[PATCH RESEND] kernel.bbclass: fix emit_depmod_pkgdata() workdir

[PATCH RESEND] kernel.bbclass: fix emit_depmod_pkgdata() workdir

[Ioan-Adrian Ratiu]

bitbake commit 67a7b8b02 "build: don't use $B as the default cwd for
functions" breaks the assumption that this function is running under
${B}. This causes build failures because System.map is under ${B}.

Signed-off-by: Ioan-Adrian Ratiu <adrian.ratiu@ni.com>
---
 meta/classes/kernel.bbclass | 1 +
 1 file changed, 1 insertion(+)

diff --git a/meta/classes/kernel.bbclass b/meta/classes/kernel.bbclass
index 062a0df..1a9b404 100644
--- a/meta/classes/kernel.bbclass
+++ b/meta/classes/kernel.bbclass
@@ -356,6 +356,7 @@ do_shared_workdir_setscene () {
 }
 
 emit_depmod_pkgdata() {
+	cd ${B}
 	# Stash data for depmod
 	install -d ${PKGDESTWORK}/${KERNEL_PACKAGE_NAME}-depmod/
 	echo "${KERNEL_VERSION}" > ${PKGDESTWORK}/${KERNEL_PACKAGE_NAME}-depmod/${KERNEL_PACKAGE_NAME}-abiversion
-- 
2.9.2

Re: [PATCH RESEND] kernel.bbclass: fix emit_depmod_pkgdata() workdir

[Burton, Ross]

[-- Attachment #1: Type: text/plain, Size: 397 bytes --]

On 10 August 2016 at 16:58, Ioan-Adrian Ratiu <adrian.ratiu@ni.com> wrote:

> bitbake commit 67a7b8b02 "build: don't use $B as the default cwd for
> functions" breaks the assumption that this function is running under
> ${B}. This causes build failures because System.map is under ${B}.
>

For clarity, can the function simply always use absolute paths instead of
relying on cwd?

Ross

[-- Attachment #2: Type: text/html, Size: 820 bytes --]

Re: [PATCH RESEND] kernel.bbclass: fix emit_depmod_pkgdata() workdir

[Ioan-Adrian Ratiu]

On Wed, 10 Aug 2016, "Burton, Ross" <ross.burton@intel.com> wrote:
> On 10 August 2016 at 16:58, Ioan-Adrian Ratiu <adrian.ratiu@ni.com> wrote:
>
>> bitbake commit 67a7b8b02 "build: don't use $B as the default cwd for
>> functions" breaks the assumption that this function is running under
>> ${B}. This causes build failures because System.map is under ${B}.
>>
>
> For clarity, can the function simply always use absolute paths instead of
> relying on cwd?

Yes, but other functions inside this bbclass also rely on cwd being
${B}; should I modify only this one to absolute paths or modify all of
these functions (do_shared_workdir, or do_strip) to maintain consistency?

>
> Ross

Re: [PATCH RESEND] kernel.bbclass: fix emit_depmod_pkgdata() workdir

[Ioan-Adrian Ratiu]

On Thu, 11 Aug 2016, Ioan-Adrian Ratiu <adrian.ratiu@ni.com> wrote:
> On Wed, 10 Aug 2016, "Burton, Ross" <ross.burton@intel.com> wrote:
>> On 10 August 2016 at 16:58, Ioan-Adrian Ratiu <adrian.ratiu@ni.com> wrote:
>>
>>> bitbake commit 67a7b8b02 "build: don't use $B as the default cwd for
>>> functions" breaks the assumption that this function is running under
>>> ${B}. This causes build failures because System.map is under ${B}.
>>>
>>
>> For clarity, can the function simply always use absolute paths instead of
>> relying on cwd?
>
> Yes, but other functions inside this bbclass also rely on cwd being
> ${B}; should I modify only this one to absolute paths or modify all of
> these functions (do_shared_workdir, or do_strip) to maintain consistency?
>

Ah nvm, I've looked closer at them functions and their use of cwd = $B
makes sense so it doesn't need modyfing. I'll resubmit only this patch.

>>
>> Ross
> -- 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core