Precise Puppy RC2, October 20, 2012

Under development: PCMCIA, wireless, etc.
Message
Author
Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#571 Post by Sage »

Couple of other features have appeared in the last two betae:
PS/2 mouse remains stationary after first liveCD load; generally works at second ask (no saved data, though). USB wi-fi undetected until second USB item plugged in.
Weird and very new.

User avatar
shinobar
Posts: 2672
Joined: Thu 28 May 2009, 09:26
Location: Japan
Contact:

Re: bold or normal

#572 Post by shinobar »

tiangeng wrote:
shinobar wrote:
tiangeng wrote:how do I adjust for the bold style in precise?
http://www.murga-linux.com/puppy/viewto ... 207#650207
Thanks!

Very good!

But I want to keep English font unchanged, and only need to adjust Chinese font, how should do?
What happened on your english fonts?
Downloads for Puppy Linux [url]http://shino.pos.to/linux/downloads.html[/url]

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

Driver for DGC USB analog modems

#573 Post by rerwin »

Barry, et al,
Now that I think I have completed the HSF modem package testing and fixing, I now attach similar packages for the DGC (CDC ACM) USB modems. The dgcmodem driver is actually only a "wrapper" around the cdc_acm module, in that it apparently handles "off-spec" models that use the Conexant DGC chipset. It seems to pass through "in-spec" models, so can apparently handle all cdc-acm devices.

My test modem works with- or without the DGC driver, so must be a normal model. So, I cannot verify that it actually makes the non-standard models work. I see in ubuntu threads that the Zoom 3095 has needed the DGC driver. If that is still true, and any of you have such a modem, could you test precise without- and with these upgraded dgcmodem packages?

In this upgrade for kernel 3.x, I have made minor corrections, one of which you might notice. Previously, a ttyACMx modem might not be detected during first boot-up, resulting in the need to tick a box during setup to force detection of the ttyACMx modem. Now the ttyACMx should always appear as detected on the first boot of a new puppy environment. The old behavior was a mistake on my part due to confusion over the addition of the hotplug script.

Another correction relates to the HSO modems, but makes only a subtle difference, if any. During hotplugging, the HSO modem should now be detected before any other that might also be plugged in. This matches the boot-time detection priority. (Since I cannot test this, I updated only the order of probing in 2 variable strings, and moved the HSO code section to first in the section group, eliminating the need for a "flag".)

The third correction maintains DGC-modem functioning as it was, compensating for a change in data available to the DGC initialization script and corresponding pupdial...hotplug... section. A message needed for detection of a DGC modem is no longer present in the message log, making the Dgcmodem init script ineffective. In fact, I suspect that it has not been effective for a long time, because the Cdcacm script executed before it and always found the modem. Therefore, I removed the Dgcmodem script altogether, along with its pupdial...hotplug... counterpart. The dgcmodem pinstall script now simply copies the Cdcacm script into /etc/init.d from the cdcacm firmware/tarball directory, avoiding maintenance of a duplicate script.

I hope that these packages as well as the HSF set can be considered for woof. Tester feedback might build confidence in these upgraded packages.
Richard
Attachments
rerwin_woof_fixes-delta-3d.pet
Updated pupdial_init_hotpluggable script, to match DGC changes
(3.69 KiB) Downloaded 485 times
dgcmodem-1.13-k3.2.28.pet
Driver module only
(8.24 KiB) Downloaded 491 times
dgcmodem-1.13-adapted-k3.tar.gz
Source, updated for kernel 3.x, at least up to precise's kernel.
(72.5 KiB) Downloaded 474 times
Last edited by rerwin on Wed 12 Sep 2012, 00:31, edited 1 time in total.

majorfoo
Posts: 448
Joined: Mon 07 Mar 2011, 22:27
Location: Wish I knew

#574 Post by majorfoo »

I have a problem trying to run xscreensaver in Precise Puppy

When I first try to start xscreensaver, I receive the xscreensaver-1.png error

If I select ok on the xscreensaver-1.png, I then receive xscreensaver-2.png error

I have installed xscreensaver in many versions of the luci/lucid series, Dpup series, Slacko series, etc and program has been working great on all of these until recently. Possible this has something to do with the latest kernel, since it will not work on Slacko-5.3.5.3 and now the latest Precise series.

