Bug 44382 - Output from compiler broken up
Summary: Output from compiler broken up
Alias: None
Product: Red Hat Linux
Classification: Retired
Component: gcc   
(Show other bugs)
Version: 7.1
Hardware: i386
OS: Linux
Target Milestone: ---
Assignee: Jakub Jelinek
QA Contact: David Lawrence
Depends On:
TreeView+ depends on / blocked
Reported: 2001-06-13 12:30 UTC by André Johansen
Modified: 2007-04-18 16:33 UTC (History)
0 users

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of:
Last Closed: 2001-06-13 12:30:04 UTC
Type: ---
Regression: ---
Mount Type: ---
Documentation: ---
Verified Versions:
Category: ---
oVirt Team: ---
RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: ---

Attachments (Terms of Use)

Description André Johansen 2001-06-13 12:30:01 UTC
When compiling code with the GCC version that comes with RHL v7.1,
it breaks output lines (probably to stay inside 80 column-wide

This is very annoying when compiling in environments that support
wider lines (like within Emacs), as the output gets much harder to
read.  When compiling C++ code, you often end up with half the function
signature on one line, the rest on the following line, and warnings
including STL code spans loads of lines.

I'd suggest going back to the old output (like in GCC v2.95.3), and
have the new line-breaking output hidden behind an option-switch to GCC.

Comment 1 Jakub Jelinek 2001-06-13 12:32:07 UTC
Use -fmessage-length=0 if it annoys you.

Comment 2 André Johansen 2001-06-13 13:24:37 UTC
Wasn't aware of that one (not mentioned in "g++ --help", found it in the info

However, I still think this is a broken change.
With the new switch, one can go back to the old style output, but then
you'll get annoying output if compiling with other versions of GCC (like
the one on RHL v6.2).
$ g++ -fmessage-length=0 str.cpp
cc1plus: Invalid option `-fmessage-length=0'

We try to maintain code for several different platforms/compilers
in one simple makefile-system, and that breaks with this change.

The best solution I can come up with now, is to set system-wide
env-variables containing that switch on RHL v7.1 machines, but that is
cumbersome...  Is it possible to set the default to 0 in the spec file or
somewhere else?

Comment 3 Jakub Jelinek 2001-06-13 13:30:15 UTC
gcc-3.0 (to be released on friday) uses the same settings, so you wouldn't
gain anything if we changed in in 2.96-RH, since you'd have to adapt your
Makefiles for 3.0 anyway.

Note You need to log in before you can comment on or make changes to this bug.