Page 9 of 9
Re: Building Blocks
Posted: 11 Feb 2016, 21:21
by Jack
I did what you ask me to do. Here is all the way I try it.
1. CD it did work
2. fat32 and didn't work
3. ext4 and didn't work
4. ntfs and didn't work
2, 3 and 4 stop in the same place.
And for my 0.02cent that don't mean I was right. It was what I was thinking it could be.
Code: Select all
guest@porteus:~$ sudo mount /dev/sdb1 /mnt/sdb1
guest@porteus:~$ ls /mnt/sdb1/porteus/
base/ modules/ optional/ ptest.sgn* rootcopy/
And if you look at Boot Info above, you will see proteus data was specified to be on /dev/sdb1 in the Kernel Command Line. Also in porteus-livedbg listing it gives Porteus Data as being found on /dev/sdb1, sdb1 being in vfat format.
I can see all the files are there on the CD plus all the USB I made for test for you.
UPDATE The only way I run or use Porteus is ext4 and I never use CD rom. I haven't use CD rom for years.
Re: Building Blocks
Posted: 11 Feb 2016, 21:49
by Bogomips
Jack wrote:I did what you ask me to do. Here is all the way I try it.
1. CD it did work
2. fat32 and didn't work
3. ext4 and didn't work
4. ntfs and didn't work
Jack, having trouble booting from usb porteus. The bios only sees floppy, but have no floppy drive. So cannot do test for booting from usb. Made usb with neko tool, and this is what was working for me, in that it could get porteus data from usb, and start. Could you try neko tool with this new iso, and get bios to boot the usb created? Thanks.
Re: Building Blocks
Posted: 11 Feb 2016, 22:02
by Jack
Bogomips wrote:Jack wrote:I did what you ask me to do. Here is all the way I try it.
1. CD it did work
2. fat32 and didn't work
3. ext4 and didn't work
4. ntfs and didn't work
Jack, having trouble booting from usb porteus. The bios only sees floppy, but have no floppy drive. So cannot do test for booting from usb. Made usb with neko tool, and this is what was working for me, in that it could get porteus data from usb, and start. Could you try neko tool with this new iso, and get bios to boot the usb created? Thanks.
I all ready try neko tool and it did the same thing as 2, 3 and 4 it didn't work.
Re: Building Blocks
Posted: 12 Feb 2016, 01:52
by Bogomips
Ed_P wrote:And all the files they search for and use begin inside the mounted ISO.
Only the boot loaders access the NTFS formatted partition. The kernel and initrd are read from the ISO and read files inside the mounted ISO file's files.
No acquaintance with grub4dos, but have some idea about Grub2.
Grub Procedure
- Kernel and Initrd Extraction
- Grub1 User
- Mounts Iso on Loop
- Extracts Kernel and Initrd to directories on same partition
- Using Menu.lst informs bootloader of their whereabouts
- Grub2 Program
- Mounts Iso on Loop
- Extracts Kernel and Initrd
- Passes pointer to them to bootloader routine
- Bootloader Routine
- Loads Kernel into Memory
- Places Initrd in a Special Memory Fille
- Grub bows out and hands over to Kernel
- Kernel decompresses Initrd
- Kernel executes Init (here I'm not sure which is which), which then executes Init Script in Initrd
- Init Script has to be able to find Porteus Data sgnfile, and only able to do so if can read filesystem on which iso resides in order to be able to loop mount the iso, in the case of isos.
Re: Building Blocks
Posted: 12 Feb 2016, 04:31
by Ed_P
Bogomips wrote:Init Script has to be able to find Porteus Data sgnfile,
To this we agree.
Bogomips wrote:and only able to do so if can read filesystem on which iso resides in order to be able to loop mount the iso, in the case of isos.
Then why does the bootloader do this?
Bogomips wrote:Places Initrd in a Special Memory Fille
And this?
Bogomips wrote:Mounts Iso on Loop
Everything is already in memory, there is no need to read the filesystem to loop the iso again.
BTW Have tried the 100216 iso, still the same problem. I even added a map of firaDisk.gz to a menu like I use to read my BartPE.iso and it didn't help either. Very strange. Maybe a missing symlink?
Re: Building Blocks
Posted: 12 Feb 2016, 17:19
by Bogomips
@ Jack Can you check this out?
- Using neko tool have formatted usb which was tested before. On the usb formatted by neko tool, see if checksums are the same as on mine?
