On Tue, Jul 6, 2010 at 4:57 AM, Ian Jackson
<Ian.Jackson@xxxxxxxxxxxxx> wrote:
Bruce Edge writes ("Re: [Xen-devel] State of gdbsx in xen-4.0-testing.hg and debugger documentation."):
> Here is a patch to enable gdbsx by default.
Thanks. Is it against 4.0-testing as it seems to be ? I think we
probably want this against xen-unstable, against which your patch
doesn't apply.
Ian, here's an unstable patch:
diff -Naur xen-unstable.hg-2010-07-06/tools/Makefile xen-unstable.hg/tools/Makefile
--- xen-unstable.hg-2010-07-06/tools/Makefile 2010-07-06 14:40:54.000000000 -0700
+++ xen-unstable.hg/tools/Makefile 2010-07-06 16:15:15.000000000 -0700
@@ -35,6 +35,7 @@
SUBDIRS-y += libxl
SUBDIRS-y += remus
SUBDIRS-$(CONFIG_X86) += xenpaging
+SUBDIRS-y += debugger/gdbsx
# These don't cross-compile
ifeq ($(XEN_COMPILE_ARCH),$(XEN_TARGET_ARCH))
@@ -114,3 +115,9 @@
$(buildmakevars2shellvars); \
$(MAKE) -C ioemu-dir clean; \
fi
+
+subdir-clean-debugger/gdbsx:
+ $(MAKE) -C debugger/gdbsx clean
+
+subdir-install-debugger/gdbsx:
+ $(MAKE) -C debugger/gdbsx install
-------------cut----------------
Thanks
-Bruce
> I suppose it would be cleaner to eliminate the debugger dir as gdbsx is the
> only one under there that's used and that would enable the tools/Rules.mk
> templates to cover the
> subdir-clean-gdbsx:
> subdir-install-gdbsx
> targets and not require them to be explicit in tools/Makefile
Well, yes, until we got another debugger again.
> Also, I don't know if
> SUBDIRS-y += debugger/gdbsx
> should be conditional on any type of target, CONFIG_X86 or?
I'm happy to enable it and get the people who find it breaks to
tell us what conditions mean it needs to be disabled :-).
Ian.