Older blog entries for mjg59 (starting at number 224)

Intel IGD opregion and GMA500

A while back, Intel defined a specification for binding ACPI-defined methods for controlling hardware to the OS-specific driver, ensuring that the two don't get out of synchronisation. I added support for this to the in-kernel i915 driver last year, and after a couple of awkwardnesses it works well now. One consequence of this that showed up slightly later is that it's necessary to do some of the setup from the i915 driver rather than the ACPI driver, which meant that we had to defer the ACPI driver from binding until the drm driver had done that setup.

The problem with GMA500 is that it also implements the IGD Opregion spec, and the ACPI video driver detects this and refuses to bind. But the GMA500 kernel driver doesn't implement support for the spec and so doesn't call the function that triggers the ACPI video registration. Working around this is simple - just add acpi_video_register() to the init function of the GMA500 drm. But note that this means that you're failing to implement the spec properly, and there's potential for stuff to be broken. A full implementation of the spec for GMA500 wouldn't be especially difficult, but there's no docs and I have no hardware so I'm not going to do it myself.

The reason I bring this up is that various people have been approaching this problem in a different way. It's easy to assume that the check in the acpi driver was naively assuming that all Intel hardware was driven by i915 and that this patch was broken. It's actually entirely correct and the (out of tree) GMA500 driver was broken. If Intel had made the effort to get their code properly upstream, it'd have been fixed there when the original change was made and nobody would ever have had a problem. Just say no to out of tree drivers.

Syndicated 2009-09-24 18:24:23 from Matthew Garrett

Portland

I'm off to Boston in under 16 hours, and I'll be getting into Portland around lunchtime on Monday. I'll be talking at Linuxcon about how we're broadening power management on Linux to be applicable from phones through netbooks up to supercomputers - that's 10:15 on Tuesday. At 10AM on Friday I'll be presenting at the Linux Plumbers conference on how userspace can express its requirements to the kernel more clearly, thereby allowing the kernel to be smarter about powering down hardware. And after a short hop down to SF for the weekend, I'll be back in Portland at the X developers conference talking about the role of X in providing relevant information to the kernel and using that to facilitate more aggressive power management.

Three talks in under 10 days. I'll even do my best to ensure that there's new jokes for each of them.

Syndicated 2009-09-17 00:04:36 from Matthew Garrett

Bye

I'm moving to the US on Thursday, so I will be here on Wednesday evening from about 7. If your presence is unlikely to make me stupifyingly angry, feel free to join me.

Syndicated 2009-09-15 11:16:36 from Matthew Garrett

Good thing: Vicodin
Bad thing: The broken arm that necessitates the vicodin

Syndicated 2009-09-07 14:13:20 from Matthew Garrett

I moved out today, and in a bit over a month should be firmly relocated in Boston. Which means a bit over a month of living and working out of bags, but I'm in a park, it's sunny and I'm relaxed enough that even working through Bugzilla isn't making me sad.

Syndicated 2009-08-25 11:31:12 from Matthew Garrett

Defective by Design

You know what's Defective by Design? Thinking that this kind of functionality is a good thing, resulting in this.

Syndicated 2009-08-11 20:18:04 from Matthew Garrett

Though people might be forgiven for having gained the opposite impression, I don't yearn for a future where open source development involves spending 25% of the time coding and 75% of the time where we have earnest conversations about the precise social nature of our utopian society and hold votes on whether to send polite but firm telegrams to people who accidently swore at some point in the recent past. I've never really yearned to live in a commune. I hate the idea of singing around campfires and there are many people who I never, ever want to be polite to.

No. What I want is a future where I can say this person is being a dick, or this person is being a dick or that this person is being a dick and not have to engage in lengthy explanations as to why before people agree. A future where calling someone on their behaviour makes people examine that person's motives rather than yours. A future where you're allowed to criticise without being perfect. A future where people don't feel scared to speak up because the consequences of doing so may include people posting their address and phone number and trying to get them fired. A future where disagreements about behaviour aren't so often intrinsically linked to gender. A future where I get to stop writing about any of this because it's taken for granted anyway.

And I want some awesomely good arguments for why this isn't an excellent future that we should all be encouraging.

