Opened 9 years ago

Closed 9 years ago

#9338 closed enhancement (fixed)

upgrade PyCrypto to upstream 2.1.0

Reported by: mvngu Owned by: tbd
Priority: major Milestone: sage-4.5.3
Component: packages: standard Keywords:
Cc: Merged in: sage-4.5.3.alpha0
Authors: Minh Van Nguyen Reviewers: David Kirkby
Report Upstream: N/A Work issues:
Branch: Commit:
Dependencies: Stopgaps:

Description (last modified by mvngu)

As the subject says. The latest version also fixes the issue with ARC2 reported at http://www.securityfocus.com/bid/33674/info. Currently, the PyCrypto spkg maintains patches for this issue. Upgrading to the latest upstream version would mean we no longer need to maintain those patches in the spkg itself.

Apply:

http://sage.math.washington.edu/home/mvngu/spkg/standard/pycrypto/pycrypto-2.1.0.spkg

Changes in this upgraded spkg include:

  • Upgrade to PyCrypto 2.1.0.
  • The bug at http://www.securityfocus.com/bid/33674/info is fixed in PyCrypto 2.1.0, so remove patches/ARC2.c and patches/ARC2.c.patch.
  • Flesh out and update the file SPKG.txt.
  • Add spkg-check to run the test suite of PyCrypto.
  • Add a check to spkg-install to check that PyCrypto installs fine.

Make sure to update #9281 when spkg-check runs the test suite fine.

Change History (8)

comment:1 Changed 9 years ago by mvngu

  • Authors set to Minh Van Nguyen
  • Description modified (diff)

I'm still testing this spkg, so it's not ready for review.

comment:2 Changed 9 years ago by mvngu

  • Description modified (diff)

comment:3 Changed 9 years ago by mvngu

  • Status changed from new to needs_review

The test results are summarized below. I only ran the spkg-check script of the PyCrypto spkg. This was tested with Sage 4.4.4. With the exception of cicero.skynet, all machines reported here are 64-bit.

  1. bsd.math: Mac OS X 10.6.4, GCC 4.2.1, Dual-Core Intel Xeon @ 2.66 GHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  2. cicero.skynet: 32-bit Fedora 12, GCC 4.5.0, Intel(R) Pentium(R) 4 CPU @ 2.66GHz
    • build: yes
    • doctest: one failure in libs/mwrank/mwrank.pyx, as reported here
    • spkg-check: pass
  3. cleo.skynet: Red Hat Enterprise Linux Server 5.3, GCC 4.5.0, IA-64 Itanium 2 @ 1594.000726 MHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  4. eno.skynet: Fedora 12, GCC 4.5.0, Intel(R) Xeon(R) CPU E5345 @ 2.33GHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  5. flavius.skynet: Fedora 12, GCC 4.5.0, AMD Opteron(tm) Processor 248 @ 2193.570 MHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  6. gcc11.fsffrance: Debian 5.0, GCC 4.3.2, Dual-Core AMD Opteron(tm) Processor 2212 @ 2000.085 MHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  7. gcc16.fsffrance: Debian 5.0, GCC 4.3.2, Quad-Core AMD Opteron(tm) Processor 8354 @ 2194.496 MHz
    • build: yes
    • doctest: 2 failures in schemes/elliptic_curves/lseries_ell.py
    • spkg-check: pass
  8. gcc100.fsffrance: Debian 5.0, GCC 4.3.2, AMD Opteron(tm) Processor 252 @ 2600.011 MHz
    • build: yes
    • doctest: failures in modules/free_module.py
    • spkg-check: pass
  9. iras.skynet: SUSE Linux Enterprise Server 10 SP1, GCC 4.5.0, IA-64 @ 1594.000683 MHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  10. lena.skynet: Fedora 12, GCC 4.5.0, AMD Phenom(tm) II X4 940 Processor @ 3000.000 MHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  11. menas.skynet: openSUSE 11.1, GCC 4.5.0, Intel(R) Core(TM)2 Quad CPU Q6600 @ 2.40GHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  12. rh.math: Ubuntu 10.04 LTS, GCC 4.4.3, Six-Core AMD Opteron(tm) Processor 8439 SE @ 800.000 MHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  13. sage.math: Ubuntu 8.04.4 LTS, GCC 4.2.4, Intel(R) Xeon(R) CPU X7460 @ 2.66GHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  14. sextus.skynet: Fedora 12, GCC 4.5.0, Intel(R) Xeon(TM) CPU @ 3.60GHz
    • build: yes
    • doctest: pass
    • spkg-check: pass
  15. taurus.skynet: Fedora 12, GCC 4.5.0, Intel(R) Xeon(R) CPU X5570 @ 2.93GHz
    • build: no, due to Linbox failing to build on taurus. This is a known issue. But forcing an installation of the PyCrypto spkg with "./sage -f <...>", and then running spkg-check, worked fine.
    • doctest: N/A since Sage 4.4.4 fails to build on taurus
    • spkg-check: pass

comment:4 Changed 9 years ago by mvngu

  • Status changed from needs_review to needs_info

I'm still waiting for build/test results on t2.math.

comment:5 Changed 9 years ago by mvngu

  • Status changed from needs_info to needs_review

Build fine on t2.math and spkg-check of PyCrypto passes. This is now ready for review.

comment:6 Changed 9 years ago by mvngu

Also build fine on Cygwin (winxp2) and spkg-check pass.

comment:7 Changed 9 years ago by drkirkby

  • Reviewers set to David Kirkby
  • Status changed from needs_review to positive_review

You have clearly tested this thoroughly Minh - I wish all Sage developers were like you.

I just tested it on two systems:

  • OpenSolaris 2009.06 on a 3.33 GHz Intel W3580 Xeon. 64-bit build. All tests in spkg-check pass. Since Sage crashes immediately on startup, I can't comment on doctests.
  • Solaris 10 on a Sun Blade 2000, with two Sun UltraSPARC III+ processors. 64-bit build. All tests in spkg-check pass. Since Sage is unstable on 64-bit Solaris 10 on SPARC, it's pointless me running any doctests. (Sage does now just about work on 64-bit SPARC. I got it running for the first time yesterday, so needless to say, it is far from perfect).

hg status

shows no problems, so positive review.

Dave

comment:8 Changed 9 years ago by mpatel

  • Merged in set to sage-4.5.3.alpha0
  • Resolution set to fixed
  • Status changed from positive_review to closed
Note: See TracTickets for help on using tickets.