Code: Select all
guest@porteus:~$ md5sum /mnt/sdb1/boot/syslinux/vmlinuz
c0377a5c8c07fb47be5e714c7ce3a78d /mnt/sdb1/boot/syslinux/vmlinuz
guest@porteus:~$ md5sum /mnt/sdb1/boot/syslinux/initrd.img
f5090165416f6711e0ad2f87014e567f /mnt/sdb1/boot/syslinux/initrd.img
- If they are the same, please try:
- Boot the usb (or cd if usb doesn't work)
- At the boot screen select TEXT mode
- Hit TAB key and add to Command Line: from=/dev/sdb1
- Hit Enter to boot
- If above test has worked, I would ask you to try:
- Then copy abSlak_100216.iso to /bld/ab
- Check Copy:
Code: Select all
guest@porteus:~$ ls -lh /mnt/sdb2/bld/ab/abSlak_100216.iso
-rw------- 1 guest guest 240M Feb 11 01:07 /mnt/sdb2/bld/ab/abSlak_100216.iso
- As in 2 above, add to Command Line: from=/dev/sdb2/bld/ab/abSlak_100216.iso and boot.
Thanks.
Results from #3 (after logging into darkstar):
Code: Select all
root@darkstar:~# cat /proc/cmdline
base_only noauto from=/dev/sdb2/bld/ab/abSlak_100216.iso
root@darkstar:~# cat /var/log/porteus-livedbg
# Recognized devices:
/dev/sr0: LABEL="Porteus" TYPE="iso9660"
/dev/sda1: UUID="ca6ce5e1-1bbd-4170-87f7-5fd0448e5f56" TYPE="reiserfs"
/dev/sda3: SEC_TYPE="msdos" UUID="4B46-EEEA" TYPE="vfat"
/dev/sda4: UUID="966b254d-20de-4376-b736-ddf89150a6b3" TYPE="reiserfs"
/dev/sda5: LABEL="kix" UUID="e737efd6-32f4-4a80-bbc1-a6ad3114b4d5" TYPE="ext2"
/dev/sda6: UUID="f554b881-2089-4c0f-8de2-70900bdc4e14" TYPE="ext2"
/dev/sda7: UUID="0d2c9e21-5e47-4dd3-a279-38551711f3dd" TYPE="swap"
/dev/sda8: UUID="AF61-DAE9" TYPE="vfat"
/dev/sda9: LABEL="iso" UUID="5bf4078e-3d9c-46e6-8a43-2dfc0eef3aeb" TYPE="ext2"
/dev/sda10: LABEL="p_10" UUID="ed3f74cd-26bb-47bd-86b0-5e690663dd55" TYPE="ext2"
/dev/sda11: UUID="6a0167f6-5cf7-412d-b85f-0076ca590371" TYPE="swap"
/dev/sdb1: SEC_TYPE="msdos" UUID="12D4-A2F1" TYPE="vfat"
/dev/sdb2: UUID="ec468a0e-7848-416d-9d29-3c74a1bfb486" TYPE="ext4"
# Booting device:
/mnt/sdb2
# Porteus data found in:
/mnt/isoloop/porteus
# Changes are stored in:
memory
# Non standard /rootcopy dir:
none
# Modules activated during boot time:
/mnt/isoloop/porteus/base/000-kernel.xzm
/mnt/isoloop/porteus/base/001-core.xzm
Re: Building Blocks
Posted: 12 Feb 2016, 18:13
by Jack
You need to find a computer that will boot up with USB so you can see. I know the CD dose work.
When I have time I will do what you want me to do. I have all kinds of problem at my home right now.
Re: Building Blocks
Posted: 12 Feb 2016, 23:18
by Bogomips
Ed_P wrote:Bogomips wrote:and only able to do so if can read filesystem on which iso resides in order to be able to loop mount the iso, in the case of isos.
Then why does the bootloader do this?
Bogomips wrote:Places Initrd in a Special Memory Fille
And this?
Bogomips wrote:Mounts Iso on Loop
Everything is already in memory, there is no need to read the filesystem to loop the iso again.
The way I understand it, it's like the division of labour, where there are strict union demarcated lines. The Kernel says, it's not its job to know about an iso mounted on a loop, which is somewhere in memory. The boot loader says it's my job to extract the kernel, decompress it and place it in memory together with initrd in a special memory file. Once done the only other responsibility is to unmount everything and exit. The Kernel's job at bootup is to just decompress the Initrd and invoke the init program which then executes the linuxrc script of Initrd.
Somewhat simplistic picture of my understanding of things.
Re: Building Blocks
Posted: 12 Feb 2016, 23:51
by brokenman
Tested from a USB drive and found to be working.
