[OpenWrt-Devel] [PATCH v2] build: refactor JSON info files to `profiles.json`
Petr Štetiar
ynezz at true.cz
Wed Mar 4 05:43:11 EST 2020
Paul Spooren <mail at aparcar.org> [2020-03-03 17:02:07]:
> On 02.03.20 23:12, Petr Štetiar wrote:
> > Paul Spooren <mail at aparcar.org> [2020-03-02 16:19:05]:
> >
> > > - .IGNORE: $(BIN_DIR)/$(call IMAGE_NAME,$(1),$(2))
> > >
> > > [...]
> > >
> > > $(BIN_DIR)/$(call IMAGE_NAME,$(1),$(2)): $(KDIR)/tmp/$(call
> > > IMAGE_NAME,$(1),$(2))
> > > - cp $$^ $$@
> > > + -cp $$^ $$@
> > >
> > > The prefixed dash ignores a failure.
>
> > This change seems like a band-aid as I really don't see a reason to touch the
> > existing code just because you've put the JSON generation in that place
> > initially.
>
> How about I create an independent patch like the following?
>
> [ -f $$^ ] && cp $$^ $$@ || true
>
> It seems to be bad style ignoring cp errors, what if the destination is no
> longer writable or storage full? This way we can remove the .IGNORE.
I don't know the original intention of the .IGNORE, so I just assume, that
it's there because some of the targets abuse the image generation code for
stuff like DTBs or other artifacts.
So I'm not sure, whether this misuse means broken only source file/dir or it
might mean problem with the destination file/dir as well.
This would probably need more checks (and testing) so it might be probably
better to bite the bullet and fix the targets causing the problems and fix
them, then removing the .IGNORE altogether without any other additional
band-aids.
-- ynezz
_______________________________________________
openwrt-devel mailing list
openwrt-devel at lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
More information about the openwrt-devel
mailing list