Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Sat, Aug 09, 2008 at 09:41:22AM +0200, Brice Goglin wrote: > This time, this is very likely to be fixed upstream for real. I just wanted to confirm that VT switching works again. Thank you. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
forwarded 435040 https://bugs.freedesktop.org/show_bug.cgi?id=13994 tags 435040 +fixed-upstream thank you On Sat, Jul 28, 2007 at 10:29:29PM +0300, Kari Pahula wrote: > Package: xserver-xorg-video-ati > Version: 1:6.6.192-1 > Severity: important > > Changing the VT from X to console and back to X locks up the computer > completely, unless I'm using the console framebuffer device. With > console fbdev, just the screen freezes and I can still ssh to it or type > blindly. This time, this is very likely to be fixed upstream for real. https://bugs.freedesktop.org/show_bug.cgi?id=13994 http://gitweb.freedesktop.org/?p=xorg/driver/xf86-video-ati.git;a=commit;h=268c848130ec1770bb645a74197b6aca7fc95abc Brice -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, Aug 06, 2007 at 05:56:44PM +0300, Kari Pahula wrote:
> 1 write(0, "(II) RADEON(0): RADEONRestoreMem"..., 50) = 50
> 1 write(0, "(II) RADEON(0): MC_FB_LOCATION"..., 48) = 48
> 1 write(0, "(II) RADEON(0): MC_AGP_LOCATIO"..., 48) = 48
> 199 nanosleep({0, 1000}, NULL) = 0
Sounds like the CRTC2 wait-for-vblank disaster, which is fixed in git.
Cheers,
Daniel
signature.asc
Description: Digital signature
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, 2007-08-06 at 18:36 +0300, Daniel Stone wrote:
> On Mon, Aug 06, 2007 at 05:56:44PM +0300, Kari Pahula wrote:
> > 1 write(0, "(II) RADEON(0): RADEONRestoreMem"..., 50) = 50
> > 1 write(0, "(II) RADEON(0): MC_FB_LOCATION"..., 48) = 48
> > 1 write(0, "(II) RADEON(0): MC_AGP_LOCATIO"..., 48) = 48
> > 199 nanosleep({0, 1000}, NULL) = 0
>
> Sounds like the CRTC2 wait-for-vblank disaster, which is fixed in git.
If it was, it would be fixed in 6.6.193 as well.
--
Earthling Michel Dänzer | http://tungstengraphics.com
Libre software enthusiast | Debian, X and DRI developer
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, Aug 06, 2007 at 11:48:14AM +0200, Michel Dänzer wrote:
> On Mon, 2007-08-06 at 11:43 +0300, Kari Pahula wrote:
> > There's a delay of more than 15 seconds (mostly calling FBIOPUTCMAP
> > ioctl and gettimeofday, apparently) after the V_BIOS entry and before
> > X is usable, but at least the VT switches without any screen freezes
> > with int10. But those delays are a separate issue altogether.
>
> Does not using Option "UseFBDev" make any difference for this?
Commenting out the UseFBDev option made a difference, in that it made
starting X faster. But it also brought the VT switching freeze back,
and this time it would lock down the kernel and not just the screen.
Doesn't respond to pings or anything. I still had the int10 module
loaded.
I did some debugging and noticed that running X with
nice -n -10 strace X 2>stracedump
made the freezes go away... If I took away the nice command with the
negative priority, the system would freeze. Looks like X really likes
to have a bit longer delay at some point. I suppose the delay(s)
caused by FBDev could cause the same thing. I know that I'm happily
making only guesses about this... I don't know why it would behave
like it does.
Some of the time X didn't lock up the system even without that nice
thing I did above, but at least I think I could identify where the
system gets locked up.
This time I didn't even see anything as distinctive in the strace dump
as I did last time... I attached the output of
tail -n 400 ~/stracedump | uniq -c
This is from a run where strace wasn't niced and I got the freeze.
