Page 5 of 7

Posted: Fri 12 Jul 2019, 10:22
by peebee
N.B. this update requires both initrd.gz and puppy_upupdd_19.04.sfs replacing

28-jul-2019

56e47951e118452d1cbad062cf782845 upupdd-19.04+13-uefi.iso and delta updates

Posted: Sat 13 Jul 2019, 02:16
by Marv
My 19.04 +10 upupdd Grub4Dos install to SSD on the i5 based Fujitsu S761 updated to 19.04 +11 uneventfully. I was running with the 64b lxde 5.2.0 kernel in +10 and retained that in +11. Save file scrubbed and checked as usual. Dmesg failures or warnings examined and no changes found. CPU % and mem use at idle unchanged, video FPS excellent, SSD/memory throughput unchanged and only slightly slower than LxPupSc64 on this hardware.

Early loading of microcode from a shared folder continues to work correctly and mitigation on the second gen i5 (with the 5.2.0 kernel) is:

Code: Select all

l1tf:Mitigation: PTE Inversion
mds:Mitigation: Clear CPU buffers; SMT vulnerable
meltdown:Mitigation: PTI
spec_store_bypass:Mitigation: Speculative Store Bypass disabled via prctl and seccomp
spectre_v1:Mitigation: __user pointer sanitization
spectre_v2:Mitigation: Full generic retpoline, IBPB: conditional, IBRS_FW, STIBP: conditional, RSB filling
Thanks for carrying on with upupdd. It's marvelously snappy on this hardware.

Posted: Sat 13 Jul 2019, 12:47
by peebee
+11 didn't last very long :cry:

+12 has woof-ce updates + Radky app updates

first boot problem

Posted: Wed 17 Jul 2019, 20:02
by foxpup
I have tried out 19.04+12 on a i5 machine.
I tried to boot from fugal install on ssd with efi/grub2 and from usb frugal install with grub4dos.

On both I had the same problem:
If there is not yet a pupsave, the boot stops just before X.
So I have to boot nox on the first boot, then go to X in a second step, then make a pupsave on first shutdown.
After that, everything goes well, also with multiple pupsaves.

Re: first boot problem

Posted: Wed 17 Jul 2019, 20:41
by peebee
foxpup wrote:I have tried out 19.04+12 on a i5 machine.
On both I had the same problem:
If there is not yet a pupsave, the boot stops just before X.
Hi foxpup

I don't see that problem.........what graphics do you have? - I have Nvidia on my desktop and Intel on my laptop and both are OK.......

Not sure what to suggest............

Re: first boot problem

Posted: Wed 17 Jul 2019, 21:21
by sheldonisaac
peebee (in part) wrote:
foxpup wrote:I have tried out 19.04+12 on a i5 machine. On both I had the same problem:
If there is not yet a pupsave, the boot stops just before X.
I don't see that problem.........what graphics do you have? - I have Nvidia on my desktop and Intel on my laptop and both are OK...
Oh, I will boot 19.04+12 on this laptop with Intel and report.
----------------------------
Booted without the pupsave, got to a GUI desktop OK.

Thanks,
Sheldon

Posted: Wed 17 Jul 2019, 21:32
by foxpup
Hello peebee.

I have intel card and nvidia graphic card on this machine. It is Optimus machine, but Puppy just uses the intel integrated gc.

I don't know what the reason for this can be, but thought I should just mention it.
There has been changes on the init in woof, so maybe I should try an earlier version with significantly different init. Can you suggest such version?

Re: first boot problem

Posted: Wed 17 Jul 2019, 22:59
by Marv
foxpup wrote:I have tried out 19.04+12 on a i5 machine.
I tried to boot from fugal install on ssd with efi/grub2 and from usb frugal install with grub4dos.

