> > On 12/25/11 11:49, Leif Middelschulte wrote: >> Hi Mike, >> >> sorry for this late reply, but am kept busy with rl. >> Find comments in their respective context.
Hi Leif, no problem, a busy time of the year. My comments are below as well. >> 2011/12/24 mh<[email protected]>: >>> >>>>>> On Dec 23, 2011, at 12:27 PM, Leif Middelschulte wrote: >>>>>> >>>>>>> 2011/12/23 mh<[email protected]>: >>>>>>>> Salute: >>>>>>>> >>>>>>>> I'm having the same issue using two different distros on the same >>>>>>>> hardware (two drives). >>>>>>>> >>>>>>>> When I'm running Arch, I'm running xorg-server 1.11.3-1 , using >>>>>>>> xf86-video-ati 6.14.3-1 >>>>>>>> >>>>>>>> On Debian, sid, I have xserver-xorg-core 1.11.2.902-1, using >>>>>>>> xserver-xorg-video-radeon 6.14.3-2 >>>>>>>> >>>>>>> I'd like you to start e via gdb (compile e with debugging symbols, see >>>>>>> our wiki). Then break on _e_randr_crtc_info_set and step through that >>>>>>> function until bad things (crash) happens. >>>>>> >>>>>> Will do and let you know what I find here. >>>>> >>>>> Ok, I installed gdb, downloaded the latest 66482 from svn. Followed the >>>>> wiki >>>>> and rebuilt the EFLs after export CFLAG=-g. Now I'm stuck. From the wiki >>>>> it >>>>> seems that I should start e, then when it crashes start gdb and break on >>>>> _e_randr_crtc_info_set. Not sure how to do that. Help appreciated. >>>> >>>> On Dec 23, 2011, at 9:41 PM, Carsten Haitzler (The Rasterman) wrote: >>>> >>>> you're going to need to start an xserver raw without any "gdm". >>>> >>>> Xorg -ac :1& >>>> >>>> run that in a text console somewhere. >>>> in another text console do: >>>> >>>> export E_START=1 >>>> gdb enlightenment >> Thanks to raster for helping out here :-) >> But shouldn't this be >> gdb enlightenment_start instead? > > No. gdb enlightenment is correct. > > dh > > >>>> br _e_randr_crtc_info_set >>>> r >>>> >>>>> But, e never crashes though. What happens is that I'll run startx from >>>>> vt1, >>>> >>>> it won't - you have to make it STOP when it hits that function with a >>>> breakpoint (br). you have to set it up before e starts like above. >>>> >>> >>> Thanks! Attached are three sequential console shots, after hitting the >>> breakpoint three times . These seem to look exactly the messages I saw >>> earlier. >> >> Sadly you did something wrong when trying to run e via gdb. Ecore_X >> fails to initialize, which is way before e_randr initialization. I'm not sure what went wrong, I followed raster's directions on how to do this, as you can see in the console 1 screenshot earlier. But I agree that Ecore_X is failing to initialize. This is what always seems to happen. I wonder if there was some other change, besides to randr stuff, between 66218 and 66220? 66219 failed to compile here, I think it was e that failed, but it was a while ago when I tried that. Andrea, have you had a chance to try any of this? It would be interesting to see if you have the same result, which I think you might from your original stderr output, which looked very similar to me. I'm going to try recompiling 66218 with debug symbols and break on _e_randr_crtc_info_set just to see what I should expect it to do. It probably won't have any impact here, but a learning opportunity for me. Still ready to try other ideas! >>>>> and then in vt1 there are a series of messages like I included earlier in >>>>> the >>>>> file called "start-results". The messages say "Enlightenment relocation >>>>> handling", then many ESTART messages, then "Enlighenment relocation >>>>> handling" >>>>> again, followed by the same ESTART messages, then "<Enlightenment Error> >>>>> A >>>>> previous instance of Enlightenment is still active on this