Syndicated 2009-07-27 18:32:04 from Matthew Garrett

Intel graphics in rawhide

I've just committed some new Intel driver code to rawhide, which should be in the next kernel build. There's a chance that some people with Intel laptops will see some screen flickering. If you do, could you please file a bug against the kernel in the Red Hat bugzilla and make sure you Cc: me (mjg@redhat.com). Include the output of the xrandr command.

Syndicated 2009-07-27 03:09:55 from Matthew Garrett

LCA 2010

Depending on timezones, you currently have somewhere between 8[1] and 33[2] hours to submit a presentation for this year's Linuxconf Australia[3]. If you think you have something interesting to talk about, then do it and enjoy an unarguably top-tier Linux conference. Also, New Zealand is excellent.

[1] Kiribati is weird.
[2] For all those of you stuck on Howland or Baker islands. If you start swimming now you might reach civilisation before the conference starts.
[3] In New Zealand, obviously

Syndicated 2009-07-24 02:47:16 from Matthew Garrett

RMS and virgins

Many of the comments here and here are disheartening, but part of the problem is that many people didn't see the presentation. Dave linked to a previous iteration of the same presentation[1]. Here's a transcription (errors are mine and mine alone):

I am Saint Ignucius of the church of emacs. I bless your computer, my child. Emacs started out as a text editor. An extensible text editor, which became a way of life for many users because it was extended so much they could do all their computing work without ever exiting from emacs. And then, it became a religion, with the launch of the newsgroup alt.religion.emacs. Tday in the church of emacs we have a great schism between several rival versions of emacs, and we also have saints. But fortunately no god, instead of gods we worship an editor. To be a member of the chuch of emacs, you must recite the confession of the faith. You must say there is no system but gnu, and linux is one of its kernels.

Then if you become a hacker you can celebrate that by having a foobar mitzvah, a ceremony in which the new hacker stands in front of the assembled congregation of hackers and chants through the lines of the system source code. And we also have the cult of the virgin of emacs. The virgin of emacs is any female who has not yet learned how to use emacs. And in the church of emacs we believe that taking her emacs virginity away is a blessed act.

The church of emacs has certain advantages compared with with other churches I won't mention. For instance, to be a saint in the church of emacs does not require celibacy. Although for some of us hackers we wouldn't notice the difference. But it does require living a life of moral purity. You must exorcise whatever proprietary evil operating systems have posessed the computers under your control or set up for your use and then install a wholy free operating system and then only install free software on the system. If you make that vow and you live by it then you too will be a saint, and you too will have the right to wear a halo. If you can find one, because they don't make them any more.

Sometimes people ask me whether it a sin in the church of emacs to use the other editor, vi. It's true that vi vi vi is the editor of the beast. But using a free implementation of vi is not a sin, it's a penance. And sometimes people will ask whether my halo is really an old computer disk. This is no computer disk, this is my halo! But it was a computer disk in a previous life. So thank you very much.


One of the frequent counterarguments against this being sexist is that RMS has often spoken out against sexism (see here, for example). It's very easy to claim to be free of sexism. It's much harder to perform the degree of introspection required to understand whether any of your actions are motivated by viewing genders differently. Do I believe that Richard is attempting to deliberately denigrate women? Not in the slightest. But I also don't believe that someone entirely gender-blind would have made the above joke.

My point here isn't to claim that he's a bad person as a result. I've got personality flaws large enough that you could probably drive a bus through them, but I'd be slightly upset if people thought I was evil because of them. My point is that nobody is above criticism, and if someone behaves in a way that offends a large subset of the community then they should to be criticised. Failing to do so sends the signal that we don't care about those who were offended, and at the same time provides no incentive for people to change their behaviour as a result. And yes, I think those who have high profile positions in the community should be held to higher standards than others - Richard's comments on Mono carry more weight because of who he is, but the cost of this is that everything else he says does as well. And if one of our nominal leaders is perceived as sexist then that reflects badly on all of us.

[1] Note that his GCDS presentation did not entirely consist of this routine - there was also significant discussion of Mono and why he believes that adopting it is dangerous. I don't entirely disagree with him, but that's really not the point here. I'm not involved in any Mono development. I work for a company that ships Mono in the community distribution it supports, but not in the enterprise distribution that pays my wages. If anyone brings up Mono in any comments here they'll be blocked and the thread deleted, because it is not relevant to this discussion.

Syndicated 2009-07-16 00:44:30 from Matthew Garrett

215 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!