Since late last year I have been running several (different variety) machines as diskless X-windows ( up to 12 not for profit) operator workstations. These systems have 256MB RAM and use internal 128MB CF cards connected to the mother board's IDE connector similar to the way discussed by Barry K..
http://www.murga.org/%7Epuppy/viewtopic.php?t=1360
Starting with PL 0.9.7 through and including PL 1.0.3 these cards always have booted without any problems. However, Pl 1.0.4 fails to boot with the "Boot failed" message. The 'syslinux.cfg' file as configured by the Puppy USB install script is exactly like Barry recommends. I have never had to modify 'syslinux.cfg' or '/etc/rc.d/rc.sysinit' through PL 1.0.3
The last PL 1.0.3 CF - IDE card I configured was "bladehunter's" 'Xfree86.4.4.0' version, but unfortunately that configuration failed to complete the boot for other reasons:
attempt to access beyond end of device
01:00 rw=0 want=14725 limit=11264
See Barry K.'s comments
Since my systems use a varitey of monitors and graphics adapters, I am still hoping to see the latest "Xfree86" offering in mainstream Puppy someday.
Hope this report proves useful....
CF card - IDE HD look alike fails to boot Puppy 1.0.4
-
BarryK
- Puppy Master
- Posts: 9392
- Joined: Mon 09 May 2005, 09:23
- Location: Perth, Western Australia
- Contact:
Just a quick reply, another contributor, sverdy, sent me a private message that when he installed Puppy to the CF card, he executed the syslinux program with "-ma" options:
# syslinux -ma <drive name>
I haven't checked this out at all, have no idea what those options do, but am passing the info on, something that worked for him.
# syslinux -ma <drive name>
I haven't checked this out at all, have no idea what those options do, but am passing the info on, something that worked for him.
Barry.... Thanks for the reply. While I didn't spend a whole lot of time, I couldn't find a reference to a "-ma" option for 'syslinux'. I decided to go ahead and test it anyway, but when attempting to edit 'install-usb.sh' I was reminded that '/usr/sbin/install-usb.sh' resides in a "read-only" file system.
I guess this is one you'll have to check out?
I guess this is one you'll have to check out?
Update.... I had a little time so I created an HD verison of PL 1.0.4. This allowed me to edit '/usr/sbin/install-usb.sh', and add the '-ma' option to the 'syslinux' statement. Then I performed a "USB" install, and rebooted the system. The boot failed with the now familiar "Boot failed" message.
If I have time I will compare 'install-usb.sh' with an older version. Of course I can be doing something wrong, but for now I guess that I am still concluding that PL 1.0.4 cannot boot a CF card - IDE HD look alike as it could in previous versions??
If I have time I will compare 'install-usb.sh' with an older version. Of course I can be doing something wrong, but for now I guess that I am still concluding that PL 1.0.4 cannot boot a CF card - IDE HD look alike as it could in previous versions??
-
BarryK
- Puppy Master
- Posts: 9392
- Joined: Mon 09 May 2005, 09:23
- Location: Perth, Western Australia
- Contact:
Re: CF card - IDE HD look alike fails to boot Puppy 1.0.4
It looks like you are using the wrong kernel.spark wrote: attempt to access beyond end of device
01:00 rw=0 want=14725 limit=11264
Pup 1.0.4 has a 2.4.29 kernel that is compiled with a ramdisk size of 12288K.
The kernels used in previous pups have ramdisk size of 11264K.
image.gz is a 12288K filesystem when expanded, and this loads into the ramdisk at bootup.
-
Guest
Barry, thanks again for the reply....
I may have confused things by reporting one problem, and mentioning another problem that I was wrestling with, both in the same post, but I am going to try it again.
The first problem:
This is probably of more interest to you because it is in mainstream Puppy 1.0.4....
Starting with PL 0.9.7 through and including PL 1.0.3, CF - IDE HD look alike cards have always configured and booted without any problems. However, PL 1.0.4 fails to boot with the "Boot failed" message. I have never had to modify 'syslinux.cfg' or '/etc/rc.d/rc.sysinit' through PL 1.0.3.
The second problem which I believe is a separte issue:
This involves "bladehunter's" 'Xfree86.4.4.0' version, where I got the following message:
attempt to access beyond end of device
01:00 rw=0 want=14725 limit=11264
I was able to run the 'Xfree86.4.4.0' version as a "live CD", but I could not successfully boot it from a CF due to the "attempt to access beyond....." error described above. While it may be true that "bladehunter's" 'Xfree86.4.4.0' version (which uses PL 1.0.3) might be using the wrong kernel, I currently have no control over that. I wouldn't know how to create Puppy with a kernel other than what is in the download files.
Please re-read my first post for more info. Sorry if I confused things.
I may have confused things by reporting one problem, and mentioning another problem that I was wrestling with, both in the same post, but I am going to try it again.
The first problem:
This is probably of more interest to you because it is in mainstream Puppy 1.0.4....
Starting with PL 0.9.7 through and including PL 1.0.3, CF - IDE HD look alike cards have always configured and booted without any problems. However, PL 1.0.4 fails to boot with the "Boot failed" message. I have never had to modify 'syslinux.cfg' or '/etc/rc.d/rc.sysinit' through PL 1.0.3.
The second problem which I believe is a separte issue:
This involves "bladehunter's" 'Xfree86.4.4.0' version, where I got the following message:
attempt to access beyond end of device
01:00 rw=0 want=14725 limit=11264
I was able to run the 'Xfree86.4.4.0' version as a "live CD", but I could not successfully boot it from a CF due to the "attempt to access beyond....." error described above. While it may be true that "bladehunter's" 'Xfree86.4.4.0' version (which uses PL 1.0.3) might be using the wrong kernel, I currently have no control over that. I wouldn't know how to create Puppy with a kernel other than what is in the download files.
Please re-read my first post for more info. Sorry if I confused things.