Browsed by
Author: kepstin

Windows 8 Benchmarking Timing?

Windows 8 Benchmarking Timing?

It’s kind of curious that there’s a bit of a ruckus going on with regards to benchmarking accuracy on Windows 8. From what I’ve heard, the actual cause of the inaccuracy is quite simple:

Windows 8 appears to use the processor’s TSC as a timing source on modern processors. There is absolutely nothing wrong with this: it provides a very fast high resolution timing source as long as the TSC on the processor runs constantly and at the same rate. Modern Intel processors set CPUID bits that indicate this to be the case (nonstop_tsc and constant_tsc). Linux has supported using the TSC for timing for several releases, with no issues.

The only catch is that to turn the TSC into a clock source, it has to be calibrated against an external clock. Linux does this once at boot, calculating the value based on the processor’s reported frequency and double-checking against a realtime clock. Presumably Windows 8 does something similar.

The problem with the benchmarks comes in only on overclocked (or underclocked) systems where the clock speed of the processor is changed while the system is running, using for example a utility provided by the motherboard manufacturer. These utilities usually change the base clock rate that the processor is using as a reference. The key point: This changes the speed at which the TSC runs. And the utilities don’t bother telling the operating system to recalibrate the clock; presumably there is no standard API to do so.

So, what can be done? I don’t know. But if motherboard manufacturers want to push in-OS overclocking solutions, they’re presumably going to have to pressure Microsoft into offering an API to tell the Windows timesource to recalibrate the TSC clock, or even just tell Windows to switch to an older, less performant timing mechanism.

Migrated to WordPress

Migrated to WordPress

So, my site is now looking a little different from how it used to. I’ve actually changed how I’m hosting the service, completely. My original site was on Google Sites via my Google Apps account, but I’ve never really been satisfied with Google Sites and the limited control that was offered. So I took the plunge and got myself a “Droplet” on DigitalOcean, and installed WordPress.

The site may look different, but barring a couple of articles that I haven’t migrated yet (in particular, a couple of my old Anime Tierlist posts are still missing for now), all of the links from my previous site should still be active – either at the same location, or with a redirect. Please let me know if that’s not the case!

For the time being, my old Google Sites page will remain operational, but it will go away once I’m sure all content has been moved here.

Network UPS Tools (nut) and systemd

Network UPS Tools (nut) and systemd

Thanks to the Fedora developers, nut does actually ship some unit files for systemd. Unfortunately, those unit files provide a somewhat minimal functionality, with fairly fixed startup order and without taking advantage of systemd’s unit activation features. This caused some problems when I was building my new server a while back. This new machine is quite wicked fast with a quad-core IVB and an SSD—so fast, in fact, that systemd was starting nut’s UPS driver before the UPS itself was probed on the USB bus. As a result, startup failed quite badly, and it didn’t automatically recover.

So, obviously, we have to delay the startup of the nut driver until the USB device has been probed. But what about the other services? As it turns out, nut-server(upsd) and nut-monitor(upsmon) are actually perfectly happy to start up in any order with no more than a few warning messages in the logs as they wait for the other services to become available. So lets simplify the unit files for those. First, /etc/systemd/system/nut-server.service:

[Unit]
Description=Network UPS Tools - power devices information server

[Service]
ExecStart=/usr/sbin/upsd
Type=forking

[Install]
WantedBy=multi-user.target

If you have configured upsd to listen to a particular network interface or on a particular network address (in upsd.conf) it almost certainly needs the interface to be configured first; you should add in After=network.target.

Now, we can tweak the unit file for the monitoring process: /etc/systemd/system/nut-monitor.service:

[Unit]
Description=Network UPS Tools - power device monitor and shutdown controller
After=nut-server.service

[Service]
ExecStart=/usr/sbin/upsmon
PIDFile=/run/nut/upsmon.pid
Type=forking

[Install]
WantedBy=multi-user.target

I’ve left in the After=nut-server.service for only one reason: To reduce the warning messages in the log output when starting both upsd and upsmon on the same system. It’s not strictly required.

