My RFID stack now reads the ATS via T=CL

Lots of T=CL features such as chaining are still missing, but the code evolves constantly, as is the API (which now starts to become easy and nice). I'm constantly committing to svn.gnumonks.org, for those of you who can't wait.

I'm now also in (temporary) possession of two other readers, as well as a 14443 B-type passport sample (in addition to my 14443 A) sample.

Meanwhile I've also confirmed that the Omnikey 5121 windows reader driver has the same (or a similar) bug as the Linux driver, too. It also refuses to work with any MTCOS based card. I hope the MTCOS sample card I sent them will help them debugging - even though I don't need their proprietary drivers anymore at this point.

Network performance woes continue: MMIO read latency

Some low-level networking guys (Lennert Buytenhek, Robert Olsson, ..) have figured yet another reason why network performance with high pps (packets per second) rates sucks so much on commodity hardware (all PCI / PCI-x / PCI express based systems).

The 'new' culprit is MMIO read latency. When you're inside a network driver interrupt handler (well, same is true for about any such handler), the first thing you usually do is read the devices' "Interrupt Status Register(s)" to find out whether the device really originated that interrupt, and which condition (TX completion, RX completion, ...) caused it.

Depending on the NIC and driver design, you do multiple reads (and writes, but writes are not that bad) within the IRQ handler.

Lennert has hacked up a tool called mmio_test to benchmark the number of CPU cycles spent. Robert improved it a bit, and I've now added support for multiple network adapters, scheduling on multiple CPUs and other bits.

In case you're interested, it is (as usual) available from my svn server. In case you want to send me some numbers, please always include /proc/cpuinfo and "lspci -v -t" output, otherwise the numbers are useless.

Impressions from ph-neutral

I've been invited by multiple people to visit ph-neutral, a small but nice meeting of hackers organized by phenoelit. Since I've already been invited (and registered) last year but somehow missed it, I had to be there this year.

The strength of the event seems to be in the "meeting, having fun" part, since at least those two talks/presentations that I've been to were a huge disappointment. I don't want to be more specific and hurt anyone's feelings... but in both cases most of the audience knew more than the self-designated "expert".