Hi, It's a bug that every body have or is it dependent of my machine ? Feedback would be welcome.
Sylvere Teissier a écrit : > > > Tristan Gingold wrote: >> On Wed, Jul 29, 2009 at 10:33:25AM +0200, Sylvere Teissier wrote: >> >>> Le vendredi 20 février 2009 à 23:30 +0100, Sylvere Teissier a écrit : >>> >>>> Arnim Läuger wrote: >>>> >>>>> What is the exact error message? >>>>> I just gave it a try with ghdl 0.27 and it analyzed a file without any >>>>> problems: >>>>> $ ghdl -a --std=93c --workdir=work -Wc,-ftest-coverage >>>>> -Wc,-fprofile-arcs ../src/T80/T80_MCode.vhd >>>>> $ ll >>>>> work >>>>> T80_MCode.gcno >>>>> >>>> Analysis works too on my platform (debian lenny - >>>> ghdl0.27+svn110+gcc4.3.2), >>>> but I have a "seg fault" when running gcov after design execution :( >>>> >>> Am I the only one to have this problem, or is it a general bug ? >>> >>> Code Coverage is an important feature, if this is a bug how can I help >>> to solve it? >>> >> >> Hi Sylvere, >> >> I have tried gcov with ghdl 0.28 and it worked. One possible issue is >> that >> your gcc version must be the same as the one used by ghdl (ie 4.3.x). >> Can you check that ? >> Maybe this constraint is too hard, and maybe I should include a ghdl-gcov >> tool. >> > It nearly works: > > ---------------------------- > test.gcno:no lines for 'work__test__ELAB' > File 'test.vhd' > Lines executed:91.67% of 12 > test.vhd:creating 'test.vhd.gcov' > > File '' > No executable lines > :creating '.gcov' > :cannot open source file > ------------------------- > > The generated test.vhd.gcov is correct. But there is an error in the > gcno in "work__test__ELAB": > if I run gcov-dump I get: > > > test.gcno:note:magic `gcno':version `403*' > test.gcno:stamp 3898939573 > test.gcno:3: 01000000: 13:FUNCTION ident=3, checksum=0xee98a117, > `work__test__ELAB' test.vhd:0 > test.gcno:18: 01410000: 3:BLOCKS 3 blocks > test.gcno:20: 0 0000 0000 0000 > test.gcno:23: 01430000: 3:ARCS 1 arcs > test.gcno:26: block 0: 1:0004 > test.gcno:28: 01430000: 3:ARCS 1 arcs > test.gcno:31: block 1: 2:0001 > test.gcno:33: 01450000: 10:LINES > test.gcno:36: block 1:`test.vhd' > test.gcno:41: block 1:`' > test.gcno:record size mismatch 1 bytes overread > test.gcno:45: 01000000: 15:FUNCTION ident=4, checksum=0x54cfe1c5, > `work__test__ARCH__test__f' test.vhd:1 > test.gcno:62: 01410000: 3:BLOCKS 3 blocks > test.gcno:64: 0 0000 0000 0000 > test.gcno:67: 01430000: 3:ARCS 1 arcs > test.gcno:70: block 0: 1:0004 > test.gcno:72: 01430000: 3:ARCS 1 arcs > test.gcno:75: block 1: 2:0001 > test.gcno:77: 01450000: 10:LINES > test.gcno:80: block 1:`test.vhd':1, 8 > test.gcno:89: 01000000: 17:FUNCTION ident=5, checksum=0x6e31d39d, > `work__test__ARCH__test__P0__PROC' test.vhd:8 > test.gcno:108: 01410000: 28:BLOCKS 28 blocks > test.gcno:110: 0 0000 0000 0000 0000 0000 0000 0000 0000 > test.gcno:118: 8 0000 0000 0000 0000 0000 0000 0000 0000 > test.gcno:126: 16 0000 0000 0000 0000 0000 0000 0000 0000 > test.gcno:134: 24 0000 0000 0000 0000 > test.gcno:138: 01430000: 3:ARCS 1 arcs > test.gcno:141: block 0: 1:0005 > > As you can see there an error at test.gcno:41 and that's why gcov search > and not find < File '' > > > -------various informations: > > $ghdl -v > GHDL 0.28 (20090917) [Sokcho edition] > Compiled with GNAT Version: 4.3.4 > GCC back-end code generator > > -------------------------------------------- > > $/usr/lib/ghdl/bin/gcov -v > gcov (GCC) 4.3.3 > > --------------------------------------- > build steps: > > ghdl -a -Wc,-ftest-coverage -Wc,-fprofile-arcs test.vhd > ghdl -e -Wl,-lgcov test > ./test > /usr/lib/ghdl/bin/gcov test > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > Ghdl-discuss mailing list > [email protected] > https://mail.gna.org/listinfo/ghdl-discuss -- Sylvère Teissier INVIA Europarc Sainte Victoire - Bât 6 13590 MEYREUIL FRANCE Tel +33 4 42 24 50 77 Fax +33 4 42 24 50 71 Web: www.invia.fr
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Ghdl-discuss mailing list [email protected] https://mail.gna.org/listinfo/ghdl-discuss
