Re: Trying to use Radeon driver on Apple eMac
On Mon, Dec 22, 2008 at 2:48 AM, Kerry Hall kerryh...@gmail.com wrote: Hey Alex! Yes, this is another 1.25 Ghz eMac, with a Radeon 9200. I need to compile the ATI driver from git? What version of the driver would this be, and how do I compile it? I tried the Connector Table option and the Modelines from the eMac thread. I tried them all in various combinations, with none of them working. Do I need BOTH the bleeding edge ATI driver and the modeline and connectortable options? well, assuming your eMac is wired up the same way (Joseph's eMac has a radeon 7500), the current code in ati git should just work. If it's wired up differently or has an EDID for the internal monitor we'll need to make some changes. So if you want to use the driver from git: 1. install the xorg header packages from your distro 2. install git-core 3. git clone git://anongit.freedesktop.org/xorg/driver/xf86-video-ati 4. cd xf86-video-ati 5. ./autogen.sh --prefix=/usr 6. make 7. (as root) make install It should autodetect your mac and do the right thing. If not, add: Option MacModel emac to the device section of your config. To use your current driver: grab the xorg,conf from Joseph's site: http://www.funsitelots.com/pub/xorg_emacG4_working/ if neither works, Can you send me the output of /proc/cpuinfo? then try the following (from earlier in this thread): Try the following option and let me know which combinations work: Option ConnectorTable i,1,0,1,j,2,0,1 where i and j are one of the following: 96, 100, 104, or 108 and i != j. and send me one of the logs with the option enabled. Alex Thank you so much Alex, this is really driving me nuts. Kerry On Sun, Dec 21, 2008 at 9:39 PM, Alex Deucher alexdeuc...@gmail.com wrote: On Fri, Dec 19, 2008 at 6:11 PM, Kerry Hall kerryh...@gmail.com wrote: Hey Alex! I basically tried everything that Joseph tried. The screen flickers twice, then is black. The only thing that has made any difference in visual output has been trying various connector table options, but I am not sure which ones to use. When using a connector table option, a few white lines appear on the screen but then it goes black. Also, I recompiled the xserver-xorg-video-ati package, I am using version 6.8.0. I am almost at my wit's end with this. Posting Xorg.0.log and my xorg.conf. Thanks!! Hey Kerry, What card/system are you trying to get working? If this is another eMac, you 'll need to use the ati driver from git, or add the ConnectorTable option and modeline(s) from the eMac thread. Alex a href=http://www.filefactory.com/file/abd425/n/xorg_conf;xorg.conf/abr / a href=http://www.filefactory.com/file/80cbb0/n/Xorg_0_log;Xorg.0.log/abr / -- (A) // (E) ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
Hey Alex! I basically tried everything that Joseph tried. The screen flickers twice, then is black. The only thing that has made any difference in visual output has been trying various connector table options, but I am not sure which ones to use. When using a connector table option, a few white lines appear on the screen but then it goes black. Also, I recompiled the xserver-xorg-video-ati package, I am using version 6.8.0. I am almost at my wit's end with this. Posting Xorg.0.log and my xorg.conf. Thanks!! a href=http://www.filefactory.com/file/abd425/n/xorg_conf;xorg.conf/abr / a href=http://www.filefactory.com/file/80cbb0/n/Xorg_0_log;Xorg.0.log/abr / ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
On Tue, 2008-12-16 at 17:37 -0500, Alex Deucher wrote: The EDID for the built-in monitor of all PowerPC Macs can be found under /proc/device-tree. Since it's hard for anyone but Apple to know about every single type of monitor they've put into their computers, wouldn't it be better for the radeon driver to find the EDID at runtime by looking in the OpenFirmware device tree rather than relying on a possibly inaccurate and out of date static quirk list? I think the xresprobe utility has a file called ddcprobe/of.c that does exactly that (ie. parses the OpenFirmware device tree and extracts the EDID). Why doesn't the radeon driver just use ddcprobe/of.c? This would seem the most robust way of getting EDID on all Apple PPC computers (iMac, Emac, Clamshell, Powerbook, PowerMac, etc). I didn't realize this existed. Does the OF tree have hardcoded