| Message ID | 20170907084230.26493-1-famz@redhat.com |
|---|---|
| State | New |
| Headers | show |
| Series | buildsys: Move rdma libs to per object | expand |
* Fam Zheng (famz@redhat.com) wrote: > Signed-off-by: Fam Zheng <famz@redhat.com> OK, I've not actually got a preference as to whether it's per-object or not - I don't really see any advantage. But since it doesn't look like it'll break it; Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> > --- > configure | 2 +- > migration/Makefile.objs | 1 + > 2 files changed, 2 insertions(+), 1 deletion(-) > > diff --git a/configure b/configure > index fb7e34a901..470b8a085b 100755 > --- a/configure > +++ b/configure > @@ -2818,7 +2818,6 @@ EOF > rdma_libs="-lrdmacm -libverbs" > if compile_prog "" "$rdma_libs" ; then > rdma="yes" > - libs_softmmu="$libs_softmmu $rdma_libs" > else > if test "$rdma" = "yes" ; then > error_exit \ > @@ -6035,6 +6034,7 @@ echo "CONFIG_TRACE_FILE=$trace_file" >> $config_host_mak > > if test "$rdma" = "yes" ; then > echo "CONFIG_RDMA=y" >> $config_host_mak > + echo "RDMA_LIBS=$rdma_libs" >> $config_host_mak > fi > > if test "$have_rtnetlink" = "yes" ; then > diff --git a/migration/Makefile.objs b/migration/Makefile.objs > index 1c7770da46..99e038024d 100644 > --- a/migration/Makefile.objs > +++ b/migration/Makefile.objs > @@ -11,3 +11,4 @@ common-obj-$(CONFIG_RDMA) += rdma.o > > common-obj-$(CONFIG_LIVE_BLOCK_MIGRATION) += block.o > > +rdma.o-libs := $(RDMA_LIBS) > -- > 2.13.5 > -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
On Thu, Sep 07, 2017 at 04:42:30PM +0800, Fam Zheng wrote: > Signed-off-by: Fam Zheng <famz@redhat.com> Reviewed-by: Peter Xu <peterx@redhat.com> > --- > configure | 2 +- > migration/Makefile.objs | 1 + > 2 files changed, 2 insertions(+), 1 deletion(-) > > diff --git a/configure b/configure > index fb7e34a901..470b8a085b 100755 > --- a/configure > +++ b/configure > @@ -2818,7 +2818,6 @@ EOF > rdma_libs="-lrdmacm -libverbs" > if compile_prog "" "$rdma_libs" ; then > rdma="yes" > - libs_softmmu="$libs_softmmu $rdma_libs" > else > if test "$rdma" = "yes" ; then > error_exit \ > @@ -6035,6 +6034,7 @@ echo "CONFIG_TRACE_FILE=$trace_file" >> $config_host_mak > > if test "$rdma" = "yes" ; then > echo "CONFIG_RDMA=y" >> $config_host_mak > + echo "RDMA_LIBS=$rdma_libs" >> $config_host_mak > fi > > if test "$have_rtnetlink" = "yes" ; then > diff --git a/migration/Makefile.objs b/migration/Makefile.objs > index 1c7770da46..99e038024d 100644 > --- a/migration/Makefile.objs > +++ b/migration/Makefile.objs > @@ -11,3 +11,4 @@ common-obj-$(CONFIG_RDMA) += rdma.o > > common-obj-$(CONFIG_LIVE_BLOCK_MIGRATION) += block.o > > +rdma.o-libs := $(RDMA_LIBS) > -- > 2.13.5 >
On Thu, 09/07 10:37, Dr. David Alan Gilbert wrote: > * Fam Zheng (famz@redhat.com) wrote: > > Signed-off-by: Fam Zheng <famz@redhat.com> > > OK, I've not actually got a preference as to whether it's > per-object or not - I don't really see any advantage. Thanks for the review. You're right this probably doesn't make a difference except for a bit more consistency, until we want to make rdma a module (as in --enable-modules) like the ones in block layer. The -libs and -cflags variables were initially added just for that. While we are talking about it, is there any reason why that will not be a good idea? There are other libraries used by QEMU outside block layer that are overdue to be converted to modules, like ui (gtk, sdl, etc.), rdma seems to be a candidate too. Fam