Anyway, I would appreciate any help that would point me in the right direction.
Attachments
xscreensaver-1.png
(8.27 KiB) Downloaded 1395 times
xscreensaver-2.png
(7.34 KiB) Downloaded 1302 times
Last edited by majorfoo on Wed 12 Sep 2012, 00:43, edited 2 times in total.

majorfoo
Posts: 448
Joined: Mon 07 Mar 2011, 22:27
Location: Wish I knew

#575 Post by majorfoo »

Found my problem with above post.
Missing library = LibXxf86misc.so.1.1.0

Now works
Last edited by majorfoo on Fri 14 Sep 2012, 12:24, edited 1 time in total.

User avatar
rerwin
Posts: 2017
Joined: Wed 24 Aug 2005, 22:50
Location: Maine, USA

Culmination of the Broadcom support project of 2012

#576 Post by rerwin »

Barry, Broadcom wifi users,
After about 8 months of experimentation, I (with much help from peebee, stu91 and others) have arrived with a "puppy infrastructre" solution to the various issues around Broadcom 43xx wifi cards. It originated with Piratesmack's concept of unloading then reloading conflicting drivers to ensure that competing modules are always loaded in a particular sequence, to avoid interference among them. For the setup to be ready for automated network setup, the activity must be accomplished very early during module loading. The idea was tested thoroughly using complex modprobe configuration files to perform the reloads at the time a priority module is modprobed. The attached solution moves most of the config-file logic into the module loader, pup_event_backend_modprobe.

The logic's inclusion in backend_modprobe makes it available for all conflicting modules represented by preference specifications in MODULESCONFIG, to allow any combination of a preference set to be active but always loaded in priority order. This is appropriate for cases where modules' device coverage overlaps but not completely, to allow both (preferred and nonpreferred) devices to operate. In testing, I found a case where it also ensures module order during initial use of a device that requires extensive firmware/pinstall setup -- the dgcmodem driver is intended to load before cdc_acm but is preceded by cdc_acm while the pinstall processing occurs, which the "reload" logic now corrects.

The reload logic should be a "last resort" for ensuring module load order. To minimize the need for it, a 1-second delay is added to the loading of a "nonpreferred" module to allow time for a possible preferred counterpart to load. To allow both to load, the preference-determination code is changed to avoid session-permanent blacklisting of nonpreferreds. All of this should reduce instances of rare-but-unexpected interference among modules during startup.

The Broadcom wifi devices use modules (ssb, bcma) that get loaded very early, before the normal event-driven loading takes place, leaving no option other than reloading them when necessary (when wl loads).

To track special actions taken for modules related to "preferences", the activities are logged with millisecond time tags to /tmp/pup_event_backend/preferences.log. (That's how I discovered the dgcmodem-cdc_acm reversal.)

The backend_modprobe changes are accompanied by small modprobe-configuration files for b43, b43legacy, brcmsmac and wl, to handle their special needs. The included pmodemdiag diagnostic script adds the preferences log to the diag file.

This package should be tried for any case where a module-preference is in effect. usually for wifi devices suported by multiple drivers. Broadcom brcm43xx users can install peebee's wl driver and verify that it works as well as it did with the "experiments". Instructions for obtaining and installing peebee's "bare-bones" multi-kernel driver packages are here:
http://murga-linux.com/puppy/viewtopic. ... 5&start=36

I hope that after some testing and validation by interested parties this package can be considered for addition to woof. Thank you to all who have assisted in getting this implementation to this point, and to those who now use it.
Richard

How does this work with the preference list? ADDED:
  • - To determine driver priority for a particular device, the preference list (PREFLIST in MODULESCONFIG) is scanned to find an entry for the "detected" module. The scan has depended on a single entry for each module with alternatives -- multiple entries could cuse an unexpected result. This upgrade enforces that assumption, so that any subsequent entry for the same detected module is ignored.

    - Determination of the "reload" priority uses the list "in reverse". It is effectively scanned right-to-left for all instances of a module as the most preferrred. The modules included in each such preference entry are candidates for being reloaded if already loaded. Although this may require entries specifically for the "reload" operation, the current preferences will trigger a reload if necessary.

    - For the Broadcom case, the preferences are "ssb:wl bcma:wl ssb:bcma", because wl is preferred over both ssb and bcma. While the "ssb:bcma" has not been necessary because no conflict exists between them, it indicates that if both are loaded, bcma must load first. The "reverse scan" for wl will find bcma, then ssb, to be loaded in that order; the scan for bcma will find ssb.

    - Considering that the user could inadvertently change the order of the preference entries and change the reload sequence, that risk could be reduced by limiting the reverse-scan to only the last entry specifed, forcing the full sequence to be contained in the last preference for (say) wl. The preference entries would then be: "bcma:wl ssb:bcma:wl", again with the "ssb:bcma" part never being effective in preference determination, only for reloads. To make this change, line 244 should include "| tail -n 1 " to become:

    Code: Select all

    NONPREFS="$(echo -n "$PREFLIST" | tr ' ' '\n' | grep -w -o ".*:${MODULE}$" | tail -n 1 | sed -e 's/:[^:]*$//' | tr : '\n')"
UPDATE 9/13/2012: Replaced the fix package with corrected version 3f. Those of you who downloaded fix-3e, please replace it with this version. Otherwise, module-preference processing will not work. I inadvertantly omitted a critical argument during merging of the mods into the latest version of the backend_modprobe script.

Update 9/14/12: Re-uploaded 3f after downloads #4 and #1 (diff) to add fix-ID annotation to corrected statement. No functional change.
Attachments
rerwin_woof_fixes-delta-3f.pet
pup_event_backend_modprobe with extended preference processing.
Configuration files for Broadcom modules.
(7.93 KiB) Downloaded 488 times
diff-pup_event_backend_modprobe-3f.txt.gz
Difference file
(1.96 KiB) Downloaded 457 times
Last edited by rerwin on Sat 15 Sep 2012, 22:19, edited 5 times in total.

User avatar
James C
Posts: 6618
Joined: Thu 26 Mar 2009, 05:12
Location: Kentucky

#577 Post by James C »

Manual frugal install an an old P4 test box. All basics working and correct on initial boot.


# report-video
Report Video 1.3 - Tue 11 Sep 2012 on Precise Puppy 5.3.93 - Linux 3.2.28 i686

Chip description:
2.0 VGA compatible controller: Intel Corporation 82865G Integrated Graphics Controller (rev 02)
oem: Intel(r)865G Graphics Chip Accelerated VGA BIOS
product: Intel(r)865G Graphics Controller Hardware Version 0.0

X Server: Xorg Driver used: intel

X.Org version: 1.11.3
dimensions: 1024x768 pixels (270x203 millimeters)
depth of root window: 16 planes


...the above also recorded in /tmp/root/ as report-video,
and archived with xorg.conf and Xorg.0.log as report-video-full.gz
# glxgears
Running synchronized to the vertical refresh. The framerate should be
approximately the same as the monitor refresh rate.
376 frames in 5.0 seconds = 75.020 FPS
376 frames in 5.0 seconds = 75.137 FPS
376 frames in 5.0 seconds = 75.143 FPS
376 frames in 5.0 seconds = 75.148 FPS
376 frames in 5.0 seconds = 75.143 FPS
376 frames in 5.0 seconds = 75.142 FPS

-Computer-
Processor : Intel(R) Pentium(R) 4 CPU 2.80GHz
Memory : 512MB (149MB used)
Operating System : Unknown distribution
User Name : root (root)
Date/Time : Tue 11 Sep 2012 11:09:14 PM CDT
-Display-
Resolution : 1024x768 pixels
OpenGL Renderer : Mesa DRI Intel(R) 865G x86/MMX/SSE2
X11 Vendor : The X.Org Foundation
-Multimedia-
Audio Adapter : ICH4 - Intel ICH5

User avatar
Ray MK
Posts: 774
Joined: Tue 05 Feb 2008, 09:10
Location: UK

#578 Post by Ray MK »

Working very well on my 10yr old Acer laptop - thanks for a luvly Puppy.
[b]Asus[/b] 701SD. 2gig ram. 8gb SSD. [b]IBM A21m[/b] laptop. 192mb ram. PIII Coppermine proc. [b]X60[/b] T2400 1.8Ghz proc. 2gig ram. 80gb hdd. [b]T41[/b] Pentium M 1400Mhz. 512mb ram.

tiangeng
Posts: 73
Joined: Thu 23 Jul 2009, 04:23
Location: æ²³å￾— China

Re: bold or normal

#579 Post by tiangeng »

shinobar wrote:What happened on your english fonts?
Image
Image

LateAdopter
Posts: 361
Joined: Fri 27 May 2011, 17:21
Location: Reading UK

#580 Post by LateAdopter »

Hello

There are a couple of kernel modules (cpuid & msr) that are not included in this puppy.

I would like to compile them, but I have never compiled anything before...

The source seems to be present in: /usr/src/linux-3.2.28/arch/x86/kernel

I had a look at m-a but I couldn't work out how to tell it what to build.

I'm sure it should only take a few lines of gibberish, typed into a terminal, to do it.

Would an expert tell me what they are, please?

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#581 Post by Sage »

Any chance some kind soul could add Netgear FA311revD2 NIC to the list, please?

User avatar
pemasu
Posts: 5474
Joined: Wed 08 Jul 2009, 12:26
Location: Finland

#582 Post by pemasu »

Sage. That NIC uses natsemi.ko module.
/lib/modules/3.2.28/kernel/drivers/net/ethernet/natsemi/natsemi.ko

So....the driver has been included but it seems to have reputation of only partial support to the devices it should support. I am afraid there isnt solution. Or....someone should create better driver...and since that card seems to be at least 10 years old...the support might be not so expected.
The scyld seems to be the organization behind that support.

Modinfo in my dpup, but the natsemi info is generic.
# modinfo natsemi
filename: /lib/modules/3.4.2-dpup/kernel/drivers/net/ethernet/natsemi/natsemi.ko
license: GPL
description: National Semiconductor DP8381x series PCI Ethernet driver
author: Donald Becker <becker@scyld.com>
alias: pci:v0000100Bd00000020sv*sd*bc*sc*i*
alias: pci:v0000100Bd00000020sv000012D9sd0000000Cbc*sc*i*
depends:
intree: Y
vermagic: 3.4.2-dpup SMP mod_unload modversions CORE2
parm: mtu:DP8381x MTU (all boards) (int)
parm: debug:DP8381x default debug level (int)
parm: rx_copybreak:DP8381x copy breakpoint for copy-only-tiny-frames (int)
parm: dspcfg_workaround:DP8381x: control DspCfg workaround (int)
parm: options:DP8381x: Bits 0-3: media type, bit 17: full duplex (array of int)
parm: full_duplex:DP8381x full duplex setting(s) (1) (array of int)

feelsub
Posts: 39
Joined: Wed 20 Jun 2012, 17:29

#583 Post by feelsub »

Thanks Bary
beta 6 works nearly fine, but

audio bugs:
1/ http://murga-linux.com/puppy/viewtopic.php?p=651953
2/ http://murga-linux.com/puppy/viewtopic.php?t=80580
3/ enhancement suggestion
http://murga-linux.com/puppy/viewtopic.php?t=79829

Non US char errors in either copying or reading either on FAT or NTFS, no issues under WinXP or Win7
File name is not a vaild UTF-8. You should rename it
In Rox these files/directories appear in red colour.
Names includes for instance : é è à ç or &
For instance, GNOME Player can't access those files.

Suggestion for laptop: a pop-up window when battery reach 9% and 3%, because I experienced the case to have a black-out (you don't necessarly notice the red battery indicator flashing).

I'll try again on next release.

Thanks

User avatar
pemasu
Posts: 5474
Joined: Wed 08 Jul 2009, 12:26
Location: Finland

#584 Post by pemasu »

Names includes for instance : é è à ç or &
For instance, GNOME Player can't access those files.
I have a testfile called ææøøøääööÅÅååÄÄ.JPG in ntfs partition.

It shows right in Precise Puppy 5.3.93 and also opens with Viewnior.

It looks like that it is accented character problem, not non US character problem.

For laptop usage with verbose warnings at certain battery state, vattery-acpitool with acpitool is the way to go...for instance. I had it in Upup Precise.

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#585 Post by Sage »

pemasu : many thanks for that info. Dug it out of the depths of a rather old box!

Sage
Posts: 5536
Joined: Tue 04 Oct 2005, 08:34
Location: GB

#586 Post by Sage »

Oh dear - no legacy drivers in ALSA for ISA cards any longer. Is this a kernel thing?

User avatar
666philb
Posts: 3615
Joined: Sun 07 Feb 2010, 12:27
Location: wales ... by the sea

#587 Post by 666philb »

I didn't see one posted so here's my attempt at the 'amd fglrx' driver for precise.
I think i got everything and it seems to work.
I used the slacko driver as a template for making the .pet(includes the slacko pinstall / punstall scripts).

you can download here http://www.mediafire.com/file/x5vqjnski ... 3.2.28.pet 54mb
MD5 bf40a19065f7f97626d1e1027ab2212d
Bionicpup64 built with bionic beaver packages http://murga-linux.com/puppy/viewtopic.php?t=114311
Xenialpup64, built with xenial xerus packages http://murga-linux.com/puppy/viewtopic.php?t=107331

artsown
Posts: 403
Joined: Wed 12 Sep 2012, 18:35

#588 Post by artsown »

Posting as a reply to Barry's Beta 6 ....

Beta 6 destroys the wireless capability I've enjoyed with previous versions.
I've tried both the simple connection and the complex setup. Neither result
consistent connects. If I do get a "successful connect" report I sometimes
can't browse ... it behaves like a DNS lookup failure.

My old Linksys router does not offer WPA2 ... only the WPA Personal that
it's been set at ok for many years.

User avatar
pemasu
Posts: 5474
Joined: Wed 08 Jul 2009, 12:26
Location: Finland

#589 Post by pemasu »

666philb. Thank you. Your ATI driver fglrx works great for me, with my difficult dual graphics. Usually I dont get stable graphichs and I need hard poweroff at sometime due to freezing. Now it seems to behave....and I get HDMI video outputted to my lcd TV and also....HDMI audio using ATI inbuild audio system !!! First time in my Puppy history I get HDMI audio...
Not bad....well...it means great.

Now I need to remaster this...which I usually dont do at all...so that I have this great HDMI friendly setup made permanent. I have already umplayer, smplayer, vlc, qt-4.8.2 installed.
Very promising.

Touchpad has the usual drunkenness as it has had with this Xorg version all the time. But wireless mouse should work ok.

User avatar
mikeslr
Posts: 3890
Joined: Mon 16 Jun 2008, 21:20
Location: 500 seconds from Sol

Frisbee on beta6 working well

#590 Post by mikeslr »

Frisbee on beta6 is working well.

Thanks peebee.

mikesLr

Post Reply