edids, or does it just read the edid from the device and store it in the tree? if it just reads from the device (which I suspect it does), then it's doing the same thing the driver is already doing. Most built in apple screens (ibooks, powerbooks, imacs, etc.) have an edid, it's just the eMac that seems to not. The preferred thing to do for this case is to try the normal DDC path first, and fall back to platform methods after if DDC fails. I had some code to do similar for ACPI a while ago, I should look at reviving that. There are definitely a few laptops where DDC doesn't work on LVDS but the firmware does have a vaguely plausible EDID somewhere. - ajax signature.asc Description: This is a digitally signed message part ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
Hi, --- On Tue, 12/16/08, Alex Deucher alexdeuc...@gmail.com wrote: The EDID for the built-in monitor of all PowerPC Macs can be found under /proc/device-tree. Since it's hard for anyone but Apple to know about every single type of monitor they've put into their computers, wouldn't it be better for the radeon driver to find the EDID at runtime by looking in the OpenFirmware device tree rather than relying on a possibly inaccurate and out of date static quirk list? I think the xresprobe utility has a file called ddcprobe/of.c that does exactly that (ie. parses the OpenFirmware device tree and extracts the EDID). Why doesn't the radeon driver just use ddcprobe/of.c? This would seem the most robust way of getting EDID on all Apple PPC computers (iMac, Emac, Clamshell, Powerbook, PowerMac, etc). I didn't realize this existed. Does the OF tree have hardcoded edids, or does it just read the edid from the device and store it in the tree? if it just reads from the device (which I suspect it does), then it's doing the same thing the driver is already doing. Most built in apple screens (ibooks, powerbooks, imacs, etc.) have an edid, it's just the eMac that seems to not. I'm not sure how OpenFirmware is implemented to know the EDID. On my iMac G3, the EDID is located at: /proc/device-tree/p...@f000/ATY,rage128p...@10/EDID Joseph, does your eMac have an EDID file in a similar location? I agree with ajax that X should fall back to reading the EDID from the firmware (OpenFirmware or ACPI) if DDC doesn't work. And since reading from the firmware doesn't require X.org's hardware drivers, this mechanism could be used as a fallback for _all_ drivers, not just radeon. Apple seems to have placed information about their built-in displays (including a bunch of EDIDs) in a package called AppleDisplays that comes with OpenDarwin (the latest version is AppleDisplays-160.0.9), which can be downloaded from http://src.macosforge.org/Projects . For Apple computers in particular, a second fallback could be to use the information provided by the AppleDisplays package. Thanks, Stan ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
I'm not sure how OpenFirmware is implemented to know the EDID. On my iMac G3,
the EDID is located at:
/proc/device-tree/p...@f000/ATY,rage128p...@10/EDID
Joseph, does your eMac have an EDID file in a similar location?
Yes. This is what I get on my eMac 1GHz with my external monitor plugged in:
$ find /proc/device-tree | grep -i edid
/proc/device-tree/p...@f000/ATY,moonrakerpar...@10/ATY,moonrake...@1/EDID
/proc/device-tree/p...@f000/ATY,moonrakerpar...@10/ATY,moonrake...@0/override-no-edid
/proc/device-tree/p...@f000/ATY,moonrakerpar...@10/ATY,moonrake...@0/has-static-EDID
/proc/device-tree/p...@f000/ATY,moonrakerpar...@10/ATY,moonrake...@0/EDID
@0 refers to the built-in monitor. EDID and override-no-edid are
exactly the same and 128 bytes, and has-static-EDID is the 4 bytes
{0,0,0,1}. monitor-parse-edid prints @0/EDID as:
Name: iMac
EISA ID: APP9d07
Screen size: 33.0 cm x 24.0 cm (16.06 inches, aspect ratio 4/3 or 3/2 = 1.38)
Gamma: 2.11
Analog signal
Max video bandwidth: 130 MHz
HorizSync 71-73
VertRefresh 70-140
# Monitor preferred modeline (89.0 Hz vsync, 72.1 kHz hsync, ratio 4/3)
ModeLine 1024x768 99.19 1024 1072 1168 1376 768 769 772 810 +hsync
+vsync
# Monitor preferred modeline (71.9 Hz vsync, 72.1 kHz hsync, ratio 4/3)
ModeLine 1280x960 122.24 1280 1328 1424 1696 960 961 964 1002 +hsync
+vsync
(yes, it does say iMac)
I tested these modes through xrandr and both of them worked.
