Article View: NET.general
Article #1711/44 float bug
From: phsdennis
Date: Mon, 25 May 1981 14:41
Date: Mon, 25 May 1981 14:41
13 lines
742 bytes
742 bytes
There is a bug in the implementation of the 11/44 float box. It affects only the stexp instruction (as far as I've been able to discover). The problem is that when you use auto-increment or -decrement as the destination mode, the register gets changed by the current size of the float word, NOT by 2. This causes havoc if you do it to the sp, for instance. That actual code doesn't occur in v7 anywhere; it was discovered when I ported fc to v7 (we needed it for our 11/60 satellite sys- tem). This has been reported to DEC, but no action has been taken.
Technical Details
Message-ID:
<anews.Aphs.109>Path:
arc1-ca.pugleaf.net!utzoo!duke!phs!dennisShow Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!duke!phs!dennis From: phs!dennis Newsgroups: NET.general Subject: 11/44 float bug Date: Mon, 25 May 1981 14:41:21 +0000 Message-ID: <anews.Aphs.109>