Re: Bug#821793: fusermount should not attempt to use /etc/mtab
Hi, Can someone clarify if this is a bug in upstream fuse (which assumes existence of /etc/mtab) or in the software that creates the chroots without /etc/mtab? If someone can give me a decent source that states that /etc/mtab does not need to exist, I'll be happy to change fusermount accordingly. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 15 2015, Holger Levsen holger-Yq+U+nADzX/cal+tvhy...@public.gmane.org
wrote:
Hi Nikolaus,
On Samstag, 14. März 2015, Nikolaus Rath wrote:
I can also reproduce this with s3ql_2.11.1+dfsg-2.dsc pbuilding for
jessie.
Alright, I'm at a loss:
[..]
The current package (2.13+dfsg) builds just as fine.
what's the setting of BINDMOUNTS in your pbuilderrc?
$ cat ~/.pbuilderrc
AUTO_DEBSIGN=${AUTO_DEBSIGN:-no}
HOOKDIR=/var/cache/pbuilder/hooks
BINDMOUNTS=/var/cache/pbuilder/result
EXTRAPACKAGES=eatmydata
export LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libeatmydata.so
$ cat /etc/pbuilderrc | grep -i BINDMOUNTS
(nothing)
Best,
-Nikolaus
--
GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F
Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F
»Time flies like an arrow, fruit flies like a Banana.«
signature.asc
Description: PGP signature
___
Reproducible-builds mailing list
Reproducible-builds@lists.alioth.debian.org
http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 10 2015, Holger Levsen holger-Yq+U+nADzX/cal+tvhy...@public.gmane.org wrote: Hi, So, same result. Which probably means I can reproduce it locally on my laptop as well. Me tries... and gets the same result. So I suppose an (at least) important bug is in order... I can also reproduce this with s3ql_2.11.1+dfsg-2.dsc pbuilding for jessie. Alright, I'm at a loss: $ dget -x http://ftp.de.debian.org/debian/pool/main/s/s3ql/s3ql_2.11.1+dfsg-2.dsc [...] $ apt-cache policy pbuilder | grep -i installed Installed: 0.215+nmu3 $ date Fri Mar 13 20:29:57 PDT 2015 $ sudo pbuilder --update [...] $ cd s3ql-2.11.1+dfsg/ $ pdebuild [...] # Some tests will try to create log files in OME mkdir -p debian/tmphome HOME=/tmp/buildd/s3ql-2.11.1+dfsg/debian/tmphome py.test-3 tests/ = test session starts == platform linux -- Python 3.4.2 -- py-1.4.25 -- pytest-2.6.3 -- /usr/bin/python3 collecting ... collected 568 items tests/t1_backends.py::test_read_write[plain-local] PASSED tests/t1_backends.py::test_read_write[plain-mock_s3c] PASSED tests/t1_backends.py::test_read_write[plain-mock_gs] PASSED tests/t1_backends.py::test_read_write[aes-local] PASSED tests/t1_backends.py::test_read_write[aes-mock_s3c] PASSED tests/t1_backends.py::test_read_write[aes-mock_gs] PASSED [...] dpkg-buildpackage: binary and diff upload (original source NOT included) I: user script /var/cache/pbuilder/build//4638/tmp/hooks/B90lintian starting [...] The current package (2.13+dfsg) builds just as fine. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 10 2015, Holger Levsen holger-Yq+U+nADzX/cal+tvhy...@public.gmane.org wrote: did you try to reproduce the problem with pbuilder locally? I think it's more likely a problem with pbuilder than with the reproducible stuff on jenkins. (But I'm offline atm so I cannot test myself.) I tried, but for me it works just fine with pbuilder. let me try with regular pbuilder on the machine... [...] tests/t1_backends.py::test_read_write[local/aes] PASSED tests/t1_backends.py::test_read_write[mock-s3c/aes] FAILED Odd. I've never seen this problem before. So, same result. Which probably means I can reproduce it locally on my laptop as well. Me tries... and gets the same result. So I suppose an (at least) important bug is in order... Nikolaus, which version of pbuilder are you using? I'm using the wheezy one on wheezy... (on both jenkins.d.n as well as my laptop.) I'm using jessie, so maybe that's it. Though I still fail to understand why this would make a difference. Anyway, I'll try to install the wheezy pbuilder and reproduce. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 01 2015, Mattia Rizzolo mattia-p8aev1vil9rafugrpc6...@public.gmane.org wrote: Anyway, I performed a local build and this is what debbindiff outputs: http://volatile.mapreri.org/2015-03-01/b46eb200e612629e11c7b6a1c925da36/s3ql.dbd.html e.g. timestamps in sphinx docs. Is this with #776443 applied? Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