With a niced strace, the part where the freeze apparently happens goes
like this:
ioctl(6, VIDIOC_S_FMT or VT_RELDISP, 0x1) = 0
shutdown(5, 2 /* send and receive */) = 0
close(5)= 0
iopl(0) = 0
ioperm(0, 0x400, 0) = 0
gettimeofday({1186408789, 97309}, NULL) = 0
gettimeofday({1186408789, 97409}, NULL) = 0
select(256, [1 3 4], NULL, NULL, {111, 851000}) = ? ERESTARTNOHAND (To be restar
ted)
--- SIGUSR1 (User defined signal 1) @ 0 (0) ---
rt_sigaction(SIGUSR1, {0x80b0f60, [USR1], SA_RESTART}, {0x80b0f60, [USR1], SA_RE
START}, 8) = 0
sigreturn() = ? (mask now [IO])
ioctl(6, VIDIOC_S_FMT or VT_RELDISP, 0x2) = 0
rt_sigprocmask(SIG_BLOCK, [IO], [IO], 8) = 0
socket(PF_FILE, SOCK_STREAM, 0) = 5
connect(5, {sa_family=AF_FILE, path="/var/run/acpid.socket"}, 110) = 0
write(0, "(II) Open ACPI successful (/var/"..., 50) = 50
1 read(10, "9-9\nDejaVuSansCondensed.ttf -dej"..., 4096) = 4096
1 read(10, "0-iso8859-2\nDejaVuSansMono.ttf -"..., 4096) = 4096
1 read(10, " -dejavu-dejavu serif-medium-o-c"..., 4096) = 786
2 read(10, "", 4096) = 0
1 close(10) = 0
1 munmap(0xb7f72000, 4096)= 0
1 open("/var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType/fonts.alias",
O_RDONLY) = 10
2 fstat64(10, {st_mode=S_IFREG|0644, st_size=0, ...}) = 0
1 mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1,
0) = 0xb7f72000
1 read(10, "", 4096) = 0
1 close(10) = 0
1 munmap(0xb7f72000, 4096)= 0
1 open("/usr/share/fonts/X11/misc/6x13-ISO8859-1.pcf.gz", O_RDONLY) = 10
1 read(10, "\37\213\10\0I4\246F\0\3\355\\it\33\327u\376\260I\264\265"...,
8192) = 4637
1 read(10, "", 8192) = 0
1 close(10) = 0
1 open("/usr/share/fonts/X11/misc/cursor.pcf.gz", O_RDONLY) = 10
1 read(10, "\37\213\10\0:4\246F\0\3\355\233{p\335\305u\307\217\356"...,
8192) = 5225
1 read(10, "", 8192) = 0
1 brk(0x83a6000) = 0x83a6000
1 close(10) = 0
1 rt_sigprocmask(SIG_BLOCK, [IO], [], 8) = 0
1 rt_sigprocmask(SIG_UNBLOCK, [IO], NULL, 8) = 0
1 gettimeofday({1186401940, 291870}, NULL) = 0
1 gettimeofday({1186401940, 291958}, NULL) = 0
1 gettimeofday({1186401940, 292053}, NULL) = 0
1 gettimeofday({1186401940, 292151}, NULL) = 0
1 select(256, [1 3 4 5 6], NULL, NULL, {600, 0}) = 1 (in [6], left {594,
704000})
1 rt_sigprocmask(SIG_BLOCK, [IO], [], 8) = 0
1 read(6, "\35", 64) = 1
1 gettimeofday({1186401945, 589552}, NULL) = 0
1 rt_sigprocmask(SIG_BLOCK, [], [IO], 8) = 0
1 rt_sigprocmask(SIG_UNBLOCK, [IO], NULL, 8) = 0
1 setitimer(ITIMER_REAL, {it_interval={0, 2}, it_value={0, 2}},
NULL) = 0
1 gettimeofday({1186401945, 589917}, NULL) = 0
1 select(256, [1 3 4 5 6], NULL, NULL, {594, 703000}) = 1 (in [6], left
{594, 696000})
1 rt_sigprocmask(SIG_BLOCK, [IO], [], 8) = 0
1 read(6, "8", 64)= 1
1 gettimeofday({1186401945, 595499}, NULL) = 0
1 rt_sigprocmas
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, 2007-08-06 at 11:43 +0300, Kari Pahula wrote: > On Mon, Aug 06, 2007 at 10:14:35AM +0200, Michel Dänzer wrote: > > Does explicitly loading the "int10" module in the xorg.conf section > > "Module" work around it? > > Yeah, that seemed to fix it. This is what I see in the log after > switching the VT to X: > > (II) Open ACPI successful (/var/run/acpid.socket) > (WW) RADEON(0): zero MEMSIZE, probably at D3cold. Re-POSTing via int10. > (II) RADEON(0): Primary V_BIOS segment is: 0xc000 > (II) Configured Mouse: ps2EnableDataReporting: succeeded So apparently the driver doesn't make sure the module is loaded before calling its function. > There's a delay of more than 15 seconds (mostly calling FBIOPUTCMAP > ioctl and gettimeofday, apparently) after the V_BIOS entry and before > X is usable, but at least the VT switches without any screen freezes > with int10. But those delays are a separate issue altogether. Does not using Option "UseFBDev" make any difference for this? -- Earthling Michel Dänzer | http://tungstengraphics.com Libre software enthusiast | Debian, X and DRI developer
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, Aug 06, 2007 at 10:14:35AM +0200, Michel Dänzer wrote: > Does explicitly loading the "int10" module in the xorg.conf section > "Module" work around it? Yeah, that seemed to fix it. This is what I see in the log after switching the VT to X: (II) Open ACPI successful (/var/run/acpid.socket) (WW) RADEON(0): zero MEMSIZE, probably at D3cold. Re-POSTing via int10. (II) RADEON(0): Primary V_BIOS segment is: 0xc000 (II) Configured Mouse: ps2EnableDataReporting: succeeded There's a delay of more than 15 seconds (mostly calling FBIOPUTCMAP ioctl and gettimeofday, apparently) after the V_BIOS entry and before X is usable, but at least the VT switches without any screen freezes with int10. But those delays are a separate issue altogether. Thanks.
