Since the INIT_MANAGER variable has been introduced, there is no need
to append the distro features and set the init manager manually.
Replace the busybox/mdev and systemd examples with the 4 values
currently supported for the INIT_MANAGER variable.
(From meta-yocto rev: 2433a0f78a7fdb8450ef9ec64fa044274ac7a9e1)
Signed-off-by: Vivien Didelot <vdidelot@pbsc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This reverts commit 84d3a90557444d8cd83d780b1c3f2b278ab07af2.
This shouldn't have been merged, its not ready yet.
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
One mmap patch was merged upstream, the other is in progress but after
discussion has been amended and resent. Update the patches as such.
(From OE-Core rev: c68349a3334542e51f6c1c23f4f4342e4fce2b81)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
rpm packaging has shown good numbers wrt reproducible builds as well
when testing core-image-minimal locally:
Reproducibility summary for rpm: same=4671 different=0 missing=0 total=4671
Thus enable it in the autobuilder to be more widely tested.
(From OE-Core rev: 84d3a90557444d8cd83d780b1c3f2b278ab07af2)
Signed-off-by: Jan-Simon Möller <jsmoeller@linuxfoundation.org>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
These two header fixes are not needed with the current source as the
headers are there through other header files. Drop the patches.
(From OE-Core rev: ece3efe43b6b9de5426c23f29db15dd709bcd749)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixes reproducibility when build acpica yacc sources by instructing
bison on how to do the file prefix mapping
(From OE-Core rev: d13a1df7bd173fca3fc2cdf889d5eb487e35d6fd)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
File name mapping was added to bison, but needs to be applied in more
cases. Add a patch to enable file name mapping for symbols
(From OE-Core rev: 5a213506377b0db4e88ca094215eadf08c21a2e1)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Now that native recipes have PACKAGES, this DEPENDS construct doesn't
work. It applies to target and nativesdk recipes so adjust accordingly.
(From OE-Core rev: 5d459ba1c13c89b246a0f8d743027e1de93da910)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Remove the redundant 'of' word in the INITRAMFS_IMAGE comment.
(From meta-yocto rev: 200b05182c74f3f4f778907d5ee92831adb3d77e)
Signed-off-by: Vivien Didelot <vdidelot@pbsc.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Adds a Dockerfile for build the reference hash equivalence server in a
container
(Bitbake rev: d9a0a88db5888039bfbb6ce5129b43350f79b1c1)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add ruby-ri-docs and meson back as not reproducible, remove watchdog,
xorg-minimal-fonts and xmlto as issues fixed.
(From OE-Core rev: 5e209a7455a231342cded99985b4a6caff424e4f)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When installing the font files, preserve their timestamp rather than using
the current time which fixes reproducibility issues.
(From OE-Core rev: a0c4f4b0182a995f0eb2709cc9b3c852527ab936)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add some temporary bumps to versions to change the output hash so
the fix applies correctly. Can be dropped next time we update the
recipe and the output changes.
(From OE-Core rev: 69205edcdff865048e55a6b7feaf82064ebc10c5)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The build was injecting the path of the host's sendmail binary. Set
this deterministically to match OE's path for it.
(From OE-Core rev: 476328b91ea4417160580d28df4fcc1147d85ae2)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The reproducible build tests can take a long time, so having more
logging messages at various points in the build can help debug where the
build is taking a long time.
(From OE-Core rev: 6b792afe8759d62af8e713b86dad8f6721961a05)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The SSTATE_MIRRORS variable was misspelled, which allowed the "clean"
test build to pull from the mirror.
(From OE-Core rev: e42497bd84d0bb370a9f7b0448bff29f01fd1b0c)
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When building vim it tries to rebuild files using iconv. If this fails
the build continues anyway but the output is not determnistic as builds
using a hosttools tarball are different from builds where there isn't a
hosttools tarball. Add the needed gconvs to the tarball when iconv is
present to become determistic and generate vim locales consistently.
(From OE-Core rev: b945652a088f430a2adec6b968cd00c5928d4272)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
I've seen local differences in the quilt output depending on whether the column
binary was available in the sysroot. Fix determinism issues by being specific
about configuration.
(From OE-Core rev: 0cd5fba8634bcc679518f98cc25be66a51081372)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When building SDKs with the deb backend you could see errors like:
Setting up nativesdk-python3-ndg-httpsclient (0.5.1-r0) ...
mkdir: cannot create directory ‘/usr/lib/opkg’: Permission denied
dpkg: error processing package nativesdk-python3-ndg-httpsclient (--configure):
which is due to environment misconfiguration when removing packages.
Fix this by setting the same environment variables as used for installation.
(From OE-Core rev: f23c7e319a192ada14bb9a82822ef2967309aaea)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The LINGUAS file can be written by two different Makefile targets
and if they race, the desktop file contents isn't deterministic.
Fix the makfile to avoid this.
(From OE-Core rev: 416bc7b697764075fbf73683cd8bddf36d839244)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a couple of configure options to avoid determism issues in the vim build.
This can happen due to the addition of glib-2.0 to the native sysroot through
later task additions to the sysroot through indirect dependencies.
(From OE-Core rev: 914f86054f5ea0a115767c1b3d9cdb4c4ef9545b)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The makefile injects the current date into the version file. Do this
deterministically with SOURCE_DATE_EPOCH.
(From OE-Core rev: 31f2ad739ea776a1e11b5cef5434df188007c7bf)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Now that 022 is the default in BB_DEFAULT_UMASK in bitbake.conf, we
don't need any of these task flags, clean up.
(From OE-Core rev: 816fca781943a7dbf40391d9db34c7bf12711962)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Rather than trying to set the umask of every task to the correct
value for determinism, set one value globally. This uses a new
bitbake variable so bump the minimum version to match.
This fixes strange determinism issues in at least quilt-ptest,
valgrind-ptest and kernel-devsrc.
(From OE-Core rev: b07b7deeae3d519d9998d583592f3e4f8f2802b3)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The two duplicate lines are not needed. The existence is confusing.
(From OE-Core rev: 36bdb4faa90dc18bc020481eba82ee570b968c39)
Signed-off-by: Thomas Viehweger <patchesThomas.Vie@web.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
As mentioned during review, this patch can be less invasive and hence easier
to maintain. Improve as such.
(From OE-Core rev: 3a1bd768e62a493aa73e82f5c443ca28b108af51)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This allows metadata to rely upon BB_DEFAULT_UMASK.
(Bitbake rev: 969ac64adab236ce2d5196bcc294005a497913ae)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
