On Fri, May 20, 2022 at 2:04 AM Ranjitsinh Rathod
<[email protected]> wrote:
>
> On Tue, May 10, 2022 at 10:50 AM, Steve Sakoman wrote:
>
> Extra newline above silently inhibits application of the following two hunks!
>
> Hi Steve,
>
> Can we raise a bug in the Yocto for this? Because it seems a serious issue 
> where the patch is actually not applied and bitbake not giving any 
> warning/error for this.
> Mostly if sometimes it should fail with the do_compilation only, but 
> sometimes if the dependency is not with do_compile it will not fail which 
> happened in this case.
>
> I think we should file a bug for this and some expert in patching tool should 
> take a look at that.
>
> Richard, Steve: Suggestions, please.

This particular case is a troublesome one, because technically that is
a well formed patch -- the new line causes the following lines to be
interpreted as a patch footer.

I tried a couple of patch checkers and they did not indicate an error
with your patch. So I am not sure how one would test for this without
generating a lot of false positives on patch footers.

So I think that the real lesson here is to not hand edit a patch
without really carefully verifying that it does exactly as intended.
It is much safer to use something like git format-patch to generate
the patch. Yes, more work, but the odds of a malformed patch are
astronomically lower!

Steve
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#165940): 
https://lists.openembedded.org/g/openembedded-core/message/165940
Mute This Topic: https://lists.openembedded.org/mt/90983830/21656
Group Owner: [email protected]
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub 
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to