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, 2007-08-06 at 05:29 +0300, Kari Pahula wrote:
>
> (II) Configured Mouse: ps2EnableDataReporting: succeeded
> (II) Open ACPI successful (/var/run/acpid.socket)
> (WW) RADEON(0): zero MEMSIZE, probably at D3cold. Re-POSTing via int10.
>
> Where the last two entries get into the log when I switch the VT.
[...]
> I ran /usr/bin/X with strace. The whole dump would be bigger than
> what I'd care to attach, but I saw this (relevant?) line at the end of
> it.
>
> writev(2, [{"X", 1}, {": ", 2}, {"symbol lookup error", 19}, {": ", 2},
> {"/usr/lib/xorg/modules/drivers//r"..., 44}, {": ", 2}, {"undefined symbol:
> xf86InitInt10", 31}, {"", 0}, {"", 0}, {"\n", 1}], 10X: symbol lookup error:
> /usr/lib/xorg/modules/drivers//radeon_drv.so: undefined symbol: xf86InitInt10
> ) = 102
Quite relevant indeed (in general, if the log file ends abruptly, the X
server's stderr output should be checked).
Does explicitly loading the "int10" module in the xorg.conf section
"Module" work around it?
--
Earthling Michel Dänzer | http://tungstengraphics.com
Libre software enthusiast | Debian, X and DRI developer
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
On Mon, Aug 06, 2007 at 12:03:41AM +0200, Brice Goglin wrote:
> Kari Pahula wrote:
> > Package: xserver-xorg-video-ati
> > Version: 1:6.6.192-1
> > Severity: important
> >
> > Changing the VT from X to console and back to X locks up the computer
> > completely, unless I'm using the console framebuffer device. With
> > console fbdev, just the screen freezes and I can still ssh to it or type
> > blindly.
> >
>
> Does 1:6.6.193-1 (in experimental now) help?
No, I'm still seeing this bug. With the new version, the last few
lines of Xorg.0.log become
(II) Configured Mouse: ps2EnableDataReporting: succeeded
(II) Open ACPI successful (/var/run/acpid.socket)
(WW) RADEON(0): zero MEMSIZE, probably at D3cold. Re-POSTing via int10.
Where the last two entries get into the log when I switch the VT.
I debugged this issue a bit further. Apparently switching the VT
kills the X process and leaves the screen frozen. Restarting X via
ssh unfreezes the screen (not that surprising really, IMHO). I tried
VT switching again after disabling ACPI support from the kernel, which
didn't help any. Only the log messages changed a bit.
(II) Configured Mouse: ps2EnableDataReporting: succeeded
(WW) Open ACPI failed (/var/run/acpid.socket) (No such file or directory)
(II) No APM support in BIOS or kernel
(WW) RADEON(0): zero MEMSIZE, probably at D3cold. Re-POSTing via int10.
I ran /usr/bin/X with strace. The whole dump would be bigger than
what I'd care to attach, but I saw this (relevant?) line at the end of
it.