On both I had the same problem:
If there is not yet a pupsave, the boot stops just before X.
So I have to boot nox on the first boot, then go to X in a second step, then make a pupsave on first shutdown.
After that, everything goes well, also with multiple pupsaves.
I'm running a Grub4Dos frugal install to SSD of 19.04 +12 on the second gen i5 hardware here:

Code: Select all

System:    Host: puppypc25146 Kernel: 5.2.1-lxpup64 x86_64 bits: 64 Desktop: JWM 2.3.7 Distro: UPupDD 19.04
Machine:   Device: laptop System: FUJITSU product: LIFEBOOK S761 serial: R1Y00453
           Mobo: FUJITSU model: FJNB225 serial: N/A
           UEFI [Legacy]: FUJITSU // Phoenix v: Version 1.17 date: 03/14/2012
Battery    CMB1: charge: 72.4 Wh 100.0% condition: 72.4/72.4 Wh (100%)
CPU:       Dual core Intel Core i5-2520M (-HT-MCP-) speed/max: 811/3200 MHz
Graphics:  Card: Intel 2nd Generation Core Processor Family Integrated Graphics Controller
           Display Server: X.Org 1.20.4 drivers: modesetting (unloaded: fbdev,vesa) Resolution: 1366x768@60.00hz
           OpenGL: renderer: Mesa DRI Intel Sandybridge Mobile x86/MMX/SSE2 version: 3.3 Mesa 19.0.2
Network:   Card: Qualcomm Atheros AR9287 Wireless Network Adapter (PCI-Express) driver: ath9k
Drives:    HDD Total Size: NA (-)
Info:      Processes: 147 Uptime: 0 min Memory: 212.1/5794.3MB Client: Shell (bash) inxi: 2.3.40 
I hid my savefile and did a boot both with the stock kernel and with the 64b 5.2.1 kernel that I have been running (also in the current LxPupSc and LxPupSc64). Both instances booted normally to X. My Grub4Dos menu entry for upupdd is:

Code: Select all

title DingoPup 19.04 +12 with 5.2.1 64b kernel  
  uuid 2988-C05D
  kernel /DingoPup/vmlinuz pmedia=ataflash pdev1=sda1 psubdir=DingoPup pfix=copy,fsck ipv6.disable=1
  initrd /microcode/microcode-update-20190514a.cpio /DingoPup/initrd.gz 
Unhid my pupsave and rebooted to post cause I'm lazy. I've been running 19.04 +12 as my daily since it posted. Mundane use but no issues here.

Could it be the presence of two video cards?

Just to note, I don't have the adrv in that boot folder as I don't use it and forgot to put it in for the test. Shouldn't think that would be the issue but...

Posted: Sun 28 Jul 2019, 18:28
by peebee
N.B. this update requires both initrd.gz and puppy_upupdd_19.04.sfs replacing

28-jul-2019

56e47951e118452d1cbad062cf782845 upupdd-19.04+13-uefi.iso and delta updates

check savefile

Posted: Tue 30 Jul 2019, 20:41
by sszindian
Just curious...

Why is the 'savefile check' function missing in startup selection menu in the last few updates of dd-pup? and... what actually does it do say, a savefile has something attached or added that shouldn't be there... I don't ever recall anyone in the past explaining its function and worth?

Thanks

>>>---Indian------>

Re: check savefile

Posted: Mon 05 Aug 2019, 07:49
by peebee
sszindian wrote:Why is the 'savefile check' function missing in startup selection menu in the last few updates of dd-pup?
Woof-CE changes on 27 June:
https://github.com/puppylinux-woof-CE/w ... ef2843f38a
update bootloader menu

remove "check filesystem" item. all items include pfix=fsck
for ext3 and ext4 this isn't a problem, it quickly checks and continues (0.1 secs)

Posted: Mon 05 Aug 2019, 15:00
by watchdog
How to download a package and all dependencies to make an sfs?

