Re: [Dri-devel] Kernel back-merge ...
On Thu, 25 Sep 2003, Dave Airlie wrote: If Linus thinks your patch goes far enough for him.. I actually usually don't much like whitespace changes, and the one I sent was purely in response to the fact that there already was a lot of whitespace changes in the i810 driver. Don't worry too much about it. The i810 whitespace fixes were fine - it just became obvious that when they fixed something like if(x) to if (x) they still left strange mixes of spaces and tabs on the same line. All my whitespace changes was _literally_ from just editing the DRI CVS merge diff (any lines added) before applying to the kernel tree. I'll happily take more whitespace fixes, I just ask that as long as you play around with whitespace you don't do any other changes. A big whitespace diff that also has a subtle semantic change embedded in it is just too painful. Linus --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] Kernel back-merge ...
On Thu, 2003-09-25 at 22:58, Linus Torvalds wrote: On Thu, 25 Sep 2003, Dave Airlie wrote: If Linus thinks your patch goes far enough for him.. I actually usually don't much like whitespace changes, and the one I sent was purely in response to the fact that there already was a lot of whitespace changes in the i810 driver. Don't worry too much about it. The i810 whitespace fixes were fine - it just became obvious that when they fixed something like if(x) to if (x) they still left strange mixes of spaces and tabs on the same line. All my whitespace changes was _literally_ from just editing the DRI CVS merge diff (any lines added) before applying to the kernel tree. I'll happily take more whitespace fixes, I just ask that as long as you play around with whitespace you don't do any other changes. A big whitespace diff that also has a subtle semantic change embedded in it is just too painful. To me, if I was going to fix part of the whitespace in i810, I might as well fix all of the issues of the same type. diff -bu said that one was non-functional changes. Sorry about the SiS diff. I had done it all in separate stages in my perforce tree, but I didn't commit it to CVS until the work was all done. I probably should have done a two-stage commit there, I just didn't think this project cared about that. If you wanted to review the SiS stuff (which I would love for anyone to do), I have a diff of most of it. The big whitespace cleanup came after a bit of initial porting work, which involved axing a bit of dead code and using some os-independence macros. This should cover the functional changes that came after that: http://people.freebsd.org/~anholt/dri/files/sisdrm-functional.diff -- Eric Anholt[EMAIL PROTECTED] http://people.freebsd.org/~anholt/ [EMAIL PROTECTED] --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] MGA G550 w/ kernel mga module version 3.1.0 hard lock
On Thu, Sep 25, 2003 at 01:24:21PM -0400, Malcolm Mallardi wrote: Second concern. With any build of the drivers, certain applications and games when in hardware accellerated mode will not display the mouse cursor. Two games that readily come to mind are Unreal Tournament and Soldier of Fortune. Could this be an SDL bug rather than a bug with the matrox drivers? I suspect it's a matrox issue, as I've seen those games work with other video card drivers just fine. I am also unsure at this time how to get more information on this bug. I doubt it's an SDL bug because I think that most games render their own mouse cursor. I've seen this bug with freespace2 under XFree86. It works fine under DirectFB. A quick glance at the freespace2 code suggests that it renders the cursor with OpenGL. Third concern. Using XFree86 4.3.0 I have been unable to use any of the mga driver builds more recent than 09/12/2003, due to the fact that from the 13th on, X dumps me out with an ABI minor version number mismatch error. The XFree86 merge was finished around that time. Maybe this breakage is intentional? -- Ville Syrjälä [EMAIL PROTECTED] http://www.sci.fi/~syrjala/ --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
26 2003 07:48 Rafael Maximo : At 07:38 PM 25/9/2003, Felix Khling wrote: Hi Rafael, (a bit confused now, which one is your first name ;-) I'm just back from vacation and read about your progress. I'm eager to start testing and debugging the new driver. I think the best thing would be if you could commit your changes to the branch so that we both (and more people) could work with up-to-date code. I'm sure Eric Anholt would give you commit access on dri.freedesktop.org if you sent him your ssh key(s). Felix, I would like to commit the changes i made but it's not even working, that's why i didn't talk to Erich Anholt about commiting my changes because i still need to work on changing ioctl to drmCommand* calls and i didn't have time to work on it this week. I have downoaded savage-0-0-1-branch from dri.freedesktop.org cvs at 23 september Does the code I got have all changes you did ? BTW - my first name is Rafael :) Dimitry --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
