https://blog.kempniu.pl/en/2015/04/29/idrac-virtual-console-linux-and-segfaults.html
TL;DR: just upgrade your iDRAC firmware to version 2.10.10.10 or later.

Thanks very much, Kempniu. We have over one hundred machines for which 1.66.65 is the latest available from support.dell.com, so this will ...

Kempniu wrote:I'll let you guys know once (if?) they give me the go-ahead.

You mentioned that there may be a more elegant work-around than deleting /usr/lib/libssl.so. Is it OK for you to share it?

Kempniu: You are a lifesaver! Your procedure fixed the issue for me also on 1.56.55 (Build 05) and 1.57.57 (Build 04).

I would be very interested to learn how you figured out what the problem is!

Tried using my desktop's cable and port (which work) with no luck.

All of a sudden my ThinkPad's Intel 82573L Ethernet card stopped working at 1000Mb/s.

Some packets seem to get through, but it's very slow (ssh lags) and ethtool reports CRC errors on received packets.

Lights on the switch's port where laptop is connected blink at high rate, indicating activity ...

Taking a guess at some theories:

- A lot of emerge -p time is spent opening and reading the thousands of small files in the tree. Perhaps this is probably as much a measure of how well python and portage avoid stomping on the parts of the CPU cache that the kernel like to use, as it is of speed of ...

I noticed that dstat in the dsk/total column seems to report twice the amount of data transferred to and from disk.

For example, if I am transferring a file over 100Mbps network, scp will report about 10MB/s, dstat's network "send" column will report about 10MB/s, but the disk read column is ...

All this proves is that things compile faster with -mtune=i386, it gives you no indication of how well the resulting code will perform,
and that makes perfect sense, i386 being more generic, hence less optimized, and optimizations do increase compile time.
I am not measuring compile time. I am ...

I AM NOT TIMING COMPILING TIME!

Q: What is the best CFLAGS?*
A: CFLAGS="-mtune=i386 -O2"

* -- assuming emerge is a good indicator of overall performance :)

I tested various gcc optimizations using emerge (i.e. using python) and it turns out that -mtune=i386 produces the fastest code. I also did ...

I am doing some benchmarks that require me to frequently re-emerge a lot of tiny packages. confcache would really speed things up for me. Does anyone know where I can find its source?

http://gentooexperimental.org/~ferringb/confcache/ no longer exists.

I like your idea.

The tree structure seems a bit complicated though. I think it would be better to have a moderated thread per package where people would post their CFLAGS, hardware, package version and benchmarks. The first post in the thread could summarize available tools to benchmark that ...

Having experienced performance issues in X I benchmarked effects various gcc optimizations. Results are below.


CPU: Core 2 Duo T7200, 2GHz, FSB: 667Mhz, L2: 4MB
Kernel: 2.6.25.8 x86_64 SMP, 100MHz clock, no preemption
Gentoo: gcc 4.3.1; glibc 2.8_p20080602; xorg-server-1.4.2; xf86-video-i810-2 ...

Checkout http://madwifi.org/wiki/Compatibility
Also, pretty much any card by Ubiquiti should work

try:

I am having some performance issues in X because of using 3 monitors and xinerama. I'd like to see if gcc's profile-directed optimization will improve performance. To do this, I need to know which library or libraries are responsible for most CPU usage. Does anybody know how to do this?

For example ...

Scoth wrote:Bah, that's it. I knew it was something simple. I wonder who decided to name those like that...

Linus

If anyone's interested, here's a VMware virtual machine with KDE4 emerged. KDE4 still needs setting up, but maybe some people will find it to be a useful starting point.

size: ~900MB
users: root, kde4
pass: gentoo4lyfe (both accounts)
http://pages.cs.wisc.edu/~vladimir/kde4.tar.gz


Instructions ...

i92guboj wrote:A good way to check if the value reported by top is true or false is to look at the temperature.

If the cpu is 100% occupied for a long time, then your cpu temperature will be a lot higher. The difference is sensible enough.

Will temperature rise even if the CPU is in wait state?

You should look for processes with state D from top or ps aux.

Other things to check:

Is the system paging/disks being hit hard.
Are their a lot of network connections/network IO.

The issue happens when the machine is completely idle -- only X and top are running, no network, no disk activity ...

I am on an core 2 system and I've noticed that often one of the cores spends almost all its time in io wait even though no process appears to be using it.

That is, top's CpuX row shows near 100% wait (or split 50-50 between cores) but shows no process that is causing that. The %CPU column either ...

Well, apparently npviewer.bin segfaulted and for some reason wrote a bunch of crap into my .viminfo. Removing that file fixed the problem. Not sure what the hell happened though...

For some reason, when I run vi or gvim as non-root user I get this message:

E575: viminfo: Illegal starting char in line: which: no soundwrapper in (/usr/bin:/bin:/usr/X11R6/bin:/usr/kde/3.5/bin:/usr/local/bin:/usr/bin:/bin:/opt/bin:/usr/x86_64-pc-linux-gnu/gcc-bin/4.1.2:/opt/blackdown-jdk-1.4.2 ...

I guess you mean /etc/conf.d/net? There is an option to timeout the association but I don't know if that's what I'm looking for. For the moment it doesn't wait 60 seconds (the default) before aborting. On the other side, I don't want my booting to freeze on wireless discovery.
Sorry, my bad. I ...

What you are trying to do will create a new security risk and you shouldn't do it.

With that out of the way, here are my 2 cents.

firewalls all traffic, except for port 80
Most likely DNS is not firewalled either. If you are hardcore,and have a little free time, you could come up with a way to ...

/etc/init.d/net?