On Wed, Jun 29, 2016 at 06:20:25PM +0200, Lukas Slebodnik wrote:
> On (24/06/16 10:16), Jakub Hrozek wrote:
> >On Thu, Jun 23, 2016 at 04:23:12PM -0400, Simo Sorce wrote:
> >> On Thu, 2016-06-23 at 21:36 +0200, Jakub Hrozek wrote:
> >> > On Thu, Apr 21, 2016 at 09:21:17AM +0200, Jakub Hrozek wrote:
> >> > > Given that Lukas says "http_parser can provide pkgconfig in
future", I
> >> > > read his mail as a preference to keep the pkg-check test. And
actually I
> >> > > agree, it doesn't hurt, let's keep it in.
> >> >
> >> > I wanted to push these patches:
> >> >
https://github.com/jhrozek/sssd/tree/secrets-review
> >> > but..I can't find http_parser_strict on Debian, only http-parser
and the
> >> > patches seem to require the _strict version. I really don't know
the
> >> > difference between the two, can we fallback to the non-strict?
> >>
> >> If it not too hard to detect if strict is present I would try to use it
> >> and fallback to not strict only of not available.
> >
> >Thanks, I will prepare a fallback patch. I agree about the preference,
> >but was wondering if we should require strict or not.
> I did a very brief test of latest patches from github and I can see
> few errors/warnings.
>
> src/responder/common/responder_common.c: In function ‘activate_unix_sockets’:
> src/responder/common/responder_common.c:795:13: error: too many arguments for format
[-Werror=format-extra-args]
> DEBUG(SSSDBG_CRIT_FAILURE,
> ^~~~~
> src/responder/common/responder_common.c:808:17: error: too many arguments for format
[-Werror=format-extra-args]
> DEBUG(SSSDBG_CRIT_FAILURE,
> ^~~~~
>
>
> src/responder/secrets/proxy.c: In function ‘ph_on_status’:
> src/responder/secrets/proxy.c:648:29: error: passing argument 2 of
‘ph_append_string’ from incompatible pointer type [-Werror=incompatible-pointer-types]
> ph_append_string(reply, &reply->reason_phrase, at, length);
> ^
> src/responder/secrets/proxy.c:624:13: note: expected ‘char **’ but argument is of
type ‘const char **’
> static void ph_append_string(TALLOC_CTX *memctx, char **dest,
> ^~~~~~~~~~~~~~~~
>
> CCLD test_cert_utils
> /usr/bin/ld: src/util/test_cert_utils-util.o: undefined reference to symbol
'hash_enter@(a)DHASH_0.4.3'
> /usr/lib64/libdhash.so.1: error adding symbols: DSO missing from command line
> Previous linking error is fixed by following diff
> diff --git a/Makefile.am b/Makefile.am
> index 622b0d2..7416141 100644
> --- a/Makefile.am
> +++ b/Makefile.am
> @@ -2895,7 +2895,6 @@ test_krb5_wait_queue_LDADD = \
>
> test_cert_utils_SOURCES = \
> src/tests/cmocka/test_cert_utils.c \
> - src/util/util.c \
> $(NULL)
> test_cert_utils_CFLAGS = \
> $(AM_CFLAGS) \
>
>
>
> CCLD responder_common-tests
> /usr/bin/ld:
../../../src/responder/common/responder_common_tests-responder_common.o: undefined
reference to symbol 'sd_listen_fds@@LIBSYSTEMD_209'
> /usr/lib64/libsystemd.so.0: error adding symbols: DSO missing from command line
> collect2: error: ld returned 1 exit status
>
> CCLD negcache-tests
> /usr/bin/ld: ../../../src/responder/common/negcache_tests-responder_common.o:
undefined reference to symbol 'sd_listen_fds@@LIBSYSTEMD_209'
> /usr/lib64/libsystemd.so.0: error adding symbols: DSO missing from command line
> collect2: error: ld returned 1 exit status
>
> and previous two linking error with following diff
> diff --git a/src/tests/cwrap/Makefile.am b/src/tests/cwrap/Makefile.am
> index 8005d99..d8a49f1 100644
> --- a/src/tests/cwrap/Makefile.am
> +++ b/src/tests/cwrap/Makefile.am
> @@ -145,6 +145,7 @@ responder_common_tests_LDADD = \
> $(CMOCKA_LIBS) \
> $(SSSD_LIBS) \
> $(SELINUX_LIBS) \
> + $(SYSTEMD_DAEMON_LIBS) \
> $(abs_top_builddir)/libsss_util.la \
> $(abs_top_builddir)/libsss_debug.la \
> $(abs_top_builddir)/libsss_test_common.la \
> @@ -162,6 +163,7 @@ negcache_tests_LDADD = \
> $(CMOCKA_LIBS) \
> $(SSSD_LIBS) \
> $(SELINUX_LIBS) \
> + $(SYSTEMD_DAEMON_LIBS) \
> $(abs_top_builddir)/libsss_util.la \
> $(abs_top_builddir)/libsss_debug.la \
> $(abs_top_builddir)/libsss_test_common.la \
OK, squashed and submitted to CI. Thanks, if the CI passes this time,
I'll push the patches and *finally* release the beta.
Our public CI log viewer machine has some issues, so I sent the CI link
which points directly to our Jenkins instance to our internal list.
ACK
For posterity, attached are patches I'm about to push.