On Thu, Mar 18, 1999 at 01:50:29AM +0100, Wichert Akkerman wrote:
> Indeed. I would *much* rather have a system that does only save
> detection (ie PCI and PnP scan) and simple skip everything else. That
PnP is not sure as I already pointed out. Scanning only PCI is not a good
solution since a lot of people are using Linux on an old system (take me as an
example ;)).
> way I can be sure that my system doesn't crash during the detection. If
> people want non-safe detection it could be an option, but I really don't
> want to support that since we can never get that right anyway.
I think we could get it at least better than Micro$oft. When we had some code
we could post it to -devel and even ./ and ask the community to test it. We
need an option to save the information how the detection has to be reordered
to be save. If we know the detection crashes either way we would ask the user
for each probe which might cause a crash.
> > 1. we know how many tests we are running. This way we could just fill the file
> > with this much of null bytes. Then we can start writing at the beginning and
> > sync after each test but ONLY the file contents.
>
> Doesn't work, you can't only sync everything with Linux, not a single
> device or a single file.
Are you sure? I have the source so I can patch it to do what I want ;) Okay,
so let's see if the standard code does thrash my hard drive if I do something
like that. I guess Linux is smart enough to detect unchanged blocks.
cu
Torsten
Attachment:
pgpb6u7FJMZo2.pgp
Description: PGP signature