One thing to double-check: Ensure that the nut packages on your system have installed the file /usr/lib/systemd/system-shutdown/nutshutdown. This file doesn’t need any modifications, but it’s required to handle emergency poweroffs and test reboots correctly. It should look something like this, and it must be executable:

#!/bin/sh
/usr/sbin/upsmon -K >/dev/null 2>&1 && /usr/sbin/upsdrvctl shutdown

Now for the fun bit; handling starting up the nut driver for the USB UPS automatically once it’s been probed by the kernel. Ideally, we want to make this independent of things like which USB port it happens to be plugged into—this means creating a device file symlink with an appropriate stable name. Lets see what we have available…

# udevadm info --attribute-walk --name /dev/bus/usb/003/003

  looking at device '/devices/pci0000:00/0000:00:1a.0/usb3/3-1/3-1.1':
    KERNEL=="3-1.1"
    SUBSYSTEM=="usb"
    DRIVER=="usb"
[...]
    ATTR{idVendor}=="051d"
    ATTR{quirks}=="0x0"
    ATTR{serial}=="4B1111P37995  "
    ATTR{version}==" 1.10"
    ATTR{urbnum}=="13527"
    ATTR{ltm_capable}=="no"
    ATTR{manufacturer}=="APC"
    ATTR{removable}=="removable"
    ATTR{idProduct}=="0002"
    ATTR{bDeviceClass}=="00"
    ATTR{product}=="Back-UPS ES 550 FW:843.K2 .D USB FW:K2 "
[...]

I’ve stripped out a fair bit of the output, but as you can see we have some things to work with. The idVendor and idProduct will identify the device as a UPS, and the serial number will give a nice unique id for this UPS. Lets write a udev rule! I stuck it into the file /etc/udev/rules.d/53-nut-usbups-systemd.rules

ACTION!="add|change", GOTO="nut-usbups-systemd_rules_end"
SUBSYSTEM=="usb_device", GOTO="nut-usbups-systemd_rules_real"
SUBSYSTEM=="usb", GOTO="nut-usbups-systemd_rules_real"
SUBSYSTEM!="usb", GOTO="nut-usbups-systemd_rules_end"

LABEL="nut-usbups-systemd_rules_real"

ATTR{idVendor}=="051d", ATTR{idProduct}=="0002", TAG+="systemd", SYMLINK+="ups/$attr{serial}"

LABEL="nut-usbups-systemd_rules_end"

With this, my UPS is now available with device file /dev/ups/4B1111P37995. And since I’ve tagged it with “systemd”, we will be able hook it up to our systemd unit. In order to keep things matching up, I defined the driver in my ups.conf as follows:

[apc]
        driver = usbhid-ups
        port = auto
        serial = "4B1111P37995"

Ok, lets go and write the systemd unit file to start up the driver for my “apc” UPS device. I’m going to make use of another cool systemd feature here, template unit files and instances. Create a unit file named nut-driver@.service, with the following content:

[Unit]
Description=Network UPS Tools - power device driver controller

[Service]
ExecStart=/usr/sbin/upsdrvctl start %i
ExecStop=/usr/sbin/upsdrvctl stop %i
Type=forking

So far, so good! Now here’s the clever bit. Right now, systemd is exposing a unit for the device, but it’s using the name sys-devices-pci0000:00-0000:00:1a.0-usb3-3\x2d1-3\x2d1.1.device which is not only very ugly, but it’s also dependent on the usb path and port probe order. Lets create a unit with our new device name, in the file /etc/systemd/system/dev-ups-4B1111P37995.device:

[Unit]
Description=APC Back-UPS ES 550_FW

You don’t actually need any more in the file; systemd will handle hooking it up to the appropriate real device automatically. But now we can hook up our unit file to be automatically started when the ups becomes available:

# mkdir /etc/systemd/system/dev-ups-4B1111P37995.device.wants
# ln -s ../nut-driver@.service /etc/systemd/system/dev-ups-4B1111P37995.device.wants/nut-driver@apc.service

Note that in the symlink that was created, I’ve added the string “apc” after the @ sign. When the unit is started, this will be substituted for the %i in the nut-driver@ template unit file. This will cause the upsdrvctl command to look for the driver named “apc” in ups.conf, thus completing the chain and starting the the appropriate driver.

And now my UPS drivers reliably start, even when my computer can boot faster than USB can probe.

