In article <Pine.LNX.4.21.0004040008590.2245-100000@winds.org> you
write:
> As modules [un]register new major IDs, their entries would
> appear/disappear from the /dev list on the fly without any mknod
> managing.
Actually you want something a bit more complicated, since modules may
be loaded "on the fly". You would like to have nodes not only for
registered devices but also for potentially registered ones. This
becomes some sort of nightmare with partitions: you cannot know the
number of partitions on this Zip disk before the driver for the Tekram
scsi card is loaded.
By the way: I compiled with gcc-2.95.2 a 2.2.14 kernel, with two
patches: the client-side nfsv3 from Olaf and Trond (0.20.8) and
ext3-0.0.2d (this requires some little tweaking to integrate both
patches, but it seems to work). This is on an Alpha (21164a).
When I insert the tmscsim.o module, I get an awful oops, something
which finishes with the death of an interrupt handler. This freezes
the machines, and Alpha have many registers, so I have only the
end of the story (I cannot scroll back). Is gcc-2.95.2 correct
for compiling kernels on an Alpha ?
--Thomas Pornin
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 21:00:11 EST