Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Integrated Graphics. starx returns no screen. ?
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2, 3, 4  Next  
Reply to topic    Gentoo Forums Forum Index Desktop Environments
View previous topic :: View next topic  
Author Message
DONAHUE
Watchman
Watchman


Joined: 09 Dec 2006
Posts: 7550
Location: Goose Creek SC

PostPosted: Fri Jun 05, 2015 12:32 am    Post subject: Reply with quote

First eliminate /etc/X11/xorg.conf best chance is automatic script
second check the kernel config in use is the one Neddy approved
third try startx again
capture the screen above and including "Caught signal 4 (illegal instruction). Server aborting" camera, cell phone, whatever
Code:
emerge --info | wgetpaste
wgetpaste /var/log/Xorg.0.log
zcat /proc/config.gz | wgetpaste
wgetpaste /proc/cpuinfo
and post the url's returned
_________________
Defund the FCC.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Fri Jun 05, 2015 6:49 am    Post subject: Reply with quote

I messed with a few kernel options to see if it might help:
I tried to slim it down. There were open menus for other driver support that I eliminated just in case.
EFI-based framebiffer support disabled
(reading up on it, it disables VESA AGP support that goes hand and hand with SiS.)
PCI resource reallocation detection enable

after make menuconfig and reboot...
Same error.

Xorg.0.log: https://bpaste.net/show/30f34445a494
emerge --info https://bpaste.net/show/49129f73b8e1
zcat /proc/config.gz https://bpaste.net/show/ec29d0a4746b
/proc/cpuinfo https://bpaste.net/show/d1ff25e2166c

http://i.imgur.com/6jaknst.png
Hope this helps.

Thanks a bunch,
Jay
Back to top
View user's profile Send private message
John R. Graham
Administrator
Administrator


Joined: 08 Mar 2005
Posts: 10156
Location: Somewhere over Atlanta, Georgia

PostPosted: Fri Jun 05, 2015 1:52 pm    Post subject: Reply with quote

jwm224 wrote:
...after make menuconfig and reboot...
Same error.
Did you merely do "make menuconfig" and not recompile kernel & modules, and then install the new kernel and modules? If so then this is not a valid test.

- John
_________________
I can confirm that I have received between 0 and 499 National Security Letters.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Fri Jun 05, 2015 2:08 pm    Post subject: Reply with quote

After it was finished I ran
Code:
make modules_install && make install


I need to run either

Code:
emerge -ev @world


or

break it up again as you had me do last time (I had to edit that I got confused for a moment):

Code:
emerge -1v linux-headers glibc gcc
emerge -ev @ system
emerge -ev @ selected


again, don't I?
I was thinking I might have to do so.

Jay
Back to top
View user's profile Send private message
John R. Graham
Administrator
Administrator


Joined: 08 Mar 2005
Posts: 10156
Location: Somewhere over Atlanta, Georgia

PostPosted: Fri Jun 05, 2015 2:32 pm    Post subject: Reply with quote

Emerge doesn't recompile the kernel but, then again, the kernel compile process doesn't use the CFLAGS from make.conf so I didn't give you any instructions for the kernel.

You're missing a step. To build the kernel after changes made to the kernel configuration, you must run
Code:
make
from the /usr/src/linux directory. After that, the two commands you cited will install the modules and the kernel binary.

I don't think you need to recompile everything at this point, so no "emerge -e" for now.

- John
_________________
I can confirm that I have received between 0 and 499 National Security Letters.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Fri Jun 05, 2015 2:52 pm    Post subject: Reply with quote

I ran
Code:
 make
from /usr/src/linux

I had my last two kernel configuration files in that directory:
Code:
jay2.config
jay3.config
Does that affect anything?
I'm still getting the same results.

I have an appointment. I should be back in about an hour or so. I'll read up more on rebuilding the kernel. I ran 'make' last time too. I remember it prompting me to.