The unfortunate part is that a lot of this work is fairly system-specific. I’m not sure that it is all suitable for use in the upstream nut package, due to the customization required. I’d appreciate any comments or thoughts.

Release version 1.0.0! (And 1.0.1, you know, because)

Release version 1.0.0! (And 1.0.1, you know, because)

There is now an official release version (or two) of the image-id tool!

This release is the final result of both porting image-id to use the latest 2.0 API version of libmirage, and adding support for reading MCN and ISRC numbers from disc images. There is some work in progress that might allow these to be automatically submitted to musicbrainz using a future third-party tool.

The release notes and download files are available on the Github release page.

Note that 1.0.1 is a sort of brown-paper-bag release; I’d forgotten to pull before tagging. There is no reason for you to use 1.0.0; please go with 1.0.1 instead.

Summer 2013

Summer 2013

I have now rejoined the ranks of the anime-watching, I guess! Here’s the list of things that I care about in the upcoming season. Expect this page to change over the course of the season.

Just to note, this list is way absurdly too long. I wonder what I’m going to end up dropping.

Continued from last season

Wait, what, I actually watched things last season? Well, no, not really. But I went and marathoned a few shows to catch up.

New this season

  • Stella Women’s Academy, High School Division Class C3 – it has Makino Yui, And if I’ve learned anything from past shows like C -Control- or Sora o Kakeru Shoujo or Tsubasa Chronicles, that doesn’t mean very much. But it has at least one character with a voice actress I’m a fan of :)
  • NEW GENSHIKEN
  • Some more The World God Only Knows, too.
  • Fate/kaleid liner PRISMA ILLYA. There is no way this can not be awesome. It’s as if they’d made a third season of Nanoha. Possibly almost as good as a Tsukihime anime, should they have made one.

Things where I used to be more than a season behind and had to catch up first

  • The second season of the “Monogatari” series. I’m kind of surprised how far behind I am on this; I have to watch Nise- and Neko- (Black) to catch up before starting on this. Which is probably a good thing, because knowing SHAFT it’s not going to come out on schedule anyways. (I’m now caught up on the previous shows, but Crunchyroll has this on a delay…)
  • There’s new Rozen Maiden, but I haven’t seen any of the previous work. So, yeah. (I’ve now watched the entire first season, and have started this. It’s awesome!)

Things that aren’t actually from this season at all.

  • Captain Harlock
  • My Little Monster (Tonari no Kaibutsu-kun). Picked it at random, but it’s surprisingly sweet.

Things that I thought I wouldn’t watch, but upon further reflection changed my mind about

Things I’m probably definitely not watching

  • Fantasista Doll – it has surprisingly well-animated plaid skirts. But that’s about it.
No More Google Reader: Part 2, The Old Reader

No More Google Reader: Part 2, The Old Reader

This project has actually been around for quite a while now; the name comes from it being a clone of “the old Google Reader”, prior to the removal of the social features. As I mentioned in the introduction, I never actually used those features—so we’ll just see how it goes as a RSS reader application.

Signup is pretty easy—you can just log in through your Google or Facebook accounts. Transferring your feed selection from Google Reader to The Old Reader isn’t as seamless as Feedly, however: You have to export your feeds via Google Takeout and then upload them to The Old Reader.

I’m not off to a great start, though: “There are 35171 users in the import queue ahead of you.” The service has been very heavily hit with the sudden influx of users migrating away from Google Reader. They do load up a Trending feed that lets me go through the user interface immediately:

The Old Reader

And, well, it does pretty much exactly what I want. Full article views and show only unread are supported; it even defaults to my preferred ‘oldest first’ on the All Items view.

There are minor quirks, of course. One of the happiest days of my Google Reader usage was when I finally figured out how to hide the sidebar (see the screenshot in the introduction). And The Old Reader brings the sidebar back, in full unremovable glory. I don’t really mind the space lost; the article page is big enough. The real problem is simply that if the mouse cursor is over the sidebar, the scroll-wheel on my mouse won’t scroll the page!

If I could hide the sidebar, The Old Reader would be perfect. As it is, it’s very good. One question is left: will they recover from the influx of Google Reader users and get the feed refreshes back up to speed? I hope they do.