Posted: Mon 05 Aug 2019, 15:29
by peebee
watchdog wrote:How to download a package and all dependencies to make an sfs?
I use createsfs - but you have to obtain all the dependencies (which can be pets, debs, txz or other sfs) yourself....

http://www.murga-linux.com/puppy/viewtopic.php?p=272291

Code: Select all

#!/bin/sh
#
# createsfs - A script to create .sfs files from .pet/.sfs files
#
# Short Description:
# 
# Its a simple script which will untar .pet or unsquash .sfs files
# and copy those files to a temp directory. Finally it will create a
# new .sfs file from that temp directory thats all
#
# License:
# Use it whatever way you want. If anything you want to share with 
# me, send a mail to mohan43u@gmail.com

# Defining variables

USAGE="
[USAGE]

	createsfs -f sfsfile [-r XXX] pet_or_sfs_file1 [pet_or_sfs_file2 ...]
	
[DESCRIPTION]

	-f sfsfile		Final SFS file to be created
	-r XXX			Three digit puppy version number (current version file: /etc/puppyversion)
	pet_or_sfs_file		Source .pet or SFS file
"

# Parsing Options

while getopts "f:r:h" OPTION
do
	case "${OPTION}" in
	f)FINALSFS="${OPTARG}";;
	r)SFSVERSION="${OPTARG}";;
	h)echo "${USAGE}" && exit 0;;
	\?)echo "${USAGE}" && exit 1;;
	esac
done

# Check for final sfs filename

test -z "${FINALSFS}" && \
echo "Error: No name given for the final sfs file" && \
echo "${USAGE}" && exit 1

if test -z "${SFSVERSION}"
then
	FINALSFS="${FINALSFS}_`cat /etc/puppyversion`"
else
	FINALSFS="${FINALSFS}_${SFSVERSION}"
fi
echo "Final SFS filename: ${FINALSFS}"
mkdir -p "${FINALSFS}"

shift $((OPTIND - 1))

# Check for atleast one .pet of .sfs file

test -z "${*}" && \
echo "Error: At least one .pet/.sfs file needed" && \
echo "${USAGE}" && exit 1

for FILE
do
	# Checking for source file present in present directory
	FILE=`basename "${FILE}"`
	test ! -f "${FILE}" \
	&& echo "${FILE} not found in current directory." \
	"Move to the directory where ${FILE} present" \
	"and run this command." \
	"exitting.." && rm -fr "${FINALSFS}" && exit 1

	# Handling .pet file
	
	if file "${FILE}" | grep -i "gzip compressed"
	then
		mv "${FILE}" "${FILE%.pet}.tar.gz"
		tar xzf "${FILE%.pet}.tar.gz" 2>/dev/null
		rm -f "${FILE%.pet}"/*.files "${FILE%.pet}"/*.pet.specs
		cp -R "${FILE%.pet}"/* "${FINALSFS}"
		rm -fr "${FILE%.pet}"
		mv "${FILE%.pet}.tar.gz" "${FILE}"
	fi

	# Handling .sfs file
	
	if file "${FILE}" | grep -i "Squashfs"
	then
		unsquashfs -d "${FINALSFS}_unsquashed" "${FILE}"
		cp -R "${FINALSFS}_unsquashed"/* "${FINALSFS}"
		rm -fr "${FINALSFS}_unsquashed"
	fi
	
	# Handling .deb file
	
	if file "${FILE}" | grep -i "Debian"
	then
		dpkg-deb -x "${FILE}" "${FINALSFS}_unsquashed"
		cp -R "${FINALSFS}_unsquashed"/* "${FINALSFS}"
		rm -fr "${FINALSFS}_unsquashed"
	fi
	
	# Handling .txz file
	
	if file "${FILE}" | grep -i "xz compressed"
	then
		mkdir "${FILE}_out"
		cd "${FILE}_out"
		tar xJf ../"${FILE}" 2>/dev/null
		test -f ./install/doinst.sh && chmod 755 ./install/doinst.sh && ./install/doinst.sh
		rm -r install
		cd ..
		cp -R "${FILE}_out"/* "${FINALSFS}"
		rm -r "${FILE}_out"
	fi

done

savefile check

Posted: Mon 05 Aug 2019, 18:27
by sszindian
@ watchdog

If you ever get an .sfs made up for this 'fix' I sure would appreciate a copy!

