Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Michael Biebl
Am 09.11.18 um 20:22 schrieb Holger Levsen: > On Fri, Nov 09, 2018 at 07:56:27PM +0100, Michael Biebl wrote: >>> │ │ │ ├── ./usr/lib/systemd/tests/test-hexdecoct >> ... >>> │ │ │ │ │0x0878 48415245 44002f6c 69622f73 79737465 HARED./lib/syste >>> │ │ │ │ │ - 0x0888 6d640068 61726564 00

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Holger Levsen
On Fri, Nov 09, 2018 at 07:56:27PM +0100, Michael Biebl wrote: > Thanks! > Looks like this did indeed trigger a failure: > https://tests.reproducible-builds.org/debian/dbdtxt/unstable/amd64/systemd_239-11.diffoscope.txt.gz > > I do see the expected RUNPATH related diffs like e.g. in: > > > │ │ │

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Michael Biebl
Am 09.11.18 um 15:12 schrieb Holger Levsen: > On Fri, Nov 09, 2018 at 03:09:01PM +0100, Michael Biebl wrote: >> Another test case should be src:systemd >> When built in a merged-usr environment, meson will embed a wrong RUNPATH >> into the test-binaries. >> This should show up in the diffoscope out

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Holger Levsen
On Fri, Nov 09, 2018 at 03:09:01PM +0100, Michael Biebl wrote: > Another test case should be src:systemd > When built in a merged-usr environment, meson will embed a wrong RUNPATH > into the test-binaries. > This should show up in the diffoscope output. triggered as well. (you can do this too: jus

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Michael Biebl
Am 09.11.18 um 14:10 schrieb Holger Levsen: > do any of you know a package which currently builds differently > depending on this? I'd like to test this change now :) Another test case should be src:systemd When built in a merged-usr environment, meson will embed a wrong RUNPATH into the test-bin

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Holger Levsen
On Fri, Nov 09, 2018 at 01:21:13PM +, Simon McVittie wrote: > The version of quilt in buster (0.65-2) builds differently, in simple > enough ways to give nice diffoscope output. That bug is fixed in unstable > (0.65-3), so please try building both - I hope you will find that quilt > is currentl

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Holger Levsen
On Fri, Nov 09, 2018 at 01:21:13PM +, Simon McVittie wrote: > The version of quilt in buster (0.65-2) builds differently, in simple > enough ways to give nice diffoscope output. That bug is fixed in unstable > (0.65-3), so please try building both - I hope you will find that quilt > is currentl

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Simon McVittie
On Fri, 09 Nov 2018 at 13:10:26 +, Holger Levsen wrote: > On Thu, Nov 08, 2018 at 06:41:18PM +, Simon McVittie wrote: > > On Sat, 14 Jul 2018 at 11:59:37 +0200, Michael Biebl wrote: > > > If we could auto-detect any problems resulting from that via the > > > reproducible builds effort, this

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-09 Thread Holger Levsen
Hi, On Thu, Nov 08, 2018 at 06:41:18PM +, Simon McVittie wrote: > On Sat, 14 Jul 2018 at 11:59:37 +0200, Michael Biebl wrote: > > If we could auto-detect any problems resulting from that via the > > reproducible builds effort, this would be great. > The attached patch is entirely untested, but

Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-08 Thread Simon McVittie
On Sat, 14 Jul 2018 at 11:59:37 +0200, Michael Biebl wrote: > If we could auto-detect any problems resulting from that via the > reproducible builds effort, this would be great. The attached patch is entirely untested, but hopefully works. smcv >From d5412972e90a57c0bdf9126101ecfa1ecdfe09c2 M

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-11-08 Thread Simon McVittie
On Sat, 14 Jul 2018 at 18:39:36 +0200, Philipp Kern wrote: > On 14.07.2018 13:01, Michael Biebl wrote: > > I would have expected that buildds use debootstrap from unstable, but > > I'm not a buildd admin, so I trust you to know more about that then me. > > Thanks for correcting me on that and sorry

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-07-14 Thread Philipp Kern
On 14.07.2018 13:01, Michael Biebl wrote: > Am 14.07.2018 um 12:28 schrieb Mattia Rizzolo: >> On Sat, Jul 14, 2018 at 11:59:37AM +0200, Michael Biebl wrote: >>> debootstrap was changed (again) in 1.0.102 to now default to >>> --merged-usr [1]. >>> >>> Systems are not forcefully upgraded to a merged

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-07-14 Thread Michael Biebl
Am 14.07.2018 um 12:28 schrieb Mattia Rizzolo: > On Sat, Jul 14, 2018 at 11:59:37AM +0200, Michael Biebl wrote: >> debootstrap was changed (again) in 1.0.102 to now default to >> --merged-usr [1]. >> >> Systems are not forcefully upgraded to a merged-usr setup atm though. >> buildds are regularly r

Bug#901473: [Qa-jenkins-dev] Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-07-14 Thread Mattia Rizzolo
On Sat, Jul 14, 2018 at 11:59:37AM +0200, Michael Biebl wrote: > debootstrap was changed (again) in 1.0.102 to now default to > --merged-usr [1]. > > Systems are not forcefully upgraded to a merged-usr setup atm though. > buildds are regularly re-created from scratch and will thus have a > merged-

Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-07-14 Thread Michael Biebl
debootstrap was changed (again) in 1.0.102 to now default to --merged-usr [1]. Systems are not forcefully upgraded to a merged-usr setup atm though. buildds are regularly re-created from scratch and will thus have a merged-usr setup. This means, binaries that were built in a merged-usr buildd envi

Bug#901473: jenkins.debian.org: Vary merged-usr in reproducibility testing?

2018-06-13 Thread Chris Lamb
Package: jenkins.debian.org Severity: wishlist X-Debbugs-CC: jrt...@debian.org >From #debian-devel: < jrtc27> does r-b do merged/non-merged [usr] variation? < lamby> No. Do you expect it to find *r-b* variations, or simply FTBFS? < jrtc27> lamby: both, but