I copied the files onto a partition on the USB drive. Changed into the boot directory and ran the porteus installer (as per a normal porteus install). Jack may need to employ the 'delay' cheat code if it is not finding the sgn file. The USB device probably needs some time to settle.
Not sure if you are interested (or it has already been posted here) in the following but I will post it anyway in case you missed it.
line 24,36,28 in rc.M throws errors on boot.
Other boot warnings/errors include:
Can't find /home/guest/p01 in fstab
Line 63 in /etc/rc.d/rc.M /usr/sbin/acpid No such file or directory
Re: Building Blocks
Posted: 12 Feb 2016, 23:59
by Jack
brokenman wrote:Tested from a USB drive and found to be working.
I copied the files onto a partition on the USB drive. Changed into the boot directory and ran the porteus installer (as per a normal porteus install). Jack may need to employ the 'delay' cheat code if it is not finding the sgn file. The USB device probably needs some time to settle.
Not sure if you are interested (or it has already been posted here) in the following but I will post it anyway in case you missed it.
line 24,36,28 in rc.M throws errors on boot.
Other boot warnings/errors include:
Can't find /home/guest/p01 in fstab
Line 63 in /etc/rc.d/rc.M /usr/sbin/acpid No such file or directory
It finds the sgn file but in the next line it count down then I see the error. Maybe my USB thumb drive are to slow.
Re: Building Blocks
Posted: 13 Feb 2016, 05:37
by Ed_P
brokenman wrote:Tested from a USB drive and found to be working.
I copied the files onto a partition on the USB drive. Changed into the boot directory and ran the porteus installer
In the two cases where this is working, your test and the CD test, something else is involved in making the ISO work, the app used to create the CD and the Porteus install to create the USB drive. The files to boot are apparently in the ISO but the ISO itself appears not to be bootable. It's like it is just a big zip file. That's why even with the drivers
Bogomips has added to the newest ISO it doesn't boot.
Re: Building Blocks
Posted: 13 Feb 2016, 09:20
by francois
It seems that you guys are working hard here. Have you adressed the creation of the devel module necessary to build vbox guest additions? Or you are not there yet.
Is is possible to get a summary of what has been achieved yet?
Re: Building Blocks
Posted: 13 Feb 2016, 14:22
by Bogomips
^ As our friend from down under would say: do refer to 1st Post of Thread, then the link and some posts leading up to it, and after, which is the whole point of the exercise.
Jack wrote:It finds the sgn file but in the next line it count down then I see the error. Maybe my USB thumb drive are to slow.
Probably. At the boot screen you need to add perhaps
delay=12 to Kernel Command Line. Have an iffy usb so put 20, which gives enough time to pull it out and push it in again so that
Ready Light shows.
Kriss wrote:But if I can be of assistance with adding some new feature or improving something, just tell me and I'll try... =)
Edited linuxrc for overlayfs. You could certainly help with checking copy2ram, ramsize=, zram=. This is what I got:
Code: Select all
root@darkstar:~# date
Sat Feb 13 23:00:09 2016
root@darkstar:~# cat /proc/cmdline
initrd=initrd.img 3 copy2ram ramsize=30% noauto from=/dev/sdb2/bld/ab/abSlak_100216.iso zram=30% BOOT_IMAGE=vmlinuz
root@darkstar:~# cat /var/log/porteus-livedbg
# Recognized devices:
/dev/sda1: UUID="ca6ce5e1-1bbd-4170-87f7-5fd0448e5f56" TYPE="reiserfs"
/dev/sda3: SEC_TYPE="msdos" UUID="4B46-EEEA" TYPE="vfat"
/dev/sda4: UUID="966b254d-20de-4376-b736-ddf89150a6b3" TYPE="reiserfs"
/dev/sda5: LABEL="kix" UUID="e737efd6-32f4-4a80-bbc1-a6ad3114b4d5" TYPE="ext2"
/dev/sda6: UUID="f554b881-2089-4c0f-8de2-70900bdc4e14" TYPE="ext2"
/dev/sda7: UUID="0d2c9e21-5e47-4dd3-a279-38551711f3dd" TYPE="swap"
/dev/sda8: UUID="AF61-DAE9" TYPE="vfat"
/dev/sda9: LABEL="iso" UUID="5bf4078e-3d9c-46e6-8a43-2dfc0eef3aeb" TYPE="ext2"
/dev/sda10: LABEL="p_10" UUID="ed3f74cd-26bb-47bd-86b0-5e690663dd55" TYPE="ext2"
/dev/sda11: UUID="6a0167f6-5cf7-412d-b85f-0076ca590371" TYPE="swap"
/dev/sdb1: SEC_TYPE="msdos" UUID="12D4-A2F1" TYPE="vfat"
/dev/sdb2: UUID="ec468a0e-7848-416d-9d29-3c74a1bfb486" TYPE="ext4"
# Booting device:
/mnt/sdb2
# Porteus data found in:
/mnt/isoloop/porteus
# Changes are stored in:
memory
# Non standard /rootcopy dir:
none
# Modules activated during boot time:
/memory/copy2ram/000-kernel.xzm
/memory/copy2ram/001-core.xzm
root@darkstar:~# df -h .
