Message ID | 1390530729-1830-1-git-send-email-danomimanchego123@gmail.com |
---|---|
State | Rejected |
Headers | show |
On 24/01/14 03:32, Danomi Manchego wrote: > Nothing prevents the skeleton overlays from having pre-built Small nit: it's not a skeleton overlay, but a rootfs overlay (I first thought you were talking about the skeleton, which is copied _before_ the build). > executables or libraries in them. Therefore, we should install > the overlays prior to stripping, so that any such pre-built stuff > gets stripped too. I'm not sure if I agree with this change. I like the current situation where the rootfs-overlay is really copied verbatim and you can trust it. If there are binaries in there, we can assume that the user has the cross-compiler available to strip them if he wants that. That said, I can't think of a situation where it would be really essential that the binaries are not stripped, and anyway in that case it is still possible to use BR2_STRIP_EXCLUDE_*. So I don't feel very strongly about this. Regards, Arnout > > Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com> > --- > Makefile | 13 +++++++------ > 1 file changed, 7 insertions(+), 6 deletions(-) > > diff --git a/Makefile b/Makefile > index 9dfb1e0..9235c31 100644 > --- a/Makefile > +++ b/Makefile > @@ -513,6 +513,13 @@ ifeq ($(BR2_PACKAGE_PYTHON_PYC_ONLY),y) > find $(TARGET_DIR)/usr/lib/ -name '*.py' -print0 | xargs -0 rm -f > endif > rm -rf $(TARGET_DIR)/usr/lib/luarocks > + > + @$(foreach d, $(call qstrip,$(BR2_ROOTFS_OVERLAY)), \ > + $(call MESSAGE,"Copying overlay $(d)"); \ > + rsync -a $(RSYNC_VCS_EXCLUSIONS) \ > + --exclude .empty --exclude '*~' \ > + $(d)/ $(TARGET_DIR)$(sep)) > + > $(STRIP_FIND_CMD) | xargs $(STRIPCMD) 2>/dev/null || true > if test -d $(TARGET_DIR)/lib/modules; then \ > find $(TARGET_DIR)/lib/modules -type f -name '*.ko' | \ > @@ -545,12 +552,6 @@ endif > echo "PRETTY_NAME=\"Buildroot $(BR2_VERSION)\"" \ > ) > $(TARGET_DIR)/etc/os-release > > - @$(foreach d, $(call qstrip,$(BR2_ROOTFS_OVERLAY)), \ > - $(call MESSAGE,"Copying overlay $(d)"); \ > - rsync -a $(RSYNC_VCS_EXCLUSIONS) \ > - --exclude .empty --exclude '*~' \ > - $(d)/ $(TARGET_DIR)$(sep)) > - > @$(foreach s, $(call qstrip,$(BR2_ROOTFS_POST_BUILD_SCRIPT)), \ > $(call MESSAGE,"Executing post-build script $(s)"); \ > $(USER_HOOKS_EXTRA_ENV) $(s) $(TARGET_DIR) $(call qstrip,$(BR2_ROOTFS_POST_SCRIPT_ARGS))$(sep)) >
>>>>> "Arnout" == Arnout Vandecappelle <arnout@mind.be> writes: > On 24/01/14 03:32, Danomi Manchego wrote: >> Nothing prevents the skeleton overlays from having pre-built > Small nit: it's not a skeleton overlay, but a rootfs overlay (I first > thought you were talking about the skeleton, which is copied _before_ > the build). >> executables or libraries in them. Therefore, we should install >> the overlays prior to stripping, so that any such pre-built stuff >> gets stripped too. > I'm not sure if I agree with this change. I like the current > situation where the rootfs-overlay is really copied verbatim and you > can trust it. If there are binaries in there, we can assume that the > user has the cross-compiler available to strip them if he wants that. I agree. As it is now the user has full control, which is IMHO nice.
On Mon, Jan 27, 2014 at 9:11 AM, Peter Korsgaard <jacmet@uclibc.org> wrote: >>>>>> "Arnout" == Arnout Vandecappelle <arnout@mind.be> writes: > > > On 24/01/14 03:32, Danomi Manchego wrote: > >> Nothing prevents the skeleton overlays from having pre-built > > > Small nit: it's not a skeleton overlay, but a rootfs overlay (I first > > thought you were talking about the skeleton, which is copied _before_ > > the build). > > >> executables or libraries in them. Therefore, we should install > >> the overlays prior to stripping, so that any such pre-built stuff > >> gets stripped too. > > > I'm not sure if I agree with this change. I like the current > > situation where the rootfs-overlay is really copied verbatim and you > > can trust it. If there are binaries in there, we can assume that the > > user has the cross-compiler available to strip them if he wants that. > > I agree. As it is now the user has full control, which is IMHO nice. Okay, I can see that it might be expected that rootfs overlay content should be transferred as-is, I suppose. Let's consider this patch retracted, then. Danomi -
diff --git a/Makefile b/Makefile index 9dfb1e0..9235c31 100644 --- a/Makefile +++ b/Makefile @@ -513,6 +513,13 @@ ifeq ($(BR2_PACKAGE_PYTHON_PYC_ONLY),y) find $(TARGET_DIR)/usr/lib/ -name '*.py' -print0 | xargs -0 rm -f endif rm -rf $(TARGET_DIR)/usr/lib/luarocks + + @$(foreach d, $(call qstrip,$(BR2_ROOTFS_OVERLAY)), \ + $(call MESSAGE,"Copying overlay $(d)"); \ + rsync -a $(RSYNC_VCS_EXCLUSIONS) \ + --exclude .empty --exclude '*~' \ + $(d)/ $(TARGET_DIR)$(sep)) + $(STRIP_FIND_CMD) | xargs $(STRIPCMD) 2>/dev/null || true if test -d $(TARGET_DIR)/lib/modules; then \ find $(TARGET_DIR)/lib/modules -type f -name '*.ko' | \ @@ -545,12 +552,6 @@ endif echo "PRETTY_NAME=\"Buildroot $(BR2_VERSION)\"" \ ) > $(TARGET_DIR)/etc/os-release - @$(foreach d, $(call qstrip,$(BR2_ROOTFS_OVERLAY)), \ - $(call MESSAGE,"Copying overlay $(d)"); \ - rsync -a $(RSYNC_VCS_EXCLUSIONS) \ - --exclude .empty --exclude '*~' \ - $(d)/ $(TARGET_DIR)$(sep)) - @$(foreach s, $(call qstrip,$(BR2_ROOTFS_POST_BUILD_SCRIPT)), \ $(call MESSAGE,"Executing post-build script $(s)"); \ $(USER_HOOKS_EXTRA_ENV) $(s) $(TARGET_DIR) $(call qstrip,$(BR2_ROOTFS_POST_SCRIPT_ARGS))$(sep))
Nothing prevents the skeleton overlays from having pre-built executables or libraries in them. Therefore, we should install the overlays prior to stripping, so that any such pre-built stuff gets stripped too. Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com> --- Makefile | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)