Ahem... I downloaded and printed out the file at the Github link peebee provided above but it was so darn pretty and colorful I hung it out on my flag-pole, didn't know what else to do with it :D

>>>---Indian------>

Posted: Tue 06 Aug 2019, 02:05
by watchdog
peebee wrote:
watchdog wrote:How to download a package and all dependencies to make an sfs?
I use createsfs - but you have to obtain all the dependencies (which can be pets, debs, txz or other sfs) yourself....
I have built vlc-3.0.7.1_i386_19.04.sfs which I share:

https://drive.google.com/file/d/1PTyju6 ... sp=sharing

I used an old trick. Put in /root a file .wgetrc with the following content:

Code: Select all

dir_prefix = /mnt/home/Downloads
and install step by step your package by ppm: the install fails but download the deb packages in that dir. Then I used alien2puppy (a modded version). Delete /root/.wgetrc at the end.

Posted: Wed 28 Aug 2019, 18:05
by peebee
N.B. this update needs the updated initrd.gz just like +13

28-Aug-2019

aad9512621560726cb807a95a6e02132 upupdd-19.04+14-uefi.iso and delta updates

Posted: Thu 29 Aug 2019, 11:44
by Marv
upupdd 19.04 +13 frugal Grub4Dos install on i5 based Fujitsu S761 updated to +14 using the delta. md5sum of generated iso correct, update continues to have clean boot and system checks, good memory/SSD throughput, good video FPS, clean lidsuspend, and overall wonderfully frisky behavior. Mitigation continues unchanged, I haven't had time to go fishing for more current 2nd gen i5 microcode, I'm still early loading the 20190619 code with the following mitigation running the 5.2.10 kernel:

l1tf:Mitigation: PTE Inversion
mds:Mitigation: Clear CPU buffers; SMT vulnerable
meltdown:Mitigation: PTI
spec_store_bypass:Mitigation: Speculative Store Bypass disabled via prctl and seccomp
swapgs barriers and __user pointer sanitization
spectre_v2:Mitigation: Full generic retpoline, IBPB: conditional, IBRS_FW, STIBP: conditional, RSB filling

I have run both the stock kernel and peebees 64b 5.2.10 with good results. Running and posting from +14 with the 5.2.10 kernel now.

Cheers, and Thanks,

Samsung Laptop - disappearing cursor

Posted: Wed 04 Sep 2019, 00:28
by davids45
G'day,

I have an odd start-up problem on my Samsung laptop with recent Ubuntu-derived 32-bit Pups (Disco, Bionic) but not with other Pups (Slacko, Buster, etc) nor with these Ubuntu Pups on my desktops.

Bionic/Disco starts up and the expected desktop displays with a cursor, but with or without movement, about 10 seconds later, the cursor disappears or becomes invisible.

Wiggling the mouse and desktop icons are affected by the passage of the now-invisible cursor.

Ctrl-Alt-backspace and then xwin at the prompt and Bionic/Disco restart, with a now permanent and normally working cursor.

How do I stop this odd start-up cursor disappearance in these particular Pups on just my Samsung laptop?

David S.

UPupEE preview

Posted: Sat 14 Sep 2019, 16:29
by peebee
Sneeky peek..........

Looks like Ubuntu is providing all the necessary 32-bit components to build UPupEE

Posted: Sun 15 Sep 2019, 16:54
by peebee
Pre-release....

UPupEE-19.10-PR1 + delta to PR9

md5 = 1d62714dfee465c2bc4b886ecd164263 upupee-19.10-PR9-uefi.iso