On Mon, Aug 13, 2018 at 01:58:56PM +0100, Daniel P. Berrangé wrote:
On Mon, Aug 13, 2018 at 02:54:00PM +0200, Ján Tomko wrote:
> On Mon, Aug 13, 2018 at 01:55:10PM +0200, Ján Tomko wrote:
> > For QEMU, we need a JSON parser that is able to handle its non-compliant
> > JSON usage:
> >
https://bugzilla.redhat.com/show_bug.cgi?id=1614569
> >
> > Unfortunately, this does not seem to be possible with Jansson.
> >
> > Revert back to using yajl, which also lets us get rid of the 'dlopen'
> > hacks and their bugfixes.
> >
> > (The QEMU default changes are not strictly required, but the patches
> > specifically use JANSSON)
> >
> > Also available on:
> > git://repo.or.cz/libvirt/jtomko.git revert
> >
> > Ján Tomko (16):
> > Revert "src: Move DLOPEN_LIBS to libraries introducing the
dependency"
>
> Note that this patch broke the build with clang by adding DLOPEN_LIBS to
> libvirt_setuid_rpc_client_la_CFLAGS instead of LDFLAGS:
>
https://www.redhat.com/archives/libvir-list/2018-August/msg00604.html
> So technically this series is a build breaker fix ;)
>
> > Revert "Fix link errors in tools/nss and tests"
> > Revert "remote: daemon: Make sure that JSON symbols are properly
> > loaded at startup"
> > Revert "util: jsoncompat: Stub out virJSONInitialize when compiling
> > without jansson"
> > Revert "tests: qemucapsprobe: Fix output after switching to jansson"
> > Revert "util: avoid symbol clash between json libraries"
> > Revert "tests: also skip qemuagenttest with old jansson"
> > Revert "m4: Introduce STABLE_ORDERING_JANSSON"
> > Revert "build: require Jansson if QEMU driver is enabled"
> > Revert "build: switch --with-qemu default from yes to check"
> > Revert "Remove virJSONValueNewStringLen"
> > Revert "build: remove references to WITH_YAJL for SETUID_RPC_CLIENT"
> > Revert "Remove functions using yajl"
> > Revert "Switch from yajl to Jansson"
> > Revert "build: undef WITH_JANSSON for SETUID_RPC_CLIENT"
>
> > Revert "build: add --with-jansson"
>
> And this patch was the only one that did not revert cleanly.
Great, on the basis that these are all clean reverts, and the last
looks sane:
Reviewed-by: Daniel P. Berrangé <berrange(a)redhat.com>