I went ahead and ran `sudo find /proc | grep -i edid` on my other
OpenFirmware Mac, an iBook Clamshell running Debian Etch, but it
turned up nothing. It runs the r128 driver just fine in Debian and
Ubuntu Gutsy IIRC, but I may have had to manually add a modeline to
get it to run on Fedora 9 which used some early version of xorg 1.5.
___
xorg mailing list
xorg@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
I played around with gtf and found that my purportedly 1280x960
monitor can display higher resolutions quite nicely (no missing
pixels, apparently). Here is a list of the resolutions and refresh
rates I tested:
//These flicker pretty badly
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
//These are tolerable
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
[EMAIL PROTECTED]
//This has noticeable horizontal bars
[EMAIL PROTECTED]
Resolutions with refresh rates below 50 or so flickered noticeably
while resolutions with heights smaller than 480 tended to have
horizontal black bars. Other than that, I was able to choose just
about any resolution I wanted, but I had to adjust the refresh rate
for the image to appear and not fall over the screen. I would say the
range of tolerability of these resolutions is probably
512x384...1600x1200.
Here is a new RADEONeMacModes that includes these modes when the
'super' mode is set to nonzero and uses canonical values when 'super'
is set to 0:
static DisplayModePtr RADEONeMacModes(xf86OutputPtr output, int super)
{
ScrnInfoPtr pScrn = output-scrn;
DisplayModePtr last=NULL, new=NULL, first=NULL;
int i, *modep;
int modecount = super ? 21 : 5;
static const char *modenames[] = {
640x480, 800x600, 1024x768, 1152x864, 1280x960,
320x240, 512x384, 1280x1024, 1280x768, 1280x800,
1360x768, 1368x768, 1400x1050,
1400x900, 1600x1024, 1600x1200, 1792x1344, 1856x1392,
1920x1200, 1920x1440, 2048x1536
};
static int modes[] = {
62120, 640,680,752,864, 480,481,484,521,
76840, 800,848,936,1072, 600,601,604,640,
99070, 1024,1088,1200,1376, 768,769,772,809,
112360, 1152,1224,1352,1552, 864,865,868,905,
124540, 1280,1368,1504,1728, 960,961,964,1001,
32280, 320, 344, 376, 432, 240, 241, 244, 282,
49710, 512, 544, 600, 688, 384, 385, 388, 425,
123190, 1280, 1368, 1504, 1728, 1024, 1025, 1028, 1064,
128770, 1280, 1368, 1504, 1728, 768, 769, 772, 810,
128040, 1280, 1368, 1504, 1728, 800, 801, 804, 842,
137120, 1360, 1456, 1600, 1840, 768, 769, 772, 810,
137710, 1368, 1464, 1608, 1848, 768, 769, 772, 810,
138590, 1400, 1496, 1648, 1896, 1050, 1051, 1054, 1091,
139160, 1400, 1496, 1648, 1896, 900, 901, 904, 941,
156430, 1600, 1704, 1880, 2160, 1024, 1025, 1028, 1065,
155470, 1600, 1704, 1880, 2160, 1200, 1201, 1204, 1241,
174000, 1792, 1912, 2104, 2416, 1344, 1345, 1348, 1385,
178710, 1856, 1976, 2176, 2496, 1392, 1393, 1396, 1432,
186570, 1920, 2048, 2256, 2592, 1200, 1201, 1204, 1241,
184140, 1920, 2048, 2256, 2592, 1440, 1441, 1444, 1480,
195170, 2048, 2176, 2400, 2752, 1536, 1537, 1540, 1576
};
modep = modes;
for (i=0; imodecount; i++) {
new = xnfcalloc(1, sizeof (DisplayModeRec));
if (new) {
new-name = xnfalloc(strlen(modenames[i]) + 1);
strcpy(new-name, modenames[i]);
new-Clock = *modep++;
new-HDisplay = *modep++;
new-HSyncStart = *modep++;
new-HSyncEnd = *modep++;
new-HTotal = *modep++;
new-VDisplay = *modep++;
new-VSyncStart = *modep++;
new-VSyncEnd = *modep++;
new-VTotal = *modep++;
new-Flags = 0;
new-type = M_T_DRIVER;
if (i==2)
new-type |= M_T_PREFERRED;
new-next = NULL;
new-prev = last;
if (last) last-next = new;
last = new;
if (!first) first = new;
xf86DrvMsg(pScrn-scrnIndex, X_INFO, Added eMac mode
%s\n, modenames[i]);
}
}
return first;
}
Of course, some of these resolutions require a raised Virtual setting.