writev(2, [{"X", 1}, {": ", 2}, {"symbol lookup error", 19}, {": ", 2},
{"/usr/lib/xorg/modules/drivers//r"..., 44}, {": ", 2}, {"undefined symbol:
xf86InitInt10", 31}, {"", 0}, {"", 0}, {"\n", 1}], 10X: symbol lookup error:
/usr/lib/xorg/modules/drivers//radeon_drv.so: undefined symbol: xf86InitInt10
) = 102
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
Kari Pahula wrote: > Package: xserver-xorg-video-ati > Version: 1:6.6.192-1 > Severity: important > > Changing the VT from X to console and back to X locks up the computer > completely, unless I'm using the console framebuffer device. With > console fbdev, just the screen freezes and I can still ssh to it or type > blindly. > Does 1:6.6.193-1 (in experimental now) help? Brice -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#435040: xserver-xorg-video-ati: switching VT from console to X freezes computer with X22 (Radeon Mobility M6 LY)
Package: xserver-xorg-video-ati Version: 1:6.6.192-1 Severity: important Changing the VT from X to console and back to X locks up the computer completely, unless I'm using the console framebuffer device. With console fbdev, just the screen freezes and I can still ssh to it or type blindly. -- Package-specific info: Contents of /var/lib/x11/X.roster: xserver-xorg /etc/X11/X target does not match checksum in /var/lib/x11/X.md5sum. X server symlink status: lrwxrwxrwx 1 root root 13 Apr 13 2006 /etc/X11/X -> /usr/bin/Xorg -rwxr-xr-x 1 root root 1736504 Jul 14 18:30 /usr/bin/Xorg Contents of /var/lib/x11/xorg.conf.roster: xserver-xorg VGA-compatible devices on PCI bus: 01:00.0 VGA compatible controller: ATI Technologies Inc Radeon Mobility M6 LY /etc/X11/xorg.conf does not match checksum in /var/lib/x11/xorg.conf.md5sum. Xorg X server configuration file status: -rw-r--r-- 1 root root 2251 Jul 28 19:11 /etc/X11/xorg.conf Contents of /etc/X11/xorg.conf: # xorg.conf (xorg X Window System server configuration file) # # This file was generated by dexconf, the Debian X Configuration tool, using # values from the debconf database. # # Edit this file with caution, and see the xorg.conf manual page. # (Type "man xorg.conf" at the shell prompt.) # # This file is automatically updated on xserver-xorg package upgrades *only* # if it has not been modified since the last upgrade of the xserver-xorg # package. # # If you have edited this file but would like it to be automatically updated # again, run the following command: # sudo dpkg-reconfigure -phigh xserver-xorg Section "Files" EndSection Section "InputDevice" Identifier "Generic Keyboard" Driver "kbd" Option "CoreKeyboard" Option "XkbRules" "xorg" Option "XkbModel" "pc105" Option "XkbLayout" "fi" EndSection Section "InputDevice" Identifier "Configured Mouse" Driver "mouse" Option "CorePointer" Option "Device""/dev/input/mice" Option "Protocol" "ImPS/2" EndSection Section "InputDevice" Identifier "Synaptics Touchpad" Driver "synaptics" Option "SendCoreEvents""true" Option "Device""/dev/psaux" Option "Protocol" "auto-dev" Option "HorizScrollDelta" "0" EndSection Section "Device" Identifier "ATI Technologies Inc Radeon Mobility M6 LY" # Driver "fbdev" Driver "ati" BusID "PCI:1:0:0" Option "UseFBDev" "true" EndSection Section "Monitor" Identifier "Generic Monitor" Option "DPMS" HorizSync 28-51 VertRefresh 43-60 EndSection Section "Screen" Identifier "Default Screen" Device "ATI Technologies Inc Radeon Mobility M6 LY" Monitor "Generic Monitor" DefaultDepth24 SubSection "Display" Depth 1 Modes "1024x768" EndSubSection SubSection "Display" Depth 4 Modes "1024x768" EndSubSection SubSection "Display" Depth 8 Modes "1024x768" EndSubSection SubSection "Display" Depth 15 Modes "1024x768" EndSubSection SubSection "Display" Depth 16 Modes "1024x768" EndSubSection SubSection "Display" Depth 24 Modes "1024x768" EndSubSection EndSection Section "ServerLayout" Identifier "Default Layout" Screen "Default Screen" InputDevice "Generic Keyboard" InputDevice "Configured Mouse" InputDevice "Synaptics Touchpad" EndSection Xorg X server log files on system: -rw-r--r-- 1 root root 38859 Jul 28 19:14 /var/log/Xorg.0.log Contents of most recent Xorg X server log file /var/log/Xorg.0.log: X Window System Version 1.3.0 Release Date: 19 April 2007 X Protocol Version 11, Revision 0, Release 1.3 Build Operating System: Linux Debian Current Operating System: Linux sininen 2.6.21.4 #2 Sat Jun 9 01:44:55 EEST 2007 i686 Build Date: 14 July 2007 Before reporting problems, check http://wiki.x.org to make sure that you have the latest version. Module Loader present 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: Sat Jul 28 19:14:32 2007 (==) Using co