Did you find anything in any of that information I sent out, or did that all look normal, aside from the startx error?
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 42581
Location: 56N 3W

PostPosted: Fri Jun 05, 2015 9:41 pm    Post subject: Reply with quote

jwm224,

Code:
[    77.094] (EE) Backtrace:
[    77.094] (EE) 0: /usr/bin/X (xorg_backtrace+0x4f) [0x81e8ebf]
[    77.094] (EE) 1: /usr/bin/X (0x8048000+0x1a5126) [0x81ed126]
[    77.094] (EE) 2: /lib/libc.so.6 (0xb72b3000+0x2e080) [0xb72e1080]
[    77.094] (EE) 3: /usr/lib/libXfont.so.1 (FontParseXLFDName+0x277) [0xb74ca7b7]
[    77.094] (EE) 4: /usr/lib/libXfont.so.1 (FontFileAddFontFile+0x3d7) [0xb74a83f7]
[    77.094] (EE) 5: /usr/lib/libXfont.so.1 (FontFileReadDirectory+0x291) [0xb74a6881]
[    77.094] (EE) 6: /usr/lib/libXfont.so.1 (FontFileInitFPE+0x2d) [0xb74a888d]
[    77.094] (EE) 7: /usr/bin/X (0x8048000+0x33de4) [0x807bde4]
[    77.095] (EE) 8: /usr/bin/X (SetDefaultFontPath+0x126) [0x807da36]
[    77.095] (EE) 9: /usr/bin/X (0x8048000+0x36346) [0x807e346]
[    77.095] (EE) 10: /usr/bin/X (0x8048000+0x1f77e) [0x806777e]
[    77.095] (EE) 11: /lib/libc.so.6 (__libc_start_main+0xf3) [0xb72cc923]
[    77.095] (EE) 12: /usr/bin/X (0x8048000+0x1f7a4) [0x80677a4]
[    77.095] (EE)
[    77.095] (EE) Illegal instruction at address 0xb74ca7b7
Points to /usr/lib/libXfont.so.1 being built with the wrong CFLAGS for your CPU

Code:
$ equery b /usr/lib/libXfont.so.1
 * Searching for /usr/lib/libXfont.so.1 ...
x11-libs/libXfont-1.5.1 (/usr/lib64/libXfont.so.1.4.1)
tells that that library belongs to x11-libs/libXfont.
Code:
emerge -1 x11-libs/libXfont
is worth a try after you have fixed your CFLAGS
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Sat Jun 06, 2015 2:30 am    Post subject: Reply with quote

I've been reading into cflags for some time now with little understanding of what changes to make.

I have:
Code:
CFLAGS= "-march=native -O2 -pipe


Should this be changed to the default:

Code:
CFLAGS="-O2 -pipe"

or
Code:
CFLAGS="-march=native-O1 -pipe"


I know last time this decision cost me a 20 hour recompile.
Also, is this a good time to add the line:
ACCEPT_LICENSE="*@EULA"
I figured this wouldn't hurt, but better safe than sorry.

Thanks too Needy.
Back to top
View user's profile Send private message
DONAHUE
Watchman
Watchman


Joined: 09 Dec 2006
Posts: 7550
Location: Goose Creek SC

PostPosted: Sat Jun 06, 2015 3:37 am    Post subject: Reply with quote

Code:
gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 | wgetpaste -t
and post url; should show what cflags -march=native causes gcc to use
Code:
cat /proc/cpuinfo
should show what flags the cpu supports, previously provided as /proc/cpuinfo https://bpaste.net/show/d1ff25e2166c
Suggest comparing the lists
your emerge --info in pastebin has CPU_FLAGS_X86="mmx sse sse2" . Those flags are supported by your cpu. Running
Code:
emerge cpuinfo2cpuflags
cpuinfo2cpuflags-x86
should provide a list derived from your cpuinfo which may be more extensive and which can be inserted in /etc/portage/make.conf.
Your cflags, CFLAGS="-march=native -O2 -pipe" seem right to my untrained eye.
re-emerging libXfont is a good idea
sempron 3000 had 32 bit, k7, socket 462 and 64 bit, k8, socket 754 versions; your cpuinfo suggests 64 bit, k8, socket 754 version in which case I believe -march=k8 would be an alternate to -march=native which would allow more optimization than not specifying -march.
_________________
Defund the FCC.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Sat Jun 06, 2015 5:40 am    Post subject: Reply with quote

