Qucs / qucs

Qucs Project official mirror
http://qucs.sourceforge.net/
GNU General Public License v2.0
1.15k stars 213 forks source link

0.0.18 build warnings openSUSE #52

Closed rdmitry closed 9 years ago

rdmitry commented 10 years ago

During preparing rpm packages for openSUSE I got next errors (our build system detects this warnings as build failure):

[  944s] ... testing for serious compiler warnings
[  944s]     (using /usr/lib/build/checks-data/check_gcc_output)
[  944s]     (using //.build.log)
[  944s] 
[  944s] I: Program causes undefined operation
[  944s]    (likely same variable used twiceand post/pre incremented in the same expression).
[  944s]    e.g. x = x++; Split it in two operations.
[  944s] W: qucs sequence-point tridiag.cpp:246
[  944s] 
[  944s] I: Program is likely to break with new gcc. Try -fno-strict-aliasing.
[  944s] W: qucs strict-aliasing-punning matlab_producer.cpp:50
[  944s] 
[  944s] I: Program returns random data in a function
[  944s] E: qucs no-return-in-nonvoid-function exportdialog.cpp:341

guitorri commented 10 years ago

Is that so that only the last one results in an error (E:)? For this error you can find a patch on the above commit.

I am looking at the other warnings.

I am interested in having the full build log. I appreciate if you can put it on a gist or elsewhere that I can access it.

Where can I find the compiler flags used buy the openSUSE packaging system?

guitorri commented 10 years ago

@rdmitry, can you please try to apply this patch

rdmitry commented 10 years ago

Sorry for delay. Compiler flags:

CFLAGS='-fmessage-length=0 -grecord-gcc-switches -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables'
CXXFLAGS='-fmessage-length=0 -grecord-gcc-switches -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables'

Full logs without patch: https://build.opensuse.org/build/home:Dmitry_R:branches:electronics/openSUSE_13.1/x86_64/qucs/_log

guitorri commented 10 years ago

I guess you are right. I wonder if c vector shouldn't be initialized instead.

guitorri commented 9 years ago

@rdmitry, you can find links to the patches on the checklist below the issue description.

guitorri commented 9 years ago

Patches merged.