Re: [XFree86] Cirrus Logic GD 7548 Woes
Richard Lynch writes: Unfortunately you may be pretty much on your own. the Cirrus driver is pretty much unsupported as noone who could support it has the HW. Would it be practical to: A) Provide SSH root access to somebody @XFree86.org, or B) Ship somebody the laptop as a loaner C) ??? Obviously A) is much easier, but I dunno what you can tell without seeing the screen, and I don't have a web-cam handy... Though I do need to set one up eventually anyway, so I reckon I could move that higher on the priority list. That's exactly the problem. And, of course, the odds on XFree86 folks having the time to mess with this are pretty small unless a lot of people have these things and so it's worth it to you to support them. It depends on the time one has to spend to figure out what's going on. Sending equipment around - especially laptops - is not a good solution either. It may get lost or broken - in which case you won't need the fix any more If you find a solution please let us know. The spikes and crackles usually appear when the memory bus is congested. Sometimes this can get fixed by reducing the pixel clock or increasing the memory clock. I'll try that. Pixel clock is the dot clock right? Yes. What's the memory clock? The clock whith which the memory access is timed. Unfortunately there is support for chaning the memory clock in this driver. [no] support right? Guess I don't really care what a memory clock is then, eh? Better to know for tomorrow's laptop though. :-) Well, there you can expect that the chipset is better supported. Setting the option PCI_Retry may also help however the cirrus driver has no support for this. We have a similar report about a Cirrus Laguna chip in our bugzilla. To be sure I understand -- My only course of action on my own is to try successively lower dot clocks in various ModeLines until I find one that works. Right? Only if you are in a high range. What I see from your logs however suggests to me that there may be something different that's causing this problem. Your clocks are already very low. Egbert. ___ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86
Re: [XFree86] Cirrus Logic GD 7548 Woes
Unfortunately you may be pretty much on your own. the Cirrus driver is pretty much unsupported as noone who could support it has the HW. If you find a solution please let us know. The spikes and crackles usually appear when the memory bus is congested. Sometimes this can get fixed by reducing the pixel clock or increasing the memory clock. Unfortunately there is support for chaning the memory clock in this driver. Setting the option PCI_Retry may also help however the cirrus driver has no support for this. We have a similar report about a Cirrus Laguna chip in our bugzilla. Egbert. Richard Lynch writes: Please CC me with any replies... I confess I'm not a regular reader, much less contributor, but have contributed much in the distant past to the PHP mailing list, FWIW. And I'll certainly add the info to the on-line database of your choice if we get this working! Hardware: Canon Innova Book 490 CDT (laptop) I can get X to sort of run at both 640x480 and 800x600. But it's all crackly -- like watching TV in the middle of a heavy wind-storm with a 1950's antenna. Lots of streaks and speckles, mostly black/white when the background is grey or white, and indigo blue when the background is black... It's all kinda speckly when it starts up as well, dividing the screen into chunks and the X cursor is, like, repeated six-fold in same-size array 3x2, with missing pixels in each repetition. Hope that makes sense. I've tried poking at the Config file and disabling accel and MMIO based on on-line research to no effect, and I'm running out of things to try, other than random ModeLine settings. I've uploaded my Config file and X log files to: http://www.l-i-e.com/xfree86/ so those of you who understand all that stuff could maybe take a peek, without cluttering up the list with them. Some ramblings that may be useful. Or not. I seem to recall that I fought this fight before, and won, but only by dumping the window manager that comes with the X downloadable binaries, and compiling everything from source, errr, maybe on another box?, and ... But maybe that was on my OTHER laptop (that got stolen). It's been s long since I tried to use X on this laptop, and I had to rip it out to make room for something else, and now I've lost my notes, and the on-line contribution I made back when there were only two (2) X laptop databases to contribute to seems to be AWOL... I can't get enough hard drive space to compile from source, which I usually prefer... While I've got your ear, some other issues: I'm not real fond of the Sparc-like X behaviour, but I've only got so much disk space and so much patience for compiling on a 133... Recommendations for a lightweight window manager that's more like Gnome would be most welcome. Can I compile X on a desktop and just copy all the stuff over?... It's not going to detect all the hardware properly, though, is it? Or is that all done at run-time? I don't really grok when X decides which modules/libraries to build and how. What would I have to do to get my fast (faster anyway) desktop to compile X for my laptop without screwing up the desktop settings? I've read some stuff that seems to apply, but it wasn't simplistic enough for my poor brain. The trackpad only seems to work every other re-boot, and when I plug in an external mouse, not at all. xf86cfg core dumps when trying to configure the second mouse -- Like Music? http://l-i-e.com/artists.htm ___ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86 ___ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86
Re: [XFree86] Cirrus Logic GD 7548 Woes
Egbert Eich writes: We have a similar report about a Cirrus Laguna chip in our bugzilla. ID #348. I forgot to mention this. Egbert. ___ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86
Re: [XFree86] Cirrus Logic GD 7548 Woes
Unfortunately you may be pretty much on your own. the Cirrus driver is pretty much unsupported as noone who could support it has the HW. Would it be practical to: A) Provide SSH root access to somebody @XFree86.org, or B) Ship somebody the laptop as a loaner C) ??? Obviously A) is much easier, but I dunno what you can tell without seeing the screen, and I don't have a web-cam handy... Though I do need to set one up eventually anyway, so I reckon I could move that higher on the priority list. And, of course, the odds on XFree86 folks having the time to mess with this are pretty small unless a lot of people have these things and so it's worth it to you to support them. If you find a solution please let us know. The spikes and crackles usually appear when the memory bus is congested. Sometimes this can get fixed by reducing the pixel clock or increasing the memory clock. I'll try that. Pixel clock is the dot clock right? What's the memory clock? Unfortunately there is support for chaning the memory clock in this driver. [no] support right? Guess I don't really care what a memory clock is then, eh? Better to know for tomorrow's laptop though. :-) Setting the option PCI_Retry may also help however the cirrus driver has no support for this. We have a similar report about a Cirrus Laguna chip in our bugzilla. To be sure I understand -- My only course of action on my own is to try successively lower dot clocks in various ModeLines until I find one that works. Right? Any recommended increments? Or is there some kind of algorithm for guessing these things, based on the one that kinda sorta works? Or should I attack it right at the minimum: (==) CIRRUS(0): Min pixel clock is 12 MHz (--) CIRRUS(0): Max pixel clock is 80 MHz (II) CIRRUS(0): Clock range: 12.00 to 80.10 MHz THANKS!!! -- Like Music? http://l-i-e.com/artists.htm ___ XFree86 mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/xfree86
