Article View: gnu.gcc.help
Article #50445Floating point operations Broken in 2.4.5 or a bad build?
From: Edward Benyukhis
Date: Thu, 30 Sep 1993 20:11
Date: Thu, 30 Sep 1993 20:11
28 lines
1390 bytes
1390 bytes
I am attepting to build GCC 2.4.5 on an Integrated Micro Products XM fault tolerant machine. This is a Motorola 68040 computer running Unix SVR3.2 OS. When I change TARGET_DEFAULT #define in mot3300.h to '01000' to indicate that I am building for a MC68040_ONLY environment, the build blows up somewhere around archiving 'floating point' primitives inti the 'tmplibgcc2.a'. Everything works OK if I define the TARGET_DEFAULT to m68020 with 68881/82 co-processor, but after I test the floating point operations, they do not work as expected producing INCORRECT results. Ex: a = 923.2344; b = 12.9344 a/b produces 30274224176.101364 Can anyone assist in resolution of this issue? Thank you, Edward Benyukhis
Technical Details
Message-ID:
<1993Sep30.201119.27139@lmpsbbs.comm.mot.com>Path:
arc1-ca.pugleaf.net!gmd.de!newsserver.jvnc.net!udel!gatech!swrinde!cs.utexas.edu!uunet!ftpbox!mothost!lmpsbbs!benyukhShow Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 11318c,af15299a742d6e0,start X-Google-Attributes: gid11318c,public X-Google-ArrivalTime: 1993-09-30 16:44:10 PST Newsgroups: gnu.gcc.help Path: arc1-ca.pugleaf.net!gmd.de!newsserver.jvnc.net!udel!gatech!swrinde!cs.utexas.edu!uunet!ftpbox!mothost!lmpsbbs!benyukh From: benyukh@ssd.comm.mot.com (Edward Benyukhis) Subject: Floating point operations Broken in 2.4.5 or a bad build? Organization: Motorola Date: Thu, 30 Sep 1993 20:11:19 GMT Message-ID: <1993Sep30.201119.27139@lmpsbbs.comm.mot.com> Sender: news@lmpsbbs.comm.mot.com (Net News) Nntp-Posting-Host: 145.1.185.34 Lines: 30