| --- Log opened Sat Mar 29 00:00:41 2014 | ||
| stekern | _franck__: heh, I bet that can be confusing if you don't know ;) | 03:50 |
|---|---|---|
| stekern | was it that that killed your mem-controller, or was it just a red herring? | 03:50 |
| stekern | nobody can accuse me for not writing verbose commit messages at least, the latest orpsoc-cores pull-request has a 10/1 lines commit message/changed code | 04:24 |
| stekern | olofk will be happy, I have qsys support in the quartus backend now, so I'll be able to upstream the sockit port soon too | 07:22 |
| _franck__ | stekern: I found the problem in my memory controller now. This self modifying code was just a false alert | 11:11 |
| _franck__ | stekern: do you know how to change hsync_len, right_margin, etc.. from the bootargs string ? | 11:12 |
| stekern | bootargs string? | 11:52 |
| stekern | ah, you mean from the kernel args? | 11:53 |
| _franck__ | bootargs = "console=tty0 console=uart,mmio,0x90000000,115200 video=ocfb:800x480-16@60"; | 11:53 |
| stekern | you can't | 11:53 |
| stekern | or, I mean, you can't set them directly there, only use the string like you show there | 11:54 |
| stekern | the alternative would be to support the device-tree bindings for it, but that's not implemented (at least not yet) | 11:55 |
| stekern | so, yes, if you have some "non-vesa" resolution, it's not going to pick it up right | 11:56 |
| _franck__ | ok, I'll hard code my values to make it work for now | 11:56 |
| stekern | yeah, that's what I did on the de0_nano as well ;) | 11:57 |
| _franck__ | ah, endianess problem here :) | 12:02 |
| _franck__ | I should use your upstreamed driver | 12:02 |
| _franck__ | ok it works now, however I have a lot of flicker at 800x480 16bpp | 12:06 |
| stekern | yeah, there are a lot of other bugfixes in the upstream one | 12:15 |
| stekern | is that together with mor1kx? | 12:17 |
| stekern | because, I get a lot of flicker on de0_nano after I put the storebuffer in there | 12:17 |
| stekern | mor1kx is too good at stealing bw... | 12:18 |
| _franck__ | yes I'm using the mor1kx | 12:18 |
| stekern | iirc, I tried to hack in a bit fairness in wb_sdram_ctrl, but it never improved to the point where there would be no flicker | 12:19 |
| _franck__ | I don't know if the cpu don't write the fb fast enough or the vga core reading it fast enough | 12:20 |
| _franck__ | is it possible to activate double buffering in linux fb ? | 12:20 |
| stekern | oh, it's the vga core that don't get enough bw | 12:20 |
| stekern | nah, that's another future improvement | 12:21 |
| _franck__ | like having two frame buffer and switch during vertical blanking | 12:21 |
| _franck__ | ok I'll work on this later. I now have to watch "plane" on TV with my son ;) | 12:22 |
| stekern | I doubt that double buffering will help for this particular problem though. | 12:26 |
| stekern | ah, Planes... it was like watching Cars, but with...Planes | 12:27 |
| stekern | http://pastie.org/8978613 | 16:31 |
| stekern | quartus_asm, go home, you're drunk... | 16:31 |
| _franck__ | :) | 17:09 |
| _franck__ | no, double buffering won't solve this problem as I have flicker on static images | 17:10 |
| blueCmd | stekern: if you're gonne fail, fail by doing rando m stuff | 23:19 |
| olofk | Yeah! fusesoc is getting it's first decorator :) | 23:28 |
| blueCmd | hmmmmmmmmmm | 23:30 |
| --- Log closed Sun Mar 30 00:00:42 2014 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!