Possible new category for non-reproducible builds: --build-id=sha1

[Santiago Torres Arias]

On Sat, Apr 24, 2021 at 05:59:07PM +0200, Roland Clobus wrote:
> Hello list,
> 
> I've looked the reproducible report for apt-cacher-ng [1].
> It looks like it is caused by a linker flag: -Wl,--build-id=sha1
> 
> I did not see a category for this type of difference. There are 31
> packages [2] that use the same linker flag.
> 
> How to fix this category of issue? As I understand it, this linker flag
> was introduced at Red Hat [3] to have unique identifier for builds of a
> package. The formula appears not to be very clear [4].
> 
> It could be solved by changing the linker, or by changing the build
> configuration of each package.

Hi Roland!


My understanding is this is somewhat akin to this family of issues[1]. 
I may be mistaken though, if so, how do you think they differ?

Cheers!
-Santiago

[1] https://tests.reproducible-builds.org/debian/issues/unstable/build_id_differences_only_issue.html
> 
> With kind regards,
> Roland Clobus
> 
> [1]
> https://tests.reproducible-builds.org/debian/rb-pkg/unstable/i386/diffoscope-results/apt-cacher-ng.html
> [2] https://codesearch.debian.net/search?q=-Wl%2C--build-id%3Dsha1
> [3]
> https://access.redhat.com/documentation/en-us/red_hat_enterprise_linux/6/html/developer_guide/compiling-build-id
> [4]
> https://stackoverflow.com/questions/41743295/elf-build-id-is-there-a-utility-to-recompute-it
> 



-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.reproducible-builds.org/pipermail/rb-general/attachments/20210424/35ac1f96/attachment.sig>