No More Google Reader: Part 1, Feedly

No More Google Reader: Part 1, Feedly

Feedly is the first in the list of cloud RSS reader software that I’m taking a look at as a replacement for Google Reader (see the Introduction for the index of all the parts.)

Somewhat ironically, Feedly might be at least indirectly responsible for Google Reader’s shutdown. It’s been around for quite a while, but Feedly doesn’t actually run its own RSS aggregator. Instead, it’s simply a front end that accesses your feeds in Google Reader via an undocumented and apparently private API. They have recently announced that they are writing their own RSS aggregator backend, and will have it deployed in time to seamlessly switch over when Google turns off Reader.

A curious result of their use of the private Google Reader API is that Feedly isn’t actually an application hosted on a web page. Instead, it requires installation into your browser as an extension, available for Firefox and Chrome. This is an issue to me, as I regularly switch browsers—and interesting hardware like my ebook reader simply can’t use it at all.

At first glance, it certainly is prettier than Google Reader ever was:

Feedly

And it is quite full featured, and can be configured to replicate my preferred view from Google Reader: only show unread, with oldest messages first. There’s minor bugs in various places—e.g. if I have no unread articles, it will instead show a seemingly random set of previously read articles—but they seem responsive to user feedback. They haven’t switched away from the Google Reader backend yet, so how well they will scale is still an unknown.

The layout at the moment is somewhat annoying; the narrow columns cause some images—comics in particular—to be rescaled at smaller than original size, making text difficult to read. I suspect that advertising will grow at some point, although they may offer a paid subscription service in the future. One curious thing is that they hope to open up their backend (which has a Google Reader compatible API) to allow other services to use it!

Feedly is nice, but the requirement for browser extensions and special mobile applications is somewhat limiting. I’m still searching for the perfect Google Reader replacement…

Well, So Much for That

I haven’t really watched any anime series as they came out in ages, so I guess you could say this page is dead. Maybe some day when I have more time, the Anime Tierlist will rise again.

No More Google Reader: Introduction

No More Google Reader: Introduction

Like many other people online, I’m a heavy user of the Google Reader application now left to find some replacement for this aggregator of interesting news that I’ve subscribed to. I know that Reader had some social features at some point, but I’ve never actually used them. There was a single thing that I valued about Reader: it allowed me to leave off checking my feeds on one computer, and resume later on a different box. It’s a “cloud-based” RSS reader, to use the modern parlance.

Since Google has announced that they’re cutting the service off, I’m now looking for an alternative. But first, let’s take a look at what I have right now:

Google Reader

There’s my feed. I use Google Reader in its most simple mode; a continuous list of entries ordered from oldest to newest with previously-read entries hidden and with article content inline. As I scroll past each message, it’s automatically marked as read; and I use the “Keep unread” option to mark things that I pass by now, but want to take a look at when I have more time.

While much of this could be gotten via social media, the experience is completely different. Social Media is constantly running in a ”push” stream. It always shows you the newest thing first, popping that down and out of view as soon as something else comes in. Social media is limited to tiny sound bite clips. With the RSS experience I use, it’s the other way around; things wait for me to come view them, and I see everything come in the order that it happened. (Particularly important if I, say, don’t check one day and miss a strip from a daily comic!) The longer form allows you to get at minimum a detailed introduction into what is linked; and many sites will inline entire articles and full-sized images so you can browse without even opening another tab.

I’m sure lots of other services are popping up now with alternatives to the soon-to-be-gone Google Reader. I’m trying a few of them, and I’ll let you know how they go, in these follow-up articles:

A MusicBrainz report on Pseudo-Releases

A MusicBrainz report on Pseudo-Releases

As an implementation of support for translating or transliterating the tracklists on foreign releases, MusicBrainz has something called “Pseudo-Releases.” They are meant to be used alongside the database entry for the original, real release to contain alternate tracklists.

Unfortunately, many MusicBrainz users weren’t quite sure on when the “Pseudo-Release” status should be used, and incorrectly set it instead of the correct status like “Official”, “Bootleg”, or “Other” on some releases. And a few translated tracklists have the status set correctly—but aren’t linked back to the original version of the release.

In order to help find and fix these issues in the database, I’ve created a new report: “pseudo-releases without transl*tion relationships”, in code review now. It looks pretty sweet:

But 2261 releases to fix up means there’s a ways to go.