Feedback for Porteus version 2.0 RC2

[Rava]

Back to my nVidia issue...

I re-booted the same machine, and again at boot time the module was not loaded, but in X I used "activate" as root from the terminal, and it loaded just okay.
Sure the nvidia-settings told me that X was not running using the nvidia driver.
I checked if the /etc/X11/xorg.conf was the one from the nVidia module. And since it was, I re-started XFCe via "LogOut" using guest account for login into X.

But again, nvidia-settings told me that X was not running using the nvidia driver.

Soo.... why is Porteus not loading the module at boot time? Seems all is fine with the module or I could not have run nvidia-settings...
And me thinks it not works that way since now the nouveau driver is still active and not kinda blacklisted, using the right /etc/X11/xorg.conf seems like a 0-trick pony here...


Any ideas?

[beny]

sorry..but you have this file in modprobe.d:# generated by nvidia-installer
blacklist nouveau
options nouveau modeset=0
i use to put in before install the nvidia driver, also xzm driver package

[brokenman]

I had to comment out the slacky = http://darkstar.ist.utl.pt/slackware/ad ... are-13.37/ line and insert http://slack.isper.sk/pub/slackware-13.37 to have slackyd working
I recommend using slackware-14.0 with Porteus-v2.0. Also this issue with slackyd was resolved in v2.0rc2. Please check that darkstar is not the first repo listed in slackyd.conf. Searching the forum shows a few threads mentioning the problem.

[claude]

@fanthom

i had the same problem after upgrade form 1.2 to 2.0-rc1.
must be some incompatibility in PPM configs.
if brokenman wont give you better advice which file to delete then i would boot in 'always fresh' and check if PPM works.
if that resolves PPM issue then i would delete whole /etc/ppm and /opt/ppm folders from your changes (please make sure you are doing it in 'always fresh' mode so aufs wont remember that these folders were deleted) and then boot as normal.

I always boot in "always fresh" mode... Booting an always fresh system is my preference, which is why I elect to use a live system as opposed to regular linux on HD....

@brokenman

Yes the PPM from rc1 suffered many bugs due to the gtkdialog upgrade. These have been addressed in the latest PPM. Please download from one of the following links.

32bit
64bit

Our modules are slowly growing in the repo. 64bit has more than 32bit but i will catch up for final.

I don't have much time on my hands these days. I'll be fine with my current setup until Porteus 2.0 final release is out, then go for 64bit for my main machine....I think that will resolve a lot of issues....
Looking forward to 2.0 final

[jesmith]

I saw this in RC1 as well, but forgot to tell you.

When I boot to Text Mode I just get a blank screen. No idea what's going on there. I can boot to X just fine.

As you know, I rebuild the kernel as part of my onboarding process for a new distro, and after I did that with RC1 I was able to get to text mode after that. That makes me suspect the nouveau drivers, since part of my process involves blacklisting those and switching to the genuine NV drivers. (Needless to say, the box has a discrete NV card in it.)

Just thought you should know...

-Joshua

[johnel]

@fanthom

Link to system dump: http://pastebin.com/0DnCbrxN

(Link to spec: http://www.expertreviews.co.uk/laptops/ ... ifications)

btw: I quickly decompressed 001-core.xzm, commented-out the "discrete disable" bit in rc.M, recompressed the module and rebooted - doesn't cure the problem graphics are still slow.

[Rava]

sorry..but you have this file in modprobe.d:# generated by nvidia-installer
blacklist nouveau
options nouveau modeset=0
i use to put in before install the nvidia driver, also xzm driver package

Ohhhh that issue again, and me hoped by now this little and necessary gem would be put in all nVidia modules by default... (The discussion about that we had like a year ago or so, but I don't bother looking for the old threads...)

I just hoped it was already a kind of rule that all nVidia drivers need the above to run and so have the modules include it by default.

Yes, I put that in bold out of believing that this should really be a Porteus rule, otherwise using any nVidia drivers would result in a Porteus not running / being not able to use that very driver by default and we would get all the noobies posts and complains in the forums once again..
(And me myself acted quite noobish to not recall and cross check for the very blacklist... I thought about blacklisting nouveau, but also thought by now the modules would include this by default.)

I put the above in the rootcopy/ folder and hope it will work then.... I also renamed to make the file name a bit shorter but could not imagine why that could be any issue about loading or not loading it via boot time...

_____________________


And, beny, did you get my PM about the slackyd man page? By now I included the draft, not yet formatted in the right way of course for a man page source, for you to check if all is fine with my translation, if you would cooperate with me on that...

[fanthom]

@bdheeman

why do we have both a /run and /var/run?

most of packages still rely on /var/run.
/run is used mainly by udev and udisk2. not sure if anything else.

How about adding/including a PDF printer driver for cups, see http://www.cups-pdf.de/

not sure if it would be popular among the users (first time i hear about it) but it could be added to the repo.

@Rava

Soo.... why is Porteus not loading the module at boot time?

you must have use 'noload=nVidia' cheatcode or such.
just in case - please generate psinfo report and post pastebin.com url here.

I just hoped it was already a kind of rule that all nVidia drivers need the above to run and so have the modules include it by default.

all nVidia modules have nouveau blacklisted by default.
nouveau is not blacklisted for you because your nVidia module is not loading during boot.
mount nVidia with 'mloop' and check /mnt/loop/etc/modprobe.d folder.

there is also another possibility:
you have dozen of porteus installations, your are not using 'from=UUID:' cheatcode -> porteus is not booting from the device you think.
once again - please generate psinfo report, upload also /var/log/porteus-livedbg and i'll give you an answer.

@johnel
logs looks ok to me.
the very last commit before rc2 was to switch acceleration method of intel driver back from SNA to UXA as i got some artifacts on gtk apps on my laptop (ironlake GPU) when SNA was enabled by default.
please create /etc/X11/xorg.conf.d/20-intel.conf with the following content:

Code: Select all

Section "Device"
   Identifier  "Intel Graphics"
   Driver      "intel"
   Option      "AccelMethod"  "sna"
EndSection

then as root run 'init 3' -> login as root -> run 'init 4'
if that wont help then please generate report once again and link here (i want to be sure that SNA was enabled correctly).
you could also downgrade to official package from slackware-14.0:
link
please boot to text mode -> run 'removepkg xf86-video-intel' -> run 'installpkg /path_to/xf86-video-intel-2.19.0-i486-1.txz -> init 4'
let me know if any of above helped.
(there is no much time left to porteus-2.0 final and i want to have all these bugs resolved asap)

@jesmith

When I boot to Text Mode I just get a blank screen. No idea what's going on there. I can boot to X just fine.

first time i hear about such case - usually people get black screen in text mode and X.
please run with stock rc2 (no nVidia driver, no other modifications) with 'acpi_osi=Linux acpi_backlight=vendor" cheatcodes and let me know if that helped.
if not then please boot to GUI, generate "Porteus System Info" report, upload to pastebin.com and post an url here.
would be also nice to have your kernel config to find differences and eventually fix porteus one.

[chyang0527]

I am thinking this question might be a silly problem, but i still don't know why.
I have tested the Porteus version 2.0 RC2 and RC1. After successfully booting into Porteus, i found that the hard disk does not be mounted automatically.
For example, my hard disk is recognized under /dev/sda, /dev/sda1, i have to manually mount this device to /mnt/sda1. This behavior is different from Porteus 1.2, which could mount the /dev/sda1 to /mnt/sda1 automatically while boot time.
Does this behavior change cause by upgrading linux kernel? or any config setting?
Thanks!

[Rava]

you must have use 'noload=nVidia' cheatcode or such.
just in case - please generate psinfo report and post pastebin.com url here.

Nope I haven't. All I changed from the 2.0 iso of XFCe version prior copying it to the USB stick was adding the nVidia driver..

all nVidia modules have nouveau blacklisted by default.

Brilliant, cause there was a time when that was not the case...

nouveau is not blacklisted for you because your nVidia module is not loading during boot.
mount nVidia with 'mloop' and check /mnt/loop/etc/modprobe.d folder.

*facepalms* Jep, sure thing. I forgot that lil detail...

there is also another possibility:
you have dozen of porteus installations, your are not using 'from=UUID:' cheatcode -> porteus is not booting from the device you think.

Nope, there is only one Porteus on that very USB stick, and no other media is present, I switched off the internal harddisk.

once again - please generate psinfo report, upload also /var/log/porteus-livedbg and i'll give you an answer.

Will do, thanks so far.

[johnel]

@fanthom

The following change from UXA to SNA worked

@johnel
logs looks ok to me.
the very last commit before rc2 was to switch acceleration method of intel driver back from SNA to UXA as i got some artifacts on gtk apps on my laptop (ironlake GPU) when SNA was enabled by default.
please create /etc/X11/xorg.conf.d/20-intel.conf with the following content:

Code: Select all

Section "Device"
   Identifier  "Intel Graphics"
   Driver      "intel"
   Option      "AccelMethod"  "sna"
EndSection

then as root run 'init 3' -> login as root -> run 'init 4'

I'm not sure what you want to do to fix it or maybe leave it as is. I guess it depends which AccelMethod is most popular. Is the UXA method newer and supports more variations? Just curious!


Thanks for your help and look forward to final release!
(If you need any more info or help please let me know)

[Rava]

So... All I did was renaming the nVidia module to 010-nVidia-310.32-v2.0-rc2-x86_64.xzm
(I prefer giving my always loaded modules numbers anyway...)

And sure, this time it was loaded just fine... :sneaky:

I will re-name it to its original name and see if all is well again, or not. If all is well then, too, I just file that mentally into the unexplainable esoteric stuff folder... stuff PCs just do once in a while. I think quite some of you know what I mean by that. xD
_________________________

The mplayer is having its fits...

Now, I have no clue if that's because I have the mp3 on a USB stick... or if the glitch is somewhere else
All gnome-mplayer gives me is this:

Code: Select all

GMLIB-Message: in media state change with state = play dontplaynext = 0
GMLIB-Message: in media state change with state = unknown dontplaynext = 1

Will change its cache stuff to see if it improves...

*stops Pachelbel_Canon_in_D_major*
*changes cache settings*
*starts up Pachelbel_Canon_in_D_major at a later so far not played part*

Arghh the sound still has gliches / breaks,

I set up the cache to 4K, the default mplayer gave me was the 2k... Me thinks 4k should be enough to work with a mp3...

And the machine, Dual Athlon is sure quick enough to handle silly plain mp3... be it 320KB/s...

So, and ideas what I also can check about the sound glitch.
Will psinfo be helpful here, too?
Or should I gather some aplay info?

[fanthom]

@chyang0527
porteus mounts every drive automatically unless you use 'noauto' or 'nohd' cheatcodes.
please provide output of 'blkid' and 'cat /proc cmdline' commands.

@Rava
so i assume nVidia issue is resolved?
please provide psinfo log for an audio issue. do you have a sound in standalone player (audacious i guess?)

@johnel
SNA is never and still experimental acceleration but i think it's less evil to have artifacts (on some GPU's) than broken rendering (on some GPU's) so we'll stick to it for final.
thanks.

[chyang0527]

@fanthom

Thanks for your information.
I think my problem might only happened in my VM test machine in Windows. (VMWare Workstation 8.0)
I have used my usb stick to test on my physical machine, the hard disk could be mounted automatically.
The wierd part is the hard disk in VM machine could be auto mounted in Porteus 1.2 with linux kernel 3.4.4 before.

below is the output of 'blkid' and 'cat /proc/cmdline'

guest@porteus:~$ su
Password:
root@porteus:/home/guest# cd /mnt/
root@porteus:/mnt# ls
live/ sr0/
root@porteus:/mnt# cd /
root@porteus:/# blkid
/dev/sr0: LABEL="Porteus" TYPE="iso9660"
/dev/loop0: TYPE="squashfs"
/dev/loop1: TYPE="squashfs"
/dev/loop2: TYPE="squashfs"
/dev/loop3: TYPE="squashfs"
/dev/loop4: TYPE="squashfs"
/dev/loop5: TYPE="squashfs"
/dev/loop6: TYPE="squashfs"
/dev/loop7: TYPE="squashfs"
/dev/sda1: UUID="B6A41AC9A41A8BCD" TYPE="ntfs"
root@porteus:/# cat /proc/cmdline
quiet initrd=initrd.xz changes=/porteus BOOT_IMAGE=vmlinuz
root@porteus:/#

Thanks.

[fanthom]

@chyang0527
in 2.0 i have compiled all SCSI controllers as kernel modules - forgotten about possibility of using porteus in vmware.
could you:
- attach all controllers (SCSI, SATA, IDE) to vmware virtual session
- boot porteus in that session
- post output of 'lspci -k > lspci' and 'lsmod > lsmod' commands.

i'll compile all necessary drivers into kernel so this problem will be resolved.
thanks