screen. >>>>> Aborting >>>>> startup." And then it repeats this process over and over until there's a >>>>> final error and X shuts down. >>>>> >>>>> So there doesn't seem to be a PID of enlightenment to run gdb on. I guess. >>>>> I've never really run gdb before. >>>>> >>>>> Any suggestions/hints would be appreciated, I'm willing to try and help >>>>> track >>>>> down the problem with this damn ATI hardware. >>>>> >>>>> >>>>>> mike >>>>>>>> >>>>>>>> >>>>>>>> On Dec 23, 2011, at 3:52 AM, Andrea Suisani wrote: >>>>>>>> >>>>>>>>> Hi all, >>>>>>>>> >>>>>>>>> sorry for not being able to join >>>>>>>>> the irc debug session last evening... >>>>>>>>> >>>>>>>>> though it seems we have have narrowed >>>>>>>>> the problem down to a peculiar comibination >>>>>>>>> of HW (ATI HD 5770/5750), e17 ver, >>>>>>>>> xorg version and x org drivere version. >>>>>>>>> >>>>>>>>> just trying get a more clear picture >>>>>>>>> I'm running X.Org X Server 1.10.4 >>>>>>>>> with fglrx used as video driver (defined >>>>>>>>> in xorg.conf) installed using >>>>>>>>> >>>>>>>>> apt-get install fglrx-updates >>>>>>>>> >>>>>>>>> on an ubuntu box ver 11.10, that's >>>>>>>>> the output of uname -a: >>>>>>>>> >>>>>>>>> Linux suino 3.0.0-14-server #23-Ubuntu SMP Mon Nov 21 20:49:05 UTC >>>>>>>>> 2011 >>>>>>>>> x86_64 x86_64 x86_64 GNU/Linux >>>>>>>>> >>>>>>>>> Mike what's your xorg version and >>>>>>>>> which driver are you using for your >>>>>>>>> 5770? >>>>>>>>> >>>>>>>>> I'm going to try to configure my Xorg >>>>>>>>> dropping the binary only driver and using >>>>>>>>> instead the one provided by: >>>>>>>>> >>>>>>>>> xserver-xorg-video-radeon >>>>>>>>> >>>>>>>>> i.e. >>>>>>>>> >>>>>>>>> http://www.x.org/wiki/radeon >>>>>>>>> >>>>>>>>> I'll report back >>>>>>>>> >>>>>>>>> >>>>>>>>> Andrea >>>>>>>>> >>>>>>>>> On 12/22/2011 08:49 PM, mh wrote: >>>>>>>>>> Leif, >>>>>>>>>> >>>>>>>>>> Following up on our irc session today: I >>>>>>>>>> deleted /opt/e17/lib/enlightenment/conf_randr and restarted e. Still >>>>>>>>>> had the black screen with gkrellm. >>>>>>>>>> >>>>>>>>>> Decided to try something else: I took the 5770 video card out of the >>>>>>>>>> box and tried the onboard video, which is a Radeon HD 4200. Restarted >>>>>>>>>> e, and the display worked properly. Recompiled e to rev 66466, and it >>>>>>>>>> worked fine. Shutdown, the reinstalled the Radeon HD 5770 based video >>>>>>>>>> card and restarted, back to a black display with gkrellm still >>>>>>>>>> working. >>>>>>>>>> >>>>>>>>>> So, from there I went back to the vt and recompiled e from v. 66218 >>>>>>>>>> and >>>>>>>>>> restarted e. Display and system worked fine. So that's where I'm at >>>>>>>>>> right now. >>>>>>>>>> >>>>>>>>>> mike >>> >>> > > > > ------------------------------------------------------------------------------ > Write once. Port to many. > Get the SDK and tools to simplify cross-platform app development. Create > new or port existing apps to sell to consumers worldwide. Explore the > Intel AppUpSM program developer opportunity. appdeveloper.intel.com/join > http://p.sf.net/sfu/intel-appdev > _______________________________________________ > enlightenment-devel mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/enlightenment-devel ------------------------------------------------------------------------------ Write once. Port to many. Get the SDK and tools to simplify cross-platform app development. Create new or port existing apps to sell to consumers worldwide. Explore the Intel AppUpSM program developer opportunity. appdeveloper.intel.com/join http://p.sf.net/sfu/intel-appdev _______________________________________________ enlightenment-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