Okay. Before I do all that, I went back into make menuconfig from /usr/src/linux and made these changes. They seemed, from what I read to be accurate, but experience goes along way so you'd know better. I did all this before I read your post:

    Low address space to protect from user allocation --- changed from 4096 to 65536
    processor family (586/k5/5x86/6x86/6x86MX) --- changed from pentium pro
    Symmetric muti-processor support --- disabled
    Enable MPS table --- disabled
    Memtest --- enabled
    High Memory Support --- changed from 4G to off


After make menuconfig, I ran make and rebooted.

gcc -march=native -E -v - </dev/null 2>&1 | grep cc1 | wgetpaste -t
https://bpaste.net/show/5d9c6f474417

cat /proc/cpuinfo
https://bpaste.net/show/5486205c253d

Code:
emerge cpuinfo2cpuflags
cpuinfo2cpuflags-x86
CPU_FLAGS_X86="3dnow cdnowext mmx mmxext sse sse2"

Yeah, that added new flags. Cool. I added them to /etc/portage/make.conf

Should I change -march=native to -march=k8...?
Thanks a bunch,
Jay
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Sat Jun 06, 2015 5:47 am    Post subject: Reply with quote

the only difference in the /proc/cpuinfo from before and now is a slight change in these to readings:

cpu MHz : 1808.350
to
cpu MHz : 1808.362

and,

bogomips : 3616.70
to
bogomips : 3616.72

I looked them both over thoroughly.
Jay
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Sat Jun 06, 2015 6:14 am    Post subject: Reply with quote

I ran:
Code:
equery b /usr/lib/libXfont.so.1

I get:

    x11-libs/libXfont-1.5.1 (/usr/lib/libXfont.so.1 -> libXfont.so.1.4.1)
    x11-libs/libXfont-1.5.1 (/usr/lib/libXfont.so.1.4.1)


I ran:
Code:
 emerge -1 x11-libs/libXfont


I rebooted.
I tried startx.
I think we may have got a little further. I don't get the same error I was getting. Now it's about the keyboard...
Here's
Code:
 cat /var/log/Xorg.0.log

https://bpaste.net/show/61ea3d899898

Thanks,
Jay
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 42581
Location: 56N 3W

PostPosted: Sat Jun 06, 2015 7:51 am    Post subject: Reply with quote

jwm224,

Your cpu MHz and bogomips variations are due to the operating frequency of the crystal oscillator controlling your motherboard changing with temperature.
Its all perfectly normal. Well spotted though

Your Xorg is better.
Code:
[    85.420] (EE) Error compiling keymap (server-0)
[    85.420] (EE) XKB: Couldn't compile keymap
[    85.420] (EE) XKB: Failed to load keymap. Loading default keymap instead.
[    85.431] (EE) Error compiling keymap (server-0)
[    85.431] (EE) XKB: Couldn't compile keymap

XKB is the old keyboard driver. and your log says
Code:
[    84.264] (II) The server relies on udev to provide the list of input devices.
   If no devices become available, reconfigure udev or disable AutoAddDevices.
so I was expecting evdev to be used for your input devices.
It appears it did not try to load. evdev is in two pieces, A kernel piece and an xorg piece.