Filesystem Size Used Avail Use% Mounted on
overlay 435M 231M 204M 54% /
root@darkstar:~# free -m
total used free shared buff/cache available
Mem: 868 27 505 236 335 578
Swap: 905 0 905
root@darkstar:~# df -h
Filesystem Size Used Avail Use% Mounted on
tmpfs 435M 6.4M 428M 2% /run
devtmpfs 431M 0 431M 0% /dev
tmpfs 435M 231M 204M 54% /mnt/live/memory
/dev/loop1 88M 88M 0 100% /mnt/live/memory/images/000-kernel
/dev/loop2 140M 140M 0 100% /mnt/live/memory/images/001-core
overlay 435M 231M 204M 54% /
/dev/sda9 77G 72G 1.4G 99% /root//mnt/sda9
In comparison my normal Porteus:
Code: Select all
guest@porteus:~$ df -h .
Filesystem Size Used Avail Use% Mounted on
aufs 265M 6.6M 258M 3% /
guest@porteus:~$ free -m
total used free shared buffers cached
Mem: 880 865 14 0 26 454
-/+ buffers/cache: 384 495
Swap: 644 9 635
guest@porteus:~$ df -h
Filesystem Size Used Avail Use% Mounted on
aufs 265M 6.6M 258M 3% /
devtmpfs 440M 4.0K 440M 1% /dev
/dev/sda4 3.9G 3.8G 101M 98% /mnt/sda4
/mnt/live/run 441M 347M 94M 79% /run
/dev/sda9 77G 72G 1.4G 99% //mnt/sda9
There seems to be a 54M shortfall in the overlay system.
Re: Building Blocks
Posted: 13 Feb 2016, 19:40
by francois
I understand that the essential info is in the first post.
Thanks.
Re: Building Blocks
Posted: 14 Feb 2016, 21:29
by Bogomips
brokenman wrote:Tested from a USB drive and found to be working.
I copied the files onto a partition on the USB drive. Changed into the boot directory and ran the porteus installer (as per a normal porteus install).
Thanks for verification. Now, on to ntfs. Modprobed ntfs.ko, so think should adjust linuxrc to use
ntfs instead of
ntfs-3g.
Using neko's iso 2 bootable usb tool, have been been able to boot usb from bios
(thanks to brainstorm suggestions contributed by your compatriot in my other thread), which also worked: Booting from USB
- Porteus data Extracted from Iso
Code: Select all
Sat Feb 13 00:55:36 2016
root@darkstar:~# cat /proc/cmdline
initrd=initrd.img 3 from=/dev/sdb1 BOOT_IMAGE=vmlinuz
root@darkstar:~# cat /var/log/porteus-livedbg
# Recognized devices:
/dev/sda1: UUID="ca6ce5e1-1bbd-4170-87f7-5fd0448e5f56" TYPE="reiserfs"
/dev/sda3: SEC_TYPE="msdos" UUID="4B46-EEEA" TYPE="vfat"
/dev/sda4: UUID="966b254d-20de-4376-b736-ddf89150a6b3" TYPE="reiserfs"
/dev/sda5: LABEL="kix" UUID="e737efd6-32f4-4a80-bbc1-a6ad3114b4d5" TYPE="ext2"
/dev/sda6: UUID="f554b881-2089-4c0f-8de2-70900bdc4e14" TYPE="ext2"
/dev/sda7: UUID="0d2c9e21-5e47-4dd3-a279-38551711f3dd" TYPE="swap"
/dev/sda8: UUID="AF61-DAE9" TYPE="vfat"
/dev/sda9: LABEL="iso" UUID="5bf4078e-3d9c-46e6-8a43-2dfc0eef3aeb" TYPE="ext2"
/dev/sda10: LABEL="p_10" UUID="ed3f74cd-26bb-47bd-86b0-5e690663dd55" TYPE="ext2"
/dev/sda11: UUID="6a0167f6-5cf7-412d-b85f-0076ca590371" TYPE="swap"
/dev/sdb1: SEC_TYPE="msdos" UUID="12D4-A2F1" TYPE="vfat"
/dev/sdb2: UUID="ec468a0e-7848-416d-9d29-3c74a1bfb486" TYPE="ext4"
# Booting device:
/mnt/sdb1
# Porteus data found in:
/mnt/sdb1/porteus
# Changes are stored in:
memory
