Forgot your password?

Comment: Re:Seems incorrect (Score 1) 80

by Ben Hutchings (#48119543) Attached to: NVIDIA Presents Plans To Support Mir and Wayland On Linux

You need at least:

1. Kernel driver for hardware init, power management, mode setting, GPU buffer management and command submission

2. Userland library for GPU buffer management and command submission

3. OpenGL implementation

In the open source graphics stack, the kernel driver exposes KMS and DRM interfaces and potentially others. Parts 2 and 3 are part of libdrm and Mesa respectively. The display server can (I think) be built on top of KMS, libdrm and OpenGL and be independent of the hardware. However it will need an extension to OpenGL called EGL which will be specific to each display server protocol.

Currently X doesn't usually work that way for historical reasons - it used 2D acceleration first and still supports hardware that has only 2D acceleration. So it has hardware-specific drivers for each family of GPUs. However there is the 'Glamor' library that supports 2D acceleration genericallly on top of OpenGL, and I would expect to see a gradual move to that, not least because it's the only option for 2D acceleration in XWayland.

Getting back to Nvidia, their problem currently is that they don't implement the same interfaces as the open source stack and therefore don't work with the new display servers that depend on those interfaces. Implementing KMS gets them a long way there. However it sounds like they still need to reimplement the EGL, not because it's hardware-specific but because their OpenGL implementation is entirely independent of Mesa.

Comment: Re:One bad apple spoils the barrel (Score 3, Interesting) 1134

by Ben Hutchings (#47829027) Attached to: Combating Recent, Ugly Incidents of Misogyny In Gamer Culture

1. The allegations against Quinn are insinuations with no evidence behind them.
2. Sarkeesian has been loudly contradicted and claimed to be a con-woman by people that can't take criticism and are annoyed by the success of her Kickstarter.
3. This is being called "misogyny" in gaming because it is directed specifically at women.
4. The Social Justice Warriors have all supported these women because they oppose misogyny.
5. It's so cheap and easy to brand gamers basement dwelling vrigin men-children than it is to look at the facts. This is stereotyping, but it is nothing like the harrassment, online bullying, doxxing or death threats made by some gamers against feminist critics.

Fixed that for you.

Comment: Re:This does pose the question: (Score 1) 195

by Ben Hutchings (#47664551) Attached to: Facebook Seeks Devs To Make Linux Network Stack As Good As FreeBSD's

ASICs generally aren't flexible enough that you could simply emulate another controller in firmware, while FPGAs suck too much power to use on commodity network adapters. Writing a new driver (or bringing an existing neglected driver up to scratch) is going to be quicker than trying to make hardware that's compatible enough to work with a driver written for another vendor's controller.

(Besides which, as that other driver is probably maintained by your competitor, do you really think they're going to make an effort to ensure that their later updates are compatible with your clone controller? You'll still have to maintain your own fork.)

I have often wondered why there isn't a vendor-neutral register-level standard for Ethernet controllers, along the lines of AHCI and xHCI. There is the virtio networking standard, but as it's designed for VMs I assume it does not cover Ethernet link management. I seem to remember that VMware tried to promote a common interface for SR-IOV virtual functions at one time, but that didn't get very far. Again that would not have included link management.

Comment: Meanwhile, in reality... (Score 2) 201

by Ben Hutchings (#45692771) Attached to: Under the Hood of SteamOS
It's a stock Debian kernel with some minor packaging changes and support for a new game controller. All those realtime patches? Not actually used by default. The full list of exciting changes:
  • Make the binnmu regexp also reconize our build suffixes
  • New XBox controller driver
  • Disable Intel P-State driver as it causes issues with sound being choppy during BigPicture trailer video playback.
  • Hard-code parallel build for now since our OBS infrastructure doesn't know how to set these options yet.
  • Add postinst step to touch /var/run/reboot-required

Comment: Re:But can SVN merge a branch yet? (Score 1) 378

by Ben Hutchings (#44056933) Attached to: Subversion 1.8 Released But Will You Still Use Git?
Repeated merges have worked well for a while now (maybe since 1.6?). It's not quite as good at merging as git is, but it works well enough. But I have to agree with the general sentiment against merging from release to devel branches. Merging should be considered an expert-only operation (not expert in version control, but in the code base). Cherry-picking/backporting fixes from devel to release is safer because then you know exactly what you're changing.

Comment: Re:Online Advertising Response (Score 1) 369

by Ben Hutchings (#42995725) Attached to: Firefox Will Soon Block Third-Party Cookies
Whenever a web site has a form, some other site can set up another (hidden) form pointing to the same URL and with any values they like. Someone who visits both sites can unintentionally submit that form (together with their cookies from the first site, so it's properly authenticated). This is 'Cross Site Request Forgery' and the usual way to avoid is to check the Referer header.

Comment: Who leaked my card details? (Score 1) 163

I personally haven't experienced abuse of my card details - so far as I know. But if I did, how could I tell who was responsible - especially when there are vast leaks like this? It seems like it would be more fair to have an industry-wide fund to compensate victims, which the leaking companies would pay into proportionately to the number of valid details leaked.

Comment: Re:Damn you kids, get off my lawn. (Score 1) 207

by Ben Hutchings (#39630205) Attached to: Demoscene: 64k Intros At Revision Demoparty

There's not 64k of assembly pumping bytes into a framebuffer and twiddling the PC speaker port to synthesize digital audio.

Of course. But all the creative work is squeezed into 64K.

One thing I couldn't find in there (and I've been out of the scene for a LONG time, so I don't know how this works on new-fangled fancy computers...) -- do these write directly to the video hardware? Or do they use OS services like DirectX11, etc?

They use DirectX, because that is the only way to support a reasonable range of hardware. (Also, you can't hit the hardware without installing a new driver or exploiting a kernel bug. Neither of which is very friendly.)

But are people still getting down and counting clock cycles?

Cycle counts aren't even documented today. Now it's all about avoiding cache misses and cache invalidation.

Comment: Re:Where's the 10GbE? (Score 1) 96

by Ben Hutchings (#39271331) Attached to: Intel Releases Sandy Bridge-based Xeon E5 Series
I think you're mistaken about PCIe link speeds, and you're also failing to account for the transaction layer protocol (TLP) overhead which can be quite substantial. PCIe 2.0 link speed is 5 GT/s but that's with 8b10 encoding; the data rate is only 4 Gb/s. PCIe 3.0 doubles the data rate to 8 Gb/s. Two 10G ports require 20 Gb/s plus some overhead, hence 8 lanes on PCIe 2.0 or potentially 4 lanes on PCIe 3.0.

"It is easier to fight for principles than to live up to them." -- Alfred Adler