RHEL Engineering is moving the tracking of its product development work on RHEL 6 through RHEL 9 to Red Hat Jira (issues.redhat.com). If you're a Red Hat customer, please continue to file support cases via the Red Hat customer portal. If you're not, please head to the "RHEL project" in Red Hat Jira and file new tickets here. Individual Bugzilla bugs in the statuses "NEW", "ASSIGNED", and "POST" are being migrated throughout September 2023. Bugs of Red Hat partners with an assigned Engineering Partner Manager (EPM) are migrated in late September as per pre-agreed dates. Bugs against components "kernel", "kernel-rt", and "kpatch" are only migrated if still in "NEW" or "ASSIGNED". If you cannot log in to RH Jira, please consult article #7032570. That failing, please send an e-mail to the RH Jira admins at rh-issues@redhat.com to troubleshoot your issue as a user management inquiry. The email creates a ServiceNow ticket with Red Hat. Individual Bugzilla bugs that are migrated will be moved to status "CLOSED", resolution "MIGRATED", and set with "MigratedToJIRA" in "Keywords". The link to the successor Jira issue will be found under "Links", have a little "two-footprint" icon next to it, and direct you to the "RHEL project" in Red Hat Jira (issue links are of type "https://issues.redhat.com/browse/RHEL-XXXX", where "X" is a digit). This same link will be available in a blue banner at the top of the page informing you that that bug has been migrated.
Bug 1661597 - Under podman, python recompiles sources even if they are compiled in build time
Summary: Under podman, python recompiles sources even if they are compiled in build time
Keywords:
Status: CLOSED ERRATA
Alias: None
Product: Red Hat Enterprise Linux 8
Classification: Red Hat
Component: podman
Version: 8.0
Hardware: Unspecified
OS: Unspecified
unspecified
unspecified
Target Milestone: rc
: 8.0
Assignee: Nalin Dahyabhai
QA Contact: atomic-bugs@redhat.com
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2018-12-21 17:45 UTC by Jan Pazdziora
Modified: 2020-11-14 08:57 UTC (History)
7 users (show)

Fixed In Version: podman-1.0
Doc Type: If docs needed, set a value
Doc Text:
Clone Of:
Environment:
Last Closed: 2019-11-05 21:00:57 UTC
Type: Bug
Target Upstream Version:
Embargoed:

Attachments (Terms of Use)


Links
System ID Private Priority Status Summary Last Updated
Red Hat Product Errata RHSA-2019:3403 0 None None None 2019-11-05 21:01:19 UTC

Description Jan Pazdziora 2018-12-21 17:45:36 UTC 
Description of problem:

In freeipa-container we use the patch python sources && python -m compileall approach to make needed changes to the python sources yet do all the changes in build time.

Under podman, that no longer works.

Version-Release number of selected component (if applicable):

podman-0.11.1.1-2.git594495d.module+el8+2237+43d1adf6.x86_64

How reproducible:

Deterministic.

Steps to Reproduce:
1. Have Dockerfile where a python source gets updated and compiled:

FROM registry.access.redhat.com/rhel7
RUN rm /usr/lib64/python2.7/gzip.py[co]
RUN touch /usr/lib64/python2.7/gzip.py && python -m compileall /usr/lib64/python2.7/gzip.py

2. Build the image: podman build -t diffit .
3. Run container: podman run --name diffit-c --rm -ti diffit bash
4. In the container, check the files: 
[root@305616dff5ba /]# stat /usr/lib64/python2.7/gzip.py*
  File: '/usr/lib64/python2.7/gzip.py'
  Size: 18694     	Blocks: 40         IO Block: 4096   regular file
Device: 37h/55d	Inode: 101856031   Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-21 17:09:43.000000000 +0000
Modify: 2018-12-21 17:09:43.000000000 +0000
Change: 2018-12-21 17:09:42.638174210 +0000
 Birth: -
  File: '/usr/lib64/python2.7/gzip.pyc'
  Size: 15071     	Blocks: 32         IO Block: 4096   regular file
Device: 37h/55d	Inode: 101856032   Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-21 17:09:43.000000000 +0000
Modify: 2018-12-21 17:09:43.000000000 +0000
Change: 2018-12-21 17:09:42.638174210 +0000
 Birth: -

5. Note that the timestamps on both .py and .pyc are the same.
6. In the container, use the python module:
[root@305616dff5ba /]# python -c 'import gzip'

7. In the container, stat the files again: stat /usr/lib64/python2.7/gzip.py*
8. In different terminal, use podman diff: podman diff diffit-c

Actual results:

  File: '/usr/lib64/python2.7/gzip.py'
  Size: 18694     	Blocks: 40         IO Block: 4096   regular file
Device: 37h/55d	Inode: 101856031   Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-21 17:09:43.000000000 +0000
Modify: 2018-12-21 17:09:43.000000000 +0000
Change: 2018-12-21 17:09:42.638174210 +0000
 Birth: -
  File: '/usr/lib64/python2.7/gzip.pyc'
  Size: 15071     	Blocks: 32         IO Block: 4096   regular file
Device: 37h/55d	Inode: 68517034    Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-21 17:43:02.338619328 +0000
Modify: 2018-12-21 17:43:02.338619328 +0000
Change: 2018-12-21 17:43:02.338619328 +0000
 Birth: -

and

C /usr/lib64/python2.7
C /usr/lib64/python2.7/gzip.pyc

Expected results:

Stat should still show the same timestamp for both .py and .pyc, podman diff should not show anything.

Additional info:
Comment 1 Daniel Walsh 2018-12-21 20:23:32 UTC 
I know that Nalin was playing around with some timestamp issues, that might fix this problem.

Could you by any change check this out with podman 0.12?
Comment 2 Jan Pazdziora 2018-12-22 20:54:37 UTC 
With podman-0.12.1.2-4.git9551f6b.module+el8+2594+933f97c7.x86_64, the behaviour is the same:

host# podman run --name diffit-c --rm -ti diffit bash
[root@b9e3cc619f91 /]# stat /usr/lib64/python2.7/gzip.py*
  File: '/usr/lib64/python2.7/gzip.py'
  Size: 18694     	Blocks: 40         IO Block: 4096   regular file
Device: 2ch/44d	Inode: 33937730    Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-22 20:52:34.000000000 +0000
Modify: 2018-12-22 20:52:34.000000000 +0000
Change: 2018-12-22 20:52:33.869019692 +0000
 Birth: -
  File: '/usr/lib64/python2.7/gzip.pyc'
  Size: 15071     	Blocks: 32         IO Block: 4096   regular file
Device: 2ch/44d	Inode: 33937731    Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-22 20:52:34.000000000 +0000
Modify: 2018-12-22 20:52:34.000000000 +0000
Change: 2018-12-22 20:52:33.870019707 +0000
 Birth: -
[root@b9e3cc619f91 /]# python -c 'import gzip'
[root@b9e3cc619f91 /]# stat /usr/lib64/python2.7/gzip.py*
  File: '/usr/lib64/python2.7/gzip.py'
  Size: 18694     	Blocks: 40         IO Block: 4096   regular file
Device: 2ch/44d	Inode: 33937730    Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-22 20:52:34.000000000 +0000
Modify: 2018-12-22 20:52:34.000000000 +0000
Change: 2018-12-22 20:52:33.869019692 +0000
 Birth: -
  File: '/usr/lib64/python2.7/gzip.pyc'
  Size: 15071     	Blocks: 32         IO Block: 4096   regular file
Device: 2ch/44d	Inode: 68104574    Links: 1
Access: (0644/-rw-r--r--)  Uid: (    0/    root)   Gid: (    0/    root)
Access: 2018-12-22 20:53:08.261520518 +0000
Modify: 2018-12-22 20:53:08.262520533 +0000
Change: 2018-12-22 20:53:08.262520533 +0000
 Birth: -

host# podman diff diffit-c
C /usr/lib64/python2.7
C /usr/lib64/python2.7/gzip.pyc
Comment 3 Nalin Dahyabhai 2019-01-10 22:11:25 UTC 
Yes, this looks like it's another result of golang 1.10's archive/tar package rounding timestamps instead of truncating them, this time tripping up the python interpreter, which decides whether or not a .py file needs to be byte compiled by comparing the mtime on the .py file with a 32-bit value that was stored in the .pyc file, which I expect it's arriving at by truncating a value.

We added a workaround in containers/storage#252 to force truncation when creating layer diffs, but libpod v0.12.1.2 was built using a version of a containers/storage from just before we merged that workaround, so it's affected.  libpod was updated to a sufficiently new version in libpod#2071, so its next release should avoid creating images that cause this to happen.
Comment 4 Jan Pazdziora 2019-01-11 09:49:08 UTC 
I confirm that I no longer see the podman diff issue in Travis CI: https://travis-ci.org/adelton/freeipa-container/builds/477234614, so at least upstream seems to be fixed now.
Comment 5 Daniel Walsh 2019-01-11 14:32:19 UTC 
Fixed in podman 1.0.
Comment 8 Joy Pu 2019-09-27 06:43:06 UTC 
Can reproduce with podman-0.10.1.3-5.gitdb08685.el8+2131+7e3e9e07.x86_64. Test with podman-1.4.2-5.module+el8.1.0+4240+893c1ab8.x86_64. It works as expect and podman diff will not show anything now. So set this to verified.
Comment 10 errata-xmlrpc 2019-11-05 21:00:57 UTC 
Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory, and where to find the updated
files, follow the link below.

If the solution does not work for you, open a new bug report.

https://access.redhat.com/errata/RHSA-2019:3403
Note You need to log in before you can comment on or make changes to this bug.