Re: X0-4 (vivante) GPU driver development sponsoring
On Jun 26, 2013, at 9:21 AM, Christian Gmeiner wrote: > I have heard that your X0-4 is powered by an Vivante GC1000 GPU. Cause of this > fact I am looking for a partner to sponsor the development of > etnaviv[1]. Together with the current maintainer we have a roadmap that looks > like this: Unfortunately, I can't directly sponsor this project right now. I have brought it to the attention of some people who might, and can offer you hardware and help with testing the result. Regards, wad ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
X0-4 (vivante) GPU driver development sponsoring
Hi. I have heard that your X0-4 is powered by an Vivante GC1000 GPU. Cause of this fact I am looking for a partner to sponsor the development of etnaviv[1]. Together with the current maintainer we have a roadmap that looks like this: - Mesa integration (3 weeks) - GC2000 support - if not already done (2 weeks) - implement context handling (3 weeks) - Checking and fixing corner cases for the shader compiler (2 months) 2 months may be on the long side, Then again, testing with various software and drilling down to the problem is time-intensive, and chromium probably uses a challenging subset of OpenGL ES. So it may be a safe estimate. # at this point a mesa driver should be useable with Vivante's GPL kernel driver and # the imx-drm may could be used in combination of the xorg modesetting driver to get # a full blown desktop environment up and running. This will be without 2D acceleration at this point. For optimal performance, we may need very basic 2D support to blit the final rendered image to the screen in a window, but that's hard to say at this point. - start the mesa-mainline process (depending on how many iterations are needed ~ 1 Month) - start thinking about the kernel side of the driver (v2 vs v4 kernel interface) and begin with a kernel driver. (2 Months) - adapt the current user space driver if needed. (x weeks) - mainline the kernel side of the driver (depending on how many iterations are needed ~ 1 Month) # a full open source (kernel and userspace) is done, which is fully integrated into the # linux graphics stack. After 4 months the user space driver is ready. After 5 months the user space driver is mainlined. After 8 months the kernel driver is ready. After 9 months the kernel driver is mainlined. What do you think? [1] https://github.com/laanwj/etna_viv thanks -- Christian Gmeiner, MSc ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [OLPC New Zealand] Māori Macrons & olpc keyboard
On 26/06/13 10:52, Walter Bender wrote: The widget was broken until recently. Not sure my patches have landed in any releases yet. But we should be able to make a layout that works for both English and Maori without any end-user intervention. Will try to get to it this week. Attached is a patch to the output of xkbcomp on an XO-4 with mechanical keyboard. Sorry it's not in style of the files in /usr/share/X11/xkb, I still haven't been able to unravel how those files work. I think I can make an mi file that just has the changed lines, which layout should I use as a base? I'm guessing I might need to do something different for the membrane and mechanical keyboards? What is the best way to modify the laptops? Copying modified output of xkbcomp to /usr/share/X11/xkb/symbols/mi and running setxkbmap mi does the trick, but setting the content of /home/olpc/.Xkbmap to mi does not make this the default after reboot. --- orig.xkb 2013-06-26 09:28:21.770500621 + +++ test.xkb 2013-06-26 09:28:07.050500619 + @@ -1188,7 +1188,7 @@ key { type= "FOUR_LEVEL_ALPHABETIC", repeat= No, -symbols[Group1]= [ e, E, oe, OE ] +symbols[Group1]= [ e, E, U0113, U0112 ] }; key { type= "FOUR_LEVEL_SEMIALPHABETIC", @@ -1208,17 +1208,17 @@ key { type= "FOUR_LEVEL_SEMIALPHABETIC", repeat= No, -symbols[Group1]= [ u, U, U032D, U032D ] +symbols[Group1]= [ u, U, U016B, U016A ] }; key { type= "FOUR_LEVEL_SEMIALPHABETIC", repeat= No, -symbols[Group1]= [ i, I, U032C, U032C ] +symbols[Group1]= [ i, I, U012B, U012A ] }; key { type= "FOUR_LEVEL_SEMIALPHABETIC", repeat= No, -symbols[Group1]= [ o, O, U0323, U0323 ] +symbols[Group1]= [ o, O, U014D, U014C ] }; key { type= "FOUR_LEVEL_SEMIALPHABETIC", @@ -1246,7 +1246,7 @@ key { type= "FOUR_LEVEL_ALPHABETIC", repeat= No, -symbols[Group1]= [ a, A, ae, AE ] +symbols[Group1]= [ a, A, U0101, U0100 ] }; key { type= "FOUR_LEVEL_SEMIALPHABETIC", ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel
Re: [OLPC New Zealand] Māori Macrons & olpc keyboard
On 26/06/13 00:22, Barry Vercoe wrote: > Tom, it's not correct to say the laptops will remain in Maori. Sorry, I should have clarified that I was referring to the keyboard only. I think the difference between the current olpc keyboard and a customized Maori one will be minor. We'll replace the æ œ ̬ ̣ ̭ characters with the macron vowels. I don't think this will present any problems if we make it difficult to reverse until a Maori keyboard and a keyboard chooser are built into the operating system. I don't know how to do the windows style grave modifier on linux and for consistency with other Linux Maori keyboards, I think alt-gr - vowel will be a better starting place than the default setup. Switching the rest of the laptop to other languages is a different matter and I wouldn't propose to stop that! ___ Devel mailing list Devel@lists.laptop.org http://lists.laptop.org/listinfo/devel