On Fri, 26 Sep 2003 14:03:39 +0600 Dimitry N. Naldaev [EMAIL PROTECTED] wrote: Ð ÑообÑении Ð¾Ñ 26 СенÑÑбÑÑ 2003 07:48 Rafael Maximo напиÑал: At 07:38 PM 25/9/2003, Felix Kühling wrote: Hi Rafael, (a bit confused now, which one is your first name ;-) I'm just back from vacation and read about your progress. I'm eager to start testing and debugging the new driver. I think the best thing would be if you could commit your changes to the branch so that we both (and more people) could work with up-to-date code. I'm sure Eric Anholt would give you commit access on dri.freedesktop.org if you sent him your ssh key(s). Felix, I would like to commit the changes i made but it's not even working, that's why i didn't talk to Erich Anholt about commiting my changes because i still need to work on changing ioctl to drmCommand* calls and i didn't have time to work on it this week. I have downoaded savage-0-0-1-branch from dri.freedesktop.org cvs at 23 september Does the code I got have all changes you did ? This is the wrong branch. It contains the start of a savage driver from scratch and is not working at this point. The S3-based code is on the savage-1_0_0-branch. It does not contain Rafaels changes yet. BTW - my first name is Rafael :) Dimitry Felix __\|/_____ ___ - Felix ___\_e -_/___/ __\___/ __\_ You can do anything, Kühling (_\Ä// /_/ /) just not everything [EMAIL PROTECTED] \___/ \___/ Uat the same time. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
26 2003 14:31 Felix Khling : On Fri, 26 Sep 2003 14:03:39 +0600 Dimitry N. Naldaev [EMAIL PROTECTED] wrote: 26 2003 07:48 Rafael Maximo : At 07:38 PM 25/9/2003, Felix Khling wrote: Hi Rafael, (a bit confused now, which one is your first name ;-) I'm just back from vacation and read about your progress. I'm eager to start testing and debugging the new driver. I think the best thing would be if you could commit your changes to the branch so that we both (and more people) could work with up-to-date code. I'm sure Eric Anholt would give you commit access on dri.freedesktop.org if you sent him your ssh key(s). Felix, I would like to commit the changes i made but it's not even working, that's why i didn't talk to Erich Anholt about commiting my changes because i still need to work on changing ioctl to drmCommand* calls and i didn't have time to work on it this week. I have downoaded savage-0-0-1-branch from dri.freedesktop.org cvs at 23 september Does the code I got have all changes you did ? This is the wrong branch. It contains the start of a savage driver from scratch and is not working at this point. The S3-based code is on the savage-1_0_0-branch. Oh yes I have misspellesd branch name once more :-( the savage-0-0-1-branch is branch I had before and now I have updated up to savage-1_0_0-branch. It does not contain Rafaels changes yet. :-( may be he send his changes in the mail list? --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
[Dri-devel] Problem with color using R200 radeons
Hello! I'm developing a graphic driver for radeon cards as part of a research OS (PLURIX) at an university. We're already supporting the R100-based radeon cards (including the (simple) TCL) and want also to support the R200-based cards. But with these cards I have a big problem: I can draw primitives with CP type-3 packets but the color is missing. In other words I can use the CP drawing commands (3D_DRAW_IMMD or 3D_DRAW_IMMD_2) and the primitives appear at the specified position but only in black, the color information included in the command packet (and in the vertex format definition) is completely ignored. (...and 3D-graphics ist not very exciting if it's only black on black:-) When I write directly in the visible area or use for instance the 2D-command PAINT_MULTI the color value is normaly used. Are there some extra registers (or one) which needs to be initialized in the beginning so that the color information will be used or have I missed another thing? I've already spent quite a time debugging and serching for a solution but now I'm running out of ideas (and ATI won't answer on mails although we're registered developers:-(. I would be glad if somebody could help me, even if it's perhaps a little bit off-topic. Thanks, Markus Fakler -- Markus Fakler *** [EMAIL PROTECTED] --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] Website - Wiki (RFC)