# Non standard /rootcopy dir:
none
# Modules activated during boot time:
/mnt/sdb1/porteus/base/000-kernel.xzm
/mnt/sdb1/porteus/base/001-core.xzm
- Porteus data in Iso on USB
Code: Select all
root@darkstar:~# date
Sat Feb 13 01:01:37 2016
root@darkstar:~# cat /proc/cmdline
initrd=initrd.img 3 from=/dev/sdb2/bld/ab/abSlak_100216.iso BOOT_IMAGE=vmlinuz
root@darkstar:~# cat /var/log/porteus-livedbg
# Recognized devices:
/dev/sda1: UUID="ca6ce5e1-1bbd-4170-87f7-5fd0448e5f56" TYPE="reiserfs"
/dev/sda3: SEC_TYPE="msdos" UUID="4B46-EEEA" TYPE="vfat"
/dev/sda4: UUID="966b254d-20de-4376-b736-ddf89150a6b3" TYPE="reiserfs"
/dev/sda5: LABEL="kix" UUID="e737efd6-32f4-4a80-bbc1-a6ad3114b4d5" TYPE="ext2"
/dev/sda6: UUID="f554b881-2089-4c0f-8de2-70900bdc4e14" TYPE="ext2"
/dev/sda7: UUID="0d2c9e21-5e47-4dd3-a279-38551711f3dd" TYPE="swap"
/dev/sda8: UUID="AF61-DAE9" TYPE="vfat"
/dev/sda9: LABEL="iso" UUID="5bf4078e-3d9c-46e6-8a43-2dfc0eef3aeb" TYPE="ext2"
/dev/sda10: LABEL="p_10" UUID="ed3f74cd-26bb-47bd-86b0-5e690663dd55" TYPE="ext2"
/dev/sda11: UUID="6a0167f6-5cf7-412d-b85f-0076ca590371" TYPE="swap"
/dev/sdb1: SEC_TYPE="msdos" UUID="12D4-A2F1" TYPE="vfat"
/dev/sdb2: UUID="ec468a0e-7848-416d-9d29-3c74a1bfb486" TYPE="ext4"
# Booting device:
/mnt/sdb2
# Porteus data found in:
/mnt/isoloop/porteus
# Changes are stored in:
memory
# Non standard /rootcopy dir:
none
# Modules activated during boot time:
/mnt/isoloop/porteus/base/000-kernel.xzm
/mnt/isoloop/porteus/base/001-core.xzm
Now roadie who cannot run 64 bit informed me in a pm on 6th Feb, which I hope he does not mind me quoting from:
I'm using a 4.3.3 kernel I compiled some time ago to use in an early Nemesis, and using your initrd.img. The kernel doesn't have overlay compiled in.
I threw a current base together and it boots fine to a login, actually went straight to an Xfce when I added it in. Looking at your lsmod, I see that overlay is loaded, as it is in mine.
And in a later pm:
Code: Select all
bash-4.3# df -h
Filesystem Size Used Avail Use% Mounted on
tmpfs 499M 6.1M 493M 2% /mnt/live
devtmpfs 497M 0 497M 0% /dev
/dev/sda5 6.5G 4.8G 1.5G 77% /mnt/sda5
/dev/sdb1 30G 21G 7.5G 74% /mnt/sdb1
tmpfs 499M 441M 58M 89% /mnt/live/memory
/dev/loop0 33M 33M 0 100% /mnt/live/memory/images/000-kernel
/dev/loop1 63M 63M 0 100% /mnt/live/memory/images/001-core
/dev/loop2 256K 256K 0 100% /mnt/live/memory/images/lxterminal-0.2.0-i486-1ponce
overlay 499M 441M 58M 89% /
tmpfs 499M 240K 499M 1% /run
Looks like Porteus format Overlay could have made it right through to XFCE!
brokenman wrote:Not sure if you are interested (or it has already been posted here) in the following but I will post it anyway in case you missed it.
line 24,36,28 in rc.M throws errors on boot.
Other boot warnings/errors include:
Can't find /home/guest/p01 in fstab
Line 63 in /etc/rc.d/rc.M /usr/sbin/acpid No such file or directory
Wasn't sure if worth pursuing, but since you mention it, will pursue it myself or maybe others could for that matter?