Also, the default modes that are added when Xorg starts sometimes
conflict with these modes. For instance, a [EMAIL PROTECTED] mode
overrides my [EMAIL PROTECTED] The former does not fit into the screen
quite as nicely as mine.
These extra modes (particularly the
Re: Trying to use Radeon driver on Apple eMac
On Thu, Dec 11, 2008 at 2:17 AM, Joseph Adams [EMAIL PROTECTED] wrote: My Xorg.0.log and output of xrandr (in the absence of an xorg.conf) can be found at http://www.funsitelots.com/pub/emac_almostthere . It looks to me like Xorg is adding default modelines that aren't appropriate for the internal CRT monitor. Since my LCD monitor supports 1280x1024 resolution that matches one of the default modes picked erroneously for the internal CRT monitor, Xorg picks that resolution and uses it. A 1280x1024 external/1280x960 internal configuration set up through xrandr does work fine, though. This is something of an xserver issue. Most multi-sync CRT's can support just about any common mode timings, so the xserver adds them since many users find them useful. Unfortunately the eMac is particularly picky about it's modes on the internal monitor. We may want to add am output flag the driver can use to tell the server not to add the default modes for cases like this. A workaround for the moment would be to add a filter for eMacs to radeon_mode_valid() to deny any modes that will not work on the eMac. Could get tricky though. What's happening is that the xserver is trying to set up the largest clone mode available. it's matching the 1280x1024 preferred mode from the external monitor's edid and adding the default 1280x1024 mode (from the default mode added by the server). unfortunately the internal monitor doesn't like this mode. Alex ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
Oops, forgot to send to xorg as well
-- Forwarded message --
From: Joseph Adams [EMAIL PROTECTED]
Date: Thu, Dec 11, 2008 at 12:49 AM
Subject: Re: Trying to use Radeon driver on Apple eMac
To: Alex Deucher [EMAIL PROTECTED]
This patch works when my external monitor is unplugged. When it is
plugged in, X selects the highest resolution the external monitor
supports (1280x1024) and does not turn on the internal monitor (I can
turn it on with xrandr --output VGA-1 --size 1024x768).
I did have to fix a tiny mistake in RADEONeMacNativeMode, however:
new-name = xnfalloc(strlen(1024x768) + 1);
new-name = 1024x768;
should be:
new-name = xnfalloc(strlen(1024x768) + 1);
strcpy(new-name, 1024x768);
This was causing Xorg to crash whenever xrandr was called.
I went ahead and made a function that adds all the modes for my
monitor based on and replacing RADEONeMacNativeMode, in case this
makes things easier for you:
static DisplayModePtr RADEONeMacModes(xf86OutputPtr output)
{
ScrnInfoPtr pScrn = output-scrn;
DisplayModePtr last=NULL, new=NULL, first=NULL;
int i, *modep;
static const char *modenames[5] = {
640x480, 800x600, 1024x768, 1152x864, 1280x960
};
static int modes[9*5] = {
62120, 640,680,752,864, 480,481,484,521,
76840, 800,848,936,1072, 600,601,604,640,
99070, 1024,1088,1200,1376, 768,769,772,809,
112360, 1152,1224,1352,1552, 864,865,868,905,
124540, 1280,1368,1504,1728, 960,961,964,1001
};
modep = modes;
for (i=0; i5; i++) {
new = xnfcalloc(1, sizeof (DisplayModeRec));
if (new) {
new-name = xnfalloc(strlen(modenames[i]) + 1);
strcpy(new-name, modenames[i]);
new-Clock = *modep++;
new-HDisplay = *modep++;
new-HSyncStart = *modep++;
new-HSyncEnd = *modep++;
new-HTotal = *modep++;
new-VDisplay = *modep++;
new-VSyncStart = *modep++;
new-VSyncEnd = *modep++;
new-VTotal = *modep++;
new-Flags = 0;
new-type = M_T_DRIVER;
if (i==2)
new-type |= M_T_PREFERRED;
new-next = NULL;
new-prev = last;
if (last) last-next = new;
last = new;
if (!first) first = new;
xf86DrvMsg(pScrn-scrnIndex, X_INFO, Added
eMac mode %s\n, modenames[i]);
}
}
return first;
}
I tested this on all my screen's resolutions and they all work fine.