If
Code:
emerge -1av xf86-input-evdev
includes
Code:
[ebuild  N     ] x11-drivers/xf86-input-evdev-2.9.2::gentoo
in the output, the Xorg part of evdev is not installed, so let it run.
If the
Code:
 [ebuild  N     ]
is
Code:
[ebuild  R     ]
then in a Rebuild, not New, so you already have it.
Also fix your INPUT_DEVICES=

The kernel part is provided by the CONFIG_INPUT_EVDEV option
Code:
grep EVDEV /usr/src/linux/.config
CONFIG_INPUT_EVDEV=y
Its on by default. That means unless you have turned it off.

A better test is
Code:
zgrep EVDEV /proc/config.gz
as that tests the running kernel if you have /proc/config.gz the former tests the kernel at /usr/src/linux/, which you may not be running.
The fix is to rebuild the kernel.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Sun Jun 07, 2015 1:02 am    Post subject: Reply with quote

Everything looked good as far as those input settings/ no changes from default according to the output: CONFIG_INPUT_EVDEV=y
I ran
Code:
emerge -1av xf86-input-evdev
in several stages of change.
I changed /etc/portage/make.conf from:
    INPUT_DEVICES="evdev keyboard mouse"
    VIDEO_CARDS="sis"
    to:
    INPUT_DEVICES="evdev synaptics"
    VIDEO_CARD="sis vesa"

I did "make menuconfig" to configure the kernel with the configuration on udev on Gentoo wiki.
I did the same thing to get the right use flags for xorg/drivers for keyboard configuration.

I ran
Code:
emerge -pv x11-base/xorg-drivers

it showed 1 installation and 1 re-install. [R]
I was hesitant, because I spent so much of my time working on the xf86-video-sis driver to re-emerge.
Isn't there something I can do to protect that package, before I proceed?

Unless, there's something else you guys have in mind.

Xorg.0.log directly after startx was ran https://bpaste.net/show/24db86449d6f

My generated and edited /etc/X11/xorg.conf https://bpaste.net/show/f129ece0660f
I made the changes trying to load evdev from the file directly, and commited it out when there was no change.

I installed lshw (I thought it might help with processor configuration and keyboard settings) https://bpaste.net/show/eb84479ee3c9
I was also wondering... if CFLAGS="-march=k8 -O2 -pipe" is a better setting, and if I did an
Code:
emerge -ev @world
with that setting would that be worth the outcome and possibly even a fix?
I wasn't sure if I should try and change it or not.
If I can get anyone anymore information with wgetpaste or answer any other questions, please ask.
Thanks, Jay

(I edited and fixed my links... If they didn't work and you saw them. Sorry about that.)
Back to top
View user's profile Send private message
DONAHUE
Watchman
Watchman


Joined: 09 Dec 2006
Posts: 7550
Location: Goose Creek SC

PostPosted: Sun Jun 07, 2015 3:05 am    Post subject: Reply with quote

Recommend
Code:
mv  /etc/X11/xorg.conf /etc/X11/xorg.conf.manual
startx
see if auto can handle it
Quote:
processor family (586/k5/5x86/6x86/6x86MX) --- changed from pentium pro
this likely got you past the illegal instruction, although the running kernel config you pasted yesterday did not have pentium pro --- next time you make a kernel, suggest you try
Quote:
Processor type and features --->
Processor family (Opteron/Athlon64/Hammer/K8) --->
[*] Generic x86 support
Which should optimize the kernel for your processor and then for using it in 32 bit mode

I think /etc/portage/make.conf should have VIDEO_CARDS="vesa sis" and INPUT_DEVICES="evdev" for what you are trying to do. VIDEO_CARDS="sis" and INPUT_DEVICES="keyboard evdev mouse" was present yesterday. Including keyboard and mouse should not be an operational problem but may slow xorg-server. Might be good to make any needed changes to make.conf and
Code:
emerge -1 xorg-server xf86-input-evdev xf86-video-vesa
.

An update
Code:
emerge -s udev systemd xf86-video-vesa xf86-video-sis  xf86-input-evdev  | wgetpaste -t
zcat /proc/config.gz | wgetpaste -t
ls -l /etc/X11 | wgetpaste -t
post url's. Could be useful in keeping us in the peanut gallery in touch with your system state.

if you ran:
Code:
nano /usr/portage/x11-base/xorg-drivers/xorg-drivers-1.16.ebuild
edit last line !<=x11-drivers/xf86-video-sis-0.10.7 to read !<x11-drivers/xf86-video-sis-0.10.7
Code:
ebuild /usr/portage/x11-base/xorg-drivers/xorg-drivers-1.16.ebuild digest
Code:
nano /usr/portage/x11-base/xorg-drivers/xorg-drivers-1.17.ebuild
edit next to last line !<=x11-drivers/xf86-video-sis-0.10.7 to read !<x11-drivers/xf86-video-sis-0.10.7
Code:
ebuild /usr/portage/x11-base/xorg-drivers/xorg-drivers-1.17.ebuild digest
You should be able to emerge xf86-video-sis at will until a new version of xf86-video-sis or xorg-drivers becomes effective. The patch at /etc/portage/patches/x11-drivers/xf86-video-sis-0.10.7 will be applied each time xf86-video-sis-0.10.7 is emerged.
_________________
Defund the FCC.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Mon Jun 08, 2015 1:41 pm    Post subject: Reply with quote

I would've got this to you sooner. I kept doing a make menuconfig. I configured 12 different kernels. None would support my wifi adapter, because of the b43/ssb bug.
I finally had to use genkernel --menuconfig all
Why that version is stable and the alternative isn't, I'm unsure. It about drove me crazy.
Code:

emerge -s udev systemd xf86-video-vesa xf86-video-sis  xf86-input-evdev  | wgetpaste -t

https://bpaste.net/show/5e5f985294c6

Code:
zcat /proc/config.gz | wgetpaste -t

https://bpaste.net/show/2315e16fa963

Code:
ls -l /etc/X11 | wgetpaste -t

https://bpaste.net/show/02df55fdce59

I never did have systemd installed. I have openrc. From what I understood when I went through the installation process was you should chose one or another. I'll have to read up more on that.
I was swapping back and forth between xorg.conf scripts and auto, and I still couldn't get startx to work. I was slightly altering and playing around with the self-generated scripts and what's on wiki.
No luck.

I can't wait to figure all this out.
Thanks, Jay

(Edited to fix 1st link. Sorry about that.)


Last edited by jwm224 on Mon Jun 08, 2015 6:47 pm; edited 1 time in total
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 42581
Location: 56N 3W

PostPosted: Mon Jun 08, 2015 5:29 pm    Post subject: Reply with quote

jwm224,

Your latest Xorg.0.log is a must too.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Mon Jun 08, 2015 6:41 pm    Post subject: Reply with quote

Xorg.0.log
https://bpaste.net/show/f2c70c2e9807

I picked up a PNY 3200 DDR 400Mhz 1Gb stick of memory today for $5.30. There was only two slots. Now, this monster has almost 1.5Gb. lol... I think it takes 128Mib of system memory for video. It only registers like 1.38.

Monetary issues and garbage equipment aside, I was just investigating the rc-status and init.d of what needs to be running for startx, but that's to get it running from boot, correct? I also was curious... I've been downloading and compiling things I'll need further down the road, like thunar, slim, slim-themes, openbox..., rather than my machine just sitting idle.
It's only a few clicks to delete. I figured it was better to give you a complete picture of what's going on.

I also set
Code:
Device Drivers -> Input device support -> <*>  Event interface

and
Code:
Automount devtmpfs at /dev, after the kernel mounted the root partition

in my latest kernel modification.

Jay
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 42581
Location: 56N 3W

PostPosted: Mon Jun 08, 2015 8:01 pm    Post subject: Reply with quote

jwm224,

You are still building, or have built somethung for the wrong CPU.
Your CFLAGS from make.conf can be changed by both ebuilds and by the packages provided build system, so it may not be something you are setting.
We need to determine the package causing the illegal instruction and fix it.

From your log
Code:
[  1657.714] (WW) Warning, couldn't open module sis
[  1657.714] (II) UnloadModule: "sis"
[  1657.714] (II) Unloading sis
[  1657.714] (EE) Failed to load module "sis" (module does not exist, 0)
Thats a bit disappointing. The sis driver is not installed.
However, vesa loaded - it works on any graphics card produced since 1998 and a few before then too.
Code:
 [  1657.715] (II) Module vesa: vendor="X.Org Foundation"
[  1657.715]    compiled for 1.16.4, module version = 2.3.3
[  1657.715]    Module class: X.Org Video Driver
[  1657.715]    ABI class: X.Org Video Driver, version 18.0
[  1657.715] (II) VESA: driver for VESA chipsets: vesa
[  1657.715] (--) using VT number 7
so far so good, then

Code:
[  1658.083] (EE) Error compiling keymap (server-0)
[  1658.083] (EE) XKB: Couldn't compile keymap
[  1658.083] (EE) XKB: Failed to load keymap. Loading default keymap instead.
[  1658.094] (EE) Error compiling keymap (server-0)
Thats 'game over'. Thats odd on a number of counts,
udev should have been adding devices ...

What does
Code:
uname -a
say?
I get
Code:
$ uname -a
Linux NeddySeagoon_Static 4.0.3-gentoo #1 SMP PREEMPT Sun May 17 13:21:36 BST 2015 x86_64 AMD Phenom(tm) II X6 1090T Processor AuthenticAMD GNU/Linux
The date, Sun May 17 13:21:36 BST 2015 is the build time of the running kernel. Yours will be different. Does it look right to you?
As your .config from /proc/config.gz looks OK, it won't matter - thats your running kernel.

Is udev actually running?
Code:
$ ps -Alf | grep udev
5 S root      1559     1  0  80   0 -  7090      - Feb09 ?        00:00:00 /sbin/udevd --daemon
0 S roy      19876 19869  0  80   0 -  1679      - 19:56 pts/4    00:00:00 grep --colour=auto udev

The /sbin/udevd --daemon is the important bit.

dmesg may tell what was killed by thu illegal instruction exception. If you can find it , rebuild it and try again.
If not, post dmesg.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
John R. Graham
Administrator
Administrator


Joined: 08 Mar 2005
Posts: 10156
Location: Somewhere over Atlanta, Georgia

PostPosted: Mon Jun 08, 2015 8:24 pm    Post subject: Reply with quote

Probably not the issue, but I'm compelled to ask. I had him fix his CFLAGS and reccompile everything. @jwm224, have you rebooted since then?

- John
_________________
I can confirm that I have received between 0 and 499 National Security Letters.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Mon Jun 08, 2015 9:01 pm    Post subject: Reply with quote

uname -a
https://bpaste.net/show/8e2914326e90

ps -Alf | grep udev
https://bpaste.net/show/b374ee4e33b3

There are systemd components showing here. I don't remember building any, and I'm positive it's not checked in the genkernel, just openrc. Though, I'll go check the option, if you think it'll help.
I want to go through it again anyway. I learned so much prepping the other kernel and troubleshooting for b43 / ssb catastrophe all of yesterday. I know there're changes to be made here and there.

dmesg
https://bpaste.net/show/84b23297304b

Yeah, I reboot after almost everything I compile or build. Maybe too much.

Somewhere over the last day I got lost on where I was with kernel options, lost the point I was at. I'm just about back there now.
I noticed that too with the SiS driver.. I saw Xorg.0.log files that had [SiS] listed all through it, and just the same, [vesa]. It seems that either will work, although I don't know the stability factor is with either one. I'm thinking it's kernel verses genkernel, because that's what I had to fall back on.

