Intel Is Working On A Linux Networking Stack For Small Systems

[phoronix]

Phoronix: Intel Is Working On A Linux Networking Stack For Small Systems

For supporting Linux networking on very small embedded systems like the Intel Quark, developers at Intel are working on a lightweight networking stack to fit on such systems...

http://www.phoronix.com/vr.php?view=MTY4MzM

[droidhacker]

Huh.

Their reference design "galileo" specs 256 MB DRAM.
Given the cost of RAM these days, anything less than that and the vendor is just being SO absurdly cheap that they aren't worth dealing with.

... not that it isn't a good idea to chop off unnecessary cruft, but this is getting slightly ridiculous.

https://communities.intel.com/docs/DOC-21835

[cbamber85]

I thought most embedded systems used lwip, is that too big for Intel!?

[vadix]

Huh.

Their reference design "galileo" specs 256 MB DRAM.
Given the cost of RAM these days, anything less than that and the vendor is just being SO absurdly cheap that they aren't worth dealing with.

... not that it isn't a good idea to chop off unnecessary cruft, but this is getting slightly ridiculous.

https://communities.intel.com/docs/DOC-21835

Slow RAM is cheap, fast RAM is expensive. The less RAM you work with, the faster, smaller, and more power efficient it can be. It also allows you to do things you couldn't otherwise. No reason to bash trying to improve something.

[Spittie]

Unless the maintainer changes his mind, it doesn't seems that this will get merged.

[okias]

Unless the maintainer changes his mind, it doesn't seems that this will get merged.

There seems to be lot developers interested in minimal networking stack, so maybe after some changes it pass

[FLHerne]

I thought most embedded systems used lwip, is that too big for Intel!?

It seems they'd rather not use it:

There were proposals to instead use LWIP in user space. LWIP with
its socket interface comes in at a bit over 100k overhead per application.

I maintain that the Linux network stack is actually not that bloated,
it just has a lot of features :-) The goal of this project was to
subset it in a sensible way so that the native kernel stack becomes
competitive with LWIP.

[Spittie]

There seems to be lot developers interested in minimal networking stack, so maybe after some changes it pass

I wouldn't bet on it. Some quotes from the mailing list

I'm not applying this.

> What parts would you remove to get the foot print down for a 2MB
> single purpose machine?

I wouldn't use Linux, end of story.

Maybe two decades ago, but not now, those days are over.

Making 2MB RAM machines today makes no sense at all.

[Luke_Wolf]

Huh.

Their reference design "galileo" specs 256 MB DRAM.
Given the cost of RAM these days, anything less than that and the vendor is just being SO absurdly cheap that they aren't worth dealing with.

... not that it isn't a good idea to chop off unnecessary cruft, but this is getting slightly ridiculous.

https://communities.intel.com/docs/DOC-21835

this ^

If you're making 2MB/4MB machines in this day and age, you're doing something wrong.

[droidhacker]

this ^

If you're making 2MB/4MB machines in this day and age, you're doing something wrong.

Absolutely.
Even CHEAP stuff now comes with AT LEAST 32 MB.
For example;
http://wiki.openwrt.org/toh/tp-link/tl-wr841nd

Those things cost under $20 on sale, and its not just an embedded system, its an entire router. They even bumped the newest version to 64 MB.