By the way, what should be the preferred or native mode of my
display? I noticed you used 1024x768 in the patch instead of the
maximum size of 1280x960, so I mirrored that here. Is this because it
is the maximum size that allows for dual-head with the default
Virtual?
By the way, thanks for responding so quickly and helping get this
resolved :) Great maintainer, would work with again! A
___
xorg mailing list
xorg@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
My Xorg.0.log and output of xrandr (in the absence of an xorg.conf) can be found at http://www.funsitelots.com/pub/emac_almostthere . It looks to me like Xorg is adding default modelines that aren't appropriate for the internal CRT monitor. Since my LCD monitor supports 1280x1024 resolution that matches one of the default modes picked erroneously for the internal CRT monitor, Xorg picks that resolution and uses it. A 1280x1024 external/1280x960 internal configuration set up through xrandr does work fine, though. I guess the solution to this is to remove the default modelines and use the ones I posted (created by gtf) instead. I'm curious to see if my monitor actually supports those odd in-between resolutions (832x624, 720x400, 640x400, 640x635) that Mac OS X doesn't report. I'll go find out. ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
I wrote a little script and got the Xorg.0.log of every combination. In some of the tests, I heard Fedora 10 GNOME's startup sound, and other times I didn't. IIRC the snippet posted farther down is from a test where I heard the startup sound. I also got my dmesg (after rebooting) which may contain some useful information: http://www.funsitelots.com/pub/emac_connectortabletests.tar.bz2 http://www.funsitelots.com/pub/emac_dmesg I'll try to briefly post the interesting pieces: This is the Xorg.0.log of the ConnectorTable option 96,1,0,1,100,2,0,1: (WW) RADEON(0): Video BIOS not detected, using default clock settings! (II) RADEON(0): Probed PLL values: xtal: 27.00 Mhz, sclk: 261.00 Mhz, mclk: 180.00 Mhz (II) RADEON(0): PLL parameters: rf=2700 rd=12 min=12500 max=35000; xclk=36000 (II) RADEON(0): Output VGA-1 has no monitor section (II) RADEON(0): I2C bus VGA-1 initialized. (II) RADEON(0): Output VGA-0 has no monitor section (II) RADEON(0): I2C bus VGA-0 initialized. (II) RADEON(0): Port0: Monitor -- AUTO Connector -- VGA DAC Type -- Primary TMDS Type -- None DDC Type -- 0x60 (II) RADEON(0): Port1: Monitor -- AUTO Connector -- VGA DAC Type -- TVDAC/ExtDAC TMDS Type -- None DDC Type -- 0x64 (II) RADEON(0): I2C device VGA-1:E-EDID segment register registered at address 0x60. (II) RADEON(0): I2C device VGA-1:ddc2 registered at address 0xA0. (II) RADEON(0): Output: VGA-1, Detected Monitor Type: 0 (II) RADEON(0): Found color CRT connected to primary DAC finished output detect: 0 (II) RADEON(0): I2C device VGA-0:E-EDID segment register registered at address 0x60. (II) RADEON(0): I2C device VGA-0:ddc2 registered at address 0xA0. (II) RADEON(0): Output: VGA-0, Detected Monitor Type: 0 finished output detect: 1 finished all detect before xf86InitialConfiguration (II) RADEON(0): Output: VGA-1, Detected Monitor Type: 0 (II) RADEON(0): Found color CRT connected to primary DAC in RADEONProbeOutputModes The other ones look very similar, except some don't have two (II) RADEON(0): I2C device VGA-0:E-EDID segment register registered at address 0x** lines. ... SELinux: Initializing. SELinux: Starting in permissive mode (comment: I thought I disabled this during installation) ... radeonfb :00:10.0: enabling device (0086 - 0087) radeonfb (:00:10.0): Invalid ROM signature 0 should be 0xaa55 radeonfb: Retrieved PLL infos from Open Firmware radeonfb: Reference=27.00 MHz (RefDiv=12) Memory=180.00 