On Fri, 26 Sep 2003 00:53:13 +0100 José Fonseca [EMAIL PROTECTED] wrote: BTW, have u seen http://dri.sourceforge.net/cgi-bin/moin.cgi/ConfigurationInfrastructure ? ;-) Feel free to add any more info you see fit. Also, I noticed I put your name wrong there too: The umlaut ü-ue thingy I didn't knew again..! Ah, I missed that one before. I fixed my name and added a home page. I'm going to add more information to ConfigurationInfrastructure later. José Fonseca PS: Still have to sort the config-*-* branch snapshots issues :-/ IIRC the only issue was the missing xdriinfo executable. Felix __\|/_____ ___ - Felix ___\_e -_/___/ __\___/ __\_ You can do anything, Kühling (_\Ä// /_/ /) just not everything [EMAIL PROTECTED] \___/ \___/ Uat the same time. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
At 06:28 AM 26/9/2003, Dimitry N. Naldaev wrote: Ð ÑообÑении Ð¾Ñ 26 СенÑÑбÑÑ 2003 14:31 Felix Kühling напиÑал: On Fri, 26 Sep 2003 14:03:39 +0600 Dimitry N. Naldaev [EMAIL PROTECTED] wrote: àÃÂþþñÃÂõýøø þà26 áõýÃÂÃÂñÃÂà2003 07:48 Rafael Maximo ýðÿøÃÂðû: At 07:38 PM 25/9/2003, Felix Kühling wrote: Hi Rafael, (a bit confused now, which one is your first name ;-) I'm just back from vacation and read about your progress. I'm eager to start testing and debugging the new driver. I think the best thing would be if you could commit your changes to the branch so that we both (and more people) could work with up-to-date code. I'm sure Eric Anholt would give you commit access on dri.freedesktop.org if you sent him your ssh key(s). Felix, I would like to commit the changes i made but it's not even working, that's why i didn't talk to Erich Anholt about commiting my changes because i still need to work on changing ioctl to drmCommand* calls and i didn't have time to work on it this week. I have downoaded savage-0-0-1-branch from dri.freedesktop.org cvs at 23 september Does the code I got have all changes you did ? This is the wrong branch. It contains the start of a savage driver from scratch and is not working at this point. The S3-based code is on the savage-1_0_0-branch. Oh yes I have misspellesd branch name once more :-( the savage-0-0-1-branch is branch I had before and now I have updated up to savage-1_0_0-branch. It does not contain Rafaels changes yet. :-( may be he send his changes in the mail list? I didn't commit any change to CVS yet, basicaly because all problems i got. For now the 2D driver need to be fixed (i didn't have time to look deep in this problem) and the code of the savage driver from S3 is a little bit confusing and sometimes very different from the other drivers and everything need to be fixed (2D, 3D and probably the kernel module). Besides all these problems, i'm not a very experienced programer witch make things e little harder for me. Any help will be very welcome. bye. Rafael Máximo --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
On Fri, 26 Sep 2003 07:36:02 -0300 Rafael Maximo [EMAIL PROTECTED] wrote: At 06:28 AM 26/9/2003, Dimitry N. Naldaev wrote: Ð_ Ñ_ообÑ_ении оÑ_ 26 СенÑ_Ñ_бÑ_Ñ_ 2003 14:31 Felix Kühling напиÑ_ал: On Fri, 26 Sep 2003 14:03:39 +0600 Dimitry N. Naldaev [EMAIL PROTECTED] wrote: Ã_Â_ Ã_Â_Ã_¾Ã_¾Ã_±Ã_Â_Ã_µÃ_½Ã_¸Ã_¸ Ã_¾Ã_Â_ 26 Ã_¡Ã_µÃ_½Ã_Â_Ã_Â_Ã_±Ã_Â_Ã_Â_ 2003 07:48 Rafael Maximo Ã_½Ã_°Ã_¿Ã_¸Ã_Â_Ã_°Ã_»: At 07:38 PM 25/9/2003, Felix Kühling wrote: Hi Rafael, (a bit confused now, which one is your first name ;-) I'm just back from vacation and read about your progress. I'm eager to start testing and debugging the new driver. I think the best thing would be if you could commit your changes to the branch so that we both (and more people) could work with up-to-date code. I'm sure Eric Anholt would give you commit access on dri.freedesktop.org if you sent him your ssh key(s). Felix, I would like to commit the changes i made but it's not even working, that's why i didn't talk to Erich Anholt about commiting my changes because i still need to work on changing ioctl to drmCommand* calls and i didn't have time to work on it this week. I have downoaded savage-0-0-1-branch from dri.freedesktop.org cvs at 23 september Does the code I got have all changes you did ? This is the wrong branch. It contains the start of a savage driver from scratch and is not working at this point. The S3-based code is on the savage-1_0_0-branch. Oh yes I have misspellesd branch name once more :-( the savage-0-0-1-branch is branch I had before and now I have updated up to savage-1_0_0-branch. It does not contain Rafaels changes yet. :-( may be he send his changes in the mail list? I didn't commit any change to CVS yet, basicaly because all problems i got. For now the 2D driver need to be fixed (i didn't have time to look deep in this problem) and the code of the savage driver from S3 is a little bit confusing and sometimes very different from the other drivers and everything need to be fixed (2D, 3D and probably the kernel module). Besides all these problems, i'm not a very experienced programer witch make things e little harder for me. Any help will be very welcome. Ok. I'm just checking out the savage-1_0_0-branch. I'll see if I can get 2D working here. bye. Rafael Máximo Felix __\|/_____ ___ - Felix ___\_e -_/___/ __\___/ __\_ You can do anything, Kühling (_\Ä// /_/ /) just not everything [EMAIL PROTECTED] \___/ \___/ Uat the same time. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
