'm doing something useful but trying should not hurt, I guess. :P

D'OH!

gperftools on windows have been ported only to a minimal libTmalloc, no sight of the CPU profiler :-X

It is unlikely that Google shares your distaste for capitalism. - Derezo
If one had the eternity of time, one would do things later. - Johan Halmén

Edgar Reynaldo
Member #8,592
May 2007
avatar

YAPTDWOW. (Yet another profiler that doesn't work on windows) :P

That's disappointing. I'm beginning to think it would be better to profile opengl on Linux and try to improve it's functioning there in the hopes of improving it on Windows too. :/

pkrcel
Member #14,001
February 2012

I'm beginning to think it would be better to profile opengl on Linux and try to improve it's functioning there in the hopes of improving it on Windows too. :/

I have the same impression, but me personally can only unse a VM and I doubt that could help a lot.

Nevertheless, the cumulative call stack could help....I discovered only now that I should have compiled with '-f-no-omit-frame-pointer' ...

It is unlikely that Google shares your distaste for capitalism. - Derezo
If one had the eternity of time, one would do things later. - Johan Halmén

Go to: