Upup Raring 3.9.9.2 with non-PAE 3.9.9 kernel. 13 July 2013

Message

mikeslr · #61 Post by **mikeslr** » Fri 19 Jul 2013, 00:00

Hi pemasu,

Just wanted to thank you for another great puplet. It's working great via a USB-key install* on my Thinkpad T42 with vicmz's OpenboxPlus installed. http://murga-linux.com/puppy/viewtopic. ... 730#703730.

Except for one problem: no sound. But that's neither yours nor vicmz's fault. I attribute that to Newton's Third Law of Physics and my cat.

Reminds me of a skit.

Patient: Doctor, will I be able to play the violin after the operation?
Doctor: I don't see why not.
Patient: Well, I was never able to before.

mikesLr

* Will move it onto the hard drive now that I see it works.

linux28 · #62 Post by **linux28** » Fri 19 Jul 2013, 01:33

Thank you!

tlchost · #63 Post by **tlchost** » Fri 19 Jul 2013, 06:36

anikin wrote:That is why it's there ... a line that flies in the face of reason but gets the job done.

As George Carlin said "It's a mystery Father, a mystery"

Thanks

charlie6 · #64 Post by **charlie6** » Fri 19 Jul 2013, 20:11

Hi Pemasu,
further infos to my previous post:
http://murga-linux.com/puppy/viewtopic. ... 4&start=50
falling back to the previous /usr/lib/xorg/modules/drivers/intel_drv.so offers a workaround.
I have unsquashfs'ed the former puppy_raring-3.9.9.sfs to pick up its intel_drv.so and inject it into the new puppy_raring-3.9.9.sfs (and mksquashfs'ed it ...

) ...for intel-video-chip-Brookdale-82845G purpose ...
Cheers, Charlie

charlie6 · #65 Post by **charlie6** » Fri 19 Jul 2013, 20:29

Hi (again) Pemasu,
considering this on Dpup-wheezy-3.5.2.7:
http://www.murga-linux.com/puppy/viewtopic.php?t=80546

pemasu wrote:... Libgl1-mesa-dri is the package for graphics spesific dri drivers. I intended to include them....but woof2 template stripped them to swrast_dri.so only. I might create it as pet ...

...how to reproduce this on Upup Raring-3.9.9 ...? is it feasable ?

Charlie

charlie6 · #66 Post by **charlie6** » Fri 19 Jul 2013, 20:50

Hi again Pemasu,
i encountered this several (and enough) times to be worth a post:
once quick-setup settings (here to fr_BE + be_latin1 keyboard) are made and X restarted, rox keeps being with US keyboard settings; while abiword, geany, gnumeric, etc... work as expected.
See shot attached: rox displays q for a; a for q; = for -; / for =; etc...
Thanks for considering this!
Charlie

pemasu · #67 Post by **pemasu** » Sat 20 Jul 2013, 07:41

Charlie6. If I understand correctly you should have azerty in Rox. Instead you have qwerty.

I have created 3.9.9.1 version and that problem does not exist in it. I have created usb-modeswitch-1.2.6 package wrongly in 3.9.9. Also I have compiled Osmo 96dpi patched version according Barrys guidelines. I have downloaded fresh Ubuntu Raring packages and fresh woof. I believe that I have used same woof platform too many times because there hasnt been any woof updates in last weeks. I may have slight platform corruption. Just hunch.

Anyway...I am returning today from island. I will upload quite soon 3.9.9.1 version.

About xorg intel_drv.so driver problem with brookdale graphics. I am sorry to hear that my compile is not compatible with it. Instead, it allows Sandybridge and IvyBridge architecture intel graphics to work. Brookdale is known to be problematic. Ubuntu choice to use sna acceleration in 2.21.6 version was not good choice. I compiled the latest 2.21.12 with default uxa acceleration. I could test compile it with sna acceleration also. Ubuntu might have essential patches also, would need to check. But....right now it seems to be choice between these newer architecture or brookdale.

edoc · #68 Post by **edoc** » Sat 20 Jul 2013, 18:11

One problem:

devx seems to interfere with loading LibreOffice 4. (I moved it to a subfolder and rebooted and LibreOffice 4 loaded fine.)

WDYT?

anikin · #69 Post by **anikin** » Sat 20 Jul 2013, 18:38

IT, that I keep mine as shown in the image below. No interference so far.

edoc · #70 Post by **edoc** » Sat 20 Jul 2013, 18:55

I'm using devx_raring_3.9.9.sfs

anikin · #71 Post by **anikin** » Sat 20 Jul 2013, 19:32

And it's only loaded when needed, and you unload it before using LibreOffice, or any other SFS. The way it's supposed to be used, correct?

edoc · #72 Post by **edoc** » Sat 20 Jul 2013, 20:29

anikin wrote:And it's only loaded when needed, and you unload it before using LibreOffice, or any other SFS. The way it's supposed to be used, correct?

One is allowed to have six SFS's load at boot time - I've always included the matching devx as one of the six.

In this case it does not appear to be playing nicely with one of the other SFS's - LibreOffice.

Naughty puppy!

anikin · #73 Post by **anikin** » Sat 20 Jul 2013, 21:16

This reminds me of: "I love my cigar too, but I take it out once in a while."

pemasu · #74 Post by **pemasu** » Sat 20 Jul 2013, 21:36

Upup Raring 3.9.9.1 has been uploaded.

Kernel is same 3.9.9 non-PAE.

I have repackaged usb-modeswitch-1.2.6 with recent usb-modeswitch-data. There is now about 30 new modem data files more since latest 1.2.4 which was in puppy repo. I hope it is functionally repackaged now. Osmo has been recompiled with 96dpi patch. It does not fill the whole screen now. Geany-1.22 from ubuntu repo. You can download the geany-plugins straight from repo. There are a lot of them. Fresh download of Ubuntu Raring packages for building. Redownloaded latest woof.

Ppower accompanied scripts has been simplified. Due to recent forum posts. There is still cpu governor switching available as tray app and automatic switch to ondemand governor when in battery.

Xorg intel_drv.so is the new 2.21.12 with uxa acceleration. Sna acceleration which ubuntu now uses would need fuller xorg.conf with separate intel device section which defines sna acceleration. Puppy does not atm automatically add that, so I have uxa acceleration in driver as default. Intel SandyBridge and IvyBridge graphics seems to be happy with this choice. Brookdale might complain unfortunately.

No default choice for Frisbee in this build. You can choose which net connection gui you use when you first time click the connect icon or tray network icon. Choose which you prefer. I think it is easier this way.
Pupdsk.flg mechanism has been still included. It has flaws, but it would be better to report of them than exclude it. You can manually check if pupdsk.flg has been left over in ext partition after X has popped up and you have finished the Quick Setup with ok. If pupdsk.flg still lurks there, remove it. I think that the mechanism is not yet foolproof.

There is now delta file for upgrading previous 3.9.9 iso.

Download link: http://www.smokey01.com/pemasu/UpupRari ... aring3991/

Kernel source sfs, devx sfs and proprietary drivers for ATI and Nvidia are in Quickpet, icon is in Pinboard.

edoc · #75 Post by **edoc** » Sat 20 Jul 2013, 21:50

Sounds cool!

Got it downloaded and will test on the Samsung NB30 Netbook.

BTW: I've never applied a Delta before - not even sure how.

Any gotchas?

pemasu · #76 Post by **pemasu** » Sat 20 Jul 2013, 21:56

Cut and copy from this post: http://www.murga-linux.com/puppy/viewtopic.php?t=52232

What to do... for upgrading an iso image.

* Download the correct .delta file for the iso image you want to upgrade:eg if you have pup_stardust-005.iso get pup_stardust-005.iso:pup_stardust-006.iso.delta. Of course check the checksum
Code:
md5sum pup_stardust-005.iso:pup_stardust-006.iso.delta
.This will produce a long string of alpha and numeric characters, it must be exactly the same as the one posted by the developer.
* Save the original .iso and the new .delta in the same directory, this is not essential but will save some searching time
* Simply click the .delta file and the Xdelta gui will appear

* Click GENERATE
* You should see a green splash screen indicating that the new file is generated and there will be a new .iso file in the same directory as the .delta file. Again, compare the checksum with the one provided by the developer. All should be well.

edoc · #77 Post by **edoc** » Sat 20 Jul 2013, 23:21

I went back and grabbed the iso and devx and updated manually.

All is well ... even LibreOffice and devx now play nicely together.

Good puppies!

Thanks for the upgrade.

Billtoo · #78 Post by **Billtoo** » Sun 21 Jul 2013, 02:57

I did a manual frugal install of 3.9.9.1 to a 16gb SDHC card, running
on a Compaq S6700NX desktop pc.
Computer
Processor 2x Intel(R) Pentium(R) 4 CPU 2.80GHz
Memory 1553MB (199MB used)
Machine Type Physical machine
Operating System Upup Raring - 3.9.9.1
User Name root (root)
Date/Time Sat 20 Jul 2013 10:48:50 PM EDT
Display
Resolution 1280x1024 pixels
OpenGL Renderer Mesa DRI R200 (RV280 5961) x86/MMX/SSE2 TCL DRI2
X11 Vendor The X.Org Foundation
Audio Devices
Audio Adapter CMI8738-MC6 - C-Media CMI8738
Input Devices
AT Translated Set 2 keyboard
Logitech USB Receiver
Logitech USB Receiver
Power Button
Power Button
PC Speaker
Linksys USB wireless adapter [RT73]

I installed lots of applications in ppm, added openbox_plus-1.5.3.pet
and a few others.

Thanks for the new version.

EDIT: I found that pmusic makes the processor run at 80%+ usage, clementine only uses 20% or so with Seamonkey active at the same time.
EDIT:I just tried pmusic again and it's okay now.

anikin · #79 Post by **anikin** » Sun 21 Jul 2013, 12:15

xorg.log wrote:(WW) Open ACPI failed (/var/run/acpid.socket) (No such file or directory)
LoadModule: "intel"
Loading /usr/lib/xorg/modules/drivers/intel_drv.so
Module intel: vendor="X.Org Foundation"
compiled for 1.13.3, module version = 2.21.12
Module class: X.Org Video Driver
ABI class: X.Org Video Driver, version 13.1
UnloadModule: "intel"
Unloading intel
Failed to load module "intel" (already loaded, -1217513310)
LoadModule: "vesa"
Loading /usr/lib/xorg/modules/drivers/vesa_drv.so
Module vesa: vendor="X.Org Foundation"
compiled for 1.12.99.902, module version = 2.3.2
Module class: X.Org Video Driver
ABI class: X.Org Video Driver, version 13.0
UnloadModule: "vesa"
Unloading vesa
Failed to load module "vesa" (already loaded, -1217513310)
LoadModule: "modesetting"
(WW) Warning, couldn't open module modesetting
UnloadModule: "modesetting"
Unloading modesetting
(EE) Failed to load module "modesetting" (module does not exist, 0)
LoadModule: "fbdev"
Loading /usr/lib/xorg/modules/drivers/fbdev_drv.so
Module fbdev: vendor="X.Org Foundation"
compiled for 1.12.99.902, module version = 0.4.3
Module class: X.Org Video Driver
ABI class: X.Org Video Driver, version 13.0
UnloadModule: "fbdev"
Unloading fbdev
Failed to load module "fbdev" (already loaded, 0)
Loading sub module "dri2"
LoadModule: "dri2"
Module "dri2" already built-in

I have already asked this question. Why modules are loaded if the are already loaded?
We didn't have anything like this in earlier builds. In the past, this section of the log was and should normally be whistle-clean - modules are just loaded, get along nothing to see here ... something ain't right. Regarding SNA, Barry's raring has it I beleive. Works well for me, Ubuntu has accepted it, Barry is reverting to UXA only because he aims at older hardware. Is upupRaring targeting the same segment?

For my eeepc (atom n270/945) the first video driver worked better. This one, doesn't give me a chance of doing some very basic xorg.conf settings. For a long time, I've been applying this to all my puppies:

Code: Select all

Section "Monitor"
	Identifier   "Monitor0"
	VendorName   "HSD"
	ModelName    "3e9"
	HorizSync    30-41
	VertRefresh  56-65
	DisplaySize	220	129
EndSection
	
Section "Modes"
	Identifier "Modes0"
EndSection

Section "Device"
    Identifier  "Card0"
    Driver      "intel"
    VendorName  "Intel Corporation"
    BoardName   "Intel Corporation Mobile 945GME Express Integrated Graphics"
    Option      "AccelMethod" "sna"
	Option      "DRI" "true"	
	BusID       "PCI:0:2:0"
EndSection

Section "Screen"
	Identifier "Screen0"
	Device     "Card0"
	Monitor    "Monitor0"
	DefaultDepth 24
		Subsection "Display"
		Depth       24
		Virtual     2048 2048
		Modes       "1024x600"
	EndSubsection
EndSection

Section "DRI"
       Group  "video"
       Mode   0660
EndSection

Some puppies give better font rendering if I specify the display size, others will do better if I comment out that line. But overall, these settings work perfectly well, because xorg has everything it needs, already predefined. However, this raring rejects my legitimate and innocent settings! It gives a square screen, that can only be adjusted (via quicksetup) to what you'll see in the second image. Note also, this is a 945 chip - very basic one, absolutely unassuming, humble and unpretentious. Not a performance beast. It just works and never fails on any puppy.

pemasu · #80 Post by **pemasu** » Sun 21 Jul 2013, 12:49

Anikin. I dont know about that xorg.log. The new mechanism how xwin - xorgwizard - xorgwizard-automatic works is not yet quite familiar to me. You should ask straight from Barry K.

About intel_drv.so choice. I know surely that IvyBridge and Sandybridge intel graphics didnt work with 3.9.9. Sandybridge intel graphics didnt even get to the X. JWM crashed. IvyBridge was slower than snail. That is why I reverted to the uxa acceleration as default. I decided to test the newest version at the same time.

Beside. Sna acceleration is not in use, even though it has been compiled as default, if you dont have that xorg.conf device section, which defines sna acceleration. Just like in Anikin has tried to do.
The intel_drv.so driver has been compiled with sna and uxa in Ubuntu version and in mine. The default choice is just different in mine and Ubuntu version.

I knew that there are some which would like to use default sna accelerated intel_drv.so. For them fix is easy and quite small.
Just download from PPM: xserver-xorg-video-intel_2.21.6 .deb. Install it and restart X.
Done.

(old)Puppy Linux Discussion Forum

(old)Puppy Linux Discussion Forum

Upup Raring 3.9.9.2 with non-PAE 3.9.9 kernel. 13 July 2013

Working great on a Thinkpad T42

Re: new upload raring-3.9.9 by 130717

rox not set to locale setting