From mboxrd@z Thu Jan 1 00:00:00 1970 From: Konrad Rzeszutek Wilk Subject: Re: [PATCH 28/28 v2] Separate runtime debug output from debug symbols Date: Fri, 26 Sep 2014 15:12:41 -0400 Message-ID: <20140926191241.GB8786@laptop.dumpdata.com> References: <1411658423-18877-29-git-send-email-olaf@aepfle.de> <1411731631-26415-1-git-send-email-olaf@aepfle.de> <5425708D02000078000399B7@mail.emea.novell.com> <20140926140256.GA9255@aepfle.de> <542597A70200007800039C0E@mail.emea.novell.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Content-Disposition: inline In-Reply-To: <542597A70200007800039C0E@mail.emea.novell.com> List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Sender: xen-devel-bounces@lists.xen.org Errors-To: xen-devel-bounces@lists.xen.org To: Jan Beulich Cc: Olaf Hering , Wei Liu , Ian Campbell , Stefano Stabellini , Tim Deegan , Ian Jackson , xen-devel@lists.xen.org, Samuel Thibault , Keir Fraser , David Scott List-Id: xen-devel@lists.xenproject.org On Fri, Sep 26, 2014 at 03:43:19PM +0100, Jan Beulich wrote: > >>> On 26.09.14 at 16:02, wrote: > > On Fri, Sep 26, Jan Beulich wrote: > > > >> >>> On 26.09.14 at 13:40, wrote: > >> > --- a/xen/Rules.mk > >> > +++ b/xen/Rules.mk > >> > @@ -10,6 +10,10 @@ lock_profile ?= n > >> > crash_debug ?= n > >> > frame_pointer ?= n > >> > lto ?= n > >> > +debug_xen ?= y > >> > +debuginfo_xen ?= y > >> > +debug := $(debug_xen) > >> > +debug_symbols := $(debuginfo_xen) > >> > >> I'm not sure I want to see this do away with the old mechanism > >> altogether - various of my testing build flavors depend on overriding > >> the "debug" variable, and I wouldn't want to change all their .config > >> files. At least for a certain period of time the latter two should > >> probably continue to use ?=. > > > > I think .config will continue to work with a "debug := y" and also 'make > > debug=y'. > > Ah, right, you do this before the inclusion point of Config.mk. > > >> > @@ -43,7 +47,10 @@ ALL_OBJS-$(x86) += $(BASEDIR)/crypto/built_in.o > >> > > >> > CFLAGS += -fno-builtin -fno-common > >> > CFLAGS += -Werror -Wredundant-decls -Wno-pointer-arith > >> > -CFLAGS += -pipe -g -D__XEN__ -include $(BASEDIR)/include/xen/config.h > >> > +ifeq ($(debug_symbols),y) > >> > +CFLAGS += -g > >> > +endif > >> > +CFLAGS += -pipe -D__XEN__ -include $(BASEDIR)/include/xen/config.h > >> > CFLAGS += -nostdinc > >> > > >> > CFLAGS-$(XSM_ENABLE) += -DXSM_ENABLE > >> > >> The context here shows that CFLAGS-y gets used elsewhere, so > >> please do so too above. And if possible this should be considered > >> elsewhere too. > > > > I'm not sure what you mean? The hunk just makes "-g" conditional. > > Looks like you nevertheless understood, as your shrunk down > patch does what I was asking for here (quoted below). > > >> In any event I don't see this change as important enough to > >> warrant overriding the feature freeze. > > > > Regarding feature freeze, that small version below for 'make debug=yn > > debug_symbols=yn' will work for me as well. > > This indeed seems more acceptable at this point. But the final > say would be with Konrad anyway. Release-Acked-by: Konrad Rzeszutek Wilk Reviewed-by: Konrad Rzeszutek Wilk I don't see this introducing an regressions. If there is an regression it will be during build time so pretty easy to spot. > > > --- a/xen/Rules.mk > > +++ b/xen/Rules.mk > > @@ -43,7 +43,8 @@ ALL_OBJS-$(x86) += $(BASEDIR)/crypto/built_in.o > > > > CFLAGS += -fno-builtin -fno-common > > CFLAGS += -Werror -Wredundant-decls -Wno-pointer-arith > > -CFLAGS += -pipe -g -D__XEN__ -include $(BASEDIR)/include/xen/config.h > > +CFLAGS-$(debug_symbols) += -g > > (This one.) > > Jan >