Mhz, System=260.00 MHz radeonfb: PLL min 12000 max 35000 i2c-adapter i2c-2: unable to read EDID block. i2c-adapter i2c-2: unable to read EDID block. i2c-adapter i2c-2: unable to read EDID block. i2c-adapter i2c-3: unable to read EDID block. i2c-adapter i2c-3: unable to read EDID block. i2c-adapter i2c-3: unable to read EDID block. radeonfb: Monitor 1 type CRT found radeonfb: EDID probed radeonfb: Monitor 2 type CRT found Console: switching to colour frame buffer device 128x48 radeonfb (:00:10.0): ATI Radeon 5157 QW Generic non-volatile memory driver v1.1 Linux agpgart interface v0.103 agpgart-uninorth :00:0b.0: Apple UniNorth 1.5 chipset agpgart-uninorth :00:0b.0: configuring for size idx: 8 agpgart-uninorth :00:0b.0: AGP aperture is 32M @ 0x0 ... If you want me to boot into OF to get more information, you'll have to tell me how :) ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
On Tue, Dec 9, 2008 at 7:39 AM, Joseph Adams [EMAIL PROTECTED] wrote: I wrote a little script and got the Xorg.0.log of every combination. In some of the tests, I heard Fedora 10 GNOME's startup sound, and other times I didn't. IIRC the snippet posted farther down is from a test where I heard the startup sound. I also got my dmesg (after rebooting) which may contain some useful information: http://www.funsitelots.com/pub/emac_connectortabletests.tar.bz2 http://www.funsitelots.com/pub/emac_dmesg Logs look good. We'll have to do more testing when you get an external monitor in order to determine the ddc line used for the external port. For now it looks like the monitor doesn't like the default mode the xserver is choosing: (II) RADEON(0): Printing probed modes for output VGA-1 (II) RADEON(0): Modeline 1280x960x60.0 108.00 1280 1376 1488 1800 960 961 964 1000 +hsync +vsync (60.0 kHz) (II) RADEON(0): Modeline 1280x960x59.9 101.25 1280 1360 1488 1696 960 963 967 996 -hsync +vsync (59.7 kHz) (II) RADEON(0): Modeline 1024x768x60.0 65.00 1024 1048 1184 1344 768 771 777 806 -hsync -vsync (48.4 kHz) (II) RADEON(0): Modeline 800x600x60.3 40.00 800 840 968 1056 600 601 605 628 +hsync +vsync (37.9 kHz) (II) RADEON(0): Modeline 640x480x59.9 25.18 640 656 752 800 480 490 492 525 -hsync -vsync (31.5 kHz) (II) RADEON(0): Output: VGA-0, Detected Monitor Type: 0 (II) RADEON(0): EDID for output VGA-0 (II) RADEON(0): Output VGA-1 connected (II) RADEON(0): Output VGA-0 disconnected (II) RADEON(0): Using user preference for initial modes (II) RADEON(0): Output VGA-1 using initial mode 1280x960 Does it work if you switch to a different mode with xrandr, e.g., xrandr --output VGA-1 --mode 1024x768 or adding your mode: xrandr --newmode 1024x768j 99.198 1024 1072 1168 1376 768 769 772 810 +HSync +VSync xrandr --addmode VGA-1 1024x768j xrandr --output VGA-1 --mode 1024x768j When we find a mode you want to use you can make it your preferred mode in your xorg.conf. Make sure to set your monitor Identifier to VGA-1 or associate it with whatever your monitor section is called Option Monitor-VGA-1 my monitor name. see this page for more info on xrandr config: http://wiki.debian.org/XStrikeForce/HowToRandR12 Make sure you test with one of the ConnectorTable options I suggested. Alex ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Trying to use Radeon driver on Apple eMac