[Dri-devel] R128 broken
I think R128 support is broken in the current 2.6 bk kernel. All 3D apps segfault. Can anyone else verify? Radeon support works without problem. = Jon Smirl [EMAIL PROTECTED] __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] R128 broken
On Fri, 2003-09-26 at 14:06, Jon Smirl wrote: I think R128 support is broken in the current 2.6 bk kernel. All 3D apps segfault. Can anyone else verify? Have you verified that the same 3D driver works with another DRM? IIRC you posted somewhere that it segfaults in the Mesa SSE check, which is unlikely to be related to the DRM. -- Earthling Michel Dnzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] MGA G550 w/ kernel mga module version 3.1.0 hard lock
On Fri, Sep 26, 2003 at 10:49:03AM +0300, Ville Syrj?l? wrote: On Thu, Sep 25, 2003 at 01:24:21PM -0400, Malcolm Mallardi wrote: I doubt it's an SDL bug because I think that most games render their own mouse cursor. I've seen this bug with freespace2 under XFree86. It works fine under DirectFB. A quick glance at the freespace2 code suggests that it renders the cursor with OpenGL. Yes, I figured this, but I felt that I should throw the idea out there just in case. :) Third concern. Using XFree86 4.3.0 I have been unable to use any of the mga driver builds more recent than 09/12/2003, due to the fact that from the 13th on, X dumps me out with an ABI minor version number mismatch error. The XFree86 merge was finished around that time. Maybe this breakage is intentional? I resolved this issue yesterday after sending the first mail. Definitely an intentional breakage. Now using the build from yesterday 09/25/03 The OpenGL mouse issue, and console switching issues with kernel module 3.1.0 still exist. -- Malcolm D. Mallardi - Dark Freak At Large Captain, we are receiving two-hundred eighty-five THOUSAND hails. AOL: Nuark UIN: 11084092 Y!: Magamo Jabber: [EMAIL PROTECTED] http://ranka.2y.net:8008/~magamo/index.htm Unsolicited Advertisements are NOT welcome. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
[Dri-devel] Radeon/r200 texture engine question
I'm thinking about implementing a textured video option on radeon for mergedfb (or non-mergedfb too for that matter). It would be useful since video windows could span heads in mergedfb mode and you could also have multiple Xv windows at the same time rather than just one with the overlay unit. My questions are as follows: 1. looking at the mga textured video code, it disables the DRI when it is in use. is there any reason for this? is it possible to use the texture engine for Xv and DRI at the smae time. It seems like it should. I think directfb does this for 3D and alpha blended windows already. When implemeting this on radeon, would I have to use the CP to program the texture registers or could I use MMIO? I've been toying with the idea of converting the existing Xv code to CP anyway, however if you disabled the DRI and wanted to use the overlay or the texture engine you'd need an MMIO code path at least. 2. I don't really understand the 3D driver. Could some one explain to me which registers I would need to manipulate to use the texture engine for Xv? or at least how it should work for 3D so I can attempt it for Xv? I assume they are defined in radeon_reg.h? They seem to be... 3. Is there a limit on the number of textures on radeon? How many Xv ports should I allow for the texture engine? I think mga allows 32, although that might have been an arbitrary decision. 4. Can the overlay and the texture engine Xv adapters co-exist? for instance when I register the Xv adapters can I register the texture Xv adapter and the overlay Xv adapter? that way I could choose to use one or the other by specifying a different Xv port in Xv apps. I don't see why I couldn't, I'm just wondering if there are any issues with doing that. 5. r200 vs r100 - how different are the texture engines? 6. Anything I'm missing? Thanks, Alex __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
I found at least a workaround for the problem. Set Option DisableTile on in the Device section of XF86Config-4. I think this has something to do with the frame buffer layout. Accellerated stuff seems to get it right. But non-accelerated stuff accesses the frame buffer in the wrong way. Try Option Accel off and you'll see what I mean. Basically the screen is squeezed vertically to half the height. I havn't found a real fix yet. Does someone with more 2D experience have a pointer? Felix On Fri, 26 Sep 2003 07:36:02 -0300 Rafael Maximo [EMAIL PROTECTED] wrote: [snip] I didn't commit any change to CVS yet, basicaly because all problems i got. For now the 2D driver need to be fixed (i didn't have time to look deep in this problem) and the code of the savage driver from S3 is a little bit confusing and sometimes very different from the other drivers and everything need to be fixed (2D, 3D and probably the kernel module). Besides all these problems, i'm not a very experienced programer witch make things e little harder for me. Any help will be very welcome. bye. Rafael Máximo --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel __\|/_____ ___ - Felix ___\_e -_/___/ __\___/ __\_ You can do anything, Kühling (_\Ä// /_/ /) just not everything [EMAIL PROTECTED] \___/ \___/ Uat the same time. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
I think I found it. Try this patch with patch -p0 in xc/programs/Xserver/hw/xfree86/drivers/savage: --- ./savage_driver.c.~1.10.10.5.~ 2003-09-26 14:00:23.0 +0200 +++ ./savage_driver.c 2003-09-26 19:03:30.0 +0200 @@ -2601,7 +2601,7 @@ pScrn = xf86Screens[pScreen-myNum]; psav = SAVPTR(pScrn); -displayWidth = pScrn-displayWidth; +displayWidth = psav-ulAperturePitch / (pScrn-bitsPerPixel 3); if (psav-rotate) { height = pScrn-virtualX; The ulAperturePitch wasn't used otherwise and this appears to be the right place. It works for me with and without tile mode. Felix On Fri, 26 Sep 2003 17:35:09 +0200 Felix Kühling [EMAIL PROTECTED] wrote: I found at least a workaround for the problem. Set Option DisableTile on in the Device section of XF86Config-4. I think this has something to do with the frame buffer layout. Accellerated stuff seems to get it right. But non-accelerated stuff accesses the frame buffer in the wrong way. Try Option Accel off and you'll see what I mean. Basically the screen is squeezed vertically to half the height. I havn't found a real fix yet. Does someone with more 2D experience have a pointer? Felix On Fri, 26 Sep 2003 07:36:02 -0300 Rafael Maximo [EMAIL PROTECTED] wrote: [snip] I didn't commit any change to CVS yet, basicaly because all problems i got. For now the 2D driver need to be fixed (i didn't have time to look deep in this problem) and the code of the savage driver from S3 is a little bit confusing and sometimes very different from the other drivers and everything need to be fixed (2D, 3D and probably the kernel module). Besides all these problems, i'm not a very experienced programer witch make things e little harder for me. Any help will be very welcome. bye. Rafael Máximo __\|/_____ ___ - Felix ___\_e -_/___/ __\___/ __\_ You can do anything, Kühling (_\Ä// /_/ /) just not everything [EMAIL PROTECTED] \___/ \___/ Uat the same time. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] DRI savage driver status
Hello Dimitry, Thanks for the patch. The BCI_LINE_MISC thing is already corrected on the savage-1_0_0-branch. The other things look like attempts to enable DRI on a chip for which DRI is not supported at this point. As was said before, the S3 driver has DRI support for the Savage4 series but not the SavageMX. When we have Savage4 working we (you?) may be able to enable DRI on the SavageMX. But this is apperently not just a matter of changing a few flags. I don't know how big the differences are. It was mentioned that UtahGLX has a working SavageMX driver. Regards, Felix On Fri, 26 Sep 2003 13:46:01 +0600 Dimitry N. Naldaev [EMAIL PROTECTED] wrote: Ð ÑообÑении Ð¾Ñ 26 СенÑÑбÑÑ 2003 04:40 Felix Kühling напиÑал: On Mon, 22 Sep 2003 11:26:50 +0600 Dimitry N. Naldaev [EMAIL PROTECTED] wrote: Hi, I decided to focus on 2D problem first but i don't know how I can debug the 2D driver and found where is the problem. I would appreciate any information or docs about debugging this kind of problem. I have fixed some problems for Xfree 4.2.0 and savage MX but there are still more :-( so I can try to help you Could you send a patch of the changes you made to the 2D driver? Yes I could :-) I try to enable DRI on Savage MX with S3 XFree86 4.2.0 3D driver After Instaling S3 driver on XFree86 4.2.0 source, I patch the source up to 4.2.1 because this is version of X instaled on my notebook. The patch have been applied cleanly and do not touch any files from S3 After that I have done make World mainly to make Makefiles (and what are needded besides) But I do not Install whole new build of X instead I go in S3 savage driver dirs and make insttall as root in them (I have made backup copy of /usr/X11R6/lib/modules/drivers/savage_drv.o before) This patch need to be applied in xc/programs/Xserver/hw/xfree86/drivers/savage directory with command patch -p1 This patch solve the next problems: 1) It force enabling DRI on all Savage chips but to do this you need to define symbol FORCE_DRI in compile time. This can be done, for example, with command make EXTRA_DEFINES=-DFORCE_DRI So errors (EE) SAVAGE(0): DRI isn't enabled (EE) SAVAGE(0): Direct rendering disabled go away from the XFee86.log 2)I fix init screen corruption on Savage MX bug when DRI forced. The problem was caused by the nex code which is part of savage_dri.c/SAVAGEDRIFinishScreenInit() function { unsigned int value = 0; OUTREG(0x850C,(INREG(0x850C) | 0x8000)); /* we don't use Y range flag,so comment it */ /* if(pSAVAGEDRI-width = 1024) value |= (129); */ if(pSAVAGEDRI-cpp == 2) { value |= (((pSAVAGEDRI-width + 0x3F) 0xFFC0) 6) 20; value |= 230; } else { value |= (((pSAVAGEDRI-width + 0x1F) 0xFFE0) 5) 20; value |= 330; } OUTREG(0x48C40, value|(pSAVAGEDRI-frontOffset 5) );/* front */ OUTREG(0x48C44, value|(pSAVAGEDRI-backOffset 5) );/* back */ OUTREG(0x48C48, value|(pSAVAGEDRI-depthOffset 5) );/* depth */ } Are this code needed at all ? The problem of the code is the grep 0x48C40 *.c in the savage 2D driver dir provide _exactly_ one match located in the above code and SAVAGEDRIFinishScreenInit() function is called from SavageScreenInit after SavageSaveRestoreState() call !!! so after VT switch ie after SavageLeaveVT() and SavageEnterVT() calls we restore original state of the registers...(Am I right or I am mistaken?) It is why VT switch fix screen corrution on savage MX So test my patch on oficialy DRI enebled chips please and report if it work... 3) It fix Memory manager initialization fall bug. So error (EE) SAVAGE(0): Memory manager initialization to (0,0) (1024,-1) failed. go away from the XFee86.log. The ful log is attached glxinfo report dri enabled see attachment and glxgears now do not crash Xserver but eat all processor power so system practicaly do not react on the keyboard and mouse button events and unusable :-( there is nothing appear in glxrears window. The output of glxgears attached. anormal high fps was when I hid the glxgears window by another window... whre I need to look / dig so solve this ptoblem ? I haven't any idia yet :-( 4) The patch DO NOT solve the enlightenmetn screen corruption problem I have reported before :-( this poblem need more investigation and can bee related with problem reported by Rafael. Are you using enlightenment window manager? or what WM you are using? AFAICT Rafael didn't change the 2D driver yet, so your patch should apply cleanly. My patch probably do not fix Rafael problem (BTW what video hardware you are using?) Now I have updated my work copy of savage-0-0-1-branch to savage-1_0_0-branch and notice that some files in savage catalogue
Re: [Dri-devel] DRI savage driver status
I believe there was some discussion of this on the directfb mailing list. I think the savage chips require a tile based framebuffer for acceleration, but this is incompatible with things that assume a linear based framebuffer. Alex --- Felix K�hling [EMAIL PROTECTED] wrote: I found at least a workaround for the problem. Set Option DisableTile on in the Device section of XF86Config-4. I think this has something to do with the frame buffer layout. Accellerated stuff seems to get it right. But non-accelerated stuff accesses the frame buffer in the wrong way. Try Option Accel off and you'll see what I mean. Basically the screen is squeezed vertically to half the height. I havn't found a real fix yet. Does someone with more 2D experience have a pointer? Felix On Fri, 26 Sep 2003 07:36:02 -0300 Rafael Maximo [EMAIL PROTECTED] wrote: [snip] I didn't commit any change to CVS yet, basicaly because all problems i got. For now the 2D driver need to be fixed (i didn't have time to look deep in this problem) and the code of the savage driver from S3 is a little bit confusing and sometimes very different from the other drivers and everything need to be fixed (2D, 3D and probably the kernel module). Besides all these problems, i'm not a very experienced programer witch make things e little harder for me. Any help will be very welcome. bye. Rafael M�ximo --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel __\|/_____ ___ - Felix ___\_e -_/___/ __\___/ __\_ You can do anything, K�hling (_\�// /_/ /) just not everything [EMAIL PROTECTED] \___/ \___/ Uat the same time. --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
[Dri-devel] drm vs. drm-4.0
In the linux 2.4 kernel tree under drivers/char there is both a drm subdirectory and a drm-4.0 subdirectory. I've looked high and low for an explanation of drm vs. drm-4.0, I assume its a version difference, but I've come up dry. Can someone either explain to me what drm-4.0 is or send me a pointer to something that documents it. I'd like to know why one implementation is picked over the other, are there version dependencies, why it exists as parallel to drm and what its trying to fix. Thanks, John -- John Dennis [EMAIL PROTECTED] --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] R128 broken
Here's what I know: Redhat Rawhide Xfree works on 2.4 uni and smp with R128 DRM Redhat Rawhide Xfree fails on 2.6 uni and smp, using kernel R128 DRM DRI CVS fails on 2.6 uni and smp, using kernel R128 DRM All combinations work on Radeon All failures are the same, immediate segfault of 3D apps. I am not getting a failure from the DRM driver, the fault is in r128_dri.so I am using a PCI Rage128 so it is uncommon. Does anyone else have a Rage128 that can give this a quick check? I tried using gdb on this but I don't know how to get by the SSE segfault which apparently is supposed to happen. --- Michel Dänzer [EMAIL PROTECTED] wrote: On Fri, 2003-09-26 at 14:06, Jon Smirl wrote: I think R128 support is broken in the current 2.6 bk kernel. All 3D apps segfault. Can anyone else verify? Have you verified that the same 3D driver works with another DRM? IIRC you posted somewhere that it segfaults in the Mesa SSE check, which is unlikely to be related to the DRM. -- Earthling Michel Dänzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer = Jon Smirl [EMAIL PROTECTED] __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] drm vs. drm-4.0
The DRM changed between xfree86 4.0.x and 4.1.0 in an incompatible way. some kernel trees retained both DRMs so that you could build the appropriate DRM depending on what version of xfree86 you were using. Alex --- John Dennis [EMAIL PROTECTED] wrote: In the linux 2.4 kernel tree under drivers/char there is both a drm subdirectory and a drm-4.0 subdirectory. I've looked high and low for an explanation of drm vs. drm-4.0, I assume its a version difference, but I've come up dry. Can someone either explain to me what drm-4.0 is or send me a pointer to something that documents it. I'd like to know why one implementation is picked over the other, are there version dependencies, why it exists as parallel to drm and what its trying to fix. Thanks, John -- John Dennis [EMAIL PROTECTED] __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
[Dri-devel] radeon 9200se
hi all, my system configuration: P4 2,4Ghz fsb 800 HT super micro p4spa+ mainboard radeon 9200se graphics card running debian woody with kernel 2.4.21 XFree86 4.3.99-12 compiled from source recently i bought a radeon 9200se graphics card. it took me a very long time to get it work in X. its because the ati driver only knows about the 9200 but not the 9200se. the pci-chip-id is differen (9200 = 5963 / 9200se = 5964) i found a patch on the internet that makes the driver recognize that chip. (http://www.mail-archive.com/[EMAIL PROTECTED]/msg12830.html) i recompiled XFree86 and it worked perfectly. now i want dri support too. i downloaded the radeon package (radeon-20030926-linux.i386.tar.bz2) run install.sh and everything worked great but when i run startx i get the same message as when the card wasnt detected by the driver before. [quote /var/log/XF86log.0.log] ...snip (II) Primary Device is: PCI 01:00:0 (--) Assigning device section with no busID to primary device (EE) No devices detected. [/quote] so my question is: does the radeon dri package support the 9200se and if not is it possible to apply the above patch to make it support it? secondly is it possible to get dri support right when building XFree86 using my existing and working driver? thanks for your help m --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] radeon 9200se
you can either patch DRI cvs with the 9200se patch you tried or force the chipID to another radeon 9200. any of these chould work: 0x5963 0x5962 0x5961 0x5960 Also I believe Michel added the patch to CVS so it should show up in the snapshots soon. Alex --- maestro [EMAIL PROTECTED] wrote: hi all, my system configuration: P4 2,4Ghz fsb 800 HT super micro p4spa+ mainboard radeon 9200se graphics card running debian woody with kernel 2.4.21 XFree86 4.3.99-12 compiled from source recently i bought a radeon 9200se graphics card. it took me a very long time to get it work in X. its because the ati driver only knows about the 9200 but not the 9200se. the pci-chip-id is differen (9200 = 5963 / 9200se = 5964) i found a patch on the internet that makes the driver recognize that chip. (http://www.mail-archive.com/[EMAIL PROTECTED]/msg12830.html) i recompiled XFree86 and it worked perfectly. now i want dri support too. i downloaded the radeon package (radeon-20030926-linux.i386.tar.bz2) run install.sh and everything worked great but when i run startx i get the same message as when the card wasnt detected by the driver before. [quote /var/log/XF86log.0.log] ...snip (II) Primary Device is: PCI 01:00:0 (--) Assigning device section with no busID to primary device (EE) No devices detected. [/quote] so my question is: does the radeon dri package support the 9200se and if not is it possible to apply the above patch to make it support it? secondly is it possible to get dri support right when building XFree86 using my existing and working driver? thanks for your help m --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] R128 broken