Jay
Back to top
View user's profile Send private message
DONAHUE
Watchman
Watchman


Joined: 09 Dec 2006
Posts: 7550
Location: Goose Creek SC

PostPosted: Mon Jun 08, 2015 9:32 pm    Post subject: Reply with quote

The sis module appears when the running kernel was eselected and emerge xf86-video-sis was run against it.
The sis module disappears when a new kernel is booted that has not had emerge xf86-video-sis run against it.
Usually the module is persistent as long as the kernel version is unchanged and the changes are minor. Possibly not the case for the kludged sis module.
_________________
Defund the FCC.
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 42581
Location: 56N 3W

PostPosted: Mon Jun 08, 2015 10:02 pm    Post subject: Reply with quote

jwm224,

Your dmesg is full of your wifi restarting, so its not useful.
Everything else looks good.

The vesa driver is a generic, uses the BIOS on the video card to provide very limited 2D acceleration driver. It will work with almost any video card.
The sis driver is designed to take advantage of your video card, so its performace is better.

Vesa is maintated to work with the latest Xorg. sis is/was not, hence the patch that DONAHUE found for you.

Regardless, we need to find the source of the illegal instruction exception and fix it.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
DONAHUE
Watchman
Watchman


Joined: 09 Dec 2006
Posts: 7550
Location: Goose Creek SC

PostPosted: Tue Jun 09, 2015 12:56 am    Post subject: Reply with quote

From the peanut gallery: It appears from the last couple of Xorg.0.log that the Caught signal 4 Illegal Instruction failure mode has been cleared.

My Xorg.0.log has many appearances of evdev and of udev and 10 of XINPUT. jwm224 has 0 evdev, 0 udev, and 1 XINPUT. As Neddy has stated repeatedly there is a udev problem causing an evdev problem.
Code:
emerge -1pv gcc glibc xorg-server xf86-video-sis xf86-input-evdev | wgetpaste -t
rc-update show | wgetpaste -t
rc-status | wgetpaste -t
post url's; look for udev use flags

Following courtesy of google and https://wiki.archlinux.org/index.php/Xorg :

Quote:
X failed to start: Keyboard initialization failed

If your hard disk is full, startx will fail. /var/log/Xorg.0.log will end with:
(EE) Error compiling keymap (server-0)
(EE) XKB: Could not compile keymap
(EE) XKB: Failed to load keymap. Loading default keymap instead.
(EE) Error compiling keymap (server-0)
(EE) XKB: Could not compile keymap
XKB: Failed to compile keymap
Keyboard initialization failed. This could be a missing or incorrect setup of xkeyboard-config.
Fatal server error:
Failed to activate core devices.
Please consult the The X.Org Foundation support at http://wiki.x.org
for help.
Please also check the log file at "/var/log/Xorg.0.log" for additional information.
(II) AIGLX: Suspending AIGLX clients for VT switch
Which exactly matchesthe statements at the end of our Xorg.0.log. The suggested fix is:
Quote:
Make some free space on your root partition and X will start.
@ jwm224:
Can you see any way this could be true?
Can anyone see how whatever process is trying to compile a keymap could conclude that the root partition is full if it isn't?
_________________
Defund the FCC.
Back to top
View user's profile Send private message
jwm224
Tux's lil' helper
Tux's lil' helper


Joined: 27 May 2015
Posts: 118
Location: Pennsylvania

PostPosted: Tue Jun 09, 2015 6:34 am    Post subject: Reply with quote

Code:
emerge -1pv gcc glibc xorg-server xf86-video-sis xf86-input-evdev | wgetpaste -t

https://bpaste.net/show/b9704e8afb6d
Code:
rc-update show | wgetpaste

https://bpaste.net/show/d1c31c49c61f
Code:
rc-status | wgetpaste -t

https://bpaste.net/show/ad53d4a00efd

I read through that same arch forum earlier today.
Here's my free space, using
Code:
df -h

https://bpaste.net/show/e2be331fe513
Lots of space in the root.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Desktop Environments All times are GMT
Goto page Previous  1, 2, 3, 4  Next
Page 2 of 4

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum