Closed Bug 1246040 Opened 9 years ago Closed 9 years ago

CentOS bootstrapping broken due to manual yasm rpm install

Categories

(Firefox Build System :: General, defect)

defect
Not set
normal

Tracking

(firefox47 fixed)

RESOLVED FIXED
mozilla47
Tracking Status
firefox47 --- fixed

People

(Reporter: gps, Assigned: gps)

References

Details

Attachments

(1 file)

We get a similar bootstrap failure on both CentOS 6 and 7. Executing as root: rpm -ivh http://pkgs.repoforge.org/yasm/yasm-1.1.0-1.el6.rf.i686.rpm Retrieving http://pkgs.repoforge.org/yasm/yasm-1.1.0-1.el6.rf.i686.rpm warning: /var/tmp/rpm-tmp.d4tAO2: Header V3 DSA/SHA1 Signature, key ID 6b8d79e6: NOKEY error: Failed dependencies: libc.so.6 is needed by yasm-1.1.0-1.el6.rf.i686 libc.so.6(GLIBC_2.0) is needed by yasm-1.1.0-1.el6.rf.i686 libc.so.6(GLIBC_2.1) is needed by yasm-1.1.0-1.el6.rf.i686 libc.so.6(GLIBC_2.3) is needed by yasm-1.1.0-1.el6.rf.i686 libc.so.6(GLIBC_2.3.4) is needed by yasm-1.1.0-1.el6.rf.i686 libc.so.6(GLIBC_2.4) is needed by yasm-1.1.0-1.el6.rf.i686 libc.so.6(GLIBC_2.7) is needed by yasm-1.1.0-1.el6.rf.i686 Traceback (most recent call last): File "bootstrap.py", line 161, in <module> sys.exit(main(sys.argv)) File "bootstrap.py", line 152, in main dasboot.bootstrap() File "/tmp/tmpBTtDK5/mozboot/bootstrap.py", line 137, in bootstrap parser.add_option('--no-interactive', dest='no_interactive', action='store_true', File "/tmp/tmpBTtDK5/mozboot/centos.py", line 62, in install_browser_packages File "/tmp/tmpBTtDK5/mozboot/base.py", line 160, in run_as_root File "/usr/lib64/python2.7/subprocess.py", line 542, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '['rpm', '-ivh', 'http://pkgs.repoforge.org/yasm/yasm-1.1.0-1.el6.rf.i686.rpm']' returned non-zero exit status 1
This might be due to running inside Docker. Our bootstrap code is looking at platform.uname()[2]. Under Docker: >>> platform.uname() ('Linux', '9ca9b44854ba', '4.1.17-boot2docker', '#1 SMP Mon Feb 1 16:37:01 UTC 2016', 'x86_64', 'x86_64')
Assignee: nobody → gps
Status: NEW → ASSIGNED
We shouldn't even look at platform.uname()[2]. At best, it should be platform.uname()[4] or platform.machine(), but in fact, platform.architecture is about as likely to be correct, if not more. (the system could be using a 64-bits kernel, but 32-bits user-space, in which case the architecture of the currently running python would be correct, but not platform.machine)
Comment on attachment 8716119 [details] MozReview Request: Bug 1246040 - Change architecture detection on CentOS; r?glandium Review request updated; see interdiff: https://reviewboard.mozilla.org/r/33725/diff/1-2/
Attachment #8716119 - Attachment description: MozReview Request: Bug 1246040 - More robustly detect 64-bit architecture when bootstrapping CentOS; r?mshal → MozReview Request: Bug 1246040 - Change architecture detection on CentOS; r?glandium
Attachment #8716119 - Flags: review?(mshal) → review?(mh+mozilla)
Comment on attachment 8716119 [details] MozReview Request: Bug 1246040 - Change architecture detection on CentOS; r?glandium https://reviewboard.mozilla.org/r/33725/#review30441
Attachment #8716119 - Flags: review?(mh+mozilla) → review+
Status: ASSIGNED → RESOLVED
Closed: 9 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla47
Product: Core → Firefox Build System
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: