Lucid Puppy 5.1 Bug Reports and Fixes

[DaveS]

Found this in /tmp/bootsysinit.log:
/etc/init.d/start_cpu_freq: line 9: [: too many arguments

Does that error happen because of a dual-core processor?

Nah... happens in dysfunctional homes

[kevin bowers]

This is the first time I've actually used this installation for something serious, and I'm now in a panic. This is a .doc file, a government download that I filled in with OO.Org writer and needed to print as-is. swriter exhibits exactly the same symptoms as ePDFView when it won't print: the printer spits out one blank page no matter how many pages the document has.

Again, a .doc created within a Puppy prints fine. I tried changing the permissions on the file; I tried saving it under a different filename. Still just one blank page for a 10-page document.

I did finally get it printed: export to PDF, import to GIMP, open pages as images (not layers), print each image. Laborious to say the least, and 10 GIMP image windows fill up even a large screen in a hurry.

I think I'm finally convinced to take rcrsn51's advice and install the earlier version of CUPS over this one, have to look up his post again. A high-powered word processor isn't much good if it won't print!

[rcrsn51]

Before you do that, try the fix from here posted today at 3:16 PM

It makes sense that any external document, not just a PDF, would have trouble printing until the font problem is resolved.

[kevin bowers]

From just a quick test, I think you can consider the font problem resolved. The .doc file I had trouble with printed with no problems, except that for some reason swriter had to do a "recovery" on the file. My fill-ins even seem intact. ePDFView printed the acrobat.pdf test file but sliced off the guy's necktie and the Adobe logo at the bottom. I don't think this is a problem, I think I have my printer set to preserve dimensions rather than fit-to-page-margins.

rcrsn51, you're a life-saver. Or at least a Puppy-saver!

BTW, I'm coming to believe as you do that this newest CUPS is not the best ever. Its feature of automatically discovering network printers seems to be its main advantage, and that isn't reliable!

Thanks for the good work once again. Couldn't have come at a better time either!

Perhaps this one is big enough to warrant including it in the next update?

[rcrsn51]

What did you do - downgrade CUPS or install the font patch?

[kevin bowers]

Sorry I didn't make that clear. Just installed the font patch. Also installed gsfonts-8.11-lup.pet, figured it couldn't hurt.

[rcrsn51]

To test this properly, you need to install just the font patch, since the larger gsfonts package duplicates it.

Could you start a fresh frugal install (puppy pfix=ram) and test your doc file again?

[kevin bowers]

rcrsn51,

OK, here you go. Fresh frugal install on same ext4 partition, added the proper lines to menu.1st. Boot it up, install update 3. Reboot, create 1 GB save file. Install printer. Check ePDFView, as expected prints blanks. Install OpenOffice, also prints blanks. Install Fontmap-8.11-lupu, reboot. ePDFView now prints Acrobat.pdf fine; OpenOffice now prints my government .doc file fine with my edits.

"A's well, a's well, an' all manner of things is well!" Was it Pip that said that? I hate to admit how long ago it was that I read Dickens.

Once again thanks!

[rcrsn51]

Excellent. The real credit goes to shinobar and you might want to get his improved version that is posted in the other thread. It provides better substitutions for some of the missing fonts. You can just install it over top.

[edoc]

Before you do that, try the fix from here posted today at 3:16 PM

It makes sense that any external document, not just a PDF, would have trouble printing until the font problem is resolved.

Just installed the font-patch PET and PDF is working fine.

Lucid 5.1 with 003 update Frugal ext3 (... and a partridge in a pear tree ...)

THANKS!

[8-bit]

I have been trying to figure out why I am getting a file system check of my ext2 40gig partition even though I do not have a check as an option when I boot. It only happens some time, but the sometime is most of the time.
It also checks the lupusave.2fs file immediately afterward.
I booted from CD with pfix=ram, set locale, time zone, and then did an e2fsck of the partition.
It says the file system was not cleanly unmounted, check forced.
But when the check is done, it finds 5 large files, 2.7% non-contiguous files.
Then if I run the check again, it reports clean.
But on rebooting to the hard drive frual install, a file system check is done again.
Someone, I think tasmod, said that it may be due to an SFS or ISO I mounted and did not unmount before rebooting setting up a flag.
Any Idea where I can find that flag so I can clear it?
With multiple frugal installs on that partition, and some frugal installed removed, I really do not want to have to reinstall everything with all the customizations made.
Right now, I am booting all lupu versions with a modified initrd.gz file taken from Lighthouse Pup version D.
Lighthouse, Puppy 431 SCSI, Puppy 421, and Puppy 412 all boot without the file system check taking place unless I specify it to happen.
Also, in case you should ask, the 40gig partition has 4.4gig free on it.
I have 2 each 1.7 gig save files as well as a 2gig save file on different versions of Puppy.

Also, my clocksource is reported as unstable by Puppy cpu scaling and it is reported that the Nvidia video driver taints the kernel.
The only problem I have had prior to this was a failed attempt to resize my pupsave file in Puppy 5.0.
A second attempt with a backed up pupsave file dropped into the directory succeeded.
So any suggestions short of a complete reinstall of everything would be welcome.
Thank you in advance.

[8-bit]

I have booted luci-204 a number of times and it does not do a filesystem check on booting.
And that is even with a fsckerr file on /mnt/home.
But lupu 5.1 does a check if that file exists and that file never gets removed as per the init script.
I am trying to understand the boot process as to how it progresses and figure out if I really have a hard drive failing or it is a software issue.
If I boot from luci-204 on the hard drive and then reboot from a live cd with "pfix=ram" my boot partition checks clean.
But if I boot from lupu 500 or lupu 5.1 from the hard drive, and then boot from a live CD to check the partition, I get a message "file system not cleanly unmounted check forced".
My lupusave files always check clean on both versions.

I realize that not everyone has the file system check running on booting, so is this a combination of hardware problem for me?

When I boot, my old pci scsi card displays a message, waits for scsi drives to spin up, and then takes me to a Grub for dos menu with 2 selections of windows and puppy.
I select Puppy and a selection menu is shown with menu.lst located on the 3rd partition of a second hard drive.
From there, I select the Puppy version I want to run.
Also, I shrank an NTFS partition on the second drive, created a new ext2 formatted partition from the blank area, created a directory for a Puppy version, copied the 3 files needed from a mounted ISO file, unmounted the iso file, and added an entry for it in menu.lst.
That one does not boot from the menu with an error 2 message from grub of "Bad file or directory".
But a frugal on the NTFS formatted partition booted fine.

And since no one else seems to be able to duplicate my problem, help is only in the form of suggestions.

[8-bit]

I may have found my problem, but how do I remedy it?
To further explain, if I do a pfix=fsck on booting and then check the /initrd/tmp/chkret file, it says the last time the system was used is in the future.

So is this causing a file system check to be flagged for bootup?
My /initrd/tmp/chkret file looks like this:

Code: Select all

Performing filesystem check on /dev/sdb3...
Superblock last write time (Sun Sep  5 07:28:43 2010,
	now = Sat Sep  4 01:30:18 2010) is in the future.
Fix? yes

/dev/sdb3 was not cleanly unmounted, check forced.
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 3A: Optimizing directories
Pass 4: Checking reference counts
Pass 5: Checking group summary information

/dev/sdb3: ***** FILE SYSTEM WAS MODIFIED *****
/dev/sdb3: 11997/5128192 files (2.7% non-contiguous), 9374888/10238130 blocks
Performing filesystem check on /dev/sdb4...
Superblock last mount time (Sun Sep  5 07:22:33 2010,
	now = Sat Sep  4 01:31:57 2010) is in the future.
Fix? yes

Superblock last write time (Sun Sep  5 07:22:44 2010,
	now = Sat Sep  4 01:31:57 2010) is in the future.
Fix? yes

/dev/sdb4: clean, 15/3203072 files, 247402/12800970 blocks
Superblock last mount time (Sun Sep  5 07:28:42 2010,
	now = Sat Sep  4 01:31:57 2010) is in the future.
Fix? yes

Superblock last write time (Sun Sep  5 07:28:42 2010,
	now = Sat Sep  4 01:31:57 2010) is in the future.
Fix? yes

Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 3A: Optimizing directories
Pass 4: Checking reference counts
Pass 5: Checking group summary information

/dev/loop1: ***** FILE SYSTEM WAS MODIFIED *****
/dev/loop1: 3812/65536 files (7.4% non-contiguous), 72249/262144 blocks

[tasmod]

OK, I did a pfix=fsck for you on my system.

Here's the report.

Code: Select all

Superblock last write time (Sun Sep  5 13:15:40 2010,
	now = Sat Sep  4 15:16:24 2010) is in the future.
Fix? yes

Pass 1: Checking inodes, blocks, and sizes
Deleted inode 18 has zero dtime.  Fix? yes

Deleted inode 21 has zero dtime.  Fix? yes

Deleted inode 27 has zero dtime.  Fix? yes

Deleted inode 31 has zero dtime.  Fix? yes

Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
Pass 3A: Optimizing directories
Failed to optimize directory /tmp (58929): Invalid argument
Pass 4: Checking reference counts
Pass 5: Checking group summary information
Block bitmap differences:  -(5633--5636) -(5641--5676) -(5681--5695) -5697 -(5699--5701) -(5705--5751) -5762 -5764 -(5766--5776) -(5790--5795) -5806 -5809
Fix? yes

Free blocks count wrong for group #0 (4572, counted=4699).
Fix? yes

Free blocks count wrong (824067, counted=824194).
Fix? yes

Inode bitmap differences:  -18 -21 -27 -31
Fix? yes

Free inodes count wrong for group #0 (2001, counted=2005).
Fix? yes

Free inodes count wrong (246223, counted=246227).
Fix? yes


/dev/loop1: ***** FILE SYSTEM WAS MODIFIED *****
/dev/loop1: 1677/247904 files (7.6% non-contiguous), 170310/994504 blocks

[tasmod]

OK, cursory checks for now. these appear to be the two pieces of code that checks and sets for a file system check.

/etc/rc.d/rc.sysinit

Code: Select all

#100316 improper shutdown check. see above, also rc.shutdown and /sbin/init...
if [ ! -d /initrd ];then #if no /initrd, full hd install.
 [ ! $PDEV1 ] && PDEV1="`df | grep ' /$' | grep '^/dev/' | cut -f 1 -d ' ' | cut -f 3 -d '/'`"
 [ ! $DEV1FS ] && DEV1FS="`mount | grep ' on / ' | grep '^/dev/' | cut -f 5 -d ' '`"
 if [ "$PDEV1" -a "$DEV1FS" ];then
  [ "`grep '^PDEV1' /etc/rc.d/PUPSTATE`" = "" ] && echo "PDEV1='$PDEV1'" >> /etc/rc.d/PUPSTATE
  [ "`grep '^DEV1FS' /etc/rc.d/PUPSTATE`" = "" ] && echo "DEV1FS='$DEV1FS'" >> /etc/rc.d/PUPSTATE
  #this file gets removed by rc.shutdown if a proper shutdown...
  echo -n "${PDEV1},${DEV1FS}," > /fsckme.err #ex: sda7,ext3,
 fi
else
 [ "$PUPSAVE" ] && echo -n "$PUPSAVE" > /initrd${PUP_HOME}/fsckme.err #ex: sda7,ext3,/pupsave-10MAR2010.2fs
fi

/etc/rc.d/rc.shutdown

Code: Select all

#100315 improper shutdown check. see /etc/rc.d/rc.sysinit, /init in initramfs, and /sbin/init...
[ -f /fsckme.err ] && rm -f /fsckme.err
[ -f /initrd${PUP_HOME}/fsckme.err ] && rm -f /initrd${PUP_HOME}/fsckme.err

ORIGPUPMODE=$PUPMODE #v2.22

echo "Executing /etc/rc.d/rc.shutdown..."

the other place you are sent to is /sbin/init which is where the check is done. In particular the file fsck.err is removed by this line when check is done.

Code: Select all

 rm -f /fsckme.err

[8-bit]

tasmod,
I had been selecting US Pacific as my locale or time zone.
I changed that to GMT -8 and guess what.
I am now getting clean shutdowns and no file system check on booting unless I specify one to be done.
So is this a bug as to selecting US Pacific instead of GMT -8?
Anyway, I am crossing my fingers and toes and hoping that was the problem.

[edoc]

That is interesting as somewhere I recall something recommending the use of the area e.g. "US Pacific Coast" rather than the fixed "GMT-6".

I am currently using "US East Coast" with no problems lke yours.

I am seeing some odd crashing problems that have the feel of a memory leak.

I will also be interested in the informed recommendation for this.

[dawnsboy]

I have read reports in this thread about Lucid not saving wireless network settings between boots. I have a full hd install of Lucid Puppy 5.1 on my desktop and connect to the internet with an ethernet card.

Lucid Puppy 5.1 will auto-connect to this network when booting from CD. However it will not remember the network settings between boots for frugal or full hard drive installs. Everything else works just fine so far but I have to manually connect to the internet everytime I boot the computer.

Any suggestions for getting it to remember the wired network settings?

[kevin bowers]

I have read reports in this thread about Lucid not saving wireless network settings between boots. I have a full hd install of Lucid Puppy 5.1 on my desktop and connect to the internet with an ethernet card.

Lucid Puppy 5.1 will auto-connect to this network when booting from CD. However it will not remember the network settings between boots for frugal or full hard drive installs. Everything else works just fine so far but I have to manually connect to the internet everytime I boot the computer.

Any suggestions for getting it to remember the wired network settings?

Several. There is a newer version of DHCPD available, I forgot where I downloaded it from but search for dhcpd-5.2.6-p5 in the forums and repositories in Puppy Package Manager. You didn't say what browser you were using but if it's Firefox you should set Edit>preferences>network>connection>settings to "no proxy". Seamonkey has a similar setting but is set to no proxy by default; I don't know about Opera. Try switching between Barry's Simple Network Setup and Dougal's Internet Connection Wizard. And one more: install jemimah's Pwireless2 and use it to configure your connection, I believe it now supports wired connections as well. Good luck!

[grover]

Hi, I have installed 5.1 (very nice - great work, everybody) but cannot get Google Chrome to work.

Hi Simes, Here is a pet of Iron. I had forgotten how much I liked it--it seems very snappy. The p2 version is not needed anymore, they say, because this version will run with p3. (That confused me too, but I hope it works.)

http://www.diddywahdiddy.net/Puppy500/I ... -Lucid.pet

thank you
I came looking with the same problem. Firefox has grown so sluggish that I couldn't imagine using anything else!