Yay! I modified my xorg.conf to add the modelines and use the ConnectorTable, and it displays fine (though there is a bit of flickering whenever xrandr is called) See http://www.funsitelots.com/pub/xorg_emacG4_working for logs, screenshots, etc. My xorg.conf is as follows: # Xorg configuration created by system-config-display Section ServerLayout Identifier single head configuration Screen 0 Screen0 0 0 InputDeviceKeyboard0 CoreKeyboard EndSection Section InputDevice Identifier Keyboard0 Driver kbd Option XkbModel pc105 Option XkbLayout us Option XkbVariant dvorak EndSection Section Device Identifier Videocard0 Driver radeon #Driver fbdev #Option UseFBDev true Option ConnectorTable 96,1,0,1,100,2,0,1 #Option MacModel powerbook-internal EndSection #Both of these xrandr modes worked: #xrandr --newmode 1024x768j 99.198 1024 1072 1168 1376 768 769 772 810 +HSync +VSync #xrandr --newmode 1280x960j 124.54 1280 1368 1504 1728 960 961 964 1001 +HSync +Vsync # $ gtf 1280 960 72 # 1280x960 @ 72.00 Hz (GTF) hsync: 72.07 kHz; pclk: 124.54 MHz # Modeline 1280x960_72.00 124.54 1280 1368 1504 1728 960 961 964 1001 -HSync +Vsync # (note that -HSync should be +HSync, or the image is about 70% over to the left (according to my test) Section Monitor Identifier eMac Monitor #HorizSync 71-73 Modeline 1280x960 124.54 1280 1368 1504 1728 960 961 964 1001 +HSync +VSync EndSection Section Screen Identifier Screen0 Device Videocard0 MonitoreMac Monitor DefaultDepth 24 SubSection Display Viewport 0 0 Depth 24 EndSubSection EndSection I went ahead and tested 3D and video playing. 3D works fine until I VT out and back in (then 3D programs like glxgears and neverball lock up). Playing an ogg vorbis through Epiphany works fine. Playing an flv or an ogg in mplayer displays a garbled image (see screenshots). Video playing does not seem to be affected by VT switching. Although these are likely separate issues, I just wanted to bring them to your attention in case they're related to this problem. I guess now we have to figure out how to get Xorg to autoprobe the correct modes? ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Trying to use Radeon driver on Apple eMac
I'm struggling to get the Radeon driver to work on my eMac G4's Radeon 7500 card (the fbdev driver works fine at 1024x768 with no acceleration). On both Fedora 8 and Fedora 10 with default xorg.conf, the screen is blank (the CRT doesn't glow or anything). I have a CRT monitor that supports up to 1280x960. The eMac also has a video out port. I suspect that video may show up on an external monitor, but I can't test that until my adapter arrives. With an unmodified configuration, Xorg accepts clients and the /var/log/Xorg.0.log looks good. However, the screen is blank. I suspect this may be bogus modelines. I tried a modeline I grabbed from fbset in the working radeonfb Linux terminal: $ fbset -x Mode 1024x768 # D: 99.197 MHz, H: 72.090 kHz, V: 89.001 Hz DotClock 99.198 HTimings 1024 1072 1168 1376 VTimings 768 769 772 810 Flags+HSync +VSync EndMode I tried adding this mode through xrandr as follows: $ xrandr --newmode 1024x768j 99.198 1024 1072 1168 1376 768 769 772 810 +HSync +VSync $ xrandr --addmode DVI-0 1024x768j $ xrandr Screen 0: minimum 320 x 200, current 1024 x 768, maximum 2048 x 1024 DVI-1 disconnected (normal left inverted right x axis y axis) DVI-0 connected 1024x768+0+0 (normal left inverted right x axis y axis) 0mm x 0mm 1024x768 59.9*+ 800x60060.3 640x48059.9 1024x768j 89.0 S-video disconnected (normal left inverted right x axis y axis) $ xrandr --output DVI-0 --mode 1024x768j xrandr: Configure crtc 0 failed Just trying to xrandr to a different resolution fails: $ xrandr -s 800x600 Failed to change the screen configuration! And VT-switching to 7 (where Xorg is running) results in Xorg segfaulting whether I do any of the above or not: xauth: creating new authority file /home/joey/.serverauth.19680 X.Org X Server 1.5.3 Release Date: 5 November 2008 X Protocol Version 11, Revision 0 Build Operating System: Linux 2.6.18-92.1.10.el5 ppc Current Operating System: Linux cello 2.6.27.5-117.fc10.ppc #1 Tue Nov 18 11:50:30 EST 2008 ppc Build Date: 