94011 – madwifi-driver 0.1_pre20050420 has performance issues

Bug 94011 - madwifi-driver 0.1_pre20050420 has performance issues

Summary: madwifi-driver 0.1_pre20050420 has performance issues

Status:	RESOLVED UPSTREAM

Alias:	None

Product:	Gentoo Linux
Classification:	Unclassified
Component:	New packages (show other bugs)
Hardware:	x86 Linux

Importance:	High normal (vote)
Assignee:	Mobile Herd (OBSOLETE)

URL:
Whiteboard:
Keywords:

Depends on:	94914
Blocks:
	Show dependency tree

Reported:	2005-05-25 16:23 UTC by Olliver Schinagl
Modified:	2005-06-04 16:30 UTC (History)
CC List:	2 users (show)

See Also:
Package list:
Runtime testing required:	---

Attachments
Add an attachment (proposed patch, testcase, etc.)

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

Description Olliver Schinagl 2005-05-25 16:23:17 UTC

I've been using madwifi-driver 0.1_pre20050420 since well 20050420 and it works
well as long as I don't transfer data at maximum speed for more then 10 seconds.
After that the connection drops and keeps on dropping after a second or two of
connection/data transfer.
This problem didn't exist with the previous version, so maybe a new snapshot is
a good idea?

Reproducible: Always
Steps to Reproduce:
scp a large file (600mb for example) from a remote host. be sure to be able to
max out the wireless link (e.g. have a 100mbit backend that can supply 100mbit
with ease.
Actual Results:  
Connection drops/reconnects/drops etc etc

Expected Results:  
Connection remains stable and the transfer of the file is possible without a
problem.

Comment 1 solar (RETIRED) gentoo-dev

2005-06-04 05:01:18 UTC

We did not write the software but rather are the packagers.
Please check with upstream.

Comment 2 Henrik Brix Andersen 2005-06-04 05:07:33 UTC

Closing as UPSTREAM.

Comment 3 Olliver Schinagl 2005-06-04 07:37:49 UTC

which is why i'm saying the current cvs snapshot package doesn't work properly
under load. We need to either go back one version, or go up one if possible.
This clearly goes under gentoo packaging.

Comment 4 Henrik Brix Andersen 2005-06-04 07:44:58 UTC

Are you saying that this is fixed in upstream CVS?

Comment 5 Olliver Schinagl 2005-06-04 07:48:22 UTC

I don't know. I do know that the current version has this problem and it should
be masked or similar. My proposal therefore is go back one version.

I don't know if it is fixed in the current CVS. But since nobody can even
confirm this issue ...

Comment 6 Bob Ketterhagen (RETIRED) gentoo-dev

2005-06-04 09:02:19 UTC

What card are you powering with madwifi?

Comment 7 Bob Ketterhagen (RETIRED) gentoo-dev

2005-06-04 09:08:35 UTC

It has already been explained in the mailing list (see the archive). 
The result is  that development in madwifi head cvs has stopped and 
only moves on in madwifi-bsd. The old developers of madwifi-head are 
not active in madwifi project (workin on bsd atheros driver version) 
anymore and now we have chosen to take over the responsibility for that.

Furthermore, anyone who thinks he is able to contribute (coding, 
documentation, howtos, bug reports) to madwifi is welcome.


-- 
Regards,
Sebastian Weitzel

Hey guys,

I think it's update time.

Comment 8 Olliver Schinagl 2005-06-04 10:40:19 UTC

I have a:
0000:02:02.0 Ethernet controller: Atheros Communications, Inc. AR5212 802.11abg
NIC (rev 01)

which came with my IBM T42 thinkpad.

Comment 9 Bob Ketterhagen (RETIRED) gentoo-dev

2005-06-04 10:51:09 UTC

Please do this.
emerge sync
emerge unmerge madwifi-driver madwifi-tools
ACCEPT_KEYWORDS="~x86" emerge -v madwifi-bsd
Using madwifi-bsd should correct your problem.
We are currently working on updating madwifi*

Comment 10 Olliver Schinagl 2005-06-04 16:30:34 UTC

horton, thanks.

I think this answer is sufficent enough to close this bug.