I'm pretty sure this was discussed previously on the list.
This is my understanding of H. Peter Anvin's opinion/view. Due to the
unpredictable nature of the quality of different BIOSs, this has been
explicitly not implemented. If you happen to have a bad BIOS, you
might (at worst) nuke your entire hard drive by overwriting it with
corrupt data when you only wanted to write to one small text file.
This is the reason that I wrote ROSH, a Read-Only SHell. I designed
it to be able to easily navigate a file system, read a file, but then
continue booting into an operating system to continue with anything
else. And yes, it is still not where I'd consider it complete or
ready for general use.
As an alternative, I believe there has been work to implement
alternate functionality as writing to a TFTP server or a serial port.
As far as attempting to implement it in versions 4 or 5, I would
suggest asking for a response from HPA although I doubt he's changed
this opinion.
On Wed, Mar 10, 2010 at 5:08 PM, Don Hiatt <donhiatt at gmail.com>
wrote:> Hello,
>
> Pardon the newbee question but I'd like to be able to log to disk from
> within syslinux (extlinux actually, as it's a ext2 fs).
fopen("w")
> does not appear to be implemented (returns EINVAL).
>
> Has anyone written the filesystem write support, and if not, any
> suggested routes? Should I just implement the lower level BIOS disk
> write support, add the FS layer, and go from there?
>
> Thanks for your time.
>
> Cheers,
>
> don
--
-Gene
"No one ever says, 'I can't read that ASCII E-mail you sent
me.'"