Improve travis build #311
Improve travis build #311
Conversation
sjaeckel
force-pushed
the
improve/mpi_selection
branch
from
bc6718c to
9d5a471
Compare
sjaeckel
force-pushed
the
improve/travis_build
branch
from
3e05077 to
221b925
Compare
sjaeckel
force-pushed
the
improve/travis_build
branch
from
221b925 to
fa9be2d
Compare
| install: | ||
| - sudo apt-get update -qq | ||
| - sudo apt-get install libtommath-dev | ||
| - sudo apt-get install libtommath-dev libgmp-dev valgrind |
There was a problem hiding this comment.
This is what was needed in ecc branch (latest ltm + tfm)
https://github.com/libtom/libtomcrypt/blob/pr/ecc-asn1-part/.travis.yml#L22
There was a problem hiding this comment.
yeah, I'd like to stay on standard-built packages as long as possible as the packages that are installed in the ecc branch are built by me ;)
| make clean &>/dev/null | ||
|
|
||
| make CFLAGS="-fsanitize=address -fno-omit-frame-pointer -static-libasan $2 $CFLAGS $4" EXTRALIBS="-lasan $5" test LTC_DEBUG=1 1>gcc_1.txt 2>gcc_2.txt | ||
| ASAN_OPTIONS=verbosity=1 ./test t ltm 1>test_std.txt 2> test_err.txt |
There was a problem hiding this comment.
shouldn't ASAN go into run_clang ?
There was a problem hiding this comment.
I thought so too ... then I wasn't able to get the ASAN build done with clang which can be installed in xenial ... but it worked with gcc, so I thought let's give it a try...
| } | ||
| #endif | ||
|
|
||
| int crypt_mp_init(const char* mpi) |
There was a problem hiding this comment.
I am still not sure if we need this. I am not strongly against, just in doubts.
There was a problem hiding this comment.
No we absolutely don't need it, but I thought it's better than the other init functions that exist, so we just remove them ASAP (that's also why I marked them already as deprecated)
There was a problem hiding this comment.
damn, now I know why this change is in this PR, I changed the base branch :(
There was a problem hiding this comment.
let's focus on the other PR first, I've still to get this one to build
| #endif | ||
|
|
||
| #ifndef LTC_DEPRECATED | ||
| #error "You need to define LTC_DEPRECATED for this compiler" |
There was a problem hiding this comment.
I do not like this. It is very unfriendly to non-gcc/non-msvc compilers.
There was a problem hiding this comment.
I hoped you're going to build this on "the other compilers" and then either fill it in or define it to nothing if the "deprecation" feature doesn't exist for "the other compilers"
There was a problem hiding this comment.
Consider:
#ifdef __GNUC__
#define LTC_DEPRECATED __attribute__((deprecated))
#elif defined(_MSC_VER)
#define LTC_DEPRECATED __declspec(deprecated)
#else
#define LTC_DEPRECATED
#endif
It might also cause troubles on older gcc as I do not know since what version this attribute is supported.There was a problem hiding this comment.
It seems to be supported on gcc 3.1+
There was a problem hiding this comment.
#if (__GNUC__ > 3) || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)
| @@ -1,5 +1,11 @@ | |||
| #!/bin/bash | |||
There was a problem hiding this comment.
Is it worth a separate script (scan_build.sh)? What about incorporating the stuff into meta_builds.sh?
There was a problem hiding this comment.
nope, I just re-used it as it mostly worked OOTB - I'm going to incorporate it.
sjaeckel
force-pushed
the
improve/mpi_selection
branch
from
343a46d to
d709e3d
Compare
|
closed? what about |
|
github closed this when I deleted the base branch and somehow I can't re-open... @github |
|
try to rebase on develop locally and |
|
did that already (did it now once again) but the reopen button is greyed out |
Introduce a new meta_build where we build several things in one spin-up of a travis instance