[Search for users]
[Overall Top Noters]
[List of all Conferences]
[Download this site]
| Title: | USG buildhelp questions/answers |
|
| Moderator: | SMURF::FILTER |
|
| Created: | Mon Apr 26 1993 |
| Last Modified: | Mon Jan 20 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 2763 |
| Total number of notes: | 5802 |
2559.0. "Is this a problem waiting to happen ?" by AOSG::FILTER (Automatic Posting Software - mail to flume::puck) Wed Sep 11 1996 13:12
Date Of Receipt: 11-SEP-1996 11:14:29.25
From: FLUME::"[email protected]" "11-Sep-1996 1049"
To: [email protected]
CC: [email protected]
Subj: Is this a problem waiting to happen ?
We now have links from support pool to support pool, so that a
single source file change will be sufficient to allow propagation of
a fix across several versions. However we do not have any way to
determine the sccsid of .h files that were used to build a .o file.
The .h file could have a change that significantly alters the way
the code performs (we must avoid .h file changes if at all possible
- we've known that, but I've seen some go through, but now it's
imperative to avoid it at all costs).
The sccsid of the .o files built with two different .h files would
still be the same. Further, if someone takes the .o from one version
and puts it on another, if the sccsid is the same there is no easy
way to spot in a vmunix that you have the wrong patch (thank god for
checksums, but we don't usually ask for checksums on all .o files
when we get a CLD with a crash, and how about .mods? have we started
issuing Checksums on .mod files ? We'd better.).
No reason to panic.
Charlie
| T.R | Title | User | Personal
Name | Date | Lines
|
|---|