On Fri, 2003-09-26 at 21:17, Jon Smirl wrote: Redhat Rawhide Xfree works on 2.4 uni and smp with R128 DRM Redhat Rawhide Xfree fails on 2.6 uni and smp, using kernel R128 DRM DRI CVS fails on 2.6 uni and smp, using kernel R128 DRM All combinations work on Radeon Could still be another difference between 2.4 and 2.6 kernel. All failures are the same, immediate segfault of 3D apps. I am not getting a failure from the DRM driver, the fault is in r128_dri.so I am using a PCI Rage128 so it is uncommon. Does anyone else have a Rage128 that can give this a quick check? I tried using gdb on this but I don't know how to get by the SSE segfault which apparently is supposed to happen. The floating point exception is normal, but not the segfault. Does it work with the MESA_NO_SSE environment variable set? -- Earthling Michel Dnzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] radeon 9200se
On Fri, 2003-09-26 at 22:30, Alex Deucher wrote: Also I believe Michel added the patch to CVS so it should show up in the snapshots soon. It's only in XFree86 CVS yet, but overriding the chip ID to another 9200 should be fine. -- Earthling Michel Dnzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
[Dri-devel] Re: CVS Update: xc (branch: trunk)
On Fri, 2003-09-26 at 19:26, Alan Hourihane wrote: Log message: no longer need radeon_pci.h since the XFree86 merge Then it was never needed. I added it in order to * try and confine radeon driver merges between XFree86 and DRI CVS to the ati driver directory (as mentioned in the CVS log of the initial revision) * allow the driver to build against an older SDK -- Earthling Michel Dnzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] R128 broken
I believe this is a problem isolated to the Redhat Rawhide Xfree binaries, specifically their r128_dri.so modules. No version of the DRM module is causing problems. I was tricked earlier when I though I was running DRI CVS but in fact I wasn't. RH has a tls enabled version in /usr/X11R6/lib/modules/dri/tls. After I copied the DRI tree in it was still using the tls version which I was unaware of. After I deleted the tls version DRI CVS works now. --- Michel Dänzer [EMAIL PROTECTED] wrote: On Fri, 2003-09-26 at 21:17, Jon Smirl wrote: Redhat Rawhide Xfree works on 2.4 uni and smp with R128 DRM Redhat Rawhide Xfree fails on 2.6 uni and smp, using kernel R128 DRM DRI CVS fails on 2.6 uni and smp, using kernel R128 DRM All combinations work on Radeon Could still be another difference between 2.4 and 2.6 kernel. All failures are the same, immediate segfault of 3D apps. I am not getting a failure from the DRM driver, the fault is in r128_dri.so I am using a PCI Rage128 so it is uncommon. Does anyone else have a Rage128 that can give this a quick check? I tried using gdb on this but I don't know how to get by the SSE segfault which apparently is supposed to happen. The floating point exception is normal, but not the segfault. Does it work with the MESA_NO_SSE environment variable set? -- Earthling Michel Dänzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-develceforge.net https://lists.sourceforge.net/lists/listinfo/dri-devel = Jon Smirl [EMAIL PROTECTED] __ Do you Yahoo!? The New Yahoo! Shopping - with improved product search http://shopping.yahoo.com --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel
Re: [Dri-devel] Radeon/r200 texture engine question
On Fri, 2003-09-26 at 18:16, Alex Deucher wrote: 1. looking at the mga textured video code, it disables the DRI when it is in use. is there any reason for this? is it possible to use the texture engine for Xv and DRI at the smae time. It seems like it should. I agree. When implemeting this on radeon, would I have to use the CP to program the texture registers or could I use MMIO? Using the 3D engine without the CP is hard if not impossible. I've been toying with the idea of converting the existing Xv code to CP anyway, however if you disabled the DRI and wanted to use the overlay or the texture engine you'd need an MMIO code path at least. The register write macros could be changed (or new ones added) to use the CP transparently when appropriate. 3. Is there a limit on the number of textures on radeon? How many Xv ports should I allow for the texture engine? I think mga allows 32, although that might have been an arbitrary decision. Yes, I don't think there's a limitation other than the amount of video RAM. 4. Can the overlay and the texture engine Xv adapters co-exist? I don't see why they couldn't. AFAIK other drivers offer both types of adaptors simultaneously. I hope others will fill in the questions I didn't feel competent enough to answer. As a bonus, once you've solved all the issues involved, adding some RENDER extension acceleration should be trivial. :) I played with that a little but never found the time to get it very far. -- Earthling Michel Dnzer \ Debian (powerpc), XFree86 and DRI developer Software libre enthusiast \ http://svcs.affero.net/rm.php?r=daenzer --- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf ___ Dri-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/dri-devel