* Fam Zheng (famz@redhat.com) wrote: > On Thu, 09/07 10:37, Dr. David Alan Gilbert wrote: > > * Fam Zheng (famz@redhat.com) wrote: > > > Signed-off-by: Fam Zheng <famz@redhat.com> > > > > OK, I've not actually got a preference as to whether it's > > per-object or not - I don't really see any advantage. > > Thanks for the review. You're right this probably doesn't make a difference > except for a bit more consistency, until we want to make rdma a module (as in > --enable-modules) like the ones in block layer. The -libs and -cflags variables > were initially added just for that. > > While we are talking about it, is there any reason why that will not be a good > idea? There are other libraries used by QEMU outside block layer that are > overdue to be converted to modules, like ui (gtk, sdl, etc.), rdma seems to be a > candidate too. I don't think we've ever tried to make the migration code modular; it probably wouldn't be impossible to split it out - it's wired into a few places but it should be doable. Dave > Fam -- Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
Fam Zheng <famz@redhat.com> wrote: > Signed-off-by: Fam Zheng <famz@redhat.com> Reviewed-by: Juan Quintela <quintela@redhat.com> > --- > configure | 2 +- > migration/Makefile.objs | 1 + > 2 files changed, 2 insertions(+), 1 deletion(-) > > diff --git a/configure b/configure > index fb7e34a901..470b8a085b 100755 > --- a/configure > +++ b/configure > @@ -2818,7 +2818,6 @@ EOF > rdma_libs="-lrdmacm -libverbs" > if compile_prog "" "$rdma_libs" ; then > rdma="yes" > - libs_softmmu="$libs_softmmu $rdma_libs" > else > if test "$rdma" = "yes" ; then > error_exit \ > @@ -6035,6 +6034,7 @@ echo "CONFIG_TRACE_FILE=$trace_file" >> $config_host_mak > > if test "$rdma" = "yes" ; then > echo "CONFIG_RDMA=y" >> $config_host_mak > + echo "RDMA_LIBS=$rdma_libs" >> $config_host_mak > fi > > if test "$have_rtnetlink" = "yes" ; then > diff --git a/migration/Makefile.objs b/migration/Makefile.objs > index 1c7770da46..99e038024d 100644 > --- a/migration/Makefile.objs > +++ b/migration/Makefile.objs > @@ -11,3 +11,4 @@ common-obj-$(CONFIG_RDMA) += rdma.o > > common-obj-$(CONFIG_LIVE_BLOCK_MIGRATION) += block.o > > +rdma.o-libs := $(RDMA_LIBS)
On Thu, Sep 07, 2017 at 07:37:42PM +0800, Fam Zheng wrote: > On Thu, 09/07 10:37, Dr. David Alan Gilbert wrote: > > * Fam Zheng (famz@redhat.com) wrote: > > > Signed-off-by: Fam Zheng <famz@redhat.com> > > > > OK, I've not actually got a preference as to whether it's > > per-object or not - I don't really see any advantage. > > Thanks for the review. You're right this probably doesn't make a difference > except for a bit more consistency, until we want to make rdma a module (as in > --enable-modules) like the ones in block layer. The -libs and -cflags variables > were initially added just for that. > > While we are talking about it, is there any reason why that will not be a good > idea? There are other libraries used by QEMU outside block layer that are > overdue to be converted to modules, like ui (gtk, sdl, etc.), rdma seems to be a > candidate too. Since we have per-module flags, it makes sense to use them whereever it is reasonable todo so. The global flags should only be needed for things which are truely globally used, which is (almost) only glib2. Regards, Daniel
On Thu, 09/07 16:42, Fam Zheng wrote:
> Signed-off-by: Fam Zheng <famz@redhat.com>
I will include this in a coming pull request. Thanks.
Fam
diff --git a/configure b/configure index fb7e34a901..470b8a085b 100755 --- a/configure +++ b/configure @@ -2818,7 +2818,6 @@ EOF rdma_libs="-lrdmacm -libverbs" if compile_prog "" "$rdma_libs" ; then rdma="yes" - libs_softmmu="$libs_softmmu $rdma_libs" else if test "$rdma" = "yes" ; then error_exit \ @@ -6035,6 +6034,7 @@ echo "CONFIG_TRACE_FILE=$trace_file" >> $config_host_mak if test "$rdma" = "yes" ; then echo "CONFIG_RDMA=y" >> $config_host_mak + echo "RDMA_LIBS=$rdma_libs" >> $config_host_mak fi if test "$have_rtnetlink" = "yes" ; then diff --git a/migration/Makefile.objs b/migration/Makefile.objs index 1c7770da46..99e038024d 100644 --- a/migration/Makefile.objs +++ b/migration/Makefile.objs @@ -11,3 +11,4 @@ common-obj-$(CONFIG_RDMA) += rdma.o common-obj-$(CONFIG_LIVE_BLOCK_MIGRATION) += block.o +rdma.o-libs := $(RDMA_LIBS)
Signed-off-by: Fam Zheng <famz@redhat.com> --- configure | 2 +- migration/Makefile.objs | 1 + 2 files changed, 2 insertions(+), 1 deletion(-)