16 November 2008 08:30:07PM Build ID: xorg-x11-server 1.5.3-5.fc10 Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. Markers: (--) probed, (**) from config file, (==) default setting, (++) from command line, (!!) notice, (II) informational, (WW) warning, (EE) error, (NI) not implemented, (??) unknown. (==) Log file: /var/log/Xorg.0.log, Time: Mon Dec 8 13:33:24 2008 (==) Using config file: /etc/X11/xorg.conf ** snip ** Leaving Restore TV ** Message: another SSH agent is running at: /tmp/ssh-igRxa19911/agent.19911 Window manager warning: Failed to read saved session file /home/joey/.config/metacity/sessions/1042b409dc9a775ef712287612087368560199110021.ms: Failed to open file '/home/joey/.config/metacity/sessions/1042b409dc9a775ef712287612087368560199110021.ms': No such file or directory Backtrace: 0: /usr/bin/X(xorg_backtrace+0x50) [0x100ffb90] 1: /usr/bin/X(xf86SigHandler+0x68) [0x10082ec8] 2: [0x100364] 3: [0x48002000] 4: /usr/lib/xorg/modules/drivers//radeon_drv.so(radeon_update_dri_buffers+0x8c) [0xe815ffc] 5: /usr/lib/xorg/modules/drivers//radeon_drv.so(RADEONEnterVT+0xb4) [0xe7e4d94] 6: /usr/lib/xorg/modules//libxaa.so [0xe6e2dc8] 7: /usr/bin/X [0x1009cf70] 8: /usr/bin/X [0x1008cf68] 9: /usr/bin/X(xf86Wakeup+0x4dc) [0x10083e2c] 10: /usr/bin/X(WakeupHandler+0x84) [0x10045d44] 11: /usr/bin/X(WaitForSomething+0x210) [0x100fc570] 12: /usr/bin/X(Dispatch+0xf8) [0x100412e8] 13: /usr/bin/X(main+0x43c) [0x1002261c] 14: /lib/libc.so.6 [0xfe0ebc4] 15: /lib/libc.so.6 [0xfe0ed80] Fatal server error: Caught signal 11. Server aborting ** snip ** waiting for X server to shut down I tried adding my magic modeline via xorg.conf, but Xorg doesn't seem to be taking it. Here is my xorg.conf: # Xorg configuration created by system-config-display Section ServerLayout Identifier single head configuration Screen 0 Screen0 0 0 InputDeviceKeyboard0 CoreKeyboard EndSection Section InputDevice Identifier Keyboard0 Driver kbd Option XkbModel pc105 Option XkbLayout us Option XkbVariant dvorak EndSection Section Device Identifier Videocard0 Driver radeon #Option Accel off Option AccelMethod XAA Option AccelDFS0 Option AGPMode 4 Option AGPFastWrite 0 Option GARTSize 64 Option EnablePageFlip 1 Option ColorTiling 1 #Option TVDACLoadDetect TRUE #Option TVStandard ntsc #Option monitor-S-video TV-monitor #Option DPMS 0 #Option UseFBDev true EndSection Section Monitor Identifier eMac Monitor
Re: Trying to use Radeon driver on Apple eMac
On Mon, Dec 8, 2008 at 2:01 PM, Joseph Adams [EMAIL PROTECTED] wrote: I'm struggling to get the Radeon driver to work on my eMac G4's Radeon 7500 card (the fbdev driver works fine at 1024x768 with no acceleration). On both Fedora 8 and Fedora 10 with default xorg.conf, the screen is blank (the CRT doesn't glow or anything). I have a CRT monitor that supports up to 1280x960. The eMac also has a video out port. I suspect that video may show up on an external monitor, but I can't test that until my adapter arrives. Mac cards don't have a connector table and are always wired up strangely. We'll probably need to add a custom connector table for your card as it's currently falling back to the default which is probably wrong for your card. Can you send me the output of /proc/cpuinfo? Try the following option and let me know which combinations work: Option ConnectorTable i,1,0,1,j,2,0,1 where i and j are one of the following: 96, 100, 104, or 108 and i != j. i and j specify the gpio line used for ddc, so when you get the right combination, you should see the driver picking up an edid if your monitor supports it. If the built in monitor doesn't supply an edid, we'll have to add a quirk for that as well. Alex ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
