Activity for Giulio Benetti
-
Giulio Benetti committed [55ade1] on Git
server.c: fix build failure as static library
-
feat(plat/zynqmp): fix section `coherent_ram' will not fit in region `RAM'
-
Fix build failure as static library
-
configure: fix AC_CHECK_FUNCS(EVP_sha224 EVP_sha384 ...) failure on static linking
-
configure: fix AC_CHECK_FUNCS(TLS_method TLSv1_method ...) failure on static linking
-
fix AC_CHECK_FUNCS(...) failure on static linking
-
fix AC_CHECK_FUNCS(...) failure on static linking
-
configure: fix AC_CHECK_FUNCS(EVP_sha224 EVP_sha384 ...) failure on static linking
-
configure: fix AC_CHECK_FUNCS(TLS_method TLSv1_method ...) failure on static linking
-
configure: fix AC_CHECK_FUNCS(EVP_sha224 EVP_sha384 ...) failure on static linking
-
configure: fix AC_CHECK_FUNCS(TLS_method TLSv1_method ...) failure on static linking
-
configure: fix AC_CHECK_FUNCS(EVP_sha224 EVP_sha384 ...) failure on
-
configure: fix AC_CHECK_FUNCS(TLS_method TLSv1_method ...) failure on
-
configure: Invert AC_CHECK_LIB(EVP_md5,..) without -lz with -lz
-
This patch is wrong. EVP_md5 is found correctly in -lcrypto. The problem comes after with AC_CHECK_FUNCS() I close this PR and after fixing correctly I will reopen another one. Best regards
-
ci/Appveyor: Also build against MinGW32
-
MinGW32: Fix OpenSSL build
-
ci, MinGW32: Use the MSYS OpenSSL library
-
Hello, any news about this patch? This shows up only in static linking. Best regards
-
configure: Invert AC_CHECK_LIB(EVP_md5,..) without -lz with -lz
-
configure: Invert AC_CHECK_LIB(EVP_md5,..) without -lz with -lz
-
configure: Invert AC_CHECK_LIB(EVP_md5,..) without -lz with -lz
-
Fix static linking build failure on DTLS
-
configure: Invert AC_CHECK_LIB(EVP_md5,..) without -lz with -lz
-
Well, using AC_MSG_NOTICE() as logger, I've found the problem. I open another Merge Request. There shouldn't be other drawbacks.
-
Pardon, you were right. configure recognizes correctly EVP_md5 linking it with -lz But still getting error: The DTLS based transports require the libssl library from OpenSSL to be available and support DTLS I dig more.
-
And of course my patch on top of latest commit of master you've pointed me.
-
I mean I recompile Buildroot from scratch with same Architecture, static and without zlib package. All with my latest patch inverting the testing of -lz. So let's see if that rules(nested AC_CHECK_LIB()) works or not. If I find it tries to link with -lz, there is something wrong, otherwise it should be ok.
-
I explained myself bad. I've already tried, but stil gives problem with DTLS. So: Build static with master and AC_CHECK_LIB() with -lz inverted works. Now I'm going to compile without -lz, so let's see what happens Any other idea?
-
I explained myself bad. I've already tried, but stil gives problem with DTLS. So: Build static with master and AC_CHECK_LIB() with -lz inverted works. Now I'm going to compile with shared, so let's see what happens Any other idea?
-
configure: Fix -lcrypto -lz check during configure with static build
-
And invert AC_CHECK_LIB like below too? Because otherwise, with latest master we've got the same -lz dependency problem. if test x$CRYPTO = x; then AC_CHECK_LIB([crypto], [EVP_md5], [CRYPTO="crypto"; LIBCRYPTO="-lcrypto -lz"], [ AC_CHECK_LIB([crypto], [EVP_md5], [CRYPTO="crypto"; LIBCRYPTO="-lcrypto"])], [], [-lz]) else LIBCRYPTO="-l${CRYPTO}" fi
-
configure: Fix -lcrypto -lz test
-
configure: Fix -lcrypto -lz test
-
configure, static linking: Fix CRYPTO checks
-
Do you mean something like this? if test x$CRYPTO = x; then AC_CHECK_LIB([crypto], [EVP_md5], [CRYPTO="crypto"; LIBCRYPTO="-lcrypto -lz"], [ AC_CHECK_LIB([crypto], [EVP_md5], [CRYPTO="crypto"; LIBCRYPTO="-lcrypto"])], [], [-lz]) else LIBCRYPTO="-l${CRYPTO}" fi Because in this way it works fine.
-
Do you mean something like this? if test x$CRYPTO = x; then AC_CHECK_LIB([crypto], [EVP_md5], [CRYPTO="crypto"; LIBCRYPTO="-lcrypto -lz"], [ AC_CHECK_LIB([crypto], [EVP_md5], [CRYPTO="crypto"; LIBCRYPTO="-lcrypto"])], [], [-lz]) else LIBCRYPTO="-l${CRYPTO}" fi
-
I give it a try. I didn't think about it, good idea. I let you know. Thanks
-
Here is the Merge Request: https://sourceforge.net/p/net-snmp/code/merge-requests/18/
-
Fix static linking build failure on DTLS
-
configure, static linking: Fix CRYPTO checks
-
configure, static linking: Fix CRYPTO checks
-
Found the problem. There was a mix of issues. Going to send a patch for libz check in static linking.
-
Sorry, same problem as before, but we're almost there. Now it's missing only zlib on "checking for DTLS_method" I'm going to try to fix your patch too. Here you have a Merge request for the same problem though: https://sourceforge.net/p/net-snmp/code/merge-requests/17/
-
Sorry, same problem as before, but we're almost there. I'm going to try to fix your patch too. Here you have a Merge request for the same problem though: https://sourceforge.net/p/net-snmp/code/merge-requests/17/
-
Sorry, same problem as before. Here you have a Merge request for the same problem: https://sourceforge.net/p/net-snmp/code/merge-requests/17/ I'm going to try to fix your patch too.
-
Oh yes. I've just realized somebody else took exactly that patch for Buildroot. Here is mailing list mail: http://buildroot-busybox.2317881.n4.nabble.com/PATCH-1-1-netsnmp-fix-static-build-with-openssl-td203429.html So we can close this ticket! Thank you.
-
Hello, this is a only workaround. It doesn't take into account autotools. You can notice I don't touch configure.ac. Anyway here is the log of a build failure with Buildroot: http://autobuild.buildroot.net/results/e27/e2787d15f72949cbb347e8a1d344f5f80b4d7697/build-end.log And here they're discussing a better patch: https://patchwork.ozlabs.org/patch/967726/ So what I propose is something temporary before the one proposed on buildroot patchwork. Kind regards
-
configure: fix checking for DTLS_method
-
Hi Robert, first of all, great job with this plugin. The x's stay for "don't care"...
-
Hi Robert, first of all, great job with this plugin. The x's stay for "don't care"...
-
Hi Robert, first of all, great job with this plugin. The x's stay for "don't care"...
-
Fixed bug in binary values
1