Hi, As Ian Jackson asked I am sending build system fixes based on patches applied a few weeks ago to Xen unstable. Earlier patches which did not required any changes were applied to Xen 4.2.3. This patch series contains only applicable patches which must be modified to be applied. List of patches: - stubdom: Clean extras/mini-os directory during make clean, - stubdom: Remove xenstore directory during make crossclean, - Silently ignore include errors during make distclean. Daniel
Daniel Kiper
2013-Jul-24 14:10 UTC
[PATCH 1/3] stubdom: Clean extras/mini-os directory during make clean
Signed-off-by: Daniel Kiper <daniel.kiper@oracle.com> Acked-by: Samuel Thibault <samuel.thibault@ens-lyon.org> (based on commit 2bf87074e17e77ca761e3490e5f054175e082e5b) --- stubdom/Makefile | 1 + 1 file changed, 1 insertion(+) diff --git a/stubdom/Makefile b/stubdom/Makefile index 75829cd..481201a 100644 --- a/stubdom/Makefile +++ b/stubdom/Makefile @@ -405,6 +405,7 @@ clean: rm -fr mini-os-$(XEN_TARGET_ARCH)-caml rm -fr mini-os-$(XEN_TARGET_ARCH)-grub rm -fr mini-os-$(XEN_TARGET_ARCH)-xenstore + $(MAKE) DESTDIR= -C $(MINI_OS) clean $(MAKE) DESTDIR= -C caml clean $(MAKE) DESTDIR= -C c clean rm -fr grub-$(XEN_TARGET_ARCH) -- 1.7.10.4
Daniel Kiper
2013-Jul-24 14:10 UTC
[PATCH 2/3] stubdom: Remove xenstore directory during make crossclean
Signed-off-by: Daniel Kiper <daniel.kiper@oracle.com> Acked-by: Samuel Thibault <samuel.thibault@ens-lyon.org> (based on commit 43be6972f977660ba9aaf15f78e39c025c9550d8) --- stubdom/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/stubdom/Makefile b/stubdom/Makefile index 481201a..fbe5169 100644 --- a/stubdom/Makefile +++ b/stubdom/Makefile @@ -420,7 +420,7 @@ crossclean: clean rm -fr $(CROSS_ROOT) rm -fr newlib-$(XEN_TARGET_ARCH) rm -fr zlib-$(XEN_TARGET_ARCH) pciutils-$(XEN_TARGET_ARCH) - rm -fr libxc-$(XEN_TARGET_ARCH) ioemu + rm -fr libxc-$(XEN_TARGET_ARCH) ioemu xenstore rm -f mk-headers-$(XEN_TARGET_ARCH) rm -fr ocaml-$(XEN_TARGET_ARCH) rm -fr include -- 1.7.10.4
Daniel Kiper
2013-Jul-24 14:10 UTC
[PATCH 3/3] Silently ignore include errors during make distclean
Signed-off-by: Daniel Kiper <daniel.kiper@oracle.com> Acked-by: Ian Jackson <ian.jackson@eu.citrix.com> (based on commit 38aabce5424710badc080decca23628442d80673) --- buildconfigs/mk.linux-2.6-common | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/buildconfigs/mk.linux-2.6-common b/buildconfigs/mk.linux-2.6-common index f3680f0..e8e97fb 100644 --- a/buildconfigs/mk.linux-2.6-common +++ b/buildconfigs/mk.linux-2.6-common @@ -29,7 +29,7 @@ LINUX_VER3 := $(LINUX_SERIES).$(word 3, $(subst ., ,$(LINUX_VER))) .PHONY: _build _build: build -include buildconfigs/src.$(XEN_LINUX_SOURCE) +-include buildconfigs/src.$(XEN_LINUX_SOURCE) LINUX_ARCH = $$(sh buildconfigs/select-linux-arch $(LINUX_SRCDIR)) IMAGE_PATH = $$(sh buildconfigs/select-linux-image $(LINUX_DIR) $(LINUX_ARCH) $(IMAGE_TARGET)) -- 1.7.10.4
Daniel Kiper writes ("[PATCH 0/3] Build system fixes for Xen 4.2.3"):> As Ian Jackson asked I am sending build system fixes based > on patches applied a few weeks ago to Xen unstable. Earlier > patches which did not required any changes were applied to > Xen 4.2.3. This patch series contains only applicable patches > which must be modified to be applied. > > List of patches: > - stubdom: Clean extras/mini-os directory during make clean, > - stubdom: Remove xenstore directory during make crossclean, > - Silently ignore include errors during make distclean.I''m afraid I think it''s probably best not to apply modified build system backport patches, unless they''re really important. These involve more risk of downstream build breakages, and the benefit is very small because most people already using the stable release will have already developed workarounds for any build system bugs they have encountered. Thanks, Ian.
On Wed, Aug 07, 2013 at 12:14:37PM +0100, Ian Jackson wrote:> Daniel Kiper writes ("[PATCH 0/3] Build system fixes for Xen 4.2.3"): > > As Ian Jackson asked I am sending build system fixes based > > on patches applied a few weeks ago to Xen unstable. Earlier > > patches which did not required any changes were applied to > > Xen 4.2.3. This patch series contains only applicable patches > > which must be modified to be applied. > > > > List of patches: > > - stubdom: Clean extras/mini-os directory during make clean, > > - stubdom: Remove xenstore directory during make crossclean, > > - Silently ignore include errors during make distclean. > > I''m afraid I think it''s probably best not to apply modified build > system backport patches, unless they''re really important. These involve > more risk of downstream build breakages, and the benefit is very small > because most people already using the stable release will have already > developed workarounds for any build system bugs they have encountered.I do not insist on applying them but I think that if we applied previous patches why do not apply those. We should act consistently. Daniel
>>> On 07.08.13 at 13:27, Daniel Kiper <daniel.kiper@oracle.com> wrote: > On Wed, Aug 07, 2013 at 12:14:37PM +0100, Ian Jackson wrote: >> Daniel Kiper writes ("[PATCH 0/3] Build system fixes for Xen 4.2.3"): >> > As Ian Jackson asked I am sending build system fixes based >> > on patches applied a few weeks ago to Xen unstable. Earlier >> > patches which did not required any changes were applied to >> > Xen 4.2.3. This patch series contains only applicable patches >> > which must be modified to be applied. >> > >> > List of patches: >> > - stubdom: Clean extras/mini-os directory during make clean, >> > - stubdom: Remove xenstore directory during make crossclean, >> > - Silently ignore include errors during make distclean. >> >> I''m afraid I think it''s probably best not to apply modified build >> system backport patches, unless they''re really important. These involve >> more risk of downstream build breakages, and the benefit is very small >> because most people already using the stable release will have already >> developed workarounds for any build system bugs they have encountered. > > I do not insist on applying them but I think that if we > applied previous patches why do not apply those. > We should act consistently.Being risk free is valued higher than consistency on stable trees. Jan