Hide Forgot
---Problem Description--- libdfp testsuite failures in the S/390 hardware version. ---Steps to Reproduce--- make check in the libdfp build dir RedHat requested to track further libdfp problems in separate bugzillas. This bugzilla is about a problem which occurred after the last libdfp version 1.0.6 which used RedHat for rebasing. > 1.Server architecture(s) (please list all effected) (x86/POWER6/Z/etc.): s390x > 2.Server type (9117-MMA/HS20/s390/etc.): s390 > 3.General component (desktop/kernel/base OS/dev tools/etc.): dev tools > 4.Other components involved (ixgbe/java/emulex/etc.): none > 5.Does the server have the latest GA firmware? yes > 6.Has the problem been shown to occur on more than one system? yes > 7.Is a tested patch available? yes, see attachement > If yes to the above, has it been approved upstream? yes, it is part of libdfp 1.0.7 > 8.What is the latest official Red Hat build on which this bug has been > seen? RHEL 6
Created attachment 475662 [details] libdfp-s390-fix
IBM Update: Run the libdfp testsuite in the build directory with: make check The hardware variant of libdfp on s390 produces failures in the testsuite: In the build dir after make check: ls -l *.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-amort.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-decode.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-fpclassify.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-get_digits.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-isfinite.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-isinf.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-isnan.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-log10d.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-logd.out -rw------- 1 krebbel krebbel 1804 2011-01-27 17:16 test-numdigits.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-param.out -rw------- 1 krebbel krebbel 5092 2011-01-27 17:16 test-printf.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-quantize.out -rw------- 1 krebbel krebbel 816 2011-01-27 17:16 test-round.out -rw------- 1 krebbel krebbel 0 2011-01-27 17:16 test-strtod.out The patch fixes the 3 bugs above in the s390 implementation of numdigits, setexp and __fe_dec_setround.
------- Comment From krebbel.ibm.com 2011-01-27 11:22 EDT------- (In reply to comment #7) > Hello, > Please send soem details on what the problem is that you are seeing. Any error > messages. > Please also provide steps to reproduce this issue. ------- Comment From prem.karat.ibm.com 2011-03-01 07:35 EDT------- (In reply to comment #9) > Hello, > I have sent this up for engineering to review. > > Thank You > Joe Kachuck Hi Joe Did you hear back from your engineering team? Cheers, Prem
------- Comment From mgrf.com 2011-03-31 16:22 EDT------- Hi Joe, John, based on our calls libdfp should get these fixes for R 6.1. If verification fail we should open a new BZ. Seems we can just continue to use this ?
------- Comment From rsa.com 2011-04-02 09:44 EDT------- (In reply to comment #14) > Hi Joe, John, > > based on our calls libdfp should get these fixes for R 6.1. > If verification fail we should open a new BZ. > > Seems we can just continue to use this ? I checked the src rpm for the latest RHEL 6.1 supplemental and these fixes are not in place.
The fix for this bug is included in snapshot 2 which went out to partners on April 6. Please test and provide feedback.
------- Comment From mgrf.com 2011-04-12 11:47 EDT------- This is verified on System z on R 6.1 snapshot 2 - closing on IBM site Thx
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on therefore solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHBA-2011-0659.html