[Reproducible-builds] Bug#779474: debbindiff: hide temporary paths in html and text output when comparing tarballs
On Mar 02 2015, Jérémy Bobbio lunar-8fiuurrzop0dnm+yrof...@public.gmane.org wrote: Paul Wise: The output when directly comparing two tarballs contains the temporary paths used. This leaks the debbindiff user's TMP variable into the report and is also ugly. I suggest changing to just showing filenames. Current: /tmp/user/1000/tmpFsxqeVdebbindiff/flasm_1.62-6.debian.tar vs. /tmp/user/1000/tmpEVZZmydebbindiff/flasm_1.62-7.debian.tar Desired: flasm_1.62-6.debian.tar vs. flasm_1.62-7.debian.tar I am not sure what the use case is here. Without full path for the filenames at the top-level, there is no way to differentiate a report for two builds of the same source. As this is debbindiff primary purpose, I must admit I am a bit at loss. As you say, the primary purpose of debbindiff is to compare two builds of the same source. So why does the path matter at all? I think all that matters is the difference - it's not as if there's a newer and older version or something like that. In the case where debbindiff is used to compare different sources (as apparently done above), the filename alone contains the necessary information. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 03 2015, Holger Levsen hol...@layer-acht.org wrote: Hi Nikolas, On Montag, 2. März 2015, Nikolaus Rath wrote: Almost certain. I could be certain if I could see the log with the failed test, but it seems that port 80 on reproducible.debian.net is down at the moment: it's up again, though it's port 443 :) do you still have problems getting the logs? I managed to get the logs yesterday, but right now it seems down again: $ ping -c 3 reproducible.debian.net PING reproducible.debian.net (46.16.73.183) 56(84) bytes of data. 64 bytes from ip46-16-73-183.pbiaas.com (46.16.73.183): icmp_seq=1 ttl=53 time=181 ms 64 bytes from ip46-16-73-183.pbiaas.com (46.16.73.183): icmp_seq=2 ttl=53 time=183 ms 64 bytes from ip46-16-73-183.pbiaas.com (46.16.73.183): icmp_seq=3 ttl=53 time=188 ms --- reproducible.debian.net ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2001ms rtt min/avg/max/mdev = 181.074/184.085/188.064/2.955 ms $ curl --connect-timeout 30 --verbose https://reproducible.debian.net/ * Hostname was NOT found in DNS cache * Trying 46.16.73.183... * Connection timed out after 30001 milliseconds * Closing connection 0 curl: (28) Connection timed out after 30001 milliseconds Is there any way I could ssh into your build host to debug locally? no, sorry. (there are no user accounts from ldap like on .debian.org machines on it.) did you try to reproduce the problem with pbuilder locally? I think it's more likely a problem with pbuilder than with the reproducible stuff on jenkins. (But I'm offline atm so I cannot test myself.) I tried, but for me it works just fine with pbuilder. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 02 2015, Mattia Rizzolo mattia-p8aev1vil9rafugrpc6...@public.gmane.org wrote: On Mon, Mar 02, 2015 at 08:42:32AM -0800, Nikolaus Rath wrote: On Mar 01 2015, Mattia Rizzolo mattia-p8aev1vil9rafugrpc6...@public.gmane.org wrote: On Sun, Mar 01, 2015 at 12:04:13PM -0800, Nikolaus Rath wrote: On Mar 01 2015, Mattia Rizzolo mattia-p8aev1vil9rafugrpc6...@public.gmane.org wrote: On Sat, Feb 28, 2015 at 09:59:01PM -0800, Nikolaus Rath wrote: https://reproducible.debian.net/rb-pkg/s3ql.html claims that S3QL has We performed a rebuild, and your package still FTBFS. This time it looks like it fails at a test with the error dugong.ConnectionClosed: connection closed unexpectedly which is quite nasty at the sight (I don't know what that means). Does this mean that the package tries to connect to the internet? No, it should not. It opens its own mock server on localhost using the any available port. Are you sure about the any available port part? Almost certain. I could be certain if I could see the log with the failed test, but it seems that port 80 on reproducible.debian.net is down at the moment: $weird_things_happen again. https://reproducible.debian.net/rbuild/sid/amd64/s3ql_2.13+dfsg-1.rbuild.log not sure what's wrong at that time. Yes, 100% positive. The tests do not listen to a specific port, they pick an available one at runtime. Whatever goes wrong is more complex than that (and so far specific to the reproducible-builds server). Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] Missing dependencies aren't actually missing
On Mar 01 2015, Mattia Rizzolo mat...@mapreri.org wrote: On Sun, Mar 01, 2015 at 12:04:13PM -0800, Nikolaus Rath wrote: On Mar 01 2015, Mattia Rizzolo mattia-p8aev1vil9rafugrpc6...@public.gmane.org wrote: On Sat, Feb 28, 2015 at 09:59:01PM -0800, Nikolaus Rath wrote: https://reproducible.debian.net/rb-pkg/s3ql.html claims that S3QL has We performed a rebuild, and your package still FTBFS. This time it looks like it fails at a test with the error dugong.ConnectionClosed: connection closed unexpectedly which is quite nasty at the sight (I don't know what that means). Does this mean that the package tries to connect to the internet? No, it should not. It opens its own mock server on localhost using the any available port. Are you sure about the any available port part? Almost certain. I could be certain if I could see the log with the failed test, but it seems that port 80 on reproducible.debian.net is down at the moment: $ ping reproducible.debian.net PING reproducible.debian.net (46.16.73.183) 56(84) bytes of data. 64 bytes from ip46-16-73-183.pbiaas.com (46.16.73.183): icmp_seq=1 ttl=53 time=177 ms 64 bytes from ip46-16-73-183.pbiaas.com (46.16.73.183): icmp_seq=2 ttl=53 time=178 ms 64 bytes from ip46-16-73-183.pbiaas.com (46.16.73.183): icmp_seq=3 ttl=53 time=178 ms ^C --- reproducible.debian.net ping statistics --- 3 packets transmitted, 3 received, 0% packet loss, time 2002ms rtt min/avg/max/mdev = 177.500/177.979/178.224/0.338 ms $ wget http://reproducible.debian.net/ --2015-03-02 08:40:00-- http://reproducible.debian.net/ Resolving reproducible.debian.net (reproducible.debian.net)... 46.16.73.183 Connecting to reproducible.debian.net (reproducible.debian.net)|46.16.73.183|:80... [timeout] Is there any way I could ssh into your build host to debug locally? I just became Debian Maintainer. Any chance this is also due to $weird_things? Both the reguphinxar unstable build and the last ci test from two days ago with don't have that problem (cf. http://ci.debian.net/packages/s/s3ql/unstable/amd64/). Anyway, I performed a local build and this is what debbindiff outputs: http://volatile.mapreri.org/2015-03-01/b46eb200e612629e11c7b6a1c925da36/s3ql.dbd.html e.g. timestamps in sphinx docs. I'll look into that, thanks. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
Re: [Reproducible-builds] randomness_in_objects_inv
On Mar 02 2015, Jérémy Bobbio lunar-8fiuurrzop0dnm+yrof...@public.gmane.org wrote: Nikolaus Rath: Can someone give me more information about how to reproduce this issue? I tried to change the file system order of the sphinx documents, but got identical results: […] You could try with the little LD_PRELOAD library that I had found on Stack Overflow: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=719845#39 http://stackoverflow.com/a/8866709 It will return files from readdir() in reverse order, so if you run it twice on the same set of files, you have a strong guarantee that it's going to be in a different order in a subsequent build. Did you read the example in my mail? ls -U shows the files in readdir order, and for the second build the positions of file1 and file2 were switched - yet the sphinx output unchanged. Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« signature.asc Description: PGP signature ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
[Reproducible-builds] Missing dependencies aren't actually missing
Hello, https://reproducible.debian.net/rb-pkg/s3ql.html claims that S3QL has missing dependencies. If I interpret the build log correctly, the python3-dugong package is missing. But that's actually available in sid: https://tracker.debian.org/pkg/python-dugong Is not available for reproducible builds, because it cannot be build reprocibly itself? Best, -Nikolaus -- GPG encrypted emails preferred. Key id: 0xD113FCAC3C4E599F Fingerprint: ED31 791B 2C5C 1613 AF38 8B8A D113 FCAC 3C4E 599F »Time flies like an arrow, fruit flies like a Banana.« ___ Reproducible-builds mailing list Reproducible-builds@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/reproducible-builds
