Bug 1470

Summary:	need summary
Product:	[Retired] Red Hat Linux	Reporter:	vasz
Component:	egcs	Assignee:	David Lawrence <dkl>
Status:	CLOSED WORKSFORME	QA Contact:
Severity:	high	Docs Contact:
Priority:	high
Version:	5.2
Target Milestone:	---
Target Release:	---
Hardware:	alpha
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	1999-03-18 21:51:03 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description vasz 1999-03-10 15:44:45 UTC

Either the kernel, or rather the C compiler
and library in 5.2 very slow. I do numerical
calculations and when I upgraded from 5.0 to
5.2 the speed of my calculation
substationally slowed down (about on third).
When went back I have gained the speed back.
I noticed, that on alpha from 5.1 the new,
and unfortunately experimental compiler,
the egcs is used. Since I am not an
expertise in kernel compiling, I am not sure
where the problen came from. It is because
the 5.1 and 5.2 kernel was compiled with
egcs or simply the egcs does not give the
same speed as the gcc compiler. I think this
is a serious problem, since many pepole,
like myself, use the alpha station for
numerical speed.

Comment 1 Preston Brown 1999-03-18 21:51:59 UTC

We have not noticed similar performance problems here.  Without
further information, we cannot determine if there is indeed a
problem.  Please provide more information if you do reopen this bug.

Comment 2 vasz 1999-03-23 19:58:59 UTC

I used the GROMACS molecular dynamics package when
I fund this speed problem. I have repeated the
install 5.2 and do GROMACS, install 5.0 from
scratch compile GROMACS and run severeal times and
the major speed problems comes out consistently.
As I said the speed problem only comes when one
dealing with NUMERICAL (floating point)
calculations.