[OE-core] [PATCH] grub git: workaround debugedit problems for 64 bit targets

Koen Kooi koen.kooi at linaro.org
Tue Jan 21 09:40:33 UTC 2014


On 01/16/2014 11:47 AM, Otavio Salvador wrote:
> On Thu, Jan 16, 2014 at 8:46 AM, Koen Kooi <koen.kooi at linaro.org
> <mailto:koen.kooi at linaro.org>> wrote:
>
>     Op 16 jan. 2014, om 11:38 heeft Otavio Salvador
>     <otavio at ossystems.com.br <mailto:otavio at ossystems.com.br>> het
>     volgende geschreven:
>      > On Thu, Jan 16, 2014 at 7:27 AM, Koen Kooi <koen.kooi at linaro.org
>     <mailto:koen.kooi at linaro.org>> wrote:
>      > On 01/15/2014 12:22 PM, Otavio Salvador wrote:
>      > On Wed, Jan 15, 2014 at 8:15 AM, Koen Kooi <koen.kooi at linaro.org
>     <mailto:koen.kooi at linaro.org>
>      > <mailto:koen.kooi at linaro.org <mailto:koen.kooi at linaro.org>>> wrote:
>      >
>      >     I've limited it to 64bit targets, but to be honest debug
>     splitting
>      >     doesn't make sense for grub since it won't run in your rootfs and
>      >     attaching gdb is pretty much impossible.
>      >
>      >     Signed-off-by: Koen Kooi <koen.kooi at linaro.org
>     <mailto:koen.kooi at linaro.org>
>      >     <mailto:koen.kooi at linaro.org <mailto:koen.kooi at linaro.org>>>
>      >
>      >
>      >
>      > I'd prefer to remove the debug systems than skip the split. What
>     others
>      > think?
>      >
>      > I'm not sure what you mean by that, do you mean using
>     INHIBIT_PACKAGE_STRIP or something else?
>      >
>      >
>      > Yes and be sure it has been stripped during build.
>
>     That seems to be a lot more work for the same effect. Stripping
>     isn't the problem, it's debugedit trying to do its magic that fails.
>
>
> Well, if not stripped the symbols will be left in the binaries which is
> not expected.

I've looked into package.bbclass and it's a classic case of code not 
matching docs :(

-- 
Koen Kooi
Builds and Baselines | Release Manager
Linaro.org | Open source software for ARM SoCs



More information about the Openembedded-core mailing list